{
  "best_global_step": 33200,
  "best_metric": 0.18836617469787598,
  "best_model_checkpoint": "saves/prompt-tuning/gemma-3-1b-it/train_multirc_1745950260/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": 18.86997413635254,
      "learning_rate": 0.29999999259779675,
      "loss": 11.4261,
      "num_input_tokens_seen": 7936,
      "step": 5
    },
    {
      "epoch": 0.0016314544416347174,
      "grad_norm": 7.197582721710205,
      "learning_rate": 0.29999996252634736,
      "loss": 8.4773,
      "num_input_tokens_seen": 17088,
      "step": 10
    },
    {
      "epoch": 0.002447181662452076,
      "grad_norm": 2.6555418968200684,
      "learning_rate": 0.2999999093230187,
      "loss": 6.4891,
      "num_input_tokens_seen": 26976,
      "step": 15
    },
    {
      "epoch": 0.0032629088832694347,
      "grad_norm": 2.1168947219848633,
      "learning_rate": 0.299999832987819,
      "loss": 5.3995,
      "num_input_tokens_seen": 37536,
      "step": 20
    },
    {
      "epoch": 0.004078636104086793,
      "grad_norm": 3.1743268966674805,
      "learning_rate": 0.29999973352076004,
      "loss": 5.1431,
      "num_input_tokens_seen": 46224,
      "step": 25
    },
    {
      "epoch": 0.004894363324904152,
      "grad_norm": 1.0057926177978516,
      "learning_rate": 0.2999996109218572,
      "loss": 4.4346,
      "num_input_tokens_seen": 55984,
      "step": 30
    },
    {
      "epoch": 0.005710090545721511,
      "grad_norm": 2.1299688816070557,
      "learning_rate": 0.2999994651911293,
      "loss": 3.9055,
      "num_input_tokens_seen": 65952,
      "step": 35
    },
    {
      "epoch": 0.0065258177665388694,
      "grad_norm": 0.8906856775283813,
      "learning_rate": 0.2999992963285989,
      "loss": 3.7707,
      "num_input_tokens_seen": 76256,
      "step": 40
    },
    {
      "epoch": 0.007341544987356228,
      "grad_norm": 2.4730329513549805,
      "learning_rate": 0.29999910433429194,
      "loss": 2.8248,
      "num_input_tokens_seen": 86080,
      "step": 45
    },
    {
      "epoch": 0.008157272208173586,
      "grad_norm": 1.0065432786941528,
      "learning_rate": 0.29999888920823814,
      "loss": 2.6501,
      "num_input_tokens_seen": 96528,
      "step": 50
    },
    {
      "epoch": 0.008972999428990946,
      "grad_norm": 0.543121337890625,
      "learning_rate": 0.29999865095047057,
      "loss": 1.7045,
      "num_input_tokens_seen": 106144,
      "step": 55
    },
    {
      "epoch": 0.009788726649808304,
      "grad_norm": 1.604982852935791,
      "learning_rate": 0.29999838956102604,
      "loss": 1.3896,
      "num_input_tokens_seen": 115184,
      "step": 60
    },
    {
      "epoch": 0.010604453870625663,
      "grad_norm": 1.2762478590011597,
      "learning_rate": 0.29999810503994484,
      "loss": 1.037,
      "num_input_tokens_seen": 124096,
      "step": 65
    },
    {
      "epoch": 0.011420181091443021,
      "grad_norm": 0.5977362990379333,
      "learning_rate": 0.29999779738727084,
      "loss": 0.7137,
      "num_input_tokens_seen": 133280,
      "step": 70
    },
    {
      "epoch": 0.012235908312260381,
      "grad_norm": 1.3828400373458862,
      "learning_rate": 0.29999746660305154,
      "loss": 0.8227,
      "num_input_tokens_seen": 143072,
      "step": 75
    },
    {
      "epoch": 0.013051635533077739,
      "grad_norm": 0.8577566742897034,
      "learning_rate": 0.2999971126873379,
      "loss": 0.7209,
      "num_input_tokens_seen": 151808,
      "step": 80
    },
    {
      "epoch": 0.013867362753895097,
      "grad_norm": 1.483636736869812,
      "learning_rate": 0.2999967356401845,
      "loss": 1.1607,
      "num_input_tokens_seen": 161184,
      "step": 85
    },
    {
      "epoch": 0.014683089974712456,
      "grad_norm": 0.8933716416358948,
      "learning_rate": 0.29999633546164944,
      "loss": 0.9416,
      "num_input_tokens_seen": 171328,
      "step": 90
    },
    {
      "epoch": 0.015498817195529814,
      "grad_norm": 0.18278126418590546,
      "learning_rate": 0.29999591215179444,
      "loss": 0.7742,
      "num_input_tokens_seen": 180304,
      "step": 95
    },
    {
      "epoch": 0.016314544416347172,
      "grad_norm": 1.087674617767334,
      "learning_rate": 0.2999954657106849,
      "loss": 0.8016,
      "num_input_tokens_seen": 191024,
      "step": 100
    },
    {
      "epoch": 0.017130271637164532,
      "grad_norm": 0.47661086916923523,
      "learning_rate": 0.2999949961383896,
      "loss": 0.641,
      "num_input_tokens_seen": 201056,
      "step": 105
    },
    {
      "epoch": 0.01794599885798189,
      "grad_norm": 0.6102319955825806,
      "learning_rate": 0.2999945034349809,
      "loss": 0.4737,
      "num_input_tokens_seen": 211808,
      "step": 110
    },
    {
      "epoch": 0.01876172607879925,
      "grad_norm": 0.4202670156955719,
      "learning_rate": 0.2999939876005348,
      "loss": 0.3806,
      "num_input_tokens_seen": 220128,
      "step": 115
    },
    {
      "epoch": 0.019577453299616607,
      "grad_norm": 0.2838693857192993,
      "learning_rate": 0.29999344863513094,
      "loss": 0.5035,
      "num_input_tokens_seen": 229264,
      "step": 120
    },
    {
      "epoch": 0.020393180520433967,
      "grad_norm": 0.17632564902305603,
      "learning_rate": 0.2999928865388523,
      "loss": 0.4744,
      "num_input_tokens_seen": 239120,
      "step": 125
    },
    {
      "epoch": 0.021208907741251327,
      "grad_norm": 0.5180648565292358,
      "learning_rate": 0.29999230131178567,
      "loss": 0.4443,
      "num_input_tokens_seen": 248656,
      "step": 130
    },
    {
      "epoch": 0.022024634962068683,
      "grad_norm": 0.4269920885562897,
      "learning_rate": 0.2999916929540212,
      "loss": 0.5492,
      "num_input_tokens_seen": 258848,
      "step": 135
    },
    {
      "epoch": 0.022840362182886043,
      "grad_norm": 0.6505976319313049,
      "learning_rate": 0.29999106146565285,
      "loss": 0.7212,
      "num_input_tokens_seen": 267936,
      "step": 140
    },
    {
      "epoch": 0.023656089403703402,
      "grad_norm": 0.416185200214386,
      "learning_rate": 0.29999040684677786,
      "loss": 0.498,
      "num_input_tokens_seen": 277456,
      "step": 145
    },
    {
      "epoch": 0.024471816624520762,
      "grad_norm": 0.1383996158838272,
      "learning_rate": 0.2999897290974972,
      "loss": 0.449,
      "num_input_tokens_seen": 286544,
      "step": 150
    },
    {
      "epoch": 0.025287543845338118,
      "grad_norm": 0.35335642099380493,
      "learning_rate": 0.2999890282179155,
      "loss": 0.4265,
      "num_input_tokens_seen": 297216,
      "step": 155
    },
    {
      "epoch": 0.026103271066155478,
      "grad_norm": 0.4622231125831604,
      "learning_rate": 0.29998830420814077,
      "loss": 0.5311,
      "num_input_tokens_seen": 307472,
      "step": 160
    },
    {
      "epoch": 0.026918998286972837,
      "grad_norm": 0.34515032172203064,
      "learning_rate": 0.2999875570682846,
      "loss": 0.4457,
      "num_input_tokens_seen": 317984,
      "step": 165
    },
    {
      "epoch": 0.027734725507790194,
      "grad_norm": 0.26643431186676025,
      "learning_rate": 0.2999867867984623,
      "loss": 0.3175,
      "num_input_tokens_seen": 328240,
      "step": 170
    },
    {
      "epoch": 0.028550452728607553,
      "grad_norm": 0.8447004556655884,
      "learning_rate": 0.29998599339879267,
      "loss": 0.3984,
      "num_input_tokens_seen": 336848,
      "step": 175
    },
    {
      "epoch": 0.029366179949424913,
      "grad_norm": 0.30018681287765503,
      "learning_rate": 0.29998517686939796,
      "loss": 0.3355,
      "num_input_tokens_seen": 346304,
      "step": 180
    },
    {
      "epoch": 0.030181907170242273,
      "grad_norm": 0.7179999947547913,
      "learning_rate": 0.29998433721040413,
      "loss": 0.3624,
      "num_input_tokens_seen": 356160,
      "step": 185
    },
    {
      "epoch": 0.03099763439105963,
      "grad_norm": 0.20405782759189606,
      "learning_rate": 0.29998347442194073,
      "loss": 0.3655,
      "num_input_tokens_seen": 365488,
      "step": 190
    },
    {
      "epoch": 0.031813361611876985,
      "grad_norm": 0.22518043220043182,
      "learning_rate": 0.2999825885041407,
      "loss": 0.4361,
      "num_input_tokens_seen": 375600,
      "step": 195
    },
    {
      "epoch": 0.032629088832694345,
      "grad_norm": 0.3701261579990387,
      "learning_rate": 0.29998167945714077,
      "loss": 0.3259,
      "num_input_tokens_seen": 385088,
      "step": 200
    },
    {
      "epoch": 0.032629088832694345,
      "eval_loss": 0.3690318465232849,
      "eval_runtime": 68.4254,
      "eval_samples_per_second": 39.824,
      "eval_steps_per_second": 19.919,
      "num_input_tokens_seen": 385088,
      "step": 200
    },
    {
      "epoch": 0.033444816053511704,
      "grad_norm": 0.5973607897758484,
      "learning_rate": 0.2999807472810811,
      "loss": 0.4224,
      "num_input_tokens_seen": 395168,
      "step": 205
    },
    {
      "epoch": 0.034260543274329064,
      "grad_norm": 0.22334657609462738,
      "learning_rate": 0.29997979197610536,
      "loss": 0.3258,
      "num_input_tokens_seen": 406016,
      "step": 210
    },
    {
      "epoch": 0.035076270495146424,
      "grad_norm": 0.09077087044715881,
      "learning_rate": 0.299978813542361,
      "loss": 0.3024,
      "num_input_tokens_seen": 415600,
      "step": 215
    },
    {
      "epoch": 0.03589199771596378,
      "grad_norm": 0.2089405357837677,
      "learning_rate": 0.2999778119799988,
      "loss": 0.58,
      "num_input_tokens_seen": 426560,
      "step": 220
    },
    {
      "epoch": 0.03670772493678114,
      "grad_norm": 0.4931613802909851,
      "learning_rate": 0.29997678728917326,
      "loss": 0.3987,
      "num_input_tokens_seen": 436640,
      "step": 225
    },
    {
      "epoch": 0.0375234521575985,
      "grad_norm": 0.13702410459518433,
      "learning_rate": 0.2999757394700424,
      "loss": 0.3034,
      "num_input_tokens_seen": 445712,
      "step": 230
    },
    {
      "epoch": 0.038339179378415855,
      "grad_norm": 1.611151933670044,
      "learning_rate": 0.29997466852276783,
      "loss": 1.2131,
      "num_input_tokens_seen": 454960,
      "step": 235
    },
    {
      "epoch": 0.039154906599233215,
      "grad_norm": 0.4760196805000305,
      "learning_rate": 0.29997357444751466,
      "loss": 0.9369,
      "num_input_tokens_seen": 465104,
      "step": 240
    },
    {
      "epoch": 0.039970633820050575,
      "grad_norm": 0.4496554434299469,
      "learning_rate": 0.2999724572444516,
      "loss": 1.0115,
      "num_input_tokens_seen": 474032,
      "step": 245
    },
    {
      "epoch": 0.040786361040867934,
      "grad_norm": 2.15747332572937,
      "learning_rate": 0.29997131691375095,
      "loss": 0.7011,
      "num_input_tokens_seen": 484032,
      "step": 250
    },
    {
      "epoch": 0.041602088261685294,
      "grad_norm": 0.68916916847229,
      "learning_rate": 0.2999701534555886,
      "loss": 0.79,
      "num_input_tokens_seen": 492704,
      "step": 255
    },
    {
      "epoch": 0.042417815482502653,
      "grad_norm": 0.7923445701599121,
      "learning_rate": 0.2999689668701439,
      "loss": 0.8045,
      "num_input_tokens_seen": 501584,
      "step": 260
    },
    {
      "epoch": 0.04323354270332001,
      "grad_norm": 0.23233264684677124,
      "learning_rate": 0.29996775715759993,
      "loss": 0.8237,
      "num_input_tokens_seen": 512480,
      "step": 265
    },
    {
      "epoch": 0.044049269924137366,
      "grad_norm": 0.15219654142856598,
      "learning_rate": 0.2999665243181432,
      "loss": 0.3479,
      "num_input_tokens_seen": 521520,
      "step": 270
    },
    {
      "epoch": 0.044864997144954726,
      "grad_norm": 0.3355565667152405,
      "learning_rate": 0.2999652683519638,
      "loss": 0.4829,
      "num_input_tokens_seen": 531568,
      "step": 275
    },
    {
      "epoch": 0.045680724365772085,
      "grad_norm": 0.4314683973789215,
      "learning_rate": 0.29996398925925544,
      "loss": 0.4936,
      "num_input_tokens_seen": 541408,
      "step": 280
    },
    {
      "epoch": 0.046496451586589445,
      "grad_norm": 0.4303458034992218,
      "learning_rate": 0.2999626870402154,
      "loss": 0.3807,
      "num_input_tokens_seen": 551008,
      "step": 285
    },
    {
      "epoch": 0.047312178807406804,
      "grad_norm": 0.1524202972650528,
      "learning_rate": 0.29996136169504445,
      "loss": 0.4448,
      "num_input_tokens_seen": 559984,
      "step": 290
    },
    {
      "epoch": 0.048127906028224164,
      "grad_norm": 0.15022683143615723,
      "learning_rate": 0.29996001322394694,
      "loss": 0.402,
      "num_input_tokens_seen": 569760,
      "step": 295
    },
    {
      "epoch": 0.048943633249041524,
      "grad_norm": 0.08683402091264725,
      "learning_rate": 0.29995864162713093,
      "loss": 0.3224,
      "num_input_tokens_seen": 579504,
      "step": 300
    },
    {
      "epoch": 0.049759360469858877,
      "grad_norm": 0.17401407659053802,
      "learning_rate": 0.2999572469048079,
      "loss": 0.2912,
      "num_input_tokens_seen": 590624,
      "step": 305
    },
    {
      "epoch": 0.050575087690676236,
      "grad_norm": 0.10288899391889572,
      "learning_rate": 0.29995582905719287,
      "loss": 0.3403,
      "num_input_tokens_seen": 599520,
      "step": 310
    },
    {
      "epoch": 0.051390814911493596,
      "grad_norm": 0.11537853628396988,
      "learning_rate": 0.2999543880845046,
      "loss": 0.302,
      "num_input_tokens_seen": 609808,
      "step": 315
    },
    {
      "epoch": 0.052206542132310955,
      "grad_norm": 0.12451272457838058,
      "learning_rate": 0.2999529239869652,
      "loss": 0.3287,
      "num_input_tokens_seen": 619856,
      "step": 320
    },
    {
      "epoch": 0.053022269353128315,
      "grad_norm": 0.11940126121044159,
      "learning_rate": 0.2999514367648005,
      "loss": 0.3374,
      "num_input_tokens_seen": 629152,
      "step": 325
    },
    {
      "epoch": 0.053837996573945675,
      "grad_norm": 0.08504441380500793,
      "learning_rate": 0.29994992641823987,
      "loss": 0.2651,
      "num_input_tokens_seen": 637904,
      "step": 330
    },
    {
      "epoch": 0.054653723794763034,
      "grad_norm": 0.06713485717773438,
      "learning_rate": 0.29994839294751613,
      "loss": 0.2327,
      "num_input_tokens_seen": 647856,
      "step": 335
    },
    {
      "epoch": 0.05546945101558039,
      "grad_norm": 0.07705973833799362,
      "learning_rate": 0.29994683635286584,
      "loss": 0.2375,
      "num_input_tokens_seen": 657456,
      "step": 340
    },
    {
      "epoch": 0.05628517823639775,
      "grad_norm": 0.06248573213815689,
      "learning_rate": 0.2999452566345291,
      "loss": 0.2911,
      "num_input_tokens_seen": 665968,
      "step": 345
    },
    {
      "epoch": 0.057100905457215106,
      "grad_norm": 0.03076113760471344,
      "learning_rate": 0.2999436537927494,
      "loss": 0.2299,
      "num_input_tokens_seen": 675184,
      "step": 350
    },
    {
      "epoch": 0.057916632678032466,
      "grad_norm": 0.05518700182437897,
      "learning_rate": 0.299942027827774,
      "loss": 0.2389,
      "num_input_tokens_seen": 685952,
      "step": 355
    },
    {
      "epoch": 0.058732359898849826,
      "grad_norm": 0.06942839920520782,
      "learning_rate": 0.29994037873985363,
      "loss": 0.3686,
      "num_input_tokens_seen": 694960,
      "step": 360
    },
    {
      "epoch": 0.059548087119667185,
      "grad_norm": 0.10586044937372208,
      "learning_rate": 0.29993870652924254,
      "loss": 0.3936,
      "num_input_tokens_seen": 705840,
      "step": 365
    },
    {
      "epoch": 0.060363814340484545,
      "grad_norm": 0.07195794582366943,
      "learning_rate": 0.29993701119619876,
      "loss": 0.3322,
      "num_input_tokens_seen": 715648,
      "step": 370
    },
    {
      "epoch": 0.0611795415613019,
      "grad_norm": 0.053050410002470016,
      "learning_rate": 0.2999352927409835,
      "loss": 0.2397,
      "num_input_tokens_seen": 725472,
      "step": 375
    },
    {
      "epoch": 0.06199526878211926,
      "grad_norm": 0.08435414731502533,
      "learning_rate": 0.29993355116386194,
      "loss": 0.2241,
      "num_input_tokens_seen": 734848,
      "step": 380
    },
    {
      "epoch": 0.06281099600293662,
      "grad_norm": 0.11217772215604782,
      "learning_rate": 0.29993178646510266,
      "loss": 0.2319,
      "num_input_tokens_seen": 743776,
      "step": 385
    },
    {
      "epoch": 0.06362672322375397,
      "grad_norm": 0.08071137964725494,
      "learning_rate": 0.2999299986449777,
      "loss": 0.2985,
      "num_input_tokens_seen": 752160,
      "step": 390
    },
    {
      "epoch": 0.06444245044457134,
      "grad_norm": 0.09554888308048248,
      "learning_rate": 0.29992818770376284,
      "loss": 0.2723,
      "num_input_tokens_seen": 760944,
      "step": 395
    },
    {
      "epoch": 0.06525817766538869,
      "grad_norm": 0.14936256408691406,
      "learning_rate": 0.29992635364173725,
      "loss": 0.2254,
      "num_input_tokens_seen": 770352,
      "step": 400
    },
    {
      "epoch": 0.06525817766538869,
      "eval_loss": 0.26653409004211426,
      "eval_runtime": 67.9275,
      "eval_samples_per_second": 40.116,
      "eval_steps_per_second": 20.065,
      "num_input_tokens_seen": 770352,
      "step": 400
    },
    {
      "epoch": 0.06607390488620606,
      "grad_norm": 0.06688330322504044,
      "learning_rate": 0.2999244964591839,
      "loss": 0.2556,
      "num_input_tokens_seen": 779504,
      "step": 405
    },
    {
      "epoch": 0.06688963210702341,
      "grad_norm": 0.04414146766066551,
      "learning_rate": 0.2999226161563891,
      "loss": 0.2678,
      "num_input_tokens_seen": 789264,
      "step": 410
    },
    {
      "epoch": 0.06770535932784078,
      "grad_norm": 0.1638306975364685,
      "learning_rate": 0.2999207127336429,
      "loss": 0.2371,
      "num_input_tokens_seen": 799248,
      "step": 415
    },
    {
      "epoch": 0.06852108654865813,
      "grad_norm": 0.10321413725614548,
      "learning_rate": 0.2999187861912387,
      "loss": 0.2562,
      "num_input_tokens_seen": 809456,
      "step": 420
    },
    {
      "epoch": 0.06933681376947548,
      "grad_norm": 0.13290010392665863,
      "learning_rate": 0.2999168365294737,
      "loss": 0.3084,
      "num_input_tokens_seen": 818848,
      "step": 425
    },
    {
      "epoch": 0.07015254099029285,
      "grad_norm": 0.09715215116739273,
      "learning_rate": 0.29991486374864856,
      "loss": 0.1975,
      "num_input_tokens_seen": 828656,
      "step": 430
    },
    {
      "epoch": 0.0709682682111102,
      "grad_norm": 0.05678919702768326,
      "learning_rate": 0.29991286784906745,
      "loss": 0.2412,
      "num_input_tokens_seen": 838576,
      "step": 435
    },
    {
      "epoch": 0.07178399543192757,
      "grad_norm": 0.09300313889980316,
      "learning_rate": 0.2999108488310382,
      "loss": 0.2583,
      "num_input_tokens_seen": 847760,
      "step": 440
    },
    {
      "epoch": 0.07259972265274492,
      "grad_norm": 0.17148126661777496,
      "learning_rate": 0.29990880669487213,
      "loss": 0.3182,
      "num_input_tokens_seen": 857152,
      "step": 445
    },
    {
      "epoch": 0.07341544987356229,
      "grad_norm": 0.056695181876420975,
      "learning_rate": 0.29990674144088425,
      "loss": 0.2663,
      "num_input_tokens_seen": 866784,
      "step": 450
    },
    {
      "epoch": 0.07423117709437964,
      "grad_norm": 0.10688269883394241,
      "learning_rate": 0.299904653069393,
      "loss": 0.3222,
      "num_input_tokens_seen": 876960,
      "step": 455
    },
    {
      "epoch": 0.075046904315197,
      "grad_norm": 0.07041271030902863,
      "learning_rate": 0.29990254158072044,
      "loss": 0.2617,
      "num_input_tokens_seen": 886144,
      "step": 460
    },
    {
      "epoch": 0.07586263153601436,
      "grad_norm": 0.0586092583835125,
      "learning_rate": 0.2999004069751921,
      "loss": 0.3015,
      "num_input_tokens_seen": 896528,
      "step": 465
    },
    {
      "epoch": 0.07667835875683171,
      "grad_norm": 0.046219486743211746,
      "learning_rate": 0.2998982492531373,
      "loss": 0.2176,
      "num_input_tokens_seen": 907168,
      "step": 470
    },
    {
      "epoch": 0.07749408597764908,
      "grad_norm": 0.03769886866211891,
      "learning_rate": 0.2998960684148887,
      "loss": 0.2866,
      "num_input_tokens_seen": 916656,
      "step": 475
    },
    {
      "epoch": 0.07830981319846643,
      "grad_norm": 0.10739131271839142,
      "learning_rate": 0.29989386446078264,
      "loss": 0.2309,
      "num_input_tokens_seen": 925648,
      "step": 480
    },
    {
      "epoch": 0.0791255404192838,
      "grad_norm": 0.051703136414289474,
      "learning_rate": 0.299891637391159,
      "loss": 0.2668,
      "num_input_tokens_seen": 936096,
      "step": 485
    },
    {
      "epoch": 0.07994126764010115,
      "grad_norm": 0.07048793137073517,
      "learning_rate": 0.2998893872063612,
      "loss": 0.3499,
      "num_input_tokens_seen": 946464,
      "step": 490
    },
    {
      "epoch": 0.08075699486091852,
      "grad_norm": 0.04133843630552292,
      "learning_rate": 0.2998871139067363,
      "loss": 0.2859,
      "num_input_tokens_seen": 954448,
      "step": 495
    },
    {
      "epoch": 0.08157272208173587,
      "grad_norm": 0.06871169805526733,
      "learning_rate": 0.2998848174926348,
      "loss": 0.2087,
      "num_input_tokens_seen": 964720,
      "step": 500
    },
    {
      "epoch": 0.08238844930255322,
      "grad_norm": 0.058107029646635056,
      "learning_rate": 0.2998824979644109,
      "loss": 0.2036,
      "num_input_tokens_seen": 974240,
      "step": 505
    },
    {
      "epoch": 0.08320417652337059,
      "grad_norm": 0.08431321382522583,
      "learning_rate": 0.29988015532242224,
      "loss": 0.3015,
      "num_input_tokens_seen": 983152,
      "step": 510
    },
    {
      "epoch": 0.08401990374418794,
      "grad_norm": 0.052418194711208344,
      "learning_rate": 0.29987778956703015,
      "loss": 0.2394,
      "num_input_tokens_seen": 993216,
      "step": 515
    },
    {
      "epoch": 0.08483563096500531,
      "grad_norm": 0.14298206567764282,
      "learning_rate": 0.2998754006985994,
      "loss": 0.3134,
      "num_input_tokens_seen": 1002480,
      "step": 520
    },
    {
      "epoch": 0.08565135818582266,
      "grad_norm": 0.09891901910305023,
      "learning_rate": 0.29987298871749846,
      "loss": 0.2951,
      "num_input_tokens_seen": 1012192,
      "step": 525
    },
    {
      "epoch": 0.08646708540664003,
      "grad_norm": 0.1226806715130806,
      "learning_rate": 0.2998705536240992,
      "loss": 0.2651,
      "num_input_tokens_seen": 1021824,
      "step": 530
    },
    {
      "epoch": 0.08728281262745738,
      "grad_norm": 0.09256590902805328,
      "learning_rate": 0.2998680954187772,
      "loss": 0.2569,
      "num_input_tokens_seen": 1032384,
      "step": 535
    },
    {
      "epoch": 0.08809853984827473,
      "grad_norm": 0.06631482392549515,
      "learning_rate": 0.2998656141019115,
      "loss": 0.2864,
      "num_input_tokens_seen": 1041888,
      "step": 540
    },
    {
      "epoch": 0.0889142670690921,
      "grad_norm": 0.09103129804134369,
      "learning_rate": 0.2998631096738848,
      "loss": 0.2289,
      "num_input_tokens_seen": 1052032,
      "step": 545
    },
    {
      "epoch": 0.08972999428990945,
      "grad_norm": 0.07785353064537048,
      "learning_rate": 0.29986058213508326,
      "loss": 0.2516,
      "num_input_tokens_seen": 1063104,
      "step": 550
    },
    {
      "epoch": 0.09054572151072682,
      "grad_norm": 0.09148314595222473,
      "learning_rate": 0.29985803148589674,
      "loss": 0.3094,
      "num_input_tokens_seen": 1074016,
      "step": 555
    },
    {
      "epoch": 0.09136144873154417,
      "grad_norm": 0.08607068657875061,
      "learning_rate": 0.2998554577267185,
      "loss": 0.2517,
      "num_input_tokens_seen": 1083776,
      "step": 560
    },
    {
      "epoch": 0.09217717595236154,
      "grad_norm": 0.06277384608983994,
      "learning_rate": 0.2998528608579455,
      "loss": 0.2123,
      "num_input_tokens_seen": 1092896,
      "step": 565
    },
    {
      "epoch": 0.09299290317317889,
      "grad_norm": 0.08404625952243805,
      "learning_rate": 0.2998502408799781,
      "loss": 0.2468,
      "num_input_tokens_seen": 1101920,
      "step": 570
    },
    {
      "epoch": 0.09380863039399624,
      "grad_norm": 0.18898604810237885,
      "learning_rate": 0.2998475977932205,
      "loss": 0.2649,
      "num_input_tokens_seen": 1111904,
      "step": 575
    },
    {
      "epoch": 0.09462435761481361,
      "grad_norm": 0.08054934442043304,
      "learning_rate": 0.29984493159808023,
      "loss": 0.2821,
      "num_input_tokens_seen": 1121744,
      "step": 580
    },
    {
      "epoch": 0.09544008483563096,
      "grad_norm": 0.19562818109989166,
      "learning_rate": 0.29984224229496836,
      "loss": 0.2065,
      "num_input_tokens_seen": 1130400,
      "step": 585
    },
    {
      "epoch": 0.09625581205644833,
      "grad_norm": 0.11274368315935135,
      "learning_rate": 0.2998395298842998,
      "loss": 0.3041,
      "num_input_tokens_seen": 1140848,
      "step": 590
    },
    {
      "epoch": 0.09707153927726568,
      "grad_norm": 0.05425263196229935,
      "learning_rate": 0.29983679436649263,
      "loss": 0.2522,
      "num_input_tokens_seen": 1151008,
      "step": 595
    },
    {
      "epoch": 0.09788726649808305,
      "grad_norm": 0.09642237424850464,
      "learning_rate": 0.2998340357419689,
      "loss": 0.2463,
      "num_input_tokens_seen": 1160480,
      "step": 600
    },
    {
      "epoch": 0.09788726649808305,
      "eval_loss": 0.2561352550983429,
      "eval_runtime": 68.0346,
      "eval_samples_per_second": 40.053,
      "eval_steps_per_second": 20.034,
      "num_input_tokens_seen": 1160480,
      "step": 600
    },
    {
      "epoch": 0.0987029937189004,
      "grad_norm": 0.0868828222155571,
      "learning_rate": 0.29983125401115385,
      "loss": 0.3145,
      "num_input_tokens_seen": 1170144,
      "step": 605
    },
    {
      "epoch": 0.09951872093971775,
      "grad_norm": 0.7048643827438354,
      "learning_rate": 0.29982844917447654,
      "loss": 0.3612,
      "num_input_tokens_seen": 1180064,
      "step": 610
    },
    {
      "epoch": 0.10033444816053512,
      "grad_norm": 0.07285650074481964,
      "learning_rate": 0.2998256212323695,
      "loss": 0.3116,
      "num_input_tokens_seen": 1190224,
      "step": 615
    },
    {
      "epoch": 0.10115017538135247,
      "grad_norm": 0.07756111025810242,
      "learning_rate": 0.29982277018526887,
      "loss": 0.3517,
      "num_input_tokens_seen": 1199504,
      "step": 620
    },
    {
      "epoch": 0.10196590260216984,
      "grad_norm": 0.07323414087295532,
      "learning_rate": 0.2998198960336143,
      "loss": 0.4011,
      "num_input_tokens_seen": 1209120,
      "step": 625
    },
    {
      "epoch": 0.10278162982298719,
      "grad_norm": 0.18309904634952545,
      "learning_rate": 0.299816998777849,
      "loss": 0.2212,
      "num_input_tokens_seen": 1218432,
      "step": 630
    },
    {
      "epoch": 0.10359735704380456,
      "grad_norm": 0.11869360506534576,
      "learning_rate": 0.2998140784184197,
      "loss": 0.3199,
      "num_input_tokens_seen": 1227376,
      "step": 635
    },
    {
      "epoch": 0.10441308426462191,
      "grad_norm": 0.030079295858740807,
      "learning_rate": 0.2998111349557769,
      "loss": 0.1832,
      "num_input_tokens_seen": 1235536,
      "step": 640
    },
    {
      "epoch": 0.10522881148543926,
      "grad_norm": 0.22688305377960205,
      "learning_rate": 0.29980816839037444,
      "loss": 0.3864,
      "num_input_tokens_seen": 1245152,
      "step": 645
    },
    {
      "epoch": 0.10604453870625663,
      "grad_norm": 0.09628093987703323,
      "learning_rate": 0.2998051787226698,
      "loss": 0.2445,
      "num_input_tokens_seen": 1254768,
      "step": 650
    },
    {
      "epoch": 0.10686026592707398,
      "grad_norm": 0.06357354670763016,
      "learning_rate": 0.29980216595312403,
      "loss": 0.28,
      "num_input_tokens_seen": 1264864,
      "step": 655
    },
    {
      "epoch": 0.10767599314789135,
      "grad_norm": 0.10401547700166702,
      "learning_rate": 0.29979913008220177,
      "loss": 0.2687,
      "num_input_tokens_seen": 1273696,
      "step": 660
    },
    {
      "epoch": 0.1084917203687087,
      "grad_norm": 0.06523094326257706,
      "learning_rate": 0.2997960711103711,
      "loss": 0.2515,
      "num_input_tokens_seen": 1281904,
      "step": 665
    },
    {
      "epoch": 0.10930744758952607,
      "grad_norm": 0.07449512928724289,
      "learning_rate": 0.29979298903810386,
      "loss": 0.2517,
      "num_input_tokens_seen": 1292144,
      "step": 670
    },
    {
      "epoch": 0.11012317481034342,
      "grad_norm": 0.0795942023396492,
      "learning_rate": 0.29978988386587524,
      "loss": 0.318,
      "num_input_tokens_seen": 1302000,
      "step": 675
    },
    {
      "epoch": 0.11093890203116077,
      "grad_norm": 0.0757814347743988,
      "learning_rate": 0.2997867555941642,
      "loss": 0.3024,
      "num_input_tokens_seen": 1312016,
      "step": 680
    },
    {
      "epoch": 0.11175462925197814,
      "grad_norm": 0.06944116950035095,
      "learning_rate": 0.299783604223453,
      "loss": 0.2514,
      "num_input_tokens_seen": 1320928,
      "step": 685
    },
    {
      "epoch": 0.1125703564727955,
      "grad_norm": 0.08203577995300293,
      "learning_rate": 0.29978042975422786,
      "loss": 0.2739,
      "num_input_tokens_seen": 1330464,
      "step": 690
    },
    {
      "epoch": 0.11338608369361286,
      "grad_norm": 0.3575659692287445,
      "learning_rate": 0.29977723218697816,
      "loss": 0.2844,
      "num_input_tokens_seen": 1340896,
      "step": 695
    },
    {
      "epoch": 0.11420181091443021,
      "grad_norm": 0.08346308022737503,
      "learning_rate": 0.299774011522197,
      "loss": 0.273,
      "num_input_tokens_seen": 1350128,
      "step": 700
    },
    {
      "epoch": 0.11501753813524758,
      "grad_norm": 0.11895722150802612,
      "learning_rate": 0.29977076776038114,
      "loss": 0.2803,
      "num_input_tokens_seen": 1359552,
      "step": 705
    },
    {
      "epoch": 0.11583326535606493,
      "grad_norm": 0.06554309278726578,
      "learning_rate": 0.2997675009020307,
      "loss": 0.3166,
      "num_input_tokens_seen": 1370448,
      "step": 710
    },
    {
      "epoch": 0.11664899257688228,
      "grad_norm": 0.030779998749494553,
      "learning_rate": 0.2997642109476496,
      "loss": 0.262,
      "num_input_tokens_seen": 1380528,
      "step": 715
    },
    {
      "epoch": 0.11746471979769965,
      "grad_norm": 0.01954539865255356,
      "learning_rate": 0.299760897897745,
      "loss": 0.3133,
      "num_input_tokens_seen": 1389328,
      "step": 720
    },
    {
      "epoch": 0.118280447018517,
      "grad_norm": 0.04499318450689316,
      "learning_rate": 0.29975756175282803,
      "loss": 0.3075,
      "num_input_tokens_seen": 1398944,
      "step": 725
    },
    {
      "epoch": 0.11909617423933437,
      "grad_norm": 0.09267327934503555,
      "learning_rate": 0.29975420251341306,
      "loss": 0.3235,
      "num_input_tokens_seen": 1407728,
      "step": 730
    },
    {
      "epoch": 0.11991190146015172,
      "grad_norm": 0.11061648279428482,
      "learning_rate": 0.29975082018001814,
      "loss": 0.392,
      "num_input_tokens_seen": 1417744,
      "step": 735
    },
    {
      "epoch": 0.12072762868096909,
      "grad_norm": 0.16188117861747742,
      "learning_rate": 0.2997474147531648,
      "loss": 0.3254,
      "num_input_tokens_seen": 1428016,
      "step": 740
    },
    {
      "epoch": 0.12154335590178644,
      "grad_norm": 0.07655633985996246,
      "learning_rate": 0.29974398623337833,
      "loss": 0.3014,
      "num_input_tokens_seen": 1438784,
      "step": 745
    },
    {
      "epoch": 0.1223590831226038,
      "grad_norm": 0.06820513308048248,
      "learning_rate": 0.2997405346211873,
      "loss": 0.3091,
      "num_input_tokens_seen": 1449120,
      "step": 750
    },
    {
      "epoch": 0.12317481034342116,
      "grad_norm": 0.061615049839019775,
      "learning_rate": 0.2997370599171241,
      "loss": 0.2812,
      "num_input_tokens_seen": 1459312,
      "step": 755
    },
    {
      "epoch": 0.12399053756423851,
      "grad_norm": 0.09007983654737473,
      "learning_rate": 0.2997335621217246,
      "loss": 0.2658,
      "num_input_tokens_seen": 1469744,
      "step": 760
    },
    {
      "epoch": 0.12480626478505588,
      "grad_norm": 0.016299240291118622,
      "learning_rate": 0.29973004123552816,
      "loss": 0.241,
      "num_input_tokens_seen": 1479376,
      "step": 765
    },
    {
      "epoch": 0.12562199200587323,
      "grad_norm": 0.032591622322797775,
      "learning_rate": 0.2997264972590777,
      "loss": 0.2824,
      "num_input_tokens_seen": 1488816,
      "step": 770
    },
    {
      "epoch": 0.1264377192266906,
      "grad_norm": 0.03721635788679123,
      "learning_rate": 0.29972293019291973,
      "loss": 0.3243,
      "num_input_tokens_seen": 1497056,
      "step": 775
    },
    {
      "epoch": 0.12725344644750794,
      "grad_norm": 0.01886346936225891,
      "learning_rate": 0.2997193400376045,
      "loss": 0.2275,
      "num_input_tokens_seen": 1506304,
      "step": 780
    },
    {
      "epoch": 0.12806917366832532,
      "grad_norm": 0.027662381529808044,
      "learning_rate": 0.2997157267936854,
      "loss": 0.2465,
      "num_input_tokens_seen": 1515328,
      "step": 785
    },
    {
      "epoch": 0.12888490088914267,
      "grad_norm": 0.03939088433980942,
      "learning_rate": 0.2997120904617199,
      "loss": 0.2625,
      "num_input_tokens_seen": 1525440,
      "step": 790
    },
    {
      "epoch": 0.12970062810996003,
      "grad_norm": 0.019169233739376068,
      "learning_rate": 0.29970843104226863,
      "loss": 0.2517,
      "num_input_tokens_seen": 1533792,
      "step": 795
    },
    {
      "epoch": 0.13051635533077738,
      "grad_norm": 0.02763618342578411,
      "learning_rate": 0.2997047485358959,
      "loss": 0.253,
      "num_input_tokens_seen": 1543296,
      "step": 800
    },
    {
      "epoch": 0.13051635533077738,
      "eval_loss": 0.2580569088459015,
      "eval_runtime": 68.013,
      "eval_samples_per_second": 40.066,
      "eval_steps_per_second": 20.04,
      "num_input_tokens_seen": 1543296,
      "step": 800
    },
    {
      "epoch": 0.13133208255159476,
      "grad_norm": 0.021913060918450356,
      "learning_rate": 0.2997010429431697,
      "loss": 0.2386,
      "num_input_tokens_seen": 1553120,
      "step": 805
    },
    {
      "epoch": 0.1321478097724121,
      "grad_norm": 0.042528264224529266,
      "learning_rate": 0.29969731426466134,
      "loss": 0.2523,
      "num_input_tokens_seen": 1562320,
      "step": 810
    },
    {
      "epoch": 0.13296353699322946,
      "grad_norm": 0.04152572527527809,
      "learning_rate": 0.299693562500946,
      "loss": 0.2465,
      "num_input_tokens_seen": 1572624,
      "step": 815
    },
    {
      "epoch": 0.13377926421404682,
      "grad_norm": 0.04038878157734871,
      "learning_rate": 0.29968978765260207,
      "loss": 0.2922,
      "num_input_tokens_seen": 1582400,
      "step": 820
    },
    {
      "epoch": 0.13459499143486417,
      "grad_norm": 0.036423731595277786,
      "learning_rate": 0.2996859897202118,
      "loss": 0.2486,
      "num_input_tokens_seen": 1591136,
      "step": 825
    },
    {
      "epoch": 0.13541071865568155,
      "grad_norm": 0.03515270724892616,
      "learning_rate": 0.2996821687043609,
      "loss": 0.2511,
      "num_input_tokens_seen": 1600544,
      "step": 830
    },
    {
      "epoch": 0.1362264458764989,
      "grad_norm": 0.026680519804358482,
      "learning_rate": 0.2996783246056384,
      "loss": 0.2416,
      "num_input_tokens_seen": 1609968,
      "step": 835
    },
    {
      "epoch": 0.13704217309731626,
      "grad_norm": 0.01570208929479122,
      "learning_rate": 0.29967445742463744,
      "loss": 0.2185,
      "num_input_tokens_seen": 1619232,
      "step": 840
    },
    {
      "epoch": 0.1378579003181336,
      "grad_norm": 0.014991573058068752,
      "learning_rate": 0.29967056716195417,
      "loss": 0.2314,
      "num_input_tokens_seen": 1628448,
      "step": 845
    },
    {
      "epoch": 0.13867362753895096,
      "grad_norm": 0.021561339497566223,
      "learning_rate": 0.2996666538181885,
      "loss": 0.327,
      "num_input_tokens_seen": 1639536,
      "step": 850
    },
    {
      "epoch": 0.13948935475976834,
      "grad_norm": 0.01680625043809414,
      "learning_rate": 0.29966271739394407,
      "loss": 0.2938,
      "num_input_tokens_seen": 1649840,
      "step": 855
    },
    {
      "epoch": 0.1403050819805857,
      "grad_norm": 0.023056387901306152,
      "learning_rate": 0.29965875788982776,
      "loss": 0.2591,
      "num_input_tokens_seen": 1659552,
      "step": 860
    },
    {
      "epoch": 0.14112080920140305,
      "grad_norm": 0.03358907252550125,
      "learning_rate": 0.2996547753064503,
      "loss": 0.2148,
      "num_input_tokens_seen": 1668592,
      "step": 865
    },
    {
      "epoch": 0.1419365364222204,
      "grad_norm": 0.029203729704022408,
      "learning_rate": 0.29965076964442583,
      "loss": 0.316,
      "num_input_tokens_seen": 1679968,
      "step": 870
    },
    {
      "epoch": 0.14275226364303778,
      "grad_norm": 0.17586879432201385,
      "learning_rate": 0.299646740904372,
      "loss": 0.2386,
      "num_input_tokens_seen": 1691664,
      "step": 875
    },
    {
      "epoch": 0.14356799086385513,
      "grad_norm": 0.019177429378032684,
      "learning_rate": 0.29964268908691016,
      "loss": 0.2614,
      "num_input_tokens_seen": 1701088,
      "step": 880
    },
    {
      "epoch": 0.14438371808467249,
      "grad_norm": 0.03965653106570244,
      "learning_rate": 0.29963861419266513,
      "loss": 0.2357,
      "num_input_tokens_seen": 1709904,
      "step": 885
    },
    {
      "epoch": 0.14519944530548984,
      "grad_norm": 0.07168698310852051,
      "learning_rate": 0.29963451622226533,
      "loss": 0.3085,
      "num_input_tokens_seen": 1719072,
      "step": 890
    },
    {
      "epoch": 0.1460151725263072,
      "grad_norm": 0.15158477425575256,
      "learning_rate": 0.29963039517634277,
      "loss": 0.3428,
      "num_input_tokens_seen": 1727680,
      "step": 895
    },
    {
      "epoch": 0.14683089974712457,
      "grad_norm": 0.0205735694617033,
      "learning_rate": 0.2996262510555328,
      "loss": 0.2516,
      "num_input_tokens_seen": 1736832,
      "step": 900
    },
    {
      "epoch": 0.14764662696794192,
      "grad_norm": 0.03210659325122833,
      "learning_rate": 0.2996220838604746,
      "loss": 0.2773,
      "num_input_tokens_seen": 1747056,
      "step": 905
    },
    {
      "epoch": 0.14846235418875928,
      "grad_norm": 0.021218515932559967,
      "learning_rate": 0.29961789359181085,
      "loss": 0.2518,
      "num_input_tokens_seen": 1758064,
      "step": 910
    },
    {
      "epoch": 0.14927808140957663,
      "grad_norm": 0.042214684188365936,
      "learning_rate": 0.29961368025018764,
      "loss": 0.2597,
      "num_input_tokens_seen": 1767616,
      "step": 915
    },
    {
      "epoch": 0.150093808630394,
      "grad_norm": 0.02204306051135063,
      "learning_rate": 0.2996094438362548,
      "loss": 0.2545,
      "num_input_tokens_seen": 1776752,
      "step": 920
    },
    {
      "epoch": 0.15090953585121136,
      "grad_norm": 0.03769299387931824,
      "learning_rate": 0.2996051843506657,
      "loss": 0.3277,
      "num_input_tokens_seen": 1786800,
      "step": 925
    },
    {
      "epoch": 0.15172526307202872,
      "grad_norm": 0.03295835852622986,
      "learning_rate": 0.299600901794077,
      "loss": 0.2613,
      "num_input_tokens_seen": 1796816,
      "step": 930
    },
    {
      "epoch": 0.15254099029284607,
      "grad_norm": 0.03611009195446968,
      "learning_rate": 0.29959659616714923,
      "loss": 0.2617,
      "num_input_tokens_seen": 1807472,
      "step": 935
    },
    {
      "epoch": 0.15335671751366342,
      "grad_norm": 0.03183411434292793,
      "learning_rate": 0.2995922674705464,
      "loss": 0.2633,
      "num_input_tokens_seen": 1816944,
      "step": 940
    },
    {
      "epoch": 0.1541724447344808,
      "grad_norm": 0.018565218895673752,
      "learning_rate": 0.2995879157049361,
      "loss": 0.2575,
      "num_input_tokens_seen": 1826208,
      "step": 945
    },
    {
      "epoch": 0.15498817195529815,
      "grad_norm": 0.015836656093597412,
      "learning_rate": 0.2995835408709893,
      "loss": 0.2809,
      "num_input_tokens_seen": 1836544,
      "step": 950
    },
    {
      "epoch": 0.1558038991761155,
      "grad_norm": 0.02514466643333435,
      "learning_rate": 0.29957914296938076,
      "loss": 0.2528,
      "num_input_tokens_seen": 1845696,
      "step": 955
    },
    {
      "epoch": 0.15661962639693286,
      "grad_norm": 0.020541472360491753,
      "learning_rate": 0.2995747220007886,
      "loss": 0.2363,
      "num_input_tokens_seen": 1856832,
      "step": 960
    },
    {
      "epoch": 0.1574353536177502,
      "grad_norm": 0.014091912657022476,
      "learning_rate": 0.2995702779658947,
      "loss": 0.2439,
      "num_input_tokens_seen": 1865776,
      "step": 965
    },
    {
      "epoch": 0.1582510808385676,
      "grad_norm": 0.022600017488002777,
      "learning_rate": 0.29956581086538425,
      "loss": 0.2273,
      "num_input_tokens_seen": 1874704,
      "step": 970
    },
    {
      "epoch": 0.15906680805938495,
      "grad_norm": 0.021195370703935623,
      "learning_rate": 0.2995613206999462,
      "loss": 0.2622,
      "num_input_tokens_seen": 1884384,
      "step": 975
    },
    {
      "epoch": 0.1598825352802023,
      "grad_norm": 0.02615443430840969,
      "learning_rate": 0.29955680747027297,
      "loss": 0.2278,
      "num_input_tokens_seen": 1894160,
      "step": 980
    },
    {
      "epoch": 0.16069826250101965,
      "grad_norm": 0.019718749448657036,
      "learning_rate": 0.2995522711770607,
      "loss": 0.2357,
      "num_input_tokens_seen": 1903488,
      "step": 985
    },
    {
      "epoch": 0.16151398972183703,
      "grad_norm": 0.02013973519206047,
      "learning_rate": 0.2995477118210087,
      "loss": 0.237,
      "num_input_tokens_seen": 1913728,
      "step": 990
    },
    {
      "epoch": 0.16232971694265438,
      "grad_norm": 0.017630673944950104,
      "learning_rate": 0.29954312940282024,
      "loss": 0.245,
      "num_input_tokens_seen": 1922976,
      "step": 995
    },
    {
      "epoch": 0.16314544416347174,
      "grad_norm": 0.03780928626656532,
      "learning_rate": 0.29953852392320196,
      "loss": 0.2537,
      "num_input_tokens_seen": 1931808,
      "step": 1000
    },
    {
      "epoch": 0.16314544416347174,
      "eval_loss": 0.23750628530979156,
      "eval_runtime": 68.0837,
      "eval_samples_per_second": 40.024,
      "eval_steps_per_second": 20.019,
      "num_input_tokens_seen": 1931808,
      "step": 1000
    },
    {
      "epoch": 0.1639611713842891,
      "grad_norm": 0.043303925544023514,
      "learning_rate": 0.2995338953828641,
      "loss": 0.2612,
      "num_input_tokens_seen": 1942560,
      "step": 1005
    },
    {
      "epoch": 0.16477689860510644,
      "grad_norm": 0.03866589814424515,
      "learning_rate": 0.2995292437825204,
      "loss": 0.2447,
      "num_input_tokens_seen": 1950912,
      "step": 1010
    },
    {
      "epoch": 0.16559262582592382,
      "grad_norm": 0.017779001966118813,
      "learning_rate": 0.29952456912288816,
      "loss": 0.3101,
      "num_input_tokens_seen": 1961696,
      "step": 1015
    },
    {
      "epoch": 0.16640835304674118,
      "grad_norm": 0.017399419099092484,
      "learning_rate": 0.2995198714046884,
      "loss": 0.2188,
      "num_input_tokens_seen": 1971280,
      "step": 1020
    },
    {
      "epoch": 0.16722408026755853,
      "grad_norm": 0.011151296086609364,
      "learning_rate": 0.2995151506286454,
      "loss": 0.2359,
      "num_input_tokens_seen": 1981200,
      "step": 1025
    },
    {
      "epoch": 0.16803980748837588,
      "grad_norm": 0.026728691533207893,
      "learning_rate": 0.2995104067954873,
      "loss": 0.2502,
      "num_input_tokens_seen": 1990400,
      "step": 1030
    },
    {
      "epoch": 0.16885553470919323,
      "grad_norm": 0.020308909937739372,
      "learning_rate": 0.2995056399059456,
      "loss": 0.2378,
      "num_input_tokens_seen": 2000016,
      "step": 1035
    },
    {
      "epoch": 0.16967126193001061,
      "grad_norm": 0.02437741495668888,
      "learning_rate": 0.2995008499607554,
      "loss": 0.2248,
      "num_input_tokens_seen": 2009600,
      "step": 1040
    },
    {
      "epoch": 0.17048698915082797,
      "grad_norm": 0.017737720161676407,
      "learning_rate": 0.2994960369606554,
      "loss": 0.236,
      "num_input_tokens_seen": 2019696,
      "step": 1045
    },
    {
      "epoch": 0.17130271637164532,
      "grad_norm": 0.016439391300082207,
      "learning_rate": 0.2994912009063878,
      "loss": 0.2448,
      "num_input_tokens_seen": 2029776,
      "step": 1050
    },
    {
      "epoch": 0.17211844359246267,
      "grad_norm": 0.032079122960567474,
      "learning_rate": 0.29948634179869843,
      "loss": 0.2734,
      "num_input_tokens_seen": 2037728,
      "step": 1055
    },
    {
      "epoch": 0.17293417081328005,
      "grad_norm": 0.022181715816259384,
      "learning_rate": 0.29948145963833656,
      "loss": 0.232,
      "num_input_tokens_seen": 2047616,
      "step": 1060
    },
    {
      "epoch": 0.1737498980340974,
      "grad_norm": 0.016368746757507324,
      "learning_rate": 0.29947655442605514,
      "loss": 0.2136,
      "num_input_tokens_seen": 2057440,
      "step": 1065
    },
    {
      "epoch": 0.17456562525491476,
      "grad_norm": 0.012836229987442493,
      "learning_rate": 0.2994716261626106,
      "loss": 0.2148,
      "num_input_tokens_seen": 2066528,
      "step": 1070
    },
    {
      "epoch": 0.1753813524757321,
      "grad_norm": 0.030164530500769615,
      "learning_rate": 0.2994666748487629,
      "loss": 0.2343,
      "num_input_tokens_seen": 2076288,
      "step": 1075
    },
    {
      "epoch": 0.17619707969654946,
      "grad_norm": 0.01976616308093071,
      "learning_rate": 0.2994617004852756,
      "loss": 0.2852,
      "num_input_tokens_seen": 2085920,
      "step": 1080
    },
    {
      "epoch": 0.17701280691736684,
      "grad_norm": 0.020619258284568787,
      "learning_rate": 0.2994567030729159,
      "loss": 0.2928,
      "num_input_tokens_seen": 2095344,
      "step": 1085
    },
    {
      "epoch": 0.1778285341381842,
      "grad_norm": 0.021567746996879578,
      "learning_rate": 0.29945168261245436,
      "loss": 0.26,
      "num_input_tokens_seen": 2104928,
      "step": 1090
    },
    {
      "epoch": 0.17864426135900155,
      "grad_norm": 0.026928987354040146,
      "learning_rate": 0.29944663910466524,
      "loss": 0.27,
      "num_input_tokens_seen": 2117072,
      "step": 1095
    },
    {
      "epoch": 0.1794599885798189,
      "grad_norm": 0.011312106624245644,
      "learning_rate": 0.2994415725503263,
      "loss": 0.2462,
      "num_input_tokens_seen": 2126640,
      "step": 1100
    },
    {
      "epoch": 0.18027571580063625,
      "grad_norm": 0.016007540747523308,
      "learning_rate": 0.29943648295021885,
      "loss": 0.2713,
      "num_input_tokens_seen": 2136112,
      "step": 1105
    },
    {
      "epoch": 0.18109144302145364,
      "grad_norm": 0.012129426933825016,
      "learning_rate": 0.2994313703051278,
      "loss": 0.2431,
      "num_input_tokens_seen": 2145904,
      "step": 1110
    },
    {
      "epoch": 0.181907170242271,
      "grad_norm": 0.023522427305579185,
      "learning_rate": 0.29942623461584156,
      "loss": 0.2989,
      "num_input_tokens_seen": 2157072,
      "step": 1115
    },
    {
      "epoch": 0.18272289746308834,
      "grad_norm": 0.03529580309987068,
      "learning_rate": 0.29942107588315214,
      "loss": 0.2092,
      "num_input_tokens_seen": 2166624,
      "step": 1120
    },
    {
      "epoch": 0.1835386246839057,
      "grad_norm": 0.014797593466937542,
      "learning_rate": 0.29941589410785513,
      "loss": 0.2165,
      "num_input_tokens_seen": 2176224,
      "step": 1125
    },
    {
      "epoch": 0.18435435190472307,
      "grad_norm": 0.0161149725317955,
      "learning_rate": 0.29941068929074954,
      "loss": 0.2627,
      "num_input_tokens_seen": 2185856,
      "step": 1130
    },
    {
      "epoch": 0.18517007912554043,
      "grad_norm": 0.08100460469722748,
      "learning_rate": 0.2994054614326381,
      "loss": 0.3159,
      "num_input_tokens_seen": 2194608,
      "step": 1135
    },
    {
      "epoch": 0.18598580634635778,
      "grad_norm": 0.06770962476730347,
      "learning_rate": 0.29940021053432686,
      "loss": 0.2846,
      "num_input_tokens_seen": 2203312,
      "step": 1140
    },
    {
      "epoch": 0.18680153356717513,
      "grad_norm": 0.03201225772500038,
      "learning_rate": 0.29939493659662575,
      "loss": 0.2632,
      "num_input_tokens_seen": 2214048,
      "step": 1145
    },
    {
      "epoch": 0.18761726078799248,
      "grad_norm": 0.043043773621320724,
      "learning_rate": 0.299389639620348,
      "loss": 0.2112,
      "num_input_tokens_seen": 2222032,
      "step": 1150
    },
    {
      "epoch": 0.18843298800880987,
      "grad_norm": 0.047011177986860275,
      "learning_rate": 0.29938431960631046,
      "loss": 0.2417,
      "num_input_tokens_seen": 2231232,
      "step": 1155
    },
    {
      "epoch": 0.18924871522962722,
      "grad_norm": 0.026658790186047554,
      "learning_rate": 0.2993789765553335,
      "loss": 0.4252,
      "num_input_tokens_seen": 2240128,
      "step": 1160
    },
    {
      "epoch": 0.19006444245044457,
      "grad_norm": 0.045968927443027496,
      "learning_rate": 0.2993736104682412,
      "loss": 0.2582,
      "num_input_tokens_seen": 2250016,
      "step": 1165
    },
    {
      "epoch": 0.19088016967126192,
      "grad_norm": 0.13606494665145874,
      "learning_rate": 0.299368221345861,
      "loss": 0.2912,
      "num_input_tokens_seen": 2260192,
      "step": 1170
    },
    {
      "epoch": 0.19169589689207928,
      "grad_norm": 0.0242098867893219,
      "learning_rate": 0.29936280918902397,
      "loss": 0.3079,
      "num_input_tokens_seen": 2270288,
      "step": 1175
    },
    {
      "epoch": 0.19251162411289666,
      "grad_norm": 0.17886605858802795,
      "learning_rate": 0.2993573739985648,
      "loss": 0.4878,
      "num_input_tokens_seen": 2280112,
      "step": 1180
    },
    {
      "epoch": 0.193327351333714,
      "grad_norm": 0.04239461570978165,
      "learning_rate": 0.2993519157753216,
      "loss": 0.3486,
      "num_input_tokens_seen": 2290160,
      "step": 1185
    },
    {
      "epoch": 0.19414307855453136,
      "grad_norm": 0.04345610365271568,
      "learning_rate": 0.2993464345201361,
      "loss": 0.3035,
      "num_input_tokens_seen": 2298560,
      "step": 1190
    },
    {
      "epoch": 0.19495880577534871,
      "grad_norm": 0.12641026079654694,
      "learning_rate": 0.2993409302338536,
      "loss": 0.2136,
      "num_input_tokens_seen": 2307312,
      "step": 1195
    },
    {
      "epoch": 0.1957745329961661,
      "grad_norm": 0.06344906985759735,
      "learning_rate": 0.2993354029173229,
      "loss": 0.3811,
      "num_input_tokens_seen": 2315744,
      "step": 1200
    },
    {
      "epoch": 0.1957745329961661,
      "eval_loss": 0.2576513886451721,
      "eval_runtime": 68.0148,
      "eval_samples_per_second": 40.065,
      "eval_steps_per_second": 20.04,
      "num_input_tokens_seen": 2315744,
      "step": 1200
    },
    {
      "epoch": 0.19659026021698345,
      "grad_norm": 0.07322939485311508,
      "learning_rate": 0.2993298525713965,
      "loss": 0.2945,
      "num_input_tokens_seen": 2325456,
      "step": 1205
    },
    {
      "epoch": 0.1974059874378008,
      "grad_norm": 0.05088430643081665,
      "learning_rate": 0.29932427919693017,
      "loss": 0.2667,
      "num_input_tokens_seen": 2334768,
      "step": 1210
    },
    {
      "epoch": 0.19822171465861815,
      "grad_norm": 0.06809491664171219,
      "learning_rate": 0.2993186827947834,
      "loss": 0.3504,
      "num_input_tokens_seen": 2345232,
      "step": 1215
    },
    {
      "epoch": 0.1990374418794355,
      "grad_norm": 0.025568358600139618,
      "learning_rate": 0.2993130633658194,
      "loss": 0.2264,
      "num_input_tokens_seen": 2355424,
      "step": 1220
    },
    {
      "epoch": 0.1998531691002529,
      "grad_norm": 0.038411110639572144,
      "learning_rate": 0.29930742091090456,
      "loss": 0.2062,
      "num_input_tokens_seen": 2364576,
      "step": 1225
    },
    {
      "epoch": 0.20066889632107024,
      "grad_norm": 0.037040602415800095,
      "learning_rate": 0.29930175543090914,
      "loss": 0.3315,
      "num_input_tokens_seen": 2375488,
      "step": 1230
    },
    {
      "epoch": 0.2014846235418876,
      "grad_norm": 0.04835139960050583,
      "learning_rate": 0.2992960669267068,
      "loss": 0.2723,
      "num_input_tokens_seen": 2386272,
      "step": 1235
    },
    {
      "epoch": 0.20230035076270494,
      "grad_norm": 0.03658295050263405,
      "learning_rate": 0.29929035539917476,
      "loss": 0.2598,
      "num_input_tokens_seen": 2395952,
      "step": 1240
    },
    {
      "epoch": 0.2031160779835223,
      "grad_norm": 0.0809701532125473,
      "learning_rate": 0.2992846208491938,
      "loss": 0.4136,
      "num_input_tokens_seen": 2404944,
      "step": 1245
    },
    {
      "epoch": 0.20393180520433968,
      "grad_norm": 0.04176202788949013,
      "learning_rate": 0.2992788632776483,
      "loss": 0.238,
      "num_input_tokens_seen": 2414656,
      "step": 1250
    },
    {
      "epoch": 0.20474753242515703,
      "grad_norm": 0.03963752090930939,
      "learning_rate": 0.29927308268542613,
      "loss": 0.2925,
      "num_input_tokens_seen": 2423456,
      "step": 1255
    },
    {
      "epoch": 0.20556325964597438,
      "grad_norm": 0.02597838081419468,
      "learning_rate": 0.2992672790734187,
      "loss": 0.2499,
      "num_input_tokens_seen": 2434368,
      "step": 1260
    },
    {
      "epoch": 0.20637898686679174,
      "grad_norm": 0.02145605906844139,
      "learning_rate": 0.299261452442521,
      "loss": 0.2583,
      "num_input_tokens_seen": 2444240,
      "step": 1265
    },
    {
      "epoch": 0.20719471408760912,
      "grad_norm": 0.023017995059490204,
      "learning_rate": 0.29925560279363167,
      "loss": 0.2685,
      "num_input_tokens_seen": 2453456,
      "step": 1270
    },
    {
      "epoch": 0.20801044130842647,
      "grad_norm": 0.03476202115416527,
      "learning_rate": 0.29924973012765266,
      "loss": 0.2866,
      "num_input_tokens_seen": 2463504,
      "step": 1275
    },
    {
      "epoch": 0.20882616852924382,
      "grad_norm": 0.04005228728055954,
      "learning_rate": 0.29924383444548974,
      "loss": 0.2628,
      "num_input_tokens_seen": 2473408,
      "step": 1280
    },
    {
      "epoch": 0.20964189575006117,
      "grad_norm": 0.022579027339816093,
      "learning_rate": 0.299237915748052,
      "loss": 0.2801,
      "num_input_tokens_seen": 2482832,
      "step": 1285
    },
    {
      "epoch": 0.21045762297087853,
      "grad_norm": 0.024342620745301247,
      "learning_rate": 0.2992319740362522,
      "loss": 0.2695,
      "num_input_tokens_seen": 2492720,
      "step": 1290
    },
    {
      "epoch": 0.2112733501916959,
      "grad_norm": 0.03749721869826317,
      "learning_rate": 0.2992260093110066,
      "loss": 0.2383,
      "num_input_tokens_seen": 2503136,
      "step": 1295
    },
    {
      "epoch": 0.21208907741251326,
      "grad_norm": 0.030770806595683098,
      "learning_rate": 0.2992200215732352,
      "loss": 0.2369,
      "num_input_tokens_seen": 2511568,
      "step": 1300
    },
    {
      "epoch": 0.2129048046333306,
      "grad_norm": 0.05151071399450302,
      "learning_rate": 0.2992140108238611,
      "loss": 0.2473,
      "num_input_tokens_seen": 2522256,
      "step": 1305
    },
    {
      "epoch": 0.21372053185414797,
      "grad_norm": 0.027141008526086807,
      "learning_rate": 0.2992079770638115,
      "loss": 0.2482,
      "num_input_tokens_seen": 2532016,
      "step": 1310
    },
    {
      "epoch": 0.21453625907496532,
      "grad_norm": 0.036679357290267944,
      "learning_rate": 0.29920192029401677,
      "loss": 0.2495,
      "num_input_tokens_seen": 2540800,
      "step": 1315
    },
    {
      "epoch": 0.2153519862957827,
      "grad_norm": 0.07340593636035919,
      "learning_rate": 0.2991958405154109,
      "loss": 0.205,
      "num_input_tokens_seen": 2551200,
      "step": 1320
    },
    {
      "epoch": 0.21616771351660005,
      "grad_norm": 0.015467182733118534,
      "learning_rate": 0.29918973772893154,
      "loss": 0.2202,
      "num_input_tokens_seen": 2561312,
      "step": 1325
    },
    {
      "epoch": 0.2169834407374174,
      "grad_norm": 0.026317765936255455,
      "learning_rate": 0.29918361193551973,
      "loss": 0.2328,
      "num_input_tokens_seen": 2570784,
      "step": 1330
    },
    {
      "epoch": 0.21779916795823476,
      "grad_norm": 0.020446252077817917,
      "learning_rate": 0.29917746313612026,
      "loss": 0.2381,
      "num_input_tokens_seen": 2581168,
      "step": 1335
    },
    {
      "epoch": 0.21861489517905214,
      "grad_norm": 0.0749962329864502,
      "learning_rate": 0.29917129133168124,
      "loss": 0.2373,
      "num_input_tokens_seen": 2590416,
      "step": 1340
    },
    {
      "epoch": 0.2194306223998695,
      "grad_norm": 0.056200698018074036,
      "learning_rate": 0.2991650965231546,
      "loss": 0.3164,
      "num_input_tokens_seen": 2601056,
      "step": 1345
    },
    {
      "epoch": 0.22024634962068684,
      "grad_norm": 0.09676844626665115,
      "learning_rate": 0.29915887871149544,
      "loss": 0.2878,
      "num_input_tokens_seen": 2611152,
      "step": 1350
    },
    {
      "epoch": 0.2210620768415042,
      "grad_norm": 0.3465587794780731,
      "learning_rate": 0.2991526378976628,
      "loss": 0.2677,
      "num_input_tokens_seen": 2620240,
      "step": 1355
    },
    {
      "epoch": 0.22187780406232155,
      "grad_norm": 0.09490274637937546,
      "learning_rate": 0.29914637408261896,
      "loss": 0.2493,
      "num_input_tokens_seen": 2630912,
      "step": 1360
    },
    {
      "epoch": 0.22269353128313893,
      "grad_norm": 0.11033900082111359,
      "learning_rate": 0.29914008726733,
      "loss": 0.2769,
      "num_input_tokens_seen": 2640560,
      "step": 1365
    },
    {
      "epoch": 0.22350925850395628,
      "grad_norm": 0.15545983612537384,
      "learning_rate": 0.2991337774527653,
      "loss": 0.2913,
      "num_input_tokens_seen": 2650864,
      "step": 1370
    },
    {
      "epoch": 0.22432498572477363,
      "grad_norm": 0.04786358401179314,
      "learning_rate": 0.2991274446398981,
      "loss": 0.2258,
      "num_input_tokens_seen": 2661920,
      "step": 1375
    },
    {
      "epoch": 0.225140712945591,
      "grad_norm": 0.04562338814139366,
      "learning_rate": 0.29912108882970484,
      "loss": 0.2199,
      "num_input_tokens_seen": 2671648,
      "step": 1380
    },
    {
      "epoch": 0.22595644016640834,
      "grad_norm": 0.031428609043359756,
      "learning_rate": 0.2991147100231657,
      "loss": 0.2683,
      "num_input_tokens_seen": 2681520,
      "step": 1385
    },
    {
      "epoch": 0.22677216738722572,
      "grad_norm": 0.0442502424120903,
      "learning_rate": 0.2991083082212644,
      "loss": 0.29,
      "num_input_tokens_seen": 2690720,
      "step": 1390
    },
    {
      "epoch": 0.22758789460804307,
      "grad_norm": 0.04914036765694618,
      "learning_rate": 0.2991018834249881,
      "loss": 0.2413,
      "num_input_tokens_seen": 2701280,
      "step": 1395
    },
    {
      "epoch": 0.22840362182886043,
      "grad_norm": 0.05887686833739281,
      "learning_rate": 0.29909543563532764,
      "loss": 0.2942,
      "num_input_tokens_seen": 2710208,
      "step": 1400
    },
    {
      "epoch": 0.22840362182886043,
      "eval_loss": 0.2469376027584076,
      "eval_runtime": 68.0174,
      "eval_samples_per_second": 40.063,
      "eval_steps_per_second": 20.039,
      "num_input_tokens_seen": 2710208,
      "step": 1400
    },
    {
      "epoch": 0.22921934904967778,
      "grad_norm": 0.04807290807366371,
      "learning_rate": 0.29908896485327746,
      "loss": 0.2811,
      "num_input_tokens_seen": 2721632,
      "step": 1405
    },
    {
      "epoch": 0.23003507627049516,
      "grad_norm": 0.16183260083198547,
      "learning_rate": 0.29908247107983527,
      "loss": 0.2247,
      "num_input_tokens_seen": 2731008,
      "step": 1410
    },
    {
      "epoch": 0.2308508034913125,
      "grad_norm": 0.059330087155103683,
      "learning_rate": 0.29907595431600253,
      "loss": 0.5385,
      "num_input_tokens_seen": 2740128,
      "step": 1415
    },
    {
      "epoch": 0.23166653071212986,
      "grad_norm": 0.023337719962000847,
      "learning_rate": 0.29906941456278424,
      "loss": 0.2707,
      "num_input_tokens_seen": 2749328,
      "step": 1420
    },
    {
      "epoch": 0.23248225793294722,
      "grad_norm": 0.1111445352435112,
      "learning_rate": 0.2990628518211889,
      "loss": 0.3067,
      "num_input_tokens_seen": 2757456,
      "step": 1425
    },
    {
      "epoch": 0.23329798515376457,
      "grad_norm": 0.04879055917263031,
      "learning_rate": 0.2990562660922286,
      "loss": 0.2854,
      "num_input_tokens_seen": 2765200,
      "step": 1430
    },
    {
      "epoch": 0.23411371237458195,
      "grad_norm": 0.022454475983977318,
      "learning_rate": 0.2990496573769189,
      "loss": 0.2685,
      "num_input_tokens_seen": 2775072,
      "step": 1435
    },
    {
      "epoch": 0.2349294395953993,
      "grad_norm": 0.02485763467848301,
      "learning_rate": 0.29904302567627894,
      "loss": 0.2376,
      "num_input_tokens_seen": 2784896,
      "step": 1440
    },
    {
      "epoch": 0.23574516681621666,
      "grad_norm": 0.02832135185599327,
      "learning_rate": 0.2990363709913314,
      "loss": 0.268,
      "num_input_tokens_seen": 2794272,
      "step": 1445
    },
    {
      "epoch": 0.236560894037034,
      "grad_norm": 0.026127714663743973,
      "learning_rate": 0.29902969332310264,
      "loss": 0.2297,
      "num_input_tokens_seen": 2803072,
      "step": 1450
    },
    {
      "epoch": 0.23737662125785136,
      "grad_norm": 0.01945984549820423,
      "learning_rate": 0.2990229926726223,
      "loss": 0.312,
      "num_input_tokens_seen": 2812528,
      "step": 1455
    },
    {
      "epoch": 0.23819234847866874,
      "grad_norm": 0.02968493476510048,
      "learning_rate": 0.29901626904092365,
      "loss": 0.2096,
      "num_input_tokens_seen": 2822464,
      "step": 1460
    },
    {
      "epoch": 0.2390080756994861,
      "grad_norm": 0.010884271003305912,
      "learning_rate": 0.2990095224290438,
      "loss": 0.2732,
      "num_input_tokens_seen": 2832832,
      "step": 1465
    },
    {
      "epoch": 0.23982380292030345,
      "grad_norm": 0.01663624681532383,
      "learning_rate": 0.29900275283802297,
      "loss": 0.2619,
      "num_input_tokens_seen": 2844336,
      "step": 1470
    },
    {
      "epoch": 0.2406395301411208,
      "grad_norm": 0.013137167319655418,
      "learning_rate": 0.2989959602689051,
      "loss": 0.2281,
      "num_input_tokens_seen": 2854176,
      "step": 1475
    },
    {
      "epoch": 0.24145525736193818,
      "grad_norm": 0.026567043736577034,
      "learning_rate": 0.2989891447227379,
      "loss": 0.2448,
      "num_input_tokens_seen": 2863888,
      "step": 1480
    },
    {
      "epoch": 0.24227098458275553,
      "grad_norm": 0.015590973198413849,
      "learning_rate": 0.29898230620057215,
      "loss": 0.2509,
      "num_input_tokens_seen": 2872640,
      "step": 1485
    },
    {
      "epoch": 0.24308671180357289,
      "grad_norm": 0.020808259025216103,
      "learning_rate": 0.2989754447034626,
      "loss": 0.2226,
      "num_input_tokens_seen": 2881488,
      "step": 1490
    },
    {
      "epoch": 0.24390243902439024,
      "grad_norm": 0.0325363390147686,
      "learning_rate": 0.2989685602324673,
      "loss": 0.322,
      "num_input_tokens_seen": 2892336,
      "step": 1495
    },
    {
      "epoch": 0.2447181662452076,
      "grad_norm": 0.019168391823768616,
      "learning_rate": 0.298961652788648,
      "loss": 0.2482,
      "num_input_tokens_seen": 2901600,
      "step": 1500
    },
    {
      "epoch": 0.24553389346602497,
      "grad_norm": 0.020617805421352386,
      "learning_rate": 0.29895472237306986,
      "loss": 0.2084,
      "num_input_tokens_seen": 2911664,
      "step": 1505
    },
    {
      "epoch": 0.24634962068684232,
      "grad_norm": 0.023142334073781967,
      "learning_rate": 0.29894776898680164,
      "loss": 0.2471,
      "num_input_tokens_seen": 2920976,
      "step": 1510
    },
    {
      "epoch": 0.24716534790765968,
      "grad_norm": 0.024718603119254112,
      "learning_rate": 0.29894079263091566,
      "loss": 0.2884,
      "num_input_tokens_seen": 2930864,
      "step": 1515
    },
    {
      "epoch": 0.24798107512847703,
      "grad_norm": 0.014776915311813354,
      "learning_rate": 0.2989337933064877,
      "loss": 0.2283,
      "num_input_tokens_seen": 2940944,
      "step": 1520
    },
    {
      "epoch": 0.24879680234929438,
      "grad_norm": 0.020364873111248016,
      "learning_rate": 0.29892677101459725,
      "loss": 0.2504,
      "num_input_tokens_seen": 2950768,
      "step": 1525
    },
    {
      "epoch": 0.24961252957011176,
      "grad_norm": 0.013527288101613522,
      "learning_rate": 0.2989197257563272,
      "loss": 0.209,
      "num_input_tokens_seen": 2959600,
      "step": 1530
    },
    {
      "epoch": 0.2504282567909291,
      "grad_norm": 0.018316270783543587,
      "learning_rate": 0.2989126575327639,
      "loss": 0.2489,
      "num_input_tokens_seen": 2970000,
      "step": 1535
    },
    {
      "epoch": 0.25124398401174647,
      "grad_norm": 0.01657470315694809,
      "learning_rate": 0.29890556634499754,
      "loss": 0.2312,
      "num_input_tokens_seen": 2980640,
      "step": 1540
    },
    {
      "epoch": 0.2520597112325638,
      "grad_norm": 0.015306548215448856,
      "learning_rate": 0.2988984521941216,
      "loss": 0.2487,
      "num_input_tokens_seen": 2990448,
      "step": 1545
    },
    {
      "epoch": 0.2528754384533812,
      "grad_norm": 0.014050760306417942,
      "learning_rate": 0.29889131508123307,
      "loss": 0.2745,
      "num_input_tokens_seen": 2999568,
      "step": 1550
    },
    {
      "epoch": 0.2536911656741985,
      "grad_norm": 0.028635557740926743,
      "learning_rate": 0.2988841550074327,
      "loss": 0.2431,
      "num_input_tokens_seen": 3009136,
      "step": 1555
    },
    {
      "epoch": 0.2545068928950159,
      "grad_norm": 0.015945468097925186,
      "learning_rate": 0.2988769719738246,
      "loss": 0.2275,
      "num_input_tokens_seen": 3018272,
      "step": 1560
    },
    {
      "epoch": 0.2553226201158333,
      "grad_norm": 0.02153211086988449,
      "learning_rate": 0.29886976598151666,
      "loss": 0.2583,
      "num_input_tokens_seen": 3027856,
      "step": 1565
    },
    {
      "epoch": 0.25613834733665064,
      "grad_norm": 0.020287469029426575,
      "learning_rate": 0.29886253703161986,
      "loss": 0.2688,
      "num_input_tokens_seen": 3037584,
      "step": 1570
    },
    {
      "epoch": 0.256954074557468,
      "grad_norm": 0.021463366225361824,
      "learning_rate": 0.29885528512524917,
      "loss": 0.2361,
      "num_input_tokens_seen": 3045856,
      "step": 1575
    },
    {
      "epoch": 0.25776980177828535,
      "grad_norm": 0.019537057727575302,
      "learning_rate": 0.29884801026352287,
      "loss": 0.2891,
      "num_input_tokens_seen": 3053872,
      "step": 1580
    },
    {
      "epoch": 0.2585855289991027,
      "grad_norm": 0.024346167221665382,
      "learning_rate": 0.2988407124475629,
      "loss": 0.2483,
      "num_input_tokens_seen": 3064560,
      "step": 1585
    },
    {
      "epoch": 0.25940125621992005,
      "grad_norm": 0.027995428070425987,
      "learning_rate": 0.2988333916784945,
      "loss": 0.2701,
      "num_input_tokens_seen": 3075408,
      "step": 1590
    },
    {
      "epoch": 0.2602169834407374,
      "grad_norm": 0.0284879170358181,
      "learning_rate": 0.2988260479574468,
      "loss": 0.2496,
      "num_input_tokens_seen": 3086240,
      "step": 1595
    },
    {
      "epoch": 0.26103271066155476,
      "grad_norm": 0.011892419308423996,
      "learning_rate": 0.2988186812855523,
      "loss": 0.2568,
      "num_input_tokens_seen": 3095216,
      "step": 1600
    },
    {
      "epoch": 0.26103271066155476,
      "eval_loss": 0.2531068027019501,
      "eval_runtime": 68.0074,
      "eval_samples_per_second": 40.069,
      "eval_steps_per_second": 20.042,
      "num_input_tokens_seen": 3095216,
      "step": 1600
    },
    {
      "epoch": 0.2618484378823721,
      "grad_norm": 0.013485066592693329,
      "learning_rate": 0.29881129166394693,
      "loss": 0.2449,
      "num_input_tokens_seen": 3105056,
      "step": 1605
    },
    {
      "epoch": 0.2626641651031895,
      "grad_norm": 0.01846809685230255,
      "learning_rate": 0.29880387909377026,
      "loss": 0.2954,
      "num_input_tokens_seen": 3116064,
      "step": 1610
    },
    {
      "epoch": 0.26347989232400687,
      "grad_norm": 0.014095965772867203,
      "learning_rate": 0.2987964435761655,
      "loss": 0.2283,
      "num_input_tokens_seen": 3126272,
      "step": 1615
    },
    {
      "epoch": 0.2642956195448242,
      "grad_norm": 0.024438492953777313,
      "learning_rate": 0.29878898511227925,
      "loss": 0.3124,
      "num_input_tokens_seen": 3135520,
      "step": 1620
    },
    {
      "epoch": 0.2651113467656416,
      "grad_norm": 0.018092269077897072,
      "learning_rate": 0.2987815037032617,
      "loss": 0.2996,
      "num_input_tokens_seen": 3145680,
      "step": 1625
    },
    {
      "epoch": 0.26592707398645893,
      "grad_norm": 0.014274140819907188,
      "learning_rate": 0.29877399935026655,
      "loss": 0.2156,
      "num_input_tokens_seen": 3156032,
      "step": 1630
    },
    {
      "epoch": 0.2667428012072763,
      "grad_norm": 0.027956411242485046,
      "learning_rate": 0.2987664720544511,
      "loss": 0.2816,
      "num_input_tokens_seen": 3164912,
      "step": 1635
    },
    {
      "epoch": 0.26755852842809363,
      "grad_norm": 0.013874692842364311,
      "learning_rate": 0.2987589218169761,
      "loss": 0.2417,
      "num_input_tokens_seen": 3175056,
      "step": 1640
    },
    {
      "epoch": 0.268374255648911,
      "grad_norm": 0.023834269493818283,
      "learning_rate": 0.29875134863900604,
      "loss": 0.2269,
      "num_input_tokens_seen": 3184976,
      "step": 1645
    },
    {
      "epoch": 0.26918998286972834,
      "grad_norm": 0.017312226817011833,
      "learning_rate": 0.29874375252170865,
      "loss": 0.2281,
      "num_input_tokens_seen": 3195552,
      "step": 1650
    },
    {
      "epoch": 0.27000571009054575,
      "grad_norm": 0.018434591591358185,
      "learning_rate": 0.2987361334662553,
      "loss": 0.2392,
      "num_input_tokens_seen": 3205312,
      "step": 1655
    },
    {
      "epoch": 0.2708214373113631,
      "grad_norm": 0.01933959499001503,
      "learning_rate": 0.29872849147382113,
      "loss": 0.2441,
      "num_input_tokens_seen": 3215216,
      "step": 1660
    },
    {
      "epoch": 0.27163716453218045,
      "grad_norm": 0.02780141495168209,
      "learning_rate": 0.2987208265455845,
      "loss": 0.2764,
      "num_input_tokens_seen": 3225024,
      "step": 1665
    },
    {
      "epoch": 0.2724528917529978,
      "grad_norm": 0.02729896642267704,
      "learning_rate": 0.29871313868272753,
      "loss": 0.2507,
      "num_input_tokens_seen": 3234032,
      "step": 1670
    },
    {
      "epoch": 0.27326861897381516,
      "grad_norm": 0.03337512165307999,
      "learning_rate": 0.29870542788643567,
      "loss": 0.2473,
      "num_input_tokens_seen": 3244160,
      "step": 1675
    },
    {
      "epoch": 0.2740843461946325,
      "grad_norm": 0.019930874928832054,
      "learning_rate": 0.2986976941578981,
      "loss": 0.2422,
      "num_input_tokens_seen": 3254144,
      "step": 1680
    },
    {
      "epoch": 0.27490007341544986,
      "grad_norm": 0.015144549310207367,
      "learning_rate": 0.29868993749830747,
      "loss": 0.2367,
      "num_input_tokens_seen": 3264400,
      "step": 1685
    },
    {
      "epoch": 0.2757158006362672,
      "grad_norm": 0.024211153388023376,
      "learning_rate": 0.2986821579088598,
      "loss": 0.2426,
      "num_input_tokens_seen": 3273456,
      "step": 1690
    },
    {
      "epoch": 0.27653152785708457,
      "grad_norm": 0.017512239515781403,
      "learning_rate": 0.29867435539075504,
      "loss": 0.2572,
      "num_input_tokens_seen": 3283680,
      "step": 1695
    },
    {
      "epoch": 0.2773472550779019,
      "grad_norm": 0.01602892577648163,
      "learning_rate": 0.2986665299451963,
      "loss": 0.2423,
      "num_input_tokens_seen": 3292960,
      "step": 1700
    },
    {
      "epoch": 0.27816298229871933,
      "grad_norm": 0.01579761691391468,
      "learning_rate": 0.29865868157339037,
      "loss": 0.2311,
      "num_input_tokens_seen": 3302160,
      "step": 1705
    },
    {
      "epoch": 0.2789787095195367,
      "grad_norm": 0.02392672747373581,
      "learning_rate": 0.2986508102765476,
      "loss": 0.235,
      "num_input_tokens_seen": 3312400,
      "step": 1710
    },
    {
      "epoch": 0.27979443674035404,
      "grad_norm": 0.018131501972675323,
      "learning_rate": 0.2986429160558818,
      "loss": 0.2432,
      "num_input_tokens_seen": 3322352,
      "step": 1715
    },
    {
      "epoch": 0.2806101639611714,
      "grad_norm": 0.017074281349778175,
      "learning_rate": 0.2986349989126104,
      "loss": 0.2532,
      "num_input_tokens_seen": 3331536,
      "step": 1720
    },
    {
      "epoch": 0.28142589118198874,
      "grad_norm": 0.01833076775074005,
      "learning_rate": 0.29862705884795426,
      "loss": 0.2228,
      "num_input_tokens_seen": 3341632,
      "step": 1725
    },
    {
      "epoch": 0.2822416184028061,
      "grad_norm": 0.017974432557821274,
      "learning_rate": 0.2986190958631379,
      "loss": 0.2158,
      "num_input_tokens_seen": 3351200,
      "step": 1730
    },
    {
      "epoch": 0.28305734562362345,
      "grad_norm": 0.02273714914917946,
      "learning_rate": 0.29861110995938933,
      "loss": 0.2863,
      "num_input_tokens_seen": 3360480,
      "step": 1735
    },
    {
      "epoch": 0.2838730728444408,
      "grad_norm": 0.016222745180130005,
      "learning_rate": 0.29860310113794,
      "loss": 0.2283,
      "num_input_tokens_seen": 3368848,
      "step": 1740
    },
    {
      "epoch": 0.28468880006525815,
      "grad_norm": 0.02996375784277916,
      "learning_rate": 0.29859506940002506,
      "loss": 0.3003,
      "num_input_tokens_seen": 3377680,
      "step": 1745
    },
    {
      "epoch": 0.28550452728607556,
      "grad_norm": 0.012816586531698704,
      "learning_rate": 0.298587014746883,
      "loss": 0.3291,
      "num_input_tokens_seen": 3388528,
      "step": 1750
    },
    {
      "epoch": 0.2863202545068929,
      "grad_norm": 0.021305255591869354,
      "learning_rate": 0.298578937179756,
      "loss": 0.2327,
      "num_input_tokens_seen": 3398448,
      "step": 1755
    },
    {
      "epoch": 0.28713598172771027,
      "grad_norm": 0.02489680051803589,
      "learning_rate": 0.29857083669988976,
      "loss": 0.2483,
      "num_input_tokens_seen": 3407248,
      "step": 1760
    },
    {
      "epoch": 0.2879517089485276,
      "grad_norm": 0.015895960852503777,
      "learning_rate": 0.29856271330853346,
      "loss": 0.218,
      "num_input_tokens_seen": 3416640,
      "step": 1765
    },
    {
      "epoch": 0.28876743616934497,
      "grad_norm": 0.012696351855993271,
      "learning_rate": 0.2985545670069398,
      "loss": 0.3001,
      "num_input_tokens_seen": 3426016,
      "step": 1770
    },
    {
      "epoch": 0.2895831633901623,
      "grad_norm": 0.0215870663523674,
      "learning_rate": 0.29854639779636505,
      "loss": 0.2666,
      "num_input_tokens_seen": 3436160,
      "step": 1775
    },
    {
      "epoch": 0.2903988906109797,
      "grad_norm": 0.019986232742667198,
      "learning_rate": 0.298538205678069,
      "loss": 0.2293,
      "num_input_tokens_seen": 3446384,
      "step": 1780
    },
    {
      "epoch": 0.29121461783179703,
      "grad_norm": 0.013448834419250488,
      "learning_rate": 0.298529990653315,
      "loss": 0.2566,
      "num_input_tokens_seen": 3455392,
      "step": 1785
    },
    {
      "epoch": 0.2920303450526144,
      "grad_norm": 0.019912201911211014,
      "learning_rate": 0.29852175272336984,
      "loss": 0.227,
      "num_input_tokens_seen": 3464144,
      "step": 1790
    },
    {
      "epoch": 0.2928460722734318,
      "grad_norm": 0.010350093245506287,
      "learning_rate": 0.29851349188950405,
      "loss": 0.2943,
      "num_input_tokens_seen": 3473504,
      "step": 1795
    },
    {
      "epoch": 0.29366179949424914,
      "grad_norm": 0.015879981219768524,
      "learning_rate": 0.2985052081529914,
      "loss": 0.2702,
      "num_input_tokens_seen": 3483504,
      "step": 1800
    },
    {
      "epoch": 0.29366179949424914,
      "eval_loss": 0.25768935680389404,
      "eval_runtime": 68.1181,
      "eval_samples_per_second": 40.004,
      "eval_steps_per_second": 20.009,
      "num_input_tokens_seen": 3483504,
      "step": 1800
    },
    {
      "epoch": 0.2944775267150665,
      "grad_norm": 0.013112787157297134,
      "learning_rate": 0.29849690151510944,
      "loss": 0.2496,
      "num_input_tokens_seen": 3493856,
      "step": 1805
    },
    {
      "epoch": 0.29529325393588385,
      "grad_norm": 0.010621834546327591,
      "learning_rate": 0.2984885719771392,
      "loss": 0.2523,
      "num_input_tokens_seen": 3502848,
      "step": 1810
    },
    {
      "epoch": 0.2961089811567012,
      "grad_norm": 0.014842388220131397,
      "learning_rate": 0.2984802195403651,
      "loss": 0.2373,
      "num_input_tokens_seen": 3513424,
      "step": 1815
    },
    {
      "epoch": 0.29692470837751855,
      "grad_norm": 0.019436072558164597,
      "learning_rate": 0.2984718442060752,
      "loss": 0.2486,
      "num_input_tokens_seen": 3520816,
      "step": 1820
    },
    {
      "epoch": 0.2977404355983359,
      "grad_norm": 0.020666010677814484,
      "learning_rate": 0.2984634459755611,
      "loss": 0.236,
      "num_input_tokens_seen": 3531376,
      "step": 1825
    },
    {
      "epoch": 0.29855616281915326,
      "grad_norm": 0.011250318959355354,
      "learning_rate": 0.29845502485011793,
      "loss": 0.2239,
      "num_input_tokens_seen": 3542144,
      "step": 1830
    },
    {
      "epoch": 0.2993718900399706,
      "grad_norm": 0.01494759414345026,
      "learning_rate": 0.2984465808310444,
      "loss": 0.2085,
      "num_input_tokens_seen": 3551120,
      "step": 1835
    },
    {
      "epoch": 0.300187617260788,
      "grad_norm": 0.012555479072034359,
      "learning_rate": 0.29843811391964253,
      "loss": 0.2333,
      "num_input_tokens_seen": 3560800,
      "step": 1840
    },
    {
      "epoch": 0.3010033444816054,
      "grad_norm": 0.00963589083403349,
      "learning_rate": 0.2984296241172182,
      "loss": 0.2229,
      "num_input_tokens_seen": 3569280,
      "step": 1845
    },
    {
      "epoch": 0.3018190717024227,
      "grad_norm": 0.013810699805617332,
      "learning_rate": 0.29842111142508043,
      "loss": 0.2756,
      "num_input_tokens_seen": 3578704,
      "step": 1850
    },
    {
      "epoch": 0.3026347989232401,
      "grad_norm": 0.02547137811779976,
      "learning_rate": 0.29841257584454217,
      "loss": 0.2561,
      "num_input_tokens_seen": 3587760,
      "step": 1855
    },
    {
      "epoch": 0.30345052614405743,
      "grad_norm": 0.025683585554361343,
      "learning_rate": 0.29840401737691963,
      "loss": 0.2585,
      "num_input_tokens_seen": 3596496,
      "step": 1860
    },
    {
      "epoch": 0.3042662533648748,
      "grad_norm": 0.016908416524529457,
      "learning_rate": 0.29839543602353263,
      "loss": 0.2666,
      "num_input_tokens_seen": 3605744,
      "step": 1865
    },
    {
      "epoch": 0.30508198058569214,
      "grad_norm": 0.011391855776309967,
      "learning_rate": 0.2983868317857046,
      "loss": 0.2413,
      "num_input_tokens_seen": 3615648,
      "step": 1870
    },
    {
      "epoch": 0.3058977078065095,
      "grad_norm": 0.011192231439054012,
      "learning_rate": 0.2983782046647623,
      "loss": 0.2186,
      "num_input_tokens_seen": 3625456,
      "step": 1875
    },
    {
      "epoch": 0.30671343502732684,
      "grad_norm": 0.01300743967294693,
      "learning_rate": 0.2983695546620362,
      "loss": 0.2359,
      "num_input_tokens_seen": 3635424,
      "step": 1880
    },
    {
      "epoch": 0.3075291622481442,
      "grad_norm": 0.011352844536304474,
      "learning_rate": 0.2983608817788603,
      "loss": 0.2711,
      "num_input_tokens_seen": 3644816,
      "step": 1885
    },
    {
      "epoch": 0.3083448894689616,
      "grad_norm": 0.01399542111903429,
      "learning_rate": 0.29835218601657193,
      "loss": 0.2429,
      "num_input_tokens_seen": 3654592,
      "step": 1890
    },
    {
      "epoch": 0.30916061668977896,
      "grad_norm": 0.01364215463399887,
      "learning_rate": 0.29834346737651224,
      "loss": 0.2355,
      "num_input_tokens_seen": 3663568,
      "step": 1895
    },
    {
      "epoch": 0.3099763439105963,
      "grad_norm": 0.013546525500714779,
      "learning_rate": 0.29833472586002563,
      "loss": 0.2296,
      "num_input_tokens_seen": 3673520,
      "step": 1900
    },
    {
      "epoch": 0.31079207113141366,
      "grad_norm": 0.011985162273049355,
      "learning_rate": 0.29832596146846024,
      "loss": 0.239,
      "num_input_tokens_seen": 3684416,
      "step": 1905
    },
    {
      "epoch": 0.311607798352231,
      "grad_norm": 0.013788538984954357,
      "learning_rate": 0.2983171742031676,
      "loss": 0.2357,
      "num_input_tokens_seen": 3694736,
      "step": 1910
    },
    {
      "epoch": 0.31242352557304837,
      "grad_norm": 0.011426234617829323,
      "learning_rate": 0.2983083640655028,
      "loss": 0.2374,
      "num_input_tokens_seen": 3704400,
      "step": 1915
    },
    {
      "epoch": 0.3132392527938657,
      "grad_norm": 0.019780969247221947,
      "learning_rate": 0.29829953105682455,
      "loss": 0.2294,
      "num_input_tokens_seen": 3714160,
      "step": 1920
    },
    {
      "epoch": 0.31405498001468307,
      "grad_norm": 0.015699950978159904,
      "learning_rate": 0.29829067517849495,
      "loss": 0.2717,
      "num_input_tokens_seen": 3724864,
      "step": 1925
    },
    {
      "epoch": 0.3148707072355004,
      "grad_norm": 0.01651962660253048,
      "learning_rate": 0.2982817964318797,
      "loss": 0.2305,
      "num_input_tokens_seen": 3735088,
      "step": 1930
    },
    {
      "epoch": 0.31568643445631783,
      "grad_norm": 0.013423462398350239,
      "learning_rate": 0.298272894818348,
      "loss": 0.2452,
      "num_input_tokens_seen": 3745360,
      "step": 1935
    },
    {
      "epoch": 0.3165021616771352,
      "grad_norm": 0.024575838819146156,
      "learning_rate": 0.2982639703392726,
      "loss": 0.2615,
      "num_input_tokens_seen": 3755968,
      "step": 1940
    },
    {
      "epoch": 0.31731788889795254,
      "grad_norm": 0.01579880528151989,
      "learning_rate": 0.29825502299602974,
      "loss": 0.2384,
      "num_input_tokens_seen": 3764672,
      "step": 1945
    },
    {
      "epoch": 0.3181336161187699,
      "grad_norm": 0.023108242079615593,
      "learning_rate": 0.2982460527899993,
      "loss": 0.2374,
      "num_input_tokens_seen": 3774208,
      "step": 1950
    },
    {
      "epoch": 0.31894934333958724,
      "grad_norm": 0.03025459311902523,
      "learning_rate": 0.29823705972256453,
      "loss": 0.2703,
      "num_input_tokens_seen": 3784528,
      "step": 1955
    },
    {
      "epoch": 0.3197650705604046,
      "grad_norm": 0.015015807934105396,
      "learning_rate": 0.2982280437951123,
      "loss": 0.2677,
      "num_input_tokens_seen": 3793696,
      "step": 1960
    },
    {
      "epoch": 0.32058079778122195,
      "grad_norm": 0.018901316449046135,
      "learning_rate": 0.298219005009033,
      "loss": 0.2516,
      "num_input_tokens_seen": 3802848,
      "step": 1965
    },
    {
      "epoch": 0.3213965250020393,
      "grad_norm": 0.017708608880639076,
      "learning_rate": 0.29820994336572043,
      "loss": 0.2365,
      "num_input_tokens_seen": 3812592,
      "step": 1970
    },
    {
      "epoch": 0.32221225222285665,
      "grad_norm": 0.018140962347388268,
      "learning_rate": 0.2982008588665721,
      "loss": 0.2056,
      "num_input_tokens_seen": 3822224,
      "step": 1975
    },
    {
      "epoch": 0.32302797944367406,
      "grad_norm": 0.01716710440814495,
      "learning_rate": 0.2981917515129889,
      "loss": 0.2341,
      "num_input_tokens_seen": 3831824,
      "step": 1980
    },
    {
      "epoch": 0.3238437066644914,
      "grad_norm": 0.011427369900047779,
      "learning_rate": 0.2981826213063753,
      "loss": 0.2063,
      "num_input_tokens_seen": 3841392,
      "step": 1985
    },
    {
      "epoch": 0.32465943388530877,
      "grad_norm": 0.013827655464410782,
      "learning_rate": 0.2981734682481394,
      "loss": 0.2447,
      "num_input_tokens_seen": 3851200,
      "step": 1990
    },
    {
      "epoch": 0.3254751611061261,
      "grad_norm": 0.009746945463120937,
      "learning_rate": 0.29816429233969255,
      "loss": 0.2265,
      "num_input_tokens_seen": 3861664,
      "step": 1995
    },
    {
      "epoch": 0.3262908883269435,
      "grad_norm": 0.013573962263762951,
      "learning_rate": 0.2981550935824499,
      "loss": 0.2579,
      "num_input_tokens_seen": 3872976,
      "step": 2000
    },
    {
      "epoch": 0.3262908883269435,
      "eval_loss": 0.25515317916870117,
      "eval_runtime": 68.0567,
      "eval_samples_per_second": 40.04,
      "eval_steps_per_second": 20.027,
      "num_input_tokens_seen": 3872976,
      "step": 2000
    },
    {
      "epoch": 0.3271066155477608,
      "grad_norm": 0.013202173635363579,
      "learning_rate": 0.29814587197783,
      "loss": 0.2312,
      "num_input_tokens_seen": 3882896,
      "step": 2005
    },
    {
      "epoch": 0.3279223427685782,
      "grad_norm": 0.011676806025207043,
      "learning_rate": 0.29813662752725495,
      "loss": 0.2486,
      "num_input_tokens_seen": 3892848,
      "step": 2010
    },
    {
      "epoch": 0.32873806998939553,
      "grad_norm": 0.023160843178629875,
      "learning_rate": 0.29812736023215025,
      "loss": 0.2479,
      "num_input_tokens_seen": 3903072,
      "step": 2015
    },
    {
      "epoch": 0.3295537972102129,
      "grad_norm": 0.011334427632391453,
      "learning_rate": 0.29811807009394514,
      "loss": 0.2351,
      "num_input_tokens_seen": 3913776,
      "step": 2020
    },
    {
      "epoch": 0.33036952443103024,
      "grad_norm": 0.012120232917368412,
      "learning_rate": 0.2981087571140723,
      "loss": 0.2295,
      "num_input_tokens_seen": 3923568,
      "step": 2025
    },
    {
      "epoch": 0.33118525165184765,
      "grad_norm": 0.010203253477811813,
      "learning_rate": 0.2980994212939678,
      "loss": 0.2388,
      "num_input_tokens_seen": 3933280,
      "step": 2030
    },
    {
      "epoch": 0.332000978872665,
      "grad_norm": 0.007570045534521341,
      "learning_rate": 0.2980900626350715,
      "loss": 0.2244,
      "num_input_tokens_seen": 3942032,
      "step": 2035
    },
    {
      "epoch": 0.33281670609348235,
      "grad_norm": 0.009604346938431263,
      "learning_rate": 0.29808068113882646,
      "loss": 0.2158,
      "num_input_tokens_seen": 3951728,
      "step": 2040
    },
    {
      "epoch": 0.3336324333142997,
      "grad_norm": 0.009557174518704414,
      "learning_rate": 0.2980712768066795,
      "loss": 0.2253,
      "num_input_tokens_seen": 3961536,
      "step": 2045
    },
    {
      "epoch": 0.33444816053511706,
      "grad_norm": 0.014389464631676674,
      "learning_rate": 0.2980618496400809,
      "loss": 0.2398,
      "num_input_tokens_seen": 3970800,
      "step": 2050
    },
    {
      "epoch": 0.3352638877559344,
      "grad_norm": 0.015884779393672943,
      "learning_rate": 0.2980523996404844,
      "loss": 0.2465,
      "num_input_tokens_seen": 3980640,
      "step": 2055
    },
    {
      "epoch": 0.33607961497675176,
      "grad_norm": 0.01736057549715042,
      "learning_rate": 0.2980429268093473,
      "loss": 0.2235,
      "num_input_tokens_seen": 3989552,
      "step": 2060
    },
    {
      "epoch": 0.3368953421975691,
      "grad_norm": 0.016489701345562935,
      "learning_rate": 0.29803343114813047,
      "loss": 0.2597,
      "num_input_tokens_seen": 3999152,
      "step": 2065
    },
    {
      "epoch": 0.33771106941838647,
      "grad_norm": 0.012411413714289665,
      "learning_rate": 0.2980239126582983,
      "loss": 0.2477,
      "num_input_tokens_seen": 4009472,
      "step": 2070
    },
    {
      "epoch": 0.3385267966392039,
      "grad_norm": 0.013135784305632114,
      "learning_rate": 0.2980143713413186,
      "loss": 0.2248,
      "num_input_tokens_seen": 4019552,
      "step": 2075
    },
    {
      "epoch": 0.33934252386002123,
      "grad_norm": 0.01795942708849907,
      "learning_rate": 0.29800480719866274,
      "loss": 0.2339,
      "num_input_tokens_seen": 4027968,
      "step": 2080
    },
    {
      "epoch": 0.3401582510808386,
      "grad_norm": 0.017623217776417732,
      "learning_rate": 0.2979952202318057,
      "loss": 0.2444,
      "num_input_tokens_seen": 4037184,
      "step": 2085
    },
    {
      "epoch": 0.34097397830165593,
      "grad_norm": 0.011587310582399368,
      "learning_rate": 0.2979856104422259,
      "loss": 0.2543,
      "num_input_tokens_seen": 4047760,
      "step": 2090
    },
    {
      "epoch": 0.3417897055224733,
      "grad_norm": 0.007085779681801796,
      "learning_rate": 0.2979759778314052,
      "loss": 0.196,
      "num_input_tokens_seen": 4057152,
      "step": 2095
    },
    {
      "epoch": 0.34260543274329064,
      "grad_norm": 0.008528485894203186,
      "learning_rate": 0.2979663224008292,
      "loss": 0.2776,
      "num_input_tokens_seen": 4066320,
      "step": 2100
    },
    {
      "epoch": 0.343421159964108,
      "grad_norm": 0.011421105824410915,
      "learning_rate": 0.2979566441519868,
      "loss": 0.236,
      "num_input_tokens_seen": 4076464,
      "step": 2105
    },
    {
      "epoch": 0.34423688718492534,
      "grad_norm": 0.013599573634564877,
      "learning_rate": 0.29794694308637054,
      "loss": 0.2852,
      "num_input_tokens_seen": 4085456,
      "step": 2110
    },
    {
      "epoch": 0.3450526144057427,
      "grad_norm": 0.01310516707599163,
      "learning_rate": 0.2979372192054764,
      "loss": 0.2407,
      "num_input_tokens_seen": 4095104,
      "step": 2115
    },
    {
      "epoch": 0.3458683416265601,
      "grad_norm": 0.011160379275679588,
      "learning_rate": 0.297927472510804,
      "loss": 0.238,
      "num_input_tokens_seen": 4105312,
      "step": 2120
    },
    {
      "epoch": 0.34668406884737746,
      "grad_norm": 0.009764881804585457,
      "learning_rate": 0.29791770300385634,
      "loss": 0.2407,
      "num_input_tokens_seen": 4115248,
      "step": 2125
    },
    {
      "epoch": 0.3474997960681948,
      "grad_norm": 0.012135087512433529,
      "learning_rate": 0.29790791068614003,
      "loss": 0.221,
      "num_input_tokens_seen": 4124560,
      "step": 2130
    },
    {
      "epoch": 0.34831552328901216,
      "grad_norm": 0.015238882973790169,
      "learning_rate": 0.2978980955591652,
      "loss": 0.2981,
      "num_input_tokens_seen": 4133776,
      "step": 2135
    },
    {
      "epoch": 0.3491312505098295,
      "grad_norm": 0.011016452684998512,
      "learning_rate": 0.2978882576244454,
      "loss": 0.2218,
      "num_input_tokens_seen": 4142480,
      "step": 2140
    },
    {
      "epoch": 0.34994697773064687,
      "grad_norm": 0.015926729887723923,
      "learning_rate": 0.2978783968834978,
      "loss": 0.2558,
      "num_input_tokens_seen": 4151184,
      "step": 2145
    },
    {
      "epoch": 0.3507627049514642,
      "grad_norm": 0.011296140030026436,
      "learning_rate": 0.29786851333784303,
      "loss": 0.2338,
      "num_input_tokens_seen": 4160656,
      "step": 2150
    },
    {
      "epoch": 0.3515784321722816,
      "grad_norm": 0.014261254109442234,
      "learning_rate": 0.2978586069890053,
      "loss": 0.2323,
      "num_input_tokens_seen": 4171104,
      "step": 2155
    },
    {
      "epoch": 0.3523941593930989,
      "grad_norm": 0.018993090838193893,
      "learning_rate": 0.29784867783851227,
      "loss": 0.3403,
      "num_input_tokens_seen": 4180128,
      "step": 2160
    },
    {
      "epoch": 0.3532098866139163,
      "grad_norm": 0.01096315961331129,
      "learning_rate": 0.2978387258878951,
      "loss": 0.2207,
      "num_input_tokens_seen": 4188944,
      "step": 2165
    },
    {
      "epoch": 0.3540256138347337,
      "grad_norm": 0.012731043621897697,
      "learning_rate": 0.29782875113868856,
      "loss": 0.2623,
      "num_input_tokens_seen": 4199472,
      "step": 2170
    },
    {
      "epoch": 0.35484134105555104,
      "grad_norm": 0.021630607545375824,
      "learning_rate": 0.2978187535924309,
      "loss": 0.2569,
      "num_input_tokens_seen": 4208624,
      "step": 2175
    },
    {
      "epoch": 0.3556570682763684,
      "grad_norm": 0.011393987573683262,
      "learning_rate": 0.29780873325066376,
      "loss": 0.2279,
      "num_input_tokens_seen": 4217488,
      "step": 2180
    },
    {
      "epoch": 0.35647279549718575,
      "grad_norm": 0.010935657657682896,
      "learning_rate": 0.2977986901149325,
      "loss": 0.2353,
      "num_input_tokens_seen": 4226336,
      "step": 2185
    },
    {
      "epoch": 0.3572885227180031,
      "grad_norm": 0.011668156832456589,
      "learning_rate": 0.29778862418678587,
      "loss": 0.2267,
      "num_input_tokens_seen": 4235168,
      "step": 2190
    },
    {
      "epoch": 0.35810424993882045,
      "grad_norm": 0.018381550908088684,
      "learning_rate": 0.29777853546777616,
      "loss": 0.2392,
      "num_input_tokens_seen": 4244912,
      "step": 2195
    },
    {
      "epoch": 0.3589199771596378,
      "grad_norm": 0.0130192581564188,
      "learning_rate": 0.2977684239594592,
      "loss": 0.2397,
      "num_input_tokens_seen": 4254272,
      "step": 2200
    },
    {
      "epoch": 0.3589199771596378,
      "eval_loss": 0.23425935208797455,
      "eval_runtime": 68.0392,
      "eval_samples_per_second": 40.05,
      "eval_steps_per_second": 20.033,
      "num_input_tokens_seen": 4254272,
      "step": 2200
    },
    {
      "epoch": 0.35973570438045516,
      "grad_norm": 0.009225458838045597,
      "learning_rate": 0.29775828966339424,
      "loss": 0.2242,
      "num_input_tokens_seen": 4263248,
      "step": 2205
    },
    {
      "epoch": 0.3605514316012725,
      "grad_norm": 0.017328301444649696,
      "learning_rate": 0.29774813258114424,
      "loss": 0.243,
      "num_input_tokens_seen": 4273664,
      "step": 2210
    },
    {
      "epoch": 0.3613671588220899,
      "grad_norm": 0.014992055483162403,
      "learning_rate": 0.29773795271427544,
      "loss": 0.2092,
      "num_input_tokens_seen": 4283360,
      "step": 2215
    },
    {
      "epoch": 0.36218288604290727,
      "grad_norm": 0.02292153425514698,
      "learning_rate": 0.2977277500643577,
      "loss": 0.2467,
      "num_input_tokens_seen": 4293648,
      "step": 2220
    },
    {
      "epoch": 0.3629986132637246,
      "grad_norm": 0.016930097714066505,
      "learning_rate": 0.29771752463296447,
      "loss": 0.2674,
      "num_input_tokens_seen": 4303264,
      "step": 2225
    },
    {
      "epoch": 0.363814340484542,
      "grad_norm": 0.015438402071595192,
      "learning_rate": 0.29770727642167266,
      "loss": 0.2446,
      "num_input_tokens_seen": 4313136,
      "step": 2230
    },
    {
      "epoch": 0.36463006770535933,
      "grad_norm": 0.012581266462802887,
      "learning_rate": 0.29769700543206257,
      "loss": 0.2409,
      "num_input_tokens_seen": 4323424,
      "step": 2235
    },
    {
      "epoch": 0.3654457949261767,
      "grad_norm": 0.015139264054596424,
      "learning_rate": 0.2976867116657182,
      "loss": 0.2719,
      "num_input_tokens_seen": 4332848,
      "step": 2240
    },
    {
      "epoch": 0.36626152214699403,
      "grad_norm": 0.013718139380216599,
      "learning_rate": 0.2976763951242269,
      "loss": 0.2071,
      "num_input_tokens_seen": 4343344,
      "step": 2245
    },
    {
      "epoch": 0.3670772493678114,
      "grad_norm": 0.010355963371694088,
      "learning_rate": 0.29766605580917965,
      "loss": 0.229,
      "num_input_tokens_seen": 4352656,
      "step": 2250
    },
    {
      "epoch": 0.36789297658862874,
      "grad_norm": 0.011381865479052067,
      "learning_rate": 0.29765569372217093,
      "loss": 0.2339,
      "num_input_tokens_seen": 4362736,
      "step": 2255
    },
    {
      "epoch": 0.36870870380944615,
      "grad_norm": 0.009939493611454964,
      "learning_rate": 0.2976453088647987,
      "loss": 0.2196,
      "num_input_tokens_seen": 4369776,
      "step": 2260
    },
    {
      "epoch": 0.3695244310302635,
      "grad_norm": 0.009944870136678219,
      "learning_rate": 0.2976349012386644,
      "loss": 0.2634,
      "num_input_tokens_seen": 4379200,
      "step": 2265
    },
    {
      "epoch": 0.37034015825108085,
      "grad_norm": 0.013324731960892677,
      "learning_rate": 0.29762447084537297,
      "loss": 0.253,
      "num_input_tokens_seen": 4388080,
      "step": 2270
    },
    {
      "epoch": 0.3711558854718982,
      "grad_norm": 0.01948154717683792,
      "learning_rate": 0.29761401768653306,
      "loss": 0.2772,
      "num_input_tokens_seen": 4398368,
      "step": 2275
    },
    {
      "epoch": 0.37197161269271556,
      "grad_norm": 0.01119715254753828,
      "learning_rate": 0.29760354176375653,
      "loss": 0.2264,
      "num_input_tokens_seen": 4407936,
      "step": 2280
    },
    {
      "epoch": 0.3727873399135329,
      "grad_norm": 0.009485119953751564,
      "learning_rate": 0.29759304307865897,
      "loss": 0.2238,
      "num_input_tokens_seen": 4418464,
      "step": 2285
    },
    {
      "epoch": 0.37360306713435026,
      "grad_norm": 0.011827081441879272,
      "learning_rate": 0.2975825216328594,
      "loss": 0.2541,
      "num_input_tokens_seen": 4427232,
      "step": 2290
    },
    {
      "epoch": 0.3744187943551676,
      "grad_norm": 0.011764606460928917,
      "learning_rate": 0.2975719774279804,
      "loss": 0.2346,
      "num_input_tokens_seen": 4435920,
      "step": 2295
    },
    {
      "epoch": 0.37523452157598497,
      "grad_norm": 0.01510842889547348,
      "learning_rate": 0.29756141046564794,
      "loss": 0.2887,
      "num_input_tokens_seen": 4445232,
      "step": 2300
    },
    {
      "epoch": 0.3760502487968023,
      "grad_norm": 0.010933509096503258,
      "learning_rate": 0.2975508207474916,
      "loss": 0.2842,
      "num_input_tokens_seen": 4454464,
      "step": 2305
    },
    {
      "epoch": 0.37686597601761973,
      "grad_norm": 0.008649904280900955,
      "learning_rate": 0.2975402082751445,
      "loss": 0.2136,
      "num_input_tokens_seen": 4464432,
      "step": 2310
    },
    {
      "epoch": 0.3776817032384371,
      "grad_norm": 0.009670495986938477,
      "learning_rate": 0.29752957305024313,
      "loss": 0.2621,
      "num_input_tokens_seen": 4474368,
      "step": 2315
    },
    {
      "epoch": 0.37849743045925444,
      "grad_norm": 0.01305792760103941,
      "learning_rate": 0.2975189150744277,
      "loss": 0.236,
      "num_input_tokens_seen": 4483856,
      "step": 2320
    },
    {
      "epoch": 0.3793131576800718,
      "grad_norm": 0.021382711827754974,
      "learning_rate": 0.29750823434934165,
      "loss": 0.233,
      "num_input_tokens_seen": 4493104,
      "step": 2325
    },
    {
      "epoch": 0.38012888490088914,
      "grad_norm": 0.010975059121847153,
      "learning_rate": 0.29749753087663217,
      "loss": 0.2313,
      "num_input_tokens_seen": 4502848,
      "step": 2330
    },
    {
      "epoch": 0.3809446121217065,
      "grad_norm": 0.009311662055552006,
      "learning_rate": 0.29748680465794985,
      "loss": 0.2373,
      "num_input_tokens_seen": 4512384,
      "step": 2335
    },
    {
      "epoch": 0.38176033934252385,
      "grad_norm": 0.009966099634766579,
      "learning_rate": 0.29747605569494884,
      "loss": 0.2299,
      "num_input_tokens_seen": 4523760,
      "step": 2340
    },
    {
      "epoch": 0.3825760665633412,
      "grad_norm": 0.02143176645040512,
      "learning_rate": 0.29746528398928673,
      "loss": 0.2351,
      "num_input_tokens_seen": 4533472,
      "step": 2345
    },
    {
      "epoch": 0.38339179378415855,
      "grad_norm": 0.01832462288439274,
      "learning_rate": 0.2974544895426247,
      "loss": 0.2365,
      "num_input_tokens_seen": 4542896,
      "step": 2350
    },
    {
      "epoch": 0.38420752100497596,
      "grad_norm": 0.007910477928817272,
      "learning_rate": 0.29744367235662733,
      "loss": 0.2155,
      "num_input_tokens_seen": 4552032,
      "step": 2355
    },
    {
      "epoch": 0.3850232482257933,
      "grad_norm": 0.01723003201186657,
      "learning_rate": 0.29743283243296276,
      "loss": 0.255,
      "num_input_tokens_seen": 4560928,
      "step": 2360
    },
    {
      "epoch": 0.38583897544661067,
      "grad_norm": 0.007771460805088282,
      "learning_rate": 0.29742196977330276,
      "loss": 0.2413,
      "num_input_tokens_seen": 4569568,
      "step": 2365
    },
    {
      "epoch": 0.386654702667428,
      "grad_norm": 0.011069709435105324,
      "learning_rate": 0.2974110843793223,
      "loss": 0.2116,
      "num_input_tokens_seen": 4578896,
      "step": 2370
    },
    {
      "epoch": 0.38747042988824537,
      "grad_norm": 0.013832904398441315,
      "learning_rate": 0.2974001762527002,
      "loss": 0.2545,
      "num_input_tokens_seen": 4588624,
      "step": 2375
    },
    {
      "epoch": 0.3882861571090627,
      "grad_norm": 0.010663233697414398,
      "learning_rate": 0.2973892453951186,
      "loss": 0.1957,
      "num_input_tokens_seen": 4598864,
      "step": 2380
    },
    {
      "epoch": 0.3891018843298801,
      "grad_norm": 0.013633666560053825,
      "learning_rate": 0.2973782918082631,
      "loss": 0.2204,
      "num_input_tokens_seen": 4608672,
      "step": 2385
    },
    {
      "epoch": 0.38991761155069743,
      "grad_norm": 0.015416186302900314,
      "learning_rate": 0.29736731549382295,
      "loss": 0.2394,
      "num_input_tokens_seen": 4618288,
      "step": 2390
    },
    {
      "epoch": 0.3907333387715148,
      "grad_norm": 0.008424981497228146,
      "learning_rate": 0.2973563164534908,
      "loss": 0.2507,
      "num_input_tokens_seen": 4627904,
      "step": 2395
    },
    {
      "epoch": 0.3915490659923322,
      "grad_norm": 0.008355711586773396,
      "learning_rate": 0.29734529468896287,
      "loss": 0.2241,
      "num_input_tokens_seen": 4637376,
      "step": 2400
    },
    {
      "epoch": 0.3915490659923322,
      "eval_loss": 0.23106154799461365,
      "eval_runtime": 68.0548,
      "eval_samples_per_second": 40.041,
      "eval_steps_per_second": 20.028,
      "num_input_tokens_seen": 4637376,
      "step": 2400
    },
    {
      "epoch": 0.39236479321314954,
      "grad_norm": 0.015287595801055431,
      "learning_rate": 0.2973342502019388,
      "loss": 0.2424,
      "num_input_tokens_seen": 4646928,
      "step": 2405
    },
    {
      "epoch": 0.3931805204339669,
      "grad_norm": 0.0197308249771595,
      "learning_rate": 0.2973231829941219,
      "loss": 0.2349,
      "num_input_tokens_seen": 4656080,
      "step": 2410
    },
    {
      "epoch": 0.39399624765478425,
      "grad_norm": 0.011419332586228848,
      "learning_rate": 0.2973120930672188,
      "loss": 0.239,
      "num_input_tokens_seen": 4666240,
      "step": 2415
    },
    {
      "epoch": 0.3948119748756016,
      "grad_norm": 0.011904167011380196,
      "learning_rate": 0.2973009804229397,
      "loss": 0.2553,
      "num_input_tokens_seen": 4676704,
      "step": 2420
    },
    {
      "epoch": 0.39562770209641895,
      "grad_norm": 0.019225237891077995,
      "learning_rate": 0.29728984506299827,
      "loss": 0.2416,
      "num_input_tokens_seen": 4686096,
      "step": 2425
    },
    {
      "epoch": 0.3964434293172363,
      "grad_norm": 0.0273347906768322,
      "learning_rate": 0.2972786869891118,
      "loss": 0.2294,
      "num_input_tokens_seen": 4696048,
      "step": 2430
    },
    {
      "epoch": 0.39725915653805366,
      "grad_norm": 0.012955114245414734,
      "learning_rate": 0.29726750620300096,
      "loss": 0.2365,
      "num_input_tokens_seen": 4705040,
      "step": 2435
    },
    {
      "epoch": 0.398074883758871,
      "grad_norm": 0.012532271444797516,
      "learning_rate": 0.29725630270639003,
      "loss": 0.2715,
      "num_input_tokens_seen": 4715152,
      "step": 2440
    },
    {
      "epoch": 0.39889061097968836,
      "grad_norm": 0.012868068180978298,
      "learning_rate": 0.2972450765010067,
      "loss": 0.2281,
      "num_input_tokens_seen": 4725456,
      "step": 2445
    },
    {
      "epoch": 0.3997063382005058,
      "grad_norm": 0.018950965255498886,
      "learning_rate": 0.29723382758858213,
      "loss": 0.2579,
      "num_input_tokens_seen": 4735552,
      "step": 2450
    },
    {
      "epoch": 0.4005220654213231,
      "grad_norm": 0.008841073140501976,
      "learning_rate": 0.29722255597085107,
      "loss": 0.2737,
      "num_input_tokens_seen": 4745552,
      "step": 2455
    },
    {
      "epoch": 0.4013377926421405,
      "grad_norm": 0.009597498923540115,
      "learning_rate": 0.2972112616495518,
      "loss": 0.2062,
      "num_input_tokens_seen": 4755920,
      "step": 2460
    },
    {
      "epoch": 0.40215351986295783,
      "grad_norm": 0.014006749726831913,
      "learning_rate": 0.297199944626426,
      "loss": 0.2289,
      "num_input_tokens_seen": 4766544,
      "step": 2465
    },
    {
      "epoch": 0.4029692470837752,
      "grad_norm": 0.016527878120541573,
      "learning_rate": 0.2971886049032189,
      "loss": 0.2241,
      "num_input_tokens_seen": 4776032,
      "step": 2470
    },
    {
      "epoch": 0.40378497430459254,
      "grad_norm": 0.01960061304271221,
      "learning_rate": 0.29717724248167926,
      "loss": 0.2512,
      "num_input_tokens_seen": 4786848,
      "step": 2475
    },
    {
      "epoch": 0.4046007015254099,
      "grad_norm": 0.011975860223174095,
      "learning_rate": 0.29716585736355927,
      "loss": 0.2222,
      "num_input_tokens_seen": 4796960,
      "step": 2480
    },
    {
      "epoch": 0.40541642874622724,
      "grad_norm": 0.012613906525075436,
      "learning_rate": 0.2971544495506147,
      "loss": 0.2529,
      "num_input_tokens_seen": 4804992,
      "step": 2485
    },
    {
      "epoch": 0.4062321559670446,
      "grad_norm": 0.01092659030109644,
      "learning_rate": 0.2971430190446048,
      "loss": 0.2378,
      "num_input_tokens_seen": 4813328,
      "step": 2490
    },
    {
      "epoch": 0.407047883187862,
      "grad_norm": 0.011040902696549892,
      "learning_rate": 0.2971315658472921,
      "loss": 0.2333,
      "num_input_tokens_seen": 4823056,
      "step": 2495
    },
    {
      "epoch": 0.40786361040867936,
      "grad_norm": 0.01686447113752365,
      "learning_rate": 0.2971200899604431,
      "loss": 0.2527,
      "num_input_tokens_seen": 4831984,
      "step": 2500
    },
    {
      "epoch": 0.4086793376294967,
      "grad_norm": 0.013517134822905064,
      "learning_rate": 0.29710859138582735,
      "loss": 0.2503,
      "num_input_tokens_seen": 4841920,
      "step": 2505
    },
    {
      "epoch": 0.40949506485031406,
      "grad_norm": 0.018293678760528564,
      "learning_rate": 0.29709707012521813,
      "loss": 0.2459,
      "num_input_tokens_seen": 4851104,
      "step": 2510
    },
    {
      "epoch": 0.4103107920711314,
      "grad_norm": 0.010815074667334557,
      "learning_rate": 0.29708552618039213,
      "loss": 0.2698,
      "num_input_tokens_seen": 4859456,
      "step": 2515
    },
    {
      "epoch": 0.41112651929194877,
      "grad_norm": 0.00891353003680706,
      "learning_rate": 0.2970739595531296,
      "loss": 0.2502,
      "num_input_tokens_seen": 4869040,
      "step": 2520
    },
    {
      "epoch": 0.4119422465127661,
      "grad_norm": 0.014352546073496342,
      "learning_rate": 0.2970623702452143,
      "loss": 0.2418,
      "num_input_tokens_seen": 4878288,
      "step": 2525
    },
    {
      "epoch": 0.41275797373358347,
      "grad_norm": 0.014359598979353905,
      "learning_rate": 0.2970507582584334,
      "loss": 0.2308,
      "num_input_tokens_seen": 4888368,
      "step": 2530
    },
    {
      "epoch": 0.4135737009544008,
      "grad_norm": 0.00997858215123415,
      "learning_rate": 0.2970391235945776,
      "loss": 0.2363,
      "num_input_tokens_seen": 4895840,
      "step": 2535
    },
    {
      "epoch": 0.41438942817521823,
      "grad_norm": 0.008834713138639927,
      "learning_rate": 0.2970274662554412,
      "loss": 0.2365,
      "num_input_tokens_seen": 4904768,
      "step": 2540
    },
    {
      "epoch": 0.4152051553960356,
      "grad_norm": 0.010290728881955147,
      "learning_rate": 0.2970157862428218,
      "loss": 0.247,
      "num_input_tokens_seen": 4914272,
      "step": 2545
    },
    {
      "epoch": 0.41602088261685294,
      "grad_norm": 0.010890426114201546,
      "learning_rate": 0.2970040835585206,
      "loss": 0.2143,
      "num_input_tokens_seen": 4924048,
      "step": 2550
    },
    {
      "epoch": 0.4168366098376703,
      "grad_norm": 0.00936148688197136,
      "learning_rate": 0.2969923582043424,
      "loss": 0.2728,
      "num_input_tokens_seen": 4933552,
      "step": 2555
    },
    {
      "epoch": 0.41765233705848764,
      "grad_norm": 0.013436699286103249,
      "learning_rate": 0.2969806101820953,
      "loss": 0.2362,
      "num_input_tokens_seen": 4943408,
      "step": 2560
    },
    {
      "epoch": 0.418468064279305,
      "grad_norm": 0.012658478692173958,
      "learning_rate": 0.2969688394935911,
      "loss": 0.2775,
      "num_input_tokens_seen": 4954176,
      "step": 2565
    },
    {
      "epoch": 0.41928379150012235,
      "grad_norm": 0.006904165726155043,
      "learning_rate": 0.2969570461406449,
      "loss": 0.2418,
      "num_input_tokens_seen": 4962800,
      "step": 2570
    },
    {
      "epoch": 0.4200995187209397,
      "grad_norm": 0.012650647200644016,
      "learning_rate": 0.29694523012507534,
      "loss": 0.285,
      "num_input_tokens_seen": 4973200,
      "step": 2575
    },
    {
      "epoch": 0.42091524594175705,
      "grad_norm": 0.013986102305352688,
      "learning_rate": 0.2969333914487048,
      "loss": 0.2705,
      "num_input_tokens_seen": 4982160,
      "step": 2580
    },
    {
      "epoch": 0.4217309731625744,
      "grad_norm": 0.008199299685657024,
      "learning_rate": 0.2969215301133587,
      "loss": 0.2156,
      "num_input_tokens_seen": 4991168,
      "step": 2585
    },
    {
      "epoch": 0.4225467003833918,
      "grad_norm": 0.008174336515367031,
      "learning_rate": 0.29690964612086634,
      "loss": 0.2338,
      "num_input_tokens_seen": 4999760,
      "step": 2590
    },
    {
      "epoch": 0.42336242760420917,
      "grad_norm": 0.010982838459312916,
      "learning_rate": 0.2968977394730604,
      "loss": 0.2448,
      "num_input_tokens_seen": 5009984,
      "step": 2595
    },
    {
      "epoch": 0.4241781548250265,
      "grad_norm": 0.00686729745939374,
      "learning_rate": 0.296885810171777,
      "loss": 0.224,
      "num_input_tokens_seen": 5019664,
      "step": 2600
    },
    {
      "epoch": 0.4241781548250265,
      "eval_loss": 0.2280333936214447,
      "eval_runtime": 68.0791,
      "eval_samples_per_second": 40.027,
      "eval_steps_per_second": 20.021,
      "num_input_tokens_seen": 5019664,
      "step": 2600
    },
    {
      "epoch": 0.4249938820458439,
      "grad_norm": 0.007925008423626423,
      "learning_rate": 0.2968738582188558,
      "loss": 0.2551,
      "num_input_tokens_seen": 5029696,
      "step": 2605
    },
    {
      "epoch": 0.4258096092666612,
      "grad_norm": 0.011550585739314556,
      "learning_rate": 0.2968618836161399,
      "loss": 0.2452,
      "num_input_tokens_seen": 5038816,
      "step": 2610
    },
    {
      "epoch": 0.4266253364874786,
      "grad_norm": 0.009591193869709969,
      "learning_rate": 0.296849886365476,
      "loss": 0.2457,
      "num_input_tokens_seen": 5047792,
      "step": 2615
    },
    {
      "epoch": 0.42744106370829593,
      "grad_norm": 0.012185521423816681,
      "learning_rate": 0.2968378664687142,
      "loss": 0.2504,
      "num_input_tokens_seen": 5056960,
      "step": 2620
    },
    {
      "epoch": 0.4282567909291133,
      "grad_norm": 0.009001745842397213,
      "learning_rate": 0.296825823927708,
      "loss": 0.2384,
      "num_input_tokens_seen": 5066912,
      "step": 2625
    },
    {
      "epoch": 0.42907251814993064,
      "grad_norm": 0.008640185929834843,
      "learning_rate": 0.29681375874431476,
      "loss": 0.2147,
      "num_input_tokens_seen": 5076800,
      "step": 2630
    },
    {
      "epoch": 0.42988824537074805,
      "grad_norm": 0.013570677489042282,
      "learning_rate": 0.29680167092039483,
      "loss": 0.2118,
      "num_input_tokens_seen": 5085456,
      "step": 2635
    },
    {
      "epoch": 0.4307039725915654,
      "grad_norm": 0.006604646798223257,
      "learning_rate": 0.2967895604578125,
      "loss": 0.2877,
      "num_input_tokens_seen": 5095712,
      "step": 2640
    },
    {
      "epoch": 0.43151969981238275,
      "grad_norm": 0.011366731487214565,
      "learning_rate": 0.2967774273584352,
      "loss": 0.1832,
      "num_input_tokens_seen": 5106272,
      "step": 2645
    },
    {
      "epoch": 0.4323354270332001,
      "grad_norm": 0.010310967452824116,
      "learning_rate": 0.2967652716241342,
      "loss": 0.2534,
      "num_input_tokens_seen": 5115936,
      "step": 2650
    },
    {
      "epoch": 0.43315115425401746,
      "grad_norm": 0.00929252989590168,
      "learning_rate": 0.29675309325678384,
      "loss": 0.223,
      "num_input_tokens_seen": 5125248,
      "step": 2655
    },
    {
      "epoch": 0.4339668814748348,
      "grad_norm": 0.015900464728474617,
      "learning_rate": 0.29674089225826233,
      "loss": 0.2366,
      "num_input_tokens_seen": 5134832,
      "step": 2660
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 0.011237477883696556,
      "learning_rate": 0.29672866863045116,
      "loss": 0.199,
      "num_input_tokens_seen": 5144400,
      "step": 2665
    },
    {
      "epoch": 0.4355983359164695,
      "grad_norm": 0.012095736339688301,
      "learning_rate": 0.2967164223752354,
      "loss": 0.1983,
      "num_input_tokens_seen": 5154448,
      "step": 2670
    },
    {
      "epoch": 0.43641406313728687,
      "grad_norm": 0.014206239022314548,
      "learning_rate": 0.2967041534945035,
      "loss": 0.2951,
      "num_input_tokens_seen": 5163760,
      "step": 2675
    },
    {
      "epoch": 0.4372297903581043,
      "grad_norm": 0.010742598213255405,
      "learning_rate": 0.2966918619901476,
      "loss": 0.2323,
      "num_input_tokens_seen": 5173168,
      "step": 2680
    },
    {
      "epoch": 0.43804551757892163,
      "grad_norm": 0.010870699770748615,
      "learning_rate": 0.2966795478640631,
      "loss": 0.2645,
      "num_input_tokens_seen": 5183648,
      "step": 2685
    },
    {
      "epoch": 0.438861244799739,
      "grad_norm": 0.023850947618484497,
      "learning_rate": 0.29666721111814903,
      "loss": 0.3106,
      "num_input_tokens_seen": 5192800,
      "step": 2690
    },
    {
      "epoch": 0.43967697202055633,
      "grad_norm": 0.011383602395653725,
      "learning_rate": 0.2966548517543079,
      "loss": 0.2581,
      "num_input_tokens_seen": 5201536,
      "step": 2695
    },
    {
      "epoch": 0.4404926992413737,
      "grad_norm": 0.00979206245392561,
      "learning_rate": 0.29664246977444564,
      "loss": 0.2106,
      "num_input_tokens_seen": 5209584,
      "step": 2700
    },
    {
      "epoch": 0.44130842646219104,
      "grad_norm": 0.012789862230420113,
      "learning_rate": 0.2966300651804717,
      "loss": 0.2123,
      "num_input_tokens_seen": 5219792,
      "step": 2705
    },
    {
      "epoch": 0.4421241536830084,
      "grad_norm": 0.007058188784867525,
      "learning_rate": 0.296617637974299,
      "loss": 0.2504,
      "num_input_tokens_seen": 5228448,
      "step": 2710
    },
    {
      "epoch": 0.44293988090382574,
      "grad_norm": 0.01037928182631731,
      "learning_rate": 0.2966051881578441,
      "loss": 0.2523,
      "num_input_tokens_seen": 5237888,
      "step": 2715
    },
    {
      "epoch": 0.4437556081246431,
      "grad_norm": 0.020450042560696602,
      "learning_rate": 0.29659271573302676,
      "loss": 0.2334,
      "num_input_tokens_seen": 5248512,
      "step": 2720
    },
    {
      "epoch": 0.4445713353454605,
      "grad_norm": 0.008041873574256897,
      "learning_rate": 0.2965802207017705,
      "loss": 0.2355,
      "num_input_tokens_seen": 5258448,
      "step": 2725
    },
    {
      "epoch": 0.44538706256627786,
      "grad_norm": 0.007244759239256382,
      "learning_rate": 0.2965677030660021,
      "loss": 0.2348,
      "num_input_tokens_seen": 5267312,
      "step": 2730
    },
    {
      "epoch": 0.4462027897870952,
      "grad_norm": 0.010566424578428268,
      "learning_rate": 0.2965551628276521,
      "loss": 0.2399,
      "num_input_tokens_seen": 5277616,
      "step": 2735
    },
    {
      "epoch": 0.44701851700791256,
      "grad_norm": 0.010681998915970325,
      "learning_rate": 0.29654259998865423,
      "loss": 0.2551,
      "num_input_tokens_seen": 5287280,
      "step": 2740
    },
    {
      "epoch": 0.4478342442287299,
      "grad_norm": 0.011779661290347576,
      "learning_rate": 0.2965300145509458,
      "loss": 0.1972,
      "num_input_tokens_seen": 5295936,
      "step": 2745
    },
    {
      "epoch": 0.44864997144954727,
      "grad_norm": 0.009836631827056408,
      "learning_rate": 0.2965174065164678,
      "loss": 0.2489,
      "num_input_tokens_seen": 5306896,
      "step": 2750
    },
    {
      "epoch": 0.4494656986703646,
      "grad_norm": 0.007735883351415396,
      "learning_rate": 0.2965047758871644,
      "loss": 0.1678,
      "num_input_tokens_seen": 5316992,
      "step": 2755
    },
    {
      "epoch": 0.450281425891182,
      "grad_norm": 0.0158876720815897,
      "learning_rate": 0.2964921226649835,
      "loss": 0.3423,
      "num_input_tokens_seen": 5327408,
      "step": 2760
    },
    {
      "epoch": 0.4510971531119993,
      "grad_norm": 0.013109300285577774,
      "learning_rate": 0.2964794468518763,
      "loss": 0.237,
      "num_input_tokens_seen": 5337712,
      "step": 2765
    },
    {
      "epoch": 0.4519128803328167,
      "grad_norm": 0.013105546124279499,
      "learning_rate": 0.2964667484497977,
      "loss": 0.2137,
      "num_input_tokens_seen": 5347328,
      "step": 2770
    },
    {
      "epoch": 0.4527286075536341,
      "grad_norm": 0.011021101847290993,
      "learning_rate": 0.29645402746070587,
      "loss": 0.2392,
      "num_input_tokens_seen": 5357728,
      "step": 2775
    },
    {
      "epoch": 0.45354433477445144,
      "grad_norm": 0.011550218798220158,
      "learning_rate": 0.2964412838865625,
      "loss": 0.2382,
      "num_input_tokens_seen": 5367968,
      "step": 2780
    },
    {
      "epoch": 0.4543600619952688,
      "grad_norm": 0.01226454321295023,
      "learning_rate": 0.29642851772933293,
      "loss": 0.2317,
      "num_input_tokens_seen": 5378128,
      "step": 2785
    },
    {
      "epoch": 0.45517578921608615,
      "grad_norm": 0.00881945714354515,
      "learning_rate": 0.29641572899098567,
      "loss": 0.2327,
      "num_input_tokens_seen": 5386736,
      "step": 2790
    },
    {
      "epoch": 0.4559915164369035,
      "grad_norm": 0.010081736370921135,
      "learning_rate": 0.29640291767349314,
      "loss": 0.2196,
      "num_input_tokens_seen": 5397744,
      "step": 2795
    },
    {
      "epoch": 0.45680724365772085,
      "grad_norm": 0.008135959506034851,
      "learning_rate": 0.2963900837788308,
      "loss": 0.2394,
      "num_input_tokens_seen": 5406912,
      "step": 2800
    },
    {
      "epoch": 0.45680724365772085,
      "eval_loss": 0.23068054020404816,
      "eval_runtime": 68.003,
      "eval_samples_per_second": 40.072,
      "eval_steps_per_second": 20.043,
      "num_input_tokens_seen": 5406912,
      "step": 2800
    },
    {
      "epoch": 0.4576229708785382,
      "grad_norm": 0.014694398269057274,
      "learning_rate": 0.2963772273089779,
      "loss": 0.2521,
      "num_input_tokens_seen": 5415840,
      "step": 2805
    },
    {
      "epoch": 0.45843869809935556,
      "grad_norm": 0.013762556947767735,
      "learning_rate": 0.2963643482659171,
      "loss": 0.2351,
      "num_input_tokens_seen": 5424816,
      "step": 2810
    },
    {
      "epoch": 0.4592544253201729,
      "grad_norm": 0.013514005579054356,
      "learning_rate": 0.2963514466516345,
      "loss": 0.2424,
      "num_input_tokens_seen": 5433600,
      "step": 2815
    },
    {
      "epoch": 0.4600701525409903,
      "grad_norm": 0.006936051417142153,
      "learning_rate": 0.2963385224681196,
      "loss": 0.2136,
      "num_input_tokens_seen": 5444016,
      "step": 2820
    },
    {
      "epoch": 0.46088587976180767,
      "grad_norm": 0.011583271436393261,
      "learning_rate": 0.29632557571736556,
      "loss": 0.2405,
      "num_input_tokens_seen": 5453776,
      "step": 2825
    },
    {
      "epoch": 0.461701606982625,
      "grad_norm": 0.00641951197758317,
      "learning_rate": 0.2963126064013689,
      "loss": 0.2208,
      "num_input_tokens_seen": 5463456,
      "step": 2830
    },
    {
      "epoch": 0.4625173342034424,
      "grad_norm": 0.010783452540636063,
      "learning_rate": 0.29629961452212966,
      "loss": 0.2807,
      "num_input_tokens_seen": 5472496,
      "step": 2835
    },
    {
      "epoch": 0.46333306142425973,
      "grad_norm": 0.00916113331913948,
      "learning_rate": 0.2962866000816513,
      "loss": 0.2182,
      "num_input_tokens_seen": 5482560,
      "step": 2840
    },
    {
      "epoch": 0.4641487886450771,
      "grad_norm": 0.013445217162370682,
      "learning_rate": 0.2962735630819409,
      "loss": 0.2129,
      "num_input_tokens_seen": 5492368,
      "step": 2845
    },
    {
      "epoch": 0.46496451586589443,
      "grad_norm": 0.010392584837973118,
      "learning_rate": 0.2962605035250089,
      "loss": 0.2374,
      "num_input_tokens_seen": 5502048,
      "step": 2850
    },
    {
      "epoch": 0.4657802430867118,
      "grad_norm": 0.013095451518893242,
      "learning_rate": 0.29624742141286914,
      "loss": 0.241,
      "num_input_tokens_seen": 5511888,
      "step": 2855
    },
    {
      "epoch": 0.46659597030752914,
      "grad_norm": 0.009413345716893673,
      "learning_rate": 0.29623431674753925,
      "loss": 0.2323,
      "num_input_tokens_seen": 5521680,
      "step": 2860
    },
    {
      "epoch": 0.46741169752834655,
      "grad_norm": 0.007919891737401485,
      "learning_rate": 0.29622118953103993,
      "loss": 0.2292,
      "num_input_tokens_seen": 5530688,
      "step": 2865
    },
    {
      "epoch": 0.4682274247491639,
      "grad_norm": 0.0058891382068395615,
      "learning_rate": 0.2962080397653957,
      "loss": 0.2517,
      "num_input_tokens_seen": 5539872,
      "step": 2870
    },
    {
      "epoch": 0.46904315196998125,
      "grad_norm": 0.008469703607261181,
      "learning_rate": 0.29619486745263435,
      "loss": 0.2393,
      "num_input_tokens_seen": 5549248,
      "step": 2875
    },
    {
      "epoch": 0.4698588791907986,
      "grad_norm": 0.007331725209951401,
      "learning_rate": 0.2961816725947873,
      "loss": 0.2213,
      "num_input_tokens_seen": 5558736,
      "step": 2880
    },
    {
      "epoch": 0.47067460641161596,
      "grad_norm": 0.013445253483951092,
      "learning_rate": 0.29616845519388924,
      "loss": 0.3057,
      "num_input_tokens_seen": 5567936,
      "step": 2885
    },
    {
      "epoch": 0.4714903336324333,
      "grad_norm": 0.009602021425962448,
      "learning_rate": 0.2961552152519785,
      "loss": 0.2239,
      "num_input_tokens_seen": 5576576,
      "step": 2890
    },
    {
      "epoch": 0.47230606085325066,
      "grad_norm": 0.007546696811914444,
      "learning_rate": 0.29614195277109695,
      "loss": 0.2631,
      "num_input_tokens_seen": 5586144,
      "step": 2895
    },
    {
      "epoch": 0.473121788074068,
      "grad_norm": 0.007384778931736946,
      "learning_rate": 0.2961286677532897,
      "loss": 0.2353,
      "num_input_tokens_seen": 5596112,
      "step": 2900
    },
    {
      "epoch": 0.47393751529488537,
      "grad_norm": 0.013223305344581604,
      "learning_rate": 0.2961153602006055,
      "loss": 0.2304,
      "num_input_tokens_seen": 5604912,
      "step": 2905
    },
    {
      "epoch": 0.4747532425157027,
      "grad_norm": 0.013659536838531494,
      "learning_rate": 0.29610203011509656,
      "loss": 0.2344,
      "num_input_tokens_seen": 5614848,
      "step": 2910
    },
    {
      "epoch": 0.47556896973652013,
      "grad_norm": 0.004788030870258808,
      "learning_rate": 0.29608867749881856,
      "loss": 0.176,
      "num_input_tokens_seen": 5624016,
      "step": 2915
    },
    {
      "epoch": 0.4763846969573375,
      "grad_norm": 0.006110379938036203,
      "learning_rate": 0.29607530235383067,
      "loss": 0.3036,
      "num_input_tokens_seen": 5633392,
      "step": 2920
    },
    {
      "epoch": 0.47720042417815484,
      "grad_norm": 0.006674197036772966,
      "learning_rate": 0.2960619046821954,
      "loss": 0.258,
      "num_input_tokens_seen": 5642928,
      "step": 2925
    },
    {
      "epoch": 0.4780161513989722,
      "grad_norm": 0.017559483647346497,
      "learning_rate": 0.2960484844859789,
      "loss": 0.2735,
      "num_input_tokens_seen": 5651776,
      "step": 2930
    },
    {
      "epoch": 0.47883187861978954,
      "grad_norm": 0.009468899108469486,
      "learning_rate": 0.29603504176725076,
      "loss": 0.2684,
      "num_input_tokens_seen": 5661888,
      "step": 2935
    },
    {
      "epoch": 0.4796476058406069,
      "grad_norm": 0.006980549544095993,
      "learning_rate": 0.296021576528084,
      "loss": 0.221,
      "num_input_tokens_seen": 5670592,
      "step": 2940
    },
    {
      "epoch": 0.48046333306142425,
      "grad_norm": 0.00671148207038641,
      "learning_rate": 0.29600808877055507,
      "loss": 0.2425,
      "num_input_tokens_seen": 5678720,
      "step": 2945
    },
    {
      "epoch": 0.4812790602822416,
      "grad_norm": 0.008028727024793625,
      "learning_rate": 0.29599457849674404,
      "loss": 0.2326,
      "num_input_tokens_seen": 5688288,
      "step": 2950
    },
    {
      "epoch": 0.48209478750305895,
      "grad_norm": 0.012512700632214546,
      "learning_rate": 0.2959810457087343,
      "loss": 0.2585,
      "num_input_tokens_seen": 5698160,
      "step": 2955
    },
    {
      "epoch": 0.48291051472387636,
      "grad_norm": 0.009150715544819832,
      "learning_rate": 0.2959674904086128,
      "loss": 0.2233,
      "num_input_tokens_seen": 5707632,
      "step": 2960
    },
    {
      "epoch": 0.4837262419446937,
      "grad_norm": 0.008970307186245918,
      "learning_rate": 0.2959539125984699,
      "loss": 0.2195,
      "num_input_tokens_seen": 5718240,
      "step": 2965
    },
    {
      "epoch": 0.48454196916551107,
      "grad_norm": 0.00791767705231905,
      "learning_rate": 0.2959403122803996,
      "loss": 0.2268,
      "num_input_tokens_seen": 5727552,
      "step": 2970
    },
    {
      "epoch": 0.4853576963863284,
      "grad_norm": 0.015247791074216366,
      "learning_rate": 0.2959266894564991,
      "loss": 0.2479,
      "num_input_tokens_seen": 5737040,
      "step": 2975
    },
    {
      "epoch": 0.48617342360714577,
      "grad_norm": 0.009160155430436134,
      "learning_rate": 0.2959130441288692,
      "loss": 0.2141,
      "num_input_tokens_seen": 5747696,
      "step": 2980
    },
    {
      "epoch": 0.4869891508279631,
      "grad_norm": 0.010271282866597176,
      "learning_rate": 0.2958993762996143,
      "loss": 0.2097,
      "num_input_tokens_seen": 5757488,
      "step": 2985
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 0.010445434600114822,
      "learning_rate": 0.2958856859708421,
      "loss": 0.2268,
      "num_input_tokens_seen": 5766368,
      "step": 2990
    },
    {
      "epoch": 0.48862060526959783,
      "grad_norm": 0.009739860892295837,
      "learning_rate": 0.2958719731446638,
      "loss": 0.2308,
      "num_input_tokens_seen": 5776288,
      "step": 2995
    },
    {
      "epoch": 0.4894363324904152,
      "grad_norm": 0.009094782173633575,
      "learning_rate": 0.29585823782319404,
      "loss": 0.2368,
      "num_input_tokens_seen": 5786080,
      "step": 3000
    },
    {
      "epoch": 0.4894363324904152,
      "eval_loss": 0.2659660875797272,
      "eval_runtime": 68.2216,
      "eval_samples_per_second": 39.943,
      "eval_steps_per_second": 19.979,
      "num_input_tokens_seen": 5786080,
      "step": 3000
    },
    {
      "epoch": 0.4902520597112326,
      "grad_norm": 0.00998267438262701,
      "learning_rate": 0.2958444800085511,
      "loss": 0.2835,
      "num_input_tokens_seen": 5796080,
      "step": 3005
    },
    {
      "epoch": 0.49106778693204994,
      "grad_norm": 0.01331998035311699,
      "learning_rate": 0.2958306997028565,
      "loss": 0.2334,
      "num_input_tokens_seen": 5805696,
      "step": 3010
    },
    {
      "epoch": 0.4918835141528673,
      "grad_norm": 0.011096695438027382,
      "learning_rate": 0.2958168969082354,
      "loss": 0.2143,
      "num_input_tokens_seen": 5814576,
      "step": 3015
    },
    {
      "epoch": 0.49269924137368465,
      "grad_norm": 0.016975779086351395,
      "learning_rate": 0.2958030716268164,
      "loss": 0.2844,
      "num_input_tokens_seen": 5823520,
      "step": 3020
    },
    {
      "epoch": 0.493514968594502,
      "grad_norm": 0.011636693961918354,
      "learning_rate": 0.2957892238607314,
      "loss": 0.2458,
      "num_input_tokens_seen": 5832624,
      "step": 3025
    },
    {
      "epoch": 0.49433069581531935,
      "grad_norm": 0.009273558855056763,
      "learning_rate": 0.2957753536121161,
      "loss": 0.2289,
      "num_input_tokens_seen": 5841920,
      "step": 3030
    },
    {
      "epoch": 0.4951464230361367,
      "grad_norm": 0.00955773051828146,
      "learning_rate": 0.29576146088310923,
      "loss": 0.2423,
      "num_input_tokens_seen": 5851552,
      "step": 3035
    },
    {
      "epoch": 0.49596215025695406,
      "grad_norm": 0.010287829674780369,
      "learning_rate": 0.2957475456758533,
      "loss": 0.2509,
      "num_input_tokens_seen": 5861632,
      "step": 3040
    },
    {
      "epoch": 0.4967778774777714,
      "grad_norm": 0.012078281491994858,
      "learning_rate": 0.2957336079924944,
      "loss": 0.24,
      "num_input_tokens_seen": 5872000,
      "step": 3045
    },
    {
      "epoch": 0.49759360469858877,
      "grad_norm": 0.01038777083158493,
      "learning_rate": 0.2957196478351816,
      "loss": 0.236,
      "num_input_tokens_seen": 5882336,
      "step": 3050
    },
    {
      "epoch": 0.4984093319194062,
      "grad_norm": 0.009358287788927555,
      "learning_rate": 0.295705665206068,
      "loss": 0.2246,
      "num_input_tokens_seen": 5892304,
      "step": 3055
    },
    {
      "epoch": 0.4992250591402235,
      "grad_norm": 0.009502839297056198,
      "learning_rate": 0.2956916601073097,
      "loss": 0.1865,
      "num_input_tokens_seen": 5901216,
      "step": 3060
    },
    {
      "epoch": 0.5000407863610409,
      "grad_norm": 0.016559695824980736,
      "learning_rate": 0.29567763254106655,
      "loss": 0.2176,
      "num_input_tokens_seen": 5910496,
      "step": 3065
    },
    {
      "epoch": 0.5008565135818582,
      "grad_norm": 0.012207725085318089,
      "learning_rate": 0.29566358250950175,
      "loss": 0.2702,
      "num_input_tokens_seen": 5920112,
      "step": 3070
    },
    {
      "epoch": 0.5016722408026756,
      "grad_norm": 0.014961454086005688,
      "learning_rate": 0.295649510014782,
      "loss": 0.2216,
      "num_input_tokens_seen": 5929024,
      "step": 3075
    },
    {
      "epoch": 0.5024879680234929,
      "grad_norm": 0.01214258000254631,
      "learning_rate": 0.2956354150590775,
      "loss": 0.2263,
      "num_input_tokens_seen": 5938448,
      "step": 3080
    },
    {
      "epoch": 0.5033036952443103,
      "grad_norm": 0.012346784584224224,
      "learning_rate": 0.2956212976445618,
      "loss": 0.2374,
      "num_input_tokens_seen": 5947760,
      "step": 3085
    },
    {
      "epoch": 0.5041194224651276,
      "grad_norm": 0.014923950657248497,
      "learning_rate": 0.295607157773412,
      "loss": 0.1971,
      "num_input_tokens_seen": 5956624,
      "step": 3090
    },
    {
      "epoch": 0.504935149685945,
      "grad_norm": 0.02256743423640728,
      "learning_rate": 0.2955929954478087,
      "loss": 0.2584,
      "num_input_tokens_seen": 5966304,
      "step": 3095
    },
    {
      "epoch": 0.5057508769067623,
      "grad_norm": 0.011570170521736145,
      "learning_rate": 0.29557881066993585,
      "loss": 0.1957,
      "num_input_tokens_seen": 5975632,
      "step": 3100
    },
    {
      "epoch": 0.5065666041275797,
      "grad_norm": 0.017179686576128006,
      "learning_rate": 0.29556460344198093,
      "loss": 0.2639,
      "num_input_tokens_seen": 5985840,
      "step": 3105
    },
    {
      "epoch": 0.507382331348397,
      "grad_norm": 0.012892786413431168,
      "learning_rate": 0.29555037376613486,
      "loss": 0.2293,
      "num_input_tokens_seen": 5996224,
      "step": 3110
    },
    {
      "epoch": 0.5081980585692144,
      "grad_norm": 0.008668092079460621,
      "learning_rate": 0.29553612164459203,
      "loss": 0.26,
      "num_input_tokens_seen": 6005808,
      "step": 3115
    },
    {
      "epoch": 0.5090137857900318,
      "grad_norm": 0.0133508937433362,
      "learning_rate": 0.29552184707955037,
      "loss": 0.2359,
      "num_input_tokens_seen": 6015600,
      "step": 3120
    },
    {
      "epoch": 0.5098295130108492,
      "grad_norm": 0.01909661293029785,
      "learning_rate": 0.29550755007321117,
      "loss": 0.229,
      "num_input_tokens_seen": 6025440,
      "step": 3125
    },
    {
      "epoch": 0.5106452402316666,
      "grad_norm": 0.012933265417814255,
      "learning_rate": 0.29549323062777916,
      "loss": 0.2448,
      "num_input_tokens_seen": 6034560,
      "step": 3130
    },
    {
      "epoch": 0.5114609674524839,
      "grad_norm": 0.011535665020346642,
      "learning_rate": 0.29547888874546263,
      "loss": 0.2227,
      "num_input_tokens_seen": 6043312,
      "step": 3135
    },
    {
      "epoch": 0.5122766946733013,
      "grad_norm": 0.009326635859906673,
      "learning_rate": 0.2954645244284732,
      "loss": 0.246,
      "num_input_tokens_seen": 6052912,
      "step": 3140
    },
    {
      "epoch": 0.5130924218941186,
      "grad_norm": 0.011271005496382713,
      "learning_rate": 0.2954501376790261,
      "loss": 0.2312,
      "num_input_tokens_seen": 6062032,
      "step": 3145
    },
    {
      "epoch": 0.513908149114936,
      "grad_norm": 0.026902055367827415,
      "learning_rate": 0.29543572849933997,
      "loss": 0.2441,
      "num_input_tokens_seen": 6070496,
      "step": 3150
    },
    {
      "epoch": 0.5147238763357533,
      "grad_norm": 0.013357853516936302,
      "learning_rate": 0.2954212968916368,
      "loss": 0.2257,
      "num_input_tokens_seen": 6078736,
      "step": 3155
    },
    {
      "epoch": 0.5155396035565707,
      "grad_norm": 0.010646682232618332,
      "learning_rate": 0.29540684285814217,
      "loss": 0.2278,
      "num_input_tokens_seen": 6087472,
      "step": 3160
    },
    {
      "epoch": 0.516355330777388,
      "grad_norm": 0.010720113292336464,
      "learning_rate": 0.2953923664010851,
      "loss": 0.2555,
      "num_input_tokens_seen": 6097760,
      "step": 3165
    },
    {
      "epoch": 0.5171710579982054,
      "grad_norm": 0.012425745837390423,
      "learning_rate": 0.295377867522698,
      "loss": 0.2336,
      "num_input_tokens_seen": 6108160,
      "step": 3170
    },
    {
      "epoch": 0.5179867852190227,
      "grad_norm": 0.020808203145861626,
      "learning_rate": 0.2953633462252168,
      "loss": 0.2369,
      "num_input_tokens_seen": 6117440,
      "step": 3175
    },
    {
      "epoch": 0.5188025124398401,
      "grad_norm": 0.011420226655900478,
      "learning_rate": 0.2953488025108809,
      "loss": 0.2443,
      "num_input_tokens_seen": 6126880,
      "step": 3180
    },
    {
      "epoch": 0.5196182396606575,
      "grad_norm": 0.018434029072523117,
      "learning_rate": 0.295334236381933,
      "loss": 0.272,
      "num_input_tokens_seen": 6137136,
      "step": 3185
    },
    {
      "epoch": 0.5204339668814748,
      "grad_norm": 0.011367757804691792,
      "learning_rate": 0.29531964784061954,
      "loss": 0.2398,
      "num_input_tokens_seen": 6148080,
      "step": 3190
    },
    {
      "epoch": 0.5212496941022922,
      "grad_norm": 0.012197312898933887,
      "learning_rate": 0.2953050368891902,
      "loss": 0.241,
      "num_input_tokens_seen": 6157840,
      "step": 3195
    },
    {
      "epoch": 0.5220654213231095,
      "grad_norm": 0.018197685480117798,
      "learning_rate": 0.29529040352989805,
      "loss": 0.2282,
      "num_input_tokens_seen": 6167600,
      "step": 3200
    },
    {
      "epoch": 0.5220654213231095,
      "eval_loss": 0.2373400181531906,
      "eval_runtime": 68.0234,
      "eval_samples_per_second": 40.06,
      "eval_steps_per_second": 20.037,
      "num_input_tokens_seen": 6167600,
      "step": 3200
    },
    {
      "epoch": 0.5228811485439269,
      "grad_norm": 0.013610998168587685,
      "learning_rate": 0.29527574776499993,
      "loss": 0.2433,
      "num_input_tokens_seen": 6177536,
      "step": 3205
    },
    {
      "epoch": 0.5236968757647442,
      "grad_norm": 0.007256229408085346,
      "learning_rate": 0.2952610695967558,
      "loss": 0.2332,
      "num_input_tokens_seen": 6187008,
      "step": 3210
    },
    {
      "epoch": 0.5245126029855617,
      "grad_norm": 0.012115806341171265,
      "learning_rate": 0.29524636902742935,
      "loss": 0.2424,
      "num_input_tokens_seen": 6196736,
      "step": 3215
    },
    {
      "epoch": 0.525328330206379,
      "grad_norm": 0.013748062774538994,
      "learning_rate": 0.2952316460592875,
      "loss": 0.255,
      "num_input_tokens_seen": 6205408,
      "step": 3220
    },
    {
      "epoch": 0.5261440574271964,
      "grad_norm": 0.013561666011810303,
      "learning_rate": 0.29521690069460066,
      "loss": 0.2654,
      "num_input_tokens_seen": 6215248,
      "step": 3225
    },
    {
      "epoch": 0.5269597846480137,
      "grad_norm": 0.009305239655077457,
      "learning_rate": 0.29520213293564285,
      "loss": 0.2439,
      "num_input_tokens_seen": 6224512,
      "step": 3230
    },
    {
      "epoch": 0.5277755118688311,
      "grad_norm": 0.010782893747091293,
      "learning_rate": 0.29518734278469144,
      "loss": 0.1938,
      "num_input_tokens_seen": 6233536,
      "step": 3235
    },
    {
      "epoch": 0.5285912390896484,
      "grad_norm": 0.012613926082849503,
      "learning_rate": 0.29517253024402723,
      "loss": 0.2389,
      "num_input_tokens_seen": 6243584,
      "step": 3240
    },
    {
      "epoch": 0.5294069663104658,
      "grad_norm": 0.007601585704833269,
      "learning_rate": 0.2951576953159345,
      "loss": 0.1876,
      "num_input_tokens_seen": 6252672,
      "step": 3245
    },
    {
      "epoch": 0.5302226935312832,
      "grad_norm": 0.009332873858511448,
      "learning_rate": 0.29514283800270097,
      "loss": 0.2039,
      "num_input_tokens_seen": 6262320,
      "step": 3250
    },
    {
      "epoch": 0.5310384207521005,
      "grad_norm": 0.010593807324767113,
      "learning_rate": 0.2951279583066179,
      "loss": 0.3015,
      "num_input_tokens_seen": 6272448,
      "step": 3255
    },
    {
      "epoch": 0.5318541479729179,
      "grad_norm": 0.0072121648117899895,
      "learning_rate": 0.2951130562299798,
      "loss": 0.2188,
      "num_input_tokens_seen": 6281744,
      "step": 3260
    },
    {
      "epoch": 0.5326698751937352,
      "grad_norm": 0.00873977318406105,
      "learning_rate": 0.29509813177508487,
      "loss": 0.2452,
      "num_input_tokens_seen": 6291568,
      "step": 3265
    },
    {
      "epoch": 0.5334856024145526,
      "grad_norm": 0.006900184787809849,
      "learning_rate": 0.2950831849442346,
      "loss": 0.2681,
      "num_input_tokens_seen": 6301168,
      "step": 3270
    },
    {
      "epoch": 0.5343013296353699,
      "grad_norm": 0.010016541928052902,
      "learning_rate": 0.2950682157397339,
      "loss": 0.2344,
      "num_input_tokens_seen": 6311296,
      "step": 3275
    },
    {
      "epoch": 0.5351170568561873,
      "grad_norm": 0.011888167820870876,
      "learning_rate": 0.2950532241638914,
      "loss": 0.2613,
      "num_input_tokens_seen": 6322032,
      "step": 3280
    },
    {
      "epoch": 0.5359327840770046,
      "grad_norm": 0.006637689657509327,
      "learning_rate": 0.2950382102190188,
      "loss": 0.2155,
      "num_input_tokens_seen": 6332208,
      "step": 3285
    },
    {
      "epoch": 0.536748511297822,
      "grad_norm": 0.009610595181584358,
      "learning_rate": 0.2950231739074316,
      "loss": 0.234,
      "num_input_tokens_seen": 6341408,
      "step": 3290
    },
    {
      "epoch": 0.5375642385186393,
      "grad_norm": 0.0097493352368474,
      "learning_rate": 0.29500811523144843,
      "loss": 0.231,
      "num_input_tokens_seen": 6350848,
      "step": 3295
    },
    {
      "epoch": 0.5383799657394567,
      "grad_norm": 0.006327477749437094,
      "learning_rate": 0.2949930341933917,
      "loss": 0.2306,
      "num_input_tokens_seen": 6360560,
      "step": 3300
    },
    {
      "epoch": 0.539195692960274,
      "grad_norm": 0.01231988612562418,
      "learning_rate": 0.29497793079558693,
      "loss": 0.2175,
      "num_input_tokens_seen": 6370880,
      "step": 3305
    },
    {
      "epoch": 0.5400114201810915,
      "grad_norm": 0.007257470395416021,
      "learning_rate": 0.2949628050403633,
      "loss": 0.1878,
      "num_input_tokens_seen": 6380832,
      "step": 3310
    },
    {
      "epoch": 0.5408271474019088,
      "grad_norm": 0.005282142199575901,
      "learning_rate": 0.2949476569300535,
      "loss": 0.1385,
      "num_input_tokens_seen": 6390848,
      "step": 3315
    },
    {
      "epoch": 0.5416428746227262,
      "grad_norm": 0.019800640642642975,
      "learning_rate": 0.29493248646699344,
      "loss": 0.302,
      "num_input_tokens_seen": 6400128,
      "step": 3320
    },
    {
      "epoch": 0.5424586018435436,
      "grad_norm": 0.01198714878410101,
      "learning_rate": 0.29491729365352265,
      "loss": 0.2929,
      "num_input_tokens_seen": 6409824,
      "step": 3325
    },
    {
      "epoch": 0.5432743290643609,
      "grad_norm": 0.016261916607618332,
      "learning_rate": 0.29490207849198397,
      "loss": 0.2384,
      "num_input_tokens_seen": 6418976,
      "step": 3330
    },
    {
      "epoch": 0.5440900562851783,
      "grad_norm": 0.011584398336708546,
      "learning_rate": 0.29488684098472384,
      "loss": 0.2499,
      "num_input_tokens_seen": 6428384,
      "step": 3335
    },
    {
      "epoch": 0.5449057835059956,
      "grad_norm": 0.010323978029191494,
      "learning_rate": 0.2948715811340921,
      "loss": 0.2462,
      "num_input_tokens_seen": 6437888,
      "step": 3340
    },
    {
      "epoch": 0.545721510726813,
      "grad_norm": 0.00995455402880907,
      "learning_rate": 0.294856298942442,
      "loss": 0.2364,
      "num_input_tokens_seen": 6447712,
      "step": 3345
    },
    {
      "epoch": 0.5465372379476303,
      "grad_norm": 0.011504355818033218,
      "learning_rate": 0.2948409944121302,
      "loss": 0.2471,
      "num_input_tokens_seen": 6456864,
      "step": 3350
    },
    {
      "epoch": 0.5473529651684477,
      "grad_norm": 0.01429010834544897,
      "learning_rate": 0.29482566754551687,
      "loss": 0.266,
      "num_input_tokens_seen": 6465824,
      "step": 3355
    },
    {
      "epoch": 0.548168692389265,
      "grad_norm": 0.01856783963739872,
      "learning_rate": 0.2948103183449656,
      "loss": 0.2342,
      "num_input_tokens_seen": 6475584,
      "step": 3360
    },
    {
      "epoch": 0.5489844196100824,
      "grad_norm": 0.036432232707738876,
      "learning_rate": 0.2947949468128435,
      "loss": 0.2191,
      "num_input_tokens_seen": 6485264,
      "step": 3365
    },
    {
      "epoch": 0.5498001468308997,
      "grad_norm": 0.012969098053872585,
      "learning_rate": 0.2947795529515209,
      "loss": 0.2821,
      "num_input_tokens_seen": 6494736,
      "step": 3370
    },
    {
      "epoch": 0.5506158740517171,
      "grad_norm": 0.015453458763659,
      "learning_rate": 0.29476413676337193,
      "loss": 0.2268,
      "num_input_tokens_seen": 6504512,
      "step": 3375
    },
    {
      "epoch": 0.5514316012725344,
      "grad_norm": 0.014344064518809319,
      "learning_rate": 0.2947486982507738,
      "loss": 0.2299,
      "num_input_tokens_seen": 6514544,
      "step": 3380
    },
    {
      "epoch": 0.5522473284933518,
      "grad_norm": 0.013938811607658863,
      "learning_rate": 0.29473323741610735,
      "loss": 0.2445,
      "num_input_tokens_seen": 6525504,
      "step": 3385
    },
    {
      "epoch": 0.5530630557141691,
      "grad_norm": 0.01937323622405529,
      "learning_rate": 0.2947177542617569,
      "loss": 0.2371,
      "num_input_tokens_seen": 6535584,
      "step": 3390
    },
    {
      "epoch": 0.5538787829349865,
      "grad_norm": 0.023693712428212166,
      "learning_rate": 0.2947022487901101,
      "loss": 0.2505,
      "num_input_tokens_seen": 6545104,
      "step": 3395
    },
    {
      "epoch": 0.5546945101558038,
      "grad_norm": 0.01571166142821312,
      "learning_rate": 0.2946867210035581,
      "loss": 0.2228,
      "num_input_tokens_seen": 6553904,
      "step": 3400
    },
    {
      "epoch": 0.5546945101558038,
      "eval_loss": 0.23818439245224,
      "eval_runtime": 67.8978,
      "eval_samples_per_second": 40.134,
      "eval_steps_per_second": 20.074,
      "num_input_tokens_seen": 6553904,
      "step": 3400
    },
    {
      "epoch": 0.5555102373766213,
      "grad_norm": 0.012279406189918518,
      "learning_rate": 0.2946711709044954,
      "loss": 0.2701,
      "num_input_tokens_seen": 6563056,
      "step": 3405
    },
    {
      "epoch": 0.5563259645974387,
      "grad_norm": 0.01188167929649353,
      "learning_rate": 0.2946555984953202,
      "loss": 0.2403,
      "num_input_tokens_seen": 6571568,
      "step": 3410
    },
    {
      "epoch": 0.557141691818256,
      "grad_norm": 0.01957404427230358,
      "learning_rate": 0.2946400037784338,
      "loss": 0.2268,
      "num_input_tokens_seen": 6581072,
      "step": 3415
    },
    {
      "epoch": 0.5579574190390734,
      "grad_norm": 0.013173938728868961,
      "learning_rate": 0.29462438675624114,
      "loss": 0.2259,
      "num_input_tokens_seen": 6591136,
      "step": 3420
    },
    {
      "epoch": 0.5587731462598907,
      "grad_norm": 0.014503357000648975,
      "learning_rate": 0.2946087474311506,
      "loss": 0.2643,
      "num_input_tokens_seen": 6600528,
      "step": 3425
    },
    {
      "epoch": 0.5595888734807081,
      "grad_norm": 0.015622377395629883,
      "learning_rate": 0.294593085805574,
      "loss": 0.2555,
      "num_input_tokens_seen": 6609536,
      "step": 3430
    },
    {
      "epoch": 0.5604046007015254,
      "grad_norm": 0.013595963828265667,
      "learning_rate": 0.2945774018819264,
      "loss": 0.2612,
      "num_input_tokens_seen": 6619600,
      "step": 3435
    },
    {
      "epoch": 0.5612203279223428,
      "grad_norm": 0.016775913536548615,
      "learning_rate": 0.2945616956626266,
      "loss": 0.2955,
      "num_input_tokens_seen": 6628928,
      "step": 3440
    },
    {
      "epoch": 0.5620360551431601,
      "grad_norm": 0.009596568532288074,
      "learning_rate": 0.2945459671500966,
      "loss": 0.2493,
      "num_input_tokens_seen": 6637008,
      "step": 3445
    },
    {
      "epoch": 0.5628517823639775,
      "grad_norm": 0.008639994077384472,
      "learning_rate": 0.2945302163467621,
      "loss": 0.2376,
      "num_input_tokens_seen": 6646512,
      "step": 3450
    },
    {
      "epoch": 0.5636675095847948,
      "grad_norm": 0.021462898701429367,
      "learning_rate": 0.2945144432550519,
      "loss": 0.2653,
      "num_input_tokens_seen": 6657168,
      "step": 3455
    },
    {
      "epoch": 0.5644832368056122,
      "grad_norm": 0.032435208559036255,
      "learning_rate": 0.29449864787739843,
      "loss": 0.2413,
      "num_input_tokens_seen": 6666944,
      "step": 3460
    },
    {
      "epoch": 0.5652989640264295,
      "grad_norm": 0.008660942316055298,
      "learning_rate": 0.2944828302162376,
      "loss": 0.2112,
      "num_input_tokens_seen": 6677072,
      "step": 3465
    },
    {
      "epoch": 0.5661146912472469,
      "grad_norm": 0.009382870979607105,
      "learning_rate": 0.2944669902740087,
      "loss": 0.2917,
      "num_input_tokens_seen": 6686528,
      "step": 3470
    },
    {
      "epoch": 0.5669304184680642,
      "grad_norm": 0.009689198806881905,
      "learning_rate": 0.2944511280531544,
      "loss": 0.2479,
      "num_input_tokens_seen": 6696624,
      "step": 3475
    },
    {
      "epoch": 0.5677461456888816,
      "grad_norm": 0.014584225602447987,
      "learning_rate": 0.29443524355612083,
      "loss": 0.2545,
      "num_input_tokens_seen": 6705376,
      "step": 3480
    },
    {
      "epoch": 0.568561872909699,
      "grad_norm": 0.019725698977708817,
      "learning_rate": 0.29441933678535764,
      "loss": 0.2761,
      "num_input_tokens_seen": 6715168,
      "step": 3485
    },
    {
      "epoch": 0.5693776001305163,
      "grad_norm": 0.012981049716472626,
      "learning_rate": 0.29440340774331786,
      "loss": 0.2335,
      "num_input_tokens_seen": 6725008,
      "step": 3490
    },
    {
      "epoch": 0.5701933273513338,
      "grad_norm": 0.013202818110585213,
      "learning_rate": 0.2943874564324579,
      "loss": 0.2631,
      "num_input_tokens_seen": 6735520,
      "step": 3495
    },
    {
      "epoch": 0.5710090545721511,
      "grad_norm": 0.010455592535436153,
      "learning_rate": 0.2943714828552376,
      "loss": 0.2383,
      "num_input_tokens_seen": 6744992,
      "step": 3500
    },
    {
      "epoch": 0.5718247817929685,
      "grad_norm": 0.00843089073896408,
      "learning_rate": 0.29435548701412045,
      "loss": 0.223,
      "num_input_tokens_seen": 6753312,
      "step": 3505
    },
    {
      "epoch": 0.5726405090137858,
      "grad_norm": 0.017345238476991653,
      "learning_rate": 0.2943394689115731,
      "loss": 0.2297,
      "num_input_tokens_seen": 6763264,
      "step": 3510
    },
    {
      "epoch": 0.5734562362346032,
      "grad_norm": 0.02042365074157715,
      "learning_rate": 0.29432342855006577,
      "loss": 0.243,
      "num_input_tokens_seen": 6773568,
      "step": 3515
    },
    {
      "epoch": 0.5742719634554205,
      "grad_norm": 0.0160121638327837,
      "learning_rate": 0.294307365932072,
      "loss": 0.2475,
      "num_input_tokens_seen": 6782672,
      "step": 3520
    },
    {
      "epoch": 0.5750876906762379,
      "grad_norm": 0.017771650105714798,
      "learning_rate": 0.294291281060069,
      "loss": 0.2721,
      "num_input_tokens_seen": 6791728,
      "step": 3525
    },
    {
      "epoch": 0.5759034178970552,
      "grad_norm": 0.011349739506840706,
      "learning_rate": 0.29427517393653724,
      "loss": 0.2429,
      "num_input_tokens_seen": 6801856,
      "step": 3530
    },
    {
      "epoch": 0.5767191451178726,
      "grad_norm": 0.016264278441667557,
      "learning_rate": 0.29425904456396046,
      "loss": 0.2315,
      "num_input_tokens_seen": 6812272,
      "step": 3535
    },
    {
      "epoch": 0.5775348723386899,
      "grad_norm": 0.008917996659874916,
      "learning_rate": 0.2942428929448262,
      "loss": 0.2851,
      "num_input_tokens_seen": 6821552,
      "step": 3540
    },
    {
      "epoch": 0.5783505995595073,
      "grad_norm": 0.011969558894634247,
      "learning_rate": 0.2942267190816252,
      "loss": 0.2124,
      "num_input_tokens_seen": 6831040,
      "step": 3545
    },
    {
      "epoch": 0.5791663267803246,
      "grad_norm": 0.012944289483129978,
      "learning_rate": 0.2942105229768516,
      "loss": 0.2376,
      "num_input_tokens_seen": 6840832,
      "step": 3550
    },
    {
      "epoch": 0.579982054001142,
      "grad_norm": 0.011282620951533318,
      "learning_rate": 0.29419430463300306,
      "loss": 0.225,
      "num_input_tokens_seen": 6850832,
      "step": 3555
    },
    {
      "epoch": 0.5807977812219594,
      "grad_norm": 0.024840280413627625,
      "learning_rate": 0.2941780640525808,
      "loss": 0.2481,
      "num_input_tokens_seen": 6861296,
      "step": 3560
    },
    {
      "epoch": 0.5816135084427767,
      "grad_norm": 0.01072059664875269,
      "learning_rate": 0.2941618012380891,
      "loss": 0.2333,
      "num_input_tokens_seen": 6870480,
      "step": 3565
    },
    {
      "epoch": 0.5824292356635941,
      "grad_norm": 0.013209078460931778,
      "learning_rate": 0.29414551619203605,
      "loss": 0.2209,
      "num_input_tokens_seen": 6879648,
      "step": 3570
    },
    {
      "epoch": 0.5832449628844114,
      "grad_norm": 0.010932535864412785,
      "learning_rate": 0.29412920891693295,
      "loss": 0.2647,
      "num_input_tokens_seen": 6889728,
      "step": 3575
    },
    {
      "epoch": 0.5840606901052288,
      "grad_norm": 0.011359712108969688,
      "learning_rate": 0.2941128794152946,
      "loss": 0.2647,
      "num_input_tokens_seen": 6898416,
      "step": 3580
    },
    {
      "epoch": 0.5848764173260461,
      "grad_norm": 0.012798167765140533,
      "learning_rate": 0.2940965276896392,
      "loss": 0.2478,
      "num_input_tokens_seen": 6907072,
      "step": 3585
    },
    {
      "epoch": 0.5856921445468636,
      "grad_norm": 0.008221741765737534,
      "learning_rate": 0.2940801537424884,
      "loss": 0.2428,
      "num_input_tokens_seen": 6916880,
      "step": 3590
    },
    {
      "epoch": 0.5865078717676809,
      "grad_norm": 0.01670309528708458,
      "learning_rate": 0.2940637575763673,
      "loss": 0.2632,
      "num_input_tokens_seen": 6927504,
      "step": 3595
    },
    {
      "epoch": 0.5873235989884983,
      "grad_norm": 0.013672412373125553,
      "learning_rate": 0.2940473391938043,
      "loss": 0.2122,
      "num_input_tokens_seen": 6936656,
      "step": 3600
    },
    {
      "epoch": 0.5873235989884983,
      "eval_loss": 0.23214249312877655,
      "eval_runtime": 68.1142,
      "eval_samples_per_second": 40.006,
      "eval_steps_per_second": 20.011,
      "num_input_tokens_seen": 6936656,
      "step": 3600
    },
    {
      "epoch": 0.5881393262093156,
      "grad_norm": 0.006907463539391756,
      "learning_rate": 0.29403089859733145,
      "loss": 0.3014,
      "num_input_tokens_seen": 6946256,
      "step": 3605
    },
    {
      "epoch": 0.588955053430133,
      "grad_norm": 0.007664334494620562,
      "learning_rate": 0.294014435789484,
      "loss": 0.2431,
      "num_input_tokens_seen": 6955936,
      "step": 3610
    },
    {
      "epoch": 0.5897707806509503,
      "grad_norm": 0.008315635845065117,
      "learning_rate": 0.2939979507728007,
      "loss": 0.2285,
      "num_input_tokens_seen": 6965504,
      "step": 3615
    },
    {
      "epoch": 0.5905865078717677,
      "grad_norm": 0.008450627326965332,
      "learning_rate": 0.2939814435498239,
      "loss": 0.2214,
      "num_input_tokens_seen": 6974768,
      "step": 3620
    },
    {
      "epoch": 0.591402235092585,
      "grad_norm": 0.008546799421310425,
      "learning_rate": 0.29396491412309905,
      "loss": 0.2063,
      "num_input_tokens_seen": 6985872,
      "step": 3625
    },
    {
      "epoch": 0.5922179623134024,
      "grad_norm": 0.008425581268966198,
      "learning_rate": 0.2939483624951753,
      "loss": 0.2732,
      "num_input_tokens_seen": 6995248,
      "step": 3630
    },
    {
      "epoch": 0.5930336895342198,
      "grad_norm": 0.007862819358706474,
      "learning_rate": 0.2939317886686051,
      "loss": 0.1852,
      "num_input_tokens_seen": 7004928,
      "step": 3635
    },
    {
      "epoch": 0.5938494167550371,
      "grad_norm": 0.00882960669696331,
      "learning_rate": 0.2939151926459443,
      "loss": 0.2267,
      "num_input_tokens_seen": 7013552,
      "step": 3640
    },
    {
      "epoch": 0.5946651439758545,
      "grad_norm": 0.012130585499107838,
      "learning_rate": 0.2938985744297522,
      "loss": 0.2574,
      "num_input_tokens_seen": 7022720,
      "step": 3645
    },
    {
      "epoch": 0.5954808711966718,
      "grad_norm": 0.008643802255392075,
      "learning_rate": 0.29388193402259166,
      "loss": 0.2428,
      "num_input_tokens_seen": 7032624,
      "step": 3650
    },
    {
      "epoch": 0.5962965984174892,
      "grad_norm": 0.008169748820364475,
      "learning_rate": 0.29386527142702873,
      "loss": 0.2329,
      "num_input_tokens_seen": 7043456,
      "step": 3655
    },
    {
      "epoch": 0.5971123256383065,
      "grad_norm": 0.009311992675065994,
      "learning_rate": 0.293848586645633,
      "loss": 0.2472,
      "num_input_tokens_seen": 7054048,
      "step": 3660
    },
    {
      "epoch": 0.5979280528591239,
      "grad_norm": 0.007616094313561916,
      "learning_rate": 0.2938318796809775,
      "loss": 0.2544,
      "num_input_tokens_seen": 7063280,
      "step": 3665
    },
    {
      "epoch": 0.5987437800799412,
      "grad_norm": 0.006911677308380604,
      "learning_rate": 0.29381515053563867,
      "loss": 0.2277,
      "num_input_tokens_seen": 7072192,
      "step": 3670
    },
    {
      "epoch": 0.5995595073007586,
      "grad_norm": 0.00680178590118885,
      "learning_rate": 0.29379839921219636,
      "loss": 0.2447,
      "num_input_tokens_seen": 7082288,
      "step": 3675
    },
    {
      "epoch": 0.600375234521576,
      "grad_norm": 0.006667545530945063,
      "learning_rate": 0.2937816257132338,
      "loss": 0.232,
      "num_input_tokens_seen": 7093952,
      "step": 3680
    },
    {
      "epoch": 0.6011909617423934,
      "grad_norm": 0.01078128069639206,
      "learning_rate": 0.2937648300413376,
      "loss": 0.2389,
      "num_input_tokens_seen": 7101680,
      "step": 3685
    },
    {
      "epoch": 0.6020066889632107,
      "grad_norm": 0.010485882870852947,
      "learning_rate": 0.293748012199098,
      "loss": 0.2466,
      "num_input_tokens_seen": 7111984,
      "step": 3690
    },
    {
      "epoch": 0.6028224161840281,
      "grad_norm": 0.00796808023005724,
      "learning_rate": 0.29373117218910844,
      "loss": 0.2263,
      "num_input_tokens_seen": 7120272,
      "step": 3695
    },
    {
      "epoch": 0.6036381434048455,
      "grad_norm": 0.012895342893898487,
      "learning_rate": 0.2937143100139659,
      "loss": 0.2577,
      "num_input_tokens_seen": 7129456,
      "step": 3700
    },
    {
      "epoch": 0.6044538706256628,
      "grad_norm": 0.012823603115975857,
      "learning_rate": 0.29369742567627083,
      "loss": 0.2538,
      "num_input_tokens_seen": 7139824,
      "step": 3705
    },
    {
      "epoch": 0.6052695978464802,
      "grad_norm": 0.009607899934053421,
      "learning_rate": 0.29368051917862675,
      "loss": 0.2685,
      "num_input_tokens_seen": 7148640,
      "step": 3710
    },
    {
      "epoch": 0.6060853250672975,
      "grad_norm": 0.011414673179388046,
      "learning_rate": 0.2936635905236411,
      "loss": 0.2505,
      "num_input_tokens_seen": 7157920,
      "step": 3715
    },
    {
      "epoch": 0.6069010522881149,
      "grad_norm": 0.010624384507536888,
      "learning_rate": 0.2936466397139244,
      "loss": 0.2458,
      "num_input_tokens_seen": 7166928,
      "step": 3720
    },
    {
      "epoch": 0.6077167795089322,
      "grad_norm": 0.0073387399315834045,
      "learning_rate": 0.2936296667520907,
      "loss": 0.2264,
      "num_input_tokens_seen": 7176688,
      "step": 3725
    },
    {
      "epoch": 0.6085325067297496,
      "grad_norm": 0.008781729266047478,
      "learning_rate": 0.2936126716407574,
      "loss": 0.2271,
      "num_input_tokens_seen": 7185824,
      "step": 3730
    },
    {
      "epoch": 0.6093482339505669,
      "grad_norm": 0.02707662247121334,
      "learning_rate": 0.29359565438254537,
      "loss": 0.2561,
      "num_input_tokens_seen": 7197040,
      "step": 3735
    },
    {
      "epoch": 0.6101639611713843,
      "grad_norm": 0.013096301816403866,
      "learning_rate": 0.29357861498007887,
      "loss": 0.2359,
      "num_input_tokens_seen": 7207136,
      "step": 3740
    },
    {
      "epoch": 0.6109796883922016,
      "grad_norm": 0.011098435148596764,
      "learning_rate": 0.29356155343598567,
      "loss": 0.2452,
      "num_input_tokens_seen": 7216624,
      "step": 3745
    },
    {
      "epoch": 0.611795415613019,
      "grad_norm": 0.008904474787414074,
      "learning_rate": 0.2935444697528968,
      "loss": 0.2601,
      "num_input_tokens_seen": 7226320,
      "step": 3750
    },
    {
      "epoch": 0.6126111428338363,
      "grad_norm": 0.00688509875908494,
      "learning_rate": 0.2935273639334468,
      "loss": 0.232,
      "num_input_tokens_seen": 7235728,
      "step": 3755
    },
    {
      "epoch": 0.6134268700546537,
      "grad_norm": 0.008390267379581928,
      "learning_rate": 0.29351023598027365,
      "loss": 0.2223,
      "num_input_tokens_seen": 7244656,
      "step": 3760
    },
    {
      "epoch": 0.614242597275471,
      "grad_norm": 0.0080176442861557,
      "learning_rate": 0.2934930858960186,
      "loss": 0.2406,
      "num_input_tokens_seen": 7254624,
      "step": 3765
    },
    {
      "epoch": 0.6150583244962884,
      "grad_norm": 0.00992939155548811,
      "learning_rate": 0.29347591368332643,
      "loss": 0.1899,
      "num_input_tokens_seen": 7265168,
      "step": 3770
    },
    {
      "epoch": 0.6158740517171059,
      "grad_norm": 0.00897907093167305,
      "learning_rate": 0.2934587193448454,
      "loss": 0.2496,
      "num_input_tokens_seen": 7275504,
      "step": 3775
    },
    {
      "epoch": 0.6166897789379232,
      "grad_norm": 0.014790310524404049,
      "learning_rate": 0.29344150288322696,
      "loss": 0.2565,
      "num_input_tokens_seen": 7284864,
      "step": 3780
    },
    {
      "epoch": 0.6175055061587406,
      "grad_norm": 0.007081762421876192,
      "learning_rate": 0.2934242643011263,
      "loss": 0.2488,
      "num_input_tokens_seen": 7293616,
      "step": 3785
    },
    {
      "epoch": 0.6183212333795579,
      "grad_norm": 0.0076677524484694,
      "learning_rate": 0.2934070036012016,
      "loss": 0.2668,
      "num_input_tokens_seen": 7303152,
      "step": 3790
    },
    {
      "epoch": 0.6191369606003753,
      "grad_norm": 0.02821490727365017,
      "learning_rate": 0.29338972078611475,
      "loss": 0.2605,
      "num_input_tokens_seen": 7312672,
      "step": 3795
    },
    {
      "epoch": 0.6199526878211926,
      "grad_norm": 0.0057229939848184586,
      "learning_rate": 0.2933724158585311,
      "loss": 0.2201,
      "num_input_tokens_seen": 7321136,
      "step": 3800
    },
    {
      "epoch": 0.6199526878211926,
      "eval_loss": 0.23056231439113617,
      "eval_runtime": 67.9696,
      "eval_samples_per_second": 40.091,
      "eval_steps_per_second": 20.053,
      "num_input_tokens_seen": 7321136,
      "step": 3800
    },
    {
      "epoch": 0.62076841504201,
      "grad_norm": 0.007890831679105759,
      "learning_rate": 0.29335508882111916,
      "loss": 0.2247,
      "num_input_tokens_seen": 7330464,
      "step": 3805
    },
    {
      "epoch": 0.6215841422628273,
      "grad_norm": 0.007338678929954767,
      "learning_rate": 0.29333773967655097,
      "loss": 0.2081,
      "num_input_tokens_seen": 7339760,
      "step": 3810
    },
    {
      "epoch": 0.6223998694836447,
      "grad_norm": 0.0074355159886181355,
      "learning_rate": 0.2933203684275021,
      "loss": 0.2068,
      "num_input_tokens_seen": 7351008,
      "step": 3815
    },
    {
      "epoch": 0.623215596704462,
      "grad_norm": 0.007556794676929712,
      "learning_rate": 0.2933029750766513,
      "loss": 0.2415,
      "num_input_tokens_seen": 7360928,
      "step": 3820
    },
    {
      "epoch": 0.6240313239252794,
      "grad_norm": 0.01843760348856449,
      "learning_rate": 0.2932855596266809,
      "loss": 0.2289,
      "num_input_tokens_seen": 7370672,
      "step": 3825
    },
    {
      "epoch": 0.6248470511460967,
      "grad_norm": 0.007266793865710497,
      "learning_rate": 0.2932681220802765,
      "loss": 0.2037,
      "num_input_tokens_seen": 7380448,
      "step": 3830
    },
    {
      "epoch": 0.6256627783669141,
      "grad_norm": 0.007006234489381313,
      "learning_rate": 0.2932506624401274,
      "loss": 0.2466,
      "num_input_tokens_seen": 7389392,
      "step": 3835
    },
    {
      "epoch": 0.6264785055877314,
      "grad_norm": 0.011170582845807076,
      "learning_rate": 0.29323318070892584,
      "loss": 0.2564,
      "num_input_tokens_seen": 7399984,
      "step": 3840
    },
    {
      "epoch": 0.6272942328085488,
      "grad_norm": 0.009392092935740948,
      "learning_rate": 0.29321567688936784,
      "loss": 0.2344,
      "num_input_tokens_seen": 7408432,
      "step": 3845
    },
    {
      "epoch": 0.6281099600293661,
      "grad_norm": 0.009227736853063107,
      "learning_rate": 0.29319815098415275,
      "loss": 0.2412,
      "num_input_tokens_seen": 7418864,
      "step": 3850
    },
    {
      "epoch": 0.6289256872501835,
      "grad_norm": 0.01870719902217388,
      "learning_rate": 0.2931806029959832,
      "loss": 0.256,
      "num_input_tokens_seen": 7426688,
      "step": 3855
    },
    {
      "epoch": 0.6297414144710008,
      "grad_norm": 0.005223392508924007,
      "learning_rate": 0.29316303292756535,
      "loss": 0.2356,
      "num_input_tokens_seen": 7436592,
      "step": 3860
    },
    {
      "epoch": 0.6305571416918182,
      "grad_norm": 0.01269554067403078,
      "learning_rate": 0.29314544078160876,
      "loss": 0.2154,
      "num_input_tokens_seen": 7446512,
      "step": 3865
    },
    {
      "epoch": 0.6313728689126357,
      "grad_norm": 0.006091979797929525,
      "learning_rate": 0.2931278265608263,
      "loss": 0.2384,
      "num_input_tokens_seen": 7456896,
      "step": 3870
    },
    {
      "epoch": 0.632188596133453,
      "grad_norm": 0.009597817435860634,
      "learning_rate": 0.29311019026793433,
      "loss": 0.2224,
      "num_input_tokens_seen": 7466832,
      "step": 3875
    },
    {
      "epoch": 0.6330043233542704,
      "grad_norm": 0.010805976577103138,
      "learning_rate": 0.29309253190565254,
      "loss": 0.2476,
      "num_input_tokens_seen": 7476528,
      "step": 3880
    },
    {
      "epoch": 0.6338200505750877,
      "grad_norm": 0.009025256149470806,
      "learning_rate": 0.2930748514767042,
      "loss": 0.2496,
      "num_input_tokens_seen": 7485280,
      "step": 3885
    },
    {
      "epoch": 0.6346357777959051,
      "grad_norm": 0.00900591816753149,
      "learning_rate": 0.29305714898381574,
      "loss": 0.2686,
      "num_input_tokens_seen": 7494336,
      "step": 3890
    },
    {
      "epoch": 0.6354515050167224,
      "grad_norm": 0.009152676910161972,
      "learning_rate": 0.29303942442971714,
      "loss": 0.2439,
      "num_input_tokens_seen": 7503792,
      "step": 3895
    },
    {
      "epoch": 0.6362672322375398,
      "grad_norm": 0.009050874970853329,
      "learning_rate": 0.2930216778171417,
      "loss": 0.2516,
      "num_input_tokens_seen": 7514912,
      "step": 3900
    },
    {
      "epoch": 0.6370829594583571,
      "grad_norm": 0.0065284003503620625,
      "learning_rate": 0.2930039091488263,
      "loss": 0.2355,
      "num_input_tokens_seen": 7524112,
      "step": 3905
    },
    {
      "epoch": 0.6378986866791745,
      "grad_norm": 0.00556637067347765,
      "learning_rate": 0.29298611842751093,
      "loss": 0.2189,
      "num_input_tokens_seen": 7534912,
      "step": 3910
    },
    {
      "epoch": 0.6387144138999918,
      "grad_norm": 0.012394110672175884,
      "learning_rate": 0.29296830565593923,
      "loss": 0.2663,
      "num_input_tokens_seen": 7544112,
      "step": 3915
    },
    {
      "epoch": 0.6395301411208092,
      "grad_norm": 0.008634750731289387,
      "learning_rate": 0.2929504708368582,
      "loss": 0.2225,
      "num_input_tokens_seen": 7553424,
      "step": 3920
    },
    {
      "epoch": 0.6403458683416265,
      "grad_norm": 0.009948216378688812,
      "learning_rate": 0.29293261397301806,
      "loss": 0.2314,
      "num_input_tokens_seen": 7563616,
      "step": 3925
    },
    {
      "epoch": 0.6411615955624439,
      "grad_norm": 0.011783702298998833,
      "learning_rate": 0.29291473506717275,
      "loss": 0.2551,
      "num_input_tokens_seen": 7573824,
      "step": 3930
    },
    {
      "epoch": 0.6419773227832613,
      "grad_norm": 0.006150904111564159,
      "learning_rate": 0.29289683412207923,
      "loss": 0.2395,
      "num_input_tokens_seen": 7583456,
      "step": 3935
    },
    {
      "epoch": 0.6427930500040786,
      "grad_norm": 0.01144163217395544,
      "learning_rate": 0.29287891114049813,
      "loss": 0.2626,
      "num_input_tokens_seen": 7593280,
      "step": 3940
    },
    {
      "epoch": 0.643608777224896,
      "grad_norm": 0.00847548246383667,
      "learning_rate": 0.29286096612519347,
      "loss": 0.2483,
      "num_input_tokens_seen": 7602976,
      "step": 3945
    },
    {
      "epoch": 0.6444245044457133,
      "grad_norm": 0.007451524958014488,
      "learning_rate": 0.2928429990789325,
      "loss": 0.2362,
      "num_input_tokens_seen": 7613328,
      "step": 3950
    },
    {
      "epoch": 0.6452402316665307,
      "grad_norm": 0.010290395468473434,
      "learning_rate": 0.29282501000448596,
      "loss": 0.1937,
      "num_input_tokens_seen": 7623664,
      "step": 3955
    },
    {
      "epoch": 0.6460559588873481,
      "grad_norm": 0.0068085952661931515,
      "learning_rate": 0.2928069989046281,
      "loss": 0.2289,
      "num_input_tokens_seen": 7633552,
      "step": 3960
    },
    {
      "epoch": 0.6468716861081655,
      "grad_norm": 0.0183284692466259,
      "learning_rate": 0.2927889657821363,
      "loss": 0.3462,
      "num_input_tokens_seen": 7642912,
      "step": 3965
    },
    {
      "epoch": 0.6476874133289828,
      "grad_norm": 0.00683052372187376,
      "learning_rate": 0.2927709106397916,
      "loss": 0.2492,
      "num_input_tokens_seen": 7651792,
      "step": 3970
    },
    {
      "epoch": 0.6485031405498002,
      "grad_norm": 0.0058819446712732315,
      "learning_rate": 0.29275283348037834,
      "loss": 0.2475,
      "num_input_tokens_seen": 7662016,
      "step": 3975
    },
    {
      "epoch": 0.6493188677706175,
      "grad_norm": 0.009171226993203163,
      "learning_rate": 0.29273473430668423,
      "loss": 0.238,
      "num_input_tokens_seen": 7671552,
      "step": 3980
    },
    {
      "epoch": 0.6501345949914349,
      "grad_norm": 0.007400431204587221,
      "learning_rate": 0.2927166131215003,
      "loss": 0.2384,
      "num_input_tokens_seen": 7681152,
      "step": 3985
    },
    {
      "epoch": 0.6509503222122522,
      "grad_norm": 0.00847187452018261,
      "learning_rate": 0.2926984699276212,
      "loss": 0.2237,
      "num_input_tokens_seen": 7690960,
      "step": 3990
    },
    {
      "epoch": 0.6517660494330696,
      "grad_norm": 0.006566270720213652,
      "learning_rate": 0.29268030472784473,
      "loss": 0.2759,
      "num_input_tokens_seen": 7700032,
      "step": 3995
    },
    {
      "epoch": 0.652581776653887,
      "grad_norm": 0.006588206626474857,
      "learning_rate": 0.2926621175249723,
      "loss": 0.2457,
      "num_input_tokens_seen": 7709856,
      "step": 4000
    },
    {
      "epoch": 0.652581776653887,
      "eval_loss": 0.2302481234073639,
      "eval_runtime": 68.0291,
      "eval_samples_per_second": 40.056,
      "eval_steps_per_second": 20.036,
      "num_input_tokens_seen": 7709856,
      "step": 4000
    },
    {
      "epoch": 0.6533975038747043,
      "grad_norm": 0.00661895889788866,
      "learning_rate": 0.29264390832180853,
      "loss": 0.2294,
      "num_input_tokens_seen": 7719216,
      "step": 4005
    },
    {
      "epoch": 0.6542132310955217,
      "grad_norm": 0.006525194738060236,
      "learning_rate": 0.29262567712116144,
      "loss": 0.2271,
      "num_input_tokens_seen": 7729584,
      "step": 4010
    },
    {
      "epoch": 0.655028958316339,
      "grad_norm": 0.010200617834925652,
      "learning_rate": 0.29260742392584266,
      "loss": 0.2449,
      "num_input_tokens_seen": 7739328,
      "step": 4015
    },
    {
      "epoch": 0.6558446855371564,
      "grad_norm": 0.0069035557098686695,
      "learning_rate": 0.292589148738667,
      "loss": 0.2442,
      "num_input_tokens_seen": 7748544,
      "step": 4020
    },
    {
      "epoch": 0.6566604127579737,
      "grad_norm": 0.0072862980887293816,
      "learning_rate": 0.2925708515624527,
      "loss": 0.2385,
      "num_input_tokens_seen": 7758400,
      "step": 4025
    },
    {
      "epoch": 0.6574761399787911,
      "grad_norm": 0.009769286029040813,
      "learning_rate": 0.29255253240002144,
      "loss": 0.237,
      "num_input_tokens_seen": 7769088,
      "step": 4030
    },
    {
      "epoch": 0.6582918671996084,
      "grad_norm": 0.010402113199234009,
      "learning_rate": 0.2925341912541983,
      "loss": 0.242,
      "num_input_tokens_seen": 7779616,
      "step": 4035
    },
    {
      "epoch": 0.6591075944204258,
      "grad_norm": 0.0060406094416975975,
      "learning_rate": 0.2925158281278116,
      "loss": 0.2478,
      "num_input_tokens_seen": 7790928,
      "step": 4040
    },
    {
      "epoch": 0.6599233216412431,
      "grad_norm": 0.01349208690226078,
      "learning_rate": 0.29249744302369324,
      "loss": 0.2445,
      "num_input_tokens_seen": 7800032,
      "step": 4045
    },
    {
      "epoch": 0.6607390488620605,
      "grad_norm": 0.009422584436833858,
      "learning_rate": 0.29247903594467844,
      "loss": 0.2256,
      "num_input_tokens_seen": 7809184,
      "step": 4050
    },
    {
      "epoch": 0.6615547760828779,
      "grad_norm": 0.007951340638101101,
      "learning_rate": 0.2924606068936058,
      "loss": 0.2329,
      "num_input_tokens_seen": 7819232,
      "step": 4055
    },
    {
      "epoch": 0.6623705033036953,
      "grad_norm": 0.008903654292225838,
      "learning_rate": 0.2924421558733173,
      "loss": 0.2349,
      "num_input_tokens_seen": 7829392,
      "step": 4060
    },
    {
      "epoch": 0.6631862305245126,
      "grad_norm": 0.011282207444310188,
      "learning_rate": 0.2924236828866583,
      "loss": 0.2235,
      "num_input_tokens_seen": 7838640,
      "step": 4065
    },
    {
      "epoch": 0.66400195774533,
      "grad_norm": 0.008541378192603588,
      "learning_rate": 0.29240518793647763,
      "loss": 0.206,
      "num_input_tokens_seen": 7848912,
      "step": 4070
    },
    {
      "epoch": 0.6648176849661473,
      "grad_norm": 0.009274479001760483,
      "learning_rate": 0.29238667102562743,
      "loss": 0.2256,
      "num_input_tokens_seen": 7860272,
      "step": 4075
    },
    {
      "epoch": 0.6656334121869647,
      "grad_norm": 0.010251122526824474,
      "learning_rate": 0.29236813215696317,
      "loss": 0.3014,
      "num_input_tokens_seen": 7868640,
      "step": 4080
    },
    {
      "epoch": 0.666449139407782,
      "grad_norm": 0.007691596169024706,
      "learning_rate": 0.2923495713333439,
      "loss": 0.2215,
      "num_input_tokens_seen": 7878256,
      "step": 4085
    },
    {
      "epoch": 0.6672648666285994,
      "grad_norm": 0.005633846390992403,
      "learning_rate": 0.29233098855763173,
      "loss": 0.2536,
      "num_input_tokens_seen": 7886576,
      "step": 4090
    },
    {
      "epoch": 0.6680805938494168,
      "grad_norm": 0.008363613858819008,
      "learning_rate": 0.29231238383269254,
      "loss": 0.2367,
      "num_input_tokens_seen": 7896944,
      "step": 4095
    },
    {
      "epoch": 0.6688963210702341,
      "grad_norm": 0.008252461440861225,
      "learning_rate": 0.2922937571613954,
      "loss": 0.2301,
      "num_input_tokens_seen": 7907472,
      "step": 4100
    },
    {
      "epoch": 0.6697120482910515,
      "grad_norm": 0.008003192022442818,
      "learning_rate": 0.29227510854661265,
      "loss": 0.2427,
      "num_input_tokens_seen": 7917328,
      "step": 4105
    },
    {
      "epoch": 0.6705277755118688,
      "grad_norm": 0.006926637142896652,
      "learning_rate": 0.29225643799122025,
      "loss": 0.2619,
      "num_input_tokens_seen": 7927344,
      "step": 4110
    },
    {
      "epoch": 0.6713435027326862,
      "grad_norm": 0.010354756377637386,
      "learning_rate": 0.2922377454980974,
      "loss": 0.2426,
      "num_input_tokens_seen": 7935840,
      "step": 4115
    },
    {
      "epoch": 0.6721592299535035,
      "grad_norm": 0.008017371408641338,
      "learning_rate": 0.29221903107012676,
      "loss": 0.2294,
      "num_input_tokens_seen": 7945696,
      "step": 4120
    },
    {
      "epoch": 0.6729749571743209,
      "grad_norm": 0.007778766565024853,
      "learning_rate": 0.29220029471019426,
      "loss": 0.2198,
      "num_input_tokens_seen": 7955536,
      "step": 4125
    },
    {
      "epoch": 0.6737906843951382,
      "grad_norm": 0.009635258466005325,
      "learning_rate": 0.2921815364211893,
      "loss": 0.2522,
      "num_input_tokens_seen": 7965200,
      "step": 4130
    },
    {
      "epoch": 0.6746064116159556,
      "grad_norm": 0.01176320482045412,
      "learning_rate": 0.29216275620600474,
      "loss": 0.1919,
      "num_input_tokens_seen": 7975072,
      "step": 4135
    },
    {
      "epoch": 0.6754221388367729,
      "grad_norm": 0.009723697789013386,
      "learning_rate": 0.29214395406753657,
      "loss": 0.1729,
      "num_input_tokens_seen": 7984080,
      "step": 4140
    },
    {
      "epoch": 0.6762378660575903,
      "grad_norm": 0.005592124070972204,
      "learning_rate": 0.2921251300086844,
      "loss": 0.2623,
      "num_input_tokens_seen": 7993088,
      "step": 4145
    },
    {
      "epoch": 0.6770535932784078,
      "grad_norm": 0.014976056292653084,
      "learning_rate": 0.2921062840323511,
      "loss": 0.2697,
      "num_input_tokens_seen": 8002320,
      "step": 4150
    },
    {
      "epoch": 0.6778693204992251,
      "grad_norm": 0.0058735632337629795,
      "learning_rate": 0.29208741614144307,
      "loss": 0.2351,
      "num_input_tokens_seen": 8012096,
      "step": 4155
    },
    {
      "epoch": 0.6786850477200425,
      "grad_norm": 0.005836415570229292,
      "learning_rate": 0.2920685263388698,
      "loss": 0.2261,
      "num_input_tokens_seen": 8022160,
      "step": 4160
    },
    {
      "epoch": 0.6795007749408598,
      "grad_norm": 0.010030175559222698,
      "learning_rate": 0.2920496146275445,
      "loss": 0.1894,
      "num_input_tokens_seen": 8031552,
      "step": 4165
    },
    {
      "epoch": 0.6803165021616772,
      "grad_norm": 0.027855269610881805,
      "learning_rate": 0.29203068101038343,
      "loss": 0.3367,
      "num_input_tokens_seen": 8041504,
      "step": 4170
    },
    {
      "epoch": 0.6811322293824945,
      "grad_norm": 0.008845536969602108,
      "learning_rate": 0.2920117254903065,
      "loss": 0.2501,
      "num_input_tokens_seen": 8051168,
      "step": 4175
    },
    {
      "epoch": 0.6819479566033119,
      "grad_norm": 0.007844889536499977,
      "learning_rate": 0.29199274807023695,
      "loss": 0.2456,
      "num_input_tokens_seen": 8061936,
      "step": 4180
    },
    {
      "epoch": 0.6827636838241292,
      "grad_norm": 0.007143144961446524,
      "learning_rate": 0.29197374875310117,
      "loss": 0.2428,
      "num_input_tokens_seen": 8071936,
      "step": 4185
    },
    {
      "epoch": 0.6835794110449466,
      "grad_norm": 0.012344191782176495,
      "learning_rate": 0.2919547275418292,
      "loss": 0.2653,
      "num_input_tokens_seen": 8081568,
      "step": 4190
    },
    {
      "epoch": 0.6843951382657639,
      "grad_norm": 0.011056135408580303,
      "learning_rate": 0.29193568443935436,
      "loss": 0.2544,
      "num_input_tokens_seen": 8090592,
      "step": 4195
    },
    {
      "epoch": 0.6852108654865813,
      "grad_norm": 0.00888869445770979,
      "learning_rate": 0.2919166194486133,
      "loss": 0.2281,
      "num_input_tokens_seen": 8100560,
      "step": 4200
    },
    {
      "epoch": 0.6852108654865813,
      "eval_loss": 0.22748814523220062,
      "eval_runtime": 68.0588,
      "eval_samples_per_second": 40.039,
      "eval_steps_per_second": 20.027,
      "num_input_tokens_seen": 8100560,
      "step": 4200
    },
    {
      "epoch": 0.6860265927073986,
      "grad_norm": 0.00634931493550539,
      "learning_rate": 0.2918975325725461,
      "loss": 0.2458,
      "num_input_tokens_seen": 8110688,
      "step": 4205
    },
    {
      "epoch": 0.686842319928216,
      "grad_norm": 0.006010852754116058,
      "learning_rate": 0.29187842381409607,
      "loss": 0.2419,
      "num_input_tokens_seen": 8120448,
      "step": 4210
    },
    {
      "epoch": 0.6876580471490333,
      "grad_norm": 0.007342149503529072,
      "learning_rate": 0.29185929317621023,
      "loss": 0.222,
      "num_input_tokens_seen": 8130608,
      "step": 4215
    },
    {
      "epoch": 0.6884737743698507,
      "grad_norm": 0.006101836916059256,
      "learning_rate": 0.29184014066183867,
      "loss": 0.2442,
      "num_input_tokens_seen": 8140592,
      "step": 4220
    },
    {
      "epoch": 0.689289501590668,
      "grad_norm": 0.00571104884147644,
      "learning_rate": 0.2918209662739349,
      "loss": 0.2381,
      "num_input_tokens_seen": 8150176,
      "step": 4225
    },
    {
      "epoch": 0.6901052288114854,
      "grad_norm": 0.007090632803738117,
      "learning_rate": 0.29180177001545593,
      "loss": 0.2374,
      "num_input_tokens_seen": 8161568,
      "step": 4230
    },
    {
      "epoch": 0.6909209560323027,
      "grad_norm": 0.007790496107190847,
      "learning_rate": 0.29178255188936203,
      "loss": 0.2436,
      "num_input_tokens_seen": 8170896,
      "step": 4235
    },
    {
      "epoch": 0.6917366832531202,
      "grad_norm": 0.005901761818677187,
      "learning_rate": 0.2917633118986169,
      "loss": 0.2141,
      "num_input_tokens_seen": 8179616,
      "step": 4240
    },
    {
      "epoch": 0.6925524104739376,
      "grad_norm": 0.00588049553334713,
      "learning_rate": 0.2917440500461875,
      "loss": 0.2168,
      "num_input_tokens_seen": 8188608,
      "step": 4245
    },
    {
      "epoch": 0.6933681376947549,
      "grad_norm": 0.009124120697379112,
      "learning_rate": 0.29172476633504435,
      "loss": 0.2539,
      "num_input_tokens_seen": 8197952,
      "step": 4250
    },
    {
      "epoch": 0.6941838649155723,
      "grad_norm": 0.008203282952308655,
      "learning_rate": 0.2917054607681612,
      "loss": 0.2196,
      "num_input_tokens_seen": 8206256,
      "step": 4255
    },
    {
      "epoch": 0.6949995921363896,
      "grad_norm": 0.006334899924695492,
      "learning_rate": 0.29168613334851523,
      "loss": 0.2122,
      "num_input_tokens_seen": 8215408,
      "step": 4260
    },
    {
      "epoch": 0.695815319357207,
      "grad_norm": 0.006699805147945881,
      "learning_rate": 0.2916667840790869,
      "loss": 0.2207,
      "num_input_tokens_seen": 8225376,
      "step": 4265
    },
    {
      "epoch": 0.6966310465780243,
      "grad_norm": 0.009030637331306934,
      "learning_rate": 0.2916474129628603,
      "loss": 0.2215,
      "num_input_tokens_seen": 8234512,
      "step": 4270
    },
    {
      "epoch": 0.6974467737988417,
      "grad_norm": 0.007298442535102367,
      "learning_rate": 0.29162802000282245,
      "loss": 0.2409,
      "num_input_tokens_seen": 8243152,
      "step": 4275
    },
    {
      "epoch": 0.698262501019659,
      "grad_norm": 0.0097851837053895,
      "learning_rate": 0.2916086052019642,
      "loss": 0.217,
      "num_input_tokens_seen": 8253056,
      "step": 4280
    },
    {
      "epoch": 0.6990782282404764,
      "grad_norm": 0.007277856580913067,
      "learning_rate": 0.2915891685632794,
      "loss": 0.2578,
      "num_input_tokens_seen": 8262416,
      "step": 4285
    },
    {
      "epoch": 0.6998939554612937,
      "grad_norm": 0.011085834354162216,
      "learning_rate": 0.29156971008976545,
      "loss": 0.2422,
      "num_input_tokens_seen": 8271856,
      "step": 4290
    },
    {
      "epoch": 0.7007096826821111,
      "grad_norm": 0.013057921081781387,
      "learning_rate": 0.2915502297844232,
      "loss": 0.2552,
      "num_input_tokens_seen": 8280048,
      "step": 4295
    },
    {
      "epoch": 0.7015254099029284,
      "grad_norm": 0.0075372084975242615,
      "learning_rate": 0.2915307276502566,
      "loss": 0.2381,
      "num_input_tokens_seen": 8288464,
      "step": 4300
    },
    {
      "epoch": 0.7023411371237458,
      "grad_norm": 0.005617037881165743,
      "learning_rate": 0.29151120369027334,
      "loss": 0.2888,
      "num_input_tokens_seen": 8297616,
      "step": 4305
    },
    {
      "epoch": 0.7031568643445631,
      "grad_norm": 0.011242855340242386,
      "learning_rate": 0.29149165790748405,
      "loss": 0.2672,
      "num_input_tokens_seen": 8307248,
      "step": 4310
    },
    {
      "epoch": 0.7039725915653805,
      "grad_norm": 0.005111074075102806,
      "learning_rate": 0.291472090304903,
      "loss": 0.2424,
      "num_input_tokens_seen": 8316384,
      "step": 4315
    },
    {
      "epoch": 0.7047883187861979,
      "grad_norm": 0.005388911347836256,
      "learning_rate": 0.2914525008855478,
      "loss": 0.2312,
      "num_input_tokens_seen": 8326416,
      "step": 4320
    },
    {
      "epoch": 0.7056040460070152,
      "grad_norm": 0.006256022956222296,
      "learning_rate": 0.2914328896524394,
      "loss": 0.2228,
      "num_input_tokens_seen": 8336544,
      "step": 4325
    },
    {
      "epoch": 0.7064197732278326,
      "grad_norm": 0.008239945396780968,
      "learning_rate": 0.291413256608602,
      "loss": 0.2473,
      "num_input_tokens_seen": 8347376,
      "step": 4330
    },
    {
      "epoch": 0.70723550044865,
      "grad_norm": 0.013902033679187298,
      "learning_rate": 0.29139360175706336,
      "loss": 0.2241,
      "num_input_tokens_seen": 8356672,
      "step": 4335
    },
    {
      "epoch": 0.7080512276694674,
      "grad_norm": 0.00769629655405879,
      "learning_rate": 0.2913739251008544,
      "loss": 0.2457,
      "num_input_tokens_seen": 8367408,
      "step": 4340
    },
    {
      "epoch": 0.7088669548902847,
      "grad_norm": 0.007451543118804693,
      "learning_rate": 0.29135422664300964,
      "loss": 0.2316,
      "num_input_tokens_seen": 8378032,
      "step": 4345
    },
    {
      "epoch": 0.7096826821111021,
      "grad_norm": 0.007369094528257847,
      "learning_rate": 0.29133450638656677,
      "loss": 0.2352,
      "num_input_tokens_seen": 8388080,
      "step": 4350
    },
    {
      "epoch": 0.7104984093319194,
      "grad_norm": 0.008514129556715488,
      "learning_rate": 0.2913147643345669,
      "loss": 0.2501,
      "num_input_tokens_seen": 8398160,
      "step": 4355
    },
    {
      "epoch": 0.7113141365527368,
      "grad_norm": 0.00677555613219738,
      "learning_rate": 0.29129500049005447,
      "loss": 0.2385,
      "num_input_tokens_seen": 8406944,
      "step": 4360
    },
    {
      "epoch": 0.7121298637735541,
      "grad_norm": 0.0047935908660292625,
      "learning_rate": 0.2912752148560773,
      "loss": 0.2214,
      "num_input_tokens_seen": 8415424,
      "step": 4365
    },
    {
      "epoch": 0.7129455909943715,
      "grad_norm": 0.008670474402606487,
      "learning_rate": 0.2912554074356866,
      "loss": 0.2408,
      "num_input_tokens_seen": 8425488,
      "step": 4370
    },
    {
      "epoch": 0.7137613182151888,
      "grad_norm": 0.008689714595675468,
      "learning_rate": 0.2912355782319371,
      "loss": 0.2648,
      "num_input_tokens_seen": 8434704,
      "step": 4375
    },
    {
      "epoch": 0.7145770454360062,
      "grad_norm": 0.004595798905938864,
      "learning_rate": 0.2912157272478864,
      "loss": 0.2038,
      "num_input_tokens_seen": 8443488,
      "step": 4380
    },
    {
      "epoch": 0.7153927726568236,
      "grad_norm": 0.005184877198189497,
      "learning_rate": 0.291195854486596,
      "loss": 0.1948,
      "num_input_tokens_seen": 8453120,
      "step": 4385
    },
    {
      "epoch": 0.7162084998776409,
      "grad_norm": 0.0061980802565813065,
      "learning_rate": 0.2911759599511305,
      "loss": 0.2446,
      "num_input_tokens_seen": 8463088,
      "step": 4390
    },
    {
      "epoch": 0.7170242270984583,
      "grad_norm": 0.007737793959677219,
      "learning_rate": 0.29115604364455777,
      "loss": 0.2341,
      "num_input_tokens_seen": 8472368,
      "step": 4395
    },
    {
      "epoch": 0.7178399543192756,
      "grad_norm": 0.007702968083322048,
      "learning_rate": 0.2911361055699493,
      "loss": 0.2378,
      "num_input_tokens_seen": 8482208,
      "step": 4400
    },
    {
      "epoch": 0.7178399543192756,
      "eval_loss": 0.2316390722990036,
      "eval_runtime": 68.0906,
      "eval_samples_per_second": 40.02,
      "eval_steps_per_second": 20.017,
      "num_input_tokens_seen": 8482208,
      "step": 4400
    },
    {
      "epoch": 0.718655681540093,
      "grad_norm": 0.012643540278077126,
      "learning_rate": 0.2911161457303797,
      "loss": 0.2532,
      "num_input_tokens_seen": 8491040,
      "step": 4405
    },
    {
      "epoch": 0.7194714087609103,
      "grad_norm": 0.0059903208166360855,
      "learning_rate": 0.291096164128927,
      "loss": 0.231,
      "num_input_tokens_seen": 8501152,
      "step": 4410
    },
    {
      "epoch": 0.7202871359817277,
      "grad_norm": 0.007892681285738945,
      "learning_rate": 0.2910761607686727,
      "loss": 0.2304,
      "num_input_tokens_seen": 8509984,
      "step": 4415
    },
    {
      "epoch": 0.721102863202545,
      "grad_norm": 0.007947713136672974,
      "learning_rate": 0.2910561356527016,
      "loss": 0.2329,
      "num_input_tokens_seen": 8519200,
      "step": 4420
    },
    {
      "epoch": 0.7219185904233625,
      "grad_norm": 0.005165505688637495,
      "learning_rate": 0.2910360887841017,
      "loss": 0.2455,
      "num_input_tokens_seen": 8528624,
      "step": 4425
    },
    {
      "epoch": 0.7227343176441798,
      "grad_norm": 0.0068440865725278854,
      "learning_rate": 0.2910160201659645,
      "loss": 0.2254,
      "num_input_tokens_seen": 8537408,
      "step": 4430
    },
    {
      "epoch": 0.7235500448649972,
      "grad_norm": 0.003965785261243582,
      "learning_rate": 0.29099592980138494,
      "loss": 0.2508,
      "num_input_tokens_seen": 8547408,
      "step": 4435
    },
    {
      "epoch": 0.7243657720858145,
      "grad_norm": 0.0062349895015358925,
      "learning_rate": 0.29097581769346115,
      "loss": 0.2427,
      "num_input_tokens_seen": 8557360,
      "step": 4440
    },
    {
      "epoch": 0.7251814993066319,
      "grad_norm": 0.005663699470460415,
      "learning_rate": 0.29095568384529463,
      "loss": 0.2413,
      "num_input_tokens_seen": 8567088,
      "step": 4445
    },
    {
      "epoch": 0.7259972265274492,
      "grad_norm": 0.005866413936018944,
      "learning_rate": 0.2909355282599903,
      "loss": 0.222,
      "num_input_tokens_seen": 8576320,
      "step": 4450
    },
    {
      "epoch": 0.7268129537482666,
      "grad_norm": 0.008959353901445866,
      "learning_rate": 0.29091535094065635,
      "loss": 0.2467,
      "num_input_tokens_seen": 8585392,
      "step": 4455
    },
    {
      "epoch": 0.727628680969084,
      "grad_norm": 0.009705748409032822,
      "learning_rate": 0.2908951518904045,
      "loss": 0.2401,
      "num_input_tokens_seen": 8594656,
      "step": 4460
    },
    {
      "epoch": 0.7284444081899013,
      "grad_norm": 0.00633176788687706,
      "learning_rate": 0.29087493111234963,
      "loss": 0.2457,
      "num_input_tokens_seen": 8605664,
      "step": 4465
    },
    {
      "epoch": 0.7292601354107187,
      "grad_norm": 0.009484470821917057,
      "learning_rate": 0.29085468860961,
      "loss": 0.2519,
      "num_input_tokens_seen": 8616608,
      "step": 4470
    },
    {
      "epoch": 0.730075862631536,
      "grad_norm": 0.005551857873797417,
      "learning_rate": 0.2908344243853073,
      "loss": 0.2182,
      "num_input_tokens_seen": 8626272,
      "step": 4475
    },
    {
      "epoch": 0.7308915898523534,
      "grad_norm": 0.006533833686262369,
      "learning_rate": 0.2908141384425666,
      "loss": 0.2311,
      "num_input_tokens_seen": 8636544,
      "step": 4480
    },
    {
      "epoch": 0.7317073170731707,
      "grad_norm": 0.0056177969090640545,
      "learning_rate": 0.2907938307845161,
      "loss": 0.2336,
      "num_input_tokens_seen": 8646176,
      "step": 4485
    },
    {
      "epoch": 0.7325230442939881,
      "grad_norm": 0.007714913226664066,
      "learning_rate": 0.2907735014142876,
      "loss": 0.2544,
      "num_input_tokens_seen": 8656016,
      "step": 4490
    },
    {
      "epoch": 0.7333387715148054,
      "grad_norm": 0.0065268222242593765,
      "learning_rate": 0.2907531503350161,
      "loss": 0.2265,
      "num_input_tokens_seen": 8665584,
      "step": 4495
    },
    {
      "epoch": 0.7341544987356228,
      "grad_norm": 0.007250432390719652,
      "learning_rate": 0.29073277754983995,
      "loss": 0.2289,
      "num_input_tokens_seen": 8674192,
      "step": 4500
    },
    {
      "epoch": 0.7349702259564401,
      "grad_norm": 0.008573725819587708,
      "learning_rate": 0.290712383061901,
      "loss": 0.2257,
      "num_input_tokens_seen": 8683696,
      "step": 4505
    },
    {
      "epoch": 0.7357859531772575,
      "grad_norm": 0.004696494434028864,
      "learning_rate": 0.2906919668743443,
      "loss": 0.2471,
      "num_input_tokens_seen": 8692272,
      "step": 4510
    },
    {
      "epoch": 0.7366016803980748,
      "grad_norm": 0.1076296716928482,
      "learning_rate": 0.29067152899031823,
      "loss": 0.2332,
      "num_input_tokens_seen": 8701792,
      "step": 4515
    },
    {
      "epoch": 0.7374174076188923,
      "grad_norm": 0.00863946694880724,
      "learning_rate": 0.2906510694129746,
      "loss": 0.2279,
      "num_input_tokens_seen": 8710160,
      "step": 4520
    },
    {
      "epoch": 0.7382331348397096,
      "grad_norm": 0.008271307684481144,
      "learning_rate": 0.2906305881454685,
      "loss": 0.2464,
      "num_input_tokens_seen": 8720960,
      "step": 4525
    },
    {
      "epoch": 0.739048862060527,
      "grad_norm": 0.010792246088385582,
      "learning_rate": 0.2906100851909585,
      "loss": 0.2284,
      "num_input_tokens_seen": 8731152,
      "step": 4530
    },
    {
      "epoch": 0.7398645892813444,
      "grad_norm": 0.2762099802494049,
      "learning_rate": 0.29058956055260626,
      "loss": 0.3896,
      "num_input_tokens_seen": 8741888,
      "step": 4535
    },
    {
      "epoch": 0.7406803165021617,
      "grad_norm": 0.02373616397380829,
      "learning_rate": 0.2905690142335771,
      "loss": 0.2817,
      "num_input_tokens_seen": 8751728,
      "step": 4540
    },
    {
      "epoch": 0.7414960437229791,
      "grad_norm": 0.012448581866919994,
      "learning_rate": 0.29054844623703946,
      "loss": 0.2456,
      "num_input_tokens_seen": 8760544,
      "step": 4545
    },
    {
      "epoch": 0.7423117709437964,
      "grad_norm": 0.02465963363647461,
      "learning_rate": 0.2905278565661651,
      "loss": 0.266,
      "num_input_tokens_seen": 8770928,
      "step": 4550
    },
    {
      "epoch": 0.7431274981646138,
      "grad_norm": 0.01657557114958763,
      "learning_rate": 0.2905072452241293,
      "loss": 0.248,
      "num_input_tokens_seen": 8780032,
      "step": 4555
    },
    {
      "epoch": 0.7439432253854311,
      "grad_norm": 0.020372465252876282,
      "learning_rate": 0.2904866122141106,
      "loss": 0.2462,
      "num_input_tokens_seen": 8790320,
      "step": 4560
    },
    {
      "epoch": 0.7447589526062485,
      "grad_norm": 0.040291137993335724,
      "learning_rate": 0.2904659575392908,
      "loss": 0.2659,
      "num_input_tokens_seen": 8799072,
      "step": 4565
    },
    {
      "epoch": 0.7455746798270658,
      "grad_norm": 0.01582503318786621,
      "learning_rate": 0.2904452812028551,
      "loss": 0.2192,
      "num_input_tokens_seen": 8807968,
      "step": 4570
    },
    {
      "epoch": 0.7463904070478832,
      "grad_norm": 0.0191954392939806,
      "learning_rate": 0.2904245832079922,
      "loss": 0.2894,
      "num_input_tokens_seen": 8817824,
      "step": 4575
    },
    {
      "epoch": 0.7472061342687005,
      "grad_norm": 0.022451482713222504,
      "learning_rate": 0.29040386355789377,
      "loss": 0.2482,
      "num_input_tokens_seen": 8827392,
      "step": 4580
    },
    {
      "epoch": 0.7480218614895179,
      "grad_norm": 0.03343096375465393,
      "learning_rate": 0.29038312225575524,
      "loss": 0.2602,
      "num_input_tokens_seen": 8836992,
      "step": 4585
    },
    {
      "epoch": 0.7488375887103352,
      "grad_norm": 0.01411908958107233,
      "learning_rate": 0.29036235930477505,
      "loss": 0.2673,
      "num_input_tokens_seen": 8846176,
      "step": 4590
    },
    {
      "epoch": 0.7496533159311526,
      "grad_norm": 0.014735998585820198,
      "learning_rate": 0.29034157470815514,
      "loss": 0.2963,
      "num_input_tokens_seen": 8857664,
      "step": 4595
    },
    {
      "epoch": 0.7504690431519699,
      "grad_norm": 0.011472364887595177,
      "learning_rate": 0.2903207684691008,
      "loss": 0.2399,
      "num_input_tokens_seen": 8868016,
      "step": 4600
    },
    {
      "epoch": 0.7504690431519699,
      "eval_loss": 0.23213130235671997,
      "eval_runtime": 68.1119,
      "eval_samples_per_second": 40.008,
      "eval_steps_per_second": 20.011,
      "num_input_tokens_seen": 8868016,
      "step": 4600
    },
    {
      "epoch": 0.7512847703727873,
      "grad_norm": 0.009271876886487007,
      "learning_rate": 0.29029994059082054,
      "loss": 0.1982,
      "num_input_tokens_seen": 8877904,
      "step": 4605
    },
    {
      "epoch": 0.7521004975936046,
      "grad_norm": 0.02332070656120777,
      "learning_rate": 0.2902790910765264,
      "loss": 0.2342,
      "num_input_tokens_seen": 8888288,
      "step": 4610
    },
    {
      "epoch": 0.7529162248144221,
      "grad_norm": 0.021017130464315414,
      "learning_rate": 0.29025821992943346,
      "loss": 0.2545,
      "num_input_tokens_seen": 8897376,
      "step": 4615
    },
    {
      "epoch": 0.7537319520352395,
      "grad_norm": 0.020870836451649666,
      "learning_rate": 0.29023732715276046,
      "loss": 0.1936,
      "num_input_tokens_seen": 8906240,
      "step": 4620
    },
    {
      "epoch": 0.7545476792560568,
      "grad_norm": 0.011365409940481186,
      "learning_rate": 0.2902164127497293,
      "loss": 0.3029,
      "num_input_tokens_seen": 8915712,
      "step": 4625
    },
    {
      "epoch": 0.7553634064768742,
      "grad_norm": 0.009317115880548954,
      "learning_rate": 0.2901954767235652,
      "loss": 0.2332,
      "num_input_tokens_seen": 8924240,
      "step": 4630
    },
    {
      "epoch": 0.7561791336976915,
      "grad_norm": 0.01446465402841568,
      "learning_rate": 0.2901745190774968,
      "loss": 0.2628,
      "num_input_tokens_seen": 8933488,
      "step": 4635
    },
    {
      "epoch": 0.7569948609185089,
      "grad_norm": 0.00790938176214695,
      "learning_rate": 0.290153539814756,
      "loss": 0.2164,
      "num_input_tokens_seen": 8943264,
      "step": 4640
    },
    {
      "epoch": 0.7578105881393262,
      "grad_norm": 0.011107610538601875,
      "learning_rate": 0.2901325389385781,
      "loss": 0.2858,
      "num_input_tokens_seen": 8952912,
      "step": 4645
    },
    {
      "epoch": 0.7586263153601436,
      "grad_norm": 0.01733863167464733,
      "learning_rate": 0.2901115164522016,
      "loss": 0.279,
      "num_input_tokens_seen": 8963248,
      "step": 4650
    },
    {
      "epoch": 0.7594420425809609,
      "grad_norm": 0.010876244865357876,
      "learning_rate": 0.29009047235886865,
      "loss": 0.236,
      "num_input_tokens_seen": 8972608,
      "step": 4655
    },
    {
      "epoch": 0.7602577698017783,
      "grad_norm": 0.01962200738489628,
      "learning_rate": 0.2900694066618243,
      "loss": 0.2812,
      "num_input_tokens_seen": 8982064,
      "step": 4660
    },
    {
      "epoch": 0.7610734970225956,
      "grad_norm": 0.013978120870888233,
      "learning_rate": 0.2900483193643172,
      "loss": 0.2449,
      "num_input_tokens_seen": 8992240,
      "step": 4665
    },
    {
      "epoch": 0.761889224243413,
      "grad_norm": 0.02866620384156704,
      "learning_rate": 0.29002721046959934,
      "loss": 0.2382,
      "num_input_tokens_seen": 9002784,
      "step": 4670
    },
    {
      "epoch": 0.7627049514642303,
      "grad_norm": 0.015387969091534615,
      "learning_rate": 0.29000607998092587,
      "loss": 0.2464,
      "num_input_tokens_seen": 9012240,
      "step": 4675
    },
    {
      "epoch": 0.7635206786850477,
      "grad_norm": 0.011861798353493214,
      "learning_rate": 0.2899849279015555,
      "loss": 0.2328,
      "num_input_tokens_seen": 9021520,
      "step": 4680
    },
    {
      "epoch": 0.764336405905865,
      "grad_norm": 0.012409216724336147,
      "learning_rate": 0.28996375423475007,
      "loss": 0.2668,
      "num_input_tokens_seen": 9030752,
      "step": 4685
    },
    {
      "epoch": 0.7651521331266824,
      "grad_norm": 0.013385165482759476,
      "learning_rate": 0.28994255898377486,
      "loss": 0.2512,
      "num_input_tokens_seen": 9039376,
      "step": 4690
    },
    {
      "epoch": 0.7659678603474998,
      "grad_norm": 0.010826660320162773,
      "learning_rate": 0.2899213421518984,
      "loss": 0.2092,
      "num_input_tokens_seen": 9049424,
      "step": 4695
    },
    {
      "epoch": 0.7667835875683171,
      "grad_norm": 0.019819283857941628,
      "learning_rate": 0.2899001037423926,
      "loss": 0.2597,
      "num_input_tokens_seen": 9060480,
      "step": 4700
    },
    {
      "epoch": 0.7675993147891346,
      "grad_norm": 0.00931676384061575,
      "learning_rate": 0.28987884375853273,
      "loss": 0.2702,
      "num_input_tokens_seen": 9070384,
      "step": 4705
    },
    {
      "epoch": 0.7684150420099519,
      "grad_norm": 0.007049329578876495,
      "learning_rate": 0.2898575622035974,
      "loss": 0.2085,
      "num_input_tokens_seen": 9080016,
      "step": 4710
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 0.009867728687822819,
      "learning_rate": 0.2898362590808683,
      "loss": 0.2487,
      "num_input_tokens_seen": 9089440,
      "step": 4715
    },
    {
      "epoch": 0.7700464964515866,
      "grad_norm": 0.0062139299698174,
      "learning_rate": 0.2898149343936308,
      "loss": 0.2181,
      "num_input_tokens_seen": 9098416,
      "step": 4720
    },
    {
      "epoch": 0.770862223672404,
      "grad_norm": 0.01232959795743227,
      "learning_rate": 0.2897935881451734,
      "loss": 0.2421,
      "num_input_tokens_seen": 9108000,
      "step": 4725
    },
    {
      "epoch": 0.7716779508932213,
      "grad_norm": 0.010737578384578228,
      "learning_rate": 0.28977222033878797,
      "loss": 0.2197,
      "num_input_tokens_seen": 9117920,
      "step": 4730
    },
    {
      "epoch": 0.7724936781140387,
      "grad_norm": 0.01576882041990757,
      "learning_rate": 0.28975083097776966,
      "loss": 0.2181,
      "num_input_tokens_seen": 9128704,
      "step": 4735
    },
    {
      "epoch": 0.773309405334856,
      "grad_norm": 0.01634388230741024,
      "learning_rate": 0.28972942006541696,
      "loss": 0.2604,
      "num_input_tokens_seen": 9137024,
      "step": 4740
    },
    {
      "epoch": 0.7741251325556734,
      "grad_norm": 0.011233117431402206,
      "learning_rate": 0.2897079876050318,
      "loss": 0.3005,
      "num_input_tokens_seen": 9146768,
      "step": 4745
    },
    {
      "epoch": 0.7749408597764907,
      "grad_norm": 0.010627947747707367,
      "learning_rate": 0.2896865335999192,
      "loss": 0.2296,
      "num_input_tokens_seen": 9155968,
      "step": 4750
    },
    {
      "epoch": 0.7757565869973081,
      "grad_norm": 0.010008633136749268,
      "learning_rate": 0.28966505805338777,
      "loss": 0.2433,
      "num_input_tokens_seen": 9166176,
      "step": 4755
    },
    {
      "epoch": 0.7765723142181254,
      "grad_norm": 0.023072820156812668,
      "learning_rate": 0.2896435609687492,
      "loss": 0.2333,
      "num_input_tokens_seen": 9175632,
      "step": 4760
    },
    {
      "epoch": 0.7773880414389428,
      "grad_norm": 0.024460311979055405,
      "learning_rate": 0.2896220423493187,
      "loss": 0.2556,
      "num_input_tokens_seen": 9186160,
      "step": 4765
    },
    {
      "epoch": 0.7782037686597602,
      "grad_norm": 0.008037022314965725,
      "learning_rate": 0.28960050219841466,
      "loss": 0.2264,
      "num_input_tokens_seen": 9196096,
      "step": 4770
    },
    {
      "epoch": 0.7790194958805775,
      "grad_norm": 0.01447114534676075,
      "learning_rate": 0.28957894051935884,
      "loss": 0.2567,
      "num_input_tokens_seen": 9204448,
      "step": 4775
    },
    {
      "epoch": 0.7798352231013949,
      "grad_norm": 0.008128161542117596,
      "learning_rate": 0.2895573573154764,
      "loss": 0.2433,
      "num_input_tokens_seen": 9214240,
      "step": 4780
    },
    {
      "epoch": 0.7806509503222122,
      "grad_norm": 0.008366372436285019,
      "learning_rate": 0.28953575259009556,
      "loss": 0.2907,
      "num_input_tokens_seen": 9224528,
      "step": 4785
    },
    {
      "epoch": 0.7814666775430296,
      "grad_norm": 0.008502130396664143,
      "learning_rate": 0.2895141263465482,
      "loss": 0.2456,
      "num_input_tokens_seen": 9234496,
      "step": 4790
    },
    {
      "epoch": 0.7822824047638469,
      "grad_norm": 0.010587797500193119,
      "learning_rate": 0.28949247858816934,
      "loss": 0.2339,
      "num_input_tokens_seen": 9244288,
      "step": 4795
    },
    {
      "epoch": 0.7830981319846644,
      "grad_norm": 0.005814172327518463,
      "learning_rate": 0.2894708093182973,
      "loss": 0.2409,
      "num_input_tokens_seen": 9254560,
      "step": 4800
    },
    {
      "epoch": 0.7830981319846644,
      "eval_loss": 0.23202386498451233,
      "eval_runtime": 67.9455,
      "eval_samples_per_second": 40.106,
      "eval_steps_per_second": 20.06,
      "num_input_tokens_seen": 9254560,
      "step": 4800
    },
    {
      "epoch": 0.7839138592054817,
      "grad_norm": 0.010515897534787655,
      "learning_rate": 0.2894491185402737,
      "loss": 0.2298,
      "num_input_tokens_seen": 9263248,
      "step": 4805
    },
    {
      "epoch": 0.7847295864262991,
      "grad_norm": 0.00991367269307375,
      "learning_rate": 0.2894274062574437,
      "loss": 0.2319,
      "num_input_tokens_seen": 9273568,
      "step": 4810
    },
    {
      "epoch": 0.7855453136471164,
      "grad_norm": 0.010768561623990536,
      "learning_rate": 0.2894056724731554,
      "loss": 0.2288,
      "num_input_tokens_seen": 9282384,
      "step": 4815
    },
    {
      "epoch": 0.7863610408679338,
      "grad_norm": 0.010235605761408806,
      "learning_rate": 0.28938391719076056,
      "loss": 0.2274,
      "num_input_tokens_seen": 9291408,
      "step": 4820
    },
    {
      "epoch": 0.7871767680887511,
      "grad_norm": 0.010034401901066303,
      "learning_rate": 0.28936214041361413,
      "loss": 0.2382,
      "num_input_tokens_seen": 9301408,
      "step": 4825
    },
    {
      "epoch": 0.7879924953095685,
      "grad_norm": 0.008769464679062366,
      "learning_rate": 0.2893403421450743,
      "loss": 0.2278,
      "num_input_tokens_seen": 9310272,
      "step": 4830
    },
    {
      "epoch": 0.7888082225303858,
      "grad_norm": 0.013412674888968468,
      "learning_rate": 0.2893185223885026,
      "loss": 0.2255,
      "num_input_tokens_seen": 9320112,
      "step": 4835
    },
    {
      "epoch": 0.7896239497512032,
      "grad_norm": 0.009692653082311153,
      "learning_rate": 0.289296681147264,
      "loss": 0.2437,
      "num_input_tokens_seen": 9329584,
      "step": 4840
    },
    {
      "epoch": 0.7904396769720206,
      "grad_norm": 0.010012593120336533,
      "learning_rate": 0.28927481842472663,
      "loss": 0.2029,
      "num_input_tokens_seen": 9338416,
      "step": 4845
    },
    {
      "epoch": 0.7912554041928379,
      "grad_norm": 0.021018195897340775,
      "learning_rate": 0.28925293422426207,
      "loss": 0.2834,
      "num_input_tokens_seen": 9348624,
      "step": 4850
    },
    {
      "epoch": 0.7920711314136553,
      "grad_norm": 0.012827047146856785,
      "learning_rate": 0.28923102854924504,
      "loss": 0.2616,
      "num_input_tokens_seen": 9356432,
      "step": 4855
    },
    {
      "epoch": 0.7928868586344726,
      "grad_norm": 0.015252654440701008,
      "learning_rate": 0.2892091014030537,
      "loss": 0.2668,
      "num_input_tokens_seen": 9364816,
      "step": 4860
    },
    {
      "epoch": 0.79370258585529,
      "grad_norm": 0.021581675857305527,
      "learning_rate": 0.2891871527890696,
      "loss": 0.2343,
      "num_input_tokens_seen": 9374880,
      "step": 4865
    },
    {
      "epoch": 0.7945183130761073,
      "grad_norm": 0.07074106484651566,
      "learning_rate": 0.2891651827106773,
      "loss": 0.2677,
      "num_input_tokens_seen": 9383904,
      "step": 4870
    },
    {
      "epoch": 0.7953340402969247,
      "grad_norm": 0.011008139699697495,
      "learning_rate": 0.2891431911712651,
      "loss": 0.1803,
      "num_input_tokens_seen": 9393504,
      "step": 4875
    },
    {
      "epoch": 0.796149767517742,
      "grad_norm": 0.02168882079422474,
      "learning_rate": 0.2891211781742241,
      "loss": 0.2053,
      "num_input_tokens_seen": 9402576,
      "step": 4880
    },
    {
      "epoch": 0.7969654947385594,
      "grad_norm": 0.009680898860096931,
      "learning_rate": 0.2890991437229492,
      "loss": 0.2531,
      "num_input_tokens_seen": 9412592,
      "step": 4885
    },
    {
      "epoch": 0.7977812219593767,
      "grad_norm": 0.008728723973035812,
      "learning_rate": 0.2890770878208383,
      "loss": 0.2369,
      "num_input_tokens_seen": 9422368,
      "step": 4890
    },
    {
      "epoch": 0.7985969491801942,
      "grad_norm": 0.007718402426689863,
      "learning_rate": 0.28905501047129273,
      "loss": 0.2424,
      "num_input_tokens_seen": 9432352,
      "step": 4895
    },
    {
      "epoch": 0.7994126764010115,
      "grad_norm": 0.007484762463718653,
      "learning_rate": 0.289032911677717,
      "loss": 0.2343,
      "num_input_tokens_seen": 9441056,
      "step": 4900
    },
    {
      "epoch": 0.8002284036218289,
      "grad_norm": 0.005572224501520395,
      "learning_rate": 0.28901079144351915,
      "loss": 0.242,
      "num_input_tokens_seen": 9451456,
      "step": 4905
    },
    {
      "epoch": 0.8010441308426463,
      "grad_norm": 0.005396664142608643,
      "learning_rate": 0.2889886497721103,
      "loss": 0.2489,
      "num_input_tokens_seen": 9460064,
      "step": 4910
    },
    {
      "epoch": 0.8018598580634636,
      "grad_norm": 0.006026825401932001,
      "learning_rate": 0.28896648666690505,
      "loss": 0.2586,
      "num_input_tokens_seen": 9469824,
      "step": 4915
    },
    {
      "epoch": 0.802675585284281,
      "grad_norm": 0.0054048276506364346,
      "learning_rate": 0.2889443021313212,
      "loss": 0.2197,
      "num_input_tokens_seen": 9477680,
      "step": 4920
    },
    {
      "epoch": 0.8034913125050983,
      "grad_norm": 0.012607118114829063,
      "learning_rate": 0.28892209616877984,
      "loss": 0.2498,
      "num_input_tokens_seen": 9486720,
      "step": 4925
    },
    {
      "epoch": 0.8043070397259157,
      "grad_norm": 0.011951031163334846,
      "learning_rate": 0.28889986878270546,
      "loss": 0.2554,
      "num_input_tokens_seen": 9496352,
      "step": 4930
    },
    {
      "epoch": 0.805122766946733,
      "grad_norm": 0.007387479301542044,
      "learning_rate": 0.28887761997652583,
      "loss": 0.2733,
      "num_input_tokens_seen": 9505488,
      "step": 4935
    },
    {
      "epoch": 0.8059384941675504,
      "grad_norm": 0.007849081419408321,
      "learning_rate": 0.2888553497536719,
      "loss": 0.2378,
      "num_input_tokens_seen": 9515264,
      "step": 4940
    },
    {
      "epoch": 0.8067542213883677,
      "grad_norm": 0.013189757242798805,
      "learning_rate": 0.2888330581175781,
      "loss": 0.2765,
      "num_input_tokens_seen": 9525760,
      "step": 4945
    },
    {
      "epoch": 0.8075699486091851,
      "grad_norm": 0.010142638348042965,
      "learning_rate": 0.28881074507168203,
      "loss": 0.2503,
      "num_input_tokens_seen": 9536320,
      "step": 4950
    },
    {
      "epoch": 0.8083856758300024,
      "grad_norm": 0.004942759871482849,
      "learning_rate": 0.2887884106194247,
      "loss": 0.2485,
      "num_input_tokens_seen": 9546672,
      "step": 4955
    },
    {
      "epoch": 0.8092014030508198,
      "grad_norm": 0.014632015489041805,
      "learning_rate": 0.28876605476425027,
      "loss": 0.2801,
      "num_input_tokens_seen": 9554544,
      "step": 4960
    },
    {
      "epoch": 0.8100171302716371,
      "grad_norm": 0.004701409954577684,
      "learning_rate": 0.2887436775096064,
      "loss": 0.2643,
      "num_input_tokens_seen": 9564416,
      "step": 4965
    },
    {
      "epoch": 0.8108328574924545,
      "grad_norm": 0.007458566688001156,
      "learning_rate": 0.2887212788589439,
      "loss": 0.2711,
      "num_input_tokens_seen": 9574768,
      "step": 4970
    },
    {
      "epoch": 0.8116485847132718,
      "grad_norm": 0.00703044980764389,
      "learning_rate": 0.2886988588157169,
      "loss": 0.2254,
      "num_input_tokens_seen": 9585856,
      "step": 4975
    },
    {
      "epoch": 0.8124643119340892,
      "grad_norm": 0.00816765334457159,
      "learning_rate": 0.28867641738338284,
      "loss": 0.2434,
      "num_input_tokens_seen": 9595440,
      "step": 4980
    },
    {
      "epoch": 0.8132800391549067,
      "grad_norm": 0.004471231251955032,
      "learning_rate": 0.2886539545654026,
      "loss": 0.2183,
      "num_input_tokens_seen": 9605424,
      "step": 4985
    },
    {
      "epoch": 0.814095766375724,
      "grad_norm": 0.004950959701091051,
      "learning_rate": 0.28863147036524006,
      "loss": 0.2596,
      "num_input_tokens_seen": 9615808,
      "step": 4990
    },
    {
      "epoch": 0.8149114935965414,
      "grad_norm": 0.003375396365299821,
      "learning_rate": 0.2886089647863626,
      "loss": 0.2321,
      "num_input_tokens_seen": 9624432,
      "step": 4995
    },
    {
      "epoch": 0.8157272208173587,
      "grad_norm": 0.006835006643086672,
      "learning_rate": 0.288586437832241,
      "loss": 0.2794,
      "num_input_tokens_seen": 9634544,
      "step": 5000
    },
    {
      "epoch": 0.8157272208173587,
      "eval_loss": 0.2501993477344513,
      "eval_runtime": 67.9415,
      "eval_samples_per_second": 40.108,
      "eval_steps_per_second": 20.061,
      "num_input_tokens_seen": 9634544,
      "step": 5000
    },
    {
      "epoch": 0.8165429480381761,
      "grad_norm": 0.008007697761058807,
      "learning_rate": 0.28856388950634904,
      "loss": 0.2745,
      "num_input_tokens_seen": 9644320,
      "step": 5005
    },
    {
      "epoch": 0.8173586752589934,
      "grad_norm": 0.009773042052984238,
      "learning_rate": 0.288541319812164,
      "loss": 0.2691,
      "num_input_tokens_seen": 9654656,
      "step": 5010
    },
    {
      "epoch": 0.8181744024798108,
      "grad_norm": 0.03306082263588905,
      "learning_rate": 0.2885187287531665,
      "loss": 0.2574,
      "num_input_tokens_seen": 9663280,
      "step": 5015
    },
    {
      "epoch": 0.8189901297006281,
      "grad_norm": 0.012114029377698898,
      "learning_rate": 0.2884961163328402,
      "loss": 0.4029,
      "num_input_tokens_seen": 9673904,
      "step": 5020
    },
    {
      "epoch": 0.8198058569214455,
      "grad_norm": 0.010625310242176056,
      "learning_rate": 0.28847348255467237,
      "loss": 0.2495,
      "num_input_tokens_seen": 9683392,
      "step": 5025
    },
    {
      "epoch": 0.8206215841422628,
      "grad_norm": 0.010384317487478256,
      "learning_rate": 0.28845082742215333,
      "loss": 0.3002,
      "num_input_tokens_seen": 9693008,
      "step": 5030
    },
    {
      "epoch": 0.8214373113630802,
      "grad_norm": 0.013179888017475605,
      "learning_rate": 0.2884281509387769,
      "loss": 0.2402,
      "num_input_tokens_seen": 9703376,
      "step": 5035
    },
    {
      "epoch": 0.8222530385838975,
      "grad_norm": 0.015603947453200817,
      "learning_rate": 0.2884054531080399,
      "loss": 0.2811,
      "num_input_tokens_seen": 9712624,
      "step": 5040
    },
    {
      "epoch": 0.8230687658047149,
      "grad_norm": 0.013565316796302795,
      "learning_rate": 0.28838273393344277,
      "loss": 0.2216,
      "num_input_tokens_seen": 9721680,
      "step": 5045
    },
    {
      "epoch": 0.8238844930255322,
      "grad_norm": 0.01199266780167818,
      "learning_rate": 0.288359993418489,
      "loss": 0.2388,
      "num_input_tokens_seen": 9731312,
      "step": 5050
    },
    {
      "epoch": 0.8247002202463496,
      "grad_norm": 0.006828212179243565,
      "learning_rate": 0.28833723156668556,
      "loss": 0.2443,
      "num_input_tokens_seen": 9741312,
      "step": 5055
    },
    {
      "epoch": 0.8255159474671669,
      "grad_norm": 0.004951545037329197,
      "learning_rate": 0.2883144483815425,
      "loss": 0.2588,
      "num_input_tokens_seen": 9750560,
      "step": 5060
    },
    {
      "epoch": 0.8263316746879843,
      "grad_norm": 0.010093114338815212,
      "learning_rate": 0.28829164386657335,
      "loss": 0.2381,
      "num_input_tokens_seen": 9758800,
      "step": 5065
    },
    {
      "epoch": 0.8271474019088016,
      "grad_norm": 0.005833684466779232,
      "learning_rate": 0.28826881802529486,
      "loss": 0.2293,
      "num_input_tokens_seen": 9767472,
      "step": 5070
    },
    {
      "epoch": 0.827963129129619,
      "grad_norm": 0.01804928667843342,
      "learning_rate": 0.28824597086122705,
      "loss": 0.2795,
      "num_input_tokens_seen": 9777056,
      "step": 5075
    },
    {
      "epoch": 0.8287788563504365,
      "grad_norm": 0.004708562511950731,
      "learning_rate": 0.28822310237789317,
      "loss": 0.2138,
      "num_input_tokens_seen": 9785888,
      "step": 5080
    },
    {
      "epoch": 0.8295945835712538,
      "grad_norm": 0.010399055667221546,
      "learning_rate": 0.2882002125788199,
      "loss": 0.2992,
      "num_input_tokens_seen": 9796080,
      "step": 5085
    },
    {
      "epoch": 0.8304103107920712,
      "grad_norm": 0.010181044228374958,
      "learning_rate": 0.2881773014675371,
      "loss": 0.2983,
      "num_input_tokens_seen": 9804512,
      "step": 5090
    },
    {
      "epoch": 0.8312260380128885,
      "grad_norm": 0.004381890874356031,
      "learning_rate": 0.288154369047578,
      "loss": 0.2357,
      "num_input_tokens_seen": 9813984,
      "step": 5095
    },
    {
      "epoch": 0.8320417652337059,
      "grad_norm": 0.0062635913491249084,
      "learning_rate": 0.28813141532247905,
      "loss": 0.2753,
      "num_input_tokens_seen": 9823504,
      "step": 5100
    },
    {
      "epoch": 0.8328574924545232,
      "grad_norm": 0.010494950227439404,
      "learning_rate": 0.28810844029578,
      "loss": 0.2351,
      "num_input_tokens_seen": 9833296,
      "step": 5105
    },
    {
      "epoch": 0.8336732196753406,
      "grad_norm": 0.007980323396623135,
      "learning_rate": 0.2880854439710238,
      "loss": 0.268,
      "num_input_tokens_seen": 9843360,
      "step": 5110
    },
    {
      "epoch": 0.8344889468961579,
      "grad_norm": 0.003422357141971588,
      "learning_rate": 0.28806242635175694,
      "loss": 0.2402,
      "num_input_tokens_seen": 9853264,
      "step": 5115
    },
    {
      "epoch": 0.8353046741169753,
      "grad_norm": 0.008859223686158657,
      "learning_rate": 0.2880393874415289,
      "loss": 0.3407,
      "num_input_tokens_seen": 9861952,
      "step": 5120
    },
    {
      "epoch": 0.8361204013377926,
      "grad_norm": 0.008217161521315575,
      "learning_rate": 0.2880163272438926,
      "loss": 0.2363,
      "num_input_tokens_seen": 9871424,
      "step": 5125
    },
    {
      "epoch": 0.83693612855861,
      "grad_norm": 0.01426495611667633,
      "learning_rate": 0.2879932457624042,
      "loss": 0.2696,
      "num_input_tokens_seen": 9881344,
      "step": 5130
    },
    {
      "epoch": 0.8377518557794273,
      "grad_norm": 0.007599536795169115,
      "learning_rate": 0.2879701430006232,
      "loss": 0.269,
      "num_input_tokens_seen": 9891104,
      "step": 5135
    },
    {
      "epoch": 0.8385675830002447,
      "grad_norm": 0.0035155904479324818,
      "learning_rate": 0.28794701896211233,
      "loss": 0.2479,
      "num_input_tokens_seen": 9900656,
      "step": 5140
    },
    {
      "epoch": 0.839383310221062,
      "grad_norm": 0.004759588278830051,
      "learning_rate": 0.28792387365043753,
      "loss": 0.2255,
      "num_input_tokens_seen": 9908944,
      "step": 5145
    },
    {
      "epoch": 0.8401990374418794,
      "grad_norm": 0.005559355486184359,
      "learning_rate": 0.28790070706916815,
      "loss": 0.2395,
      "num_input_tokens_seen": 9917344,
      "step": 5150
    },
    {
      "epoch": 0.8410147646626968,
      "grad_norm": 0.00726032629609108,
      "learning_rate": 0.2878775192218768,
      "loss": 0.2503,
      "num_input_tokens_seen": 9925616,
      "step": 5155
    },
    {
      "epoch": 0.8418304918835141,
      "grad_norm": 0.009188043884932995,
      "learning_rate": 0.2878543101121393,
      "loss": 0.2569,
      "num_input_tokens_seen": 9936288,
      "step": 5160
    },
    {
      "epoch": 0.8426462191043315,
      "grad_norm": 0.007070805877447128,
      "learning_rate": 0.28783107974353483,
      "loss": 0.2304,
      "num_input_tokens_seen": 9946224,
      "step": 5165
    },
    {
      "epoch": 0.8434619463251488,
      "grad_norm": 0.0035427797120064497,
      "learning_rate": 0.2878078281196457,
      "loss": 0.242,
      "num_input_tokens_seen": 9954896,
      "step": 5170
    },
    {
      "epoch": 0.8442776735459663,
      "grad_norm": 0.006212121341377497,
      "learning_rate": 0.28778455524405777,
      "loss": 0.2643,
      "num_input_tokens_seen": 9964960,
      "step": 5175
    },
    {
      "epoch": 0.8450934007667836,
      "grad_norm": 0.0045789373107254505,
      "learning_rate": 0.2877612611203598,
      "loss": 0.2473,
      "num_input_tokens_seen": 9975344,
      "step": 5180
    },
    {
      "epoch": 0.845909127987601,
      "grad_norm": 0.01227855309844017,
      "learning_rate": 0.28773794575214423,
      "loss": 0.2392,
      "num_input_tokens_seen": 9984624,
      "step": 5185
    },
    {
      "epoch": 0.8467248552084183,
      "grad_norm": 0.005861755460500717,
      "learning_rate": 0.28771460914300645,
      "loss": 0.3231,
      "num_input_tokens_seen": 9995008,
      "step": 5190
    },
    {
      "epoch": 0.8475405824292357,
      "grad_norm": 0.005537101067602634,
      "learning_rate": 0.2876912512965454,
      "loss": 0.2451,
      "num_input_tokens_seen": 10005200,
      "step": 5195
    },
    {
      "epoch": 0.848356309650053,
      "grad_norm": 0.005364083219319582,
      "learning_rate": 0.287667872216363,
      "loss": 0.2242,
      "num_input_tokens_seen": 10013984,
      "step": 5200
    },
    {
      "epoch": 0.848356309650053,
      "eval_loss": 0.24980692565441132,
      "eval_runtime": 67.9925,
      "eval_samples_per_second": 40.078,
      "eval_steps_per_second": 20.046,
      "num_input_tokens_seen": 10013984,
      "step": 5200
    },
    {
      "epoch": 0.8491720368708704,
      "grad_norm": 0.004841634538024664,
      "learning_rate": 0.2876444719060647,
      "loss": 0.2603,
      "num_input_tokens_seen": 10023648,
      "step": 5205
    },
    {
      "epoch": 0.8499877640916877,
      "grad_norm": 0.005180737469345331,
      "learning_rate": 0.287621050369259,
      "loss": 0.2211,
      "num_input_tokens_seen": 10031968,
      "step": 5210
    },
    {
      "epoch": 0.8508034913125051,
      "grad_norm": 0.004641360603272915,
      "learning_rate": 0.28759760760955794,
      "loss": 0.2328,
      "num_input_tokens_seen": 10041040,
      "step": 5215
    },
    {
      "epoch": 0.8516192185333225,
      "grad_norm": 0.008801034651696682,
      "learning_rate": 0.2875741436305766,
      "loss": 0.2712,
      "num_input_tokens_seen": 10051664,
      "step": 5220
    },
    {
      "epoch": 0.8524349457541398,
      "grad_norm": 0.003834398230537772,
      "learning_rate": 0.28755065843593347,
      "loss": 0.2486,
      "num_input_tokens_seen": 10060432,
      "step": 5225
    },
    {
      "epoch": 0.8532506729749572,
      "grad_norm": 0.003887468483299017,
      "learning_rate": 0.2875271520292502,
      "loss": 0.2353,
      "num_input_tokens_seen": 10070288,
      "step": 5230
    },
    {
      "epoch": 0.8540664001957745,
      "grad_norm": 0.004212388303130865,
      "learning_rate": 0.28750362441415184,
      "loss": 0.2391,
      "num_input_tokens_seen": 10079392,
      "step": 5235
    },
    {
      "epoch": 0.8548821274165919,
      "grad_norm": 0.005119230132550001,
      "learning_rate": 0.28748007559426664,
      "loss": 0.2357,
      "num_input_tokens_seen": 10088320,
      "step": 5240
    },
    {
      "epoch": 0.8556978546374092,
      "grad_norm": 0.004051294643431902,
      "learning_rate": 0.2874565055732261,
      "loss": 0.2342,
      "num_input_tokens_seen": 10097456,
      "step": 5245
    },
    {
      "epoch": 0.8565135818582266,
      "grad_norm": 0.007505667395889759,
      "learning_rate": 0.28743291435466495,
      "loss": 0.2325,
      "num_input_tokens_seen": 10108048,
      "step": 5250
    },
    {
      "epoch": 0.8573293090790439,
      "grad_norm": 0.004756520036607981,
      "learning_rate": 0.2874093019422214,
      "loss": 0.2347,
      "num_input_tokens_seen": 10116608,
      "step": 5255
    },
    {
      "epoch": 0.8581450362998613,
      "grad_norm": 0.008230448700487614,
      "learning_rate": 0.28738566833953666,
      "loss": 0.2645,
      "num_input_tokens_seen": 10125936,
      "step": 5260
    },
    {
      "epoch": 0.8589607635206787,
      "grad_norm": 0.005201885476708412,
      "learning_rate": 0.28736201355025537,
      "loss": 0.2105,
      "num_input_tokens_seen": 10135856,
      "step": 5265
    },
    {
      "epoch": 0.8597764907414961,
      "grad_norm": 0.009215373545885086,
      "learning_rate": 0.28733833757802535,
      "loss": 0.2483,
      "num_input_tokens_seen": 10145696,
      "step": 5270
    },
    {
      "epoch": 0.8605922179623134,
      "grad_norm": 0.008922301232814789,
      "learning_rate": 0.28731464042649785,
      "loss": 0.2308,
      "num_input_tokens_seen": 10153936,
      "step": 5275
    },
    {
      "epoch": 0.8614079451831308,
      "grad_norm": 0.007179032079875469,
      "learning_rate": 0.2872909220993271,
      "loss": 0.2337,
      "num_input_tokens_seen": 10164192,
      "step": 5280
    },
    {
      "epoch": 0.8622236724039481,
      "grad_norm": 0.004676359705626965,
      "learning_rate": 0.287267182600171,
      "loss": 0.2337,
      "num_input_tokens_seen": 10173968,
      "step": 5285
    },
    {
      "epoch": 0.8630393996247655,
      "grad_norm": 0.003858218202367425,
      "learning_rate": 0.2872434219326902,
      "loss": 0.2038,
      "num_input_tokens_seen": 10184160,
      "step": 5290
    },
    {
      "epoch": 0.8638551268455829,
      "grad_norm": 0.004960728343576193,
      "learning_rate": 0.28721964010054907,
      "loss": 0.2451,
      "num_input_tokens_seen": 10194352,
      "step": 5295
    },
    {
      "epoch": 0.8646708540664002,
      "grad_norm": 0.0037625788245350122,
      "learning_rate": 0.28719583710741503,
      "loss": 0.24,
      "num_input_tokens_seen": 10202464,
      "step": 5300
    },
    {
      "epoch": 0.8654865812872176,
      "grad_norm": 0.004231672268360853,
      "learning_rate": 0.28717201295695877,
      "loss": 0.2636,
      "num_input_tokens_seen": 10211696,
      "step": 5305
    },
    {
      "epoch": 0.8663023085080349,
      "grad_norm": 0.006779567338526249,
      "learning_rate": 0.28714816765285434,
      "loss": 0.2525,
      "num_input_tokens_seen": 10221264,
      "step": 5310
    },
    {
      "epoch": 0.8671180357288523,
      "grad_norm": 0.004784373100847006,
      "learning_rate": 0.28712430119877896,
      "loss": 0.2501,
      "num_input_tokens_seen": 10232208,
      "step": 5315
    },
    {
      "epoch": 0.8679337629496696,
      "grad_norm": 0.004157423973083496,
      "learning_rate": 0.28710041359841304,
      "loss": 0.2507,
      "num_input_tokens_seen": 10242272,
      "step": 5320
    },
    {
      "epoch": 0.868749490170487,
      "grad_norm": 0.005231261253356934,
      "learning_rate": 0.28707650485544056,
      "loss": 0.2421,
      "num_input_tokens_seen": 10252912,
      "step": 5325
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.010402094572782516,
      "learning_rate": 0.28705257497354836,
      "loss": 0.2443,
      "num_input_tokens_seen": 10261584,
      "step": 5330
    },
    {
      "epoch": 0.8703809446121217,
      "grad_norm": 0.003739170031622052,
      "learning_rate": 0.28702862395642675,
      "loss": 0.2261,
      "num_input_tokens_seen": 10271456,
      "step": 5335
    },
    {
      "epoch": 0.871196671832939,
      "grad_norm": 0.002865096554160118,
      "learning_rate": 0.28700465180776935,
      "loss": 0.2651,
      "num_input_tokens_seen": 10281296,
      "step": 5340
    },
    {
      "epoch": 0.8720123990537564,
      "grad_norm": 0.0036922465078532696,
      "learning_rate": 0.2869806585312729,
      "loss": 0.2337,
      "num_input_tokens_seen": 10291408,
      "step": 5345
    },
    {
      "epoch": 0.8728281262745737,
      "grad_norm": 0.010346932336688042,
      "learning_rate": 0.28695664413063754,
      "loss": 0.2558,
      "num_input_tokens_seen": 10301232,
      "step": 5350
    },
    {
      "epoch": 0.8736438534953911,
      "grad_norm": 0.005061840172857046,
      "learning_rate": 0.28693260860956654,
      "loss": 0.2389,
      "num_input_tokens_seen": 10310864,
      "step": 5355
    },
    {
      "epoch": 0.8744595807162086,
      "grad_norm": 0.006993912160396576,
      "learning_rate": 0.2869085519717665,
      "loss": 0.2268,
      "num_input_tokens_seen": 10321840,
      "step": 5360
    },
    {
      "epoch": 0.8752753079370259,
      "grad_norm": 0.002749860752373934,
      "learning_rate": 0.28688447422094726,
      "loss": 0.2459,
      "num_input_tokens_seen": 10332208,
      "step": 5365
    },
    {
      "epoch": 0.8760910351578433,
      "grad_norm": 0.005008810665458441,
      "learning_rate": 0.2868603753608219,
      "loss": 0.2435,
      "num_input_tokens_seen": 10342848,
      "step": 5370
    },
    {
      "epoch": 0.8769067623786606,
      "grad_norm": 0.006579785607755184,
      "learning_rate": 0.28683625539510665,
      "loss": 0.2545,
      "num_input_tokens_seen": 10352304,
      "step": 5375
    },
    {
      "epoch": 0.877722489599478,
      "grad_norm": 0.003920006565749645,
      "learning_rate": 0.28681211432752135,
      "loss": 0.2271,
      "num_input_tokens_seen": 10361536,
      "step": 5380
    },
    {
      "epoch": 0.8785382168202953,
      "grad_norm": 0.009443496353924274,
      "learning_rate": 0.2867879521617887,
      "loss": 0.2387,
      "num_input_tokens_seen": 10370320,
      "step": 5385
    },
    {
      "epoch": 0.8793539440411127,
      "grad_norm": 0.0024183164350688457,
      "learning_rate": 0.28676376890163485,
      "loss": 0.1803,
      "num_input_tokens_seen": 10380288,
      "step": 5390
    },
    {
      "epoch": 0.88016967126193,
      "grad_norm": 0.004646060522645712,
      "learning_rate": 0.2867395645507891,
      "loss": 0.1903,
      "num_input_tokens_seen": 10389024,
      "step": 5395
    },
    {
      "epoch": 0.8809853984827474,
      "grad_norm": 0.00576488021761179,
      "learning_rate": 0.2867153391129842,
      "loss": 0.2163,
      "num_input_tokens_seen": 10397792,
      "step": 5400
    },
    {
      "epoch": 0.8809853984827474,
      "eval_loss": 0.2634451985359192,
      "eval_runtime": 68.0196,
      "eval_samples_per_second": 40.062,
      "eval_steps_per_second": 20.038,
      "num_input_tokens_seen": 10397792,
      "step": 5400
    },
    {
      "epoch": 0.8818011257035647,
      "grad_norm": 0.006959003396332264,
      "learning_rate": 0.28669109259195585,
      "loss": 0.3158,
      "num_input_tokens_seen": 10407952,
      "step": 5405
    },
    {
      "epoch": 0.8826168529243821,
      "grad_norm": 0.0028567647095769644,
      "learning_rate": 0.2866668249914433,
      "loss": 0.2234,
      "num_input_tokens_seen": 10416096,
      "step": 5410
    },
    {
      "epoch": 0.8834325801451994,
      "grad_norm": 0.002394329523667693,
      "learning_rate": 0.2866425363151889,
      "loss": 0.2349,
      "num_input_tokens_seen": 10426176,
      "step": 5415
    },
    {
      "epoch": 0.8842483073660168,
      "grad_norm": 0.00442560575902462,
      "learning_rate": 0.2866182265669382,
      "loss": 0.2467,
      "num_input_tokens_seen": 10435264,
      "step": 5420
    },
    {
      "epoch": 0.8850640345868341,
      "grad_norm": 0.00673436326906085,
      "learning_rate": 0.28659389575044014,
      "loss": 0.2189,
      "num_input_tokens_seen": 10444448,
      "step": 5425
    },
    {
      "epoch": 0.8858797618076515,
      "grad_norm": 0.004034572746604681,
      "learning_rate": 0.28656954386944683,
      "loss": 0.2362,
      "num_input_tokens_seen": 10455152,
      "step": 5430
    },
    {
      "epoch": 0.8866954890284688,
      "grad_norm": 0.00219884910620749,
      "learning_rate": 0.28654517092771353,
      "loss": 0.1978,
      "num_input_tokens_seen": 10464848,
      "step": 5435
    },
    {
      "epoch": 0.8875112162492862,
      "grad_norm": 0.006581251043826342,
      "learning_rate": 0.286520776928999,
      "loss": 0.2735,
      "num_input_tokens_seen": 10474704,
      "step": 5440
    },
    {
      "epoch": 0.8883269434701035,
      "grad_norm": 0.005520228762179613,
      "learning_rate": 0.286496361877065,
      "loss": 0.2257,
      "num_input_tokens_seen": 10484400,
      "step": 5445
    },
    {
      "epoch": 0.889142670690921,
      "grad_norm": 0.002975247334688902,
      "learning_rate": 0.28647192577567676,
      "loss": 0.2355,
      "num_input_tokens_seen": 10494480,
      "step": 5450
    },
    {
      "epoch": 0.8899583979117384,
      "grad_norm": 0.004317048471421003,
      "learning_rate": 0.28644746862860254,
      "loss": 0.2551,
      "num_input_tokens_seen": 10503328,
      "step": 5455
    },
    {
      "epoch": 0.8907741251325557,
      "grad_norm": 0.00373673252761364,
      "learning_rate": 0.2864229904396139,
      "loss": 0.2472,
      "num_input_tokens_seen": 10513008,
      "step": 5460
    },
    {
      "epoch": 0.8915898523533731,
      "grad_norm": 0.0030146653298288584,
      "learning_rate": 0.28639849121248573,
      "loss": 0.2189,
      "num_input_tokens_seen": 10521344,
      "step": 5465
    },
    {
      "epoch": 0.8924055795741904,
      "grad_norm": 0.002964756917208433,
      "learning_rate": 0.28637397095099615,
      "loss": 0.2084,
      "num_input_tokens_seen": 10530736,
      "step": 5470
    },
    {
      "epoch": 0.8932213067950078,
      "grad_norm": 0.0035969470627605915,
      "learning_rate": 0.28634942965892646,
      "loss": 0.2435,
      "num_input_tokens_seen": 10540992,
      "step": 5475
    },
    {
      "epoch": 0.8940370340158251,
      "grad_norm": 0.0027276775799691677,
      "learning_rate": 0.28632486734006124,
      "loss": 0.2221,
      "num_input_tokens_seen": 10551120,
      "step": 5480
    },
    {
      "epoch": 0.8948527612366425,
      "grad_norm": 0.003442558227106929,
      "learning_rate": 0.28630028399818835,
      "loss": 0.209,
      "num_input_tokens_seen": 10560752,
      "step": 5485
    },
    {
      "epoch": 0.8956684884574598,
      "grad_norm": 0.005234915297478437,
      "learning_rate": 0.2862756796370987,
      "loss": 0.226,
      "num_input_tokens_seen": 10570176,
      "step": 5490
    },
    {
      "epoch": 0.8964842156782772,
      "grad_norm": 0.0031102586071938276,
      "learning_rate": 0.2862510542605868,
      "loss": 0.2629,
      "num_input_tokens_seen": 10580992,
      "step": 5495
    },
    {
      "epoch": 0.8972999428990945,
      "grad_norm": 0.006703409366309643,
      "learning_rate": 0.2862264078724501,
      "loss": 0.2195,
      "num_input_tokens_seen": 10591296,
      "step": 5500
    },
    {
      "epoch": 0.8981156701199119,
      "grad_norm": 0.0023773331195116043,
      "learning_rate": 0.28620174047648933,
      "loss": 0.2299,
      "num_input_tokens_seen": 10600960,
      "step": 5505
    },
    {
      "epoch": 0.8989313973407292,
      "grad_norm": 0.0024247239343822002,
      "learning_rate": 0.2861770520765086,
      "loss": 0.2149,
      "num_input_tokens_seen": 10610704,
      "step": 5510
    },
    {
      "epoch": 0.8997471245615466,
      "grad_norm": 0.0020720225293189287,
      "learning_rate": 0.2861523426763151,
      "loss": 0.2194,
      "num_input_tokens_seen": 10620224,
      "step": 5515
    },
    {
      "epoch": 0.900562851782364,
      "grad_norm": 0.0017880212981253862,
      "learning_rate": 0.2861276122797194,
      "loss": 0.2375,
      "num_input_tokens_seen": 10629248,
      "step": 5520
    },
    {
      "epoch": 0.9013785790031813,
      "grad_norm": 0.0030946338083595037,
      "learning_rate": 0.28610286089053516,
      "loss": 0.2888,
      "num_input_tokens_seen": 10638400,
      "step": 5525
    },
    {
      "epoch": 0.9021943062239987,
      "grad_norm": 0.0032247542403638363,
      "learning_rate": 0.28607808851257943,
      "loss": 0.2581,
      "num_input_tokens_seen": 10648176,
      "step": 5530
    },
    {
      "epoch": 0.903010033444816,
      "grad_norm": 0.002292066579684615,
      "learning_rate": 0.28605329514967237,
      "loss": 0.2417,
      "num_input_tokens_seen": 10656848,
      "step": 5535
    },
    {
      "epoch": 0.9038257606656334,
      "grad_norm": 0.006433773785829544,
      "learning_rate": 0.2860284808056374,
      "loss": 0.2224,
      "num_input_tokens_seen": 10666944,
      "step": 5540
    },
    {
      "epoch": 0.9046414878864508,
      "grad_norm": 0.002137107076123357,
      "learning_rate": 0.28600364548430135,
      "loss": 0.2018,
      "num_input_tokens_seen": 10676720,
      "step": 5545
    },
    {
      "epoch": 0.9054572151072682,
      "grad_norm": 0.0036367017310112715,
      "learning_rate": 0.28597878918949393,
      "loss": 0.3225,
      "num_input_tokens_seen": 10686896,
      "step": 5550
    },
    {
      "epoch": 0.9062729423280855,
      "grad_norm": 0.003799343015998602,
      "learning_rate": 0.2859539119250485,
      "loss": 0.2023,
      "num_input_tokens_seen": 10696560,
      "step": 5555
    },
    {
      "epoch": 0.9070886695489029,
      "grad_norm": 0.0020030345767736435,
      "learning_rate": 0.2859290136948013,
      "loss": 0.2527,
      "num_input_tokens_seen": 10707296,
      "step": 5560
    },
    {
      "epoch": 0.9079043967697202,
      "grad_norm": 0.0034251746255904436,
      "learning_rate": 0.28590409450259197,
      "loss": 0.2421,
      "num_input_tokens_seen": 10715664,
      "step": 5565
    },
    {
      "epoch": 0.9087201239905376,
      "grad_norm": 0.004732694011181593,
      "learning_rate": 0.28587915435226346,
      "loss": 0.2372,
      "num_input_tokens_seen": 10725648,
      "step": 5570
    },
    {
      "epoch": 0.9095358512113549,
      "grad_norm": 0.003487583016976714,
      "learning_rate": 0.2858541932476617,
      "loss": 0.2435,
      "num_input_tokens_seen": 10736192,
      "step": 5575
    },
    {
      "epoch": 0.9103515784321723,
      "grad_norm": 0.0037759828846901655,
      "learning_rate": 0.2858292111926361,
      "loss": 0.2227,
      "num_input_tokens_seen": 10746176,
      "step": 5580
    },
    {
      "epoch": 0.9111673056529896,
      "grad_norm": 0.0074996682815253735,
      "learning_rate": 0.28580420819103924,
      "loss": 0.232,
      "num_input_tokens_seen": 10755200,
      "step": 5585
    },
    {
      "epoch": 0.911983032873807,
      "grad_norm": 0.0021359454840421677,
      "learning_rate": 0.2857791842467269,
      "loss": 0.2417,
      "num_input_tokens_seen": 10764816,
      "step": 5590
    },
    {
      "epoch": 0.9127987600946244,
      "grad_norm": 0.0021763660479336977,
      "learning_rate": 0.2857541393635579,
      "loss": 0.2486,
      "num_input_tokens_seen": 10774272,
      "step": 5595
    },
    {
      "epoch": 0.9136144873154417,
      "grad_norm": 0.002760158386081457,
      "learning_rate": 0.2857290735453948,
      "loss": 0.2432,
      "num_input_tokens_seen": 10784512,
      "step": 5600
    },
    {
      "epoch": 0.9136144873154417,
      "eval_loss": 0.2413567453622818,
      "eval_runtime": 68.0207,
      "eval_samples_per_second": 40.061,
      "eval_steps_per_second": 20.038,
      "num_input_tokens_seen": 10784512,
      "step": 5600
    },
    {
      "epoch": 0.9144302145362591,
      "grad_norm": 0.0023677123244851828,
      "learning_rate": 0.28570398679610276,
      "loss": 0.2518,
      "num_input_tokens_seen": 10794256,
      "step": 5605
    },
    {
      "epoch": 0.9152459417570764,
      "grad_norm": 0.003238425822928548,
      "learning_rate": 0.2856788791195506,
      "loss": 0.245,
      "num_input_tokens_seen": 10804112,
      "step": 5610
    },
    {
      "epoch": 0.9160616689778938,
      "grad_norm": 0.007297350559383631,
      "learning_rate": 0.28565375051961023,
      "loss": 0.2321,
      "num_input_tokens_seen": 10812752,
      "step": 5615
    },
    {
      "epoch": 0.9168773961987111,
      "grad_norm": 0.003917547408491373,
      "learning_rate": 0.28562860100015686,
      "loss": 0.2037,
      "num_input_tokens_seen": 10821552,
      "step": 5620
    },
    {
      "epoch": 0.9176931234195285,
      "grad_norm": 0.004224450793117285,
      "learning_rate": 0.2856034305650687,
      "loss": 0.2087,
      "num_input_tokens_seen": 10830704,
      "step": 5625
    },
    {
      "epoch": 0.9185088506403458,
      "grad_norm": 0.010193431749939919,
      "learning_rate": 0.28557823921822756,
      "loss": 0.2442,
      "num_input_tokens_seen": 10840816,
      "step": 5630
    },
    {
      "epoch": 0.9193245778611632,
      "grad_norm": 0.005025677382946014,
      "learning_rate": 0.2855530269635181,
      "loss": 0.2387,
      "num_input_tokens_seen": 10849472,
      "step": 5635
    },
    {
      "epoch": 0.9201403050819806,
      "grad_norm": 0.0020785005763173103,
      "learning_rate": 0.2855277938048284,
      "loss": 0.2177,
      "num_input_tokens_seen": 10859296,
      "step": 5640
    },
    {
      "epoch": 0.920956032302798,
      "grad_norm": 0.0034441319294273853,
      "learning_rate": 0.2855025397460498,
      "loss": 0.2284,
      "num_input_tokens_seen": 10868448,
      "step": 5645
    },
    {
      "epoch": 0.9217717595236153,
      "grad_norm": 0.003087430726736784,
      "learning_rate": 0.28547726479107666,
      "loss": 0.2615,
      "num_input_tokens_seen": 10877312,
      "step": 5650
    },
    {
      "epoch": 0.9225874867444327,
      "grad_norm": 0.0038645020686089993,
      "learning_rate": 0.2854519689438068,
      "loss": 0.217,
      "num_input_tokens_seen": 10886912,
      "step": 5655
    },
    {
      "epoch": 0.92340321396525,
      "grad_norm": 0.0022906395606696606,
      "learning_rate": 0.2854266522081412,
      "loss": 0.2251,
      "num_input_tokens_seen": 10896032,
      "step": 5660
    },
    {
      "epoch": 0.9242189411860674,
      "grad_norm": 0.0033363625407218933,
      "learning_rate": 0.28540131458798385,
      "loss": 0.235,
      "num_input_tokens_seen": 10906400,
      "step": 5665
    },
    {
      "epoch": 0.9250346684068848,
      "grad_norm": 0.002886895090341568,
      "learning_rate": 0.28537595608724226,
      "loss": 0.2258,
      "num_input_tokens_seen": 10915552,
      "step": 5670
    },
    {
      "epoch": 0.9258503956277021,
      "grad_norm": 0.00682784104719758,
      "learning_rate": 0.28535057670982705,
      "loss": 0.2279,
      "num_input_tokens_seen": 10924720,
      "step": 5675
    },
    {
      "epoch": 0.9266661228485195,
      "grad_norm": 0.00692589720711112,
      "learning_rate": 0.285325176459652,
      "loss": 0.2389,
      "num_input_tokens_seen": 10934768,
      "step": 5680
    },
    {
      "epoch": 0.9274818500693368,
      "grad_norm": 0.003647887846454978,
      "learning_rate": 0.28529975534063406,
      "loss": 0.2153,
      "num_input_tokens_seen": 10944336,
      "step": 5685
    },
    {
      "epoch": 0.9282975772901542,
      "grad_norm": 0.0061363717541098595,
      "learning_rate": 0.2852743133566936,
      "loss": 0.2512,
      "num_input_tokens_seen": 10954480,
      "step": 5690
    },
    {
      "epoch": 0.9291133045109715,
      "grad_norm": 0.005970308557152748,
      "learning_rate": 0.2852488505117541,
      "loss": 0.2292,
      "num_input_tokens_seen": 10964432,
      "step": 5695
    },
    {
      "epoch": 0.9299290317317889,
      "grad_norm": 0.0037161146756261587,
      "learning_rate": 0.28522336680974214,
      "loss": 0.239,
      "num_input_tokens_seen": 10973728,
      "step": 5700
    },
    {
      "epoch": 0.9307447589526062,
      "grad_norm": 0.0033658291213214397,
      "learning_rate": 0.2851978622545877,
      "loss": 0.2507,
      "num_input_tokens_seen": 10983376,
      "step": 5705
    },
    {
      "epoch": 0.9315604861734236,
      "grad_norm": 0.003743629204109311,
      "learning_rate": 0.285172336850224,
      "loss": 0.2117,
      "num_input_tokens_seen": 10993472,
      "step": 5710
    },
    {
      "epoch": 0.9323762133942409,
      "grad_norm": 0.002715751063078642,
      "learning_rate": 0.2851467906005871,
      "loss": 0.2222,
      "num_input_tokens_seen": 11004240,
      "step": 5715
    },
    {
      "epoch": 0.9331919406150583,
      "grad_norm": 0.005166356451809406,
      "learning_rate": 0.28512122350961683,
      "loss": 0.226,
      "num_input_tokens_seen": 11013168,
      "step": 5720
    },
    {
      "epoch": 0.9340076678358756,
      "grad_norm": 0.0031169874127954245,
      "learning_rate": 0.2850956355812559,
      "loss": 0.2594,
      "num_input_tokens_seen": 11022832,
      "step": 5725
    },
    {
      "epoch": 0.9348233950566931,
      "grad_norm": 0.0025907792150974274,
      "learning_rate": 0.28507002681945015,
      "loss": 0.2246,
      "num_input_tokens_seen": 11031376,
      "step": 5730
    },
    {
      "epoch": 0.9356391222775104,
      "grad_norm": 0.002007456496357918,
      "learning_rate": 0.28504439722814895,
      "loss": 0.2223,
      "num_input_tokens_seen": 11041152,
      "step": 5735
    },
    {
      "epoch": 0.9364548494983278,
      "grad_norm": 0.006230440456420183,
      "learning_rate": 0.28501874681130457,
      "loss": 0.246,
      "num_input_tokens_seen": 11050688,
      "step": 5740
    },
    {
      "epoch": 0.9372705767191452,
      "grad_norm": 0.004267892800271511,
      "learning_rate": 0.2849930755728727,
      "loss": 0.2408,
      "num_input_tokens_seen": 11060400,
      "step": 5745
    },
    {
      "epoch": 0.9380863039399625,
      "grad_norm": 0.006867346353828907,
      "learning_rate": 0.28496738351681217,
      "loss": 0.2914,
      "num_input_tokens_seen": 11070480,
      "step": 5750
    },
    {
      "epoch": 0.9389020311607799,
      "grad_norm": 0.0025287740863859653,
      "learning_rate": 0.284941670647085,
      "loss": 0.2368,
      "num_input_tokens_seen": 11079824,
      "step": 5755
    },
    {
      "epoch": 0.9397177583815972,
      "grad_norm": 0.0024632259737700224,
      "learning_rate": 0.2849159369676563,
      "loss": 0.2166,
      "num_input_tokens_seen": 11090752,
      "step": 5760
    },
    {
      "epoch": 0.9405334856024146,
      "grad_norm": 0.004623197950422764,
      "learning_rate": 0.2848901824824948,
      "loss": 0.2005,
      "num_input_tokens_seen": 11100608,
      "step": 5765
    },
    {
      "epoch": 0.9413492128232319,
      "grad_norm": 0.005961703136563301,
      "learning_rate": 0.284864407195572,
      "loss": 0.2436,
      "num_input_tokens_seen": 11110416,
      "step": 5770
    },
    {
      "epoch": 0.9421649400440493,
      "grad_norm": 0.0024814517237246037,
      "learning_rate": 0.28483861111086284,
      "loss": 0.2213,
      "num_input_tokens_seen": 11120256,
      "step": 5775
    },
    {
      "epoch": 0.9429806672648666,
      "grad_norm": 0.002504809992387891,
      "learning_rate": 0.2848127942323453,
      "loss": 0.2294,
      "num_input_tokens_seen": 11128672,
      "step": 5780
    },
    {
      "epoch": 0.943796394485684,
      "grad_norm": 0.0037417917046695948,
      "learning_rate": 0.2847869565640007,
      "loss": 0.2317,
      "num_input_tokens_seen": 11137248,
      "step": 5785
    },
    {
      "epoch": 0.9446121217065013,
      "grad_norm": 0.0031998504418879747,
      "learning_rate": 0.2847610981098136,
      "loss": 0.2329,
      "num_input_tokens_seen": 11146640,
      "step": 5790
    },
    {
      "epoch": 0.9454278489273187,
      "grad_norm": 0.0020457885693758726,
      "learning_rate": 0.2847352188737716,
      "loss": 0.2409,
      "num_input_tokens_seen": 11155664,
      "step": 5795
    },
    {
      "epoch": 0.946243576148136,
      "grad_norm": 0.004663546569645405,
      "learning_rate": 0.2847093188598658,
      "loss": 0.2406,
      "num_input_tokens_seen": 11165168,
      "step": 5800
    },
    {
      "epoch": 0.946243576148136,
      "eval_loss": 0.2271076738834381,
      "eval_runtime": 67.9673,
      "eval_samples_per_second": 40.093,
      "eval_steps_per_second": 20.054,
      "num_input_tokens_seen": 11165168,
      "step": 5800
    },
    {
      "epoch": 0.9470593033689534,
      "grad_norm": 0.0033537843264639378,
      "learning_rate": 0.28468339807209003,
      "loss": 0.2273,
      "num_input_tokens_seen": 11175088,
      "step": 5805
    },
    {
      "epoch": 0.9478750305897707,
      "grad_norm": 0.005453765392303467,
      "learning_rate": 0.2846574565144418,
      "loss": 0.2316,
      "num_input_tokens_seen": 11184752,
      "step": 5810
    },
    {
      "epoch": 0.9486907578105881,
      "grad_norm": 0.002491951687261462,
      "learning_rate": 0.28463149419092154,
      "loss": 0.2384,
      "num_input_tokens_seen": 11194624,
      "step": 5815
    },
    {
      "epoch": 0.9495064850314054,
      "grad_norm": 0.0024468174669891596,
      "learning_rate": 0.284605511105533,
      "loss": 0.2169,
      "num_input_tokens_seen": 11204992,
      "step": 5820
    },
    {
      "epoch": 0.9503222122522229,
      "grad_norm": 0.00293921772390604,
      "learning_rate": 0.28457950726228315,
      "loss": 0.258,
      "num_input_tokens_seen": 11215728,
      "step": 5825
    },
    {
      "epoch": 0.9511379394730403,
      "grad_norm": 0.003095361404120922,
      "learning_rate": 0.28455348266518193,
      "loss": 0.2494,
      "num_input_tokens_seen": 11226064,
      "step": 5830
    },
    {
      "epoch": 0.9519536666938576,
      "grad_norm": 0.0027347211726009846,
      "learning_rate": 0.28452743731824287,
      "loss": 0.2393,
      "num_input_tokens_seen": 11235504,
      "step": 5835
    },
    {
      "epoch": 0.952769393914675,
      "grad_norm": 0.0062113492749631405,
      "learning_rate": 0.28450137122548236,
      "loss": 0.2374,
      "num_input_tokens_seen": 11244624,
      "step": 5840
    },
    {
      "epoch": 0.9535851211354923,
      "grad_norm": 0.0021474251989275217,
      "learning_rate": 0.2844752843909201,
      "loss": 0.2306,
      "num_input_tokens_seen": 11254416,
      "step": 5845
    },
    {
      "epoch": 0.9544008483563097,
      "grad_norm": 0.002592541044577956,
      "learning_rate": 0.28444917681857923,
      "loss": 0.2273,
      "num_input_tokens_seen": 11263408,
      "step": 5850
    },
    {
      "epoch": 0.955216575577127,
      "grad_norm": 0.0022328554186969995,
      "learning_rate": 0.28442304851248557,
      "loss": 0.2313,
      "num_input_tokens_seen": 11272560,
      "step": 5855
    },
    {
      "epoch": 0.9560323027979444,
      "grad_norm": 0.0024685573298484087,
      "learning_rate": 0.2843968994766686,
      "loss": 0.2268,
      "num_input_tokens_seen": 11282704,
      "step": 5860
    },
    {
      "epoch": 0.9568480300187617,
      "grad_norm": 0.005181280896067619,
      "learning_rate": 0.28437072971516075,
      "loss": 0.2363,
      "num_input_tokens_seen": 11291712,
      "step": 5865
    },
    {
      "epoch": 0.9576637572395791,
      "grad_norm": 0.0034681763499975204,
      "learning_rate": 0.2843445392319979,
      "loss": 0.2435,
      "num_input_tokens_seen": 11301424,
      "step": 5870
    },
    {
      "epoch": 0.9584794844603964,
      "grad_norm": 0.002080300822854042,
      "learning_rate": 0.28431832803121865,
      "loss": 0.2221,
      "num_input_tokens_seen": 11310368,
      "step": 5875
    },
    {
      "epoch": 0.9592952116812138,
      "grad_norm": 0.004765456076711416,
      "learning_rate": 0.28429209611686534,
      "loss": 0.2405,
      "num_input_tokens_seen": 11320064,
      "step": 5880
    },
    {
      "epoch": 0.9601109389020311,
      "grad_norm": 0.0039916555397212505,
      "learning_rate": 0.28426584349298323,
      "loss": 0.236,
      "num_input_tokens_seen": 11329904,
      "step": 5885
    },
    {
      "epoch": 0.9609266661228485,
      "grad_norm": 0.015312276780605316,
      "learning_rate": 0.2842395701636207,
      "loss": 0.24,
      "num_input_tokens_seen": 11338848,
      "step": 5890
    },
    {
      "epoch": 0.9617423933436658,
      "grad_norm": 0.008055792190134525,
      "learning_rate": 0.28421327613282954,
      "loss": 0.2367,
      "num_input_tokens_seen": 11348112,
      "step": 5895
    },
    {
      "epoch": 0.9625581205644832,
      "grad_norm": 0.0028313708025962114,
      "learning_rate": 0.28418696140466454,
      "loss": 0.2376,
      "num_input_tokens_seen": 11356880,
      "step": 5900
    },
    {
      "epoch": 0.9633738477853006,
      "grad_norm": 0.0021750631276518106,
      "learning_rate": 0.2841606259831838,
      "loss": 0.2352,
      "num_input_tokens_seen": 11366512,
      "step": 5905
    },
    {
      "epoch": 0.9641895750061179,
      "grad_norm": 0.0025015966966748238,
      "learning_rate": 0.2841342698724486,
      "loss": 0.2386,
      "num_input_tokens_seen": 11375856,
      "step": 5910
    },
    {
      "epoch": 0.9650053022269353,
      "grad_norm": 0.003139458829537034,
      "learning_rate": 0.28410789307652334,
      "loss": 0.2324,
      "num_input_tokens_seen": 11386208,
      "step": 5915
    },
    {
      "epoch": 0.9658210294477527,
      "grad_norm": 0.003349294187501073,
      "learning_rate": 0.2840814955994756,
      "loss": 0.2397,
      "num_input_tokens_seen": 11396224,
      "step": 5920
    },
    {
      "epoch": 0.9666367566685701,
      "grad_norm": 0.003682085545733571,
      "learning_rate": 0.2840550774453763,
      "loss": 0.2292,
      "num_input_tokens_seen": 11405904,
      "step": 5925
    },
    {
      "epoch": 0.9674524838893874,
      "grad_norm": 0.002809351310133934,
      "learning_rate": 0.28402863861829947,
      "loss": 0.2332,
      "num_input_tokens_seen": 11415120,
      "step": 5930
    },
    {
      "epoch": 0.9682682111102048,
      "grad_norm": 0.003085819771513343,
      "learning_rate": 0.2840021791223222,
      "loss": 0.242,
      "num_input_tokens_seen": 11425648,
      "step": 5935
    },
    {
      "epoch": 0.9690839383310221,
      "grad_norm": 0.003844847669824958,
      "learning_rate": 0.2839756989615249,
      "loss": 0.238,
      "num_input_tokens_seen": 11436576,
      "step": 5940
    },
    {
      "epoch": 0.9698996655518395,
      "grad_norm": 0.003858511336147785,
      "learning_rate": 0.28394919813999125,
      "loss": 0.2286,
      "num_input_tokens_seen": 11446672,
      "step": 5945
    },
    {
      "epoch": 0.9707153927726568,
      "grad_norm": 0.007721780799329281,
      "learning_rate": 0.28392267666180787,
      "loss": 0.2401,
      "num_input_tokens_seen": 11456480,
      "step": 5950
    },
    {
      "epoch": 0.9715311199934742,
      "grad_norm": 0.006008715834468603,
      "learning_rate": 0.2838961345310648,
      "loss": 0.2334,
      "num_input_tokens_seen": 11465872,
      "step": 5955
    },
    {
      "epoch": 0.9723468472142915,
      "grad_norm": 0.00864906795322895,
      "learning_rate": 0.2838695717518552,
      "loss": 0.2504,
      "num_input_tokens_seen": 11476112,
      "step": 5960
    },
    {
      "epoch": 0.9731625744351089,
      "grad_norm": 0.004616532474756241,
      "learning_rate": 0.28384298832827526,
      "loss": 0.2402,
      "num_input_tokens_seen": 11486816,
      "step": 5965
    },
    {
      "epoch": 0.9739783016559262,
      "grad_norm": 0.007343799341470003,
      "learning_rate": 0.28381638426442457,
      "loss": 0.2305,
      "num_input_tokens_seen": 11496400,
      "step": 5970
    },
    {
      "epoch": 0.9747940288767436,
      "grad_norm": 0.003849360626190901,
      "learning_rate": 0.2837897595644057,
      "loss": 0.2451,
      "num_input_tokens_seen": 11505376,
      "step": 5975
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 0.007477073464542627,
      "learning_rate": 0.28376311423232475,
      "loss": 0.2316,
      "num_input_tokens_seen": 11515472,
      "step": 5980
    },
    {
      "epoch": 0.9764254833183783,
      "grad_norm": 0.0030819412786513567,
      "learning_rate": 0.2837364482722905,
      "loss": 0.2289,
      "num_input_tokens_seen": 11524208,
      "step": 5985
    },
    {
      "epoch": 0.9772412105391957,
      "grad_norm": 0.0031713047064840794,
      "learning_rate": 0.28370976168841533,
      "loss": 0.2644,
      "num_input_tokens_seen": 11534176,
      "step": 5990
    },
    {
      "epoch": 0.978056937760013,
      "grad_norm": 0.0053544798865914345,
      "learning_rate": 0.2836830544848146,
      "loss": 0.2347,
      "num_input_tokens_seen": 11544592,
      "step": 5995
    },
    {
      "epoch": 0.9788726649808304,
      "grad_norm": 0.007725570350885391,
      "learning_rate": 0.2836563266656069,
      "loss": 0.2439,
      "num_input_tokens_seen": 11553056,
      "step": 6000
    },
    {
      "epoch": 0.9788726649808304,
      "eval_loss": 0.2323310226202011,
      "eval_runtime": 68.0356,
      "eval_samples_per_second": 40.053,
      "eval_steps_per_second": 20.034,
      "num_input_tokens_seen": 11553056,
      "step": 6000
    },
    {
      "epoch": 0.9796883922016477,
      "grad_norm": 0.0059147002175450325,
      "learning_rate": 0.283629578234914,
      "loss": 0.2273,
      "num_input_tokens_seen": 11562896,
      "step": 6005
    },
    {
      "epoch": 0.9805041194224652,
      "grad_norm": 0.0057913679629564285,
      "learning_rate": 0.2836028091968608,
      "loss": 0.2444,
      "num_input_tokens_seen": 11572016,
      "step": 6010
    },
    {
      "epoch": 0.9813198466432825,
      "grad_norm": 0.0028066530358046293,
      "learning_rate": 0.28357601955557554,
      "loss": 0.2244,
      "num_input_tokens_seen": 11580560,
      "step": 6015
    },
    {
      "epoch": 0.9821355738640999,
      "grad_norm": 0.0058667403645813465,
      "learning_rate": 0.2835492093151894,
      "loss": 0.312,
      "num_input_tokens_seen": 11590672,
      "step": 6020
    },
    {
      "epoch": 0.9829513010849172,
      "grad_norm": 0.0030964871402829885,
      "learning_rate": 0.2835223784798369,
      "loss": 0.2499,
      "num_input_tokens_seen": 11600848,
      "step": 6025
    },
    {
      "epoch": 0.9837670283057346,
      "grad_norm": 0.005105667281895876,
      "learning_rate": 0.2834955270536557,
      "loss": 0.2506,
      "num_input_tokens_seen": 11611232,
      "step": 6030
    },
    {
      "epoch": 0.984582755526552,
      "grad_norm": 0.007872977294027805,
      "learning_rate": 0.2834686550407866,
      "loss": 0.2763,
      "num_input_tokens_seen": 11620832,
      "step": 6035
    },
    {
      "epoch": 0.9853984827473693,
      "grad_norm": 0.002502478426322341,
      "learning_rate": 0.28344176244537367,
      "loss": 0.2424,
      "num_input_tokens_seen": 11630464,
      "step": 6040
    },
    {
      "epoch": 0.9862142099681867,
      "grad_norm": 0.002616854850202799,
      "learning_rate": 0.28341484927156396,
      "loss": 0.2352,
      "num_input_tokens_seen": 11639632,
      "step": 6045
    },
    {
      "epoch": 0.987029937189004,
      "grad_norm": 0.0017952065682038665,
      "learning_rate": 0.28338791552350795,
      "loss": 0.2217,
      "num_input_tokens_seen": 11648144,
      "step": 6050
    },
    {
      "epoch": 0.9878456644098214,
      "grad_norm": 0.004732069559395313,
      "learning_rate": 0.28336096120535914,
      "loss": 0.286,
      "num_input_tokens_seen": 11658688,
      "step": 6055
    },
    {
      "epoch": 0.9886613916306387,
      "grad_norm": 0.003096875734627247,
      "learning_rate": 0.2833339863212741,
      "loss": 0.2315,
      "num_input_tokens_seen": 11668480,
      "step": 6060
    },
    {
      "epoch": 0.9894771188514561,
      "grad_norm": 0.006848010700196028,
      "learning_rate": 0.28330699087541283,
      "loss": 0.2457,
      "num_input_tokens_seen": 11678576,
      "step": 6065
    },
    {
      "epoch": 0.9902928460722734,
      "grad_norm": 0.0017538213869556785,
      "learning_rate": 0.2832799748719384,
      "loss": 0.2387,
      "num_input_tokens_seen": 11688480,
      "step": 6070
    },
    {
      "epoch": 0.9911085732930908,
      "grad_norm": 0.0028003316838294268,
      "learning_rate": 0.28325293831501686,
      "loss": 0.2263,
      "num_input_tokens_seen": 11698128,
      "step": 6075
    },
    {
      "epoch": 0.9919243005139081,
      "grad_norm": 0.001708903699181974,
      "learning_rate": 0.2832258812088177,
      "loss": 0.2419,
      "num_input_tokens_seen": 11708000,
      "step": 6080
    },
    {
      "epoch": 0.9927400277347255,
      "grad_norm": 0.003649214282631874,
      "learning_rate": 0.2831988035575134,
      "loss": 0.2299,
      "num_input_tokens_seen": 11716352,
      "step": 6085
    },
    {
      "epoch": 0.9935557549555428,
      "grad_norm": 0.0078019495122134686,
      "learning_rate": 0.28317170536527975,
      "loss": 0.2545,
      "num_input_tokens_seen": 11725632,
      "step": 6090
    },
    {
      "epoch": 0.9943714821763602,
      "grad_norm": 0.0029123001731932163,
      "learning_rate": 0.2831445866362956,
      "loss": 0.2006,
      "num_input_tokens_seen": 11736448,
      "step": 6095
    },
    {
      "epoch": 0.9951872093971775,
      "grad_norm": 0.004095390904694796,
      "learning_rate": 0.2831174473747429,
      "loss": 0.2123,
      "num_input_tokens_seen": 11745712,
      "step": 6100
    },
    {
      "epoch": 0.996002936617995,
      "grad_norm": 0.002200579736381769,
      "learning_rate": 0.2830902875848071,
      "loss": 0.2474,
      "num_input_tokens_seen": 11754512,
      "step": 6105
    },
    {
      "epoch": 0.9968186638388123,
      "grad_norm": 0.011766140349209309,
      "learning_rate": 0.28306310727067635,
      "loss": 0.2319,
      "num_input_tokens_seen": 11763808,
      "step": 6110
    },
    {
      "epoch": 0.9976343910596297,
      "grad_norm": 0.0036345093976706266,
      "learning_rate": 0.2830359064365423,
      "loss": 0.223,
      "num_input_tokens_seen": 11773632,
      "step": 6115
    },
    {
      "epoch": 0.998450118280447,
      "grad_norm": 0.005624611862003803,
      "learning_rate": 0.28300868508659965,
      "loss": 0.2557,
      "num_input_tokens_seen": 11783840,
      "step": 6120
    },
    {
      "epoch": 0.9992658455012644,
      "grad_norm": 0.0023894545156508684,
      "learning_rate": 0.28298144322504626,
      "loss": 0.2452,
      "num_input_tokens_seen": 11794208,
      "step": 6125
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.0067185224033892155,
      "learning_rate": 0.2829541808560832,
      "loss": 0.2265,
      "num_input_tokens_seen": 11803328,
      "step": 6130
    },
    {
      "epoch": 1.0008157272208174,
      "grad_norm": 0.003944866359233856,
      "learning_rate": 0.2829268979839146,
      "loss": 0.2548,
      "num_input_tokens_seen": 11813488,
      "step": 6135
    },
    {
      "epoch": 1.0016314544416347,
      "grad_norm": 0.004661101382225752,
      "learning_rate": 0.2828995946127479,
      "loss": 0.2446,
      "num_input_tokens_seen": 11823680,
      "step": 6140
    },
    {
      "epoch": 1.002447181662452,
      "grad_norm": 0.0014994704397395253,
      "learning_rate": 0.2828722707467936,
      "loss": 0.2178,
      "num_input_tokens_seen": 11833184,
      "step": 6145
    },
    {
      "epoch": 1.0032629088832694,
      "grad_norm": 0.0024824582505971193,
      "learning_rate": 0.2828449263902653,
      "loss": 0.2397,
      "num_input_tokens_seen": 11843792,
      "step": 6150
    },
    {
      "epoch": 1.0040786361040868,
      "grad_norm": 0.004567562136799097,
      "learning_rate": 0.28281756154738,
      "loss": 0.2456,
      "num_input_tokens_seen": 11853344,
      "step": 6155
    },
    {
      "epoch": 1.0048943633249041,
      "grad_norm": 0.0032871549483388662,
      "learning_rate": 0.28279017622235764,
      "loss": 0.2365,
      "num_input_tokens_seen": 11862416,
      "step": 6160
    },
    {
      "epoch": 1.0057100905457215,
      "grad_norm": 0.0041265180334448814,
      "learning_rate": 0.28276277041942127,
      "loss": 0.2326,
      "num_input_tokens_seen": 11872544,
      "step": 6165
    },
    {
      "epoch": 1.0065258177665388,
      "grad_norm": 0.002686036517843604,
      "learning_rate": 0.2827353441427974,
      "loss": 0.2183,
      "num_input_tokens_seen": 11882896,
      "step": 6170
    },
    {
      "epoch": 1.0073415449873562,
      "grad_norm": 0.005088973790407181,
      "learning_rate": 0.2827078973967153,
      "loss": 0.2628,
      "num_input_tokens_seen": 11892512,
      "step": 6175
    },
    {
      "epoch": 1.0081572722081735,
      "grad_norm": 0.0018431944772601128,
      "learning_rate": 0.2826804301854078,
      "loss": 0.2124,
      "num_input_tokens_seen": 11902592,
      "step": 6180
    },
    {
      "epoch": 1.0089729994289909,
      "grad_norm": 0.002508684992790222,
      "learning_rate": 0.2826529425131105,
      "loss": 0.2353,
      "num_input_tokens_seen": 11912752,
      "step": 6185
    },
    {
      "epoch": 1.0097887266498082,
      "grad_norm": 0.00627397233620286,
      "learning_rate": 0.2826254343840625,
      "loss": 0.2466,
      "num_input_tokens_seen": 11922368,
      "step": 6190
    },
    {
      "epoch": 1.0106044538706256,
      "grad_norm": 0.0038354473654180765,
      "learning_rate": 0.2825979058025059,
      "loss": 0.2319,
      "num_input_tokens_seen": 11930176,
      "step": 6195
    },
    {
      "epoch": 1.011420181091443,
      "grad_norm": 0.0052454788237810135,
      "learning_rate": 0.2825703567726858,
      "loss": 0.2439,
      "num_input_tokens_seen": 11940352,
      "step": 6200
    },
    {
      "epoch": 1.011420181091443,
      "eval_loss": 0.22833667695522308,
      "eval_runtime": 68.0329,
      "eval_samples_per_second": 40.054,
      "eval_steps_per_second": 20.034,
      "num_input_tokens_seen": 11940352,
      "step": 6200
    },
    {
      "epoch": 1.0122359083122603,
      "grad_norm": 0.004702163860201836,
      "learning_rate": 0.2825427872988508,
      "loss": 0.2491,
      "num_input_tokens_seen": 11951168,
      "step": 6205
    },
    {
      "epoch": 1.0130516355330776,
      "grad_norm": 0.0020901889074593782,
      "learning_rate": 0.28251519738525227,
      "loss": 0.2176,
      "num_input_tokens_seen": 11960720,
      "step": 6210
    },
    {
      "epoch": 1.013867362753895,
      "grad_norm": 0.0028184170369058847,
      "learning_rate": 0.28248758703614507,
      "loss": 0.232,
      "num_input_tokens_seen": 11971488,
      "step": 6215
    },
    {
      "epoch": 1.0146830899747123,
      "grad_norm": 0.0023822037037461996,
      "learning_rate": 0.28245995625578696,
      "loss": 0.209,
      "num_input_tokens_seen": 11981648,
      "step": 6220
    },
    {
      "epoch": 1.01549881719553,
      "grad_norm": 0.005676956847310066,
      "learning_rate": 0.282432305048439,
      "loss": 0.2183,
      "num_input_tokens_seen": 11989584,
      "step": 6225
    },
    {
      "epoch": 1.0163145444163473,
      "grad_norm": 0.003353028791025281,
      "learning_rate": 0.28240463341836536,
      "loss": 0.2313,
      "num_input_tokens_seen": 11999600,
      "step": 6230
    },
    {
      "epoch": 1.0171302716371646,
      "grad_norm": 0.0024410055484622717,
      "learning_rate": 0.2823769413698334,
      "loss": 0.2225,
      "num_input_tokens_seen": 12008752,
      "step": 6235
    },
    {
      "epoch": 1.017945998857982,
      "grad_norm": 0.0042627183720469475,
      "learning_rate": 0.2823492289071135,
      "loss": 0.2641,
      "num_input_tokens_seen": 12018128,
      "step": 6240
    },
    {
      "epoch": 1.0187617260787993,
      "grad_norm": 0.0030879955738782883,
      "learning_rate": 0.2823214960344793,
      "loss": 0.2468,
      "num_input_tokens_seen": 12028368,
      "step": 6245
    },
    {
      "epoch": 1.0195774532996167,
      "grad_norm": 0.001796187600120902,
      "learning_rate": 0.28229374275620756,
      "loss": 0.2424,
      "num_input_tokens_seen": 12038336,
      "step": 6250
    },
    {
      "epoch": 1.020393180520434,
      "grad_norm": 0.00307590258307755,
      "learning_rate": 0.28226596907657814,
      "loss": 0.2328,
      "num_input_tokens_seen": 12047808,
      "step": 6255
    },
    {
      "epoch": 1.0212089077412514,
      "grad_norm": 0.0025615147314965725,
      "learning_rate": 0.28223817499987414,
      "loss": 0.2308,
      "num_input_tokens_seen": 12057680,
      "step": 6260
    },
    {
      "epoch": 1.0220246349620687,
      "grad_norm": 0.0053976052440702915,
      "learning_rate": 0.2822103605303818,
      "loss": 0.2374,
      "num_input_tokens_seen": 12067088,
      "step": 6265
    },
    {
      "epoch": 1.022840362182886,
      "grad_norm": 0.0019977071788161993,
      "learning_rate": 0.2821825256723903,
      "loss": 0.2289,
      "num_input_tokens_seen": 12075392,
      "step": 6270
    },
    {
      "epoch": 1.0236560894037035,
      "grad_norm": 0.0047242348082363605,
      "learning_rate": 0.2821546704301923,
      "loss": 0.2522,
      "num_input_tokens_seen": 12084752,
      "step": 6275
    },
    {
      "epoch": 1.0244718166245208,
      "grad_norm": 0.0031259767711162567,
      "learning_rate": 0.2821267948080834,
      "loss": 0.2401,
      "num_input_tokens_seen": 12094448,
      "step": 6280
    },
    {
      "epoch": 1.0252875438453382,
      "grad_norm": 0.00792744942009449,
      "learning_rate": 0.28209889881036226,
      "loss": 0.272,
      "num_input_tokens_seen": 12105504,
      "step": 6285
    },
    {
      "epoch": 1.0261032710661555,
      "grad_norm": 0.005411970429122448,
      "learning_rate": 0.28207098244133094,
      "loss": 0.2405,
      "num_input_tokens_seen": 12114992,
      "step": 6290
    },
    {
      "epoch": 1.0269189982869729,
      "grad_norm": 0.003779832273721695,
      "learning_rate": 0.2820430457052943,
      "loss": 0.2323,
      "num_input_tokens_seen": 12125424,
      "step": 6295
    },
    {
      "epoch": 1.0277347255077902,
      "grad_norm": 0.003275296650826931,
      "learning_rate": 0.28201508860656077,
      "loss": 0.2401,
      "num_input_tokens_seen": 12134704,
      "step": 6300
    },
    {
      "epoch": 1.0285504527286076,
      "grad_norm": 0.0018992612604051828,
      "learning_rate": 0.2819871111494415,
      "loss": 0.2308,
      "num_input_tokens_seen": 12144320,
      "step": 6305
    },
    {
      "epoch": 1.029366179949425,
      "grad_norm": 0.002590994583442807,
      "learning_rate": 0.28195911333825113,
      "loss": 0.2203,
      "num_input_tokens_seen": 12153680,
      "step": 6310
    },
    {
      "epoch": 1.0301819071702423,
      "grad_norm": 0.0022935529705137014,
      "learning_rate": 0.28193109517730713,
      "loss": 0.2505,
      "num_input_tokens_seen": 12164656,
      "step": 6315
    },
    {
      "epoch": 1.0309976343910596,
      "grad_norm": 0.002159976167604327,
      "learning_rate": 0.2819030566709303,
      "loss": 0.2057,
      "num_input_tokens_seen": 12174416,
      "step": 6320
    },
    {
      "epoch": 1.031813361611877,
      "grad_norm": 0.0026839652564376593,
      "learning_rate": 0.2818749978234445,
      "loss": 0.2193,
      "num_input_tokens_seen": 12184384,
      "step": 6325
    },
    {
      "epoch": 1.0326290888326943,
      "grad_norm": 0.0024927849881350994,
      "learning_rate": 0.2818469186391768,
      "loss": 0.2352,
      "num_input_tokens_seen": 12194688,
      "step": 6330
    },
    {
      "epoch": 1.0334448160535117,
      "grad_norm": 0.003823400242254138,
      "learning_rate": 0.28181881912245743,
      "loss": 0.2515,
      "num_input_tokens_seen": 12204032,
      "step": 6335
    },
    {
      "epoch": 1.034260543274329,
      "grad_norm": 0.0070544658228755,
      "learning_rate": 0.2817906992776195,
      "loss": 0.2362,
      "num_input_tokens_seen": 12213728,
      "step": 6340
    },
    {
      "epoch": 1.0350762704951464,
      "grad_norm": 0.00523315230384469,
      "learning_rate": 0.28176255910899967,
      "loss": 0.2347,
      "num_input_tokens_seen": 12223888,
      "step": 6345
    },
    {
      "epoch": 1.0358919977159637,
      "grad_norm": 0.002454828005284071,
      "learning_rate": 0.2817343986209373,
      "loss": 0.2227,
      "num_input_tokens_seen": 12234544,
      "step": 6350
    },
    {
      "epoch": 1.036707724936781,
      "grad_norm": 0.002971520647406578,
      "learning_rate": 0.2817062178177753,
      "loss": 0.2208,
      "num_input_tokens_seen": 12244800,
      "step": 6355
    },
    {
      "epoch": 1.0375234521575984,
      "grad_norm": 0.004880780819803476,
      "learning_rate": 0.2816780167038593,
      "loss": 0.2127,
      "num_input_tokens_seen": 12254240,
      "step": 6360
    },
    {
      "epoch": 1.0383391793784158,
      "grad_norm": 0.002254561521112919,
      "learning_rate": 0.28164979528353834,
      "loss": 0.185,
      "num_input_tokens_seen": 12264800,
      "step": 6365
    },
    {
      "epoch": 1.0391549065992332,
      "grad_norm": 0.0030577117577195168,
      "learning_rate": 0.28162155356116453,
      "loss": 0.1928,
      "num_input_tokens_seen": 12274816,
      "step": 6370
    },
    {
      "epoch": 1.0399706338200505,
      "grad_norm": 0.0019832479301840067,
      "learning_rate": 0.28159329154109314,
      "loss": 0.2778,
      "num_input_tokens_seen": 12283904,
      "step": 6375
    },
    {
      "epoch": 1.0407863610408679,
      "grad_norm": 0.002695804927498102,
      "learning_rate": 0.28156500922768246,
      "loss": 0.3465,
      "num_input_tokens_seen": 12293088,
      "step": 6380
    },
    {
      "epoch": 1.0416020882616852,
      "grad_norm": 0.002346494933590293,
      "learning_rate": 0.28153670662529406,
      "loss": 0.2289,
      "num_input_tokens_seen": 12302448,
      "step": 6385
    },
    {
      "epoch": 1.0424178154825026,
      "grad_norm": 0.00209188275039196,
      "learning_rate": 0.28150838373829246,
      "loss": 0.2503,
      "num_input_tokens_seen": 12311984,
      "step": 6390
    },
    {
      "epoch": 1.04323354270332,
      "grad_norm": 0.0031161170918494463,
      "learning_rate": 0.2814800405710455,
      "loss": 0.2362,
      "num_input_tokens_seen": 12323232,
      "step": 6395
    },
    {
      "epoch": 1.0440492699241373,
      "grad_norm": 0.003089185105636716,
      "learning_rate": 0.2814516771279239,
      "loss": 0.2551,
      "num_input_tokens_seen": 12331920,
      "step": 6400
    },
    {
      "epoch": 1.0440492699241373,
      "eval_loss": 0.2284741848707199,
      "eval_runtime": 68.1059,
      "eval_samples_per_second": 40.011,
      "eval_steps_per_second": 20.013,
      "num_input_tokens_seen": 12331920,
      "step": 6400
    },
    {
      "epoch": 1.0448649971449546,
      "grad_norm": 0.0039943503215909,
      "learning_rate": 0.28142329341330186,
      "loss": 0.2281,
      "num_input_tokens_seen": 12341824,
      "step": 6405
    },
    {
      "epoch": 1.0456807243657722,
      "grad_norm": 0.0029581652488559484,
      "learning_rate": 0.2813948894315564,
      "loss": 0.2395,
      "num_input_tokens_seen": 12352224,
      "step": 6410
    },
    {
      "epoch": 1.0464964515865895,
      "grad_norm": 0.0028045119252055883,
      "learning_rate": 0.2813664651870677,
      "loss": 0.2201,
      "num_input_tokens_seen": 12360544,
      "step": 6415
    },
    {
      "epoch": 1.047312178807407,
      "grad_norm": 0.0031537858303636312,
      "learning_rate": 0.28133802068421926,
      "loss": 0.2517,
      "num_input_tokens_seen": 12370592,
      "step": 6420
    },
    {
      "epoch": 1.0481279060282243,
      "grad_norm": 0.004047194495797157,
      "learning_rate": 0.28130955592739754,
      "loss": 0.2073,
      "num_input_tokens_seen": 12380912,
      "step": 6425
    },
    {
      "epoch": 1.0489436332490416,
      "grad_norm": 0.0019247356103733182,
      "learning_rate": 0.2812810709209922,
      "loss": 0.262,
      "num_input_tokens_seen": 12390432,
      "step": 6430
    },
    {
      "epoch": 1.049759360469859,
      "grad_norm": 0.0036770340520888567,
      "learning_rate": 0.2812525656693959,
      "loss": 0.2716,
      "num_input_tokens_seen": 12400240,
      "step": 6435
    },
    {
      "epoch": 1.0505750876906763,
      "grad_norm": 0.002290488686412573,
      "learning_rate": 0.28122404017700453,
      "loss": 0.2325,
      "num_input_tokens_seen": 12409024,
      "step": 6440
    },
    {
      "epoch": 1.0513908149114937,
      "grad_norm": 0.003572594141587615,
      "learning_rate": 0.2811954944482171,
      "loss": 0.2003,
      "num_input_tokens_seen": 12419792,
      "step": 6445
    },
    {
      "epoch": 1.052206542132311,
      "grad_norm": 0.00333880796097219,
      "learning_rate": 0.2811669284874358,
      "loss": 0.2621,
      "num_input_tokens_seen": 12429600,
      "step": 6450
    },
    {
      "epoch": 1.0530222693531284,
      "grad_norm": 0.001918243826366961,
      "learning_rate": 0.2811383422990657,
      "loss": 0.2162,
      "num_input_tokens_seen": 12438064,
      "step": 6455
    },
    {
      "epoch": 1.0538379965739457,
      "grad_norm": 0.0019354472169652581,
      "learning_rate": 0.2811097358875152,
      "loss": 0.202,
      "num_input_tokens_seen": 12449232,
      "step": 6460
    },
    {
      "epoch": 1.054653723794763,
      "grad_norm": 0.002348786685615778,
      "learning_rate": 0.2810811092571959,
      "loss": 0.231,
      "num_input_tokens_seen": 12458448,
      "step": 6465
    },
    {
      "epoch": 1.0554694510155804,
      "grad_norm": 0.0020970511250197887,
      "learning_rate": 0.28105246241252224,
      "loss": 0.2254,
      "num_input_tokens_seen": 12468144,
      "step": 6470
    },
    {
      "epoch": 1.0562851782363978,
      "grad_norm": 0.002253622282296419,
      "learning_rate": 0.28102379535791194,
      "loss": 0.2137,
      "num_input_tokens_seen": 12477936,
      "step": 6475
    },
    {
      "epoch": 1.0571009054572151,
      "grad_norm": 0.0023521706461906433,
      "learning_rate": 0.2809951080977859,
      "loss": 0.2719,
      "num_input_tokens_seen": 12487888,
      "step": 6480
    },
    {
      "epoch": 1.0579166326780325,
      "grad_norm": 0.002730242209509015,
      "learning_rate": 0.28096640063656797,
      "loss": 0.2282,
      "num_input_tokens_seen": 12498176,
      "step": 6485
    },
    {
      "epoch": 1.0587323598988498,
      "grad_norm": 0.006099619437009096,
      "learning_rate": 0.2809376729786852,
      "loss": 0.2358,
      "num_input_tokens_seen": 12508560,
      "step": 6490
    },
    {
      "epoch": 1.0595480871196672,
      "grad_norm": 0.0041848174296319485,
      "learning_rate": 0.28090892512856785,
      "loss": 0.2094,
      "num_input_tokens_seen": 12517632,
      "step": 6495
    },
    {
      "epoch": 1.0603638143404845,
      "grad_norm": 0.0013809758238494396,
      "learning_rate": 0.2808801570906491,
      "loss": 0.2265,
      "num_input_tokens_seen": 12526112,
      "step": 6500
    },
    {
      "epoch": 1.061179541561302,
      "grad_norm": 0.002555985003709793,
      "learning_rate": 0.2808513688693654,
      "loss": 0.216,
      "num_input_tokens_seen": 12536544,
      "step": 6505
    },
    {
      "epoch": 1.0619952687821193,
      "grad_norm": 0.0037983458023518324,
      "learning_rate": 0.28082256046915627,
      "loss": 0.2764,
      "num_input_tokens_seen": 12546784,
      "step": 6510
    },
    {
      "epoch": 1.0628109960029366,
      "grad_norm": 0.0031800216529518366,
      "learning_rate": 0.28079373189446427,
      "loss": 0.2388,
      "num_input_tokens_seen": 12557072,
      "step": 6515
    },
    {
      "epoch": 1.063626723223754,
      "grad_norm": 0.0035818738397210836,
      "learning_rate": 0.28076488314973513,
      "loss": 0.2277,
      "num_input_tokens_seen": 12566176,
      "step": 6520
    },
    {
      "epoch": 1.0644424504445713,
      "grad_norm": 0.004811420571058989,
      "learning_rate": 0.28073601423941774,
      "loss": 0.2093,
      "num_input_tokens_seen": 12575632,
      "step": 6525
    },
    {
      "epoch": 1.0652581776653887,
      "grad_norm": 0.0033974754624068737,
      "learning_rate": 0.28070712516796403,
      "loss": 0.2561,
      "num_input_tokens_seen": 12586000,
      "step": 6530
    },
    {
      "epoch": 1.066073904886206,
      "grad_norm": 0.00410070177167654,
      "learning_rate": 0.28067821593982906,
      "loss": 0.2367,
      "num_input_tokens_seen": 12596704,
      "step": 6535
    },
    {
      "epoch": 1.0668896321070234,
      "grad_norm": 0.0020285840146243572,
      "learning_rate": 0.28064928655947097,
      "loss": 0.2342,
      "num_input_tokens_seen": 12605264,
      "step": 6540
    },
    {
      "epoch": 1.0677053593278407,
      "grad_norm": 0.003553298767656088,
      "learning_rate": 0.28062033703135103,
      "loss": 0.2249,
      "num_input_tokens_seen": 12616192,
      "step": 6545
    },
    {
      "epoch": 1.068521086548658,
      "grad_norm": 0.002598882419988513,
      "learning_rate": 0.2805913673599337,
      "loss": 0.2123,
      "num_input_tokens_seen": 12626096,
      "step": 6550
    },
    {
      "epoch": 1.0693368137694754,
      "grad_norm": 0.0019637602381408215,
      "learning_rate": 0.2805623775496864,
      "loss": 0.2214,
      "num_input_tokens_seen": 12636208,
      "step": 6555
    },
    {
      "epoch": 1.0701525409902928,
      "grad_norm": 0.002893670229241252,
      "learning_rate": 0.2805333676050797,
      "loss": 0.2173,
      "num_input_tokens_seen": 12646144,
      "step": 6560
    },
    {
      "epoch": 1.0709682682111101,
      "grad_norm": 0.003695861902087927,
      "learning_rate": 0.2805043375305873,
      "loss": 0.2504,
      "num_input_tokens_seen": 12656880,
      "step": 6565
    },
    {
      "epoch": 1.0717839954319275,
      "grad_norm": 0.00265326420776546,
      "learning_rate": 0.2804752873306861,
      "loss": 0.2209,
      "num_input_tokens_seen": 12666640,
      "step": 6570
    },
    {
      "epoch": 1.0725997226527448,
      "grad_norm": 0.0036510147619992495,
      "learning_rate": 0.2804462170098559,
      "loss": 0.2551,
      "num_input_tokens_seen": 12677264,
      "step": 6575
    },
    {
      "epoch": 1.0734154498735622,
      "grad_norm": 0.004649189300835133,
      "learning_rate": 0.2804171265725797,
      "loss": 0.2399,
      "num_input_tokens_seen": 12687808,
      "step": 6580
    },
    {
      "epoch": 1.0742311770943795,
      "grad_norm": 0.0055875014513731,
      "learning_rate": 0.28038801602334373,
      "loss": 0.2496,
      "num_input_tokens_seen": 12697024,
      "step": 6585
    },
    {
      "epoch": 1.075046904315197,
      "grad_norm": 0.0018865700112655759,
      "learning_rate": 0.28035888536663717,
      "loss": 0.2318,
      "num_input_tokens_seen": 12706304,
      "step": 6590
    },
    {
      "epoch": 1.0758626315360145,
      "grad_norm": 0.002321660751476884,
      "learning_rate": 0.2803297346069522,
      "loss": 0.2307,
      "num_input_tokens_seen": 12716768,
      "step": 6595
    },
    {
      "epoch": 1.0766783587568316,
      "grad_norm": 0.002024697605520487,
      "learning_rate": 0.28030056374878437,
      "loss": 0.2529,
      "num_input_tokens_seen": 12726352,
      "step": 6600
    },
    {
      "epoch": 1.0766783587568316,
      "eval_loss": 0.2255900502204895,
      "eval_runtime": 68.0619,
      "eval_samples_per_second": 40.037,
      "eval_steps_per_second": 20.026,
      "num_input_tokens_seen": 12726352,
      "step": 6600
    },
    {
      "epoch": 1.0774940859776492,
      "grad_norm": 0.0022745600435882807,
      "learning_rate": 0.2802713727966321,
      "loss": 0.2245,
      "num_input_tokens_seen": 12736480,
      "step": 6605
    },
    {
      "epoch": 1.0783098131984665,
      "grad_norm": 0.0023409563582390547,
      "learning_rate": 0.28024216175499717,
      "loss": 0.232,
      "num_input_tokens_seen": 12745984,
      "step": 6610
    },
    {
      "epoch": 1.0791255404192839,
      "grad_norm": 0.002539156936109066,
      "learning_rate": 0.2802129306283841,
      "loss": 0.2396,
      "num_input_tokens_seen": 12755920,
      "step": 6615
    },
    {
      "epoch": 1.0799412676401012,
      "grad_norm": 0.0036974467802792788,
      "learning_rate": 0.28018367942130074,
      "loss": 0.2337,
      "num_input_tokens_seen": 12765376,
      "step": 6620
    },
    {
      "epoch": 1.0807569948609186,
      "grad_norm": 0.0026221871376037598,
      "learning_rate": 0.28015440813825804,
      "loss": 0.2349,
      "num_input_tokens_seen": 12773776,
      "step": 6625
    },
    {
      "epoch": 1.081572722081736,
      "grad_norm": 0.006248905323445797,
      "learning_rate": 0.28012511678377006,
      "loss": 0.2144,
      "num_input_tokens_seen": 12784288,
      "step": 6630
    },
    {
      "epoch": 1.0823884493025533,
      "grad_norm": 0.0021040800493210554,
      "learning_rate": 0.28009580536235373,
      "loss": 0.2403,
      "num_input_tokens_seen": 12793264,
      "step": 6635
    },
    {
      "epoch": 1.0832041765233706,
      "grad_norm": 0.0036350837908685207,
      "learning_rate": 0.28006647387852934,
      "loss": 0.2303,
      "num_input_tokens_seen": 12801728,
      "step": 6640
    },
    {
      "epoch": 1.084019903744188,
      "grad_norm": 0.002145822625607252,
      "learning_rate": 0.28003712233682015,
      "loss": 0.2289,
      "num_input_tokens_seen": 12811952,
      "step": 6645
    },
    {
      "epoch": 1.0848356309650053,
      "grad_norm": 0.0035605565644800663,
      "learning_rate": 0.2800077507417526,
      "loss": 0.2294,
      "num_input_tokens_seen": 12821632,
      "step": 6650
    },
    {
      "epoch": 1.0856513581858227,
      "grad_norm": 0.0033981655724346638,
      "learning_rate": 0.2799783590978561,
      "loss": 0.2383,
      "num_input_tokens_seen": 12830832,
      "step": 6655
    },
    {
      "epoch": 1.08646708540664,
      "grad_norm": 0.005504513159394264,
      "learning_rate": 0.2799489474096632,
      "loss": 0.2615,
      "num_input_tokens_seen": 12841520,
      "step": 6660
    },
    {
      "epoch": 1.0872828126274574,
      "grad_norm": 0.002678654622286558,
      "learning_rate": 0.27991951568170953,
      "loss": 0.2364,
      "num_input_tokens_seen": 12852560,
      "step": 6665
    },
    {
      "epoch": 1.0880985398482748,
      "grad_norm": 0.003037855261936784,
      "learning_rate": 0.2798900639185339,
      "loss": 0.2404,
      "num_input_tokens_seen": 12861952,
      "step": 6670
    },
    {
      "epoch": 1.0889142670690921,
      "grad_norm": 0.0018843155121430755,
      "learning_rate": 0.2798605921246781,
      "loss": 0.2281,
      "num_input_tokens_seen": 12871600,
      "step": 6675
    },
    {
      "epoch": 1.0897299942899095,
      "grad_norm": 0.0030629297252744436,
      "learning_rate": 0.2798311003046871,
      "loss": 0.2257,
      "num_input_tokens_seen": 12881520,
      "step": 6680
    },
    {
      "epoch": 1.0905457215107268,
      "grad_norm": 0.002082706894725561,
      "learning_rate": 0.2798015884631089,
      "loss": 0.2418,
      "num_input_tokens_seen": 12890368,
      "step": 6685
    },
    {
      "epoch": 1.0913614487315442,
      "grad_norm": 0.006100207567214966,
      "learning_rate": 0.27977205660449445,
      "loss": 0.2111,
      "num_input_tokens_seen": 12900384,
      "step": 6690
    },
    {
      "epoch": 1.0921771759523615,
      "grad_norm": 0.005576287396252155,
      "learning_rate": 0.2797425047333981,
      "loss": 0.2331,
      "num_input_tokens_seen": 12909056,
      "step": 6695
    },
    {
      "epoch": 1.0929929031731789,
      "grad_norm": 0.002869265154004097,
      "learning_rate": 0.27971293285437715,
      "loss": 0.2355,
      "num_input_tokens_seen": 12919088,
      "step": 6700
    },
    {
      "epoch": 1.0938086303939962,
      "grad_norm": 0.0026976268272846937,
      "learning_rate": 0.2796833409719918,
      "loss": 0.2237,
      "num_input_tokens_seen": 12929504,
      "step": 6705
    },
    {
      "epoch": 1.0946243576148136,
      "grad_norm": 0.006096510216593742,
      "learning_rate": 0.27965372909080566,
      "loss": 0.2168,
      "num_input_tokens_seen": 12939520,
      "step": 6710
    },
    {
      "epoch": 1.095440084835631,
      "grad_norm": 0.0040095094591379166,
      "learning_rate": 0.27962409721538506,
      "loss": 0.1908,
      "num_input_tokens_seen": 12948832,
      "step": 6715
    },
    {
      "epoch": 1.0962558120564483,
      "grad_norm": 0.004279172047972679,
      "learning_rate": 0.27959444535029976,
      "loss": 0.2528,
      "num_input_tokens_seen": 12958176,
      "step": 6720
    },
    {
      "epoch": 1.0970715392772656,
      "grad_norm": 0.0025761707220226526,
      "learning_rate": 0.27956477350012243,
      "loss": 0.2398,
      "num_input_tokens_seen": 12967168,
      "step": 6725
    },
    {
      "epoch": 1.097887266498083,
      "grad_norm": 0.003401112277060747,
      "learning_rate": 0.27953508166942875,
      "loss": 0.2609,
      "num_input_tokens_seen": 12976096,
      "step": 6730
    },
    {
      "epoch": 1.0987029937189003,
      "grad_norm": 0.00511059258133173,
      "learning_rate": 0.27950536986279767,
      "loss": 0.2297,
      "num_input_tokens_seen": 12984880,
      "step": 6735
    },
    {
      "epoch": 1.0995187209397177,
      "grad_norm": 0.002928713569417596,
      "learning_rate": 0.2794756380848111,
      "loss": 0.2379,
      "num_input_tokens_seen": 12995232,
      "step": 6740
    },
    {
      "epoch": 1.100334448160535,
      "grad_norm": 0.002502793911844492,
      "learning_rate": 0.279445886340054,
      "loss": 0.2399,
      "num_input_tokens_seen": 13002576,
      "step": 6745
    },
    {
      "epoch": 1.1011501753813524,
      "grad_norm": 0.00400219950824976,
      "learning_rate": 0.27941611463311455,
      "loss": 0.2331,
      "num_input_tokens_seen": 13011584,
      "step": 6750
    },
    {
      "epoch": 1.1019659026021698,
      "grad_norm": 0.0035905311815440655,
      "learning_rate": 0.2793863229685839,
      "loss": 0.2336,
      "num_input_tokens_seen": 13022288,
      "step": 6755
    },
    {
      "epoch": 1.102781629822987,
      "grad_norm": 0.004720817785710096,
      "learning_rate": 0.27935651135105627,
      "loss": 0.2319,
      "num_input_tokens_seen": 13032000,
      "step": 6760
    },
    {
      "epoch": 1.1035973570438045,
      "grad_norm": 0.001553499372676015,
      "learning_rate": 0.279326679785129,
      "loss": 0.2408,
      "num_input_tokens_seen": 13040048,
      "step": 6765
    },
    {
      "epoch": 1.1044130842646218,
      "grad_norm": 0.002694274066016078,
      "learning_rate": 0.2792968282754024,
      "loss": 0.2422,
      "num_input_tokens_seen": 13049248,
      "step": 6770
    },
    {
      "epoch": 1.1052288114854392,
      "grad_norm": 0.003442636225372553,
      "learning_rate": 0.2792669568264801,
      "loss": 0.2393,
      "num_input_tokens_seen": 13057616,
      "step": 6775
    },
    {
      "epoch": 1.1060445387062567,
      "grad_norm": 0.003207291942089796,
      "learning_rate": 0.27923706544296856,
      "loss": 0.2464,
      "num_input_tokens_seen": 13066320,
      "step": 6780
    },
    {
      "epoch": 1.1068602659270739,
      "grad_norm": 0.002176588401198387,
      "learning_rate": 0.2792071541294775,
      "loss": 0.2329,
      "num_input_tokens_seen": 13075696,
      "step": 6785
    },
    {
      "epoch": 1.1076759931478914,
      "grad_norm": 0.0018910480430349708,
      "learning_rate": 0.27917722289061947,
      "loss": 0.2259,
      "num_input_tokens_seen": 13085872,
      "step": 6790
    },
    {
      "epoch": 1.1084917203687088,
      "grad_norm": 0.0029372619464993477,
      "learning_rate": 0.27914727173101034,
      "loss": 0.2111,
      "num_input_tokens_seen": 13095248,
      "step": 6795
    },
    {
      "epoch": 1.1093074475895262,
      "grad_norm": 0.001698796171694994,
      "learning_rate": 0.279117300655269,
      "loss": 0.2318,
      "num_input_tokens_seen": 13105200,
      "step": 6800
    },
    {
      "epoch": 1.1093074475895262,
      "eval_loss": 0.22426138818264008,
      "eval_runtime": 68.0034,
      "eval_samples_per_second": 40.072,
      "eval_steps_per_second": 20.043,
      "num_input_tokens_seen": 13105200,
      "step": 6800
    },
    {
      "epoch": 1.1101231748103435,
      "grad_norm": 0.0036156128626316786,
      "learning_rate": 0.2790873096680173,
      "loss": 0.236,
      "num_input_tokens_seen": 13113760,
      "step": 6805
    },
    {
      "epoch": 1.1109389020311609,
      "grad_norm": 0.005617002956569195,
      "learning_rate": 0.2790572987738802,
      "loss": 0.2041,
      "num_input_tokens_seen": 13124768,
      "step": 6810
    },
    {
      "epoch": 1.1117546292519782,
      "grad_norm": 0.0019667865708470345,
      "learning_rate": 0.27902726797748584,
      "loss": 0.1884,
      "num_input_tokens_seen": 13134608,
      "step": 6815
    },
    {
      "epoch": 1.1125703564727956,
      "grad_norm": 0.001536951633170247,
      "learning_rate": 0.2789972172834652,
      "loss": 0.1941,
      "num_input_tokens_seen": 13144000,
      "step": 6820
    },
    {
      "epoch": 1.113386083693613,
      "grad_norm": 0.0028498899191617966,
      "learning_rate": 0.2789671466964527,
      "loss": 0.2673,
      "num_input_tokens_seen": 13152928,
      "step": 6825
    },
    {
      "epoch": 1.1142018109144303,
      "grad_norm": 0.004296474624425173,
      "learning_rate": 0.2789370562210854,
      "loss": 0.2615,
      "num_input_tokens_seen": 13162288,
      "step": 6830
    },
    {
      "epoch": 1.1150175381352476,
      "grad_norm": 0.003375919768586755,
      "learning_rate": 0.27890694586200376,
      "loss": 0.2213,
      "num_input_tokens_seen": 13170192,
      "step": 6835
    },
    {
      "epoch": 1.115833265356065,
      "grad_norm": 0.003806260647252202,
      "learning_rate": 0.2788768156238511,
      "loss": 0.2458,
      "num_input_tokens_seen": 13179584,
      "step": 6840
    },
    {
      "epoch": 1.1166489925768823,
      "grad_norm": 0.002426475053653121,
      "learning_rate": 0.27884666551127385,
      "loss": 0.222,
      "num_input_tokens_seen": 13188624,
      "step": 6845
    },
    {
      "epoch": 1.1174647197976997,
      "grad_norm": 0.0054281000047922134,
      "learning_rate": 0.2788164955289217,
      "loss": 0.2449,
      "num_input_tokens_seen": 13198944,
      "step": 6850
    },
    {
      "epoch": 1.118280447018517,
      "grad_norm": 0.004752063658088446,
      "learning_rate": 0.27878630568144697,
      "loss": 0.2221,
      "num_input_tokens_seen": 13208176,
      "step": 6855
    },
    {
      "epoch": 1.1190961742393344,
      "grad_norm": 0.0021998309530317783,
      "learning_rate": 0.2787560959735056,
      "loss": 0.24,
      "num_input_tokens_seen": 13218000,
      "step": 6860
    },
    {
      "epoch": 1.1199119014601517,
      "grad_norm": 0.0048821065574884415,
      "learning_rate": 0.27872586640975616,
      "loss": 0.2248,
      "num_input_tokens_seen": 13227952,
      "step": 6865
    },
    {
      "epoch": 1.120727628680969,
      "grad_norm": 0.0027442092541605234,
      "learning_rate": 0.27869561699486045,
      "loss": 0.2442,
      "num_input_tokens_seen": 13238480,
      "step": 6870
    },
    {
      "epoch": 1.1215433559017864,
      "grad_norm": 0.0023296880535781384,
      "learning_rate": 0.2786653477334833,
      "loss": 0.2346,
      "num_input_tokens_seen": 13247488,
      "step": 6875
    },
    {
      "epoch": 1.1223590831226038,
      "grad_norm": 0.0025866245850920677,
      "learning_rate": 0.2786350586302926,
      "loss": 0.2076,
      "num_input_tokens_seen": 13256000,
      "step": 6880
    },
    {
      "epoch": 1.1231748103434211,
      "grad_norm": 0.0056564523838460445,
      "learning_rate": 0.27860474968995935,
      "loss": 0.2521,
      "num_input_tokens_seen": 13265488,
      "step": 6885
    },
    {
      "epoch": 1.1239905375642385,
      "grad_norm": 0.0019951851572841406,
      "learning_rate": 0.27857442091715756,
      "loss": 0.234,
      "num_input_tokens_seen": 13274128,
      "step": 6890
    },
    {
      "epoch": 1.1248062647850559,
      "grad_norm": 0.0028697927482426167,
      "learning_rate": 0.27854407231656425,
      "loss": 0.2288,
      "num_input_tokens_seen": 13284384,
      "step": 6895
    },
    {
      "epoch": 1.1256219920058732,
      "grad_norm": 0.002051249612122774,
      "learning_rate": 0.2785137038928596,
      "loss": 0.2165,
      "num_input_tokens_seen": 13294560,
      "step": 6900
    },
    {
      "epoch": 1.1264377192266906,
      "grad_norm": 0.004393431823700666,
      "learning_rate": 0.27848331565072687,
      "loss": 0.2422,
      "num_input_tokens_seen": 13304656,
      "step": 6905
    },
    {
      "epoch": 1.127253446447508,
      "grad_norm": 0.0044518159702420235,
      "learning_rate": 0.27845290759485225,
      "loss": 0.2167,
      "num_input_tokens_seen": 13315296,
      "step": 6910
    },
    {
      "epoch": 1.1280691736683253,
      "grad_norm": 0.0020196076948195696,
      "learning_rate": 0.278422479729925,
      "loss": 0.25,
      "num_input_tokens_seen": 13325152,
      "step": 6915
    },
    {
      "epoch": 1.1288849008891426,
      "grad_norm": 0.003301221877336502,
      "learning_rate": 0.2783920320606375,
      "loss": 0.2398,
      "num_input_tokens_seen": 13333728,
      "step": 6920
    },
    {
      "epoch": 1.12970062810996,
      "grad_norm": 0.0037529231049120426,
      "learning_rate": 0.2783615645916852,
      "loss": 0.2388,
      "num_input_tokens_seen": 13342880,
      "step": 6925
    },
    {
      "epoch": 1.1305163553307773,
      "grad_norm": 0.003912606276571751,
      "learning_rate": 0.2783310773277666,
      "loss": 0.2113,
      "num_input_tokens_seen": 13351616,
      "step": 6930
    },
    {
      "epoch": 1.1313320825515947,
      "grad_norm": 0.003384537296369672,
      "learning_rate": 0.2783005702735831,
      "loss": 0.206,
      "num_input_tokens_seen": 13361872,
      "step": 6935
    },
    {
      "epoch": 1.132147809772412,
      "grad_norm": 0.001335157547146082,
      "learning_rate": 0.2782700434338394,
      "loss": 0.2415,
      "num_input_tokens_seen": 13370336,
      "step": 6940
    },
    {
      "epoch": 1.1329635369932294,
      "grad_norm": 0.001683565555140376,
      "learning_rate": 0.278239496813243,
      "loss": 0.2424,
      "num_input_tokens_seen": 13378720,
      "step": 6945
    },
    {
      "epoch": 1.1337792642140467,
      "grad_norm": 0.002555681159719825,
      "learning_rate": 0.27820893041650463,
      "loss": 0.2172,
      "num_input_tokens_seen": 13388720,
      "step": 6950
    },
    {
      "epoch": 1.134594991434864,
      "grad_norm": 0.0023529455065727234,
      "learning_rate": 0.27817834424833804,
      "loss": 0.2317,
      "num_input_tokens_seen": 13398304,
      "step": 6955
    },
    {
      "epoch": 1.1354107186556814,
      "grad_norm": 0.0018806750886142254,
      "learning_rate": 0.27814773831345996,
      "loss": 0.2332,
      "num_input_tokens_seen": 13408368,
      "step": 6960
    },
    {
      "epoch": 1.136226445876499,
      "grad_norm": 0.0033800492528826,
      "learning_rate": 0.2781171126165902,
      "loss": 0.2364,
      "num_input_tokens_seen": 13418512,
      "step": 6965
    },
    {
      "epoch": 1.1370421730973161,
      "grad_norm": 0.0023106406442821026,
      "learning_rate": 0.2780864671624517,
      "loss": 0.2174,
      "num_input_tokens_seen": 13427296,
      "step": 6970
    },
    {
      "epoch": 1.1378579003181337,
      "grad_norm": 0.004166170954704285,
      "learning_rate": 0.27805580195577034,
      "loss": 0.2182,
      "num_input_tokens_seen": 13435264,
      "step": 6975
    },
    {
      "epoch": 1.1386736275389508,
      "grad_norm": 0.00351844634860754,
      "learning_rate": 0.2780251170012751,
      "loss": 0.201,
      "num_input_tokens_seen": 13444608,
      "step": 6980
    },
    {
      "epoch": 1.1394893547597684,
      "grad_norm": 0.001796938362531364,
      "learning_rate": 0.27799441230369787,
      "loss": 0.2433,
      "num_input_tokens_seen": 13454528,
      "step": 6985
    },
    {
      "epoch": 1.1403050819805858,
      "grad_norm": 0.005050116218626499,
      "learning_rate": 0.27796368786777387,
      "loss": 0.2381,
      "num_input_tokens_seen": 13464016,
      "step": 6990
    },
    {
      "epoch": 1.1411208092014031,
      "grad_norm": 0.0033169910311698914,
      "learning_rate": 0.277932943698241,
      "loss": 0.2445,
      "num_input_tokens_seen": 13473488,
      "step": 6995
    },
    {
      "epoch": 1.1419365364222205,
      "grad_norm": 0.00308008911088109,
      "learning_rate": 0.2779021797998406,
      "loss": 0.2397,
      "num_input_tokens_seen": 13483648,
      "step": 7000
    },
    {
      "epoch": 1.1419365364222205,
      "eval_loss": 0.23303374648094177,
      "eval_runtime": 67.9328,
      "eval_samples_per_second": 40.113,
      "eval_steps_per_second": 20.064,
      "num_input_tokens_seen": 13483648,
      "step": 7000
    },
    {
      "epoch": 1.1427522636430378,
      "grad_norm": 0.005234956741333008,
      "learning_rate": 0.2778713961773167,
      "loss": 0.2403,
      "num_input_tokens_seen": 13492992,
      "step": 7005
    },
    {
      "epoch": 1.1435679908638552,
      "grad_norm": 0.0028359463904052973,
      "learning_rate": 0.2778405928354166,
      "loss": 0.2296,
      "num_input_tokens_seen": 13502208,
      "step": 7010
    },
    {
      "epoch": 1.1443837180846725,
      "grad_norm": 0.0011770130367949605,
      "learning_rate": 0.27780976977889055,
      "loss": 0.2236,
      "num_input_tokens_seen": 13511600,
      "step": 7015
    },
    {
      "epoch": 1.14519944530549,
      "grad_norm": 0.002307882299646735,
      "learning_rate": 0.27777892701249185,
      "loss": 0.2269,
      "num_input_tokens_seen": 13521760,
      "step": 7020
    },
    {
      "epoch": 1.1460151725263072,
      "grad_norm": 0.00213801977224648,
      "learning_rate": 0.2777480645409768,
      "loss": 0.2224,
      "num_input_tokens_seen": 13530752,
      "step": 7025
    },
    {
      "epoch": 1.1468308997471246,
      "grad_norm": 0.0018503686878830194,
      "learning_rate": 0.27771718236910486,
      "loss": 0.2149,
      "num_input_tokens_seen": 13540256,
      "step": 7030
    },
    {
      "epoch": 1.147646626967942,
      "grad_norm": 0.004723282065242529,
      "learning_rate": 0.27768628050163835,
      "loss": 0.1972,
      "num_input_tokens_seen": 13551168,
      "step": 7035
    },
    {
      "epoch": 1.1484623541887593,
      "grad_norm": 0.002966300118714571,
      "learning_rate": 0.2776553589433428,
      "loss": 0.276,
      "num_input_tokens_seen": 13560608,
      "step": 7040
    },
    {
      "epoch": 1.1492780814095767,
      "grad_norm": 0.005091426428407431,
      "learning_rate": 0.27762441769898666,
      "loss": 0.2384,
      "num_input_tokens_seen": 13569424,
      "step": 7045
    },
    {
      "epoch": 1.150093808630394,
      "grad_norm": 0.001886715181171894,
      "learning_rate": 0.2775934567733415,
      "loss": 0.2255,
      "num_input_tokens_seen": 13578272,
      "step": 7050
    },
    {
      "epoch": 1.1509095358512114,
      "grad_norm": 0.0028985326644033194,
      "learning_rate": 0.2775624761711819,
      "loss": 0.2207,
      "num_input_tokens_seen": 13588192,
      "step": 7055
    },
    {
      "epoch": 1.1517252630720287,
      "grad_norm": 0.0028998423367738724,
      "learning_rate": 0.2775314758972854,
      "loss": 0.2269,
      "num_input_tokens_seen": 13596976,
      "step": 7060
    },
    {
      "epoch": 1.152540990292846,
      "grad_norm": 0.0036694819573312998,
      "learning_rate": 0.2775004559564327,
      "loss": 0.2274,
      "num_input_tokens_seen": 13608016,
      "step": 7065
    },
    {
      "epoch": 1.1533567175136634,
      "grad_norm": 0.0037324780132621527,
      "learning_rate": 0.2774694163534073,
      "loss": 0.2333,
      "num_input_tokens_seen": 13617040,
      "step": 7070
    },
    {
      "epoch": 1.1541724447344808,
      "grad_norm": 0.004811120685189962,
      "learning_rate": 0.27743835709299614,
      "loss": 0.2358,
      "num_input_tokens_seen": 13625856,
      "step": 7075
    },
    {
      "epoch": 1.1549881719552981,
      "grad_norm": 0.0023574240040034056,
      "learning_rate": 0.2774072781799888,
      "loss": 0.2764,
      "num_input_tokens_seen": 13634928,
      "step": 7080
    },
    {
      "epoch": 1.1558038991761155,
      "grad_norm": 0.0032747220247983932,
      "learning_rate": 0.27737617961917804,
      "loss": 0.2473,
      "num_input_tokens_seen": 13644736,
      "step": 7085
    },
    {
      "epoch": 1.1566196263969328,
      "grad_norm": 0.002182804048061371,
      "learning_rate": 0.27734506141535964,
      "loss": 0.2466,
      "num_input_tokens_seen": 13655136,
      "step": 7090
    },
    {
      "epoch": 1.1574353536177502,
      "grad_norm": 0.004150434397161007,
      "learning_rate": 0.2773139235733325,
      "loss": 0.2098,
      "num_input_tokens_seen": 13664656,
      "step": 7095
    },
    {
      "epoch": 1.1582510808385675,
      "grad_norm": 0.003079087007790804,
      "learning_rate": 0.2772827660978984,
      "loss": 0.2085,
      "num_input_tokens_seen": 13672880,
      "step": 7100
    },
    {
      "epoch": 1.159066808059385,
      "grad_norm": 0.002510182559490204,
      "learning_rate": 0.27725158899386226,
      "loss": 0.2493,
      "num_input_tokens_seen": 13682240,
      "step": 7105
    },
    {
      "epoch": 1.1598825352802022,
      "grad_norm": 0.005055922549217939,
      "learning_rate": 0.27722039226603196,
      "loss": 0.2326,
      "num_input_tokens_seen": 13690720,
      "step": 7110
    },
    {
      "epoch": 1.1606982625010196,
      "grad_norm": 0.0018735129851847887,
      "learning_rate": 0.2771891759192184,
      "loss": 0.227,
      "num_input_tokens_seen": 13699888,
      "step": 7115
    },
    {
      "epoch": 1.161513989721837,
      "grad_norm": 0.0030958435963839293,
      "learning_rate": 0.2771579399582355,
      "loss": 0.2302,
      "num_input_tokens_seen": 13709360,
      "step": 7120
    },
    {
      "epoch": 1.1623297169426543,
      "grad_norm": 0.004835519473999739,
      "learning_rate": 0.2771266843879004,
      "loss": 0.2317,
      "num_input_tokens_seen": 13719040,
      "step": 7125
    },
    {
      "epoch": 1.1631454441634717,
      "grad_norm": 0.0030543343164026737,
      "learning_rate": 0.2770954092130329,
      "loss": 0.2447,
      "num_input_tokens_seen": 13729328,
      "step": 7130
    },
    {
      "epoch": 1.163961171384289,
      "grad_norm": 0.0057610091753304005,
      "learning_rate": 0.27706411443845613,
      "loss": 0.2049,
      "num_input_tokens_seen": 13739792,
      "step": 7135
    },
    {
      "epoch": 1.1647768986051064,
      "grad_norm": 0.002834867686033249,
      "learning_rate": 0.27703280006899617,
      "loss": 0.2516,
      "num_input_tokens_seen": 13748912,
      "step": 7140
    },
    {
      "epoch": 1.1655926258259237,
      "grad_norm": 0.0020824852399528027,
      "learning_rate": 0.277001466109482,
      "loss": 0.2147,
      "num_input_tokens_seen": 13758656,
      "step": 7145
    },
    {
      "epoch": 1.1664083530467413,
      "grad_norm": 0.0029730245005339384,
      "learning_rate": 0.2769701125647458,
      "loss": 0.2325,
      "num_input_tokens_seen": 13768944,
      "step": 7150
    },
    {
      "epoch": 1.1672240802675584,
      "grad_norm": 0.0063051627948880196,
      "learning_rate": 0.27693873943962266,
      "loss": 0.2297,
      "num_input_tokens_seen": 13776880,
      "step": 7155
    },
    {
      "epoch": 1.168039807488376,
      "grad_norm": 0.0035773245617747307,
      "learning_rate": 0.2769073467389506,
      "loss": 0.2329,
      "num_input_tokens_seen": 13785488,
      "step": 7160
    },
    {
      "epoch": 1.1688555347091931,
      "grad_norm": 0.0017024407861754298,
      "learning_rate": 0.2768759344675709,
      "loss": 0.2239,
      "num_input_tokens_seen": 13795136,
      "step": 7165
    },
    {
      "epoch": 1.1696712619300107,
      "grad_norm": 0.0020036939531564713,
      "learning_rate": 0.27684450263032767,
      "loss": 0.2235,
      "num_input_tokens_seen": 13804720,
      "step": 7170
    },
    {
      "epoch": 1.170486989150828,
      "grad_norm": 0.0045548235066235065,
      "learning_rate": 0.2768130512320682,
      "loss": 0.2072,
      "num_input_tokens_seen": 13814272,
      "step": 7175
    },
    {
      "epoch": 1.1713027163716454,
      "grad_norm": 0.004253558348864317,
      "learning_rate": 0.27678158027764244,
      "loss": 0.2384,
      "num_input_tokens_seen": 13824624,
      "step": 7180
    },
    {
      "epoch": 1.1721184435924628,
      "grad_norm": 0.003657011082395911,
      "learning_rate": 0.27675008977190385,
      "loss": 0.2267,
      "num_input_tokens_seen": 13833328,
      "step": 7185
    },
    {
      "epoch": 1.17293417081328,
      "grad_norm": 0.002196050249040127,
      "learning_rate": 0.2767185797197086,
      "loss": 0.2171,
      "num_input_tokens_seen": 13843520,
      "step": 7190
    },
    {
      "epoch": 1.1737498980340975,
      "grad_norm": 0.004302776884287596,
      "learning_rate": 0.2766870501259159,
      "loss": 0.3034,
      "num_input_tokens_seen": 13852736,
      "step": 7195
    },
    {
      "epoch": 1.1745656252549148,
      "grad_norm": 0.005437311716377735,
      "learning_rate": 0.276655500995388,
      "loss": 0.2211,
      "num_input_tokens_seen": 13862816,
      "step": 7200
    },
    {
      "epoch": 1.1745656252549148,
      "eval_loss": 0.25151845812797546,
      "eval_runtime": 68.1554,
      "eval_samples_per_second": 39.982,
      "eval_steps_per_second": 19.998,
      "num_input_tokens_seen": 13862816,
      "step": 7200
    },
    {
      "epoch": 1.1753813524757322,
      "grad_norm": 0.002657757606357336,
      "learning_rate": 0.27662393233299015,
      "loss": 0.2461,
      "num_input_tokens_seen": 13872784,
      "step": 7205
    },
    {
      "epoch": 1.1761970796965495,
      "grad_norm": 0.004399906378239393,
      "learning_rate": 0.27659234414359074,
      "loss": 0.2282,
      "num_input_tokens_seen": 13883040,
      "step": 7210
    },
    {
      "epoch": 1.1770128069173669,
      "grad_norm": 0.0034741414710879326,
      "learning_rate": 0.27656073643206097,
      "loss": 0.2154,
      "num_input_tokens_seen": 13891632,
      "step": 7215
    },
    {
      "epoch": 1.1778285341381842,
      "grad_norm": 0.007295540999621153,
      "learning_rate": 0.27652910920327517,
      "loss": 0.2586,
      "num_input_tokens_seen": 13900928,
      "step": 7220
    },
    {
      "epoch": 1.1786442613590016,
      "grad_norm": 0.0029533091001212597,
      "learning_rate": 0.2764974624621107,
      "loss": 0.2547,
      "num_input_tokens_seen": 13910304,
      "step": 7225
    },
    {
      "epoch": 1.179459988579819,
      "grad_norm": 0.007529593538492918,
      "learning_rate": 0.2764657962134479,
      "loss": 0.2705,
      "num_input_tokens_seen": 13919824,
      "step": 7230
    },
    {
      "epoch": 1.1802757158006363,
      "grad_norm": 0.0020041130483150482,
      "learning_rate": 0.27643411046217,
      "loss": 0.2365,
      "num_input_tokens_seen": 13929184,
      "step": 7235
    },
    {
      "epoch": 1.1810914430214536,
      "grad_norm": 0.004479687195271254,
      "learning_rate": 0.27640240521316334,
      "loss": 0.242,
      "num_input_tokens_seen": 13939744,
      "step": 7240
    },
    {
      "epoch": 1.181907170242271,
      "grad_norm": 0.0024349249433726072,
      "learning_rate": 0.2763706804713174,
      "loss": 0.2361,
      "num_input_tokens_seen": 13949632,
      "step": 7245
    },
    {
      "epoch": 1.1827228974630883,
      "grad_norm": 0.0020498973317444324,
      "learning_rate": 0.2763389362415245,
      "loss": 0.2468,
      "num_input_tokens_seen": 13959872,
      "step": 7250
    },
    {
      "epoch": 1.1835386246839057,
      "grad_norm": 0.0053410702385008335,
      "learning_rate": 0.27630717252867987,
      "loss": 0.2425,
      "num_input_tokens_seen": 13969520,
      "step": 7255
    },
    {
      "epoch": 1.184354351904723,
      "grad_norm": 0.0023327579256147146,
      "learning_rate": 0.276275389337682,
      "loss": 0.2313,
      "num_input_tokens_seen": 13977680,
      "step": 7260
    },
    {
      "epoch": 1.1851700791255404,
      "grad_norm": 0.005204034969210625,
      "learning_rate": 0.2762435866734322,
      "loss": 0.2292,
      "num_input_tokens_seen": 13987440,
      "step": 7265
    },
    {
      "epoch": 1.1859858063463578,
      "grad_norm": 0.0023220896255224943,
      "learning_rate": 0.27621176454083485,
      "loss": 0.2231,
      "num_input_tokens_seen": 13996976,
      "step": 7270
    },
    {
      "epoch": 1.186801533567175,
      "grad_norm": 0.002610586117953062,
      "learning_rate": 0.2761799229447973,
      "loss": 0.2297,
      "num_input_tokens_seen": 14007424,
      "step": 7275
    },
    {
      "epoch": 1.1876172607879925,
      "grad_norm": 0.005668473429977894,
      "learning_rate": 0.27614806189023006,
      "loss": 0.2471,
      "num_input_tokens_seen": 14019120,
      "step": 7280
    },
    {
      "epoch": 1.1884329880088098,
      "grad_norm": 0.004268669988960028,
      "learning_rate": 0.27611618138204636,
      "loss": 0.2014,
      "num_input_tokens_seen": 14028976,
      "step": 7285
    },
    {
      "epoch": 1.1892487152296272,
      "grad_norm": 0.005752684082835913,
      "learning_rate": 0.2760842814251626,
      "loss": 0.2677,
      "num_input_tokens_seen": 14039200,
      "step": 7290
    },
    {
      "epoch": 1.1900644424504445,
      "grad_norm": 0.00346958776935935,
      "learning_rate": 0.2760523620244982,
      "loss": 0.2427,
      "num_input_tokens_seen": 14047984,
      "step": 7295
    },
    {
      "epoch": 1.1908801696712619,
      "grad_norm": 0.0030905783642083406,
      "learning_rate": 0.27602042318497544,
      "loss": 0.2468,
      "num_input_tokens_seen": 14057792,
      "step": 7300
    },
    {
      "epoch": 1.1916958968920792,
      "grad_norm": 0.002993278205394745,
      "learning_rate": 0.2759884649115198,
      "loss": 0.267,
      "num_input_tokens_seen": 14066048,
      "step": 7305
    },
    {
      "epoch": 1.1925116241128966,
      "grad_norm": 0.00538286566734314,
      "learning_rate": 0.2759564872090596,
      "loss": 0.2289,
      "num_input_tokens_seen": 14075744,
      "step": 7310
    },
    {
      "epoch": 1.193327351333714,
      "grad_norm": 0.0043473015539348125,
      "learning_rate": 0.2759244900825262,
      "loss": 0.2108,
      "num_input_tokens_seen": 14086224,
      "step": 7315
    },
    {
      "epoch": 1.1941430785545313,
      "grad_norm": 0.0057536400854587555,
      "learning_rate": 0.2758924735368539,
      "loss": 0.2572,
      "num_input_tokens_seen": 14097744,
      "step": 7320
    },
    {
      "epoch": 1.1949588057753486,
      "grad_norm": 0.007565231528133154,
      "learning_rate": 0.27586043757698014,
      "loss": 0.2454,
      "num_input_tokens_seen": 14106752,
      "step": 7325
    },
    {
      "epoch": 1.195774532996166,
      "grad_norm": 0.004808663856238127,
      "learning_rate": 0.27582838220784534,
      "loss": 0.2459,
      "num_input_tokens_seen": 14116256,
      "step": 7330
    },
    {
      "epoch": 1.1965902602169836,
      "grad_norm": 0.002352276584133506,
      "learning_rate": 0.27579630743439265,
      "loss": 0.2247,
      "num_input_tokens_seen": 14126608,
      "step": 7335
    },
    {
      "epoch": 1.1974059874378007,
      "grad_norm": 0.003007532563060522,
      "learning_rate": 0.2757642132615686,
      "loss": 0.2377,
      "num_input_tokens_seen": 14136064,
      "step": 7340
    },
    {
      "epoch": 1.1982217146586183,
      "grad_norm": 0.0027499052230268717,
      "learning_rate": 0.2757320996943223,
      "loss": 0.2453,
      "num_input_tokens_seen": 14146160,
      "step": 7345
    },
    {
      "epoch": 1.1990374418794354,
      "grad_norm": 0.003496755613014102,
      "learning_rate": 0.2756999667376062,
      "loss": 0.2194,
      "num_input_tokens_seen": 14155888,
      "step": 7350
    },
    {
      "epoch": 1.199853169100253,
      "grad_norm": 0.0027660687919706106,
      "learning_rate": 0.2756678143963756,
      "loss": 0.2458,
      "num_input_tokens_seen": 14164912,
      "step": 7355
    },
    {
      "epoch": 1.2006688963210703,
      "grad_norm": 0.00307522458024323,
      "learning_rate": 0.2756356426755888,
      "loss": 0.2105,
      "num_input_tokens_seen": 14174096,
      "step": 7360
    },
    {
      "epoch": 1.2014846235418877,
      "grad_norm": 0.0016604360425844789,
      "learning_rate": 0.27560345158020705,
      "loss": 0.2196,
      "num_input_tokens_seen": 14183760,
      "step": 7365
    },
    {
      "epoch": 1.202300350762705,
      "grad_norm": 0.0034727330785244703,
      "learning_rate": 0.27557124111519465,
      "loss": 0.2466,
      "num_input_tokens_seen": 14192704,
      "step": 7370
    },
    {
      "epoch": 1.2031160779835224,
      "grad_norm": 0.0056239767000079155,
      "learning_rate": 0.27553901128551883,
      "loss": 0.2466,
      "num_input_tokens_seen": 14202800,
      "step": 7375
    },
    {
      "epoch": 1.2039318052043397,
      "grad_norm": 0.0015910051297396421,
      "learning_rate": 0.2755067620961498,
      "loss": 0.251,
      "num_input_tokens_seen": 14212480,
      "step": 7380
    },
    {
      "epoch": 1.204747532425157,
      "grad_norm": 0.0033945413306355476,
      "learning_rate": 0.27547449355206094,
      "loss": 0.222,
      "num_input_tokens_seen": 14221648,
      "step": 7385
    },
    {
      "epoch": 1.2055632596459744,
      "grad_norm": 0.004281153902411461,
      "learning_rate": 0.2754422056582283,
      "loss": 0.2411,
      "num_input_tokens_seen": 14231680,
      "step": 7390
    },
    {
      "epoch": 1.2063789868667918,
      "grad_norm": 0.002069859765470028,
      "learning_rate": 0.27540989841963115,
      "loss": 0.2256,
      "num_input_tokens_seen": 14241824,
      "step": 7395
    },
    {
      "epoch": 1.2071947140876091,
      "grad_norm": 0.0025946395471692085,
      "learning_rate": 0.27537757184125167,
      "loss": 0.187,
      "num_input_tokens_seen": 14252288,
      "step": 7400
    },
    {
      "epoch": 1.2071947140876091,
      "eval_loss": 0.22993551194667816,
      "eval_runtime": 68.0442,
      "eval_samples_per_second": 40.047,
      "eval_steps_per_second": 20.031,
      "num_input_tokens_seen": 14252288,
      "step": 7400
    },
    {
      "epoch": 1.2080104413084265,
      "grad_norm": 0.006765143945813179,
      "learning_rate": 0.275345225928075,
      "loss": 0.2761,
      "num_input_tokens_seen": 14261088,
      "step": 7405
    },
    {
      "epoch": 1.2088261685292438,
      "grad_norm": 0.003755626967176795,
      "learning_rate": 0.2753128606850893,
      "loss": 0.2404,
      "num_input_tokens_seen": 14271248,
      "step": 7410
    },
    {
      "epoch": 1.2096418957500612,
      "grad_norm": 0.0070985765196383,
      "learning_rate": 0.2752804761172858,
      "loss": 0.2692,
      "num_input_tokens_seen": 14281216,
      "step": 7415
    },
    {
      "epoch": 1.2104576229708786,
      "grad_norm": 0.0019460059702396393,
      "learning_rate": 0.27524807222965836,
      "loss": 0.2634,
      "num_input_tokens_seen": 14291184,
      "step": 7420
    },
    {
      "epoch": 1.211273350191696,
      "grad_norm": 0.006526197772473097,
      "learning_rate": 0.27521564902720436,
      "loss": 0.2415,
      "num_input_tokens_seen": 14301472,
      "step": 7425
    },
    {
      "epoch": 1.2120890774125133,
      "grad_norm": 0.0038413682486861944,
      "learning_rate": 0.2751832065149236,
      "loss": 0.2182,
      "num_input_tokens_seen": 14310880,
      "step": 7430
    },
    {
      "epoch": 1.2129048046333306,
      "grad_norm": 0.004833206068724394,
      "learning_rate": 0.2751507446978193,
      "loss": 0.2321,
      "num_input_tokens_seen": 14321408,
      "step": 7435
    },
    {
      "epoch": 1.213720531854148,
      "grad_norm": 0.002738840179517865,
      "learning_rate": 0.2751182635808974,
      "loss": 0.2148,
      "num_input_tokens_seen": 14330608,
      "step": 7440
    },
    {
      "epoch": 1.2145362590749653,
      "grad_norm": 0.0032908169087022543,
      "learning_rate": 0.27508576316916694,
      "loss": 0.2483,
      "num_input_tokens_seen": 14341072,
      "step": 7445
    },
    {
      "epoch": 1.2153519862957827,
      "grad_norm": 0.0011683382326737046,
      "learning_rate": 0.2750532434676399,
      "loss": 0.214,
      "num_input_tokens_seen": 14351360,
      "step": 7450
    },
    {
      "epoch": 1.2161677135166,
      "grad_norm": 0.00419272854924202,
      "learning_rate": 0.27502070448133115,
      "loss": 0.2297,
      "num_input_tokens_seen": 14360880,
      "step": 7455
    },
    {
      "epoch": 1.2169834407374174,
      "grad_norm": 0.0026388715486973524,
      "learning_rate": 0.2749881462152587,
      "loss": 0.2017,
      "num_input_tokens_seen": 14371120,
      "step": 7460
    },
    {
      "epoch": 1.2177991679582347,
      "grad_norm": 0.006328687071800232,
      "learning_rate": 0.2749555686744434,
      "loss": 0.2871,
      "num_input_tokens_seen": 14380544,
      "step": 7465
    },
    {
      "epoch": 1.218614895179052,
      "grad_norm": 0.003436900209635496,
      "learning_rate": 0.2749229718639091,
      "loss": 0.2239,
      "num_input_tokens_seen": 14389824,
      "step": 7470
    },
    {
      "epoch": 1.2194306223998694,
      "grad_norm": 0.0021050286013633013,
      "learning_rate": 0.27489035578868265,
      "loss": 0.2425,
      "num_input_tokens_seen": 14398816,
      "step": 7475
    },
    {
      "epoch": 1.2202463496206868,
      "grad_norm": 0.003157935570925474,
      "learning_rate": 0.2748577204537939,
      "loss": 0.26,
      "num_input_tokens_seen": 14408096,
      "step": 7480
    },
    {
      "epoch": 1.2210620768415041,
      "grad_norm": 0.003020482836291194,
      "learning_rate": 0.2748250658642756,
      "loss": 0.2425,
      "num_input_tokens_seen": 14417552,
      "step": 7485
    },
    {
      "epoch": 1.2218778040623215,
      "grad_norm": 0.004075151402503252,
      "learning_rate": 0.2747923920251634,
      "loss": 0.2482,
      "num_input_tokens_seen": 14426816,
      "step": 7490
    },
    {
      "epoch": 1.2226935312831388,
      "grad_norm": 0.00269410596229136,
      "learning_rate": 0.27475969894149627,
      "loss": 0.24,
      "num_input_tokens_seen": 14435472,
      "step": 7495
    },
    {
      "epoch": 1.2235092585039562,
      "grad_norm": 0.001951803918927908,
      "learning_rate": 0.2747269866183156,
      "loss": 0.2361,
      "num_input_tokens_seen": 14445344,
      "step": 7500
    },
    {
      "epoch": 1.2243249857247736,
      "grad_norm": 0.0018858193652704358,
      "learning_rate": 0.27469425506066625,
      "loss": 0.2128,
      "num_input_tokens_seen": 14456080,
      "step": 7505
    },
    {
      "epoch": 1.225140712945591,
      "grad_norm": 0.003660524496808648,
      "learning_rate": 0.27466150427359576,
      "loss": 0.2217,
      "num_input_tokens_seen": 14465536,
      "step": 7510
    },
    {
      "epoch": 1.2259564401664083,
      "grad_norm": 0.0019315965473651886,
      "learning_rate": 0.2746287342621547,
      "loss": 0.2432,
      "num_input_tokens_seen": 14475152,
      "step": 7515
    },
    {
      "epoch": 1.2267721673872258,
      "grad_norm": 0.002497086301445961,
      "learning_rate": 0.2745959450313966,
      "loss": 0.2169,
      "num_input_tokens_seen": 14484720,
      "step": 7520
    },
    {
      "epoch": 1.227587894608043,
      "grad_norm": 0.0032041827216744423,
      "learning_rate": 0.27456313658637804,
      "loss": 0.2219,
      "num_input_tokens_seen": 14493744,
      "step": 7525
    },
    {
      "epoch": 1.2284036218288605,
      "grad_norm": 0.0029660221189260483,
      "learning_rate": 0.27453030893215846,
      "loss": 0.2123,
      "num_input_tokens_seen": 14503440,
      "step": 7530
    },
    {
      "epoch": 1.2292193490496777,
      "grad_norm": 0.001573150628246367,
      "learning_rate": 0.2744974620738003,
      "loss": 0.27,
      "num_input_tokens_seen": 14512432,
      "step": 7535
    },
    {
      "epoch": 1.2300350762704952,
      "grad_norm": 0.0019082268700003624,
      "learning_rate": 0.27446459601636897,
      "loss": 0.2206,
      "num_input_tokens_seen": 14522176,
      "step": 7540
    },
    {
      "epoch": 1.2308508034913126,
      "grad_norm": 0.0037939336616545916,
      "learning_rate": 0.2744317107649328,
      "loss": 0.22,
      "num_input_tokens_seen": 14531728,
      "step": 7545
    },
    {
      "epoch": 1.23166653071213,
      "grad_norm": 0.0016881616320461035,
      "learning_rate": 0.2743988063245631,
      "loss": 0.3062,
      "num_input_tokens_seen": 14541184,
      "step": 7550
    },
    {
      "epoch": 1.2324822579329473,
      "grad_norm": 0.0025261996779590845,
      "learning_rate": 0.2743658827003342,
      "loss": 0.2303,
      "num_input_tokens_seen": 14551232,
      "step": 7555
    },
    {
      "epoch": 1.2332979851537647,
      "grad_norm": 0.005453511606901884,
      "learning_rate": 0.27433293989732327,
      "loss": 0.239,
      "num_input_tokens_seen": 14561008,
      "step": 7560
    },
    {
      "epoch": 1.234113712374582,
      "grad_norm": 0.005678705405443907,
      "learning_rate": 0.27429997792061056,
      "loss": 0.2457,
      "num_input_tokens_seen": 14571264,
      "step": 7565
    },
    {
      "epoch": 1.2349294395953994,
      "grad_norm": 0.0077543118968605995,
      "learning_rate": 0.27426699677527927,
      "loss": 0.2324,
      "num_input_tokens_seen": 14580240,
      "step": 7570
    },
    {
      "epoch": 1.2357451668162167,
      "grad_norm": 0.003124658949673176,
      "learning_rate": 0.2742339964664154,
      "loss": 0.2257,
      "num_input_tokens_seen": 14591040,
      "step": 7575
    },
    {
      "epoch": 1.236560894037034,
      "grad_norm": 0.0041211433708667755,
      "learning_rate": 0.274200976999108,
      "loss": 0.208,
      "num_input_tokens_seen": 14600368,
      "step": 7580
    },
    {
      "epoch": 1.2373766212578514,
      "grad_norm": 0.004143411293625832,
      "learning_rate": 0.27416793837844916,
      "loss": 0.2665,
      "num_input_tokens_seen": 14609712,
      "step": 7585
    },
    {
      "epoch": 1.2381923484786688,
      "grad_norm": 0.005240399390459061,
      "learning_rate": 0.27413488060953384,
      "loss": 0.234,
      "num_input_tokens_seen": 14620128,
      "step": 7590
    },
    {
      "epoch": 1.2390080756994861,
      "grad_norm": 0.004349465016275644,
      "learning_rate": 0.27410180369745996,
      "loss": 0.2651,
      "num_input_tokens_seen": 14629488,
      "step": 7595
    },
    {
      "epoch": 1.2398238029203035,
      "grad_norm": 0.0018391602206975222,
      "learning_rate": 0.27406870764732844,
      "loss": 0.2061,
      "num_input_tokens_seen": 14638816,
      "step": 7600
    },
    {
      "epoch": 1.2398238029203035,
      "eval_loss": 0.2262483686208725,
      "eval_runtime": 68.0552,
      "eval_samples_per_second": 40.041,
      "eval_steps_per_second": 20.028,
      "num_input_tokens_seen": 14638816,
      "step": 7600
    },
    {
      "epoch": 1.2406395301411208,
      "grad_norm": 0.0027496409602463245,
      "learning_rate": 0.27403559246424297,
      "loss": 0.2299,
      "num_input_tokens_seen": 14648240,
      "step": 7605
    },
    {
      "epoch": 1.2414552573619382,
      "grad_norm": 0.002635001437738538,
      "learning_rate": 0.2740024581533105,
      "loss": 0.2367,
      "num_input_tokens_seen": 14657520,
      "step": 7610
    },
    {
      "epoch": 1.2422709845827555,
      "grad_norm": 0.00454211188480258,
      "learning_rate": 0.2739693047196406,
      "loss": 0.2768,
      "num_input_tokens_seen": 14665696,
      "step": 7615
    },
    {
      "epoch": 1.2430867118035729,
      "grad_norm": 0.0038179464172571898,
      "learning_rate": 0.27393613216834606,
      "loss": 0.2343,
      "num_input_tokens_seen": 14675392,
      "step": 7620
    },
    {
      "epoch": 1.2439024390243902,
      "grad_norm": 0.007346556521952152,
      "learning_rate": 0.2739029405045424,
      "loss": 0.2452,
      "num_input_tokens_seen": 14684496,
      "step": 7625
    },
    {
      "epoch": 1.2447181662452076,
      "grad_norm": 0.0033662801142781973,
      "learning_rate": 0.2738697297333483,
      "loss": 0.2306,
      "num_input_tokens_seen": 14693408,
      "step": 7630
    },
    {
      "epoch": 1.245533893466025,
      "grad_norm": 0.0026842504739761353,
      "learning_rate": 0.2738364998598852,
      "loss": 0.2319,
      "num_input_tokens_seen": 14703328,
      "step": 7635
    },
    {
      "epoch": 1.2463496206868423,
      "grad_norm": 0.005325646139681339,
      "learning_rate": 0.27380325088927765,
      "loss": 0.2281,
      "num_input_tokens_seen": 14713536,
      "step": 7640
    },
    {
      "epoch": 1.2471653479076596,
      "grad_norm": 0.0034498777240514755,
      "learning_rate": 0.27376998282665294,
      "loss": 0.2164,
      "num_input_tokens_seen": 14723856,
      "step": 7645
    },
    {
      "epoch": 1.247981075128477,
      "grad_norm": 0.002281180815771222,
      "learning_rate": 0.27373669567714154,
      "loss": 0.2353,
      "num_input_tokens_seen": 14734352,
      "step": 7650
    },
    {
      "epoch": 1.2487968023492944,
      "grad_norm": 0.0042783720418810844,
      "learning_rate": 0.27370338944587663,
      "loss": 0.3161,
      "num_input_tokens_seen": 14743168,
      "step": 7655
    },
    {
      "epoch": 1.2496125295701117,
      "grad_norm": 0.002050591865554452,
      "learning_rate": 0.27367006413799455,
      "loss": 0.2269,
      "num_input_tokens_seen": 14752480,
      "step": 7660
    },
    {
      "epoch": 1.250428256790929,
      "grad_norm": 0.003138219006359577,
      "learning_rate": 0.2736367197586345,
      "loss": 0.239,
      "num_input_tokens_seen": 14762096,
      "step": 7665
    },
    {
      "epoch": 1.2512439840117464,
      "grad_norm": 0.0025798517744988203,
      "learning_rate": 0.2736033563129385,
      "loss": 0.2472,
      "num_input_tokens_seen": 14773104,
      "step": 7670
    },
    {
      "epoch": 1.2520597112325638,
      "grad_norm": 0.0017434112960472703,
      "learning_rate": 0.27356997380605164,
      "loss": 0.2221,
      "num_input_tokens_seen": 14781968,
      "step": 7675
    },
    {
      "epoch": 1.2528754384533811,
      "grad_norm": 0.0013540061190724373,
      "learning_rate": 0.27353657224312194,
      "loss": 0.2074,
      "num_input_tokens_seen": 14792016,
      "step": 7680
    },
    {
      "epoch": 1.2536911656741985,
      "grad_norm": 0.0032076649367809296,
      "learning_rate": 0.2735031516293004,
      "loss": 0.2329,
      "num_input_tokens_seen": 14801440,
      "step": 7685
    },
    {
      "epoch": 1.2545068928950158,
      "grad_norm": 0.001908611156977713,
      "learning_rate": 0.2734697119697408,
      "loss": 0.2201,
      "num_input_tokens_seen": 14812496,
      "step": 7690
    },
    {
      "epoch": 1.2553226201158334,
      "grad_norm": 0.005528052803128958,
      "learning_rate": 0.27343625326959997,
      "loss": 0.2624,
      "num_input_tokens_seen": 14822144,
      "step": 7695
    },
    {
      "epoch": 1.2561383473366505,
      "grad_norm": 0.0026895359624177217,
      "learning_rate": 0.27340277553403775,
      "loss": 0.218,
      "num_input_tokens_seen": 14832000,
      "step": 7700
    },
    {
      "epoch": 1.256954074557468,
      "grad_norm": 0.0016545589314773679,
      "learning_rate": 0.2733692787682167,
      "loss": 0.2353,
      "num_input_tokens_seen": 14842064,
      "step": 7705
    },
    {
      "epoch": 1.2577698017782852,
      "grad_norm": 0.004286075942218304,
      "learning_rate": 0.27333576297730255,
      "loss": 0.2288,
      "num_input_tokens_seen": 14850880,
      "step": 7710
    },
    {
      "epoch": 1.2585855289991028,
      "grad_norm": 0.004952363204210997,
      "learning_rate": 0.2733022281664638,
      "loss": 0.2375,
      "num_input_tokens_seen": 14859456,
      "step": 7715
    },
    {
      "epoch": 1.25940125621992,
      "grad_norm": 0.003880498232319951,
      "learning_rate": 0.273268674340872,
      "loss": 0.2159,
      "num_input_tokens_seen": 14870112,
      "step": 7720
    },
    {
      "epoch": 1.2602169834407375,
      "grad_norm": 0.001773497206158936,
      "learning_rate": 0.27323510150570146,
      "loss": 0.2291,
      "num_input_tokens_seen": 14879088,
      "step": 7725
    },
    {
      "epoch": 1.2610327106615546,
      "grad_norm": 0.0015541069442406297,
      "learning_rate": 0.27320150966612966,
      "loss": 0.2545,
      "num_input_tokens_seen": 14888560,
      "step": 7730
    },
    {
      "epoch": 1.2618484378823722,
      "grad_norm": 0.0015136320143938065,
      "learning_rate": 0.2731678988273368,
      "loss": 0.2369,
      "num_input_tokens_seen": 14898720,
      "step": 7735
    },
    {
      "epoch": 1.2626641651031896,
      "grad_norm": 0.0030190900433808565,
      "learning_rate": 0.27313426899450605,
      "loss": 0.2466,
      "num_input_tokens_seen": 14908176,
      "step": 7740
    },
    {
      "epoch": 1.263479892324007,
      "grad_norm": 0.0018212745198979974,
      "learning_rate": 0.27310062017282366,
      "loss": 0.2232,
      "num_input_tokens_seen": 14918736,
      "step": 7745
    },
    {
      "epoch": 1.2642956195448243,
      "grad_norm": 0.001357552013359964,
      "learning_rate": 0.2730669523674787,
      "loss": 0.2219,
      "num_input_tokens_seen": 14926784,
      "step": 7750
    },
    {
      "epoch": 1.2651113467656416,
      "grad_norm": 0.005230552516877651,
      "learning_rate": 0.2730332655836631,
      "loss": 0.2245,
      "num_input_tokens_seen": 14936688,
      "step": 7755
    },
    {
      "epoch": 1.265927073986459,
      "grad_norm": 0.0029117451049387455,
      "learning_rate": 0.2729995598265718,
      "loss": 0.2272,
      "num_input_tokens_seen": 14946112,
      "step": 7760
    },
    {
      "epoch": 1.2667428012072763,
      "grad_norm": 0.00196373974904418,
      "learning_rate": 0.2729658351014027,
      "loss": 0.243,
      "num_input_tokens_seen": 14955904,
      "step": 7765
    },
    {
      "epoch": 1.2675585284280937,
      "grad_norm": 0.0017864431720227003,
      "learning_rate": 0.27293209141335656,
      "loss": 0.2204,
      "num_input_tokens_seen": 14965600,
      "step": 7770
    },
    {
      "epoch": 1.268374255648911,
      "grad_norm": 0.002746789250522852,
      "learning_rate": 0.27289832876763703,
      "loss": 0.226,
      "num_input_tokens_seen": 14977296,
      "step": 7775
    },
    {
      "epoch": 1.2691899828697284,
      "grad_norm": 0.002135898219421506,
      "learning_rate": 0.27286454716945074,
      "loss": 0.2407,
      "num_input_tokens_seen": 14986496,
      "step": 7780
    },
    {
      "epoch": 1.2700057100905457,
      "grad_norm": 0.003016109811142087,
      "learning_rate": 0.27283074662400725,
      "loss": 0.2159,
      "num_input_tokens_seen": 14996256,
      "step": 7785
    },
    {
      "epoch": 1.270821437311363,
      "grad_norm": 0.003320625750347972,
      "learning_rate": 0.2727969271365191,
      "loss": 0.2139,
      "num_input_tokens_seen": 15006032,
      "step": 7790
    },
    {
      "epoch": 1.2716371645321805,
      "grad_norm": 0.003703972091898322,
      "learning_rate": 0.2727630887122016,
      "loss": 0.2334,
      "num_input_tokens_seen": 15015216,
      "step": 7795
    },
    {
      "epoch": 1.2724528917529978,
      "grad_norm": 0.0027978122234344482,
      "learning_rate": 0.27272923135627314,
      "loss": 0.2402,
      "num_input_tokens_seen": 15024560,
      "step": 7800
    },
    {
      "epoch": 1.2724528917529978,
      "eval_loss": 0.2247651219367981,
      "eval_runtime": 68.1538,
      "eval_samples_per_second": 39.983,
      "eval_steps_per_second": 19.999,
      "num_input_tokens_seen": 15024560,
      "step": 7800
    },
    {
      "epoch": 1.2732686189738152,
      "grad_norm": 0.0030850847251713276,
      "learning_rate": 0.2726953550739548,
      "loss": 0.2472,
      "num_input_tokens_seen": 15034032,
      "step": 7805
    },
    {
      "epoch": 1.2740843461946325,
      "grad_norm": 0.0026968736201524734,
      "learning_rate": 0.27266145987047086,
      "loss": 0.2315,
      "num_input_tokens_seen": 15043456,
      "step": 7810
    },
    {
      "epoch": 1.2749000734154499,
      "grad_norm": 0.0014417540514841676,
      "learning_rate": 0.27262754575104836,
      "loss": 0.2176,
      "num_input_tokens_seen": 15052496,
      "step": 7815
    },
    {
      "epoch": 1.2757158006362672,
      "grad_norm": 0.0019233248895034194,
      "learning_rate": 0.27259361272091726,
      "loss": 0.2296,
      "num_input_tokens_seen": 15063120,
      "step": 7820
    },
    {
      "epoch": 1.2765315278570846,
      "grad_norm": 0.003346566576510668,
      "learning_rate": 0.27255966078531046,
      "loss": 0.2514,
      "num_input_tokens_seen": 15072192,
      "step": 7825
    },
    {
      "epoch": 1.277347255077902,
      "grad_norm": 0.0018194842850789428,
      "learning_rate": 0.2725256899494638,
      "loss": 0.2218,
      "num_input_tokens_seen": 15082784,
      "step": 7830
    },
    {
      "epoch": 1.2781629822987193,
      "grad_norm": 0.0027142404578626156,
      "learning_rate": 0.272491700218616,
      "loss": 0.2211,
      "num_input_tokens_seen": 15093520,
      "step": 7835
    },
    {
      "epoch": 1.2789787095195366,
      "grad_norm": 0.003752127755433321,
      "learning_rate": 0.27245769159800876,
      "loss": 0.2202,
      "num_input_tokens_seen": 15103696,
      "step": 7840
    },
    {
      "epoch": 1.279794436740354,
      "grad_norm": 0.004137747921049595,
      "learning_rate": 0.2724236640928865,
      "loss": 0.2094,
      "num_input_tokens_seen": 15113072,
      "step": 7845
    },
    {
      "epoch": 1.2806101639611713,
      "grad_norm": 0.002468005521222949,
      "learning_rate": 0.27238961770849673,
      "loss": 0.2617,
      "num_input_tokens_seen": 15122272,
      "step": 7850
    },
    {
      "epoch": 1.2814258911819887,
      "grad_norm": 0.003076426684856415,
      "learning_rate": 0.27235555245008997,
      "loss": 0.2185,
      "num_input_tokens_seen": 15131232,
      "step": 7855
    },
    {
      "epoch": 1.282241618402806,
      "grad_norm": 0.004243469797074795,
      "learning_rate": 0.2723214683229193,
      "loss": 0.2236,
      "num_input_tokens_seen": 15140384,
      "step": 7860
    },
    {
      "epoch": 1.2830573456236234,
      "grad_norm": 0.0018509491346776485,
      "learning_rate": 0.27228736533224107,
      "loss": 0.2404,
      "num_input_tokens_seen": 15150336,
      "step": 7865
    },
    {
      "epoch": 1.2838730728444407,
      "grad_norm": 0.0032696358393877745,
      "learning_rate": 0.27225324348331437,
      "loss": 0.2604,
      "num_input_tokens_seen": 15159408,
      "step": 7870
    },
    {
      "epoch": 1.284688800065258,
      "grad_norm": 0.003390601836144924,
      "learning_rate": 0.27221910278140116,
      "loss": 0.2431,
      "num_input_tokens_seen": 15168288,
      "step": 7875
    },
    {
      "epoch": 1.2855045272860757,
      "grad_norm": 0.002811052603647113,
      "learning_rate": 0.2721849432317664,
      "loss": 0.2427,
      "num_input_tokens_seen": 15178576,
      "step": 7880
    },
    {
      "epoch": 1.2863202545068928,
      "grad_norm": 0.005559599492698908,
      "learning_rate": 0.2721507648396779,
      "loss": 0.2453,
      "num_input_tokens_seen": 15188240,
      "step": 7885
    },
    {
      "epoch": 1.2871359817277104,
      "grad_norm": 0.005793185904622078,
      "learning_rate": 0.27211656761040653,
      "loss": 0.2354,
      "num_input_tokens_seen": 15198448,
      "step": 7890
    },
    {
      "epoch": 1.2879517089485275,
      "grad_norm": 0.003590843640267849,
      "learning_rate": 0.2720823515492257,
      "loss": 0.2317,
      "num_input_tokens_seen": 15208368,
      "step": 7895
    },
    {
      "epoch": 1.288767436169345,
      "grad_norm": 0.005590815097093582,
      "learning_rate": 0.27204811666141215,
      "loss": 0.2348,
      "num_input_tokens_seen": 15217840,
      "step": 7900
    },
    {
      "epoch": 1.2895831633901622,
      "grad_norm": 0.00306148873642087,
      "learning_rate": 0.2720138629522452,
      "loss": 0.2465,
      "num_input_tokens_seen": 15227184,
      "step": 7905
    },
    {
      "epoch": 1.2903988906109798,
      "grad_norm": 0.0027979447040706873,
      "learning_rate": 0.2719795904270073,
      "loss": 0.2283,
      "num_input_tokens_seen": 15237136,
      "step": 7910
    },
    {
      "epoch": 1.291214617831797,
      "grad_norm": 0.005080769304186106,
      "learning_rate": 0.2719452990909837,
      "loss": 0.2326,
      "num_input_tokens_seen": 15247600,
      "step": 7915
    },
    {
      "epoch": 1.2920303450526145,
      "grad_norm": 0.0057789976708590984,
      "learning_rate": 0.2719109889494625,
      "loss": 0.2315,
      "num_input_tokens_seen": 15256976,
      "step": 7920
    },
    {
      "epoch": 1.2928460722734318,
      "grad_norm": 0.005711480975151062,
      "learning_rate": 0.27187666000773475,
      "loss": 0.2399,
      "num_input_tokens_seen": 15266992,
      "step": 7925
    },
    {
      "epoch": 1.2936617994942492,
      "grad_norm": 0.024814652279019356,
      "learning_rate": 0.2718423122710944,
      "loss": 0.3146,
      "num_input_tokens_seen": 15276416,
      "step": 7930
    },
    {
      "epoch": 1.2944775267150666,
      "grad_norm": 0.014101771637797356,
      "learning_rate": 0.2718079457448384,
      "loss": 0.2414,
      "num_input_tokens_seen": 15286352,
      "step": 7935
    },
    {
      "epoch": 1.295293253935884,
      "grad_norm": 0.033974722027778625,
      "learning_rate": 0.27177356043426637,
      "loss": 0.2998,
      "num_input_tokens_seen": 15295616,
      "step": 7940
    },
    {
      "epoch": 1.2961089811567013,
      "grad_norm": 0.01690865308046341,
      "learning_rate": 0.27173915634468104,
      "loss": 0.2709,
      "num_input_tokens_seen": 15304704,
      "step": 7945
    },
    {
      "epoch": 1.2969247083775186,
      "grad_norm": 0.015345118939876556,
      "learning_rate": 0.27170473348138796,
      "loss": 0.3426,
      "num_input_tokens_seen": 15315168,
      "step": 7950
    },
    {
      "epoch": 1.297740435598336,
      "grad_norm": 0.01654205098748207,
      "learning_rate": 0.27167029184969554,
      "loss": 0.273,
      "num_input_tokens_seen": 15324912,
      "step": 7955
    },
    {
      "epoch": 1.2985561628191533,
      "grad_norm": 0.028984468430280685,
      "learning_rate": 0.27163583145491504,
      "loss": 0.296,
      "num_input_tokens_seen": 15335152,
      "step": 7960
    },
    {
      "epoch": 1.2993718900399707,
      "grad_norm": 0.13199490308761597,
      "learning_rate": 0.2716013523023608,
      "loss": 0.4031,
      "num_input_tokens_seen": 15345936,
      "step": 7965
    },
    {
      "epoch": 1.300187617260788,
      "grad_norm": 0.015362458303570747,
      "learning_rate": 0.27156685439734995,
      "loss": 0.3283,
      "num_input_tokens_seen": 15355376,
      "step": 7970
    },
    {
      "epoch": 1.3010033444816054,
      "grad_norm": 0.017133528366684914,
      "learning_rate": 0.2715323377452024,
      "loss": 0.2808,
      "num_input_tokens_seen": 15364304,
      "step": 7975
    },
    {
      "epoch": 1.3018190717024227,
      "grad_norm": 0.04358282312750816,
      "learning_rate": 0.2714978023512411,
      "loss": 0.308,
      "num_input_tokens_seen": 15374336,
      "step": 7980
    },
    {
      "epoch": 1.30263479892324,
      "grad_norm": 0.028873147442936897,
      "learning_rate": 0.2714632482207918,
      "loss": 0.2806,
      "num_input_tokens_seen": 15384128,
      "step": 7985
    },
    {
      "epoch": 1.3034505261440574,
      "grad_norm": 0.0110600171610713,
      "learning_rate": 0.2714286753591833,
      "loss": 0.3116,
      "num_input_tokens_seen": 15394496,
      "step": 7990
    },
    {
      "epoch": 1.3042662533648748,
      "grad_norm": 0.013616582378745079,
      "learning_rate": 0.27139408377174706,
      "loss": 0.2508,
      "num_input_tokens_seen": 15402992,
      "step": 7995
    },
    {
      "epoch": 1.3050819805856921,
      "grad_norm": 0.011976636946201324,
      "learning_rate": 0.27135947346381756,
      "loss": 0.262,
      "num_input_tokens_seen": 15412000,
      "step": 8000
    },
    {
      "epoch": 1.3050819805856921,
      "eval_loss": 0.24048493802547455,
      "eval_runtime": 68.0971,
      "eval_samples_per_second": 40.016,
      "eval_steps_per_second": 20.016,
      "num_input_tokens_seen": 15412000,
      "step": 8000
    },
    {
      "epoch": 1.3058977078065095,
      "grad_norm": 0.02486364170908928,
      "learning_rate": 0.2713248444407322,
      "loss": 0.2432,
      "num_input_tokens_seen": 15421504,
      "step": 8005
    },
    {
      "epoch": 1.3067134350273268,
      "grad_norm": 0.017769666388630867,
      "learning_rate": 0.27129019670783106,
      "loss": 0.229,
      "num_input_tokens_seen": 15429856,
      "step": 8010
    },
    {
      "epoch": 1.3075291622481442,
      "grad_norm": 0.019436001777648926,
      "learning_rate": 0.27125553027045746,
      "loss": 0.268,
      "num_input_tokens_seen": 15439712,
      "step": 8015
    },
    {
      "epoch": 1.3083448894689615,
      "grad_norm": 0.0130903460085392,
      "learning_rate": 0.2712208451339572,
      "loss": 0.2402,
      "num_input_tokens_seen": 15449888,
      "step": 8020
    },
    {
      "epoch": 1.309160616689779,
      "grad_norm": 0.012301189824938774,
      "learning_rate": 0.27118614130367935,
      "loss": 0.2401,
      "num_input_tokens_seen": 15459440,
      "step": 8025
    },
    {
      "epoch": 1.3099763439105963,
      "grad_norm": 0.008426617830991745,
      "learning_rate": 0.2711514187849756,
      "loss": 0.2712,
      "num_input_tokens_seen": 15468704,
      "step": 8030
    },
    {
      "epoch": 1.3107920711314136,
      "grad_norm": 0.003974124789237976,
      "learning_rate": 0.27111667758320057,
      "loss": 0.2587,
      "num_input_tokens_seen": 15477984,
      "step": 8035
    },
    {
      "epoch": 1.311607798352231,
      "grad_norm": 0.010850759223103523,
      "learning_rate": 0.27108191770371176,
      "loss": 0.22,
      "num_input_tokens_seen": 15485520,
      "step": 8040
    },
    {
      "epoch": 1.3124235255730483,
      "grad_norm": 0.009812718257308006,
      "learning_rate": 0.2710471391518697,
      "loss": 0.2358,
      "num_input_tokens_seen": 15494592,
      "step": 8045
    },
    {
      "epoch": 1.3132392527938657,
      "grad_norm": 0.0038970336318016052,
      "learning_rate": 0.2710123419330375,
      "loss": 0.2347,
      "num_input_tokens_seen": 15503136,
      "step": 8050
    },
    {
      "epoch": 1.314054980014683,
      "grad_norm": 0.005041956435889006,
      "learning_rate": 0.2709775260525816,
      "loss": 0.2449,
      "num_input_tokens_seen": 15511856,
      "step": 8055
    },
    {
      "epoch": 1.3148707072355004,
      "grad_norm": 0.005855274386703968,
      "learning_rate": 0.27094269151587075,
      "loss": 0.271,
      "num_input_tokens_seen": 15520880,
      "step": 8060
    },
    {
      "epoch": 1.315686434456318,
      "grad_norm": 0.0038751084357500076,
      "learning_rate": 0.27090783832827703,
      "loss": 0.2229,
      "num_input_tokens_seen": 15531472,
      "step": 8065
    },
    {
      "epoch": 1.316502161677135,
      "grad_norm": 0.004093600902706385,
      "learning_rate": 0.2708729664951753,
      "loss": 0.2172,
      "num_input_tokens_seen": 15540640,
      "step": 8070
    },
    {
      "epoch": 1.3173178888979526,
      "grad_norm": 0.0014585425378754735,
      "learning_rate": 0.27083807602194304,
      "loss": 0.1717,
      "num_input_tokens_seen": 15550496,
      "step": 8075
    },
    {
      "epoch": 1.3181336161187698,
      "grad_norm": 0.002987944521009922,
      "learning_rate": 0.270803166913961,
      "loss": 0.2826,
      "num_input_tokens_seen": 15559856,
      "step": 8080
    },
    {
      "epoch": 1.3189493433395874,
      "grad_norm": 0.003381680231541395,
      "learning_rate": 0.27076823917661247,
      "loss": 0.2148,
      "num_input_tokens_seen": 15568880,
      "step": 8085
    },
    {
      "epoch": 1.3197650705604045,
      "grad_norm": 0.002526611555367708,
      "learning_rate": 0.2707332928152838,
      "loss": 0.256,
      "num_input_tokens_seen": 15578944,
      "step": 8090
    },
    {
      "epoch": 1.320580797781222,
      "grad_norm": 0.006407811306416988,
      "learning_rate": 0.2706983278353641,
      "loss": 0.2505,
      "num_input_tokens_seen": 15587600,
      "step": 8095
    },
    {
      "epoch": 1.3213965250020392,
      "grad_norm": 0.00586236035451293,
      "learning_rate": 0.27066334424224553,
      "loss": 0.2538,
      "num_input_tokens_seen": 15597264,
      "step": 8100
    },
    {
      "epoch": 1.3222122522228568,
      "grad_norm": 0.00483299046754837,
      "learning_rate": 0.27062834204132297,
      "loss": 0.2416,
      "num_input_tokens_seen": 15607392,
      "step": 8105
    },
    {
      "epoch": 1.3230279794436741,
      "grad_norm": 0.00430904608219862,
      "learning_rate": 0.27059332123799407,
      "loss": 0.2619,
      "num_input_tokens_seen": 15616048,
      "step": 8110
    },
    {
      "epoch": 1.3238437066644915,
      "grad_norm": 0.006013525649905205,
      "learning_rate": 0.27055828183765956,
      "loss": 0.2566,
      "num_input_tokens_seen": 15625488,
      "step": 8115
    },
    {
      "epoch": 1.3246594338853088,
      "grad_norm": 0.0045315856114029884,
      "learning_rate": 0.270523223845723,
      "loss": 0.2405,
      "num_input_tokens_seen": 15634000,
      "step": 8120
    },
    {
      "epoch": 1.3254751611061262,
      "grad_norm": 0.00421555619686842,
      "learning_rate": 0.2704881472675907,
      "loss": 0.2577,
      "num_input_tokens_seen": 15644464,
      "step": 8125
    },
    {
      "epoch": 1.3262908883269435,
      "grad_norm": 0.0038324762135744095,
      "learning_rate": 0.270453052108672,
      "loss": 0.2131,
      "num_input_tokens_seen": 15653488,
      "step": 8130
    },
    {
      "epoch": 1.3271066155477609,
      "grad_norm": 0.007571733556687832,
      "learning_rate": 0.2704179383743789,
      "loss": 0.2541,
      "num_input_tokens_seen": 15664512,
      "step": 8135
    },
    {
      "epoch": 1.3279223427685782,
      "grad_norm": 0.00610905047506094,
      "learning_rate": 0.27038280607012644,
      "loss": 0.2324,
      "num_input_tokens_seen": 15674400,
      "step": 8140
    },
    {
      "epoch": 1.3287380699893956,
      "grad_norm": 0.002771607134491205,
      "learning_rate": 0.27034765520133247,
      "loss": 0.2593,
      "num_input_tokens_seen": 15683392,
      "step": 8145
    },
    {
      "epoch": 1.329553797210213,
      "grad_norm": 0.0067487116903066635,
      "learning_rate": 0.2703124857734177,
      "loss": 0.2603,
      "num_input_tokens_seen": 15692848,
      "step": 8150
    },
    {
      "epoch": 1.3303695244310303,
      "grad_norm": 0.008084307424724102,
      "learning_rate": 0.27027729779180565,
      "loss": 0.2477,
      "num_input_tokens_seen": 15703088,
      "step": 8155
    },
    {
      "epoch": 1.3311852516518476,
      "grad_norm": 0.006026018410921097,
      "learning_rate": 0.27024209126192283,
      "loss": 0.2617,
      "num_input_tokens_seen": 15712800,
      "step": 8160
    },
    {
      "epoch": 1.332000978872665,
      "grad_norm": 0.003991578705608845,
      "learning_rate": 0.2702068661891984,
      "loss": 0.2191,
      "num_input_tokens_seen": 15721664,
      "step": 8165
    },
    {
      "epoch": 1.3328167060934824,
      "grad_norm": 0.004080517683178186,
      "learning_rate": 0.2701716225790647,
      "loss": 0.2795,
      "num_input_tokens_seen": 15731328,
      "step": 8170
    },
    {
      "epoch": 1.3336324333142997,
      "grad_norm": 0.004250073805451393,
      "learning_rate": 0.27013636043695655,
      "loss": 0.2547,
      "num_input_tokens_seen": 15740448,
      "step": 8175
    },
    {
      "epoch": 1.334448160535117,
      "grad_norm": 0.0029597687534987926,
      "learning_rate": 0.27010107976831194,
      "loss": 0.2253,
      "num_input_tokens_seen": 15750208,
      "step": 8180
    },
    {
      "epoch": 1.3352638877559344,
      "grad_norm": 0.0029400018975138664,
      "learning_rate": 0.2700657805785715,
      "loss": 0.231,
      "num_input_tokens_seen": 15760384,
      "step": 8185
    },
    {
      "epoch": 1.3360796149767518,
      "grad_norm": 0.002990720560774207,
      "learning_rate": 0.2700304628731789,
      "loss": 0.212,
      "num_input_tokens_seen": 15770320,
      "step": 8190
    },
    {
      "epoch": 1.3368953421975691,
      "grad_norm": 0.0025358775164932013,
      "learning_rate": 0.26999512665758046,
      "loss": 0.2132,
      "num_input_tokens_seen": 15779632,
      "step": 8195
    },
    {
      "epoch": 1.3377110694183865,
      "grad_norm": 0.0036217086017131805,
      "learning_rate": 0.2699597719372256,
      "loss": 0.2524,
      "num_input_tokens_seen": 15789456,
      "step": 8200
    },
    {
      "epoch": 1.3377110694183865,
      "eval_loss": 0.2334071397781372,
      "eval_runtime": 68.1368,
      "eval_samples_per_second": 39.993,
      "eval_steps_per_second": 20.004,
      "num_input_tokens_seen": 15789456,
      "step": 8200
    },
    {
      "epoch": 1.3385267966392038,
      "grad_norm": 0.004394848830997944,
      "learning_rate": 0.26992439871756635,
      "loss": 0.2077,
      "num_input_tokens_seen": 15798384,
      "step": 8205
    },
    {
      "epoch": 1.3393425238600212,
      "grad_norm": 0.0068792481906712055,
      "learning_rate": 0.2698890070040578,
      "loss": 0.208,
      "num_input_tokens_seen": 15807008,
      "step": 8210
    },
    {
      "epoch": 1.3401582510808385,
      "grad_norm": 0.004957845434546471,
      "learning_rate": 0.2698535968021577,
      "loss": 0.2751,
      "num_input_tokens_seen": 15816640,
      "step": 8215
    },
    {
      "epoch": 1.3409739783016559,
      "grad_norm": 0.004083022475242615,
      "learning_rate": 0.26981816811732684,
      "loss": 0.2296,
      "num_input_tokens_seen": 15828048,
      "step": 8220
    },
    {
      "epoch": 1.3417897055224732,
      "grad_norm": 0.00893857330083847,
      "learning_rate": 0.26978272095502875,
      "loss": 0.297,
      "num_input_tokens_seen": 15836976,
      "step": 8225
    },
    {
      "epoch": 1.3426054327432906,
      "grad_norm": 0.008694800548255444,
      "learning_rate": 0.26974725532072974,
      "loss": 0.2977,
      "num_input_tokens_seen": 15846368,
      "step": 8230
    },
    {
      "epoch": 1.343421159964108,
      "grad_norm": 0.001961581874638796,
      "learning_rate": 0.26971177121989914,
      "loss": 0.2042,
      "num_input_tokens_seen": 15856496,
      "step": 8235
    },
    {
      "epoch": 1.3442368871849253,
      "grad_norm": 0.0033708231057971716,
      "learning_rate": 0.2696762686580091,
      "loss": 0.326,
      "num_input_tokens_seen": 15866528,
      "step": 8240
    },
    {
      "epoch": 1.3450526144057426,
      "grad_norm": 0.004713460803031921,
      "learning_rate": 0.26964074764053436,
      "loss": 0.3172,
      "num_input_tokens_seen": 15875936,
      "step": 8245
    },
    {
      "epoch": 1.3458683416265602,
      "grad_norm": 0.001700917142443359,
      "learning_rate": 0.2696052081729529,
      "loss": 0.2413,
      "num_input_tokens_seen": 15884560,
      "step": 8250
    },
    {
      "epoch": 1.3466840688473773,
      "grad_norm": 0.006739374715834856,
      "learning_rate": 0.2695696502607453,
      "loss": 0.2889,
      "num_input_tokens_seen": 15893968,
      "step": 8255
    },
    {
      "epoch": 1.347499796068195,
      "grad_norm": 0.003197186393663287,
      "learning_rate": 0.26953407390939504,
      "loss": 0.2665,
      "num_input_tokens_seen": 15903776,
      "step": 8260
    },
    {
      "epoch": 1.348315523289012,
      "grad_norm": 0.00279553746804595,
      "learning_rate": 0.26949847912438835,
      "loss": 0.2514,
      "num_input_tokens_seen": 15912208,
      "step": 8265
    },
    {
      "epoch": 1.3491312505098296,
      "grad_norm": 0.0017273295670747757,
      "learning_rate": 0.26946286591121454,
      "loss": 0.2566,
      "num_input_tokens_seen": 15922832,
      "step": 8270
    },
    {
      "epoch": 1.3499469777306468,
      "grad_norm": 0.0038956953212618828,
      "learning_rate": 0.2694272342753655,
      "loss": 0.3132,
      "num_input_tokens_seen": 15933520,
      "step": 8275
    },
    {
      "epoch": 1.3507627049514643,
      "grad_norm": 0.0020475813653320074,
      "learning_rate": 0.26939158422233617,
      "loss": 0.185,
      "num_input_tokens_seen": 15943296,
      "step": 8280
    },
    {
      "epoch": 1.3515784321722815,
      "grad_norm": 0.0022828083019703627,
      "learning_rate": 0.26935591575762413,
      "loss": 0.2287,
      "num_input_tokens_seen": 15952336,
      "step": 8285
    },
    {
      "epoch": 1.352394159393099,
      "grad_norm": 0.005713772028684616,
      "learning_rate": 0.26932022888672996,
      "loss": 0.2521,
      "num_input_tokens_seen": 15961648,
      "step": 8290
    },
    {
      "epoch": 1.3532098866139162,
      "grad_norm": 0.003554785856977105,
      "learning_rate": 0.26928452361515703,
      "loss": 0.225,
      "num_input_tokens_seen": 15971072,
      "step": 8295
    },
    {
      "epoch": 1.3540256138347337,
      "grad_norm": 0.006205619778484106,
      "learning_rate": 0.26924879994841155,
      "loss": 0.2434,
      "num_input_tokens_seen": 15980128,
      "step": 8300
    },
    {
      "epoch": 1.354841341055551,
      "grad_norm": 0.003561254357919097,
      "learning_rate": 0.2692130578920025,
      "loss": 0.2517,
      "num_input_tokens_seen": 15989584,
      "step": 8305
    },
    {
      "epoch": 1.3556570682763684,
      "grad_norm": 0.0033786753192543983,
      "learning_rate": 0.26917729745144187,
      "loss": 0.1761,
      "num_input_tokens_seen": 15998880,
      "step": 8310
    },
    {
      "epoch": 1.3564727954971858,
      "grad_norm": 0.0048704808577895164,
      "learning_rate": 0.2691415186322443,
      "loss": 0.2867,
      "num_input_tokens_seen": 16009104,
      "step": 8315
    },
    {
      "epoch": 1.3572885227180032,
      "grad_norm": 0.0027775901835411787,
      "learning_rate": 0.2691057214399273,
      "loss": 0.2643,
      "num_input_tokens_seen": 16018640,
      "step": 8320
    },
    {
      "epoch": 1.3581042499388205,
      "grad_norm": 0.0022914260625839233,
      "learning_rate": 0.2690699058800113,
      "loss": 0.2841,
      "num_input_tokens_seen": 16028688,
      "step": 8325
    },
    {
      "epoch": 1.3589199771596379,
      "grad_norm": 0.0023662084713578224,
      "learning_rate": 0.2690340719580194,
      "loss": 0.2279,
      "num_input_tokens_seen": 16039232,
      "step": 8330
    },
    {
      "epoch": 1.3597357043804552,
      "grad_norm": 0.0019008436938747764,
      "learning_rate": 0.2689982196794778,
      "loss": 0.2399,
      "num_input_tokens_seen": 16049536,
      "step": 8335
    },
    {
      "epoch": 1.3605514316012726,
      "grad_norm": 0.0018884919118136168,
      "learning_rate": 0.2689623490499153,
      "loss": 0.2432,
      "num_input_tokens_seen": 16060032,
      "step": 8340
    },
    {
      "epoch": 1.36136715882209,
      "grad_norm": 0.0019650664180517197,
      "learning_rate": 0.2689264600748636,
      "loss": 0.2403,
      "num_input_tokens_seen": 16070064,
      "step": 8345
    },
    {
      "epoch": 1.3621828860429073,
      "grad_norm": 0.004118216689676046,
      "learning_rate": 0.26889055275985724,
      "loss": 0.2416,
      "num_input_tokens_seen": 16079408,
      "step": 8350
    },
    {
      "epoch": 1.3629986132637246,
      "grad_norm": 0.004316916689276695,
      "learning_rate": 0.2688546271104335,
      "loss": 0.2424,
      "num_input_tokens_seen": 16087440,
      "step": 8355
    },
    {
      "epoch": 1.363814340484542,
      "grad_norm": 0.0018819107208400965,
      "learning_rate": 0.26881868313213275,
      "loss": 0.2292,
      "num_input_tokens_seen": 16097584,
      "step": 8360
    },
    {
      "epoch": 1.3646300677053593,
      "grad_norm": 0.003356178756803274,
      "learning_rate": 0.2687827208304978,
      "loss": 0.2434,
      "num_input_tokens_seen": 16106864,
      "step": 8365
    },
    {
      "epoch": 1.3654457949261767,
      "grad_norm": 0.0029730708338320255,
      "learning_rate": 0.26874674021107464,
      "loss": 0.2409,
      "num_input_tokens_seen": 16116656,
      "step": 8370
    },
    {
      "epoch": 1.366261522146994,
      "grad_norm": 0.003745361929759383,
      "learning_rate": 0.2687107412794118,
      "loss": 0.2347,
      "num_input_tokens_seen": 16125776,
      "step": 8375
    },
    {
      "epoch": 1.3670772493678114,
      "grad_norm": 0.0021466228645294905,
      "learning_rate": 0.26867472404106096,
      "loss": 0.2343,
      "num_input_tokens_seen": 16135216,
      "step": 8380
    },
    {
      "epoch": 1.3678929765886287,
      "grad_norm": 0.0021865745075047016,
      "learning_rate": 0.26863868850157624,
      "loss": 0.2424,
      "num_input_tokens_seen": 16144080,
      "step": 8385
    },
    {
      "epoch": 1.368708703809446,
      "grad_norm": 0.004883565474301577,
      "learning_rate": 0.26860263466651485,
      "loss": 0.2385,
      "num_input_tokens_seen": 16153392,
      "step": 8390
    },
    {
      "epoch": 1.3695244310302634,
      "grad_norm": 0.004027844872325659,
      "learning_rate": 0.26856656254143674,
      "loss": 0.2455,
      "num_input_tokens_seen": 16162784,
      "step": 8395
    },
    {
      "epoch": 1.3703401582510808,
      "grad_norm": 0.002509942976757884,
      "learning_rate": 0.2685304721319047,
      "loss": 0.2249,
      "num_input_tokens_seen": 16173616,
      "step": 8400
    },
    {
      "epoch": 1.3703401582510808,
      "eval_loss": 0.2285296469926834,
      "eval_runtime": 68.2345,
      "eval_samples_per_second": 39.936,
      "eval_steps_per_second": 19.975,
      "num_input_tokens_seen": 16173616,
      "step": 8400
    },
    {
      "epoch": 1.3711558854718982,
      "grad_norm": 0.001784664811566472,
      "learning_rate": 0.2684943634434843,
      "loss": 0.2103,
      "num_input_tokens_seen": 16183520,
      "step": 8405
    },
    {
      "epoch": 1.3719716126927155,
      "grad_norm": 0.002871142700314522,
      "learning_rate": 0.268458236481744,
      "loss": 0.2629,
      "num_input_tokens_seen": 16192192,
      "step": 8410
    },
    {
      "epoch": 1.3727873399135329,
      "grad_norm": 0.002106844913214445,
      "learning_rate": 0.2684220912522549,
      "loss": 0.2492,
      "num_input_tokens_seen": 16201888,
      "step": 8415
    },
    {
      "epoch": 1.3736030671343502,
      "grad_norm": 0.0018340088427066803,
      "learning_rate": 0.2683859277605913,
      "loss": 0.225,
      "num_input_tokens_seen": 16211760,
      "step": 8420
    },
    {
      "epoch": 1.3744187943551676,
      "grad_norm": 0.003567741485312581,
      "learning_rate": 0.2683497460123298,
      "loss": 0.2381,
      "num_input_tokens_seen": 16222256,
      "step": 8425
    },
    {
      "epoch": 1.375234521575985,
      "grad_norm": 0.0029552082996815443,
      "learning_rate": 0.26831354601305013,
      "loss": 0.2187,
      "num_input_tokens_seen": 16231856,
      "step": 8430
    },
    {
      "epoch": 1.3760502487968023,
      "grad_norm": 0.001680054352618754,
      "learning_rate": 0.26827732776833496,
      "loss": 0.2595,
      "num_input_tokens_seen": 16240752,
      "step": 8435
    },
    {
      "epoch": 1.3768659760176196,
      "grad_norm": 0.0038798744790256023,
      "learning_rate": 0.26824109128376944,
      "loss": 0.2212,
      "num_input_tokens_seen": 16250960,
      "step": 8440
    },
    {
      "epoch": 1.3776817032384372,
      "grad_norm": 0.004859786946326494,
      "learning_rate": 0.2682048365649417,
      "loss": 0.2369,
      "num_input_tokens_seen": 16260016,
      "step": 8445
    },
    {
      "epoch": 1.3784974304592543,
      "grad_norm": 0.005123807583004236,
      "learning_rate": 0.2681685636174428,
      "loss": 0.2613,
      "num_input_tokens_seen": 16270304,
      "step": 8450
    },
    {
      "epoch": 1.379313157680072,
      "grad_norm": 0.004000545479357243,
      "learning_rate": 0.2681322724468663,
      "loss": 0.2941,
      "num_input_tokens_seen": 16279664,
      "step": 8455
    },
    {
      "epoch": 1.380128884900889,
      "grad_norm": 0.0050854869186878204,
      "learning_rate": 0.2680959630588089,
      "loss": 0.2435,
      "num_input_tokens_seen": 16289184,
      "step": 8460
    },
    {
      "epoch": 1.3809446121217066,
      "grad_norm": 0.003579669166356325,
      "learning_rate": 0.26805963545886985,
      "loss": 0.2482,
      "num_input_tokens_seen": 16300224,
      "step": 8465
    },
    {
      "epoch": 1.3817603393425237,
      "grad_norm": 0.005867453757673502,
      "learning_rate": 0.26802328965265143,
      "loss": 0.2519,
      "num_input_tokens_seen": 16309536,
      "step": 8470
    },
    {
      "epoch": 1.3825760665633413,
      "grad_norm": 0.002991484245285392,
      "learning_rate": 0.26798692564575854,
      "loss": 0.2346,
      "num_input_tokens_seen": 16317600,
      "step": 8475
    },
    {
      "epoch": 1.3833917937841584,
      "grad_norm": 0.0031631754245609045,
      "learning_rate": 0.26795054344379904,
      "loss": 0.2276,
      "num_input_tokens_seen": 16327792,
      "step": 8480
    },
    {
      "epoch": 1.384207521004976,
      "grad_norm": 0.0012545025674626231,
      "learning_rate": 0.2679141430523835,
      "loss": 0.2508,
      "num_input_tokens_seen": 16338384,
      "step": 8485
    },
    {
      "epoch": 1.3850232482257934,
      "grad_norm": 0.001957341330125928,
      "learning_rate": 0.2678777244771252,
      "loss": 0.2363,
      "num_input_tokens_seen": 16347792,
      "step": 8490
    },
    {
      "epoch": 1.3858389754466107,
      "grad_norm": 0.0032101646065711975,
      "learning_rate": 0.2678412877236405,
      "loss": 0.2219,
      "num_input_tokens_seen": 16358672,
      "step": 8495
    },
    {
      "epoch": 1.386654702667428,
      "grad_norm": 0.0028474037535488605,
      "learning_rate": 0.2678048327975484,
      "loss": 0.2186,
      "num_input_tokens_seen": 16368016,
      "step": 8500
    },
    {
      "epoch": 1.3874704298882454,
      "grad_norm": 0.0017115671653300524,
      "learning_rate": 0.2677683597044706,
      "loss": 0.2609,
      "num_input_tokens_seen": 16377504,
      "step": 8505
    },
    {
      "epoch": 1.3882861571090628,
      "grad_norm": 0.0047202217392623425,
      "learning_rate": 0.2677318684500318,
      "loss": 0.2465,
      "num_input_tokens_seen": 16386528,
      "step": 8510
    },
    {
      "epoch": 1.3891018843298801,
      "grad_norm": 0.0033643292263150215,
      "learning_rate": 0.2676953590398593,
      "loss": 0.2326,
      "num_input_tokens_seen": 16397376,
      "step": 8515
    },
    {
      "epoch": 1.3899176115506975,
      "grad_norm": 0.004251459147781134,
      "learning_rate": 0.2676588314795834,
      "loss": 0.245,
      "num_input_tokens_seen": 16406304,
      "step": 8520
    },
    {
      "epoch": 1.3907333387715148,
      "grad_norm": 0.003457891521975398,
      "learning_rate": 0.26762228577483715,
      "loss": 0.2216,
      "num_input_tokens_seen": 16415424,
      "step": 8525
    },
    {
      "epoch": 1.3915490659923322,
      "grad_norm": 0.002614626893773675,
      "learning_rate": 0.2675857219312563,
      "loss": 0.2477,
      "num_input_tokens_seen": 16425536,
      "step": 8530
    },
    {
      "epoch": 1.3923647932131495,
      "grad_norm": 0.0023515361826866865,
      "learning_rate": 0.2675491399544794,
      "loss": 0.2279,
      "num_input_tokens_seen": 16433248,
      "step": 8535
    },
    {
      "epoch": 1.393180520433967,
      "grad_norm": 0.0032395930029451847,
      "learning_rate": 0.2675125398501479,
      "loss": 0.2304,
      "num_input_tokens_seen": 16442416,
      "step": 8540
    },
    {
      "epoch": 1.3939962476547842,
      "grad_norm": 0.002972619840875268,
      "learning_rate": 0.26747592162390604,
      "loss": 0.2174,
      "num_input_tokens_seen": 16452416,
      "step": 8545
    },
    {
      "epoch": 1.3948119748756016,
      "grad_norm": 0.003217783523723483,
      "learning_rate": 0.26743928528140076,
      "loss": 0.2213,
      "num_input_tokens_seen": 16461952,
      "step": 8550
    },
    {
      "epoch": 1.395627702096419,
      "grad_norm": 0.002994847483932972,
      "learning_rate": 0.26740263082828186,
      "loss": 0.2207,
      "num_input_tokens_seen": 16472432,
      "step": 8555
    },
    {
      "epoch": 1.3964434293172363,
      "grad_norm": 0.004197489004582167,
      "learning_rate": 0.2673659582702019,
      "loss": 0.2491,
      "num_input_tokens_seen": 16482288,
      "step": 8560
    },
    {
      "epoch": 1.3972591565380537,
      "grad_norm": 0.0016242963029071689,
      "learning_rate": 0.2673292676128163,
      "loss": 0.2301,
      "num_input_tokens_seen": 16492544,
      "step": 8565
    },
    {
      "epoch": 1.398074883758871,
      "grad_norm": 0.0018946180352941155,
      "learning_rate": 0.2672925588617831,
      "loss": 0.2513,
      "num_input_tokens_seen": 16502048,
      "step": 8570
    },
    {
      "epoch": 1.3988906109796884,
      "grad_norm": 0.002441722434014082,
      "learning_rate": 0.2672558320227634,
      "loss": 0.2141,
      "num_input_tokens_seen": 16511008,
      "step": 8575
    },
    {
      "epoch": 1.3997063382005057,
      "grad_norm": 0.004167723935097456,
      "learning_rate": 0.2672190871014209,
      "loss": 0.2618,
      "num_input_tokens_seen": 16520656,
      "step": 8580
    },
    {
      "epoch": 1.400522065421323,
      "grad_norm": 0.003899160074070096,
      "learning_rate": 0.267182324103422,
      "loss": 0.2276,
      "num_input_tokens_seen": 16529552,
      "step": 8585
    },
    {
      "epoch": 1.4013377926421404,
      "grad_norm": 0.003215565113350749,
      "learning_rate": 0.2671455430344362,
      "loss": 0.2268,
      "num_input_tokens_seen": 16539392,
      "step": 8590
    },
    {
      "epoch": 1.4021535198629578,
      "grad_norm": 0.005462399683892727,
      "learning_rate": 0.2671087439001355,
      "loss": 0.2411,
      "num_input_tokens_seen": 16548800,
      "step": 8595
    },
    {
      "epoch": 1.4029692470837751,
      "grad_norm": 0.002123350976034999,
      "learning_rate": 0.2670719267061948,
      "loss": 0.2501,
      "num_input_tokens_seen": 16558464,
      "step": 8600
    },
    {
      "epoch": 1.4029692470837751,
      "eval_loss": 0.25421038269996643,
      "eval_runtime": 68.0936,
      "eval_samples_per_second": 40.018,
      "eval_steps_per_second": 20.017,
      "num_input_tokens_seen": 16558464,
      "step": 8600
    },
    {
      "epoch": 1.4037849743045925,
      "grad_norm": 0.003910926636308432,
      "learning_rate": 0.2670350914582918,
      "loss": 0.2509,
      "num_input_tokens_seen": 16566832,
      "step": 8605
    },
    {
      "epoch": 1.4046007015254098,
      "grad_norm": 0.004229648504406214,
      "learning_rate": 0.26699823816210694,
      "loss": 0.2512,
      "num_input_tokens_seen": 16576912,
      "step": 8610
    },
    {
      "epoch": 1.4054164287462272,
      "grad_norm": 0.005705173593014479,
      "learning_rate": 0.26696136682332344,
      "loss": 0.2252,
      "num_input_tokens_seen": 16586544,
      "step": 8615
    },
    {
      "epoch": 1.4062321559670445,
      "grad_norm": 0.0030958522111177444,
      "learning_rate": 0.2669244774476274,
      "loss": 0.2524,
      "num_input_tokens_seen": 16595808,
      "step": 8620
    },
    {
      "epoch": 1.407047883187862,
      "grad_norm": 0.0030408361926674843,
      "learning_rate": 0.2668875700407075,
      "loss": 0.2506,
      "num_input_tokens_seen": 16604848,
      "step": 8625
    },
    {
      "epoch": 1.4078636104086795,
      "grad_norm": 0.0041437032632529736,
      "learning_rate": 0.26685064460825547,
      "loss": 0.2215,
      "num_input_tokens_seen": 16614480,
      "step": 8630
    },
    {
      "epoch": 1.4086793376294966,
      "grad_norm": 0.0018103040056303144,
      "learning_rate": 0.26681370115596553,
      "loss": 0.2485,
      "num_input_tokens_seen": 16623520,
      "step": 8635
    },
    {
      "epoch": 1.4094950648503142,
      "grad_norm": 0.002735567046329379,
      "learning_rate": 0.26677673968953497,
      "loss": 0.2276,
      "num_input_tokens_seen": 16633216,
      "step": 8640
    },
    {
      "epoch": 1.4103107920711313,
      "grad_norm": 0.003082463052123785,
      "learning_rate": 0.2667397602146636,
      "loss": 0.2322,
      "num_input_tokens_seen": 16642768,
      "step": 8645
    },
    {
      "epoch": 1.4111265192919489,
      "grad_norm": 0.003307885257527232,
      "learning_rate": 0.2667027627370542,
      "loss": 0.2445,
      "num_input_tokens_seen": 16652032,
      "step": 8650
    },
    {
      "epoch": 1.411942246512766,
      "grad_norm": 0.003473862772807479,
      "learning_rate": 0.26666574726241216,
      "loss": 0.2355,
      "num_input_tokens_seen": 16660976,
      "step": 8655
    },
    {
      "epoch": 1.4127579737335836,
      "grad_norm": 0.0018622084753587842,
      "learning_rate": 0.2666287137964458,
      "loss": 0.2262,
      "num_input_tokens_seen": 16670496,
      "step": 8660
    },
    {
      "epoch": 1.4135737009544007,
      "grad_norm": 0.004533341620117426,
      "learning_rate": 0.26659166234486614,
      "loss": 0.201,
      "num_input_tokens_seen": 16681232,
      "step": 8665
    },
    {
      "epoch": 1.4143894281752183,
      "grad_norm": 0.002568654716014862,
      "learning_rate": 0.2665545929133869,
      "loss": 0.1852,
      "num_input_tokens_seen": 16690576,
      "step": 8670
    },
    {
      "epoch": 1.4152051553960356,
      "grad_norm": 0.00591790908947587,
      "learning_rate": 0.2665175055077248,
      "loss": 0.2901,
      "num_input_tokens_seen": 16700352,
      "step": 8675
    },
    {
      "epoch": 1.416020882616853,
      "grad_norm": 0.00663463631644845,
      "learning_rate": 0.2664804001335991,
      "loss": 0.3526,
      "num_input_tokens_seen": 16711264,
      "step": 8680
    },
    {
      "epoch": 1.4168366098376703,
      "grad_norm": 0.0020682746544480324,
      "learning_rate": 0.26644327679673185,
      "loss": 0.2485,
      "num_input_tokens_seen": 16721872,
      "step": 8685
    },
    {
      "epoch": 1.4176523370584877,
      "grad_norm": 0.002901768311858177,
      "learning_rate": 0.26640613550284803,
      "loss": 0.2297,
      "num_input_tokens_seen": 16730912,
      "step": 8690
    },
    {
      "epoch": 1.418468064279305,
      "grad_norm": 0.0021388870663940907,
      "learning_rate": 0.26636897625767525,
      "loss": 0.2448,
      "num_input_tokens_seen": 16741632,
      "step": 8695
    },
    {
      "epoch": 1.4192837915001224,
      "grad_norm": 0.012069445103406906,
      "learning_rate": 0.266331799066944,
      "loss": 0.2042,
      "num_input_tokens_seen": 16750080,
      "step": 8700
    },
    {
      "epoch": 1.4200995187209398,
      "grad_norm": 0.0062945024110376835,
      "learning_rate": 0.2662946039363874,
      "loss": 0.2282,
      "num_input_tokens_seen": 16758240,
      "step": 8705
    },
    {
      "epoch": 1.420915245941757,
      "grad_norm": 0.006922286935150623,
      "learning_rate": 0.2662573908717414,
      "loss": 0.2474,
      "num_input_tokens_seen": 16768224,
      "step": 8710
    },
    {
      "epoch": 1.4217309731625745,
      "grad_norm": 0.0035130351316183805,
      "learning_rate": 0.2662201598787447,
      "loss": 0.2469,
      "num_input_tokens_seen": 16778688,
      "step": 8715
    },
    {
      "epoch": 1.4225467003833918,
      "grad_norm": 0.005494821351021528,
      "learning_rate": 0.2661829109631389,
      "loss": 0.2993,
      "num_input_tokens_seen": 16788400,
      "step": 8720
    },
    {
      "epoch": 1.4233624276042092,
      "grad_norm": 0.002773764543235302,
      "learning_rate": 0.26614564413066816,
      "loss": 0.2578,
      "num_input_tokens_seen": 16798368,
      "step": 8725
    },
    {
      "epoch": 1.4241781548250265,
      "grad_norm": 0.00511676212772727,
      "learning_rate": 0.2661083593870795,
      "loss": 0.2783,
      "num_input_tokens_seen": 16807072,
      "step": 8730
    },
    {
      "epoch": 1.4249938820458439,
      "grad_norm": 0.004461419302970171,
      "learning_rate": 0.26607105673812276,
      "loss": 0.2268,
      "num_input_tokens_seen": 16818112,
      "step": 8735
    },
    {
      "epoch": 1.4258096092666612,
      "grad_norm": 0.004388109780848026,
      "learning_rate": 0.2660337361895504,
      "loss": 0.2423,
      "num_input_tokens_seen": 16829440,
      "step": 8740
    },
    {
      "epoch": 1.4266253364874786,
      "grad_norm": 0.0065626297146081924,
      "learning_rate": 0.26599639774711775,
      "loss": 0.2333,
      "num_input_tokens_seen": 16839712,
      "step": 8745
    },
    {
      "epoch": 1.427441063708296,
      "grad_norm": 0.003023345023393631,
      "learning_rate": 0.2659590414165829,
      "loss": 0.2461,
      "num_input_tokens_seen": 16848608,
      "step": 8750
    },
    {
      "epoch": 1.4282567909291133,
      "grad_norm": 0.003088773926720023,
      "learning_rate": 0.2659216672037066,
      "loss": 0.2246,
      "num_input_tokens_seen": 16859696,
      "step": 8755
    },
    {
      "epoch": 1.4290725181499306,
      "grad_norm": 0.0023800781928002834,
      "learning_rate": 0.26588427511425244,
      "loss": 0.2405,
      "num_input_tokens_seen": 16868832,
      "step": 8760
    },
    {
      "epoch": 1.429888245370748,
      "grad_norm": 0.0021277633495628834,
      "learning_rate": 0.26584686515398676,
      "loss": 0.2123,
      "num_input_tokens_seen": 16878288,
      "step": 8765
    },
    {
      "epoch": 1.4307039725915653,
      "grad_norm": 0.0029556946828961372,
      "learning_rate": 0.2658094373286787,
      "loss": 0.2169,
      "num_input_tokens_seen": 16888624,
      "step": 8770
    },
    {
      "epoch": 1.4315196998123827,
      "grad_norm": 0.00460114935413003,
      "learning_rate": 0.2657719916441,
      "loss": 0.264,
      "num_input_tokens_seen": 16897744,
      "step": 8775
    },
    {
      "epoch": 1.4323354270332,
      "grad_norm": 0.002677340293303132,
      "learning_rate": 0.2657345281060253,
      "loss": 0.2156,
      "num_input_tokens_seen": 16907632,
      "step": 8780
    },
    {
      "epoch": 1.4331511542540174,
      "grad_norm": 0.006796613801270723,
      "learning_rate": 0.26569704672023203,
      "loss": 0.2597,
      "num_input_tokens_seen": 16916752,
      "step": 8785
    },
    {
      "epoch": 1.4339668814748348,
      "grad_norm": 0.008340748026967049,
      "learning_rate": 0.26565954749250015,
      "loss": 0.2421,
      "num_input_tokens_seen": 16926176,
      "step": 8790
    },
    {
      "epoch": 1.434782608695652,
      "grad_norm": 0.00418113125488162,
      "learning_rate": 0.2656220304286126,
      "loss": 0.2383,
      "num_input_tokens_seen": 16936336,
      "step": 8795
    },
    {
      "epoch": 1.4355983359164695,
      "grad_norm": 0.005068761296570301,
      "learning_rate": 0.265584495534355,
      "loss": 0.2079,
      "num_input_tokens_seen": 16945488,
      "step": 8800
    },
    {
      "epoch": 1.4355983359164695,
      "eval_loss": 0.23352327942848206,
      "eval_runtime": 68.0297,
      "eval_samples_per_second": 40.056,
      "eval_steps_per_second": 20.035,
      "num_input_tokens_seen": 16945488,
      "step": 8800
    },
    {
      "epoch": 1.4364140631372868,
      "grad_norm": 0.002097544027492404,
      "learning_rate": 0.2655469428155156,
      "loss": 0.2086,
      "num_input_tokens_seen": 16955408,
      "step": 8805
    },
    {
      "epoch": 1.4372297903581042,
      "grad_norm": 0.0016762653831392527,
      "learning_rate": 0.2655093722778856,
      "loss": 0.2379,
      "num_input_tokens_seen": 16965680,
      "step": 8810
    },
    {
      "epoch": 1.4380455175789217,
      "grad_norm": 0.0029917857609689236,
      "learning_rate": 0.2654717839272588,
      "loss": 0.2557,
      "num_input_tokens_seen": 16975632,
      "step": 8815
    },
    {
      "epoch": 1.4388612447997389,
      "grad_norm": 0.003587851533666253,
      "learning_rate": 0.2654341777694318,
      "loss": 0.2479,
      "num_input_tokens_seen": 16985536,
      "step": 8820
    },
    {
      "epoch": 1.4396769720205564,
      "grad_norm": 0.0035651586949825287,
      "learning_rate": 0.265396553810204,
      "loss": 0.254,
      "num_input_tokens_seen": 16994352,
      "step": 8825
    },
    {
      "epoch": 1.4404926992413736,
      "grad_norm": 0.0061755976639688015,
      "learning_rate": 0.26535891205537737,
      "loss": 0.274,
      "num_input_tokens_seen": 17004960,
      "step": 8830
    },
    {
      "epoch": 1.4413084264621912,
      "grad_norm": 0.001709987292997539,
      "learning_rate": 0.26532125251075683,
      "loss": 0.2152,
      "num_input_tokens_seen": 17015408,
      "step": 8835
    },
    {
      "epoch": 1.4421241536830083,
      "grad_norm": 0.0036495120730251074,
      "learning_rate": 0.26528357518214996,
      "loss": 0.3174,
      "num_input_tokens_seen": 17024656,
      "step": 8840
    },
    {
      "epoch": 1.4429398809038259,
      "grad_norm": 0.005090116057544947,
      "learning_rate": 0.26524588007536704,
      "loss": 0.2486,
      "num_input_tokens_seen": 17033056,
      "step": 8845
    },
    {
      "epoch": 1.443755608124643,
      "grad_norm": 0.004706103354692459,
      "learning_rate": 0.26520816719622115,
      "loss": 0.2143,
      "num_input_tokens_seen": 17042384,
      "step": 8850
    },
    {
      "epoch": 1.4445713353454606,
      "grad_norm": 0.002771947765722871,
      "learning_rate": 0.2651704365505281,
      "loss": 0.2428,
      "num_input_tokens_seen": 17052384,
      "step": 8855
    },
    {
      "epoch": 1.445387062566278,
      "grad_norm": 0.003898603143170476,
      "learning_rate": 0.26513268814410634,
      "loss": 0.2379,
      "num_input_tokens_seen": 17059168,
      "step": 8860
    },
    {
      "epoch": 1.4462027897870953,
      "grad_norm": 0.0020916881039738655,
      "learning_rate": 0.2650949219827773,
      "loss": 0.2383,
      "num_input_tokens_seen": 17069408,
      "step": 8865
    },
    {
      "epoch": 1.4470185170079126,
      "grad_norm": 0.003333034925162792,
      "learning_rate": 0.26505713807236486,
      "loss": 0.2378,
      "num_input_tokens_seen": 17080016,
      "step": 8870
    },
    {
      "epoch": 1.44783424422873,
      "grad_norm": 0.001875822781585157,
      "learning_rate": 0.26501933641869585,
      "loss": 0.2205,
      "num_input_tokens_seen": 17089968,
      "step": 8875
    },
    {
      "epoch": 1.4486499714495473,
      "grad_norm": 0.0030387358274310827,
      "learning_rate": 0.26498151702759976,
      "loss": 0.2777,
      "num_input_tokens_seen": 17099504,
      "step": 8880
    },
    {
      "epoch": 1.4494656986703647,
      "grad_norm": 0.0034199354704469442,
      "learning_rate": 0.2649436799049088,
      "loss": 0.2136,
      "num_input_tokens_seen": 17109840,
      "step": 8885
    },
    {
      "epoch": 1.450281425891182,
      "grad_norm": 0.0026816162280738354,
      "learning_rate": 0.2649058250564579,
      "loss": 0.2303,
      "num_input_tokens_seen": 17118640,
      "step": 8890
    },
    {
      "epoch": 1.4510971531119994,
      "grad_norm": 0.003198359627276659,
      "learning_rate": 0.26486795248808476,
      "loss": 0.2448,
      "num_input_tokens_seen": 17128608,
      "step": 8895
    },
    {
      "epoch": 1.4519128803328167,
      "grad_norm": 0.003700959961861372,
      "learning_rate": 0.2648300622056298,
      "loss": 0.2353,
      "num_input_tokens_seen": 17138384,
      "step": 8900
    },
    {
      "epoch": 1.452728607553634,
      "grad_norm": 0.001962856622412801,
      "learning_rate": 0.2647921542149363,
      "loss": 0.2402,
      "num_input_tokens_seen": 17148560,
      "step": 8905
    },
    {
      "epoch": 1.4535443347744514,
      "grad_norm": 0.003413478611037135,
      "learning_rate": 0.26475422852185,
      "loss": 0.2521,
      "num_input_tokens_seen": 17159232,
      "step": 8910
    },
    {
      "epoch": 1.4543600619952688,
      "grad_norm": 0.0025012302212417126,
      "learning_rate": 0.2647162851322196,
      "loss": 0.228,
      "num_input_tokens_seen": 17168640,
      "step": 8915
    },
    {
      "epoch": 1.4551757892160861,
      "grad_norm": 0.0016995141049847007,
      "learning_rate": 0.2646783240518964,
      "loss": 0.2051,
      "num_input_tokens_seen": 17178000,
      "step": 8920
    },
    {
      "epoch": 1.4559915164369035,
      "grad_norm": 0.006682902108877897,
      "learning_rate": 0.26464034528673447,
      "loss": 0.2586,
      "num_input_tokens_seen": 17188272,
      "step": 8925
    },
    {
      "epoch": 1.4568072436577209,
      "grad_norm": 0.0020187250338494778,
      "learning_rate": 0.26460234884259065,
      "loss": 0.2341,
      "num_input_tokens_seen": 17199008,
      "step": 8930
    },
    {
      "epoch": 1.4576229708785382,
      "grad_norm": 0.005155214574187994,
      "learning_rate": 0.2645643347253245,
      "loss": 0.2555,
      "num_input_tokens_seen": 17209008,
      "step": 8935
    },
    {
      "epoch": 1.4584386980993556,
      "grad_norm": 0.002621206920593977,
      "learning_rate": 0.2645263029407982,
      "loss": 0.2229,
      "num_input_tokens_seen": 17219760,
      "step": 8940
    },
    {
      "epoch": 1.459254425320173,
      "grad_norm": 0.0044502681121230125,
      "learning_rate": 0.2644882534948767,
      "loss": 0.2264,
      "num_input_tokens_seen": 17229728,
      "step": 8945
    },
    {
      "epoch": 1.4600701525409903,
      "grad_norm": 0.003044551471248269,
      "learning_rate": 0.2644501863934278,
      "loss": 0.1576,
      "num_input_tokens_seen": 17239888,
      "step": 8950
    },
    {
      "epoch": 1.4608858797618076,
      "grad_norm": 0.0033321899827569723,
      "learning_rate": 0.26441210164232193,
      "loss": 0.3212,
      "num_input_tokens_seen": 17249376,
      "step": 8955
    },
    {
      "epoch": 1.461701606982625,
      "grad_norm": 0.0051599955186247826,
      "learning_rate": 0.26437399924743216,
      "loss": 0.2509,
      "num_input_tokens_seen": 17259072,
      "step": 8960
    },
    {
      "epoch": 1.4625173342034423,
      "grad_norm": 0.0021584585774689913,
      "learning_rate": 0.26433587921463436,
      "loss": 0.2229,
      "num_input_tokens_seen": 17269680,
      "step": 8965
    },
    {
      "epoch": 1.4633330614242597,
      "grad_norm": 0.0023674210533499718,
      "learning_rate": 0.2642977415498072,
      "loss": 0.2422,
      "num_input_tokens_seen": 17278944,
      "step": 8970
    },
    {
      "epoch": 1.464148788645077,
      "grad_norm": 0.0014668174553662539,
      "learning_rate": 0.26425958625883195,
      "loss": 0.2188,
      "num_input_tokens_seen": 17288384,
      "step": 8975
    },
    {
      "epoch": 1.4649645158658944,
      "grad_norm": 0.003813209943473339,
      "learning_rate": 0.2642214133475926,
      "loss": 0.2796,
      "num_input_tokens_seen": 17298176,
      "step": 8980
    },
    {
      "epoch": 1.4657802430867117,
      "grad_norm": 0.0023153580259531736,
      "learning_rate": 0.26418322282197587,
      "loss": 0.2288,
      "num_input_tokens_seen": 17308416,
      "step": 8985
    },
    {
      "epoch": 1.466595970307529,
      "grad_norm": 0.004620824009180069,
      "learning_rate": 0.2641450146878714,
      "loss": 0.2327,
      "num_input_tokens_seen": 17317664,
      "step": 8990
    },
    {
      "epoch": 1.4674116975283464,
      "grad_norm": 0.002125695114955306,
      "learning_rate": 0.26410678895117107,
      "loss": 0.2274,
      "num_input_tokens_seen": 17328256,
      "step": 8995
    },
    {
      "epoch": 1.468227424749164,
      "grad_norm": 0.0014773732982575893,
      "learning_rate": 0.26406854561777,
      "loss": 0.1982,
      "num_input_tokens_seen": 17338800,
      "step": 9000
    },
    {
      "epoch": 1.468227424749164,
      "eval_loss": 0.23297715187072754,
      "eval_runtime": 68.0172,
      "eval_samples_per_second": 40.063,
      "eval_steps_per_second": 20.039,
      "num_input_tokens_seen": 17338800,
      "step": 9000
    },
    {
      "epoch": 1.4690431519699811,
      "grad_norm": 0.002764298114925623,
      "learning_rate": 0.26403028469356576,
      "loss": 0.1826,
      "num_input_tokens_seen": 17348048,
      "step": 9005
    },
    {
      "epoch": 1.4698588791907987,
      "grad_norm": 0.0029503877740353346,
      "learning_rate": 0.2639920061844585,
      "loss": 0.2269,
      "num_input_tokens_seen": 17356416,
      "step": 9010
    },
    {
      "epoch": 1.4706746064116158,
      "grad_norm": 0.0017609262140467763,
      "learning_rate": 0.2639537100963515,
      "loss": 0.2896,
      "num_input_tokens_seen": 17364048,
      "step": 9015
    },
    {
      "epoch": 1.4714903336324334,
      "grad_norm": 0.0018743386026471853,
      "learning_rate": 0.26391539643515033,
      "loss": 0.2292,
      "num_input_tokens_seen": 17374272,
      "step": 9020
    },
    {
      "epoch": 1.4723060608532506,
      "grad_norm": 0.001050892984494567,
      "learning_rate": 0.26387706520676346,
      "loss": 0.2113,
      "num_input_tokens_seen": 17383152,
      "step": 9025
    },
    {
      "epoch": 1.4731217880740681,
      "grad_norm": 0.00312721636146307,
      "learning_rate": 0.26383871641710205,
      "loss": 0.223,
      "num_input_tokens_seen": 17391632,
      "step": 9030
    },
    {
      "epoch": 1.4739375152948853,
      "grad_norm": 0.006124366540461779,
      "learning_rate": 0.26380035007208,
      "loss": 0.2432,
      "num_input_tokens_seen": 17400960,
      "step": 9035
    },
    {
      "epoch": 1.4747532425157028,
      "grad_norm": 0.0017689260421320796,
      "learning_rate": 0.26376196617761394,
      "loss": 0.2147,
      "num_input_tokens_seen": 17409584,
      "step": 9040
    },
    {
      "epoch": 1.4755689697365202,
      "grad_norm": 0.0023732073605060577,
      "learning_rate": 0.263723564739623,
      "loss": 0.2503,
      "num_input_tokens_seen": 17419616,
      "step": 9045
    },
    {
      "epoch": 1.4763846969573375,
      "grad_norm": 0.003065867582336068,
      "learning_rate": 0.2636851457640293,
      "loss": 0.2543,
      "num_input_tokens_seen": 17429856,
      "step": 9050
    },
    {
      "epoch": 1.477200424178155,
      "grad_norm": 0.003058659378439188,
      "learning_rate": 0.26364670925675737,
      "loss": 0.2209,
      "num_input_tokens_seen": 17441168,
      "step": 9055
    },
    {
      "epoch": 1.4780161513989722,
      "grad_norm": 0.001778429839760065,
      "learning_rate": 0.2636082552237347,
      "loss": 0.2359,
      "num_input_tokens_seen": 17449712,
      "step": 9060
    },
    {
      "epoch": 1.4788318786197896,
      "grad_norm": 0.0027494393289089203,
      "learning_rate": 0.26356978367089146,
      "loss": 0.2282,
      "num_input_tokens_seen": 17459232,
      "step": 9065
    },
    {
      "epoch": 1.479647605840607,
      "grad_norm": 0.0016203377163037658,
      "learning_rate": 0.26353129460416036,
      "loss": 0.2291,
      "num_input_tokens_seen": 17469120,
      "step": 9070
    },
    {
      "epoch": 1.4804633330614243,
      "grad_norm": 0.0013918998884037137,
      "learning_rate": 0.2634927880294769,
      "loss": 0.2355,
      "num_input_tokens_seen": 17479472,
      "step": 9075
    },
    {
      "epoch": 1.4812790602822417,
      "grad_norm": 0.002982516773045063,
      "learning_rate": 0.26345426395277927,
      "loss": 0.2079,
      "num_input_tokens_seen": 17490384,
      "step": 9080
    },
    {
      "epoch": 1.482094787503059,
      "grad_norm": 0.002352071925997734,
      "learning_rate": 0.2634157223800084,
      "loss": 0.2352,
      "num_input_tokens_seen": 17501168,
      "step": 9085
    },
    {
      "epoch": 1.4829105147238764,
      "grad_norm": 0.003919532988220453,
      "learning_rate": 0.26337716331710787,
      "loss": 0.2444,
      "num_input_tokens_seen": 17510112,
      "step": 9090
    },
    {
      "epoch": 1.4837262419446937,
      "grad_norm": 0.005136490799486637,
      "learning_rate": 0.2633385867700239,
      "loss": 0.2094,
      "num_input_tokens_seen": 17520288,
      "step": 9095
    },
    {
      "epoch": 1.484541969165511,
      "grad_norm": 0.0028198231011629105,
      "learning_rate": 0.2632999927447056,
      "loss": 0.217,
      "num_input_tokens_seen": 17530704,
      "step": 9100
    },
    {
      "epoch": 1.4853576963863284,
      "grad_norm": 0.0028377301059663296,
      "learning_rate": 0.2632613812471046,
      "loss": 0.2114,
      "num_input_tokens_seen": 17540320,
      "step": 9105
    },
    {
      "epoch": 1.4861734236071458,
      "grad_norm": 0.0042454544454813,
      "learning_rate": 0.2632227522831753,
      "loss": 0.2218,
      "num_input_tokens_seen": 17550240,
      "step": 9110
    },
    {
      "epoch": 1.4869891508279631,
      "grad_norm": 0.0021361385006457567,
      "learning_rate": 0.26318410585887475,
      "loss": 0.2562,
      "num_input_tokens_seen": 17559200,
      "step": 9115
    },
    {
      "epoch": 1.4878048780487805,
      "grad_norm": 0.004405431915074587,
      "learning_rate": 0.2631454419801627,
      "loss": 0.2279,
      "num_input_tokens_seen": 17569824,
      "step": 9120
    },
    {
      "epoch": 1.4886206052695978,
      "grad_norm": 0.0021029370836913586,
      "learning_rate": 0.2631067606530016,
      "loss": 0.2311,
      "num_input_tokens_seen": 17580400,
      "step": 9125
    },
    {
      "epoch": 1.4894363324904152,
      "grad_norm": 0.0049202400259673595,
      "learning_rate": 0.2630680618833567,
      "loss": 0.2082,
      "num_input_tokens_seen": 17590400,
      "step": 9130
    },
    {
      "epoch": 1.4902520597112325,
      "grad_norm": 0.0018802674021571875,
      "learning_rate": 0.26302934567719566,
      "loss": 0.2424,
      "num_input_tokens_seen": 17600768,
      "step": 9135
    },
    {
      "epoch": 1.4910677869320499,
      "grad_norm": 0.003677021013572812,
      "learning_rate": 0.2629906120404892,
      "loss": 0.2337,
      "num_input_tokens_seen": 17611568,
      "step": 9140
    },
    {
      "epoch": 1.4918835141528672,
      "grad_norm": 0.0037180185317993164,
      "learning_rate": 0.26295186097921036,
      "loss": 0.2467,
      "num_input_tokens_seen": 17622560,
      "step": 9145
    },
    {
      "epoch": 1.4926992413736846,
      "grad_norm": 0.00357917626388371,
      "learning_rate": 0.2629130924993351,
      "loss": 0.2044,
      "num_input_tokens_seen": 17634256,
      "step": 9150
    },
    {
      "epoch": 1.493514968594502,
      "grad_norm": 0.0017679485026746988,
      "learning_rate": 0.2628743066068421,
      "loss": 0.2567,
      "num_input_tokens_seen": 17643168,
      "step": 9155
    },
    {
      "epoch": 1.4943306958153193,
      "grad_norm": 0.003213403979316354,
      "learning_rate": 0.26283550330771244,
      "loss": 0.2411,
      "num_input_tokens_seen": 17652976,
      "step": 9160
    },
    {
      "epoch": 1.4951464230361367,
      "grad_norm": 0.002136166673153639,
      "learning_rate": 0.2627966826079303,
      "loss": 0.2021,
      "num_input_tokens_seen": 17662256,
      "step": 9165
    },
    {
      "epoch": 1.495962150256954,
      "grad_norm": 0.0017618570709601045,
      "learning_rate": 0.26275784451348216,
      "loss": 0.2236,
      "num_input_tokens_seen": 17670848,
      "step": 9170
    },
    {
      "epoch": 1.4967778774777714,
      "grad_norm": 0.001381305861286819,
      "learning_rate": 0.2627189890303574,
      "loss": 0.2177,
      "num_input_tokens_seen": 17680240,
      "step": 9175
    },
    {
      "epoch": 1.4975936046985887,
      "grad_norm": 0.007421594578772783,
      "learning_rate": 0.262680116164548,
      "loss": 0.234,
      "num_input_tokens_seen": 17691360,
      "step": 9180
    },
    {
      "epoch": 1.4984093319194063,
      "grad_norm": 0.0030669032130390406,
      "learning_rate": 0.2626412259220487,
      "loss": 0.1505,
      "num_input_tokens_seen": 17701120,
      "step": 9185
    },
    {
      "epoch": 1.4992250591402234,
      "grad_norm": 0.0032501169480383396,
      "learning_rate": 0.2626023183088568,
      "loss": 0.2401,
      "num_input_tokens_seen": 17710144,
      "step": 9190
    },
    {
      "epoch": 1.500040786361041,
      "grad_norm": 0.0016477004392072558,
      "learning_rate": 0.26256339333097234,
      "loss": 0.244,
      "num_input_tokens_seen": 17719104,
      "step": 9195
    },
    {
      "epoch": 1.5008565135818581,
      "grad_norm": 0.0016411832766607404,
      "learning_rate": 0.2625244509943981,
      "loss": 0.2529,
      "num_input_tokens_seen": 17729104,
      "step": 9200
    },
    {
      "epoch": 1.5008565135818581,
      "eval_loss": 0.2367810308933258,
      "eval_runtime": 68.0085,
      "eval_samples_per_second": 40.069,
      "eval_steps_per_second": 20.042,
      "num_input_tokens_seen": 17729104,
      "step": 9200
    },
    {
      "epoch": 1.5016722408026757,
      "grad_norm": 0.006086734123528004,
      "learning_rate": 0.2624854913051395,
      "loss": 0.2694,
      "num_input_tokens_seen": 17738768,
      "step": 9205
    },
    {
      "epoch": 1.5024879680234928,
      "grad_norm": 0.003527663182467222,
      "learning_rate": 0.26244651426920446,
      "loss": 0.2548,
      "num_input_tokens_seen": 17748864,
      "step": 9210
    },
    {
      "epoch": 1.5033036952443104,
      "grad_norm": 0.007281680591404438,
      "learning_rate": 0.26240751989260386,
      "loss": 0.2329,
      "num_input_tokens_seen": 17757872,
      "step": 9215
    },
    {
      "epoch": 1.5041194224651275,
      "grad_norm": 0.0016238788375630975,
      "learning_rate": 0.2623685081813511,
      "loss": 0.2158,
      "num_input_tokens_seen": 17768320,
      "step": 9220
    },
    {
      "epoch": 1.504935149685945,
      "grad_norm": 0.0035681510344147682,
      "learning_rate": 0.2623294791414623,
      "loss": 0.2396,
      "num_input_tokens_seen": 17776992,
      "step": 9225
    },
    {
      "epoch": 1.5057508769067622,
      "grad_norm": 0.003538289573043585,
      "learning_rate": 0.26229043277895614,
      "loss": 0.2371,
      "num_input_tokens_seen": 17786864,
      "step": 9230
    },
    {
      "epoch": 1.5065666041275798,
      "grad_norm": 0.00378765189088881,
      "learning_rate": 0.2622513690998542,
      "loss": 0.2221,
      "num_input_tokens_seen": 17797248,
      "step": 9235
    },
    {
      "epoch": 1.507382331348397,
      "grad_norm": 0.002451697364449501,
      "learning_rate": 0.26221228811018044,
      "loss": 0.232,
      "num_input_tokens_seen": 17806480,
      "step": 9240
    },
    {
      "epoch": 1.5081980585692145,
      "grad_norm": 0.0021002348512411118,
      "learning_rate": 0.2621731898159617,
      "loss": 0.2416,
      "num_input_tokens_seen": 17817376,
      "step": 9245
    },
    {
      "epoch": 1.5090137857900316,
      "grad_norm": 0.0023972545750439167,
      "learning_rate": 0.26213407422322743,
      "loss": 0.2175,
      "num_input_tokens_seen": 17826224,
      "step": 9250
    },
    {
      "epoch": 1.5098295130108492,
      "grad_norm": 0.0043663340620696545,
      "learning_rate": 0.2620949413380098,
      "loss": 0.2439,
      "num_input_tokens_seen": 17836528,
      "step": 9255
    },
    {
      "epoch": 1.5106452402316666,
      "grad_norm": 0.00450406176969409,
      "learning_rate": 0.26205579116634353,
      "loss": 0.2735,
      "num_input_tokens_seen": 17845312,
      "step": 9260
    },
    {
      "epoch": 1.511460967452484,
      "grad_norm": 0.0031248098239302635,
      "learning_rate": 0.26201662371426604,
      "loss": 0.2446,
      "num_input_tokens_seen": 17853360,
      "step": 9265
    },
    {
      "epoch": 1.5122766946733013,
      "grad_norm": 0.0031052979175001383,
      "learning_rate": 0.2619774389878175,
      "loss": 0.2428,
      "num_input_tokens_seen": 17862336,
      "step": 9270
    },
    {
      "epoch": 1.5130924218941186,
      "grad_norm": 0.0028437881264835596,
      "learning_rate": 0.2619382369930407,
      "loss": 0.2246,
      "num_input_tokens_seen": 17872224,
      "step": 9275
    },
    {
      "epoch": 1.513908149114936,
      "grad_norm": 0.003226666944101453,
      "learning_rate": 0.261899017735981,
      "loss": 0.2007,
      "num_input_tokens_seen": 17881872,
      "step": 9280
    },
    {
      "epoch": 1.5147238763357533,
      "grad_norm": 0.0023702990729361773,
      "learning_rate": 0.2618597812226866,
      "loss": 0.2251,
      "num_input_tokens_seen": 17890944,
      "step": 9285
    },
    {
      "epoch": 1.5155396035565707,
      "grad_norm": 0.002572274301201105,
      "learning_rate": 0.2618205274592082,
      "loss": 0.2473,
      "num_input_tokens_seen": 17899392,
      "step": 9290
    },
    {
      "epoch": 1.516355330777388,
      "grad_norm": 0.0030597327277064323,
      "learning_rate": 0.2617812564515992,
      "loss": 0.254,
      "num_input_tokens_seen": 17908432,
      "step": 9295
    },
    {
      "epoch": 1.5171710579982054,
      "grad_norm": 0.0019335869001224637,
      "learning_rate": 0.2617419682059158,
      "loss": 0.2061,
      "num_input_tokens_seen": 17918080,
      "step": 9300
    },
    {
      "epoch": 1.5179867852190227,
      "grad_norm": 0.0017976564122363925,
      "learning_rate": 0.26170266272821663,
      "loss": 0.2326,
      "num_input_tokens_seen": 17927280,
      "step": 9305
    },
    {
      "epoch": 1.51880251243984,
      "grad_norm": 0.0017163990996778011,
      "learning_rate": 0.26166334002456315,
      "loss": 0.2292,
      "num_input_tokens_seen": 17937456,
      "step": 9310
    },
    {
      "epoch": 1.5196182396606575,
      "grad_norm": 0.0012959004379808903,
      "learning_rate": 0.2616240001010194,
      "loss": 0.2291,
      "num_input_tokens_seen": 17947440,
      "step": 9315
    },
    {
      "epoch": 1.5204339668814748,
      "grad_norm": 0.0018748723668977618,
      "learning_rate": 0.26158464296365197,
      "loss": 0.2227,
      "num_input_tokens_seen": 17957056,
      "step": 9320
    },
    {
      "epoch": 1.5212496941022922,
      "grad_norm": 0.002026437083259225,
      "learning_rate": 0.2615452686185304,
      "loss": 0.209,
      "num_input_tokens_seen": 17966752,
      "step": 9325
    },
    {
      "epoch": 1.5220654213231095,
      "grad_norm": 0.0015500342706218362,
      "learning_rate": 0.26150587707172673,
      "loss": 0.2621,
      "num_input_tokens_seen": 17975984,
      "step": 9330
    },
    {
      "epoch": 1.5228811485439269,
      "grad_norm": 0.0013854895951226354,
      "learning_rate": 0.2614664683293154,
      "loss": 0.2107,
      "num_input_tokens_seen": 17985296,
      "step": 9335
    },
    {
      "epoch": 1.5236968757647442,
      "grad_norm": 0.004362464416772127,
      "learning_rate": 0.26142704239737397,
      "loss": 0.2504,
      "num_input_tokens_seen": 17994304,
      "step": 9340
    },
    {
      "epoch": 1.5245126029855616,
      "grad_norm": 0.0023028498981148005,
      "learning_rate": 0.26138759928198235,
      "loss": 0.2255,
      "num_input_tokens_seen": 18002768,
      "step": 9345
    },
    {
      "epoch": 1.5253283302063791,
      "grad_norm": 0.0016695758095011115,
      "learning_rate": 0.26134813898922304,
      "loss": 0.2265,
      "num_input_tokens_seen": 18010816,
      "step": 9350
    },
    {
      "epoch": 1.5261440574271963,
      "grad_norm": 0.0036874033976346254,
      "learning_rate": 0.26130866152518145,
      "loss": 0.2432,
      "num_input_tokens_seen": 18020640,
      "step": 9355
    },
    {
      "epoch": 1.5269597846480139,
      "grad_norm": 0.005033411551266909,
      "learning_rate": 0.2612691668959455,
      "loss": 0.2643,
      "num_input_tokens_seen": 18030720,
      "step": 9360
    },
    {
      "epoch": 1.527775511868831,
      "grad_norm": 0.001587557839229703,
      "learning_rate": 0.2612296551076057,
      "loss": 0.2541,
      "num_input_tokens_seen": 18040576,
      "step": 9365
    },
    {
      "epoch": 1.5285912390896486,
      "grad_norm": 0.002637596568092704,
      "learning_rate": 0.26119012616625525,
      "loss": 0.2305,
      "num_input_tokens_seen": 18049552,
      "step": 9370
    },
    {
      "epoch": 1.5294069663104657,
      "grad_norm": 0.00311761861667037,
      "learning_rate": 0.26115058007799,
      "loss": 0.2316,
      "num_input_tokens_seen": 18059104,
      "step": 9375
    },
    {
      "epoch": 1.5302226935312833,
      "grad_norm": 0.004772372078150511,
      "learning_rate": 0.26111101684890864,
      "loss": 0.2237,
      "num_input_tokens_seen": 18068736,
      "step": 9380
    },
    {
      "epoch": 1.5310384207521004,
      "grad_norm": 0.0018068391364067793,
      "learning_rate": 0.26107143648511205,
      "loss": 0.2344,
      "num_input_tokens_seen": 18077904,
      "step": 9385
    },
    {
      "epoch": 1.531854147972918,
      "grad_norm": 0.0027291150763630867,
      "learning_rate": 0.2610318389927042,
      "loss": 0.2152,
      "num_input_tokens_seen": 18088144,
      "step": 9390
    },
    {
      "epoch": 1.532669875193735,
      "grad_norm": 0.0028764952439814806,
      "learning_rate": 0.26099222437779146,
      "loss": 0.264,
      "num_input_tokens_seen": 18097536,
      "step": 9395
    },
    {
      "epoch": 1.5334856024145527,
      "grad_norm": 0.002824134426191449,
      "learning_rate": 0.26095259264648285,
      "loss": 0.2348,
      "num_input_tokens_seen": 18107328,
      "step": 9400
    },
    {
      "epoch": 1.5334856024145527,
      "eval_loss": 0.24091751873493195,
      "eval_runtime": 68.0717,
      "eval_samples_per_second": 40.031,
      "eval_steps_per_second": 20.023,
      "num_input_tokens_seen": 18107328,
      "step": 9400
    },
    {
      "epoch": 1.5343013296353698,
      "grad_norm": 0.002247612690553069,
      "learning_rate": 0.2609129438048902,
      "loss": 0.2385,
      "num_input_tokens_seen": 18117552,
      "step": 9405
    },
    {
      "epoch": 1.5351170568561874,
      "grad_norm": 0.003361280309036374,
      "learning_rate": 0.2608732778591278,
      "loss": 0.2597,
      "num_input_tokens_seen": 18126496,
      "step": 9410
    },
    {
      "epoch": 1.5359327840770045,
      "grad_norm": 0.007416905835270882,
      "learning_rate": 0.2608335948153126,
      "loss": 0.2394,
      "num_input_tokens_seen": 18136480,
      "step": 9415
    },
    {
      "epoch": 1.536748511297822,
      "grad_norm": 0.0010274003725498915,
      "learning_rate": 0.26079389467956426,
      "loss": 0.2147,
      "num_input_tokens_seen": 18146976,
      "step": 9420
    },
    {
      "epoch": 1.5375642385186392,
      "grad_norm": 0.0013363390462473035,
      "learning_rate": 0.26075417745800505,
      "loss": 0.2479,
      "num_input_tokens_seen": 18157120,
      "step": 9425
    },
    {
      "epoch": 1.5383799657394568,
      "grad_norm": 0.0020477892830967903,
      "learning_rate": 0.26071444315675985,
      "loss": 0.2402,
      "num_input_tokens_seen": 18166000,
      "step": 9430
    },
    {
      "epoch": 1.539195692960274,
      "grad_norm": 0.0018815306248143315,
      "learning_rate": 0.2606746917819562,
      "loss": 0.2326,
      "num_input_tokens_seen": 18176128,
      "step": 9435
    },
    {
      "epoch": 1.5400114201810915,
      "grad_norm": 0.0018851880449801683,
      "learning_rate": 0.2606349233397242,
      "loss": 0.2335,
      "num_input_tokens_seen": 18186416,
      "step": 9440
    },
    {
      "epoch": 1.5408271474019088,
      "grad_norm": 0.0015632875729352236,
      "learning_rate": 0.26059513783619676,
      "loss": 0.2289,
      "num_input_tokens_seen": 18196144,
      "step": 9445
    },
    {
      "epoch": 1.5416428746227262,
      "grad_norm": 0.0028908932581543922,
      "learning_rate": 0.26055533527750924,
      "loss": 0.2696,
      "num_input_tokens_seen": 18206720,
      "step": 9450
    },
    {
      "epoch": 1.5424586018435436,
      "grad_norm": 0.0015436123358085752,
      "learning_rate": 0.26051551566979964,
      "loss": 0.2194,
      "num_input_tokens_seen": 18215760,
      "step": 9455
    },
    {
      "epoch": 1.543274329064361,
      "grad_norm": 0.006186226848512888,
      "learning_rate": 0.26047567901920876,
      "loss": 0.2309,
      "num_input_tokens_seen": 18225664,
      "step": 9460
    },
    {
      "epoch": 1.5440900562851783,
      "grad_norm": 0.002337270649150014,
      "learning_rate": 0.2604358253318798,
      "loss": 0.2396,
      "num_input_tokens_seen": 18236224,
      "step": 9465
    },
    {
      "epoch": 1.5449057835059956,
      "grad_norm": 0.0012440441641956568,
      "learning_rate": 0.26039595461395876,
      "loss": 0.2239,
      "num_input_tokens_seen": 18246192,
      "step": 9470
    },
    {
      "epoch": 1.545721510726813,
      "grad_norm": 0.00143211311660707,
      "learning_rate": 0.26035606687159424,
      "loss": 0.2223,
      "num_input_tokens_seen": 18256368,
      "step": 9475
    },
    {
      "epoch": 1.5465372379476303,
      "grad_norm": 0.0014087504241615534,
      "learning_rate": 0.26031616211093733,
      "loss": 0.2463,
      "num_input_tokens_seen": 18265440,
      "step": 9480
    },
    {
      "epoch": 1.5473529651684477,
      "grad_norm": 0.0016101290239021182,
      "learning_rate": 0.26027624033814195,
      "loss": 0.2457,
      "num_input_tokens_seen": 18274320,
      "step": 9485
    },
    {
      "epoch": 1.548168692389265,
      "grad_norm": 0.002700287848711014,
      "learning_rate": 0.2602363015593645,
      "loss": 0.2316,
      "num_input_tokens_seen": 18285520,
      "step": 9490
    },
    {
      "epoch": 1.5489844196100824,
      "grad_norm": 0.0010369549272581935,
      "learning_rate": 0.26019634578076395,
      "loss": 0.2277,
      "num_input_tokens_seen": 18295280,
      "step": 9495
    },
    {
      "epoch": 1.5498001468308997,
      "grad_norm": 0.002533108228817582,
      "learning_rate": 0.26015637300850214,
      "loss": 0.244,
      "num_input_tokens_seen": 18306192,
      "step": 9500
    },
    {
      "epoch": 1.550615874051717,
      "grad_norm": 0.0015449870843440294,
      "learning_rate": 0.26011638324874325,
      "loss": 0.2278,
      "num_input_tokens_seen": 18316944,
      "step": 9505
    },
    {
      "epoch": 1.5514316012725344,
      "grad_norm": 0.0027940399013459682,
      "learning_rate": 0.2600763765076543,
      "loss": 0.2318,
      "num_input_tokens_seen": 18326992,
      "step": 9510
    },
    {
      "epoch": 1.5522473284933518,
      "grad_norm": 0.002392898313701153,
      "learning_rate": 0.2600363527914048,
      "loss": 0.2381,
      "num_input_tokens_seen": 18336256,
      "step": 9515
    },
    {
      "epoch": 1.5530630557141691,
      "grad_norm": 0.0015574011486023664,
      "learning_rate": 0.25999631210616686,
      "loss": 0.2337,
      "num_input_tokens_seen": 18346112,
      "step": 9520
    },
    {
      "epoch": 1.5538787829349865,
      "grad_norm": 0.001744559733197093,
      "learning_rate": 0.25995625445811527,
      "loss": 0.2423,
      "num_input_tokens_seen": 18355024,
      "step": 9525
    },
    {
      "epoch": 1.5546945101558038,
      "grad_norm": 0.0043938495218753815,
      "learning_rate": 0.2599161798534275,
      "loss": 0.2305,
      "num_input_tokens_seen": 18364272,
      "step": 9530
    },
    {
      "epoch": 1.5555102373766214,
      "grad_norm": 0.0030314414761960506,
      "learning_rate": 0.25987608829828346,
      "loss": 0.2401,
      "num_input_tokens_seen": 18374544,
      "step": 9535
    },
    {
      "epoch": 1.5563259645974385,
      "grad_norm": 0.00177750492002815,
      "learning_rate": 0.25983597979886586,
      "loss": 0.2133,
      "num_input_tokens_seen": 18384768,
      "step": 9540
    },
    {
      "epoch": 1.5571416918182561,
      "grad_norm": 0.002397460164502263,
      "learning_rate": 0.2597958543613599,
      "loss": 0.268,
      "num_input_tokens_seen": 18394176,
      "step": 9545
    },
    {
      "epoch": 1.5579574190390733,
      "grad_norm": 0.003178779501467943,
      "learning_rate": 0.25975571199195335,
      "loss": 0.2703,
      "num_input_tokens_seen": 18402000,
      "step": 9550
    },
    {
      "epoch": 1.5587731462598908,
      "grad_norm": 0.001495170989073813,
      "learning_rate": 0.25971555269683677,
      "loss": 0.2392,
      "num_input_tokens_seen": 18412096,
      "step": 9555
    },
    {
      "epoch": 1.559588873480708,
      "grad_norm": 0.0021417983807623386,
      "learning_rate": 0.25967537648220324,
      "loss": 0.2385,
      "num_input_tokens_seen": 18421472,
      "step": 9560
    },
    {
      "epoch": 1.5604046007015255,
      "grad_norm": 0.001915894215926528,
      "learning_rate": 0.2596351833542483,
      "loss": 0.2286,
      "num_input_tokens_seen": 18431680,
      "step": 9565
    },
    {
      "epoch": 1.5612203279223427,
      "grad_norm": 0.002705928636714816,
      "learning_rate": 0.25959497331917036,
      "loss": 0.2298,
      "num_input_tokens_seen": 18441616,
      "step": 9570
    },
    {
      "epoch": 1.5620360551431602,
      "grad_norm": 0.0022786036133766174,
      "learning_rate": 0.2595547463831703,
      "loss": 0.24,
      "num_input_tokens_seen": 18451840,
      "step": 9575
    },
    {
      "epoch": 1.5628517823639774,
      "grad_norm": 0.002247156808152795,
      "learning_rate": 0.25951450255245156,
      "loss": 0.2214,
      "num_input_tokens_seen": 18461296,
      "step": 9580
    },
    {
      "epoch": 1.563667509584795,
      "grad_norm": 0.002214344684034586,
      "learning_rate": 0.2594742418332203,
      "loss": 0.2729,
      "num_input_tokens_seen": 18471856,
      "step": 9585
    },
    {
      "epoch": 1.564483236805612,
      "grad_norm": 0.0014948261668905616,
      "learning_rate": 0.2594339642316852,
      "loss": 0.2372,
      "num_input_tokens_seen": 18479456,
      "step": 9590
    },
    {
      "epoch": 1.5652989640264297,
      "grad_norm": 0.0033908775076270103,
      "learning_rate": 0.2593936697540576,
      "loss": 0.244,
      "num_input_tokens_seen": 18488592,
      "step": 9595
    },
    {
      "epoch": 1.5661146912472468,
      "grad_norm": 0.001405567629262805,
      "learning_rate": 0.2593533584065514,
      "loss": 0.2399,
      "num_input_tokens_seen": 18497776,
      "step": 9600
    },
    {
      "epoch": 1.5661146912472468,
      "eval_loss": 0.23119552433490753,
      "eval_runtime": 68.0286,
      "eval_samples_per_second": 40.057,
      "eval_steps_per_second": 20.036,
      "num_input_tokens_seen": 18497776,
      "step": 9600
    },
    {
      "epoch": 1.5669304184680644,
      "grad_norm": 0.004202737007290125,
      "learning_rate": 0.2593130301953831,
      "loss": 0.2349,
      "num_input_tokens_seen": 18506272,
      "step": 9605
    },
    {
      "epoch": 1.5677461456888815,
      "grad_norm": 0.002438816474750638,
      "learning_rate": 0.2592726851267718,
      "loss": 0.2307,
      "num_input_tokens_seen": 18516304,
      "step": 9610
    },
    {
      "epoch": 1.568561872909699,
      "grad_norm": 0.0011981904972344637,
      "learning_rate": 0.2592323232069393,
      "loss": 0.2473,
      "num_input_tokens_seen": 18525168,
      "step": 9615
    },
    {
      "epoch": 1.5693776001305162,
      "grad_norm": 0.001919917413033545,
      "learning_rate": 0.25919194444210986,
      "loss": 0.2471,
      "num_input_tokens_seen": 18534400,
      "step": 9620
    },
    {
      "epoch": 1.5701933273513338,
      "grad_norm": 0.0013905600644648075,
      "learning_rate": 0.2591515488385103,
      "loss": 0.2334,
      "num_input_tokens_seen": 18545248,
      "step": 9625
    },
    {
      "epoch": 1.5710090545721511,
      "grad_norm": 0.0011460961541160941,
      "learning_rate": 0.2591111364023704,
      "loss": 0.2167,
      "num_input_tokens_seen": 18554128,
      "step": 9630
    },
    {
      "epoch": 1.5718247817929685,
      "grad_norm": 0.0025085043162107468,
      "learning_rate": 0.259070707139922,
      "loss": 0.2391,
      "num_input_tokens_seen": 18564288,
      "step": 9635
    },
    {
      "epoch": 1.5726405090137858,
      "grad_norm": 0.0022234872449189425,
      "learning_rate": 0.25903026105739985,
      "loss": 0.2194,
      "num_input_tokens_seen": 18573840,
      "step": 9640
    },
    {
      "epoch": 1.5734562362346032,
      "grad_norm": 0.0048591503873467445,
      "learning_rate": 0.2589897981610413,
      "loss": 0.2487,
      "num_input_tokens_seen": 18583488,
      "step": 9645
    },
    {
      "epoch": 1.5742719634554205,
      "grad_norm": 0.0022934789303690195,
      "learning_rate": 0.2589493184570863,
      "loss": 0.2455,
      "num_input_tokens_seen": 18593904,
      "step": 9650
    },
    {
      "epoch": 1.5750876906762379,
      "grad_norm": 0.002226062584668398,
      "learning_rate": 0.25890882195177717,
      "loss": 0.2353,
      "num_input_tokens_seen": 18602320,
      "step": 9655
    },
    {
      "epoch": 1.5759034178970552,
      "grad_norm": 0.000653633673209697,
      "learning_rate": 0.25886830865135907,
      "loss": 0.2233,
      "num_input_tokens_seen": 18610432,
      "step": 9660
    },
    {
      "epoch": 1.5767191451178726,
      "grad_norm": 0.0017854557372629642,
      "learning_rate": 0.25882777856207967,
      "loss": 0.2381,
      "num_input_tokens_seen": 18620896,
      "step": 9665
    },
    {
      "epoch": 1.57753487233869,
      "grad_norm": 0.0036375061608850956,
      "learning_rate": 0.2587872316901892,
      "loss": 0.2479,
      "num_input_tokens_seen": 18629376,
      "step": 9670
    },
    {
      "epoch": 1.5783505995595073,
      "grad_norm": 0.0026204430032521486,
      "learning_rate": 0.25874666804194046,
      "loss": 0.2373,
      "num_input_tokens_seen": 18638624,
      "step": 9675
    },
    {
      "epoch": 1.5791663267803246,
      "grad_norm": 0.0028532978612929583,
      "learning_rate": 0.258706087623589,
      "loss": 0.2353,
      "num_input_tokens_seen": 18648224,
      "step": 9680
    },
    {
      "epoch": 1.579982054001142,
      "grad_norm": 0.0028683445416390896,
      "learning_rate": 0.25866549044139264,
      "loss": 0.2324,
      "num_input_tokens_seen": 18658944,
      "step": 9685
    },
    {
      "epoch": 1.5807977812219594,
      "grad_norm": 0.0027263525407761335,
      "learning_rate": 0.25862487650161214,
      "loss": 0.2415,
      "num_input_tokens_seen": 18668528,
      "step": 9690
    },
    {
      "epoch": 1.5816135084427767,
      "grad_norm": 0.0022827668581157923,
      "learning_rate": 0.2585842458105106,
      "loss": 0.2548,
      "num_input_tokens_seen": 18678512,
      "step": 9695
    },
    {
      "epoch": 1.582429235663594,
      "grad_norm": 0.002139634219929576,
      "learning_rate": 0.2585435983743538,
      "loss": 0.245,
      "num_input_tokens_seen": 18688960,
      "step": 9700
    },
    {
      "epoch": 1.5832449628844114,
      "grad_norm": 0.0020701163448393345,
      "learning_rate": 0.2585029341994101,
      "loss": 0.2248,
      "num_input_tokens_seen": 18698992,
      "step": 9705
    },
    {
      "epoch": 1.5840606901052288,
      "grad_norm": 0.002391975140199065,
      "learning_rate": 0.2584622532919504,
      "loss": 0.2305,
      "num_input_tokens_seen": 18708528,
      "step": 9710
    },
    {
      "epoch": 1.5848764173260461,
      "grad_norm": 0.0025031333789229393,
      "learning_rate": 0.2584215556582482,
      "loss": 0.2317,
      "num_input_tokens_seen": 18717680,
      "step": 9715
    },
    {
      "epoch": 1.5856921445468637,
      "grad_norm": 0.002407670021057129,
      "learning_rate": 0.25838084130457967,
      "loss": 0.2252,
      "num_input_tokens_seen": 18727760,
      "step": 9720
    },
    {
      "epoch": 1.5865078717676808,
      "grad_norm": 0.0014448240399360657,
      "learning_rate": 0.2583401102372234,
      "loss": 0.2235,
      "num_input_tokens_seen": 18736416,
      "step": 9725
    },
    {
      "epoch": 1.5873235989884984,
      "grad_norm": 0.0017869352595880628,
      "learning_rate": 0.2582993624624606,
      "loss": 0.2135,
      "num_input_tokens_seen": 18745184,
      "step": 9730
    },
    {
      "epoch": 1.5881393262093155,
      "grad_norm": 0.001563594094477594,
      "learning_rate": 0.25825859798657513,
      "loss": 0.2511,
      "num_input_tokens_seen": 18754224,
      "step": 9735
    },
    {
      "epoch": 1.588955053430133,
      "grad_norm": 0.0018175685545429587,
      "learning_rate": 0.25821781681585343,
      "loss": 0.2121,
      "num_input_tokens_seen": 18765392,
      "step": 9740
    },
    {
      "epoch": 1.5897707806509502,
      "grad_norm": 0.004616181366145611,
      "learning_rate": 0.2581770189565844,
      "loss": 0.2251,
      "num_input_tokens_seen": 18775792,
      "step": 9745
    },
    {
      "epoch": 1.5905865078717678,
      "grad_norm": 0.0015151688130572438,
      "learning_rate": 0.25813620441505963,
      "loss": 0.2488,
      "num_input_tokens_seen": 18784544,
      "step": 9750
    },
    {
      "epoch": 1.591402235092585,
      "grad_norm": 0.0062131499871611595,
      "learning_rate": 0.2580953731975732,
      "loss": 0.2346,
      "num_input_tokens_seen": 18794000,
      "step": 9755
    },
    {
      "epoch": 1.5922179623134025,
      "grad_norm": 0.004002904985100031,
      "learning_rate": 0.2580545253104218,
      "loss": 0.2473,
      "num_input_tokens_seen": 18804528,
      "step": 9760
    },
    {
      "epoch": 1.5930336895342196,
      "grad_norm": 0.0015922305174171925,
      "learning_rate": 0.2580136607599047,
      "loss": 0.2588,
      "num_input_tokens_seen": 18813664,
      "step": 9765
    },
    {
      "epoch": 1.5938494167550372,
      "grad_norm": 0.001958861481398344,
      "learning_rate": 0.2579727795523238,
      "loss": 0.2435,
      "num_input_tokens_seen": 18823104,
      "step": 9770
    },
    {
      "epoch": 1.5946651439758543,
      "grad_norm": 0.0015958542935550213,
      "learning_rate": 0.25793188169398334,
      "loss": 0.2413,
      "num_input_tokens_seen": 18832592,
      "step": 9775
    },
    {
      "epoch": 1.595480871196672,
      "grad_norm": 0.0016708322800695896,
      "learning_rate": 0.25789096719119037,
      "loss": 0.2296,
      "num_input_tokens_seen": 18842640,
      "step": 9780
    },
    {
      "epoch": 1.596296598417489,
      "grad_norm": 0.003366712713614106,
      "learning_rate": 0.2578500360502544,
      "loss": 0.2365,
      "num_input_tokens_seen": 18852016,
      "step": 9785
    },
    {
      "epoch": 1.5971123256383066,
      "grad_norm": 0.001193511881865561,
      "learning_rate": 0.2578090882774876,
      "loss": 0.2253,
      "num_input_tokens_seen": 18862880,
      "step": 9790
    },
    {
      "epoch": 1.5979280528591238,
      "grad_norm": 0.0020248994696885347,
      "learning_rate": 0.25776812387920456,
      "loss": 0.2294,
      "num_input_tokens_seen": 18871312,
      "step": 9795
    },
    {
      "epoch": 1.5987437800799413,
      "grad_norm": 0.0017119103576987982,
      "learning_rate": 0.2577271428617225,
      "loss": 0.2159,
      "num_input_tokens_seen": 18881008,
      "step": 9800
    },
    {
      "epoch": 1.5987437800799413,
      "eval_loss": 0.22693748772144318,
      "eval_runtime": 67.9914,
      "eval_samples_per_second": 40.079,
      "eval_steps_per_second": 20.047,
      "num_input_tokens_seen": 18881008,
      "step": 9800
    },
    {
      "epoch": 1.5995595073007585,
      "grad_norm": 0.0014082045527175069,
      "learning_rate": 0.25768614523136124,
      "loss": 0.2202,
      "num_input_tokens_seen": 18891936,
      "step": 9805
    },
    {
      "epoch": 1.600375234521576,
      "grad_norm": 0.0020443752873688936,
      "learning_rate": 0.25764513099444314,
      "loss": 0.2596,
      "num_input_tokens_seen": 18901440,
      "step": 9810
    },
    {
      "epoch": 1.6011909617423934,
      "grad_norm": 0.002071574330329895,
      "learning_rate": 0.25760410015729307,
      "loss": 0.2271,
      "num_input_tokens_seen": 18911696,
      "step": 9815
    },
    {
      "epoch": 1.6020066889632107,
      "grad_norm": 0.0020300124306231737,
      "learning_rate": 0.2575630527262385,
      "loss": 0.2015,
      "num_input_tokens_seen": 18920416,
      "step": 9820
    },
    {
      "epoch": 1.602822416184028,
      "grad_norm": 0.0017223508330062032,
      "learning_rate": 0.25752198870760945,
      "loss": 0.2411,
      "num_input_tokens_seen": 18930256,
      "step": 9825
    },
    {
      "epoch": 1.6036381434048455,
      "grad_norm": 0.0029891026206314564,
      "learning_rate": 0.2574809081077386,
      "loss": 0.2489,
      "num_input_tokens_seen": 18940400,
      "step": 9830
    },
    {
      "epoch": 1.6044538706256628,
      "grad_norm": 0.001987969968467951,
      "learning_rate": 0.257439810932961,
      "loss": 0.2322,
      "num_input_tokens_seen": 18950032,
      "step": 9835
    },
    {
      "epoch": 1.6052695978464802,
      "grad_norm": 0.002449291292577982,
      "learning_rate": 0.2573986971896144,
      "loss": 0.2327,
      "num_input_tokens_seen": 18959056,
      "step": 9840
    },
    {
      "epoch": 1.6060853250672975,
      "grad_norm": 0.0021837467793375254,
      "learning_rate": 0.257357566884039,
      "loss": 0.2328,
      "num_input_tokens_seen": 18969296,
      "step": 9845
    },
    {
      "epoch": 1.6069010522881149,
      "grad_norm": 0.002780939918011427,
      "learning_rate": 0.25731642002257765,
      "loss": 0.2427,
      "num_input_tokens_seen": 18979920,
      "step": 9850
    },
    {
      "epoch": 1.6077167795089322,
      "grad_norm": 0.0016622469993308187,
      "learning_rate": 0.25727525661157574,
      "loss": 0.2122,
      "num_input_tokens_seen": 18989424,
      "step": 9855
    },
    {
      "epoch": 1.6085325067297496,
      "grad_norm": 0.0016419236781075597,
      "learning_rate": 0.2572340766573811,
      "loss": 0.2275,
      "num_input_tokens_seen": 18998976,
      "step": 9860
    },
    {
      "epoch": 1.609348233950567,
      "grad_norm": 0.001936365501023829,
      "learning_rate": 0.25719288016634434,
      "loss": 0.2077,
      "num_input_tokens_seen": 19007632,
      "step": 9865
    },
    {
      "epoch": 1.6101639611713843,
      "grad_norm": 0.0027846854645758867,
      "learning_rate": 0.25715166714481835,
      "loss": 0.2097,
      "num_input_tokens_seen": 19017568,
      "step": 9870
    },
    {
      "epoch": 1.6109796883922016,
      "grad_norm": 0.002531982259824872,
      "learning_rate": 0.2571104375991587,
      "loss": 0.1984,
      "num_input_tokens_seen": 19027520,
      "step": 9875
    },
    {
      "epoch": 1.611795415613019,
      "grad_norm": 0.00134281103964895,
      "learning_rate": 0.2570691915357236,
      "loss": 0.1552,
      "num_input_tokens_seen": 19035904,
      "step": 9880
    },
    {
      "epoch": 1.6126111428338363,
      "grad_norm": 0.0067843422293663025,
      "learning_rate": 0.2570279289608736,
      "loss": 0.3427,
      "num_input_tokens_seen": 19045360,
      "step": 9885
    },
    {
      "epoch": 1.6134268700546537,
      "grad_norm": 0.0019370380323380232,
      "learning_rate": 0.256986649880972,
      "loss": 0.2075,
      "num_input_tokens_seen": 19056608,
      "step": 9890
    },
    {
      "epoch": 1.614242597275471,
      "grad_norm": 0.002590143121778965,
      "learning_rate": 0.25694535430238447,
      "loss": 0.2382,
      "num_input_tokens_seen": 19066464,
      "step": 9895
    },
    {
      "epoch": 1.6150583244962884,
      "grad_norm": 0.0017215819098055363,
      "learning_rate": 0.25690404223147933,
      "loss": 0.2441,
      "num_input_tokens_seen": 19076016,
      "step": 9900
    },
    {
      "epoch": 1.615874051717106,
      "grad_norm": 0.0015770315658301115,
      "learning_rate": 0.2568627136746275,
      "loss": 0.2123,
      "num_input_tokens_seen": 19084992,
      "step": 9905
    },
    {
      "epoch": 1.616689778937923,
      "grad_norm": 0.0017925620777532458,
      "learning_rate": 0.25682136863820226,
      "loss": 0.2401,
      "num_input_tokens_seen": 19093712,
      "step": 9910
    },
    {
      "epoch": 1.6175055061587407,
      "grad_norm": 0.0012414074735715985,
      "learning_rate": 0.25678000712857957,
      "loss": 0.2301,
      "num_input_tokens_seen": 19103888,
      "step": 9915
    },
    {
      "epoch": 1.6183212333795578,
      "grad_norm": 0.0027545220218598843,
      "learning_rate": 0.2567386291521379,
      "loss": 0.2486,
      "num_input_tokens_seen": 19112640,
      "step": 9920
    },
    {
      "epoch": 1.6191369606003754,
      "grad_norm": 0.0020489317830652,
      "learning_rate": 0.2566972347152583,
      "loss": 0.2503,
      "num_input_tokens_seen": 19122304,
      "step": 9925
    },
    {
      "epoch": 1.6199526878211925,
      "grad_norm": 0.001301864511333406,
      "learning_rate": 0.2566558238243242,
      "loss": 0.2565,
      "num_input_tokens_seen": 19132032,
      "step": 9930
    },
    {
      "epoch": 1.62076841504201,
      "grad_norm": 0.0012663551606237888,
      "learning_rate": 0.25661439648572176,
      "loss": 0.2168,
      "num_input_tokens_seen": 19142432,
      "step": 9935
    },
    {
      "epoch": 1.6215841422628272,
      "grad_norm": 0.002436683513224125,
      "learning_rate": 0.25657295270583963,
      "loss": 0.2211,
      "num_input_tokens_seen": 19151840,
      "step": 9940
    },
    {
      "epoch": 1.6223998694836448,
      "grad_norm": 0.003313264576718211,
      "learning_rate": 0.25653149249106894,
      "loss": 0.2305,
      "num_input_tokens_seen": 19161680,
      "step": 9945
    },
    {
      "epoch": 1.623215596704462,
      "grad_norm": 0.005458814091980457,
      "learning_rate": 0.25649001584780323,
      "loss": 0.2219,
      "num_input_tokens_seen": 19170832,
      "step": 9950
    },
    {
      "epoch": 1.6240313239252795,
      "grad_norm": 0.001353754661977291,
      "learning_rate": 0.2564485227824389,
      "loss": 0.2497,
      "num_input_tokens_seen": 19180176,
      "step": 9955
    },
    {
      "epoch": 1.6248470511460966,
      "grad_norm": 0.001962808193638921,
      "learning_rate": 0.25640701330137466,
      "loss": 0.2334,
      "num_input_tokens_seen": 19188800,
      "step": 9960
    },
    {
      "epoch": 1.6256627783669142,
      "grad_norm": 0.0021031151991337538,
      "learning_rate": 0.2563654874110117,
      "loss": 0.23,
      "num_input_tokens_seen": 19197552,
      "step": 9965
    },
    {
      "epoch": 1.6264785055877313,
      "grad_norm": 0.0014888639561831951,
      "learning_rate": 0.256323945117754,
      "loss": 0.2477,
      "num_input_tokens_seen": 19207536,
      "step": 9970
    },
    {
      "epoch": 1.627294232808549,
      "grad_norm": 0.002816116204485297,
      "learning_rate": 0.2562823864280078,
      "loss": 0.215,
      "num_input_tokens_seen": 19217680,
      "step": 9975
    },
    {
      "epoch": 1.628109960029366,
      "grad_norm": 0.0011761856731027365,
      "learning_rate": 0.25624081134818194,
      "loss": 0.2277,
      "num_input_tokens_seen": 19227328,
      "step": 9980
    },
    {
      "epoch": 1.6289256872501836,
      "grad_norm": 0.002981193596497178,
      "learning_rate": 0.2561992198846879,
      "loss": 0.2128,
      "num_input_tokens_seen": 19237088,
      "step": 9985
    },
    {
      "epoch": 1.6297414144710007,
      "grad_norm": 0.0020584387239068747,
      "learning_rate": 0.25615761204393955,
      "loss": 0.2535,
      "num_input_tokens_seen": 19246416,
      "step": 9990
    },
    {
      "epoch": 1.6305571416918183,
      "grad_norm": 0.0021105504129081964,
      "learning_rate": 0.2561159878323534,
      "loss": 0.2534,
      "num_input_tokens_seen": 19257344,
      "step": 9995
    },
    {
      "epoch": 1.6313728689126357,
      "grad_norm": 0.0012267236597836018,
      "learning_rate": 0.2560743472563483,
      "loss": 0.2333,
      "num_input_tokens_seen": 19266960,
      "step": 10000
    },
    {
      "epoch": 1.6313728689126357,
      "eval_loss": 0.23113077878952026,
      "eval_runtime": 68.0391,
      "eval_samples_per_second": 40.051,
      "eval_steps_per_second": 20.033,
      "num_input_tokens_seen": 19266960,
      "step": 10000
    },
    {
      "epoch": 1.632188596133453,
      "grad_norm": 0.0026901368983089924,
      "learning_rate": 0.25603269032234593,
      "loss": 0.2197,
      "num_input_tokens_seen": 19276304,
      "step": 10005
    },
    {
      "epoch": 1.6330043233542704,
      "grad_norm": 0.00418771943077445,
      "learning_rate": 0.2559910170367702,
      "loss": 0.2335,
      "num_input_tokens_seen": 19286672,
      "step": 10010
    },
    {
      "epoch": 1.6338200505750877,
      "grad_norm": 0.0021358213853091,
      "learning_rate": 0.2559493274060477,
      "loss": 0.2395,
      "num_input_tokens_seen": 19296944,
      "step": 10015
    },
    {
      "epoch": 1.634635777795905,
      "grad_norm": 0.0013099719071760774,
      "learning_rate": 0.2559076214366074,
      "loss": 0.2343,
      "num_input_tokens_seen": 19307120,
      "step": 10020
    },
    {
      "epoch": 1.6354515050167224,
      "grad_norm": 0.001253866241313517,
      "learning_rate": 0.25586589913488106,
      "loss": 0.252,
      "num_input_tokens_seen": 19316384,
      "step": 10025
    },
    {
      "epoch": 1.6362672322375398,
      "grad_norm": 0.001321700750850141,
      "learning_rate": 0.2558241605073026,
      "loss": 0.2316,
      "num_input_tokens_seen": 19325792,
      "step": 10030
    },
    {
      "epoch": 1.6370829594583571,
      "grad_norm": 0.0011437886860221624,
      "learning_rate": 0.25578240556030873,
      "loss": 0.2267,
      "num_input_tokens_seen": 19336448,
      "step": 10035
    },
    {
      "epoch": 1.6378986866791745,
      "grad_norm": 0.0006757881492376328,
      "learning_rate": 0.2557406343003386,
      "loss": 0.2259,
      "num_input_tokens_seen": 19345280,
      "step": 10040
    },
    {
      "epoch": 1.6387144138999918,
      "grad_norm": 0.0013733053347095847,
      "learning_rate": 0.25569884673383375,
      "loss": 0.23,
      "num_input_tokens_seen": 19355248,
      "step": 10045
    },
    {
      "epoch": 1.6395301411208092,
      "grad_norm": 0.002181948395445943,
      "learning_rate": 0.25565704286723856,
      "loss": 0.2392,
      "num_input_tokens_seen": 19364480,
      "step": 10050
    },
    {
      "epoch": 1.6403458683416265,
      "grad_norm": 0.0012376640224829316,
      "learning_rate": 0.25561522270699955,
      "loss": 0.2372,
      "num_input_tokens_seen": 19374688,
      "step": 10055
    },
    {
      "epoch": 1.641161595562444,
      "grad_norm": 0.001221927348524332,
      "learning_rate": 0.25557338625956594,
      "loss": 0.2226,
      "num_input_tokens_seen": 19384320,
      "step": 10060
    },
    {
      "epoch": 1.6419773227832613,
      "grad_norm": 0.0012339529348537326,
      "learning_rate": 0.25553153353138947,
      "loss": 0.2379,
      "num_input_tokens_seen": 19393488,
      "step": 10065
    },
    {
      "epoch": 1.6427930500040786,
      "grad_norm": 0.003245321335271001,
      "learning_rate": 0.2554896645289243,
      "loss": 0.2509,
      "num_input_tokens_seen": 19402128,
      "step": 10070
    },
    {
      "epoch": 1.643608777224896,
      "grad_norm": 0.0017228496726602316,
      "learning_rate": 0.2554477792586272,
      "loss": 0.2247,
      "num_input_tokens_seen": 19410896,
      "step": 10075
    },
    {
      "epoch": 1.6444245044457133,
      "grad_norm": 0.0014326847158372402,
      "learning_rate": 0.25540587772695744,
      "loss": 0.2413,
      "num_input_tokens_seen": 19421360,
      "step": 10080
    },
    {
      "epoch": 1.6452402316665307,
      "grad_norm": 0.001576153328642249,
      "learning_rate": 0.2553639599403767,
      "loss": 0.2861,
      "num_input_tokens_seen": 19431200,
      "step": 10085
    },
    {
      "epoch": 1.6460559588873482,
      "grad_norm": 0.0013655048096552491,
      "learning_rate": 0.2553220259053493,
      "loss": 0.249,
      "num_input_tokens_seen": 19441264,
      "step": 10090
    },
    {
      "epoch": 1.6468716861081654,
      "grad_norm": 0.0020488747395575047,
      "learning_rate": 0.2552800756283419,
      "loss": 0.2121,
      "num_input_tokens_seen": 19451040,
      "step": 10095
    },
    {
      "epoch": 1.647687413328983,
      "grad_norm": 0.0014576752437278628,
      "learning_rate": 0.25523810911582373,
      "loss": 0.2194,
      "num_input_tokens_seen": 19461232,
      "step": 10100
    },
    {
      "epoch": 1.6485031405498,
      "grad_norm": 0.0033030351623892784,
      "learning_rate": 0.25519612637426675,
      "loss": 0.2323,
      "num_input_tokens_seen": 19472000,
      "step": 10105
    },
    {
      "epoch": 1.6493188677706176,
      "grad_norm": 0.003434407291933894,
      "learning_rate": 0.25515412741014504,
      "loss": 0.2278,
      "num_input_tokens_seen": 19481776,
      "step": 10110
    },
    {
      "epoch": 1.6501345949914348,
      "grad_norm": 0.0015406617894768715,
      "learning_rate": 0.2551121122299355,
      "loss": 0.2081,
      "num_input_tokens_seen": 19492560,
      "step": 10115
    },
    {
      "epoch": 1.6509503222122524,
      "grad_norm": 0.0023120343685150146,
      "learning_rate": 0.2550700808401173,
      "loss": 0.2661,
      "num_input_tokens_seen": 19501344,
      "step": 10120
    },
    {
      "epoch": 1.6517660494330695,
      "grad_norm": 0.002236139727756381,
      "learning_rate": 0.2550280332471722,
      "loss": 0.2335,
      "num_input_tokens_seen": 19512048,
      "step": 10125
    },
    {
      "epoch": 1.652581776653887,
      "grad_norm": 0.0013769454089924693,
      "learning_rate": 0.2549859694575845,
      "loss": 0.2366,
      "num_input_tokens_seen": 19521248,
      "step": 10130
    },
    {
      "epoch": 1.6533975038747042,
      "grad_norm": 0.0011995077366009355,
      "learning_rate": 0.254943889477841,
      "loss": 0.2355,
      "num_input_tokens_seen": 19531136,
      "step": 10135
    },
    {
      "epoch": 1.6542132310955218,
      "grad_norm": 0.0012573556741699576,
      "learning_rate": 0.25490179331443097,
      "loss": 0.2526,
      "num_input_tokens_seen": 19540816,
      "step": 10140
    },
    {
      "epoch": 1.655028958316339,
      "grad_norm": 0.004700527526438236,
      "learning_rate": 0.25485968097384615,
      "loss": 0.2578,
      "num_input_tokens_seen": 19550912,
      "step": 10145
    },
    {
      "epoch": 1.6558446855371565,
      "grad_norm": 0.001957460306584835,
      "learning_rate": 0.25481755246258075,
      "loss": 0.2408,
      "num_input_tokens_seen": 19560352,
      "step": 10150
    },
    {
      "epoch": 1.6566604127579736,
      "grad_norm": 0.002900499850511551,
      "learning_rate": 0.2547754077871315,
      "loss": 0.245,
      "num_input_tokens_seen": 19569488,
      "step": 10155
    },
    {
      "epoch": 1.6574761399787912,
      "grad_norm": 0.0030536912381649017,
      "learning_rate": 0.25473324695399774,
      "loss": 0.2176,
      "num_input_tokens_seen": 19579024,
      "step": 10160
    },
    {
      "epoch": 1.6582918671996083,
      "grad_norm": 0.001785146538168192,
      "learning_rate": 0.25469106996968105,
      "loss": 0.2468,
      "num_input_tokens_seen": 19587984,
      "step": 10165
    },
    {
      "epoch": 1.6591075944204259,
      "grad_norm": 0.0011630489025264978,
      "learning_rate": 0.2546488768406858,
      "loss": 0.2198,
      "num_input_tokens_seen": 19596608,
      "step": 10170
    },
    {
      "epoch": 1.659923321641243,
      "grad_norm": 0.0027995239943265915,
      "learning_rate": 0.25460666757351863,
      "loss": 0.2496,
      "num_input_tokens_seen": 19606400,
      "step": 10175
    },
    {
      "epoch": 1.6607390488620606,
      "grad_norm": 0.0015161311021074653,
      "learning_rate": 0.25456444217468877,
      "loss": 0.2305,
      "num_input_tokens_seen": 19614928,
      "step": 10180
    },
    {
      "epoch": 1.661554776082878,
      "grad_norm": 0.0025055173318833113,
      "learning_rate": 0.25452220065070785,
      "loss": 0.246,
      "num_input_tokens_seen": 19623488,
      "step": 10185
    },
    {
      "epoch": 1.6623705033036953,
      "grad_norm": 0.0015762062976136804,
      "learning_rate": 0.2544799430080901,
      "loss": 0.2201,
      "num_input_tokens_seen": 19631584,
      "step": 10190
    },
    {
      "epoch": 1.6631862305245126,
      "grad_norm": 0.004167820792645216,
      "learning_rate": 0.2544376692533522,
      "loss": 0.2363,
      "num_input_tokens_seen": 19641056,
      "step": 10195
    },
    {
      "epoch": 1.66400195774533,
      "grad_norm": 0.0021797558292746544,
      "learning_rate": 0.2543953793930132,
      "loss": 0.2268,
      "num_input_tokens_seen": 19650480,
      "step": 10200
    },
    {
      "epoch": 1.66400195774533,
      "eval_loss": 0.23081189393997192,
      "eval_runtime": 68.1193,
      "eval_samples_per_second": 40.003,
      "eval_steps_per_second": 20.009,
      "num_input_tokens_seen": 19650480,
      "step": 10200
    },
    {
      "epoch": 1.6648176849661473,
      "grad_norm": 0.003333036322146654,
      "learning_rate": 0.2543530734335948,
      "loss": 0.2201,
      "num_input_tokens_seen": 19660592,
      "step": 10205
    },
    {
      "epoch": 1.6656334121869647,
      "grad_norm": 0.00528361089527607,
      "learning_rate": 0.2543107513816211,
      "loss": 0.2742,
      "num_input_tokens_seen": 19670240,
      "step": 10210
    },
    {
      "epoch": 1.666449139407782,
      "grad_norm": 0.0017837763298302889,
      "learning_rate": 0.25426841324361865,
      "loss": 0.2226,
      "num_input_tokens_seen": 19679808,
      "step": 10215
    },
    {
      "epoch": 1.6672648666285994,
      "grad_norm": 0.004247843753546476,
      "learning_rate": 0.2542260590261166,
      "loss": 0.2394,
      "num_input_tokens_seen": 19690432,
      "step": 10220
    },
    {
      "epoch": 1.6680805938494168,
      "grad_norm": 0.001360388589091599,
      "learning_rate": 0.2541836887356465,
      "loss": 0.2355,
      "num_input_tokens_seen": 19699824,
      "step": 10225
    },
    {
      "epoch": 1.6688963210702341,
      "grad_norm": 0.002553654834628105,
      "learning_rate": 0.2541413023787423,
      "loss": 0.2444,
      "num_input_tokens_seen": 19709808,
      "step": 10230
    },
    {
      "epoch": 1.6697120482910515,
      "grad_norm": 0.0017325645312666893,
      "learning_rate": 0.2540988999619405,
      "loss": 0.2369,
      "num_input_tokens_seen": 19719232,
      "step": 10235
    },
    {
      "epoch": 1.6705277755118688,
      "grad_norm": 0.001382621587254107,
      "learning_rate": 0.25405648149178023,
      "loss": 0.224,
      "num_input_tokens_seen": 19729360,
      "step": 10240
    },
    {
      "epoch": 1.6713435027326862,
      "grad_norm": 0.0015915428521111608,
      "learning_rate": 0.2540140469748028,
      "loss": 0.2191,
      "num_input_tokens_seen": 19740160,
      "step": 10245
    },
    {
      "epoch": 1.6721592299535035,
      "grad_norm": 0.0013533898163586855,
      "learning_rate": 0.25397159641755224,
      "loss": 0.2481,
      "num_input_tokens_seen": 19748304,
      "step": 10250
    },
    {
      "epoch": 1.6729749571743209,
      "grad_norm": 0.0027449193876236677,
      "learning_rate": 0.2539291298265749,
      "loss": 0.2432,
      "num_input_tokens_seen": 19758464,
      "step": 10255
    },
    {
      "epoch": 1.6737906843951382,
      "grad_norm": 0.0018205833621323109,
      "learning_rate": 0.2538866472084197,
      "loss": 0.2317,
      "num_input_tokens_seen": 19768992,
      "step": 10260
    },
    {
      "epoch": 1.6746064116159556,
      "grad_norm": 0.0012347976444289088,
      "learning_rate": 0.25384414856963794,
      "loss": 0.2402,
      "num_input_tokens_seen": 19778768,
      "step": 10265
    },
    {
      "epoch": 1.675422138836773,
      "grad_norm": 0.0012882099254056811,
      "learning_rate": 0.25380163391678356,
      "loss": 0.284,
      "num_input_tokens_seen": 19788768,
      "step": 10270
    },
    {
      "epoch": 1.6762378660575903,
      "grad_norm": 0.0023111719638109207,
      "learning_rate": 0.2537591032564127,
      "loss": 0.2347,
      "num_input_tokens_seen": 19798416,
      "step": 10275
    },
    {
      "epoch": 1.6770535932784076,
      "grad_norm": 0.0011848292779177427,
      "learning_rate": 0.25371655659508424,
      "loss": 0.2275,
      "num_input_tokens_seen": 19809136,
      "step": 10280
    },
    {
      "epoch": 1.6778693204992252,
      "grad_norm": 0.002035013400018215,
      "learning_rate": 0.25367399393935935,
      "loss": 0.2463,
      "num_input_tokens_seen": 19818304,
      "step": 10285
    },
    {
      "epoch": 1.6786850477200423,
      "grad_norm": 0.0017330653499811888,
      "learning_rate": 0.25363141529580174,
      "loss": 0.2278,
      "num_input_tokens_seen": 19827120,
      "step": 10290
    },
    {
      "epoch": 1.67950077494086,
      "grad_norm": 0.0014708179514855146,
      "learning_rate": 0.2535888206709776,
      "loss": 0.2346,
      "num_input_tokens_seen": 19837520,
      "step": 10295
    },
    {
      "epoch": 1.680316502161677,
      "grad_norm": 0.0017257622675970197,
      "learning_rate": 0.2535462100714555,
      "loss": 0.2261,
      "num_input_tokens_seen": 19846960,
      "step": 10300
    },
    {
      "epoch": 1.6811322293824946,
      "grad_norm": 0.0011682553449645638,
      "learning_rate": 0.2535035835038066,
      "loss": 0.2392,
      "num_input_tokens_seen": 19856432,
      "step": 10305
    },
    {
      "epoch": 1.6819479566033118,
      "grad_norm": 0.0016494605224579573,
      "learning_rate": 0.2534609409746044,
      "loss": 0.2367,
      "num_input_tokens_seen": 19866128,
      "step": 10310
    },
    {
      "epoch": 1.6827636838241293,
      "grad_norm": 0.001570298452861607,
      "learning_rate": 0.253418282490425,
      "loss": 0.2315,
      "num_input_tokens_seen": 19875568,
      "step": 10315
    },
    {
      "epoch": 1.6835794110449465,
      "grad_norm": 0.0028389643412083387,
      "learning_rate": 0.2533756080578467,
      "loss": 0.2383,
      "num_input_tokens_seen": 19885600,
      "step": 10320
    },
    {
      "epoch": 1.684395138265764,
      "grad_norm": 0.002548340940847993,
      "learning_rate": 0.25333291768345056,
      "loss": 0.2551,
      "num_input_tokens_seen": 19895008,
      "step": 10325
    },
    {
      "epoch": 1.6852108654865812,
      "grad_norm": 0.0012697833590209484,
      "learning_rate": 0.25329021137381996,
      "loss": 0.2311,
      "num_input_tokens_seen": 19903424,
      "step": 10330
    },
    {
      "epoch": 1.6860265927073987,
      "grad_norm": 0.0017438794020563364,
      "learning_rate": 0.25324748913554074,
      "loss": 0.2347,
      "num_input_tokens_seen": 19913504,
      "step": 10335
    },
    {
      "epoch": 1.6868423199282159,
      "grad_norm": 0.001779054175131023,
      "learning_rate": 0.2532047509752013,
      "loss": 0.2323,
      "num_input_tokens_seen": 19921632,
      "step": 10340
    },
    {
      "epoch": 1.6876580471490334,
      "grad_norm": 0.0014671999961137772,
      "learning_rate": 0.25316199689939217,
      "loss": 0.2419,
      "num_input_tokens_seen": 19932256,
      "step": 10345
    },
    {
      "epoch": 1.6884737743698506,
      "grad_norm": 0.001251425826922059,
      "learning_rate": 0.2531192269147068,
      "loss": 0.2408,
      "num_input_tokens_seen": 19942064,
      "step": 10350
    },
    {
      "epoch": 1.6892895015906682,
      "grad_norm": 0.0021402444690465927,
      "learning_rate": 0.2530764410277407,
      "loss": 0.2296,
      "num_input_tokens_seen": 19952144,
      "step": 10355
    },
    {
      "epoch": 1.6901052288114853,
      "grad_norm": 0.003509371541440487,
      "learning_rate": 0.25303363924509203,
      "loss": 0.2186,
      "num_input_tokens_seen": 19962592,
      "step": 10360
    },
    {
      "epoch": 1.6909209560323029,
      "grad_norm": 0.0025024123024195433,
      "learning_rate": 0.25299082157336145,
      "loss": 0.2332,
      "num_input_tokens_seen": 19971520,
      "step": 10365
    },
    {
      "epoch": 1.6917366832531202,
      "grad_norm": 0.00139927479904145,
      "learning_rate": 0.2529479880191519,
      "loss": 0.2374,
      "num_input_tokens_seen": 19980784,
      "step": 10370
    },
    {
      "epoch": 1.6925524104739376,
      "grad_norm": 0.0024958362337201834,
      "learning_rate": 0.2529051385890689,
      "loss": 0.2314,
      "num_input_tokens_seen": 19990528,
      "step": 10375
    },
    {
      "epoch": 1.693368137694755,
      "grad_norm": 0.003127506235614419,
      "learning_rate": 0.2528622732897203,
      "loss": 0.2241,
      "num_input_tokens_seen": 20000000,
      "step": 10380
    },
    {
      "epoch": 1.6941838649155723,
      "grad_norm": 0.0014979597181081772,
      "learning_rate": 0.25281939212771654,
      "loss": 0.1996,
      "num_input_tokens_seen": 20010544,
      "step": 10385
    },
    {
      "epoch": 1.6949995921363896,
      "grad_norm": 0.0016398384468629956,
      "learning_rate": 0.2527764951096704,
      "loss": 0.2618,
      "num_input_tokens_seen": 20020704,
      "step": 10390
    },
    {
      "epoch": 1.695815319357207,
      "grad_norm": 0.001628612051717937,
      "learning_rate": 0.2527335822421971,
      "loss": 0.2435,
      "num_input_tokens_seen": 20030672,
      "step": 10395
    },
    {
      "epoch": 1.6966310465780243,
      "grad_norm": 0.0017737190937623382,
      "learning_rate": 0.25269065353191444,
      "loss": 0.2241,
      "num_input_tokens_seen": 20041120,
      "step": 10400
    },
    {
      "epoch": 1.6966310465780243,
      "eval_loss": 0.2292788177728653,
      "eval_runtime": 67.9982,
      "eval_samples_per_second": 40.075,
      "eval_steps_per_second": 20.045,
      "num_input_tokens_seen": 20041120,
      "step": 10400
    },
    {
      "epoch": 1.6974467737988417,
      "grad_norm": 0.0032972528133541346,
      "learning_rate": 0.2526477089854425,
      "loss": 0.2212,
      "num_input_tokens_seen": 20049536,
      "step": 10405
    },
    {
      "epoch": 1.698262501019659,
      "grad_norm": 0.001181602361612022,
      "learning_rate": 0.25260474860940385,
      "loss": 0.2342,
      "num_input_tokens_seen": 20058064,
      "step": 10410
    },
    {
      "epoch": 1.6990782282404764,
      "grad_norm": 0.001954946666955948,
      "learning_rate": 0.2525617724104236,
      "loss": 0.2188,
      "num_input_tokens_seen": 20066880,
      "step": 10415
    },
    {
      "epoch": 1.6998939554612937,
      "grad_norm": 0.0022260798141360283,
      "learning_rate": 0.25251878039512915,
      "loss": 0.2419,
      "num_input_tokens_seen": 20076432,
      "step": 10420
    },
    {
      "epoch": 1.700709682682111,
      "grad_norm": 0.002773147774860263,
      "learning_rate": 0.25247577257015047,
      "loss": 0.2407,
      "num_input_tokens_seen": 20085856,
      "step": 10425
    },
    {
      "epoch": 1.7015254099029284,
      "grad_norm": 0.001234592404216528,
      "learning_rate": 0.2524327489421198,
      "loss": 0.2161,
      "num_input_tokens_seen": 20096624,
      "step": 10430
    },
    {
      "epoch": 1.7023411371237458,
      "grad_norm": 0.0024376921355724335,
      "learning_rate": 0.25238970951767203,
      "loss": 0.2352,
      "num_input_tokens_seen": 20106976,
      "step": 10435
    },
    {
      "epoch": 1.7031568643445631,
      "grad_norm": 0.002119458746165037,
      "learning_rate": 0.25234665430344433,
      "loss": 0.2296,
      "num_input_tokens_seen": 20117088,
      "step": 10440
    },
    {
      "epoch": 1.7039725915653805,
      "grad_norm": 0.0018520603189244866,
      "learning_rate": 0.2523035833060764,
      "loss": 0.2199,
      "num_input_tokens_seen": 20126688,
      "step": 10445
    },
    {
      "epoch": 1.7047883187861979,
      "grad_norm": 0.0017093720380216837,
      "learning_rate": 0.2522604965322103,
      "loss": 0.1992,
      "num_input_tokens_seen": 20137424,
      "step": 10450
    },
    {
      "epoch": 1.7056040460070152,
      "grad_norm": 0.0017374349990859628,
      "learning_rate": 0.25221739398849047,
      "loss": 0.2325,
      "num_input_tokens_seen": 20147392,
      "step": 10455
    },
    {
      "epoch": 1.7064197732278326,
      "grad_norm": 0.00186823564581573,
      "learning_rate": 0.252174275681564,
      "loss": 0.2667,
      "num_input_tokens_seen": 20156176,
      "step": 10460
    },
    {
      "epoch": 1.70723550044865,
      "grad_norm": 0.004182980861514807,
      "learning_rate": 0.2521311416180802,
      "loss": 0.2381,
      "num_input_tokens_seen": 20164464,
      "step": 10465
    },
    {
      "epoch": 1.7080512276694675,
      "grad_norm": 0.00170889834407717,
      "learning_rate": 0.25208799180469094,
      "loss": 0.2581,
      "num_input_tokens_seen": 20174128,
      "step": 10470
    },
    {
      "epoch": 1.7088669548902846,
      "grad_norm": 0.0010295663960278034,
      "learning_rate": 0.2520448262480504,
      "loss": 0.2275,
      "num_input_tokens_seen": 20182496,
      "step": 10475
    },
    {
      "epoch": 1.7096826821111022,
      "grad_norm": 0.0028300981502979994,
      "learning_rate": 0.25200164495481525,
      "loss": 0.2118,
      "num_input_tokens_seen": 20191440,
      "step": 10480
    },
    {
      "epoch": 1.7104984093319193,
      "grad_norm": 0.0011893343180418015,
      "learning_rate": 0.25195844793164474,
      "loss": 0.2986,
      "num_input_tokens_seen": 20200896,
      "step": 10485
    },
    {
      "epoch": 1.711314136552737,
      "grad_norm": 0.002041210187599063,
      "learning_rate": 0.2519152351852001,
      "loss": 0.2239,
      "num_input_tokens_seen": 20210688,
      "step": 10490
    },
    {
      "epoch": 1.712129863773554,
      "grad_norm": 0.0013325675390660763,
      "learning_rate": 0.25187200672214555,
      "loss": 0.2307,
      "num_input_tokens_seen": 20220848,
      "step": 10495
    },
    {
      "epoch": 1.7129455909943716,
      "grad_norm": 0.0014581596478819847,
      "learning_rate": 0.2518287625491473,
      "loss": 0.2168,
      "num_input_tokens_seen": 20230320,
      "step": 10500
    },
    {
      "epoch": 1.7137613182151887,
      "grad_norm": 0.0009070825763046741,
      "learning_rate": 0.25178550267287425,
      "loss": 0.2311,
      "num_input_tokens_seen": 20239264,
      "step": 10505
    },
    {
      "epoch": 1.7145770454360063,
      "grad_norm": 0.0029337096493691206,
      "learning_rate": 0.2517422270999976,
      "loss": 0.2465,
      "num_input_tokens_seen": 20248768,
      "step": 10510
    },
    {
      "epoch": 1.7153927726568234,
      "grad_norm": 0.0030435039661824703,
      "learning_rate": 0.2516989358371909,
      "loss": 0.2064,
      "num_input_tokens_seen": 20258528,
      "step": 10515
    },
    {
      "epoch": 1.716208499877641,
      "grad_norm": 0.001394623308442533,
      "learning_rate": 0.25165562889113025,
      "loss": 0.2408,
      "num_input_tokens_seen": 20267536,
      "step": 10520
    },
    {
      "epoch": 1.7170242270984581,
      "grad_norm": 0.0013588534202426672,
      "learning_rate": 0.2516123062684942,
      "loss": 0.2298,
      "num_input_tokens_seen": 20277248,
      "step": 10525
    },
    {
      "epoch": 1.7178399543192757,
      "grad_norm": 0.0016353604150936007,
      "learning_rate": 0.25156896797596356,
      "loss": 0.2196,
      "num_input_tokens_seen": 20287664,
      "step": 10530
    },
    {
      "epoch": 1.7186556815400928,
      "grad_norm": 0.0013315384276211262,
      "learning_rate": 0.2515256140202216,
      "loss": 0.2159,
      "num_input_tokens_seen": 20298432,
      "step": 10535
    },
    {
      "epoch": 1.7194714087609104,
      "grad_norm": 0.0016865339130163193,
      "learning_rate": 0.25148224440795425,
      "loss": 0.2456,
      "num_input_tokens_seen": 20308320,
      "step": 10540
    },
    {
      "epoch": 1.7202871359817276,
      "grad_norm": 0.0014633703976869583,
      "learning_rate": 0.2514388591458494,
      "loss": 0.2373,
      "num_input_tokens_seen": 20318976,
      "step": 10545
    },
    {
      "epoch": 1.7211028632025451,
      "grad_norm": 0.002725760918110609,
      "learning_rate": 0.2513954582405977,
      "loss": 0.2469,
      "num_input_tokens_seen": 20328544,
      "step": 10550
    },
    {
      "epoch": 1.7219185904233625,
      "grad_norm": 0.0018906946061179042,
      "learning_rate": 0.2513520416988922,
      "loss": 0.2184,
      "num_input_tokens_seen": 20337552,
      "step": 10555
    },
    {
      "epoch": 1.7227343176441798,
      "grad_norm": 0.0026267929933965206,
      "learning_rate": 0.2513086095274281,
      "loss": 0.2358,
      "num_input_tokens_seen": 20347520,
      "step": 10560
    },
    {
      "epoch": 1.7235500448649972,
      "grad_norm": 0.0006054769619368017,
      "learning_rate": 0.25126516173290336,
      "loss": 0.2241,
      "num_input_tokens_seen": 20356368,
      "step": 10565
    },
    {
      "epoch": 1.7243657720858145,
      "grad_norm": 0.0023478211369365454,
      "learning_rate": 0.2512216983220181,
      "loss": 0.2427,
      "num_input_tokens_seen": 20363888,
      "step": 10570
    },
    {
      "epoch": 1.725181499306632,
      "grad_norm": 0.0037200257647782564,
      "learning_rate": 0.25117821930147494,
      "loss": 0.2381,
      "num_input_tokens_seen": 20373456,
      "step": 10575
    },
    {
      "epoch": 1.7259972265274492,
      "grad_norm": 0.0012208633124828339,
      "learning_rate": 0.2511347246779788,
      "loss": 0.2316,
      "num_input_tokens_seen": 20383600,
      "step": 10580
    },
    {
      "epoch": 1.7268129537482666,
      "grad_norm": 0.0010900446213781834,
      "learning_rate": 0.25109121445823723,
      "loss": 0.2272,
      "num_input_tokens_seen": 20393424,
      "step": 10585
    },
    {
      "epoch": 1.727628680969084,
      "grad_norm": 0.00251594977453351,
      "learning_rate": 0.25104768864896004,
      "loss": 0.2186,
      "num_input_tokens_seen": 20402624,
      "step": 10590
    },
    {
      "epoch": 1.7284444081899013,
      "grad_norm": 0.00361866457387805,
      "learning_rate": 0.2510041472568594,
      "loss": 0.2415,
      "num_input_tokens_seen": 20411760,
      "step": 10595
    },
    {
      "epoch": 1.7292601354107187,
      "grad_norm": 0.0015247144037857652,
      "learning_rate": 0.25096059028864987,
      "loss": 0.2057,
      "num_input_tokens_seen": 20421120,
      "step": 10600
    },
    {
      "epoch": 1.7292601354107187,
      "eval_loss": 0.22477734088897705,
      "eval_runtime": 68.1138,
      "eval_samples_per_second": 40.007,
      "eval_steps_per_second": 20.011,
      "num_input_tokens_seen": 20421120,
      "step": 10600
    },
    {
      "epoch": 1.730075862631536,
      "grad_norm": 0.0017314128344878554,
      "learning_rate": 0.25091701775104863,
      "loss": 0.2299,
      "num_input_tokens_seen": 20430096,
      "step": 10605
    },
    {
      "epoch": 1.7308915898523534,
      "grad_norm": 0.00301167881116271,
      "learning_rate": 0.250873429650775,
      "loss": 0.1975,
      "num_input_tokens_seen": 20440304,
      "step": 10610
    },
    {
      "epoch": 1.7317073170731707,
      "grad_norm": 0.001609745086170733,
      "learning_rate": 0.25082982599455095,
      "loss": 0.2493,
      "num_input_tokens_seen": 20448496,
      "step": 10615
    },
    {
      "epoch": 1.732523044293988,
      "grad_norm": 0.0023928768932819366,
      "learning_rate": 0.2507862067891006,
      "loss": 0.2477,
      "num_input_tokens_seen": 20457104,
      "step": 10620
    },
    {
      "epoch": 1.7333387715148054,
      "grad_norm": 0.0036621647886931896,
      "learning_rate": 0.25074257204115064,
      "loss": 0.2424,
      "num_input_tokens_seen": 20467424,
      "step": 10625
    },
    {
      "epoch": 1.7341544987356228,
      "grad_norm": 0.0023163678124547005,
      "learning_rate": 0.25069892175742997,
      "loss": 0.2357,
      "num_input_tokens_seen": 20479136,
      "step": 10630
    },
    {
      "epoch": 1.7349702259564401,
      "grad_norm": 0.001670380588620901,
      "learning_rate": 0.25065525594467014,
      "loss": 0.2442,
      "num_input_tokens_seen": 20489888,
      "step": 10635
    },
    {
      "epoch": 1.7357859531772575,
      "grad_norm": 0.0015359403332695365,
      "learning_rate": 0.2506115746096049,
      "loss": 0.2206,
      "num_input_tokens_seen": 20499392,
      "step": 10640
    },
    {
      "epoch": 1.7366016803980748,
      "grad_norm": 0.0021711140871047974,
      "learning_rate": 0.25056787775897055,
      "loss": 0.2128,
      "num_input_tokens_seen": 20509232,
      "step": 10645
    },
    {
      "epoch": 1.7374174076188922,
      "grad_norm": 0.0011166883632540703,
      "learning_rate": 0.2505241653995056,
      "loss": 0.2329,
      "num_input_tokens_seen": 20518112,
      "step": 10650
    },
    {
      "epoch": 1.7382331348397098,
      "grad_norm": 0.0017210873775184155,
      "learning_rate": 0.25048043753795113,
      "loss": 0.2659,
      "num_input_tokens_seen": 20527792,
      "step": 10655
    },
    {
      "epoch": 1.739048862060527,
      "grad_norm": 0.0010069379350170493,
      "learning_rate": 0.2504366941810504,
      "loss": 0.2202,
      "num_input_tokens_seen": 20537824,
      "step": 10660
    },
    {
      "epoch": 1.7398645892813445,
      "grad_norm": 0.0016365977935492992,
      "learning_rate": 0.2503929353355493,
      "loss": 0.2118,
      "num_input_tokens_seen": 20546432,
      "step": 10665
    },
    {
      "epoch": 1.7406803165021616,
      "grad_norm": 0.0034956648014485836,
      "learning_rate": 0.250349161008196,
      "loss": 0.2468,
      "num_input_tokens_seen": 20555632,
      "step": 10670
    },
    {
      "epoch": 1.7414960437229792,
      "grad_norm": 0.0024707040283828974,
      "learning_rate": 0.2503053712057409,
      "loss": 0.2189,
      "num_input_tokens_seen": 20564864,
      "step": 10675
    },
    {
      "epoch": 1.7423117709437963,
      "grad_norm": 0.00359166762791574,
      "learning_rate": 0.25026156593493715,
      "loss": 0.2426,
      "num_input_tokens_seen": 20574448,
      "step": 10680
    },
    {
      "epoch": 1.7431274981646139,
      "grad_norm": 0.001431249431334436,
      "learning_rate": 0.2502177452025399,
      "loss": 0.201,
      "num_input_tokens_seen": 20583856,
      "step": 10685
    },
    {
      "epoch": 1.743943225385431,
      "grad_norm": 0.00328878709115088,
      "learning_rate": 0.25017390901530695,
      "loss": 0.231,
      "num_input_tokens_seen": 20592784,
      "step": 10690
    },
    {
      "epoch": 1.7447589526062486,
      "grad_norm": 0.00158664770424366,
      "learning_rate": 0.2501300573799984,
      "loss": 0.2334,
      "num_input_tokens_seen": 20602400,
      "step": 10695
    },
    {
      "epoch": 1.7455746798270657,
      "grad_norm": 0.0014469340676441789,
      "learning_rate": 0.2500861903033766,
      "loss": 0.2108,
      "num_input_tokens_seen": 20612784,
      "step": 10700
    },
    {
      "epoch": 1.7463904070478833,
      "grad_norm": 0.0017409343272447586,
      "learning_rate": 0.25004230779220654,
      "loss": 0.2066,
      "num_input_tokens_seen": 20623280,
      "step": 10705
    },
    {
      "epoch": 1.7472061342687004,
      "grad_norm": 0.0014545656740665436,
      "learning_rate": 0.24999840985325542,
      "loss": 0.2613,
      "num_input_tokens_seen": 20632592,
      "step": 10710
    },
    {
      "epoch": 1.748021861489518,
      "grad_norm": 0.001966852694749832,
      "learning_rate": 0.24995449649329285,
      "loss": 0.2371,
      "num_input_tokens_seen": 20641616,
      "step": 10715
    },
    {
      "epoch": 1.7488375887103351,
      "grad_norm": 0.0012363120913505554,
      "learning_rate": 0.2499105677190908,
      "loss": 0.2229,
      "num_input_tokens_seen": 20652048,
      "step": 10720
    },
    {
      "epoch": 1.7496533159311527,
      "grad_norm": 0.00212880945764482,
      "learning_rate": 0.24986662353742364,
      "loss": 0.2351,
      "num_input_tokens_seen": 20660720,
      "step": 10725
    },
    {
      "epoch": 1.7504690431519698,
      "grad_norm": 0.0023662138264626265,
      "learning_rate": 0.24982266395506814,
      "loss": 0.2062,
      "num_input_tokens_seen": 20669680,
      "step": 10730
    },
    {
      "epoch": 1.7512847703727874,
      "grad_norm": 0.0022597750648856163,
      "learning_rate": 0.2497786889788034,
      "loss": 0.2172,
      "num_input_tokens_seen": 20678800,
      "step": 10735
    },
    {
      "epoch": 1.7521004975936045,
      "grad_norm": 0.0023018494248390198,
      "learning_rate": 0.24973469861541095,
      "loss": 0.2211,
      "num_input_tokens_seen": 20688688,
      "step": 10740
    },
    {
      "epoch": 1.752916224814422,
      "grad_norm": 0.0014088815078139305,
      "learning_rate": 0.24969069287167456,
      "loss": 0.2078,
      "num_input_tokens_seen": 20698912,
      "step": 10745
    },
    {
      "epoch": 1.7537319520352395,
      "grad_norm": 0.000693154928740114,
      "learning_rate": 0.2496466717543806,
      "loss": 0.1893,
      "num_input_tokens_seen": 20708416,
      "step": 10750
    },
    {
      "epoch": 1.7545476792560568,
      "grad_norm": 0.0017492228653281927,
      "learning_rate": 0.24960263527031762,
      "loss": 0.2635,
      "num_input_tokens_seen": 20717328,
      "step": 10755
    },
    {
      "epoch": 1.7553634064768742,
      "grad_norm": 0.0014704514760524035,
      "learning_rate": 0.24955858342627657,
      "loss": 0.2556,
      "num_input_tokens_seen": 20728272,
      "step": 10760
    },
    {
      "epoch": 1.7561791336976915,
      "grad_norm": 0.0014936260413378477,
      "learning_rate": 0.24951451622905083,
      "loss": 0.2103,
      "num_input_tokens_seen": 20738928,
      "step": 10765
    },
    {
      "epoch": 1.7569948609185089,
      "grad_norm": 0.002130186418071389,
      "learning_rate": 0.24947043368543612,
      "loss": 0.2509,
      "num_input_tokens_seen": 20750176,
      "step": 10770
    },
    {
      "epoch": 1.7578105881393262,
      "grad_norm": 0.0013634765055030584,
      "learning_rate": 0.2494263358022305,
      "loss": 0.2271,
      "num_input_tokens_seen": 20759584,
      "step": 10775
    },
    {
      "epoch": 1.7586263153601436,
      "grad_norm": 0.002186310477554798,
      "learning_rate": 0.24938222258623444,
      "loss": 0.2197,
      "num_input_tokens_seen": 20768896,
      "step": 10780
    },
    {
      "epoch": 1.759442042580961,
      "grad_norm": 0.0015529809752479196,
      "learning_rate": 0.24933809404425075,
      "loss": 0.2203,
      "num_input_tokens_seen": 20779232,
      "step": 10785
    },
    {
      "epoch": 1.7602577698017783,
      "grad_norm": 0.0012766167055815458,
      "learning_rate": 0.24929395018308453,
      "loss": 0.2626,
      "num_input_tokens_seen": 20788848,
      "step": 10790
    },
    {
      "epoch": 1.7610734970225956,
      "grad_norm": 0.0010596693027764559,
      "learning_rate": 0.24924979100954348,
      "loss": 0.2533,
      "num_input_tokens_seen": 20799216,
      "step": 10795
    },
    {
      "epoch": 1.761889224243413,
      "grad_norm": 0.000984309590421617,
      "learning_rate": 0.24920561653043735,
      "loss": 0.2239,
      "num_input_tokens_seen": 20808496,
      "step": 10800
    },
    {
      "epoch": 1.761889224243413,
      "eval_loss": 0.22534197568893433,
      "eval_runtime": 68.0664,
      "eval_samples_per_second": 40.034,
      "eval_steps_per_second": 20.025,
      "num_input_tokens_seen": 20808496,
      "step": 10800
    },
    {
      "epoch": 1.7627049514642303,
      "grad_norm": 0.0012264379765838385,
      "learning_rate": 0.24916142675257846,
      "loss": 0.2032,
      "num_input_tokens_seen": 20817600,
      "step": 10805
    },
    {
      "epoch": 1.7635206786850477,
      "grad_norm": 0.0014392082812264562,
      "learning_rate": 0.24911722168278144,
      "loss": 0.2647,
      "num_input_tokens_seen": 20826496,
      "step": 10810
    },
    {
      "epoch": 1.764336405905865,
      "grad_norm": 0.001558618969283998,
      "learning_rate": 0.24907300132786328,
      "loss": 0.2383,
      "num_input_tokens_seen": 20835952,
      "step": 10815
    },
    {
      "epoch": 1.7651521331266824,
      "grad_norm": 0.0013957723276689649,
      "learning_rate": 0.24902876569464322,
      "loss": 0.224,
      "num_input_tokens_seen": 20846800,
      "step": 10820
    },
    {
      "epoch": 1.7659678603474998,
      "grad_norm": 0.0019253146601840854,
      "learning_rate": 0.24898451478994305,
      "loss": 0.2373,
      "num_input_tokens_seen": 20855968,
      "step": 10825
    },
    {
      "epoch": 1.766783587568317,
      "grad_norm": 0.0011160647263750434,
      "learning_rate": 0.2489402486205868,
      "loss": 0.2106,
      "num_input_tokens_seen": 20867360,
      "step": 10830
    },
    {
      "epoch": 1.7675993147891345,
      "grad_norm": 0.0014835736947134137,
      "learning_rate": 0.24889596719340085,
      "loss": 0.2402,
      "num_input_tokens_seen": 20875616,
      "step": 10835
    },
    {
      "epoch": 1.768415042009952,
      "grad_norm": 0.0026930994354188442,
      "learning_rate": 0.24885167051521392,
      "loss": 0.246,
      "num_input_tokens_seen": 20883792,
      "step": 10840
    },
    {
      "epoch": 1.7692307692307692,
      "grad_norm": 0.0009651876171119511,
      "learning_rate": 0.24880735859285716,
      "loss": 0.223,
      "num_input_tokens_seen": 20893920,
      "step": 10845
    },
    {
      "epoch": 1.7700464964515867,
      "grad_norm": 0.001565054408274591,
      "learning_rate": 0.24876303143316406,
      "loss": 0.2429,
      "num_input_tokens_seen": 20902704,
      "step": 10850
    },
    {
      "epoch": 1.7708622236724039,
      "grad_norm": 0.0018197577446699142,
      "learning_rate": 0.24871868904297031,
      "loss": 0.2221,
      "num_input_tokens_seen": 20913008,
      "step": 10855
    },
    {
      "epoch": 1.7716779508932214,
      "grad_norm": 0.0012058975407853723,
      "learning_rate": 0.24867433142911416,
      "loss": 0.2119,
      "num_input_tokens_seen": 20922896,
      "step": 10860
    },
    {
      "epoch": 1.7724936781140386,
      "grad_norm": 0.0028170763980597258,
      "learning_rate": 0.24862995859843612,
      "loss": 0.2573,
      "num_input_tokens_seen": 20931664,
      "step": 10865
    },
    {
      "epoch": 1.7733094053348561,
      "grad_norm": 0.0020669244695454836,
      "learning_rate": 0.24858557055777897,
      "loss": 0.1807,
      "num_input_tokens_seen": 20941632,
      "step": 10870
    },
    {
      "epoch": 1.7741251325556733,
      "grad_norm": 0.0012074465630576015,
      "learning_rate": 0.24854116731398793,
      "loss": 0.2483,
      "num_input_tokens_seen": 20950928,
      "step": 10875
    },
    {
      "epoch": 1.7749408597764909,
      "grad_norm": 0.002337052021175623,
      "learning_rate": 0.24849674887391052,
      "loss": 0.2154,
      "num_input_tokens_seen": 20962144,
      "step": 10880
    },
    {
      "epoch": 1.775756586997308,
      "grad_norm": 0.003511512652039528,
      "learning_rate": 0.2484523152443967,
      "loss": 0.2817,
      "num_input_tokens_seen": 20972512,
      "step": 10885
    },
    {
      "epoch": 1.7765723142181256,
      "grad_norm": 0.002761411713436246,
      "learning_rate": 0.24840786643229862,
      "loss": 0.2548,
      "num_input_tokens_seen": 20981824,
      "step": 10890
    },
    {
      "epoch": 1.7773880414389427,
      "grad_norm": 0.002364045474678278,
      "learning_rate": 0.2483634024444709,
      "loss": 0.2415,
      "num_input_tokens_seen": 20991472,
      "step": 10895
    },
    {
      "epoch": 1.7782037686597603,
      "grad_norm": 0.002116560935974121,
      "learning_rate": 0.24831892328777033,
      "loss": 0.2272,
      "num_input_tokens_seen": 21001664,
      "step": 10900
    },
    {
      "epoch": 1.7790194958805774,
      "grad_norm": 0.001988199772313237,
      "learning_rate": 0.2482744289690563,
      "loss": 0.2229,
      "num_input_tokens_seen": 21010864,
      "step": 10905
    },
    {
      "epoch": 1.779835223101395,
      "grad_norm": 0.0014510821783915162,
      "learning_rate": 0.2482299194951903,
      "loss": 0.2352,
      "num_input_tokens_seen": 21020480,
      "step": 10910
    },
    {
      "epoch": 1.780650950322212,
      "grad_norm": 0.0009620221680961549,
      "learning_rate": 0.2481853948730363,
      "loss": 0.2511,
      "num_input_tokens_seen": 21029568,
      "step": 10915
    },
    {
      "epoch": 1.7814666775430297,
      "grad_norm": 0.0013423198834061623,
      "learning_rate": 0.24814085510946052,
      "loss": 0.2143,
      "num_input_tokens_seen": 21039136,
      "step": 10920
    },
    {
      "epoch": 1.7822824047638468,
      "grad_norm": 0.0009644190431572497,
      "learning_rate": 0.24809630021133158,
      "loss": 0.2267,
      "num_input_tokens_seen": 21048752,
      "step": 10925
    },
    {
      "epoch": 1.7830981319846644,
      "grad_norm": 0.002049872884526849,
      "learning_rate": 0.24805173018552037,
      "loss": 0.2292,
      "num_input_tokens_seen": 21057872,
      "step": 10930
    },
    {
      "epoch": 1.7839138592054817,
      "grad_norm": 0.0009521484607830644,
      "learning_rate": 0.2480071450389002,
      "loss": 0.2305,
      "num_input_tokens_seen": 21067264,
      "step": 10935
    },
    {
      "epoch": 1.784729586426299,
      "grad_norm": 0.0018820982659235597,
      "learning_rate": 0.24796254477834662,
      "loss": 0.2423,
      "num_input_tokens_seen": 21076304,
      "step": 10940
    },
    {
      "epoch": 1.7855453136471164,
      "grad_norm": 0.00099936465267092,
      "learning_rate": 0.24791792941073754,
      "loss": 0.2208,
      "num_input_tokens_seen": 21085696,
      "step": 10945
    },
    {
      "epoch": 1.7863610408679338,
      "grad_norm": 0.0018515229457989335,
      "learning_rate": 0.2478732989429533,
      "loss": 0.2461,
      "num_input_tokens_seen": 21095952,
      "step": 10950
    },
    {
      "epoch": 1.7871767680887511,
      "grad_norm": 0.0010470764245837927,
      "learning_rate": 0.24782865338187632,
      "loss": 0.2271,
      "num_input_tokens_seen": 21105056,
      "step": 10955
    },
    {
      "epoch": 1.7879924953095685,
      "grad_norm": 0.0009965684730559587,
      "learning_rate": 0.2477839927343916,
      "loss": 0.2295,
      "num_input_tokens_seen": 21113904,
      "step": 10960
    },
    {
      "epoch": 1.7888082225303858,
      "grad_norm": 0.0016708016628399491,
      "learning_rate": 0.2477393170073864,
      "loss": 0.2327,
      "num_input_tokens_seen": 21124224,
      "step": 10965
    },
    {
      "epoch": 1.7896239497512032,
      "grad_norm": 0.0011344641679897904,
      "learning_rate": 0.2476946262077503,
      "loss": 0.2208,
      "num_input_tokens_seen": 21134704,
      "step": 10970
    },
    {
      "epoch": 1.7904396769720206,
      "grad_norm": 0.003951904363930225,
      "learning_rate": 0.24764992034237507,
      "loss": 0.2284,
      "num_input_tokens_seen": 21144864,
      "step": 10975
    },
    {
      "epoch": 1.791255404192838,
      "grad_norm": 0.0010573134059086442,
      "learning_rate": 0.24760519941815498,
      "loss": 0.2221,
      "num_input_tokens_seen": 21155840,
      "step": 10980
    },
    {
      "epoch": 1.7920711314136553,
      "grad_norm": 0.000849330157507211,
      "learning_rate": 0.2475604634419866,
      "loss": 0.2263,
      "num_input_tokens_seen": 21166272,
      "step": 10985
    },
    {
      "epoch": 1.7928868586344726,
      "grad_norm": 0.0023040161468088627,
      "learning_rate": 0.24751571242076872,
      "loss": 0.2275,
      "num_input_tokens_seen": 21176416,
      "step": 10990
    },
    {
      "epoch": 1.79370258585529,
      "grad_norm": 0.0012356723891571164,
      "learning_rate": 0.2474709463614025,
      "loss": 0.2352,
      "num_input_tokens_seen": 21185680,
      "step": 10995
    },
    {
      "epoch": 1.7945183130761073,
      "grad_norm": 0.003314385889098048,
      "learning_rate": 0.24742616527079145,
      "loss": 0.2168,
      "num_input_tokens_seen": 21195024,
      "step": 11000
    },
    {
      "epoch": 1.7945183130761073,
      "eval_loss": 0.23574867844581604,
      "eval_runtime": 68.0733,
      "eval_samples_per_second": 40.03,
      "eval_steps_per_second": 20.023,
      "num_input_tokens_seen": 21195024,
      "step": 11000
    },
    {
      "epoch": 1.7953340402969247,
      "grad_norm": 0.0016484203515574336,
      "learning_rate": 0.24738136915584139,
      "loss": 0.2269,
      "num_input_tokens_seen": 21203376,
      "step": 11005
    },
    {
      "epoch": 1.796149767517742,
      "grad_norm": 0.0014119280967861414,
      "learning_rate": 0.24733655802346047,
      "loss": 0.234,
      "num_input_tokens_seen": 21212672,
      "step": 11010
    },
    {
      "epoch": 1.7969654947385594,
      "grad_norm": 0.0009673930471763015,
      "learning_rate": 0.24729173188055906,
      "loss": 0.257,
      "num_input_tokens_seen": 21222656,
      "step": 11015
    },
    {
      "epoch": 1.7977812219593767,
      "grad_norm": 0.0014312005368992686,
      "learning_rate": 0.24724689073404996,
      "loss": 0.2432,
      "num_input_tokens_seen": 21231456,
      "step": 11020
    },
    {
      "epoch": 1.7985969491801943,
      "grad_norm": 0.0024729419965296984,
      "learning_rate": 0.24720203459084822,
      "loss": 0.2271,
      "num_input_tokens_seen": 21241568,
      "step": 11025
    },
    {
      "epoch": 1.7994126764010114,
      "grad_norm": 0.0020969815086573362,
      "learning_rate": 0.24715716345787123,
      "loss": 0.2309,
      "num_input_tokens_seen": 21251136,
      "step": 11030
    },
    {
      "epoch": 1.800228403621829,
      "grad_norm": 0.003673459868878126,
      "learning_rate": 0.2471122773420387,
      "loss": 0.2481,
      "num_input_tokens_seen": 21262320,
      "step": 11035
    },
    {
      "epoch": 1.8010441308426461,
      "grad_norm": 0.0022751367650926113,
      "learning_rate": 0.24706737625027259,
      "loss": 0.2316,
      "num_input_tokens_seen": 21272256,
      "step": 11040
    },
    {
      "epoch": 1.8018598580634637,
      "grad_norm": 0.0014369059354066849,
      "learning_rate": 0.24702246018949725,
      "loss": 0.2354,
      "num_input_tokens_seen": 21281600,
      "step": 11045
    },
    {
      "epoch": 1.8026755852842808,
      "grad_norm": 0.0013286267640069127,
      "learning_rate": 0.2469775291666393,
      "loss": 0.2572,
      "num_input_tokens_seen": 21291072,
      "step": 11050
    },
    {
      "epoch": 1.8034913125050984,
      "grad_norm": 0.0030348654836416245,
      "learning_rate": 0.24693258318862765,
      "loss": 0.2296,
      "num_input_tokens_seen": 21300624,
      "step": 11055
    },
    {
      "epoch": 1.8043070397259156,
      "grad_norm": 0.0012513361871242523,
      "learning_rate": 0.2468876222623935,
      "loss": 0.2413,
      "num_input_tokens_seen": 21310656,
      "step": 11060
    },
    {
      "epoch": 1.8051227669467331,
      "grad_norm": 0.0009705309639684856,
      "learning_rate": 0.2468426463948705,
      "loss": 0.2353,
      "num_input_tokens_seen": 21319200,
      "step": 11065
    },
    {
      "epoch": 1.8059384941675503,
      "grad_norm": 0.00243544764816761,
      "learning_rate": 0.24679765559299438,
      "loss": 0.2141,
      "num_input_tokens_seen": 21328576,
      "step": 11070
    },
    {
      "epoch": 1.8067542213883678,
      "grad_norm": 0.0025005782954394817,
      "learning_rate": 0.24675264986370332,
      "loss": 0.264,
      "num_input_tokens_seen": 21338176,
      "step": 11075
    },
    {
      "epoch": 1.807569948609185,
      "grad_norm": 0.003744306741282344,
      "learning_rate": 0.2467076292139378,
      "loss": 0.2197,
      "num_input_tokens_seen": 21347936,
      "step": 11080
    },
    {
      "epoch": 1.8083856758300025,
      "grad_norm": 0.0022264057770371437,
      "learning_rate": 0.24666259365064055,
      "loss": 0.2455,
      "num_input_tokens_seen": 21356672,
      "step": 11085
    },
    {
      "epoch": 1.8092014030508197,
      "grad_norm": 0.0020980960689485073,
      "learning_rate": 0.24661754318075663,
      "loss": 0.22,
      "num_input_tokens_seen": 21366288,
      "step": 11090
    },
    {
      "epoch": 1.8100171302716372,
      "grad_norm": 0.0014339779736474156,
      "learning_rate": 0.2465724778112334,
      "loss": 0.2362,
      "num_input_tokens_seen": 21375024,
      "step": 11095
    },
    {
      "epoch": 1.8108328574924544,
      "grad_norm": 0.0028386267367750406,
      "learning_rate": 0.24652739754902042,
      "loss": 0.2391,
      "num_input_tokens_seen": 21385136,
      "step": 11100
    },
    {
      "epoch": 1.811648584713272,
      "grad_norm": 0.0010810961248353124,
      "learning_rate": 0.24648230240106975,
      "loss": 0.2185,
      "num_input_tokens_seen": 21393024,
      "step": 11105
    },
    {
      "epoch": 1.812464311934089,
      "grad_norm": 0.003751607146114111,
      "learning_rate": 0.2464371923743356,
      "loss": 0.2562,
      "num_input_tokens_seen": 21401520,
      "step": 11110
    },
    {
      "epoch": 1.8132800391549067,
      "grad_norm": 0.001471756142564118,
      "learning_rate": 0.24639206747577444,
      "loss": 0.2109,
      "num_input_tokens_seen": 21410544,
      "step": 11115
    },
    {
      "epoch": 1.814095766375724,
      "grad_norm": 0.000838826410472393,
      "learning_rate": 0.24634692771234515,
      "loss": 0.2269,
      "num_input_tokens_seen": 21418656,
      "step": 11120
    },
    {
      "epoch": 1.8149114935965414,
      "grad_norm": 0.0007130649755708873,
      "learning_rate": 0.2463017730910088,
      "loss": 0.2372,
      "num_input_tokens_seen": 21427584,
      "step": 11125
    },
    {
      "epoch": 1.8157272208173587,
      "grad_norm": 0.0018708629067987204,
      "learning_rate": 0.2462566036187289,
      "loss": 0.2774,
      "num_input_tokens_seen": 21436416,
      "step": 11130
    },
    {
      "epoch": 1.816542948038176,
      "grad_norm": 0.0034397945273667574,
      "learning_rate": 0.24621141930247106,
      "loss": 0.2531,
      "num_input_tokens_seen": 21444896,
      "step": 11135
    },
    {
      "epoch": 1.8173586752589934,
      "grad_norm": 0.001884732861071825,
      "learning_rate": 0.2461662201492033,
      "loss": 0.2448,
      "num_input_tokens_seen": 21453712,
      "step": 11140
    },
    {
      "epoch": 1.8181744024798108,
      "grad_norm": 0.001278682961128652,
      "learning_rate": 0.24612100616589586,
      "loss": 0.2134,
      "num_input_tokens_seen": 21462704,
      "step": 11145
    },
    {
      "epoch": 1.8189901297006281,
      "grad_norm": 0.0011790540302172303,
      "learning_rate": 0.24607577735952135,
      "loss": 0.198,
      "num_input_tokens_seen": 21471024,
      "step": 11150
    },
    {
      "epoch": 1.8198058569214455,
      "grad_norm": 0.0009693565079942346,
      "learning_rate": 0.24603053373705464,
      "loss": 0.2225,
      "num_input_tokens_seen": 21480352,
      "step": 11155
    },
    {
      "epoch": 1.8206215841422628,
      "grad_norm": 0.002764699747785926,
      "learning_rate": 0.2459852753054728,
      "loss": 0.2825,
      "num_input_tokens_seen": 21490352,
      "step": 11160
    },
    {
      "epoch": 1.8214373113630802,
      "grad_norm": 0.0015854050870984793,
      "learning_rate": 0.24594000207175526,
      "loss": 0.2138,
      "num_input_tokens_seen": 21501216,
      "step": 11165
    },
    {
      "epoch": 1.8222530385838975,
      "grad_norm": 0.0026748625095933676,
      "learning_rate": 0.2458947140428838,
      "loss": 0.2319,
      "num_input_tokens_seen": 21510704,
      "step": 11170
    },
    {
      "epoch": 1.8230687658047149,
      "grad_norm": 0.0020599381532520056,
      "learning_rate": 0.24584941122584233,
      "loss": 0.2277,
      "num_input_tokens_seen": 21520704,
      "step": 11175
    },
    {
      "epoch": 1.8238844930255322,
      "grad_norm": 0.0008336493629030883,
      "learning_rate": 0.24580409362761713,
      "loss": 0.2494,
      "num_input_tokens_seen": 21530016,
      "step": 11180
    },
    {
      "epoch": 1.8247002202463496,
      "grad_norm": 0.002256967592984438,
      "learning_rate": 0.2457587612551967,
      "loss": 0.2673,
      "num_input_tokens_seen": 21540032,
      "step": 11185
    },
    {
      "epoch": 1.825515947467167,
      "grad_norm": 0.0018488471396267414,
      "learning_rate": 0.24571341411557193,
      "loss": 0.2411,
      "num_input_tokens_seen": 21550208,
      "step": 11190
    },
    {
      "epoch": 1.8263316746879843,
      "grad_norm": 0.0012153503485023975,
      "learning_rate": 0.2456680522157359,
      "loss": 0.2309,
      "num_input_tokens_seen": 21560688,
      "step": 11195
    },
    {
      "epoch": 1.8271474019088016,
      "grad_norm": 0.0011326720705255866,
      "learning_rate": 0.245622675562684,
      "loss": 0.2592,
      "num_input_tokens_seen": 21570368,
      "step": 11200
    },
    {
      "epoch": 1.8271474019088016,
      "eval_loss": 0.22525036334991455,
      "eval_runtime": 67.9781,
      "eval_samples_per_second": 40.086,
      "eval_steps_per_second": 20.051,
      "num_input_tokens_seen": 21570368,
      "step": 11200
    },
    {
      "epoch": 1.827963129129619,
      "grad_norm": 0.0021561430767178535,
      "learning_rate": 0.24557728416341384,
      "loss": 0.243,
      "num_input_tokens_seen": 21580208,
      "step": 11205
    },
    {
      "epoch": 1.8287788563504366,
      "grad_norm": 0.0018531590467318892,
      "learning_rate": 0.24553187802492538,
      "loss": 0.2245,
      "num_input_tokens_seen": 21589920,
      "step": 11210
    },
    {
      "epoch": 1.8295945835712537,
      "grad_norm": 0.0007510501891374588,
      "learning_rate": 0.24548645715422074,
      "loss": 0.2257,
      "num_input_tokens_seen": 21598256,
      "step": 11215
    },
    {
      "epoch": 1.8304103107920713,
      "grad_norm": 0.0007960456423461437,
      "learning_rate": 0.2454410215583045,
      "loss": 0.2202,
      "num_input_tokens_seen": 21607392,
      "step": 11220
    },
    {
      "epoch": 1.8312260380128884,
      "grad_norm": 0.0030047642067074776,
      "learning_rate": 0.24539557124418332,
      "loss": 0.2316,
      "num_input_tokens_seen": 21617104,
      "step": 11225
    },
    {
      "epoch": 1.832041765233706,
      "grad_norm": 0.0012048082426190376,
      "learning_rate": 0.24535010621886624,
      "loss": 0.2279,
      "num_input_tokens_seen": 21625824,
      "step": 11230
    },
    {
      "epoch": 1.8328574924545231,
      "grad_norm": 0.002454443136230111,
      "learning_rate": 0.2453046264893646,
      "loss": 0.2508,
      "num_input_tokens_seen": 21636224,
      "step": 11235
    },
    {
      "epoch": 1.8336732196753407,
      "grad_norm": 0.0022536907345056534,
      "learning_rate": 0.24525913206269184,
      "loss": 0.2242,
      "num_input_tokens_seen": 21646432,
      "step": 11240
    },
    {
      "epoch": 1.8344889468961578,
      "grad_norm": 0.0019944547675549984,
      "learning_rate": 0.2452136229458638,
      "loss": 0.2228,
      "num_input_tokens_seen": 21656048,
      "step": 11245
    },
    {
      "epoch": 1.8353046741169754,
      "grad_norm": 0.0023328496608883142,
      "learning_rate": 0.24516809914589857,
      "loss": 0.2249,
      "num_input_tokens_seen": 21665872,
      "step": 11250
    },
    {
      "epoch": 1.8361204013377925,
      "grad_norm": 0.0022140045184642076,
      "learning_rate": 0.2451225606698165,
      "loss": 0.2388,
      "num_input_tokens_seen": 21675424,
      "step": 11255
    },
    {
      "epoch": 1.83693612855861,
      "grad_norm": 0.0015429683262482285,
      "learning_rate": 0.2450770075246402,
      "loss": 0.2366,
      "num_input_tokens_seen": 21685264,
      "step": 11260
    },
    {
      "epoch": 1.8377518557794272,
      "grad_norm": 0.0014026977587491274,
      "learning_rate": 0.24503143971739455,
      "loss": 0.2318,
      "num_input_tokens_seen": 21694384,
      "step": 11265
    },
    {
      "epoch": 1.8385675830002448,
      "grad_norm": 0.0017756834859028459,
      "learning_rate": 0.24498585725510663,
      "loss": 0.2194,
      "num_input_tokens_seen": 21704224,
      "step": 11270
    },
    {
      "epoch": 1.839383310221062,
      "grad_norm": 0.0014164319727569818,
      "learning_rate": 0.24494026014480583,
      "loss": 0.2567,
      "num_input_tokens_seen": 21713440,
      "step": 11275
    },
    {
      "epoch": 1.8401990374418795,
      "grad_norm": 0.0011511839693412185,
      "learning_rate": 0.24489464839352387,
      "loss": 0.2284,
      "num_input_tokens_seen": 21723952,
      "step": 11280
    },
    {
      "epoch": 1.8410147646626966,
      "grad_norm": 0.001755429315380752,
      "learning_rate": 0.2448490220082946,
      "loss": 0.2275,
      "num_input_tokens_seen": 21732128,
      "step": 11285
    },
    {
      "epoch": 1.8418304918835142,
      "grad_norm": 0.0026181587018072605,
      "learning_rate": 0.24480338099615415,
      "loss": 0.2481,
      "num_input_tokens_seen": 21741808,
      "step": 11290
    },
    {
      "epoch": 1.8426462191043314,
      "grad_norm": 0.0007825035718269646,
      "learning_rate": 0.244757725364141,
      "loss": 0.2282,
      "num_input_tokens_seen": 21750448,
      "step": 11295
    },
    {
      "epoch": 1.843461946325149,
      "grad_norm": 0.0010835155844688416,
      "learning_rate": 0.24471205511929583,
      "loss": 0.2062,
      "num_input_tokens_seen": 21760048,
      "step": 11300
    },
    {
      "epoch": 1.8442776735459663,
      "grad_norm": 0.0010074955644086003,
      "learning_rate": 0.24466637026866145,
      "loss": 0.2421,
      "num_input_tokens_seen": 21768592,
      "step": 11305
    },
    {
      "epoch": 1.8450934007667836,
      "grad_norm": 0.0007434137514792383,
      "learning_rate": 0.2446206708192832,
      "loss": 0.2463,
      "num_input_tokens_seen": 21779328,
      "step": 11310
    },
    {
      "epoch": 1.845909127987601,
      "grad_norm": 0.0019097595941275358,
      "learning_rate": 0.2445749567782084,
      "loss": 0.2187,
      "num_input_tokens_seen": 21788464,
      "step": 11315
    },
    {
      "epoch": 1.8467248552084183,
      "grad_norm": 0.0015017444966360927,
      "learning_rate": 0.2445292281524868,
      "loss": 0.2324,
      "num_input_tokens_seen": 21798416,
      "step": 11320
    },
    {
      "epoch": 1.8475405824292357,
      "grad_norm": 0.0013584258267655969,
      "learning_rate": 0.24448348494917022,
      "loss": 0.2133,
      "num_input_tokens_seen": 21808288,
      "step": 11325
    },
    {
      "epoch": 1.848356309650053,
      "grad_norm": 0.0014189545763656497,
      "learning_rate": 0.24443772717531295,
      "loss": 0.2008,
      "num_input_tokens_seen": 21818080,
      "step": 11330
    },
    {
      "epoch": 1.8491720368708704,
      "grad_norm": 0.0014908944722265005,
      "learning_rate": 0.24439195483797138,
      "loss": 0.2342,
      "num_input_tokens_seen": 21826880,
      "step": 11335
    },
    {
      "epoch": 1.8499877640916877,
      "grad_norm": 0.0011467617005109787,
      "learning_rate": 0.24434616794420416,
      "loss": 0.2536,
      "num_input_tokens_seen": 21836688,
      "step": 11340
    },
    {
      "epoch": 1.850803491312505,
      "grad_norm": 0.0028697482775896788,
      "learning_rate": 0.24430036650107223,
      "loss": 0.2268,
      "num_input_tokens_seen": 21845696,
      "step": 11345
    },
    {
      "epoch": 1.8516192185333225,
      "grad_norm": 0.0043199388310313225,
      "learning_rate": 0.2442545505156387,
      "loss": 0.2325,
      "num_input_tokens_seen": 21855648,
      "step": 11350
    },
    {
      "epoch": 1.8524349457541398,
      "grad_norm": 0.00137241929769516,
      "learning_rate": 0.24420871999496904,
      "loss": 0.2273,
      "num_input_tokens_seen": 21865744,
      "step": 11355
    },
    {
      "epoch": 1.8532506729749572,
      "grad_norm": 0.003268870525062084,
      "learning_rate": 0.24416287494613084,
      "loss": 0.2465,
      "num_input_tokens_seen": 21876096,
      "step": 11360
    },
    {
      "epoch": 1.8540664001957745,
      "grad_norm": 0.0015087983338162303,
      "learning_rate": 0.24411701537619399,
      "loss": 0.2115,
      "num_input_tokens_seen": 21886144,
      "step": 11365
    },
    {
      "epoch": 1.8548821274165919,
      "grad_norm": 0.002607950009405613,
      "learning_rate": 0.24407114129223062,
      "loss": 0.2154,
      "num_input_tokens_seen": 21895344,
      "step": 11370
    },
    {
      "epoch": 1.8556978546374092,
      "grad_norm": 0.0017623754683881998,
      "learning_rate": 0.2440252527013151,
      "loss": 0.2439,
      "num_input_tokens_seen": 21904192,
      "step": 11375
    },
    {
      "epoch": 1.8565135818582266,
      "grad_norm": 0.0015626635868102312,
      "learning_rate": 0.24397934961052403,
      "loss": 0.2577,
      "num_input_tokens_seen": 21914080,
      "step": 11380
    },
    {
      "epoch": 1.857329309079044,
      "grad_norm": 0.0017113563371822238,
      "learning_rate": 0.24393343202693618,
      "loss": 0.2195,
      "num_input_tokens_seen": 21922128,
      "step": 11385
    },
    {
      "epoch": 1.8581450362998613,
      "grad_norm": 0.0030662226490676403,
      "learning_rate": 0.2438874999576327,
      "loss": 0.2278,
      "num_input_tokens_seen": 21932208,
      "step": 11390
    },
    {
      "epoch": 1.8589607635206788,
      "grad_norm": 0.0016470493283122778,
      "learning_rate": 0.24384155340969688,
      "loss": 0.2418,
      "num_input_tokens_seen": 21940960,
      "step": 11395
    },
    {
      "epoch": 1.859776490741496,
      "grad_norm": 0.0016990241128951311,
      "learning_rate": 0.24379559239021423,
      "loss": 0.2493,
      "num_input_tokens_seen": 21950896,
      "step": 11400
    },
    {
      "epoch": 1.859776490741496,
      "eval_loss": 0.232594296336174,
      "eval_runtime": 68.0785,
      "eval_samples_per_second": 40.027,
      "eval_steps_per_second": 20.021,
      "num_input_tokens_seen": 21950896,
      "step": 11400
    },
    {
      "epoch": 1.8605922179623136,
      "grad_norm": 0.0020736835431307554,
      "learning_rate": 0.2437496169062725,
      "loss": 0.2405,
      "num_input_tokens_seen": 21960224,
      "step": 11405
    },
    {
      "epoch": 1.8614079451831307,
      "grad_norm": 0.0025427371729165316,
      "learning_rate": 0.24370362696496176,
      "loss": 0.2439,
      "num_input_tokens_seen": 21969552,
      "step": 11410
    },
    {
      "epoch": 1.8622236724039483,
      "grad_norm": 0.0013510604621842504,
      "learning_rate": 0.24365762257337417,
      "loss": 0.2467,
      "num_input_tokens_seen": 21977712,
      "step": 11415
    },
    {
      "epoch": 1.8630393996247654,
      "grad_norm": 0.0022646973375231028,
      "learning_rate": 0.2436116037386042,
      "loss": 0.2264,
      "num_input_tokens_seen": 21985648,
      "step": 11420
    },
    {
      "epoch": 1.863855126845583,
      "grad_norm": 0.0025007096119225025,
      "learning_rate": 0.24356557046774852,
      "loss": 0.2345,
      "num_input_tokens_seen": 21995696,
      "step": 11425
    },
    {
      "epoch": 1.8646708540664,
      "grad_norm": 0.0012452977243810892,
      "learning_rate": 0.24351952276790606,
      "loss": 0.2211,
      "num_input_tokens_seen": 22006240,
      "step": 11430
    },
    {
      "epoch": 1.8654865812872177,
      "grad_norm": 0.0025089257396757603,
      "learning_rate": 0.24347346064617797,
      "loss": 0.2333,
      "num_input_tokens_seen": 22015520,
      "step": 11435
    },
    {
      "epoch": 1.8663023085080348,
      "grad_norm": 0.0010174860944971442,
      "learning_rate": 0.24342738410966758,
      "loss": 0.2144,
      "num_input_tokens_seen": 22024528,
      "step": 11440
    },
    {
      "epoch": 1.8671180357288524,
      "grad_norm": 0.0014731746632605791,
      "learning_rate": 0.24338129316548046,
      "loss": 0.2431,
      "num_input_tokens_seen": 22033696,
      "step": 11445
    },
    {
      "epoch": 1.8679337629496695,
      "grad_norm": 0.001104783616028726,
      "learning_rate": 0.24333518782072444,
      "loss": 0.2399,
      "num_input_tokens_seen": 22043600,
      "step": 11450
    },
    {
      "epoch": 1.868749490170487,
      "grad_norm": 0.0011265702778473496,
      "learning_rate": 0.24328906808250952,
      "loss": 0.2229,
      "num_input_tokens_seen": 22053072,
      "step": 11455
    },
    {
      "epoch": 1.8695652173913042,
      "grad_norm": 0.0015552387339994311,
      "learning_rate": 0.243242933957948,
      "loss": 0.228,
      "num_input_tokens_seen": 22062160,
      "step": 11460
    },
    {
      "epoch": 1.8703809446121218,
      "grad_norm": 0.00299106165766716,
      "learning_rate": 0.24319678545415427,
      "loss": 0.2391,
      "num_input_tokens_seen": 22072000,
      "step": 11465
    },
    {
      "epoch": 1.871196671832939,
      "grad_norm": 0.0025549607817083597,
      "learning_rate": 0.24315062257824507,
      "loss": 0.2168,
      "num_input_tokens_seen": 22081488,
      "step": 11470
    },
    {
      "epoch": 1.8720123990537565,
      "grad_norm": 0.001225952641107142,
      "learning_rate": 0.24310444533733921,
      "loss": 0.2159,
      "num_input_tokens_seen": 22090576,
      "step": 11475
    },
    {
      "epoch": 1.8728281262745736,
      "grad_norm": 0.0025122491642832756,
      "learning_rate": 0.2430582537385579,
      "loss": 0.2101,
      "num_input_tokens_seen": 22098624,
      "step": 11480
    },
    {
      "epoch": 1.8736438534953912,
      "grad_norm": 0.0021821598056703806,
      "learning_rate": 0.2430120477890244,
      "loss": 0.2402,
      "num_input_tokens_seen": 22107776,
      "step": 11485
    },
    {
      "epoch": 1.8744595807162086,
      "grad_norm": 0.001547693507745862,
      "learning_rate": 0.24296582749586426,
      "loss": 0.2119,
      "num_input_tokens_seen": 22116208,
      "step": 11490
    },
    {
      "epoch": 1.875275307937026,
      "grad_norm": 0.001725568319670856,
      "learning_rate": 0.24291959286620526,
      "loss": 0.2566,
      "num_input_tokens_seen": 22126960,
      "step": 11495
    },
    {
      "epoch": 1.8760910351578433,
      "grad_norm": 0.0019001566106453538,
      "learning_rate": 0.24287334390717738,
      "loss": 0.2145,
      "num_input_tokens_seen": 22137792,
      "step": 11500
    },
    {
      "epoch": 1.8769067623786606,
      "grad_norm": 0.002225508913397789,
      "learning_rate": 0.24282708062591268,
      "loss": 0.2158,
      "num_input_tokens_seen": 22147648,
      "step": 11505
    },
    {
      "epoch": 1.877722489599478,
      "grad_norm": 0.0015056295087561011,
      "learning_rate": 0.24278080302954563,
      "loss": 0.2402,
      "num_input_tokens_seen": 22157792,
      "step": 11510
    },
    {
      "epoch": 1.8785382168202953,
      "grad_norm": 0.003960910718888044,
      "learning_rate": 0.24273451112521283,
      "loss": 0.2299,
      "num_input_tokens_seen": 22167424,
      "step": 11515
    },
    {
      "epoch": 1.8793539440411127,
      "grad_norm": 0.002400642726570368,
      "learning_rate": 0.242688204920053,
      "loss": 0.2459,
      "num_input_tokens_seen": 22176368,
      "step": 11520
    },
    {
      "epoch": 1.88016967126193,
      "grad_norm": 0.0031016424763947725,
      "learning_rate": 0.24264188442120715,
      "loss": 0.2326,
      "num_input_tokens_seen": 22185744,
      "step": 11525
    },
    {
      "epoch": 1.8809853984827474,
      "grad_norm": 0.0015337846707552671,
      "learning_rate": 0.24259554963581853,
      "loss": 0.2403,
      "num_input_tokens_seen": 22194704,
      "step": 11530
    },
    {
      "epoch": 1.8818011257035647,
      "grad_norm": 0.0010660815751180053,
      "learning_rate": 0.24254920057103257,
      "loss": 0.2225,
      "num_input_tokens_seen": 22205808,
      "step": 11535
    },
    {
      "epoch": 1.882616852924382,
      "grad_norm": 0.0032112544868141413,
      "learning_rate": 0.24250283723399685,
      "loss": 0.2263,
      "num_input_tokens_seen": 22214880,
      "step": 11540
    },
    {
      "epoch": 1.8834325801451994,
      "grad_norm": 0.00333891692571342,
      "learning_rate": 0.24245645963186108,
      "loss": 0.2269,
      "num_input_tokens_seen": 22222976,
      "step": 11545
    },
    {
      "epoch": 1.8842483073660168,
      "grad_norm": 0.001500188373029232,
      "learning_rate": 0.2424100677717774,
      "loss": 0.2317,
      "num_input_tokens_seen": 22232096,
      "step": 11550
    },
    {
      "epoch": 1.8850640345868341,
      "grad_norm": 0.002449684077873826,
      "learning_rate": 0.24236366166090004,
      "loss": 0.1621,
      "num_input_tokens_seen": 22243168,
      "step": 11555
    },
    {
      "epoch": 1.8858797618076515,
      "grad_norm": 0.00555446045473218,
      "learning_rate": 0.24231724130638527,
      "loss": 0.2696,
      "num_input_tokens_seen": 22251648,
      "step": 11560
    },
    {
      "epoch": 1.8866954890284688,
      "grad_norm": 0.0031435552518814802,
      "learning_rate": 0.2422708067153917,
      "loss": 0.2643,
      "num_input_tokens_seen": 22262208,
      "step": 11565
    },
    {
      "epoch": 1.8875112162492862,
      "grad_norm": 0.0011681071482598782,
      "learning_rate": 0.24222435789508026,
      "loss": 0.2327,
      "num_input_tokens_seen": 22273248,
      "step": 11570
    },
    {
      "epoch": 1.8883269434701035,
      "grad_norm": 0.002346003195270896,
      "learning_rate": 0.24217789485261387,
      "loss": 0.2167,
      "num_input_tokens_seen": 22283392,
      "step": 11575
    },
    {
      "epoch": 1.8891426706909211,
      "grad_norm": 0.0024451136123389006,
      "learning_rate": 0.2421314175951577,
      "loss": 0.2484,
      "num_input_tokens_seen": 22294096,
      "step": 11580
    },
    {
      "epoch": 1.8899583979117383,
      "grad_norm": 0.003457689890637994,
      "learning_rate": 0.2420849261298791,
      "loss": 0.2305,
      "num_input_tokens_seen": 22303424,
      "step": 11585
    },
    {
      "epoch": 1.8907741251325558,
      "grad_norm": 0.004218368791043758,
      "learning_rate": 0.24203842046394775,
      "loss": 0.2376,
      "num_input_tokens_seen": 22313456,
      "step": 11590
    },
    {
      "epoch": 1.891589852353373,
      "grad_norm": 0.0014535775408148766,
      "learning_rate": 0.24199190060453535,
      "loss": 0.2311,
      "num_input_tokens_seen": 22322784,
      "step": 11595
    },
    {
      "epoch": 1.8924055795741905,
      "grad_norm": 0.002577155828475952,
      "learning_rate": 0.2419453665588158,
      "loss": 0.2433,
      "num_input_tokens_seen": 22333376,
      "step": 11600
    },
    {
      "epoch": 1.8924055795741905,
      "eval_loss": 0.22560729086399078,
      "eval_runtime": 68.0311,
      "eval_samples_per_second": 40.055,
      "eval_steps_per_second": 20.035,
      "num_input_tokens_seen": 22333376,
      "step": 11600
    },
    {
      "epoch": 1.8932213067950077,
      "grad_norm": 0.001220877282321453,
      "learning_rate": 0.24189881833396523,
      "loss": 0.2411,
      "num_input_tokens_seen": 22342944,
      "step": 11605
    },
    {
      "epoch": 1.8940370340158252,
      "grad_norm": 0.0023755785077810287,
      "learning_rate": 0.24185225593716203,
      "loss": 0.2211,
      "num_input_tokens_seen": 22352128,
      "step": 11610
    },
    {
      "epoch": 1.8948527612366424,
      "grad_norm": 0.0019313113298267126,
      "learning_rate": 0.2418056793755867,
      "loss": 0.2099,
      "num_input_tokens_seen": 22362688,
      "step": 11615
    },
    {
      "epoch": 1.89566848845746,
      "grad_norm": 0.002169752959161997,
      "learning_rate": 0.24175908865642187,
      "loss": 0.225,
      "num_input_tokens_seen": 22370528,
      "step": 11620
    },
    {
      "epoch": 1.896484215678277,
      "grad_norm": 0.0013578231446444988,
      "learning_rate": 0.24171248378685248,
      "loss": 0.2373,
      "num_input_tokens_seen": 22380784,
      "step": 11625
    },
    {
      "epoch": 1.8972999428990946,
      "grad_norm": 0.0036556238774210215,
      "learning_rate": 0.24166586477406554,
      "loss": 0.252,
      "num_input_tokens_seen": 22391200,
      "step": 11630
    },
    {
      "epoch": 1.8981156701199118,
      "grad_norm": 0.0021956602577120066,
      "learning_rate": 0.24161923162525034,
      "loss": 0.2381,
      "num_input_tokens_seen": 22400288,
      "step": 11635
    },
    {
      "epoch": 1.8989313973407294,
      "grad_norm": 0.0025412393733859062,
      "learning_rate": 0.2415725843475982,
      "loss": 0.2264,
      "num_input_tokens_seen": 22409152,
      "step": 11640
    },
    {
      "epoch": 1.8997471245615465,
      "grad_norm": 0.001190902548842132,
      "learning_rate": 0.24152592294830286,
      "loss": 0.1993,
      "num_input_tokens_seen": 22418880,
      "step": 11645
    },
    {
      "epoch": 1.900562851782364,
      "grad_norm": 0.0010131472954526544,
      "learning_rate": 0.24147924743455995,
      "loss": 0.221,
      "num_input_tokens_seen": 22428048,
      "step": 11650
    },
    {
      "epoch": 1.9013785790031812,
      "grad_norm": 0.003099937690421939,
      "learning_rate": 0.24143255781356754,
      "loss": 0.2117,
      "num_input_tokens_seen": 22436944,
      "step": 11655
    },
    {
      "epoch": 1.9021943062239988,
      "grad_norm": 0.0014572225045412779,
      "learning_rate": 0.24138585409252566,
      "loss": 0.263,
      "num_input_tokens_seen": 22446960,
      "step": 11660
    },
    {
      "epoch": 1.903010033444816,
      "grad_norm": 0.0013428849633783102,
      "learning_rate": 0.24133913627863662,
      "loss": 0.2211,
      "num_input_tokens_seen": 22457728,
      "step": 11665
    },
    {
      "epoch": 1.9038257606656335,
      "grad_norm": 0.0015108260558918118,
      "learning_rate": 0.241292404379105,
      "loss": 0.2405,
      "num_input_tokens_seen": 22467712,
      "step": 11670
    },
    {
      "epoch": 1.9046414878864508,
      "grad_norm": 0.0014021629467606544,
      "learning_rate": 0.24124565840113735,
      "loss": 0.2275,
      "num_input_tokens_seen": 22477824,
      "step": 11675
    },
    {
      "epoch": 1.9054572151072682,
      "grad_norm": 0.0017521734116598964,
      "learning_rate": 0.2411988983519425,
      "loss": 0.2438,
      "num_input_tokens_seen": 22487408,
      "step": 11680
    },
    {
      "epoch": 1.9062729423280855,
      "grad_norm": 0.0018245643004775047,
      "learning_rate": 0.24115212423873145,
      "loss": 0.2423,
      "num_input_tokens_seen": 22496560,
      "step": 11685
    },
    {
      "epoch": 1.9070886695489029,
      "grad_norm": 0.0014239161973819137,
      "learning_rate": 0.24110533606871737,
      "loss": 0.2138,
      "num_input_tokens_seen": 22506608,
      "step": 11690
    },
    {
      "epoch": 1.9079043967697202,
      "grad_norm": 0.0017700030002743006,
      "learning_rate": 0.24105853384911552,
      "loss": 0.2345,
      "num_input_tokens_seen": 22515808,
      "step": 11695
    },
    {
      "epoch": 1.9087201239905376,
      "grad_norm": 0.003785554086789489,
      "learning_rate": 0.24101171758714346,
      "loss": 0.2431,
      "num_input_tokens_seen": 22524544,
      "step": 11700
    },
    {
      "epoch": 1.909535851211355,
      "grad_norm": 0.0030195945873856544,
      "learning_rate": 0.24096488729002086,
      "loss": 0.2169,
      "num_input_tokens_seen": 22534080,
      "step": 11705
    },
    {
      "epoch": 1.9103515784321723,
      "grad_norm": 0.0019349673530086875,
      "learning_rate": 0.24091804296496946,
      "loss": 0.2164,
      "num_input_tokens_seen": 22542800,
      "step": 11710
    },
    {
      "epoch": 1.9111673056529896,
      "grad_norm": 0.0016509207198396325,
      "learning_rate": 0.2408711846192133,
      "loss": 0.2768,
      "num_input_tokens_seen": 22553392,
      "step": 11715
    },
    {
      "epoch": 1.911983032873807,
      "grad_norm": 0.002670396352186799,
      "learning_rate": 0.24082431225997855,
      "loss": 0.1746,
      "num_input_tokens_seen": 22561936,
      "step": 11720
    },
    {
      "epoch": 1.9127987600946244,
      "grad_norm": 0.0022388016805052757,
      "learning_rate": 0.24077742589449344,
      "loss": 0.2275,
      "num_input_tokens_seen": 22571312,
      "step": 11725
    },
    {
      "epoch": 1.9136144873154417,
      "grad_norm": 0.003288719803094864,
      "learning_rate": 0.24073052552998844,
      "loss": 0.2528,
      "num_input_tokens_seen": 22580976,
      "step": 11730
    },
    {
      "epoch": 1.914430214536259,
      "grad_norm": 0.000948488072026521,
      "learning_rate": 0.2406836111736963,
      "loss": 0.2452,
      "num_input_tokens_seen": 22589152,
      "step": 11735
    },
    {
      "epoch": 1.9152459417570764,
      "grad_norm": 0.0009476160048507154,
      "learning_rate": 0.2406366828328517,
      "loss": 0.2222,
      "num_input_tokens_seen": 22597968,
      "step": 11740
    },
    {
      "epoch": 1.9160616689778938,
      "grad_norm": 0.0017010391457006335,
      "learning_rate": 0.2405897405146915,
      "loss": 0.2193,
      "num_input_tokens_seen": 22607760,
      "step": 11745
    },
    {
      "epoch": 1.9168773961987111,
      "grad_norm": 0.0022818802390247583,
      "learning_rate": 0.240542784226455,
      "loss": 0.2401,
      "num_input_tokens_seen": 22616816,
      "step": 11750
    },
    {
      "epoch": 1.9176931234195285,
      "grad_norm": 0.0022421996109187603,
      "learning_rate": 0.24049581397538328,
      "loss": 0.2326,
      "num_input_tokens_seen": 22626192,
      "step": 11755
    },
    {
      "epoch": 1.9185088506403458,
      "grad_norm": 0.0028900280594825745,
      "learning_rate": 0.24044882976871984,
      "loss": 0.2308,
      "num_input_tokens_seen": 22636832,
      "step": 11760
    },
    {
      "epoch": 1.9193245778611632,
      "grad_norm": 0.0011747991666197777,
      "learning_rate": 0.2404018316137102,
      "loss": 0.231,
      "num_input_tokens_seen": 22647776,
      "step": 11765
    },
    {
      "epoch": 1.9201403050819805,
      "grad_norm": 0.0041903904639184475,
      "learning_rate": 0.24035481951760204,
      "loss": 0.2382,
      "num_input_tokens_seen": 22656256,
      "step": 11770
    },
    {
      "epoch": 1.920956032302798,
      "grad_norm": 0.001576438546180725,
      "learning_rate": 0.2403077934876452,
      "loss": 0.2312,
      "num_input_tokens_seen": 22666992,
      "step": 11775
    },
    {
      "epoch": 1.9217717595236152,
      "grad_norm": 0.0008979091071523726,
      "learning_rate": 0.2402607535310918,
      "loss": 0.2239,
      "num_input_tokens_seen": 22674992,
      "step": 11780
    },
    {
      "epoch": 1.9225874867444328,
      "grad_norm": 0.002168158534914255,
      "learning_rate": 0.2402136996551959,
      "loss": 0.2248,
      "num_input_tokens_seen": 22685648,
      "step": 11785
    },
    {
      "epoch": 1.92340321396525,
      "grad_norm": 0.0017524001887068152,
      "learning_rate": 0.24016663186721376,
      "loss": 0.229,
      "num_input_tokens_seen": 22694304,
      "step": 11790
    },
    {
      "epoch": 1.9242189411860675,
      "grad_norm": 0.001449136296287179,
      "learning_rate": 0.24011955017440395,
      "loss": 0.2289,
      "num_input_tokens_seen": 22703808,
      "step": 11795
    },
    {
      "epoch": 1.9250346684068846,
      "grad_norm": 0.003584136487916112,
      "learning_rate": 0.24007245458402696,
      "loss": 0.2541,
      "num_input_tokens_seen": 22714512,
      "step": 11800
    },
    {
      "epoch": 1.9250346684068846,
      "eval_loss": 0.23513619601726532,
      "eval_runtime": 68.0186,
      "eval_samples_per_second": 40.063,
      "eval_steps_per_second": 20.039,
      "num_input_tokens_seen": 22714512,
      "step": 11800
    },
    {
      "epoch": 1.9258503956277022,
      "grad_norm": 0.0015418896218761802,
      "learning_rate": 0.2400253451033456,
      "loss": 0.2387,
      "num_input_tokens_seen": 22724432,
      "step": 11805
    },
    {
      "epoch": 1.9266661228485193,
      "grad_norm": 0.0012764851562678814,
      "learning_rate": 0.23997822173962463,
      "loss": 0.231,
      "num_input_tokens_seen": 22734736,
      "step": 11810
    },
    {
      "epoch": 1.927481850069337,
      "grad_norm": 0.0020429182332009077,
      "learning_rate": 0.23993108450013118,
      "loss": 0.2599,
      "num_input_tokens_seen": 22745440,
      "step": 11815
    },
    {
      "epoch": 1.928297577290154,
      "grad_norm": 0.001665174844674766,
      "learning_rate": 0.2398839333921343,
      "loss": 0.2223,
      "num_input_tokens_seen": 22753568,
      "step": 11820
    },
    {
      "epoch": 1.9291133045109716,
      "grad_norm": 0.0014074748614802957,
      "learning_rate": 0.23983676842290536,
      "loss": 0.2208,
      "num_input_tokens_seen": 22763680,
      "step": 11825
    },
    {
      "epoch": 1.9299290317317888,
      "grad_norm": 0.0030272426083683968,
      "learning_rate": 0.2397895895997178,
      "loss": 0.1878,
      "num_input_tokens_seen": 22773264,
      "step": 11830
    },
    {
      "epoch": 1.9307447589526063,
      "grad_norm": 0.003598329843953252,
      "learning_rate": 0.23974239692984714,
      "loss": 0.235,
      "num_input_tokens_seen": 22782784,
      "step": 11835
    },
    {
      "epoch": 1.9315604861734235,
      "grad_norm": 0.001331635401584208,
      "learning_rate": 0.2396951904205711,
      "loss": 0.1772,
      "num_input_tokens_seen": 22792064,
      "step": 11840
    },
    {
      "epoch": 1.932376213394241,
      "grad_norm": 0.0035005013924092054,
      "learning_rate": 0.23964797007916952,
      "loss": 0.2586,
      "num_input_tokens_seen": 22800560,
      "step": 11845
    },
    {
      "epoch": 1.9331919406150582,
      "grad_norm": 0.0024463750887662172,
      "learning_rate": 0.23960073591292436,
      "loss": 0.1758,
      "num_input_tokens_seen": 22811728,
      "step": 11850
    },
    {
      "epoch": 1.9340076678358757,
      "grad_norm": 0.0015488811768591404,
      "learning_rate": 0.2395534879291197,
      "loss": 0.2406,
      "num_input_tokens_seen": 22819904,
      "step": 11855
    },
    {
      "epoch": 1.934823395056693,
      "grad_norm": 0.002576378406956792,
      "learning_rate": 0.23950622613504186,
      "loss": 0.24,
      "num_input_tokens_seen": 22828592,
      "step": 11860
    },
    {
      "epoch": 1.9356391222775104,
      "grad_norm": 0.002958419034257531,
      "learning_rate": 0.2394589505379791,
      "loss": 0.2578,
      "num_input_tokens_seen": 22838464,
      "step": 11865
    },
    {
      "epoch": 1.9364548494983278,
      "grad_norm": 0.0024813751224428415,
      "learning_rate": 0.23941166114522197,
      "loss": 0.2386,
      "num_input_tokens_seen": 22847488,
      "step": 11870
    },
    {
      "epoch": 1.9372705767191452,
      "grad_norm": 0.0021476184483617544,
      "learning_rate": 0.23936435796406308,
      "loss": 0.2169,
      "num_input_tokens_seen": 22856384,
      "step": 11875
    },
    {
      "epoch": 1.9380863039399625,
      "grad_norm": 0.002901161555200815,
      "learning_rate": 0.23931704100179715,
      "loss": 0.2232,
      "num_input_tokens_seen": 22866224,
      "step": 11880
    },
    {
      "epoch": 1.9389020311607799,
      "grad_norm": 0.003072948195040226,
      "learning_rate": 0.2392697102657211,
      "loss": 0.207,
      "num_input_tokens_seen": 22876512,
      "step": 11885
    },
    {
      "epoch": 1.9397177583815972,
      "grad_norm": 0.0016710566123947501,
      "learning_rate": 0.23922236576313388,
      "loss": 0.2327,
      "num_input_tokens_seen": 22886800,
      "step": 11890
    },
    {
      "epoch": 1.9405334856024146,
      "grad_norm": 0.0017894185148179531,
      "learning_rate": 0.2391750075013366,
      "loss": 0.2581,
      "num_input_tokens_seen": 22895728,
      "step": 11895
    },
    {
      "epoch": 1.941349212823232,
      "grad_norm": 0.0015365860890597105,
      "learning_rate": 0.2391276354876326,
      "loss": 0.2001,
      "num_input_tokens_seen": 22905424,
      "step": 11900
    },
    {
      "epoch": 1.9421649400440493,
      "grad_norm": 0.0017360267229378223,
      "learning_rate": 0.23908024972932707,
      "loss": 0.2212,
      "num_input_tokens_seen": 22914016,
      "step": 11905
    },
    {
      "epoch": 1.9429806672648666,
      "grad_norm": 0.0017105786828324199,
      "learning_rate": 0.2390328502337276,
      "loss": 0.2335,
      "num_input_tokens_seen": 22924448,
      "step": 11910
    },
    {
      "epoch": 1.943796394485684,
      "grad_norm": 0.0011032808106392622,
      "learning_rate": 0.23898543700814376,
      "loss": 0.2292,
      "num_input_tokens_seen": 22933152,
      "step": 11915
    },
    {
      "epoch": 1.9446121217065013,
      "grad_norm": 0.0010060120839625597,
      "learning_rate": 0.2389380100598873,
      "loss": 0.2461,
      "num_input_tokens_seen": 22942112,
      "step": 11920
    },
    {
      "epoch": 1.9454278489273187,
      "grad_norm": 0.004239773843437433,
      "learning_rate": 0.23889056939627207,
      "loss": 0.2349,
      "num_input_tokens_seen": 22951360,
      "step": 11925
    },
    {
      "epoch": 1.946243576148136,
      "grad_norm": 0.0024050993379205465,
      "learning_rate": 0.23884311502461386,
      "loss": 0.2341,
      "num_input_tokens_seen": 22961824,
      "step": 11930
    },
    {
      "epoch": 1.9470593033689534,
      "grad_norm": 0.0023421444930136204,
      "learning_rate": 0.23879564695223088,
      "loss": 0.2541,
      "num_input_tokens_seen": 22970784,
      "step": 11935
    },
    {
      "epoch": 1.9478750305897707,
      "grad_norm": 0.002105381805449724,
      "learning_rate": 0.23874816518644332,
      "loss": 0.2377,
      "num_input_tokens_seen": 22980960,
      "step": 11940
    },
    {
      "epoch": 1.948690757810588,
      "grad_norm": 0.0028763916343450546,
      "learning_rate": 0.23870066973457335,
      "loss": 0.236,
      "num_input_tokens_seen": 22991136,
      "step": 11945
    },
    {
      "epoch": 1.9495064850314054,
      "grad_norm": 0.0019062466453760862,
      "learning_rate": 0.23865316060394545,
      "loss": 0.2392,
      "num_input_tokens_seen": 23000896,
      "step": 11950
    },
    {
      "epoch": 1.9503222122522228,
      "grad_norm": 0.0017750976840034127,
      "learning_rate": 0.2386056378018861,
      "loss": 0.2186,
      "num_input_tokens_seen": 23010528,
      "step": 11955
    },
    {
      "epoch": 1.9511379394730404,
      "grad_norm": 0.0013412637636065483,
      "learning_rate": 0.2385581013357239,
      "loss": 0.2264,
      "num_input_tokens_seen": 23018800,
      "step": 11960
    },
    {
      "epoch": 1.9519536666938575,
      "grad_norm": 0.0031687540467828512,
      "learning_rate": 0.23851055121278958,
      "loss": 0.2256,
      "num_input_tokens_seen": 23029392,
      "step": 11965
    },
    {
      "epoch": 1.952769393914675,
      "grad_norm": 0.0025338660925626755,
      "learning_rate": 0.23846298744041594,
      "loss": 0.2345,
      "num_input_tokens_seen": 23040336,
      "step": 11970
    },
    {
      "epoch": 1.9535851211354922,
      "grad_norm": 0.0029049438890069723,
      "learning_rate": 0.23841541002593802,
      "loss": 0.2303,
      "num_input_tokens_seen": 23050672,
      "step": 11975
    },
    {
      "epoch": 1.9544008483563098,
      "grad_norm": 0.002982059260830283,
      "learning_rate": 0.23836781897669276,
      "loss": 0.2507,
      "num_input_tokens_seen": 23059984,
      "step": 11980
    },
    {
      "epoch": 1.955216575577127,
      "grad_norm": 0.0013212596531957388,
      "learning_rate": 0.23832021430001926,
      "loss": 0.2403,
      "num_input_tokens_seen": 23069296,
      "step": 11985
    },
    {
      "epoch": 1.9560323027979445,
      "grad_norm": 0.0012953534023836255,
      "learning_rate": 0.2382725960032588,
      "loss": 0.2174,
      "num_input_tokens_seen": 23079568,
      "step": 11990
    },
    {
      "epoch": 1.9568480300187616,
      "grad_norm": 0.00306443078443408,
      "learning_rate": 0.23822496409375482,
      "loss": 0.2156,
      "num_input_tokens_seen": 23089536,
      "step": 11995
    },
    {
      "epoch": 1.9576637572395792,
      "grad_norm": 0.001444953610189259,
      "learning_rate": 0.2381773185788526,
      "loss": 0.2216,
      "num_input_tokens_seen": 23099888,
      "step": 12000
    },
    {
      "epoch": 1.9576637572395792,
      "eval_loss": 0.22554148733615875,
      "eval_runtime": 68.0108,
      "eval_samples_per_second": 40.067,
      "eval_steps_per_second": 20.041,
      "num_input_tokens_seen": 23099888,
      "step": 12000
    },
    {
      "epoch": 1.9584794844603963,
      "grad_norm": 0.0024417657405138016,
      "learning_rate": 0.2381296594658998,
      "loss": 0.1762,
      "num_input_tokens_seen": 23108880,
      "step": 12005
    },
    {
      "epoch": 1.959295211681214,
      "grad_norm": 0.002066638320684433,
      "learning_rate": 0.238081986762246,
      "loss": 0.3282,
      "num_input_tokens_seen": 23117376,
      "step": 12010
    },
    {
      "epoch": 1.960110938902031,
      "grad_norm": 0.0024164002388715744,
      "learning_rate": 0.23803430047524293,
      "loss": 0.2486,
      "num_input_tokens_seen": 23126624,
      "step": 12015
    },
    {
      "epoch": 1.9609266661228486,
      "grad_norm": 0.0015372349880635738,
      "learning_rate": 0.23798660061224441,
      "loss": 0.2282,
      "num_input_tokens_seen": 23135600,
      "step": 12020
    },
    {
      "epoch": 1.9617423933436657,
      "grad_norm": 0.0019268353935331106,
      "learning_rate": 0.23793888718060632,
      "loss": 0.2366,
      "num_input_tokens_seen": 23145824,
      "step": 12025
    },
    {
      "epoch": 1.9625581205644833,
      "grad_norm": 0.0018445674795657396,
      "learning_rate": 0.23789116018768675,
      "loss": 0.2225,
      "num_input_tokens_seen": 23155808,
      "step": 12030
    },
    {
      "epoch": 1.9633738477853004,
      "grad_norm": 0.0014877973590046167,
      "learning_rate": 0.2378434196408458,
      "loss": 0.2048,
      "num_input_tokens_seen": 23166720,
      "step": 12035
    },
    {
      "epoch": 1.964189575006118,
      "grad_norm": 0.00276266410946846,
      "learning_rate": 0.23779566554744563,
      "loss": 0.2375,
      "num_input_tokens_seen": 23176672,
      "step": 12040
    },
    {
      "epoch": 1.9650053022269351,
      "grad_norm": 0.0017138294642791152,
      "learning_rate": 0.23774789791485051,
      "loss": 0.2521,
      "num_input_tokens_seen": 23187248,
      "step": 12045
    },
    {
      "epoch": 1.9658210294477527,
      "grad_norm": 0.0024649319238960743,
      "learning_rate": 0.2377001167504268,
      "loss": 0.2351,
      "num_input_tokens_seen": 23196416,
      "step": 12050
    },
    {
      "epoch": 1.96663675666857,
      "grad_norm": 0.0012589477701112628,
      "learning_rate": 0.23765232206154302,
      "loss": 0.2251,
      "num_input_tokens_seen": 23204736,
      "step": 12055
    },
    {
      "epoch": 1.9674524838893874,
      "grad_norm": 0.003337721573188901,
      "learning_rate": 0.23760451385556966,
      "loss": 0.2313,
      "num_input_tokens_seen": 23215360,
      "step": 12060
    },
    {
      "epoch": 1.9682682111102048,
      "grad_norm": 0.0018955300329253078,
      "learning_rate": 0.23755669213987932,
      "loss": 0.204,
      "num_input_tokens_seen": 23224848,
      "step": 12065
    },
    {
      "epoch": 1.9690839383310221,
      "grad_norm": 0.0025169721338897943,
      "learning_rate": 0.23750885692184676,
      "loss": 0.1954,
      "num_input_tokens_seen": 23234848,
      "step": 12070
    },
    {
      "epoch": 1.9698996655518395,
      "grad_norm": 0.0015395267400890589,
      "learning_rate": 0.23746100820884875,
      "loss": 0.2277,
      "num_input_tokens_seen": 23244800,
      "step": 12075
    },
    {
      "epoch": 1.9707153927726568,
      "grad_norm": 0.0020567697938531637,
      "learning_rate": 0.23741314600826421,
      "loss": 0.2402,
      "num_input_tokens_seen": 23254336,
      "step": 12080
    },
    {
      "epoch": 1.9715311199934742,
      "grad_norm": 0.0011488941963762045,
      "learning_rate": 0.23736527032747406,
      "loss": 0.2073,
      "num_input_tokens_seen": 23264000,
      "step": 12085
    },
    {
      "epoch": 1.9723468472142915,
      "grad_norm": 0.001944348099641502,
      "learning_rate": 0.23731738117386128,
      "loss": 0.2298,
      "num_input_tokens_seen": 23273456,
      "step": 12090
    },
    {
      "epoch": 1.973162574435109,
      "grad_norm": 0.0027489778585731983,
      "learning_rate": 0.237269478554811,
      "loss": 0.2248,
      "num_input_tokens_seen": 23282624,
      "step": 12095
    },
    {
      "epoch": 1.9739783016559262,
      "grad_norm": 0.0016771730734035373,
      "learning_rate": 0.23722156247771053,
      "loss": 0.2551,
      "num_input_tokens_seen": 23292640,
      "step": 12100
    },
    {
      "epoch": 1.9747940288767436,
      "grad_norm": 0.003015250898897648,
      "learning_rate": 0.23717363294994895,
      "loss": 0.2228,
      "num_input_tokens_seen": 23301168,
      "step": 12105
    },
    {
      "epoch": 1.975609756097561,
      "grad_norm": 0.004784972872585058,
      "learning_rate": 0.2371256899789177,
      "loss": 0.2458,
      "num_input_tokens_seen": 23311648,
      "step": 12110
    },
    {
      "epoch": 1.9764254833183783,
      "grad_norm": 0.0023100473918020725,
      "learning_rate": 0.23707773357201017,
      "loss": 0.2782,
      "num_input_tokens_seen": 23321408,
      "step": 12115
    },
    {
      "epoch": 1.9772412105391957,
      "grad_norm": 0.0009223211673088372,
      "learning_rate": 0.2370297637366218,
      "loss": 0.2099,
      "num_input_tokens_seen": 23330320,
      "step": 12120
    },
    {
      "epoch": 1.978056937760013,
      "grad_norm": 0.003284366335719824,
      "learning_rate": 0.23698178048015026,
      "loss": 0.2716,
      "num_input_tokens_seen": 23340160,
      "step": 12125
    },
    {
      "epoch": 1.9788726649808304,
      "grad_norm": 0.0015566333895549178,
      "learning_rate": 0.236933783809995,
      "loss": 0.2014,
      "num_input_tokens_seen": 23349776,
      "step": 12130
    },
    {
      "epoch": 1.9796883922016477,
      "grad_norm": 0.001741049694828689,
      "learning_rate": 0.23688577373355785,
      "loss": 0.2011,
      "num_input_tokens_seen": 23360144,
      "step": 12135
    },
    {
      "epoch": 1.980504119422465,
      "grad_norm": 0.003189387731254101,
      "learning_rate": 0.23683775025824247,
      "loss": 0.2394,
      "num_input_tokens_seen": 23370128,
      "step": 12140
    },
    {
      "epoch": 1.9813198466432826,
      "grad_norm": 0.004083799198269844,
      "learning_rate": 0.2367897133914548,
      "loss": 0.2496,
      "num_input_tokens_seen": 23379584,
      "step": 12145
    },
    {
      "epoch": 1.9821355738640998,
      "grad_norm": 0.0014079594984650612,
      "learning_rate": 0.2367416631406026,
      "loss": 0.2072,
      "num_input_tokens_seen": 23388032,
      "step": 12150
    },
    {
      "epoch": 1.9829513010849174,
      "grad_norm": 0.003952261060476303,
      "learning_rate": 0.23669359951309588,
      "loss": 0.2743,
      "num_input_tokens_seen": 23396960,
      "step": 12155
    },
    {
      "epoch": 1.9837670283057345,
      "grad_norm": 0.0030602486804127693,
      "learning_rate": 0.23664552251634666,
      "loss": 0.2327,
      "num_input_tokens_seen": 23406272,
      "step": 12160
    },
    {
      "epoch": 1.984582755526552,
      "grad_norm": 0.003714726772159338,
      "learning_rate": 0.23659743215776907,
      "loss": 0.2437,
      "num_input_tokens_seen": 23416512,
      "step": 12165
    },
    {
      "epoch": 1.9853984827473692,
      "grad_norm": 0.001069550053216517,
      "learning_rate": 0.23654932844477908,
      "loss": 0.2311,
      "num_input_tokens_seen": 23425632,
      "step": 12170
    },
    {
      "epoch": 1.9862142099681868,
      "grad_norm": 0.0009289812878705561,
      "learning_rate": 0.23650121138479507,
      "loss": 0.2219,
      "num_input_tokens_seen": 23435504,
      "step": 12175
    },
    {
      "epoch": 1.987029937189004,
      "grad_norm": 0.0008036977960728109,
      "learning_rate": 0.23645308098523724,
      "loss": 0.2613,
      "num_input_tokens_seen": 23444336,
      "step": 12180
    },
    {
      "epoch": 1.9878456644098215,
      "grad_norm": 0.005185524467378855,
      "learning_rate": 0.23640493725352785,
      "loss": 0.2265,
      "num_input_tokens_seen": 23454256,
      "step": 12185
    },
    {
      "epoch": 1.9886613916306386,
      "grad_norm": 0.0012464143801480532,
      "learning_rate": 0.2363567801970913,
      "loss": 0.2413,
      "num_input_tokens_seen": 23463072,
      "step": 12190
    },
    {
      "epoch": 1.9894771188514562,
      "grad_norm": 0.0017739901086315513,
      "learning_rate": 0.236308609823354,
      "loss": 0.2431,
      "num_input_tokens_seen": 23472816,
      "step": 12195
    },
    {
      "epoch": 1.9902928460722733,
      "grad_norm": 0.0007180466200225055,
      "learning_rate": 0.23626042613974452,
      "loss": 0.2311,
      "num_input_tokens_seen": 23482400,
      "step": 12200
    },
    {
      "epoch": 1.9902928460722733,
      "eval_loss": 0.2382442206144333,
      "eval_runtime": 68.1034,
      "eval_samples_per_second": 40.013,
      "eval_steps_per_second": 20.014,
      "num_input_tokens_seen": 23482400,
      "step": 12200
    },
    {
      "epoch": 1.9911085732930909,
      "grad_norm": 0.001997825223952532,
      "learning_rate": 0.23621222915369325,
      "loss": 0.2386,
      "num_input_tokens_seen": 23491392,
      "step": 12205
    },
    {
      "epoch": 1.991924300513908,
      "grad_norm": 0.0022937434259802103,
      "learning_rate": 0.23616401887263283,
      "loss": 0.2264,
      "num_input_tokens_seen": 23501344,
      "step": 12210
    },
    {
      "epoch": 1.9927400277347256,
      "grad_norm": 0.002404562197625637,
      "learning_rate": 0.23611579530399793,
      "loss": 0.2321,
      "num_input_tokens_seen": 23510304,
      "step": 12215
    },
    {
      "epoch": 1.9935557549555427,
      "grad_norm": 0.002812295453622937,
      "learning_rate": 0.23606755845522517,
      "loss": 0.2321,
      "num_input_tokens_seen": 23520224,
      "step": 12220
    },
    {
      "epoch": 1.9943714821763603,
      "grad_norm": 0.0015628342516720295,
      "learning_rate": 0.23601930833375329,
      "loss": 0.2202,
      "num_input_tokens_seen": 23528816,
      "step": 12225
    },
    {
      "epoch": 1.9951872093971774,
      "grad_norm": 0.0013563948450610042,
      "learning_rate": 0.23597104494702312,
      "loss": 0.1973,
      "num_input_tokens_seen": 23538176,
      "step": 12230
    },
    {
      "epoch": 1.996002936617995,
      "grad_norm": 0.0030481384601444006,
      "learning_rate": 0.23592276830247744,
      "loss": 0.2623,
      "num_input_tokens_seen": 23546608,
      "step": 12235
    },
    {
      "epoch": 1.9968186638388123,
      "grad_norm": 0.003587523242458701,
      "learning_rate": 0.2358744784075611,
      "loss": 0.225,
      "num_input_tokens_seen": 23555664,
      "step": 12240
    },
    {
      "epoch": 1.9976343910596297,
      "grad_norm": 0.0013145245611667633,
      "learning_rate": 0.235826175269721,
      "loss": 0.2217,
      "num_input_tokens_seen": 23564912,
      "step": 12245
    },
    {
      "epoch": 1.998450118280447,
      "grad_norm": 0.0027680990751832724,
      "learning_rate": 0.23577785889640612,
      "loss": 0.2345,
      "num_input_tokens_seen": 23574976,
      "step": 12250
    },
    {
      "epoch": 1.9992658455012644,
      "grad_norm": 0.0014738230966031551,
      "learning_rate": 0.23572952929506744,
      "loss": 0.2142,
      "num_input_tokens_seen": 23585520,
      "step": 12255
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.004099647980183363,
      "learning_rate": 0.23568118647315803,
      "loss": 0.2667,
      "num_input_tokens_seen": 23594288,
      "step": 12260
    },
    {
      "epoch": 2.0008157272208176,
      "grad_norm": 0.0019347359193488955,
      "learning_rate": 0.23563283043813296,
      "loss": 0.2436,
      "num_input_tokens_seen": 23603248,
      "step": 12265
    },
    {
      "epoch": 2.0016314544416347,
      "grad_norm": 0.0014630977530032396,
      "learning_rate": 0.23558446119744922,
      "loss": 0.2555,
      "num_input_tokens_seen": 23611920,
      "step": 12270
    },
    {
      "epoch": 2.0024471816624523,
      "grad_norm": 0.0016480337362736464,
      "learning_rate": 0.23553607875856608,
      "loss": 0.2259,
      "num_input_tokens_seen": 23622272,
      "step": 12275
    },
    {
      "epoch": 2.0032629088832694,
      "grad_norm": 0.003517549019306898,
      "learning_rate": 0.2354876831289447,
      "loss": 0.248,
      "num_input_tokens_seen": 23631104,
      "step": 12280
    },
    {
      "epoch": 2.004078636104087,
      "grad_norm": 0.003350980347022414,
      "learning_rate": 0.23543927431604827,
      "loss": 0.2395,
      "num_input_tokens_seen": 23639328,
      "step": 12285
    },
    {
      "epoch": 2.004894363324904,
      "grad_norm": 0.0029589396435767412,
      "learning_rate": 0.23539085232734203,
      "loss": 0.2325,
      "num_input_tokens_seen": 23648416,
      "step": 12290
    },
    {
      "epoch": 2.0057100905457217,
      "grad_norm": 0.001511454931460321,
      "learning_rate": 0.2353424171702933,
      "loss": 0.2271,
      "num_input_tokens_seen": 23657184,
      "step": 12295
    },
    {
      "epoch": 2.006525817766539,
      "grad_norm": 0.0024767317809164524,
      "learning_rate": 0.23529396885237133,
      "loss": 0.2246,
      "num_input_tokens_seen": 23667312,
      "step": 12300
    },
    {
      "epoch": 2.0073415449873564,
      "grad_norm": 0.0021680735517293215,
      "learning_rate": 0.2352455073810475,
      "loss": 0.2258,
      "num_input_tokens_seen": 23677344,
      "step": 12305
    },
    {
      "epoch": 2.0081572722081735,
      "grad_norm": 0.0010388389928266406,
      "learning_rate": 0.23519703276379517,
      "loss": 0.2344,
      "num_input_tokens_seen": 23687856,
      "step": 12310
    },
    {
      "epoch": 2.008972999428991,
      "grad_norm": 0.0010070002172142267,
      "learning_rate": 0.2351485450080897,
      "loss": 0.2324,
      "num_input_tokens_seen": 23696432,
      "step": 12315
    },
    {
      "epoch": 2.0097887266498082,
      "grad_norm": 0.00140933517832309,
      "learning_rate": 0.2351000441214086,
      "loss": 0.2404,
      "num_input_tokens_seen": 23706560,
      "step": 12320
    },
    {
      "epoch": 2.010604453870626,
      "grad_norm": 0.0014014929765835404,
      "learning_rate": 0.23505153011123125,
      "loss": 0.2232,
      "num_input_tokens_seen": 23716144,
      "step": 12325
    },
    {
      "epoch": 2.011420181091443,
      "grad_norm": 0.0021885582245886326,
      "learning_rate": 0.23500300298503912,
      "loss": 0.2224,
      "num_input_tokens_seen": 23725856,
      "step": 12330
    },
    {
      "epoch": 2.0122359083122605,
      "grad_norm": 0.004394575487822294,
      "learning_rate": 0.23495446275031576,
      "loss": 0.2391,
      "num_input_tokens_seen": 23734240,
      "step": 12335
    },
    {
      "epoch": 2.0130516355330776,
      "grad_norm": 0.0018834165530279279,
      "learning_rate": 0.2349059094145466,
      "loss": 0.2317,
      "num_input_tokens_seen": 23744208,
      "step": 12340
    },
    {
      "epoch": 2.013867362753895,
      "grad_norm": 0.0009803208522498608,
      "learning_rate": 0.2348573429852192,
      "loss": 0.2115,
      "num_input_tokens_seen": 23754432,
      "step": 12345
    },
    {
      "epoch": 2.0146830899747123,
      "grad_norm": 0.001108212280087173,
      "learning_rate": 0.23480876346982313,
      "loss": 0.2127,
      "num_input_tokens_seen": 23764064,
      "step": 12350
    },
    {
      "epoch": 2.01549881719553,
      "grad_norm": 0.0028019060846418142,
      "learning_rate": 0.23476017087585,
      "loss": 0.2387,
      "num_input_tokens_seen": 23774144,
      "step": 12355
    },
    {
      "epoch": 2.016314544416347,
      "grad_norm": 0.0019674994982779026,
      "learning_rate": 0.23471156521079334,
      "loss": 0.2327,
      "num_input_tokens_seen": 23782624,
      "step": 12360
    },
    {
      "epoch": 2.0171302716371646,
      "grad_norm": 0.0018100133165717125,
      "learning_rate": 0.23466294648214875,
      "loss": 0.2213,
      "num_input_tokens_seen": 23792896,
      "step": 12365
    },
    {
      "epoch": 2.0179459988579818,
      "grad_norm": 0.0019439085153862834,
      "learning_rate": 0.2346143146974139,
      "loss": 0.2362,
      "num_input_tokens_seen": 23802864,
      "step": 12370
    },
    {
      "epoch": 2.0187617260787993,
      "grad_norm": 0.0016624316340312362,
      "learning_rate": 0.23456566986408836,
      "loss": 0.2492,
      "num_input_tokens_seen": 23810784,
      "step": 12375
    },
    {
      "epoch": 2.0195774532996165,
      "grad_norm": 0.003725034184753895,
      "learning_rate": 0.23451701198967384,
      "loss": 0.2359,
      "num_input_tokens_seen": 23820096,
      "step": 12380
    },
    {
      "epoch": 2.020393180520434,
      "grad_norm": 0.0011587865883484483,
      "learning_rate": 0.23446834108167397,
      "loss": 0.2293,
      "num_input_tokens_seen": 23830272,
      "step": 12385
    },
    {
      "epoch": 2.021208907741251,
      "grad_norm": 0.0024005172308534384,
      "learning_rate": 0.23441965714759438,
      "loss": 0.2643,
      "num_input_tokens_seen": 23839536,
      "step": 12390
    },
    {
      "epoch": 2.0220246349620687,
      "grad_norm": 0.0032516701612621546,
      "learning_rate": 0.23437096019494277,
      "loss": 0.227,
      "num_input_tokens_seen": 23849392,
      "step": 12395
    },
    {
      "epoch": 2.022840362182886,
      "grad_norm": 0.0019446242367848754,
      "learning_rate": 0.23432225023122885,
      "loss": 0.2379,
      "num_input_tokens_seen": 23860160,
      "step": 12400
    },
    {
      "epoch": 2.022840362182886,
      "eval_loss": 0.2257762998342514,
      "eval_runtime": 67.9595,
      "eval_samples_per_second": 40.097,
      "eval_steps_per_second": 20.056,
      "num_input_tokens_seen": 23860160,
      "step": 12400
    },
    {
      "epoch": 2.0236560894037035,
      "grad_norm": 0.0031195199117064476,
      "learning_rate": 0.23427352726396428,
      "loss": 0.2201,
      "num_input_tokens_seen": 23869568,
      "step": 12405
    },
    {
      "epoch": 2.0244718166245206,
      "grad_norm": 0.0026415856555104256,
      "learning_rate": 0.2342247913006628,
      "loss": 0.2799,
      "num_input_tokens_seen": 23880016,
      "step": 12410
    },
    {
      "epoch": 2.025287543845338,
      "grad_norm": 0.0022471281699836254,
      "learning_rate": 0.23417604234883999,
      "loss": 0.2421,
      "num_input_tokens_seen": 23889312,
      "step": 12415
    },
    {
      "epoch": 2.0261032710661553,
      "grad_norm": 0.0013664615107700229,
      "learning_rate": 0.23412728041601363,
      "loss": 0.2163,
      "num_input_tokens_seen": 23899536,
      "step": 12420
    },
    {
      "epoch": 2.026918998286973,
      "grad_norm": 0.0008136804099194705,
      "learning_rate": 0.23407850550970347,
      "loss": 0.2586,
      "num_input_tokens_seen": 23908304,
      "step": 12425
    },
    {
      "epoch": 2.02773472550779,
      "grad_norm": 0.0017334066797047853,
      "learning_rate": 0.23402971763743116,
      "loss": 0.2225,
      "num_input_tokens_seen": 23918992,
      "step": 12430
    },
    {
      "epoch": 2.0285504527286076,
      "grad_norm": 0.0010294633684679866,
      "learning_rate": 0.23398091680672037,
      "loss": 0.2126,
      "num_input_tokens_seen": 23927136,
      "step": 12435
    },
    {
      "epoch": 2.0293661799494247,
      "grad_norm": 0.002445748308673501,
      "learning_rate": 0.23393210302509687,
      "loss": 0.2254,
      "num_input_tokens_seen": 23937616,
      "step": 12440
    },
    {
      "epoch": 2.0301819071702423,
      "grad_norm": 0.0034633520990610123,
      "learning_rate": 0.23388327630008832,
      "loss": 0.2118,
      "num_input_tokens_seen": 23947584,
      "step": 12445
    },
    {
      "epoch": 2.03099763439106,
      "grad_norm": 0.0021060705184936523,
      "learning_rate": 0.23383443663922443,
      "loss": 0.1897,
      "num_input_tokens_seen": 23957632,
      "step": 12450
    },
    {
      "epoch": 2.031813361611877,
      "grad_norm": 0.002386168809607625,
      "learning_rate": 0.23378558405003685,
      "loss": 0.2824,
      "num_input_tokens_seen": 23967440,
      "step": 12455
    },
    {
      "epoch": 2.0326290888326946,
      "grad_norm": 0.003127741627395153,
      "learning_rate": 0.2337367185400593,
      "loss": 0.2045,
      "num_input_tokens_seen": 23976928,
      "step": 12460
    },
    {
      "epoch": 2.0334448160535117,
      "grad_norm": 0.003309275023639202,
      "learning_rate": 0.23368784011682747,
      "loss": 0.2473,
      "num_input_tokens_seen": 23987728,
      "step": 12465
    },
    {
      "epoch": 2.0342605432743293,
      "grad_norm": 0.002485667821019888,
      "learning_rate": 0.23363894878787902,
      "loss": 0.246,
      "num_input_tokens_seen": 23997568,
      "step": 12470
    },
    {
      "epoch": 2.0350762704951464,
      "grad_norm": 0.0010670217452570796,
      "learning_rate": 0.23359004456075352,
      "loss": 0.229,
      "num_input_tokens_seen": 24007568,
      "step": 12475
    },
    {
      "epoch": 2.035891997715964,
      "grad_norm": 0.0014752805000171065,
      "learning_rate": 0.23354112744299277,
      "loss": 0.2257,
      "num_input_tokens_seen": 24018032,
      "step": 12480
    },
    {
      "epoch": 2.036707724936781,
      "grad_norm": 0.002261033281683922,
      "learning_rate": 0.2334921974421403,
      "loss": 0.2228,
      "num_input_tokens_seen": 24027168,
      "step": 12485
    },
    {
      "epoch": 2.0375234521575987,
      "grad_norm": 0.002973154652863741,
      "learning_rate": 0.23344325456574178,
      "loss": 0.2603,
      "num_input_tokens_seen": 24037296,
      "step": 12490
    },
    {
      "epoch": 2.038339179378416,
      "grad_norm": 0.0019073649309575558,
      "learning_rate": 0.23339429882134477,
      "loss": 0.2414,
      "num_input_tokens_seen": 24046112,
      "step": 12495
    },
    {
      "epoch": 2.0391549065992334,
      "grad_norm": 0.0019185408018529415,
      "learning_rate": 0.23334533021649884,
      "loss": 0.2519,
      "num_input_tokens_seen": 24054768,
      "step": 12500
    },
    {
      "epoch": 2.0399706338200505,
      "grad_norm": 0.0016149911098182201,
      "learning_rate": 0.23329634875875566,
      "loss": 0.2207,
      "num_input_tokens_seen": 24065520,
      "step": 12505
    },
    {
      "epoch": 2.040786361040868,
      "grad_norm": 0.0029047310817986727,
      "learning_rate": 0.23324735445566874,
      "loss": 0.2296,
      "num_input_tokens_seen": 24076112,
      "step": 12510
    },
    {
      "epoch": 2.041602088261685,
      "grad_norm": 0.00347155868075788,
      "learning_rate": 0.2331983473147936,
      "loss": 0.2573,
      "num_input_tokens_seen": 24085824,
      "step": 12515
    },
    {
      "epoch": 2.042417815482503,
      "grad_norm": 0.0014259968884289265,
      "learning_rate": 0.23314932734368776,
      "loss": 0.2138,
      "num_input_tokens_seen": 24096304,
      "step": 12520
    },
    {
      "epoch": 2.04323354270332,
      "grad_norm": 0.0025584532413631678,
      "learning_rate": 0.2331002945499107,
      "loss": 0.2648,
      "num_input_tokens_seen": 24106192,
      "step": 12525
    },
    {
      "epoch": 2.0440492699241375,
      "grad_norm": 0.0017128167673945427,
      "learning_rate": 0.23305124894102397,
      "loss": 0.2407,
      "num_input_tokens_seen": 24116160,
      "step": 12530
    },
    {
      "epoch": 2.0448649971449546,
      "grad_norm": 0.0008927627932280302,
      "learning_rate": 0.23300219052459092,
      "loss": 0.2326,
      "num_input_tokens_seen": 24125408,
      "step": 12535
    },
    {
      "epoch": 2.045680724365772,
      "grad_norm": 0.0024587635416537523,
      "learning_rate": 0.23295311930817708,
      "loss": 0.2339,
      "num_input_tokens_seen": 24135216,
      "step": 12540
    },
    {
      "epoch": 2.0464964515865893,
      "grad_norm": 0.0019373614341020584,
      "learning_rate": 0.23290403529934972,
      "loss": 0.2357,
      "num_input_tokens_seen": 24144640,
      "step": 12545
    },
    {
      "epoch": 2.047312178807407,
      "grad_norm": 0.0012720882659778,
      "learning_rate": 0.23285493850567832,
      "loss": 0.234,
      "num_input_tokens_seen": 24153792,
      "step": 12550
    },
    {
      "epoch": 2.048127906028224,
      "grad_norm": 0.0023318789899349213,
      "learning_rate": 0.23280582893473414,
      "loss": 0.2359,
      "num_input_tokens_seen": 24162624,
      "step": 12555
    },
    {
      "epoch": 2.0489436332490416,
      "grad_norm": 0.003536105854436755,
      "learning_rate": 0.2327567065940906,
      "loss": 0.2304,
      "num_input_tokens_seen": 24171856,
      "step": 12560
    },
    {
      "epoch": 2.0497593604698587,
      "grad_norm": 0.0012392994249239564,
      "learning_rate": 0.23270757149132285,
      "loss": 0.2238,
      "num_input_tokens_seen": 24181888,
      "step": 12565
    },
    {
      "epoch": 2.0505750876906763,
      "grad_norm": 0.0028680639807134867,
      "learning_rate": 0.23265842363400827,
      "loss": 0.2628,
      "num_input_tokens_seen": 24191600,
      "step": 12570
    },
    {
      "epoch": 2.0513908149114934,
      "grad_norm": 0.004346473142504692,
      "learning_rate": 0.23260926302972595,
      "loss": 0.2263,
      "num_input_tokens_seen": 24200816,
      "step": 12575
    },
    {
      "epoch": 2.052206542132311,
      "grad_norm": 0.0034413307439535856,
      "learning_rate": 0.2325600896860572,
      "loss": 0.2501,
      "num_input_tokens_seen": 24211344,
      "step": 12580
    },
    {
      "epoch": 2.053022269353128,
      "grad_norm": 0.0017798093613237143,
      "learning_rate": 0.23251090361058505,
      "loss": 0.2353,
      "num_input_tokens_seen": 24221136,
      "step": 12585
    },
    {
      "epoch": 2.0538379965739457,
      "grad_norm": 0.004185046069324017,
      "learning_rate": 0.23246170481089476,
      "loss": 0.2598,
      "num_input_tokens_seen": 24229616,
      "step": 12590
    },
    {
      "epoch": 2.054653723794763,
      "grad_norm": 0.0014254432171583176,
      "learning_rate": 0.23241249329457317,
      "loss": 0.2294,
      "num_input_tokens_seen": 24239232,
      "step": 12595
    },
    {
      "epoch": 2.0554694510155804,
      "grad_norm": 0.0011339513584971428,
      "learning_rate": 0.23236326906920957,
      "loss": 0.239,
      "num_input_tokens_seen": 24249008,
      "step": 12600
    },
    {
      "epoch": 2.0554694510155804,
      "eval_loss": 0.22530090808868408,
      "eval_runtime": 68.0089,
      "eval_samples_per_second": 40.068,
      "eval_steps_per_second": 20.042,
      "num_input_tokens_seen": 24249008,
      "step": 12600
    },
    {
      "epoch": 2.0562851782363976,
      "grad_norm": 0.0020651929080486298,
      "learning_rate": 0.2323140321423948,
      "loss": 0.2308,
      "num_input_tokens_seen": 24258432,
      "step": 12605
    },
    {
      "epoch": 2.057100905457215,
      "grad_norm": 0.0020841415971517563,
      "learning_rate": 0.23226478252172184,
      "loss": 0.2272,
      "num_input_tokens_seen": 24268128,
      "step": 12610
    },
    {
      "epoch": 2.0579166326780323,
      "grad_norm": 0.001941184513270855,
      "learning_rate": 0.23221552021478561,
      "loss": 0.242,
      "num_input_tokens_seen": 24278496,
      "step": 12615
    },
    {
      "epoch": 2.05873235989885,
      "grad_norm": 0.0012719297083094716,
      "learning_rate": 0.232166245229183,
      "loss": 0.245,
      "num_input_tokens_seen": 24287792,
      "step": 12620
    },
    {
      "epoch": 2.059548087119667,
      "grad_norm": 0.0015722170937806368,
      "learning_rate": 0.2321169575725128,
      "loss": 0.233,
      "num_input_tokens_seen": 24295856,
      "step": 12625
    },
    {
      "epoch": 2.0603638143404845,
      "grad_norm": 0.0012563263298943639,
      "learning_rate": 0.23206765725237577,
      "loss": 0.2164,
      "num_input_tokens_seen": 24306528,
      "step": 12630
    },
    {
      "epoch": 2.0611795415613017,
      "grad_norm": 0.0028177944477647543,
      "learning_rate": 0.2320183442763747,
      "loss": 0.2368,
      "num_input_tokens_seen": 24315760,
      "step": 12635
    },
    {
      "epoch": 2.0619952687821193,
      "grad_norm": 0.001583671197295189,
      "learning_rate": 0.23196901865211422,
      "loss": 0.1986,
      "num_input_tokens_seen": 24326480,
      "step": 12640
    },
    {
      "epoch": 2.062810996002937,
      "grad_norm": 0.0018380386754870415,
      "learning_rate": 0.231919680387201,
      "loss": 0.2311,
      "num_input_tokens_seen": 24336864,
      "step": 12645
    },
    {
      "epoch": 2.063626723223754,
      "grad_norm": 0.002251619240269065,
      "learning_rate": 0.23187032948924358,
      "loss": 0.1968,
      "num_input_tokens_seen": 24346736,
      "step": 12650
    },
    {
      "epoch": 2.0644424504445715,
      "grad_norm": 0.002093652030453086,
      "learning_rate": 0.23182096596585247,
      "loss": 0.222,
      "num_input_tokens_seen": 24356000,
      "step": 12655
    },
    {
      "epoch": 2.0652581776653887,
      "grad_norm": 0.002049647271633148,
      "learning_rate": 0.23177158982464025,
      "loss": 0.244,
      "num_input_tokens_seen": 24366384,
      "step": 12660
    },
    {
      "epoch": 2.0660739048862062,
      "grad_norm": 0.00204466818831861,
      "learning_rate": 0.23172220107322122,
      "loss": 0.2587,
      "num_input_tokens_seen": 24377440,
      "step": 12665
    },
    {
      "epoch": 2.0668896321070234,
      "grad_norm": 0.003437793580815196,
      "learning_rate": 0.23167279971921184,
      "loss": 0.2214,
      "num_input_tokens_seen": 24386832,
      "step": 12670
    },
    {
      "epoch": 2.067705359327841,
      "grad_norm": 0.0015625765081495047,
      "learning_rate": 0.23162338577023034,
      "loss": 0.2336,
      "num_input_tokens_seen": 24395328,
      "step": 12675
    },
    {
      "epoch": 2.068521086548658,
      "grad_norm": 0.0015825764276087284,
      "learning_rate": 0.23157395923389704,
      "loss": 0.2268,
      "num_input_tokens_seen": 24403824,
      "step": 12680
    },
    {
      "epoch": 2.0693368137694756,
      "grad_norm": 0.0014759479090571404,
      "learning_rate": 0.2315245201178341,
      "loss": 0.2204,
      "num_input_tokens_seen": 24414448,
      "step": 12685
    },
    {
      "epoch": 2.0701525409902928,
      "grad_norm": 0.0019170455634593964,
      "learning_rate": 0.23147506842966564,
      "loss": 0.2322,
      "num_input_tokens_seen": 24424128,
      "step": 12690
    },
    {
      "epoch": 2.0709682682111104,
      "grad_norm": 0.0013799956068396568,
      "learning_rate": 0.23142560417701774,
      "loss": 0.2617,
      "num_input_tokens_seen": 24434096,
      "step": 12695
    },
    {
      "epoch": 2.0717839954319275,
      "grad_norm": 0.0034341467544436455,
      "learning_rate": 0.23137612736751845,
      "loss": 0.2146,
      "num_input_tokens_seen": 24444256,
      "step": 12700
    },
    {
      "epoch": 2.072599722652745,
      "grad_norm": 0.0017515160143375397,
      "learning_rate": 0.23132663800879766,
      "loss": 0.2318,
      "num_input_tokens_seen": 24454496,
      "step": 12705
    },
    {
      "epoch": 2.073415449873562,
      "grad_norm": 0.0015074950642883778,
      "learning_rate": 0.2312771361084873,
      "loss": 0.2225,
      "num_input_tokens_seen": 24464432,
      "step": 12710
    },
    {
      "epoch": 2.0742311770943798,
      "grad_norm": 0.0019631339237093925,
      "learning_rate": 0.23122762167422112,
      "loss": 0.2357,
      "num_input_tokens_seen": 24474816,
      "step": 12715
    },
    {
      "epoch": 2.075046904315197,
      "grad_norm": 0.0022719488479197025,
      "learning_rate": 0.23117809471363493,
      "loss": 0.2037,
      "num_input_tokens_seen": 24484432,
      "step": 12720
    },
    {
      "epoch": 2.0758626315360145,
      "grad_norm": 0.0008002606336958706,
      "learning_rate": 0.23112855523436637,
      "loss": 0.2099,
      "num_input_tokens_seen": 24493488,
      "step": 12725
    },
    {
      "epoch": 2.0766783587568316,
      "grad_norm": 0.0042917970567941666,
      "learning_rate": 0.23107900324405511,
      "loss": 0.2629,
      "num_input_tokens_seen": 24504208,
      "step": 12730
    },
    {
      "epoch": 2.077494085977649,
      "grad_norm": 0.001548335305415094,
      "learning_rate": 0.2310294387503426,
      "loss": 0.2065,
      "num_input_tokens_seen": 24513280,
      "step": 12735
    },
    {
      "epoch": 2.0783098131984663,
      "grad_norm": 0.000904514396097511,
      "learning_rate": 0.23097986176087237,
      "loss": 0.229,
      "num_input_tokens_seen": 24521920,
      "step": 12740
    },
    {
      "epoch": 2.079125540419284,
      "grad_norm": 0.0019318206468597054,
      "learning_rate": 0.23093027228328986,
      "loss": 0.22,
      "num_input_tokens_seen": 24532272,
      "step": 12745
    },
    {
      "epoch": 2.079941267640101,
      "grad_norm": 0.0020989556796848774,
      "learning_rate": 0.23088067032524226,
      "loss": 0.2151,
      "num_input_tokens_seen": 24543312,
      "step": 12750
    },
    {
      "epoch": 2.0807569948609186,
      "grad_norm": 0.0028238396625965834,
      "learning_rate": 0.23083105589437888,
      "loss": 0.2631,
      "num_input_tokens_seen": 24552816,
      "step": 12755
    },
    {
      "epoch": 2.0815727220817357,
      "grad_norm": 0.0016391811659559608,
      "learning_rate": 0.23078142899835094,
      "loss": 0.2341,
      "num_input_tokens_seen": 24562800,
      "step": 12760
    },
    {
      "epoch": 2.0823884493025533,
      "grad_norm": 0.002276328159496188,
      "learning_rate": 0.23073178964481147,
      "loss": 0.2199,
      "num_input_tokens_seen": 24572208,
      "step": 12765
    },
    {
      "epoch": 2.0832041765233704,
      "grad_norm": 0.0036441287957131863,
      "learning_rate": 0.2306821378414155,
      "loss": 0.2352,
      "num_input_tokens_seen": 24581536,
      "step": 12770
    },
    {
      "epoch": 2.084019903744188,
      "grad_norm": 0.0011009074514731765,
      "learning_rate": 0.2306324735958199,
      "loss": 0.2339,
      "num_input_tokens_seen": 24591232,
      "step": 12775
    },
    {
      "epoch": 2.084835630965005,
      "grad_norm": 0.0021505975164473057,
      "learning_rate": 0.23058279691568362,
      "loss": 0.2336,
      "num_input_tokens_seen": 24601152,
      "step": 12780
    },
    {
      "epoch": 2.0856513581858227,
      "grad_norm": 0.001541624660603702,
      "learning_rate": 0.23053310780866745,
      "loss": 0.247,
      "num_input_tokens_seen": 24610944,
      "step": 12785
    },
    {
      "epoch": 2.08646708540664,
      "grad_norm": 0.0008987695910036564,
      "learning_rate": 0.23048340628243397,
      "loss": 0.2293,
      "num_input_tokens_seen": 24620992,
      "step": 12790
    },
    {
      "epoch": 2.0872828126274574,
      "grad_norm": 0.0013202183181419969,
      "learning_rate": 0.23043369234464783,
      "loss": 0.1936,
      "num_input_tokens_seen": 24630864,
      "step": 12795
    },
    {
      "epoch": 2.0880985398482745,
      "grad_norm": 0.003055373439565301,
      "learning_rate": 0.2303839660029755,
      "loss": 0.2476,
      "num_input_tokens_seen": 24639552,
      "step": 12800
    },
    {
      "epoch": 2.0880985398482745,
      "eval_loss": 0.22860901057720184,
      "eval_runtime": 67.9743,
      "eval_samples_per_second": 40.089,
      "eval_steps_per_second": 20.052,
      "num_input_tokens_seen": 24639552,
      "step": 12800
    },
    {
      "epoch": 2.088914267069092,
      "grad_norm": 0.001993517391383648,
      "learning_rate": 0.23033422726508548,
      "loss": 0.2012,
      "num_input_tokens_seen": 24648400,
      "step": 12805
    },
    {
      "epoch": 2.0897299942899092,
      "grad_norm": 0.00228211865760386,
      "learning_rate": 0.23028447613864808,
      "loss": 0.2532,
      "num_input_tokens_seen": 24658336,
      "step": 12810
    },
    {
      "epoch": 2.090545721510727,
      "grad_norm": 0.002799722831696272,
      "learning_rate": 0.2302347126313355,
      "loss": 0.2396,
      "num_input_tokens_seen": 24667792,
      "step": 12815
    },
    {
      "epoch": 2.0913614487315444,
      "grad_norm": 0.0016514919698238373,
      "learning_rate": 0.23018493675082197,
      "loss": 0.2554,
      "num_input_tokens_seen": 24676208,
      "step": 12820
    },
    {
      "epoch": 2.0921771759523615,
      "grad_norm": 0.0010864317882806063,
      "learning_rate": 0.2301351485047835,
      "loss": 0.2459,
      "num_input_tokens_seen": 24685728,
      "step": 12825
    },
    {
      "epoch": 2.092992903173179,
      "grad_norm": 0.0016845714999362826,
      "learning_rate": 0.23008534790089813,
      "loss": 0.1958,
      "num_input_tokens_seen": 24696528,
      "step": 12830
    },
    {
      "epoch": 2.0938086303939962,
      "grad_norm": 0.0018492341041564941,
      "learning_rate": 0.2300355349468457,
      "loss": 0.2304,
      "num_input_tokens_seen": 24705808,
      "step": 12835
    },
    {
      "epoch": 2.094624357614814,
      "grad_norm": 0.0012050956720486283,
      "learning_rate": 0.22998570965030793,
      "loss": 0.226,
      "num_input_tokens_seen": 24715280,
      "step": 12840
    },
    {
      "epoch": 2.095440084835631,
      "grad_norm": 0.0030506763141602278,
      "learning_rate": 0.22993587201896862,
      "loss": 0.2034,
      "num_input_tokens_seen": 24723360,
      "step": 12845
    },
    {
      "epoch": 2.0962558120564485,
      "grad_norm": 0.003040684387087822,
      "learning_rate": 0.2298860220605133,
      "loss": 0.2708,
      "num_input_tokens_seen": 24733072,
      "step": 12850
    },
    {
      "epoch": 2.0970715392772656,
      "grad_norm": 0.0014133666409179568,
      "learning_rate": 0.22983615978262942,
      "loss": 0.2054,
      "num_input_tokens_seen": 24742896,
      "step": 12855
    },
    {
      "epoch": 2.097887266498083,
      "grad_norm": 0.001740772626362741,
      "learning_rate": 0.22978628519300648,
      "loss": 0.2031,
      "num_input_tokens_seen": 24752864,
      "step": 12860
    },
    {
      "epoch": 2.0987029937189003,
      "grad_norm": 0.0026852271985262632,
      "learning_rate": 0.22973639829933568,
      "loss": 0.2214,
      "num_input_tokens_seen": 24761392,
      "step": 12865
    },
    {
      "epoch": 2.099518720939718,
      "grad_norm": 0.0015369850443676114,
      "learning_rate": 0.22968649910931027,
      "loss": 0.2137,
      "num_input_tokens_seen": 24772000,
      "step": 12870
    },
    {
      "epoch": 2.100334448160535,
      "grad_norm": 0.001145477988757193,
      "learning_rate": 0.22963658763062528,
      "loss": 0.2223,
      "num_input_tokens_seen": 24782064,
      "step": 12875
    },
    {
      "epoch": 2.1011501753813526,
      "grad_norm": 0.0020407806150615215,
      "learning_rate": 0.22958666387097765,
      "loss": 0.2581,
      "num_input_tokens_seen": 24791968,
      "step": 12880
    },
    {
      "epoch": 2.1019659026021698,
      "grad_norm": 0.0013030635891482234,
      "learning_rate": 0.22953672783806633,
      "loss": 0.2548,
      "num_input_tokens_seen": 24802576,
      "step": 12885
    },
    {
      "epoch": 2.1027816298229873,
      "grad_norm": 0.0008591444347985089,
      "learning_rate": 0.22948677953959207,
      "loss": 0.2208,
      "num_input_tokens_seen": 24812336,
      "step": 12890
    },
    {
      "epoch": 2.1035973570438045,
      "grad_norm": 0.0020706881769001484,
      "learning_rate": 0.2294368189832575,
      "loss": 0.2303,
      "num_input_tokens_seen": 24822304,
      "step": 12895
    },
    {
      "epoch": 2.104413084264622,
      "grad_norm": 0.0010361216263845563,
      "learning_rate": 0.2293868461767672,
      "loss": 0.2061,
      "num_input_tokens_seen": 24832048,
      "step": 12900
    },
    {
      "epoch": 2.105228811485439,
      "grad_norm": 0.0015324046835303307,
      "learning_rate": 0.22933686112782758,
      "loss": 0.2667,
      "num_input_tokens_seen": 24842272,
      "step": 12905
    },
    {
      "epoch": 2.1060445387062567,
      "grad_norm": 0.0014008466387167573,
      "learning_rate": 0.22928686384414698,
      "loss": 0.2069,
      "num_input_tokens_seen": 24851408,
      "step": 12910
    },
    {
      "epoch": 2.106860265927074,
      "grad_norm": 0.002389807952567935,
      "learning_rate": 0.22923685433343552,
      "loss": 0.2088,
      "num_input_tokens_seen": 24862192,
      "step": 12915
    },
    {
      "epoch": 2.1076759931478914,
      "grad_norm": 0.0023017541971057653,
      "learning_rate": 0.22918683260340542,
      "loss": 0.2328,
      "num_input_tokens_seen": 24872704,
      "step": 12920
    },
    {
      "epoch": 2.1084917203687086,
      "grad_norm": 0.002325371140614152,
      "learning_rate": 0.2291367986617706,
      "loss": 0.2217,
      "num_input_tokens_seen": 24881232,
      "step": 12925
    },
    {
      "epoch": 2.109307447589526,
      "grad_norm": 0.005180733744055033,
      "learning_rate": 0.22908675251624697,
      "loss": 0.2641,
      "num_input_tokens_seen": 24891872,
      "step": 12930
    },
    {
      "epoch": 2.1101231748103433,
      "grad_norm": 0.0024748253636062145,
      "learning_rate": 0.22903669417455216,
      "loss": 0.2374,
      "num_input_tokens_seen": 24901840,
      "step": 12935
    },
    {
      "epoch": 2.110938902031161,
      "grad_norm": 0.0018351373728364706,
      "learning_rate": 0.22898662364440592,
      "loss": 0.1759,
      "num_input_tokens_seen": 24911120,
      "step": 12940
    },
    {
      "epoch": 2.111754629251978,
      "grad_norm": 0.0011569896014407277,
      "learning_rate": 0.2289365409335297,
      "loss": 0.2419,
      "num_input_tokens_seen": 24920784,
      "step": 12945
    },
    {
      "epoch": 2.1125703564727956,
      "grad_norm": 0.0023015791084617376,
      "learning_rate": 0.2288864460496469,
      "loss": 0.2357,
      "num_input_tokens_seen": 24930208,
      "step": 12950
    },
    {
      "epoch": 2.1133860836936127,
      "grad_norm": 0.0018736561760306358,
      "learning_rate": 0.22883633900048272,
      "loss": 0.2369,
      "num_input_tokens_seen": 24939600,
      "step": 12955
    },
    {
      "epoch": 2.1142018109144303,
      "grad_norm": 0.0015533051919192076,
      "learning_rate": 0.2287862197937644,
      "loss": 0.2385,
      "num_input_tokens_seen": 24947840,
      "step": 12960
    },
    {
      "epoch": 2.1150175381352474,
      "grad_norm": 0.0022777733393013477,
      "learning_rate": 0.2287360884372209,
      "loss": 0.2381,
      "num_input_tokens_seen": 24958032,
      "step": 12965
    },
    {
      "epoch": 2.115833265356065,
      "grad_norm": 0.002461405238136649,
      "learning_rate": 0.22868594493858307,
      "loss": 0.2364,
      "num_input_tokens_seen": 24966976,
      "step": 12970
    },
    {
      "epoch": 2.116648992576882,
      "grad_norm": 0.0020159010309726,
      "learning_rate": 0.2286357893055837,
      "loss": 0.2486,
      "num_input_tokens_seen": 24978288,
      "step": 12975
    },
    {
      "epoch": 2.1174647197976997,
      "grad_norm": 0.0012503315228968859,
      "learning_rate": 0.22858562154595746,
      "loss": 0.213,
      "num_input_tokens_seen": 24987616,
      "step": 12980
    },
    {
      "epoch": 2.118280447018517,
      "grad_norm": 0.0018800280522555113,
      "learning_rate": 0.22853544166744078,
      "loss": 0.2337,
      "num_input_tokens_seen": 24997824,
      "step": 12985
    },
    {
      "epoch": 2.1190961742393344,
      "grad_norm": 0.000802580441813916,
      "learning_rate": 0.22848524967777206,
      "loss": 0.2423,
      "num_input_tokens_seen": 25007712,
      "step": 12990
    },
    {
      "epoch": 2.1199119014601515,
      "grad_norm": 0.0011889211600646377,
      "learning_rate": 0.22843504558469152,
      "loss": 0.2295,
      "num_input_tokens_seen": 25017024,
      "step": 12995
    },
    {
      "epoch": 2.120727628680969,
      "grad_norm": 0.0018398710526525974,
      "learning_rate": 0.2283848293959413,
      "loss": 0.2214,
      "num_input_tokens_seen": 25026880,
      "step": 13000
    },
    {
      "epoch": 2.120727628680969,
      "eval_loss": 0.22608456015586853,
      "eval_runtime": 67.9243,
      "eval_samples_per_second": 40.118,
      "eval_steps_per_second": 20.066,
      "num_input_tokens_seen": 25026880,
      "step": 13000
    },
    {
      "epoch": 2.121543355901786,
      "grad_norm": 0.002536576706916094,
      "learning_rate": 0.22833460111926532,
      "loss": 0.2078,
      "num_input_tokens_seen": 25037408,
      "step": 13005
    },
    {
      "epoch": 2.122359083122604,
      "grad_norm": 0.001987575087696314,
      "learning_rate": 0.22828436076240946,
      "loss": 0.232,
      "num_input_tokens_seen": 25045760,
      "step": 13010
    },
    {
      "epoch": 2.1231748103434214,
      "grad_norm": 0.0014186057960614562,
      "learning_rate": 0.22823410833312135,
      "loss": 0.266,
      "num_input_tokens_seen": 25054896,
      "step": 13015
    },
    {
      "epoch": 2.1239905375642385,
      "grad_norm": 0.0014702529879286885,
      "learning_rate": 0.2281838438391506,
      "loss": 0.2119,
      "num_input_tokens_seen": 25064768,
      "step": 13020
    },
    {
      "epoch": 2.124806264785056,
      "grad_norm": 0.0023473664186894894,
      "learning_rate": 0.22813356728824863,
      "loss": 0.2365,
      "num_input_tokens_seen": 25075152,
      "step": 13025
    },
    {
      "epoch": 2.125621992005873,
      "grad_norm": 0.0021571223624050617,
      "learning_rate": 0.2280832786881687,
      "loss": 0.2368,
      "num_input_tokens_seen": 25085120,
      "step": 13030
    },
    {
      "epoch": 2.126437719226691,
      "grad_norm": 0.0019226871663704515,
      "learning_rate": 0.22803297804666592,
      "loss": 0.2347,
      "num_input_tokens_seen": 25095312,
      "step": 13035
    },
    {
      "epoch": 2.127253446447508,
      "grad_norm": 0.0017563545843586326,
      "learning_rate": 0.22798266537149728,
      "loss": 0.2337,
      "num_input_tokens_seen": 25104160,
      "step": 13040
    },
    {
      "epoch": 2.1280691736683255,
      "grad_norm": 0.0013521051732823253,
      "learning_rate": 0.22793234067042167,
      "loss": 0.2445,
      "num_input_tokens_seen": 25113104,
      "step": 13045
    },
    {
      "epoch": 2.1288849008891426,
      "grad_norm": 0.002413281938061118,
      "learning_rate": 0.22788200395119979,
      "loss": 0.2225,
      "num_input_tokens_seen": 25123760,
      "step": 13050
    },
    {
      "epoch": 2.12970062810996,
      "grad_norm": 0.0009241647785529494,
      "learning_rate": 0.2278316552215942,
      "loss": 0.2299,
      "num_input_tokens_seen": 25133168,
      "step": 13055
    },
    {
      "epoch": 2.1305163553307773,
      "grad_norm": 0.0013185287825763226,
      "learning_rate": 0.22778129448936918,
      "loss": 0.2374,
      "num_input_tokens_seen": 25142208,
      "step": 13060
    },
    {
      "epoch": 2.131332082551595,
      "grad_norm": 0.0015108030056580901,
      "learning_rate": 0.22773092176229118,
      "loss": 0.2451,
      "num_input_tokens_seen": 25151616,
      "step": 13065
    },
    {
      "epoch": 2.132147809772412,
      "grad_norm": 0.0018512659007683396,
      "learning_rate": 0.22768053704812816,
      "loss": 0.2302,
      "num_input_tokens_seen": 25161536,
      "step": 13070
    },
    {
      "epoch": 2.1329635369932296,
      "grad_norm": 0.0011393072782084346,
      "learning_rate": 0.22763014035465018,
      "loss": 0.2392,
      "num_input_tokens_seen": 25172032,
      "step": 13075
    },
    {
      "epoch": 2.1337792642140467,
      "grad_norm": 0.0013474065344780684,
      "learning_rate": 0.22757973168962892,
      "loss": 0.2286,
      "num_input_tokens_seen": 25180400,
      "step": 13080
    },
    {
      "epoch": 2.1345949914348643,
      "grad_norm": 0.0008260190370492637,
      "learning_rate": 0.22752931106083818,
      "loss": 0.2185,
      "num_input_tokens_seen": 25189232,
      "step": 13085
    },
    {
      "epoch": 2.1354107186556814,
      "grad_norm": 0.0012178616598248482,
      "learning_rate": 0.22747887847605341,
      "loss": 0.2311,
      "num_input_tokens_seen": 25199584,
      "step": 13090
    },
    {
      "epoch": 2.136226445876499,
      "grad_norm": 0.0016865157522261143,
      "learning_rate": 0.22742843394305184,
      "loss": 0.2083,
      "num_input_tokens_seen": 25208288,
      "step": 13095
    },
    {
      "epoch": 2.137042173097316,
      "grad_norm": 0.0035619379486888647,
      "learning_rate": 0.22737797746961272,
      "loss": 0.2217,
      "num_input_tokens_seen": 25218576,
      "step": 13100
    },
    {
      "epoch": 2.1378579003181337,
      "grad_norm": 0.0015885685570538044,
      "learning_rate": 0.22732750906351712,
      "loss": 0.2087,
      "num_input_tokens_seen": 25228032,
      "step": 13105
    },
    {
      "epoch": 2.138673627538951,
      "grad_norm": 0.0024604997597634792,
      "learning_rate": 0.22727702873254785,
      "loss": 0.2338,
      "num_input_tokens_seen": 25239392,
      "step": 13110
    },
    {
      "epoch": 2.1394893547597684,
      "grad_norm": 0.0030636214651167393,
      "learning_rate": 0.22722653648448968,
      "loss": 0.2185,
      "num_input_tokens_seen": 25248656,
      "step": 13115
    },
    {
      "epoch": 2.1403050819805856,
      "grad_norm": 0.0015207502292469144,
      "learning_rate": 0.22717603232712902,
      "loss": 0.2346,
      "num_input_tokens_seen": 25257360,
      "step": 13120
    },
    {
      "epoch": 2.141120809201403,
      "grad_norm": 0.0008285005460493267,
      "learning_rate": 0.22712551626825436,
      "loss": 0.2191,
      "num_input_tokens_seen": 25266576,
      "step": 13125
    },
    {
      "epoch": 2.1419365364222203,
      "grad_norm": 0.0016861147014424205,
      "learning_rate": 0.2270749883156559,
      "loss": 0.2227,
      "num_input_tokens_seen": 25277456,
      "step": 13130
    },
    {
      "epoch": 2.142752263643038,
      "grad_norm": 0.0032102796249091625,
      "learning_rate": 0.22702444847712563,
      "loss": 0.2375,
      "num_input_tokens_seen": 25286160,
      "step": 13135
    },
    {
      "epoch": 2.143567990863855,
      "grad_norm": 0.0010196494404226542,
      "learning_rate": 0.22697389676045743,
      "loss": 0.2128,
      "num_input_tokens_seen": 25295040,
      "step": 13140
    },
    {
      "epoch": 2.1443837180846725,
      "grad_norm": 0.0038343367632478476,
      "learning_rate": 0.22692333317344704,
      "loss": 0.2549,
      "num_input_tokens_seen": 25304656,
      "step": 13145
    },
    {
      "epoch": 2.1451994453054897,
      "grad_norm": 0.003332689171656966,
      "learning_rate": 0.22687275772389198,
      "loss": 0.2259,
      "num_input_tokens_seen": 25313616,
      "step": 13150
    },
    {
      "epoch": 2.1460151725263072,
      "grad_norm": 0.004251500591635704,
      "learning_rate": 0.22682217041959168,
      "loss": 0.2554,
      "num_input_tokens_seen": 25323664,
      "step": 13155
    },
    {
      "epoch": 2.1468308997471244,
      "grad_norm": 0.0009706949349492788,
      "learning_rate": 0.2267715712683473,
      "loss": 0.2182,
      "num_input_tokens_seen": 25332864,
      "step": 13160
    },
    {
      "epoch": 2.147646626967942,
      "grad_norm": 0.000979984411969781,
      "learning_rate": 0.22672096027796182,
      "loss": 0.2634,
      "num_input_tokens_seen": 25342512,
      "step": 13165
    },
    {
      "epoch": 2.148462354188759,
      "grad_norm": 0.0016402057372033596,
      "learning_rate": 0.22667033745624016,
      "loss": 0.2129,
      "num_input_tokens_seen": 25351296,
      "step": 13170
    },
    {
      "epoch": 2.1492780814095767,
      "grad_norm": 0.0036404223646968603,
      "learning_rate": 0.22661970281098895,
      "loss": 0.2309,
      "num_input_tokens_seen": 25360592,
      "step": 13175
    },
    {
      "epoch": 2.150093808630394,
      "grad_norm": 0.002903146203607321,
      "learning_rate": 0.22656905635001667,
      "loss": 0.2305,
      "num_input_tokens_seen": 25370096,
      "step": 13180
    },
    {
      "epoch": 2.1509095358512114,
      "grad_norm": 0.0032560713589191437,
      "learning_rate": 0.2265183980811337,
      "loss": 0.254,
      "num_input_tokens_seen": 25381392,
      "step": 13185
    },
    {
      "epoch": 2.151725263072029,
      "grad_norm": 0.0018241209909319878,
      "learning_rate": 0.22646772801215218,
      "loss": 0.2429,
      "num_input_tokens_seen": 25390992,
      "step": 13190
    },
    {
      "epoch": 2.152540990292846,
      "grad_norm": 0.001592029700987041,
      "learning_rate": 0.22641704615088598,
      "loss": 0.2226,
      "num_input_tokens_seen": 25401136,
      "step": 13195
    },
    {
      "epoch": 2.153356717513663,
      "grad_norm": 0.001946097007021308,
      "learning_rate": 0.22636635250515103,
      "loss": 0.2436,
      "num_input_tokens_seen": 25410448,
      "step": 13200
    },
    {
      "epoch": 2.153356717513663,
      "eval_loss": 0.22487516701221466,
      "eval_runtime": 67.9317,
      "eval_samples_per_second": 40.114,
      "eval_steps_per_second": 20.064,
      "num_input_tokens_seen": 25410448,
      "step": 13200
    },
    {
      "epoch": 2.1541724447344808,
      "grad_norm": 0.0018768003210425377,
      "learning_rate": 0.2263156470827648,
      "loss": 0.2314,
      "num_input_tokens_seen": 25421264,
      "step": 13205
    },
    {
      "epoch": 2.1549881719552983,
      "grad_norm": 0.0009412807412445545,
      "learning_rate": 0.22626492989154678,
      "loss": 0.2232,
      "num_input_tokens_seen": 25429792,
      "step": 13210
    },
    {
      "epoch": 2.1558038991761155,
      "grad_norm": 0.0011874469928443432,
      "learning_rate": 0.22621420093931813,
      "loss": 0.2313,
      "num_input_tokens_seen": 25438736,
      "step": 13215
    },
    {
      "epoch": 2.156619626396933,
      "grad_norm": 0.0033140352461487055,
      "learning_rate": 0.22616346023390194,
      "loss": 0.2498,
      "num_input_tokens_seen": 25447760,
      "step": 13220
    },
    {
      "epoch": 2.15743535361775,
      "grad_norm": 0.0013014695141464472,
      "learning_rate": 0.22611270778312306,
      "loss": 0.2377,
      "num_input_tokens_seen": 25458848,
      "step": 13225
    },
    {
      "epoch": 2.1582510808385678,
      "grad_norm": 0.0022138492204248905,
      "learning_rate": 0.2260619435948081,
      "loss": 0.2373,
      "num_input_tokens_seen": 25468688,
      "step": 13230
    },
    {
      "epoch": 2.159066808059385,
      "grad_norm": 0.0023548705503344536,
      "learning_rate": 0.22601116767678567,
      "loss": 0.222,
      "num_input_tokens_seen": 25478032,
      "step": 13235
    },
    {
      "epoch": 2.1598825352802025,
      "grad_norm": 0.0013410262763500214,
      "learning_rate": 0.2259603800368859,
      "loss": 0.212,
      "num_input_tokens_seen": 25486976,
      "step": 13240
    },
    {
      "epoch": 2.1606982625010196,
      "grad_norm": 0.0020485869608819485,
      "learning_rate": 0.22590958068294098,
      "loss": 0.2431,
      "num_input_tokens_seen": 25497392,
      "step": 13245
    },
    {
      "epoch": 2.161513989721837,
      "grad_norm": 0.001836107112467289,
      "learning_rate": 0.22585876962278478,
      "loss": 0.187,
      "num_input_tokens_seen": 25507872,
      "step": 13250
    },
    {
      "epoch": 2.1623297169426543,
      "grad_norm": 0.0019853420089930296,
      "learning_rate": 0.22580794686425298,
      "loss": 0.2532,
      "num_input_tokens_seen": 25517040,
      "step": 13255
    },
    {
      "epoch": 2.163145444163472,
      "grad_norm": 0.0012575011933222413,
      "learning_rate": 0.22575711241518312,
      "loss": 0.2086,
      "num_input_tokens_seen": 25526816,
      "step": 13260
    },
    {
      "epoch": 2.163961171384289,
      "grad_norm": 0.0012516113929450512,
      "learning_rate": 0.22570626628341453,
      "loss": 0.2252,
      "num_input_tokens_seen": 25535904,
      "step": 13265
    },
    {
      "epoch": 2.1647768986051066,
      "grad_norm": 0.0019871010445058346,
      "learning_rate": 0.22565540847678828,
      "loss": 0.2169,
      "num_input_tokens_seen": 25545552,
      "step": 13270
    },
    {
      "epoch": 2.1655926258259237,
      "grad_norm": 0.0014729518443346024,
      "learning_rate": 0.2256045390031473,
      "loss": 0.2259,
      "num_input_tokens_seen": 25554656,
      "step": 13275
    },
    {
      "epoch": 2.1664083530467413,
      "grad_norm": 0.0030275629833340645,
      "learning_rate": 0.22555365787033627,
      "loss": 0.227,
      "num_input_tokens_seen": 25563280,
      "step": 13280
    },
    {
      "epoch": 2.1672240802675584,
      "grad_norm": 0.001388485892675817,
      "learning_rate": 0.22550276508620173,
      "loss": 0.2196,
      "num_input_tokens_seen": 25572688,
      "step": 13285
    },
    {
      "epoch": 2.168039807488376,
      "grad_norm": 0.002057622652500868,
      "learning_rate": 0.22545186065859202,
      "loss": 0.2213,
      "num_input_tokens_seen": 25581808,
      "step": 13290
    },
    {
      "epoch": 2.168855534709193,
      "grad_norm": 0.0023283089976757765,
      "learning_rate": 0.2254009445953572,
      "loss": 0.177,
      "num_input_tokens_seen": 25591856,
      "step": 13295
    },
    {
      "epoch": 2.1696712619300107,
      "grad_norm": 0.001950370380654931,
      "learning_rate": 0.22535001690434917,
      "loss": 0.2362,
      "num_input_tokens_seen": 25601568,
      "step": 13300
    },
    {
      "epoch": 2.170486989150828,
      "grad_norm": 0.0012249659048393369,
      "learning_rate": 0.22529907759342163,
      "loss": 0.1349,
      "num_input_tokens_seen": 25611120,
      "step": 13305
    },
    {
      "epoch": 2.1713027163716454,
      "grad_norm": 0.0036712412256747484,
      "learning_rate": 0.22524812667043007,
      "loss": 0.2851,
      "num_input_tokens_seen": 25621488,
      "step": 13310
    },
    {
      "epoch": 2.1721184435924625,
      "grad_norm": 0.0029514182824641466,
      "learning_rate": 0.22519716414323177,
      "loss": 0.2795,
      "num_input_tokens_seen": 25631168,
      "step": 13315
    },
    {
      "epoch": 2.17293417081328,
      "grad_norm": 0.001877459930256009,
      "learning_rate": 0.22514619001968567,
      "loss": 0.2263,
      "num_input_tokens_seen": 25640400,
      "step": 13320
    },
    {
      "epoch": 2.1737498980340972,
      "grad_norm": 0.0011459620436653495,
      "learning_rate": 0.2250952043076528,
      "loss": 0.2284,
      "num_input_tokens_seen": 25647808,
      "step": 13325
    },
    {
      "epoch": 2.174565625254915,
      "grad_norm": 0.0015945059712976217,
      "learning_rate": 0.2250442070149957,
      "loss": 0.2548,
      "num_input_tokens_seen": 25656224,
      "step": 13330
    },
    {
      "epoch": 2.175381352475732,
      "grad_norm": 0.0009625717648304999,
      "learning_rate": 0.22499319814957885,
      "loss": 0.2089,
      "num_input_tokens_seen": 25666368,
      "step": 13335
    },
    {
      "epoch": 2.1761970796965495,
      "grad_norm": 0.0018187299137935042,
      "learning_rate": 0.2249421777192684,
      "loss": 0.1913,
      "num_input_tokens_seen": 25675968,
      "step": 13340
    },
    {
      "epoch": 2.1770128069173666,
      "grad_norm": 0.002483511110767722,
      "learning_rate": 0.22489114573193236,
      "loss": 0.2621,
      "num_input_tokens_seen": 25684800,
      "step": 13345
    },
    {
      "epoch": 2.1778285341381842,
      "grad_norm": 0.0020493490155786276,
      "learning_rate": 0.2248401021954405,
      "loss": 0.2243,
      "num_input_tokens_seen": 25693712,
      "step": 13350
    },
    {
      "epoch": 2.1786442613590014,
      "grad_norm": 0.001734612975269556,
      "learning_rate": 0.22478904711766443,
      "loss": 0.2116,
      "num_input_tokens_seen": 25702624,
      "step": 13355
    },
    {
      "epoch": 2.179459988579819,
      "grad_norm": 0.0029549282044172287,
      "learning_rate": 0.22473798050647734,
      "loss": 0.2309,
      "num_input_tokens_seen": 25711744,
      "step": 13360
    },
    {
      "epoch": 2.180275715800636,
      "grad_norm": 0.0012409868650138378,
      "learning_rate": 0.22468690236975453,
      "loss": 0.2212,
      "num_input_tokens_seen": 25720064,
      "step": 13365
    },
    {
      "epoch": 2.1810914430214536,
      "grad_norm": 0.001666469033807516,
      "learning_rate": 0.22463581271537272,
      "loss": 0.2268,
      "num_input_tokens_seen": 25730672,
      "step": 13370
    },
    {
      "epoch": 2.1819071702422708,
      "grad_norm": 0.0016734753735363483,
      "learning_rate": 0.22458471155121076,
      "loss": 0.2397,
      "num_input_tokens_seen": 25740928,
      "step": 13375
    },
    {
      "epoch": 2.1827228974630883,
      "grad_norm": 0.0015760554233565927,
      "learning_rate": 0.2245335988851489,
      "loss": 0.2099,
      "num_input_tokens_seen": 25749600,
      "step": 13380
    },
    {
      "epoch": 2.183538624683906,
      "grad_norm": 0.0021966134663671255,
      "learning_rate": 0.2244824747250695,
      "loss": 0.2341,
      "num_input_tokens_seen": 25759056,
      "step": 13385
    },
    {
      "epoch": 2.184354351904723,
      "grad_norm": 0.0016167998546734452,
      "learning_rate": 0.22443133907885646,
      "loss": 0.2128,
      "num_input_tokens_seen": 25766576,
      "step": 13390
    },
    {
      "epoch": 2.1851700791255406,
      "grad_norm": 0.0035558026283979416,
      "learning_rate": 0.22438019195439557,
      "loss": 0.2429,
      "num_input_tokens_seen": 25776640,
      "step": 13395
    },
    {
      "epoch": 2.1859858063463578,
      "grad_norm": 0.0020483648404479027,
      "learning_rate": 0.22432903335957435,
      "loss": 0.2123,
      "num_input_tokens_seen": 25785744,
      "step": 13400
    },
    {
      "epoch": 2.1859858063463578,
      "eval_loss": 0.2229584902524948,
      "eval_runtime": 67.9104,
      "eval_samples_per_second": 40.126,
      "eval_steps_per_second": 20.071,
      "num_input_tokens_seen": 25785744,
      "step": 13400
    },
    {
      "epoch": 2.1868015335671753,
      "grad_norm": 0.0013868951937183738,
      "learning_rate": 0.22427786330228214,
      "loss": 0.2313,
      "num_input_tokens_seen": 25795504,
      "step": 13405
    },
    {
      "epoch": 2.1876172607879925,
      "grad_norm": 0.0023841294459998608,
      "learning_rate": 0.22422668179040997,
      "loss": 0.2285,
      "num_input_tokens_seen": 25805776,
      "step": 13410
    },
    {
      "epoch": 2.18843298800881,
      "grad_norm": 0.002002524910494685,
      "learning_rate": 0.2241754888318507,
      "loss": 0.2086,
      "num_input_tokens_seen": 25815840,
      "step": 13415
    },
    {
      "epoch": 2.189248715229627,
      "grad_norm": 0.0030234968289732933,
      "learning_rate": 0.22412428443449886,
      "loss": 0.231,
      "num_input_tokens_seen": 25824112,
      "step": 13420
    },
    {
      "epoch": 2.1900644424504447,
      "grad_norm": 0.003555614035576582,
      "learning_rate": 0.22407306860625087,
      "loss": 0.2107,
      "num_input_tokens_seen": 25833664,
      "step": 13425
    },
    {
      "epoch": 2.190880169671262,
      "grad_norm": 0.0014595186803489923,
      "learning_rate": 0.22402184135500483,
      "loss": 0.2334,
      "num_input_tokens_seen": 25842656,
      "step": 13430
    },
    {
      "epoch": 2.1916958968920794,
      "grad_norm": 0.0025221477262675762,
      "learning_rate": 0.22397060268866067,
      "loss": 0.2521,
      "num_input_tokens_seen": 25851584,
      "step": 13435
    },
    {
      "epoch": 2.1925116241128966,
      "grad_norm": 0.0010613186750561,
      "learning_rate": 0.22391935261511994,
      "loss": 0.249,
      "num_input_tokens_seen": 25860528,
      "step": 13440
    },
    {
      "epoch": 2.193327351333714,
      "grad_norm": 0.001339250011369586,
      "learning_rate": 0.22386809114228615,
      "loss": 0.2381,
      "num_input_tokens_seen": 25870448,
      "step": 13445
    },
    {
      "epoch": 2.1941430785545313,
      "grad_norm": 0.003779228078201413,
      "learning_rate": 0.22381681827806446,
      "loss": 0.2176,
      "num_input_tokens_seen": 25880496,
      "step": 13450
    },
    {
      "epoch": 2.194958805775349,
      "grad_norm": 0.0018159312894567847,
      "learning_rate": 0.22376553403036173,
      "loss": 0.2169,
      "num_input_tokens_seen": 25891504,
      "step": 13455
    },
    {
      "epoch": 2.195774532996166,
      "grad_norm": 0.0015850787749513984,
      "learning_rate": 0.22371423840708662,
      "loss": 0.223,
      "num_input_tokens_seen": 25901488,
      "step": 13460
    },
    {
      "epoch": 2.1965902602169836,
      "grad_norm": 0.002126997569575906,
      "learning_rate": 0.22366293141614962,
      "loss": 0.2502,
      "num_input_tokens_seen": 25910608,
      "step": 13465
    },
    {
      "epoch": 2.1974059874378007,
      "grad_norm": 0.0018597830785438418,
      "learning_rate": 0.22361161306546287,
      "loss": 0.2478,
      "num_input_tokens_seen": 25920864,
      "step": 13470
    },
    {
      "epoch": 2.1982217146586183,
      "grad_norm": 0.0015354631468653679,
      "learning_rate": 0.22356028336294037,
      "loss": 0.2008,
      "num_input_tokens_seen": 25929040,
      "step": 13475
    },
    {
      "epoch": 2.1990374418794354,
      "grad_norm": 0.0026017040945589542,
      "learning_rate": 0.2235089423164977,
      "loss": 0.2274,
      "num_input_tokens_seen": 25939008,
      "step": 13480
    },
    {
      "epoch": 2.199853169100253,
      "grad_norm": 0.0015490162186324596,
      "learning_rate": 0.22345758993405243,
      "loss": 0.2612,
      "num_input_tokens_seen": 25948720,
      "step": 13485
    },
    {
      "epoch": 2.20066889632107,
      "grad_norm": 0.0012879105051979423,
      "learning_rate": 0.2234062262235236,
      "loss": 0.2502,
      "num_input_tokens_seen": 25957360,
      "step": 13490
    },
    {
      "epoch": 2.2014846235418877,
      "grad_norm": 0.002577119506895542,
      "learning_rate": 0.22335485119283222,
      "loss": 0.2402,
      "num_input_tokens_seen": 25966048,
      "step": 13495
    },
    {
      "epoch": 2.202300350762705,
      "grad_norm": 0.001702954643405974,
      "learning_rate": 0.22330346484990093,
      "loss": 0.2285,
      "num_input_tokens_seen": 25976080,
      "step": 13500
    },
    {
      "epoch": 2.2031160779835224,
      "grad_norm": 0.0014464538544416428,
      "learning_rate": 0.22325206720265425,
      "loss": 0.2384,
      "num_input_tokens_seen": 25985232,
      "step": 13505
    },
    {
      "epoch": 2.2039318052043395,
      "grad_norm": 0.0028063105419278145,
      "learning_rate": 0.2232006582590182,
      "loss": 0.2241,
      "num_input_tokens_seen": 25994080,
      "step": 13510
    },
    {
      "epoch": 2.204747532425157,
      "grad_norm": 0.005615751259028912,
      "learning_rate": 0.22314923802692077,
      "loss": 0.2528,
      "num_input_tokens_seen": 26004464,
      "step": 13515
    },
    {
      "epoch": 2.205563259645974,
      "grad_norm": 0.002561195520684123,
      "learning_rate": 0.22309780651429156,
      "loss": 0.2102,
      "num_input_tokens_seen": 26013504,
      "step": 13520
    },
    {
      "epoch": 2.206378986866792,
      "grad_norm": 0.0011649734806269407,
      "learning_rate": 0.22304636372906203,
      "loss": 0.2504,
      "num_input_tokens_seen": 26022496,
      "step": 13525
    },
    {
      "epoch": 2.207194714087609,
      "grad_norm": 0.0018606656230986118,
      "learning_rate": 0.22299490967916522,
      "loss": 0.2121,
      "num_input_tokens_seen": 26031984,
      "step": 13530
    },
    {
      "epoch": 2.2080104413084265,
      "grad_norm": 0.0017580266576260328,
      "learning_rate": 0.22294344437253602,
      "loss": 0.2311,
      "num_input_tokens_seen": 26040912,
      "step": 13535
    },
    {
      "epoch": 2.2088261685292436,
      "grad_norm": 0.0031670955941081047,
      "learning_rate": 0.22289196781711101,
      "loss": 0.2362,
      "num_input_tokens_seen": 26050544,
      "step": 13540
    },
    {
      "epoch": 2.209641895750061,
      "grad_norm": 0.002815853338688612,
      "learning_rate": 0.2228404800208286,
      "loss": 0.2617,
      "num_input_tokens_seen": 26059488,
      "step": 13545
    },
    {
      "epoch": 2.2104576229708783,
      "grad_norm": 0.0032359687611460686,
      "learning_rate": 0.22278898099162875,
      "loss": 0.2516,
      "num_input_tokens_seen": 26069264,
      "step": 13550
    },
    {
      "epoch": 2.211273350191696,
      "grad_norm": 0.002598332706838846,
      "learning_rate": 0.22273747073745337,
      "loss": 0.2336,
      "num_input_tokens_seen": 26077680,
      "step": 13555
    },
    {
      "epoch": 2.2120890774125135,
      "grad_norm": 0.0012653721496462822,
      "learning_rate": 0.22268594926624588,
      "loss": 0.2371,
      "num_input_tokens_seen": 26087632,
      "step": 13560
    },
    {
      "epoch": 2.2129048046333306,
      "grad_norm": 0.002156355418264866,
      "learning_rate": 0.22263441658595162,
      "loss": 0.223,
      "num_input_tokens_seen": 26097328,
      "step": 13565
    },
    {
      "epoch": 2.2137205318541477,
      "grad_norm": 0.0014257014263421297,
      "learning_rate": 0.2225828727045175,
      "loss": 0.2269,
      "num_input_tokens_seen": 26106288,
      "step": 13570
    },
    {
      "epoch": 2.2145362590749653,
      "grad_norm": 0.002682384103536606,
      "learning_rate": 0.22253131762989228,
      "loss": 0.2328,
      "num_input_tokens_seen": 26115792,
      "step": 13575
    },
    {
      "epoch": 2.215351986295783,
      "grad_norm": 0.002551872981712222,
      "learning_rate": 0.2224797513700264,
      "loss": 0.2368,
      "num_input_tokens_seen": 26124512,
      "step": 13580
    },
    {
      "epoch": 2.2161677135166,
      "grad_norm": 0.001429209834896028,
      "learning_rate": 0.22242817393287204,
      "loss": 0.2163,
      "num_input_tokens_seen": 26134224,
      "step": 13585
    },
    {
      "epoch": 2.2169834407374176,
      "grad_norm": 0.0019416982540860772,
      "learning_rate": 0.22237658532638305,
      "loss": 0.2131,
      "num_input_tokens_seen": 26143584,
      "step": 13590
    },
    {
      "epoch": 2.2177991679582347,
      "grad_norm": 0.002239932306110859,
      "learning_rate": 0.22232498555851513,
      "loss": 0.2225,
      "num_input_tokens_seen": 26153408,
      "step": 13595
    },
    {
      "epoch": 2.2186148951790523,
      "grad_norm": 0.0016936915926635265,
      "learning_rate": 0.22227337463722546,
      "loss": 0.2422,
      "num_input_tokens_seen": 26163104,
      "step": 13600
    },
    {
      "epoch": 2.2186148951790523,
      "eval_loss": 0.23258113861083984,
      "eval_runtime": 67.8283,
      "eval_samples_per_second": 40.175,
      "eval_steps_per_second": 20.095,
      "num_input_tokens_seen": 26163104,
      "step": 13600
    },
    {
      "epoch": 2.2194306223998694,
      "grad_norm": 0.003651247126981616,
      "learning_rate": 0.2222217525704732,
      "loss": 0.2072,
      "num_input_tokens_seen": 26171696,
      "step": 13605
    },
    {
      "epoch": 2.220246349620687,
      "grad_norm": 0.006745154038071632,
      "learning_rate": 0.22217011936621908,
      "loss": 0.2201,
      "num_input_tokens_seen": 26180912,
      "step": 13610
    },
    {
      "epoch": 2.221062076841504,
      "grad_norm": 0.002265739953145385,
      "learning_rate": 0.22211847503242566,
      "loss": 0.2026,
      "num_input_tokens_seen": 26189680,
      "step": 13615
    },
    {
      "epoch": 2.2218778040623217,
      "grad_norm": 0.001136389677412808,
      "learning_rate": 0.22206681957705704,
      "loss": 0.2913,
      "num_input_tokens_seen": 26200608,
      "step": 13620
    },
    {
      "epoch": 2.222693531283139,
      "grad_norm": 0.001200114726088941,
      "learning_rate": 0.2220151530080792,
      "loss": 0.2613,
      "num_input_tokens_seen": 26209824,
      "step": 13625
    },
    {
      "epoch": 2.2235092585039564,
      "grad_norm": 0.0009466836345382035,
      "learning_rate": 0.2219634753334598,
      "loss": 0.2222,
      "num_input_tokens_seen": 26219680,
      "step": 13630
    },
    {
      "epoch": 2.2243249857247736,
      "grad_norm": 0.004434639122337103,
      "learning_rate": 0.22191178656116817,
      "loss": 0.2275,
      "num_input_tokens_seen": 26229136,
      "step": 13635
    },
    {
      "epoch": 2.225140712945591,
      "grad_norm": 0.002247035503387451,
      "learning_rate": 0.2218600866991753,
      "loss": 0.2369,
      "num_input_tokens_seen": 26238208,
      "step": 13640
    },
    {
      "epoch": 2.2259564401664083,
      "grad_norm": 0.018890058621764183,
      "learning_rate": 0.221808375755454,
      "loss": 0.2295,
      "num_input_tokens_seen": 26247904,
      "step": 13645
    },
    {
      "epoch": 2.226772167387226,
      "grad_norm": 0.001645658165216446,
      "learning_rate": 0.22175665373797881,
      "loss": 0.2319,
      "num_input_tokens_seen": 26257776,
      "step": 13650
    },
    {
      "epoch": 2.227587894608043,
      "grad_norm": 0.0025927552487701178,
      "learning_rate": 0.22170492065472583,
      "loss": 0.2502,
      "num_input_tokens_seen": 26267936,
      "step": 13655
    },
    {
      "epoch": 2.2284036218288605,
      "grad_norm": 0.0014753993600606918,
      "learning_rate": 0.221653176513673,
      "loss": 0.2238,
      "num_input_tokens_seen": 26277840,
      "step": 13660
    },
    {
      "epoch": 2.2292193490496777,
      "grad_norm": 0.0015629766276106238,
      "learning_rate": 0.2216014213227999,
      "loss": 0.2212,
      "num_input_tokens_seen": 26287296,
      "step": 13665
    },
    {
      "epoch": 2.2300350762704952,
      "grad_norm": 0.003616185626015067,
      "learning_rate": 0.22154965509008784,
      "loss": 0.2457,
      "num_input_tokens_seen": 26298224,
      "step": 13670
    },
    {
      "epoch": 2.2308508034913124,
      "grad_norm": 0.00204605795443058,
      "learning_rate": 0.2214978778235198,
      "loss": 0.2266,
      "num_input_tokens_seen": 26307744,
      "step": 13675
    },
    {
      "epoch": 2.23166653071213,
      "grad_norm": 0.0015411617932841182,
      "learning_rate": 0.2214460895310805,
      "loss": 0.2574,
      "num_input_tokens_seen": 26316992,
      "step": 13680
    },
    {
      "epoch": 2.232482257932947,
      "grad_norm": 0.0016346696065738797,
      "learning_rate": 0.22139429022075635,
      "loss": 0.2183,
      "num_input_tokens_seen": 26327424,
      "step": 13685
    },
    {
      "epoch": 2.2332979851537647,
      "grad_norm": 0.002126029459759593,
      "learning_rate": 0.22134247990053546,
      "loss": 0.2262,
      "num_input_tokens_seen": 26338000,
      "step": 13690
    },
    {
      "epoch": 2.234113712374582,
      "grad_norm": 0.001585689838975668,
      "learning_rate": 0.2212906585784076,
      "loss": 0.247,
      "num_input_tokens_seen": 26347952,
      "step": 13695
    },
    {
      "epoch": 2.2349294395953994,
      "grad_norm": 0.0010846588993445039,
      "learning_rate": 0.22123882626236432,
      "loss": 0.2235,
      "num_input_tokens_seen": 26358000,
      "step": 13700
    },
    {
      "epoch": 2.2357451668162165,
      "grad_norm": 0.0013859542086720467,
      "learning_rate": 0.2211869829603988,
      "loss": 0.2358,
      "num_input_tokens_seen": 26367104,
      "step": 13705
    },
    {
      "epoch": 2.236560894037034,
      "grad_norm": 0.002318417886272073,
      "learning_rate": 0.22113512868050592,
      "loss": 0.2542,
      "num_input_tokens_seen": 26376848,
      "step": 13710
    },
    {
      "epoch": 2.237376621257851,
      "grad_norm": 0.002808869816362858,
      "learning_rate": 0.2210832634306822,
      "loss": 0.2256,
      "num_input_tokens_seen": 26386128,
      "step": 13715
    },
    {
      "epoch": 2.2381923484786688,
      "grad_norm": 0.0016635515494272113,
      "learning_rate": 0.22103138721892598,
      "loss": 0.2206,
      "num_input_tokens_seen": 26394688,
      "step": 13720
    },
    {
      "epoch": 2.239008075699486,
      "grad_norm": 0.0017894740449264646,
      "learning_rate": 0.22097950005323724,
      "loss": 0.2156,
      "num_input_tokens_seen": 26404608,
      "step": 13725
    },
    {
      "epoch": 2.2398238029203035,
      "grad_norm": 0.0015097224386408925,
      "learning_rate": 0.22092760194161762,
      "loss": 0.2428,
      "num_input_tokens_seen": 26413408,
      "step": 13730
    },
    {
      "epoch": 2.2406395301411206,
      "grad_norm": 0.0010505453683435917,
      "learning_rate": 0.2208756928920704,
      "loss": 0.2228,
      "num_input_tokens_seen": 26423840,
      "step": 13735
    },
    {
      "epoch": 2.241455257361938,
      "grad_norm": 0.0016228900058194995,
      "learning_rate": 0.22082377291260072,
      "loss": 0.2084,
      "num_input_tokens_seen": 26432240,
      "step": 13740
    },
    {
      "epoch": 2.2422709845827553,
      "grad_norm": 0.0020256617572158575,
      "learning_rate": 0.2207718420112152,
      "loss": 0.2148,
      "num_input_tokens_seen": 26443088,
      "step": 13745
    },
    {
      "epoch": 2.243086711803573,
      "grad_norm": 0.0013174227206036448,
      "learning_rate": 0.22071990019592228,
      "loss": 0.2065,
      "num_input_tokens_seen": 26453024,
      "step": 13750
    },
    {
      "epoch": 2.2439024390243905,
      "grad_norm": 0.001712669269181788,
      "learning_rate": 0.22066794747473198,
      "loss": 0.2694,
      "num_input_tokens_seen": 26462704,
      "step": 13755
    },
    {
      "epoch": 2.2447181662452076,
      "grad_norm": 0.0023045851849019527,
      "learning_rate": 0.2206159838556562,
      "loss": 0.1922,
      "num_input_tokens_seen": 26473200,
      "step": 13760
    },
    {
      "epoch": 2.245533893466025,
      "grad_norm": 0.004262774251401424,
      "learning_rate": 0.2205640093467082,
      "loss": 0.2709,
      "num_input_tokens_seen": 26482048,
      "step": 13765
    },
    {
      "epoch": 2.2463496206868423,
      "grad_norm": 0.0022003427147865295,
      "learning_rate": 0.22051202395590322,
      "loss": 0.2342,
      "num_input_tokens_seen": 26490816,
      "step": 13770
    },
    {
      "epoch": 2.24716534790766,
      "grad_norm": 0.0021640111226588488,
      "learning_rate": 0.22046002769125808,
      "loss": 0.2165,
      "num_input_tokens_seen": 26500928,
      "step": 13775
    },
    {
      "epoch": 2.247981075128477,
      "grad_norm": 0.001628063153475523,
      "learning_rate": 0.2204080205607912,
      "loss": 0.2219,
      "num_input_tokens_seen": 26510688,
      "step": 13780
    },
    {
      "epoch": 2.2487968023492946,
      "grad_norm": 0.0012350277975201607,
      "learning_rate": 0.22035600257252272,
      "loss": 0.2404,
      "num_input_tokens_seen": 26518704,
      "step": 13785
    },
    {
      "epoch": 2.2496125295701117,
      "grad_norm": 0.0019005165668204427,
      "learning_rate": 0.2203039737344745,
      "loss": 0.2231,
      "num_input_tokens_seen": 26527376,
      "step": 13790
    },
    {
      "epoch": 2.2504282567909293,
      "grad_norm": 0.0009811189956963062,
      "learning_rate": 0.22025193405467003,
      "loss": 0.2482,
      "num_input_tokens_seen": 26536080,
      "step": 13795
    },
    {
      "epoch": 2.2512439840117464,
      "grad_norm": 0.0027152744587510824,
      "learning_rate": 0.2201998835411345,
      "loss": 0.2263,
      "num_input_tokens_seen": 26546240,
      "step": 13800
    },
    {
      "epoch": 2.2512439840117464,
      "eval_loss": 0.220285102725029,
      "eval_runtime": 67.9993,
      "eval_samples_per_second": 40.074,
      "eval_steps_per_second": 20.044,
      "num_input_tokens_seen": 26546240,
      "step": 13800
    },
    {
      "epoch": 2.252059711232564,
      "grad_norm": 0.0023616489488631487,
      "learning_rate": 0.22014782220189474,
      "loss": 0.2276,
      "num_input_tokens_seen": 26554800,
      "step": 13805
    },
    {
      "epoch": 2.252875438453381,
      "grad_norm": 0.004534952808171511,
      "learning_rate": 0.2200957500449793,
      "loss": 0.2323,
      "num_input_tokens_seen": 26563408,
      "step": 13810
    },
    {
      "epoch": 2.2536911656741987,
      "grad_norm": 0.0022423998452723026,
      "learning_rate": 0.22004366707841827,
      "loss": 0.2621,
      "num_input_tokens_seen": 26572720,
      "step": 13815
    },
    {
      "epoch": 2.254506892895016,
      "grad_norm": 0.0032932276371866465,
      "learning_rate": 0.21999157331024358,
      "loss": 0.2351,
      "num_input_tokens_seen": 26582432,
      "step": 13820
    },
    {
      "epoch": 2.2553226201158334,
      "grad_norm": 0.0009878978598862886,
      "learning_rate": 0.21993946874848871,
      "loss": 0.2309,
      "num_input_tokens_seen": 26592336,
      "step": 13825
    },
    {
      "epoch": 2.2561383473366505,
      "grad_norm": 0.00191101327072829,
      "learning_rate": 0.2198873534011888,
      "loss": 0.2267,
      "num_input_tokens_seen": 26601136,
      "step": 13830
    },
    {
      "epoch": 2.256954074557468,
      "grad_norm": 0.001911400118842721,
      "learning_rate": 0.2198352272763808,
      "loss": 0.2452,
      "num_input_tokens_seen": 26611904,
      "step": 13835
    },
    {
      "epoch": 2.2577698017782852,
      "grad_norm": 0.0011154816020280123,
      "learning_rate": 0.2197830903821031,
      "loss": 0.2172,
      "num_input_tokens_seen": 26619888,
      "step": 13840
    },
    {
      "epoch": 2.258585528999103,
      "grad_norm": 0.0015053675742819905,
      "learning_rate": 0.21973094272639598,
      "loss": 0.198,
      "num_input_tokens_seen": 26629936,
      "step": 13845
    },
    {
      "epoch": 2.25940125621992,
      "grad_norm": 0.003979410044848919,
      "learning_rate": 0.21967878431730117,
      "loss": 0.2383,
      "num_input_tokens_seen": 26639424,
      "step": 13850
    },
    {
      "epoch": 2.2602169834407375,
      "grad_norm": 0.003483768319711089,
      "learning_rate": 0.21962661516286217,
      "loss": 0.2536,
      "num_input_tokens_seen": 26648288,
      "step": 13855
    },
    {
      "epoch": 2.2610327106615546,
      "grad_norm": 0.0014927095035091043,
      "learning_rate": 0.21957443527112414,
      "loss": 0.2634,
      "num_input_tokens_seen": 26659328,
      "step": 13860
    },
    {
      "epoch": 2.261848437882372,
      "grad_norm": 0.0024827062152326107,
      "learning_rate": 0.21952224465013384,
      "loss": 0.237,
      "num_input_tokens_seen": 26667632,
      "step": 13865
    },
    {
      "epoch": 2.2626641651031894,
      "grad_norm": 0.0031295751687139273,
      "learning_rate": 0.21947004330793976,
      "loss": 0.2293,
      "num_input_tokens_seen": 26676640,
      "step": 13870
    },
    {
      "epoch": 2.263479892324007,
      "grad_norm": 0.0026839422062039375,
      "learning_rate": 0.21941783125259198,
      "loss": 0.233,
      "num_input_tokens_seen": 26685904,
      "step": 13875
    },
    {
      "epoch": 2.264295619544824,
      "grad_norm": 0.0019456238951534033,
      "learning_rate": 0.21936560849214226,
      "loss": 0.2354,
      "num_input_tokens_seen": 26695632,
      "step": 13880
    },
    {
      "epoch": 2.2651113467656416,
      "grad_norm": 0.0030657947063446045,
      "learning_rate": 0.21931337503464404,
      "loss": 0.2193,
      "num_input_tokens_seen": 26704880,
      "step": 13885
    },
    {
      "epoch": 2.2659270739864588,
      "grad_norm": 0.0033669278491288424,
      "learning_rate": 0.21926113088815233,
      "loss": 0.2492,
      "num_input_tokens_seen": 26713872,
      "step": 13890
    },
    {
      "epoch": 2.2667428012072763,
      "grad_norm": 0.003459226805716753,
      "learning_rate": 0.2192088760607238,
      "loss": 0.1979,
      "num_input_tokens_seen": 26723056,
      "step": 13895
    },
    {
      "epoch": 2.2675585284280935,
      "grad_norm": 0.0013944347156211734,
      "learning_rate": 0.2191566105604169,
      "loss": 0.2497,
      "num_input_tokens_seen": 26732208,
      "step": 13900
    },
    {
      "epoch": 2.268374255648911,
      "grad_norm": 0.0030219750478863716,
      "learning_rate": 0.21910433439529153,
      "loss": 0.2199,
      "num_input_tokens_seen": 26741344,
      "step": 13905
    },
    {
      "epoch": 2.269189982869728,
      "grad_norm": 0.006878149230033159,
      "learning_rate": 0.2190520475734094,
      "loss": 0.2247,
      "num_input_tokens_seen": 26750592,
      "step": 13910
    },
    {
      "epoch": 2.2700057100905457,
      "grad_norm": 0.0018991775577887893,
      "learning_rate": 0.2189997501028338,
      "loss": 0.345,
      "num_input_tokens_seen": 26758928,
      "step": 13915
    },
    {
      "epoch": 2.270821437311363,
      "grad_norm": 0.0009767537703737617,
      "learning_rate": 0.2189474419916296,
      "loss": 0.2438,
      "num_input_tokens_seen": 26768400,
      "step": 13920
    },
    {
      "epoch": 2.2716371645321805,
      "grad_norm": 0.0036061119753867388,
      "learning_rate": 0.21889512324786342,
      "loss": 0.2346,
      "num_input_tokens_seen": 26776832,
      "step": 13925
    },
    {
      "epoch": 2.272452891752998,
      "grad_norm": 0.001366576412692666,
      "learning_rate": 0.21884279387960345,
      "loss": 0.2338,
      "num_input_tokens_seen": 26786336,
      "step": 13930
    },
    {
      "epoch": 2.273268618973815,
      "grad_norm": 0.001874938141554594,
      "learning_rate": 0.2187904538949195,
      "loss": 0.2363,
      "num_input_tokens_seen": 26794992,
      "step": 13935
    },
    {
      "epoch": 2.2740843461946323,
      "grad_norm": 0.0017847215058282018,
      "learning_rate": 0.2187381033018831,
      "loss": 0.2701,
      "num_input_tokens_seen": 26805328,
      "step": 13940
    },
    {
      "epoch": 2.27490007341545,
      "grad_norm": 0.0010888224933296442,
      "learning_rate": 0.2186857421085673,
      "loss": 0.2397,
      "num_input_tokens_seen": 26814480,
      "step": 13945
    },
    {
      "epoch": 2.2757158006362674,
      "grad_norm": 0.0026666056364774704,
      "learning_rate": 0.21863337032304697,
      "loss": 0.2239,
      "num_input_tokens_seen": 26824832,
      "step": 13950
    },
    {
      "epoch": 2.2765315278570846,
      "grad_norm": 0.0013376012211665511,
      "learning_rate": 0.21858098795339845,
      "loss": 0.2145,
      "num_input_tokens_seen": 26835104,
      "step": 13955
    },
    {
      "epoch": 2.2773472550779017,
      "grad_norm": 0.002089991234242916,
      "learning_rate": 0.21852859500769975,
      "loss": 0.2164,
      "num_input_tokens_seen": 26845248,
      "step": 13960
    },
    {
      "epoch": 2.2781629822987193,
      "grad_norm": 0.0027380618266761303,
      "learning_rate": 0.21847619149403044,
      "loss": 0.2749,
      "num_input_tokens_seen": 26855456,
      "step": 13965
    },
    {
      "epoch": 2.278978709519537,
      "grad_norm": 0.0019328861963003874,
      "learning_rate": 0.21842377742047195,
      "loss": 0.2308,
      "num_input_tokens_seen": 26865168,
      "step": 13970
    },
    {
      "epoch": 2.279794436740354,
      "grad_norm": 0.0027691374998539686,
      "learning_rate": 0.21837135279510705,
      "loss": 0.2291,
      "num_input_tokens_seen": 26874464,
      "step": 13975
    },
    {
      "epoch": 2.2806101639611716,
      "grad_norm": 0.0019524643430486321,
      "learning_rate": 0.21831891762602038,
      "loss": 0.2401,
      "num_input_tokens_seen": 26883664,
      "step": 13980
    },
    {
      "epoch": 2.2814258911819887,
      "grad_norm": 0.002634994452819228,
      "learning_rate": 0.21826647192129806,
      "loss": 0.2073,
      "num_input_tokens_seen": 26892976,
      "step": 13985
    },
    {
      "epoch": 2.2822416184028063,
      "grad_norm": 0.001036110450513661,
      "learning_rate": 0.21821401568902787,
      "loss": 0.2376,
      "num_input_tokens_seen": 26902896,
      "step": 13990
    },
    {
      "epoch": 2.2830573456236234,
      "grad_norm": 0.002117401920258999,
      "learning_rate": 0.21816154893729925,
      "loss": 0.2347,
      "num_input_tokens_seen": 26913536,
      "step": 13995
    },
    {
      "epoch": 2.283873072844441,
      "grad_norm": 0.001008776598609984,
      "learning_rate": 0.2181090716742032,
      "loss": 0.2249,
      "num_input_tokens_seen": 26923408,
      "step": 14000
    },
    {
      "epoch": 2.283873072844441,
      "eval_loss": 0.22506460547447205,
      "eval_runtime": 67.9131,
      "eval_samples_per_second": 40.125,
      "eval_steps_per_second": 20.07,
      "num_input_tokens_seen": 26923408,
      "step": 14000
    },
    {
      "epoch": 2.284688800065258,
      "grad_norm": 0.002311903052031994,
      "learning_rate": 0.21805658390783236,
      "loss": 0.2547,
      "num_input_tokens_seen": 26933664,
      "step": 14005
    },
    {
      "epoch": 2.2855045272860757,
      "grad_norm": 0.0013627322623506188,
      "learning_rate": 0.21800408564628107,
      "loss": 0.2173,
      "num_input_tokens_seen": 26943312,
      "step": 14010
    },
    {
      "epoch": 2.286320254506893,
      "grad_norm": 0.001640821690671146,
      "learning_rate": 0.21795157689764516,
      "loss": 0.2566,
      "num_input_tokens_seen": 26952288,
      "step": 14015
    },
    {
      "epoch": 2.2871359817277104,
      "grad_norm": 0.0013655528891831636,
      "learning_rate": 0.21789905767002216,
      "loss": 0.2255,
      "num_input_tokens_seen": 26962448,
      "step": 14020
    },
    {
      "epoch": 2.2879517089485275,
      "grad_norm": 0.0018569548847153783,
      "learning_rate": 0.2178465279715112,
      "loss": 0.2399,
      "num_input_tokens_seen": 26971168,
      "step": 14025
    },
    {
      "epoch": 2.288767436169345,
      "grad_norm": 0.0007406910299323499,
      "learning_rate": 0.21779398781021303,
      "loss": 0.224,
      "num_input_tokens_seen": 26980336,
      "step": 14030
    },
    {
      "epoch": 2.289583163390162,
      "grad_norm": 0.0013852554839104414,
      "learning_rate": 0.21774143719422998,
      "loss": 0.224,
      "num_input_tokens_seen": 26990160,
      "step": 14035
    },
    {
      "epoch": 2.29039889061098,
      "grad_norm": 0.0009024828323163092,
      "learning_rate": 0.21768887613166601,
      "loss": 0.2269,
      "num_input_tokens_seen": 26999392,
      "step": 14040
    },
    {
      "epoch": 2.291214617831797,
      "grad_norm": 0.0018640387570485473,
      "learning_rate": 0.2176363046306267,
      "loss": 0.2156,
      "num_input_tokens_seen": 27007968,
      "step": 14045
    },
    {
      "epoch": 2.2920303450526145,
      "grad_norm": 0.0015205551171675324,
      "learning_rate": 0.21758372269921925,
      "loss": 0.2106,
      "num_input_tokens_seen": 27017952,
      "step": 14050
    },
    {
      "epoch": 2.2928460722734316,
      "grad_norm": 0.0022346682380884886,
      "learning_rate": 0.21753113034555244,
      "loss": 0.2576,
      "num_input_tokens_seen": 27027264,
      "step": 14055
    },
    {
      "epoch": 2.293661799494249,
      "grad_norm": 0.001520562102086842,
      "learning_rate": 0.2174785275777367,
      "loss": 0.2141,
      "num_input_tokens_seen": 27037648,
      "step": 14060
    },
    {
      "epoch": 2.2944775267150663,
      "grad_norm": 0.0010873290011659265,
      "learning_rate": 0.21742591440388404,
      "loss": 0.23,
      "num_input_tokens_seen": 27047056,
      "step": 14065
    },
    {
      "epoch": 2.295293253935884,
      "grad_norm": 0.0013331196969375014,
      "learning_rate": 0.21737329083210802,
      "loss": 0.2247,
      "num_input_tokens_seen": 27056240,
      "step": 14070
    },
    {
      "epoch": 2.296108981156701,
      "grad_norm": 0.001278732786886394,
      "learning_rate": 0.2173206568705239,
      "loss": 0.2254,
      "num_input_tokens_seen": 27066064,
      "step": 14075
    },
    {
      "epoch": 2.2969247083775186,
      "grad_norm": 0.0030221606139093637,
      "learning_rate": 0.2172680125272485,
      "loss": 0.2426,
      "num_input_tokens_seen": 27076064,
      "step": 14080
    },
    {
      "epoch": 2.2977404355983357,
      "grad_norm": 0.0010089784627780318,
      "learning_rate": 0.2172153578104002,
      "loss": 0.2308,
      "num_input_tokens_seen": 27085744,
      "step": 14085
    },
    {
      "epoch": 2.2985561628191533,
      "grad_norm": 0.0019825943745672703,
      "learning_rate": 0.21716269272809902,
      "loss": 0.2369,
      "num_input_tokens_seen": 27095760,
      "step": 14090
    },
    {
      "epoch": 2.2993718900399704,
      "grad_norm": 0.0009985630167648196,
      "learning_rate": 0.21711001728846666,
      "loss": 0.2447,
      "num_input_tokens_seen": 27105952,
      "step": 14095
    },
    {
      "epoch": 2.300187617260788,
      "grad_norm": 0.002406940096989274,
      "learning_rate": 0.21705733149962628,
      "loss": 0.2342,
      "num_input_tokens_seen": 27115184,
      "step": 14100
    },
    {
      "epoch": 2.3010033444816056,
      "grad_norm": 0.0012337297666817904,
      "learning_rate": 0.21700463536970263,
      "loss": 0.2322,
      "num_input_tokens_seen": 27125120,
      "step": 14105
    },
    {
      "epoch": 2.3018190717024227,
      "grad_norm": 0.0009864018065854907,
      "learning_rate": 0.21695192890682222,
      "loss": 0.2208,
      "num_input_tokens_seen": 27134128,
      "step": 14110
    },
    {
      "epoch": 2.30263479892324,
      "grad_norm": 0.003253266215324402,
      "learning_rate": 0.21689921211911298,
      "loss": 0.2201,
      "num_input_tokens_seen": 27143904,
      "step": 14115
    },
    {
      "epoch": 2.3034505261440574,
      "grad_norm": 0.001636991393752396,
      "learning_rate": 0.21684648501470452,
      "loss": 0.2087,
      "num_input_tokens_seen": 27154112,
      "step": 14120
    },
    {
      "epoch": 2.304266253364875,
      "grad_norm": 0.002206297591328621,
      "learning_rate": 0.216793747601728,
      "loss": 0.2421,
      "num_input_tokens_seen": 27162736,
      "step": 14125
    },
    {
      "epoch": 2.305081980585692,
      "grad_norm": 0.0018635581946000457,
      "learning_rate": 0.21674099988831627,
      "loss": 0.2171,
      "num_input_tokens_seen": 27172016,
      "step": 14130
    },
    {
      "epoch": 2.3058977078065093,
      "grad_norm": 0.003415819490328431,
      "learning_rate": 0.21668824188260363,
      "loss": 0.2445,
      "num_input_tokens_seen": 27181696,
      "step": 14135
    },
    {
      "epoch": 2.306713435027327,
      "grad_norm": 0.0012489233631640673,
      "learning_rate": 0.21663547359272606,
      "loss": 0.2284,
      "num_input_tokens_seen": 27190848,
      "step": 14140
    },
    {
      "epoch": 2.3075291622481444,
      "grad_norm": 0.0014047103468328714,
      "learning_rate": 0.216582695026821,
      "loss": 0.2569,
      "num_input_tokens_seen": 27200608,
      "step": 14145
    },
    {
      "epoch": 2.3083448894689615,
      "grad_norm": 0.0012463792227208614,
      "learning_rate": 0.21652990619302767,
      "loss": 0.2104,
      "num_input_tokens_seen": 27211392,
      "step": 14150
    },
    {
      "epoch": 2.309160616689779,
      "grad_norm": 0.0014176212716847658,
      "learning_rate": 0.21647710709948673,
      "loss": 0.2534,
      "num_input_tokens_seen": 27220432,
      "step": 14155
    },
    {
      "epoch": 2.3099763439105963,
      "grad_norm": 0.0020345936063677073,
      "learning_rate": 0.2164242977543405,
      "loss": 0.2226,
      "num_input_tokens_seen": 27229120,
      "step": 14160
    },
    {
      "epoch": 2.310792071131414,
      "grad_norm": 0.001476513920351863,
      "learning_rate": 0.21637147816573277,
      "loss": 0.2172,
      "num_input_tokens_seen": 27239312,
      "step": 14165
    },
    {
      "epoch": 2.311607798352231,
      "grad_norm": 0.0021768868900835514,
      "learning_rate": 0.21631864834180908,
      "loss": 0.2405,
      "num_input_tokens_seen": 27248464,
      "step": 14170
    },
    {
      "epoch": 2.3124235255730485,
      "grad_norm": 0.003462795866653323,
      "learning_rate": 0.21626580829071637,
      "loss": 0.222,
      "num_input_tokens_seen": 27258096,
      "step": 14175
    },
    {
      "epoch": 2.3132392527938657,
      "grad_norm": 0.0011947249295189977,
      "learning_rate": 0.21621295802060328,
      "loss": 0.2174,
      "num_input_tokens_seen": 27269424,
      "step": 14180
    },
    {
      "epoch": 2.3140549800146832,
      "grad_norm": 0.002584664849564433,
      "learning_rate": 0.21616009753961996,
      "loss": 0.2463,
      "num_input_tokens_seen": 27278592,
      "step": 14185
    },
    {
      "epoch": 2.3148707072355004,
      "grad_norm": 0.0012560473987832665,
      "learning_rate": 0.2161072268559182,
      "loss": 0.2285,
      "num_input_tokens_seen": 27288880,
      "step": 14190
    },
    {
      "epoch": 2.315686434456318,
      "grad_norm": 0.0024789473973214626,
      "learning_rate": 0.21605434597765133,
      "loss": 0.2252,
      "num_input_tokens_seen": 27299728,
      "step": 14195
    },
    {
      "epoch": 2.316502161677135,
      "grad_norm": 0.0010892009595409036,
      "learning_rate": 0.21600145491297418,
      "loss": 0.2394,
      "num_input_tokens_seen": 27309344,
      "step": 14200
    },
    {
      "epoch": 2.316502161677135,
      "eval_loss": 0.22729814052581787,
      "eval_runtime": 68.0728,
      "eval_samples_per_second": 40.031,
      "eval_steps_per_second": 20.023,
      "num_input_tokens_seen": 27309344,
      "step": 14200
    },
    {
      "epoch": 2.3173178888979526,
      "grad_norm": 0.002378567587584257,
      "learning_rate": 0.21594855367004326,
      "loss": 0.2161,
      "num_input_tokens_seen": 27319936,
      "step": 14205
    },
    {
      "epoch": 2.31813361611877,
      "grad_norm": 0.00352457701228559,
      "learning_rate": 0.21589564225701663,
      "loss": 0.2234,
      "num_input_tokens_seen": 27329680,
      "step": 14210
    },
    {
      "epoch": 2.3189493433395874,
      "grad_norm": 0.0017832155572250485,
      "learning_rate": 0.21584272068205385,
      "loss": 0.2184,
      "num_input_tokens_seen": 27340576,
      "step": 14215
    },
    {
      "epoch": 2.3197650705604045,
      "grad_norm": 0.0013921316713094711,
      "learning_rate": 0.2157897889533161,
      "loss": 0.2105,
      "num_input_tokens_seen": 27351056,
      "step": 14220
    },
    {
      "epoch": 2.320580797781222,
      "grad_norm": 0.001896918984130025,
      "learning_rate": 0.21573684707896612,
      "loss": 0.2168,
      "num_input_tokens_seen": 27361232,
      "step": 14225
    },
    {
      "epoch": 2.321396525002039,
      "grad_norm": 0.0012676042970269918,
      "learning_rate": 0.21568389506716826,
      "loss": 0.2217,
      "num_input_tokens_seen": 27370096,
      "step": 14230
    },
    {
      "epoch": 2.3222122522228568,
      "grad_norm": 0.002596597420051694,
      "learning_rate": 0.21563093292608831,
      "loss": 0.229,
      "num_input_tokens_seen": 27379840,
      "step": 14235
    },
    {
      "epoch": 2.323027979443674,
      "grad_norm": 0.005043770186603069,
      "learning_rate": 0.21557796066389376,
      "loss": 0.2667,
      "num_input_tokens_seen": 27388928,
      "step": 14240
    },
    {
      "epoch": 2.3238437066644915,
      "grad_norm": 0.0020392572041600943,
      "learning_rate": 0.21552497828875353,
      "loss": 0.2304,
      "num_input_tokens_seen": 27398288,
      "step": 14245
    },
    {
      "epoch": 2.3246594338853086,
      "grad_norm": 0.0031236964277923107,
      "learning_rate": 0.21547198580883828,
      "loss": 0.2498,
      "num_input_tokens_seen": 27408960,
      "step": 14250
    },
    {
      "epoch": 2.325475161106126,
      "grad_norm": 0.0051279072649776936,
      "learning_rate": 0.21541898323232,
      "loss": 0.218,
      "num_input_tokens_seen": 27419296,
      "step": 14255
    },
    {
      "epoch": 2.3262908883269433,
      "grad_norm": 0.0028242068365216255,
      "learning_rate": 0.2153659705673724,
      "loss": 0.25,
      "num_input_tokens_seen": 27428912,
      "step": 14260
    },
    {
      "epoch": 2.327106615547761,
      "grad_norm": 0.0037845198530703783,
      "learning_rate": 0.2153129478221707,
      "loss": 0.2624,
      "num_input_tokens_seen": 27436144,
      "step": 14265
    },
    {
      "epoch": 2.327922342768578,
      "grad_norm": 0.0014532232889905572,
      "learning_rate": 0.21525991500489164,
      "loss": 0.2189,
      "num_input_tokens_seen": 27446272,
      "step": 14270
    },
    {
      "epoch": 2.3287380699893956,
      "grad_norm": 0.002577844774350524,
      "learning_rate": 0.21520687212371362,
      "loss": 0.2416,
      "num_input_tokens_seen": 27455536,
      "step": 14275
    },
    {
      "epoch": 2.3295537972102127,
      "grad_norm": 0.0014426634879782796,
      "learning_rate": 0.21515381918681648,
      "loss": 0.23,
      "num_input_tokens_seen": 27465584,
      "step": 14280
    },
    {
      "epoch": 2.3303695244310303,
      "grad_norm": 0.0011449131416156888,
      "learning_rate": 0.21510075620238167,
      "loss": 0.1982,
      "num_input_tokens_seen": 27475280,
      "step": 14285
    },
    {
      "epoch": 2.3311852516518474,
      "grad_norm": 0.0029393956065177917,
      "learning_rate": 0.21504768317859208,
      "loss": 0.2607,
      "num_input_tokens_seen": 27485888,
      "step": 14290
    },
    {
      "epoch": 2.332000978872665,
      "grad_norm": 0.0014445721171796322,
      "learning_rate": 0.2149946001236323,
      "loss": 0.2081,
      "num_input_tokens_seen": 27496976,
      "step": 14295
    },
    {
      "epoch": 2.3328167060934826,
      "grad_norm": 0.0011736373417079449,
      "learning_rate": 0.21494150704568848,
      "loss": 0.2305,
      "num_input_tokens_seen": 27505424,
      "step": 14300
    },
    {
      "epoch": 2.3336324333142997,
      "grad_norm": 0.0012476489646360278,
      "learning_rate": 0.21488840395294811,
      "loss": 0.2416,
      "num_input_tokens_seen": 27514336,
      "step": 14305
    },
    {
      "epoch": 2.334448160535117,
      "grad_norm": 0.0007371913525275886,
      "learning_rate": 0.21483529085360042,
      "loss": 0.2034,
      "num_input_tokens_seen": 27523456,
      "step": 14310
    },
    {
      "epoch": 2.3352638877559344,
      "grad_norm": 0.0009369839681312442,
      "learning_rate": 0.2147821677558361,
      "loss": 0.2174,
      "num_input_tokens_seen": 27531904,
      "step": 14315
    },
    {
      "epoch": 2.336079614976752,
      "grad_norm": 0.0014137940015643835,
      "learning_rate": 0.2147290346678475,
      "loss": 0.1933,
      "num_input_tokens_seen": 27541776,
      "step": 14320
    },
    {
      "epoch": 2.336895342197569,
      "grad_norm": 0.002390409354120493,
      "learning_rate": 0.21467589159782827,
      "loss": 0.2648,
      "num_input_tokens_seen": 27551648,
      "step": 14325
    },
    {
      "epoch": 2.3377110694183862,
      "grad_norm": 0.0027248805854469538,
      "learning_rate": 0.21462273855397374,
      "loss": 0.2239,
      "num_input_tokens_seen": 27560480,
      "step": 14330
    },
    {
      "epoch": 2.338526796639204,
      "grad_norm": 0.0009870784124359488,
      "learning_rate": 0.21456957554448083,
      "loss": 0.1949,
      "num_input_tokens_seen": 27570624,
      "step": 14335
    },
    {
      "epoch": 2.3393425238600214,
      "grad_norm": 0.003319978015497327,
      "learning_rate": 0.21451640257754795,
      "loss": 0.2364,
      "num_input_tokens_seen": 27580128,
      "step": 14340
    },
    {
      "epoch": 2.3401582510808385,
      "grad_norm": 0.0019344042520970106,
      "learning_rate": 0.21446321966137508,
      "loss": 0.194,
      "num_input_tokens_seen": 27589648,
      "step": 14345
    },
    {
      "epoch": 2.340973978301656,
      "grad_norm": 0.0014476771466434002,
      "learning_rate": 0.21441002680416354,
      "loss": 0.2316,
      "num_input_tokens_seen": 27598752,
      "step": 14350
    },
    {
      "epoch": 2.3417897055224732,
      "grad_norm": 0.0022540728095918894,
      "learning_rate": 0.21435682401411654,
      "loss": 0.2337,
      "num_input_tokens_seen": 27607808,
      "step": 14355
    },
    {
      "epoch": 2.342605432743291,
      "grad_norm": 0.002245118608698249,
      "learning_rate": 0.2143036112994385,
      "loss": 0.2124,
      "num_input_tokens_seen": 27618560,
      "step": 14360
    },
    {
      "epoch": 2.343421159964108,
      "grad_norm": 0.0016396386781707406,
      "learning_rate": 0.21425038866833548,
      "loss": 0.1927,
      "num_input_tokens_seen": 27629888,
      "step": 14365
    },
    {
      "epoch": 2.3442368871849255,
      "grad_norm": 0.002378955716267228,
      "learning_rate": 0.21419715612901508,
      "loss": 0.2095,
      "num_input_tokens_seen": 27639568,
      "step": 14370
    },
    {
      "epoch": 2.3450526144057426,
      "grad_norm": 0.002799259964376688,
      "learning_rate": 0.21414391368968652,
      "loss": 0.2497,
      "num_input_tokens_seen": 27649760,
      "step": 14375
    },
    {
      "epoch": 2.34586834162656,
      "grad_norm": 0.0029056682251393795,
      "learning_rate": 0.21409066135856034,
      "loss": 0.2064,
      "num_input_tokens_seen": 27660640,
      "step": 14380
    },
    {
      "epoch": 2.3466840688473773,
      "grad_norm": 0.0013372688554227352,
      "learning_rate": 0.21403739914384878,
      "loss": 0.242,
      "num_input_tokens_seen": 27670304,
      "step": 14385
    },
    {
      "epoch": 2.347499796068195,
      "grad_norm": 0.0011219182051718235,
      "learning_rate": 0.21398412705376554,
      "loss": 0.2103,
      "num_input_tokens_seen": 27680880,
      "step": 14390
    },
    {
      "epoch": 2.348315523289012,
      "grad_norm": 0.0020003283862024546,
      "learning_rate": 0.2139308450965258,
      "loss": 0.2539,
      "num_input_tokens_seen": 27690128,
      "step": 14395
    },
    {
      "epoch": 2.3491312505098296,
      "grad_norm": 0.0023818244226276875,
      "learning_rate": 0.21387755328034638,
      "loss": 0.2427,
      "num_input_tokens_seen": 27698752,
      "step": 14400
    },
    {
      "epoch": 2.3491312505098296,
      "eval_loss": 0.22608648240566254,
      "eval_runtime": 67.9228,
      "eval_samples_per_second": 40.119,
      "eval_steps_per_second": 20.067,
      "num_input_tokens_seen": 27698752,
      "step": 14400
    },
    {
      "epoch": 2.3499469777306468,
      "grad_norm": 0.0012737257638946176,
      "learning_rate": 0.2138242516134455,
      "loss": 0.2127,
      "num_input_tokens_seen": 27708048,
      "step": 14405
    },
    {
      "epoch": 2.3507627049514643,
      "grad_norm": 0.001968945376574993,
      "learning_rate": 0.2137709401040429,
      "loss": 0.2352,
      "num_input_tokens_seen": 27716912,
      "step": 14410
    },
    {
      "epoch": 2.3515784321722815,
      "grad_norm": 0.0037983988877385855,
      "learning_rate": 0.21371761876036,
      "loss": 0.2367,
      "num_input_tokens_seen": 27726640,
      "step": 14415
    },
    {
      "epoch": 2.352394159393099,
      "grad_norm": 0.00159183400683105,
      "learning_rate": 0.21366428759061956,
      "loss": 0.237,
      "num_input_tokens_seen": 27736976,
      "step": 14420
    },
    {
      "epoch": 2.353209886613916,
      "grad_norm": 0.001125660608522594,
      "learning_rate": 0.2136109466030459,
      "loss": 0.2131,
      "num_input_tokens_seen": 27746128,
      "step": 14425
    },
    {
      "epoch": 2.3540256138347337,
      "grad_norm": 0.0012744576670229435,
      "learning_rate": 0.2135575958058649,
      "loss": 0.214,
      "num_input_tokens_seen": 27756576,
      "step": 14430
    },
    {
      "epoch": 2.354841341055551,
      "grad_norm": 0.001236051437444985,
      "learning_rate": 0.2135042352073039,
      "loss": 0.2752,
      "num_input_tokens_seen": 27765344,
      "step": 14435
    },
    {
      "epoch": 2.3556570682763684,
      "grad_norm": 0.0015865042805671692,
      "learning_rate": 0.2134508648155918,
      "loss": 0.1996,
      "num_input_tokens_seen": 27775824,
      "step": 14440
    },
    {
      "epoch": 2.3564727954971856,
      "grad_norm": 0.002016364596784115,
      "learning_rate": 0.213397484638959,
      "loss": 0.2045,
      "num_input_tokens_seen": 27785184,
      "step": 14445
    },
    {
      "epoch": 2.357288522718003,
      "grad_norm": 0.0014760070480406284,
      "learning_rate": 0.21334409468563728,
      "loss": 0.1871,
      "num_input_tokens_seen": 27793456,
      "step": 14450
    },
    {
      "epoch": 2.3581042499388203,
      "grad_norm": 0.0013718964764848351,
      "learning_rate": 0.2132906949638602,
      "loss": 0.2613,
      "num_input_tokens_seen": 27804288,
      "step": 14455
    },
    {
      "epoch": 2.358919977159638,
      "grad_norm": 0.0012723186518996954,
      "learning_rate": 0.21323728548186255,
      "loss": 0.2145,
      "num_input_tokens_seen": 27815200,
      "step": 14460
    },
    {
      "epoch": 2.359735704380455,
      "grad_norm": 0.002465655328705907,
      "learning_rate": 0.21318386624788088,
      "loss": 0.1924,
      "num_input_tokens_seen": 27825072,
      "step": 14465
    },
    {
      "epoch": 2.3605514316012726,
      "grad_norm": 0.001498905010521412,
      "learning_rate": 0.21313043727015288,
      "loss": 0.1932,
      "num_input_tokens_seen": 27834512,
      "step": 14470
    },
    {
      "epoch": 2.36136715882209,
      "grad_norm": 0.0022902146447449923,
      "learning_rate": 0.2130769985569182,
      "loss": 0.2221,
      "num_input_tokens_seen": 27843232,
      "step": 14475
    },
    {
      "epoch": 2.3621828860429073,
      "grad_norm": 0.001909158076159656,
      "learning_rate": 0.21302355011641766,
      "loss": 0.1915,
      "num_input_tokens_seen": 27853456,
      "step": 14480
    },
    {
      "epoch": 2.3629986132637244,
      "grad_norm": 0.002260468900203705,
      "learning_rate": 0.21297009195689365,
      "loss": 0.1803,
      "num_input_tokens_seen": 27863680,
      "step": 14485
    },
    {
      "epoch": 2.363814340484542,
      "grad_norm": 0.002461361000314355,
      "learning_rate": 0.21291662408659015,
      "loss": 0.1801,
      "num_input_tokens_seen": 27872016,
      "step": 14490
    },
    {
      "epoch": 2.3646300677053596,
      "grad_norm": 0.004413582384586334,
      "learning_rate": 0.21286314651375254,
      "loss": 0.2056,
      "num_input_tokens_seen": 27882048,
      "step": 14495
    },
    {
      "epoch": 2.3654457949261767,
      "grad_norm": 0.003535663476213813,
      "learning_rate": 0.2128096592466278,
      "loss": 0.2689,
      "num_input_tokens_seen": 27891040,
      "step": 14500
    },
    {
      "epoch": 2.366261522146994,
      "grad_norm": 0.002386820735409856,
      "learning_rate": 0.21275616229346428,
      "loss": 0.2119,
      "num_input_tokens_seen": 27899872,
      "step": 14505
    },
    {
      "epoch": 2.3670772493678114,
      "grad_norm": 0.0013016968732699752,
      "learning_rate": 0.21270265566251184,
      "loss": 0.2563,
      "num_input_tokens_seen": 27908848,
      "step": 14510
    },
    {
      "epoch": 2.367892976588629,
      "grad_norm": 0.0018723041284829378,
      "learning_rate": 0.21264913936202193,
      "loss": 0.1879,
      "num_input_tokens_seen": 27918800,
      "step": 14515
    },
    {
      "epoch": 2.368708703809446,
      "grad_norm": 0.005379993002861738,
      "learning_rate": 0.2125956134002475,
      "loss": 0.2161,
      "num_input_tokens_seen": 27928688,
      "step": 14520
    },
    {
      "epoch": 2.3695244310302637,
      "grad_norm": 0.0044066025875508785,
      "learning_rate": 0.2125420777854428,
      "loss": 0.2631,
      "num_input_tokens_seen": 27937536,
      "step": 14525
    },
    {
      "epoch": 2.370340158251081,
      "grad_norm": 0.0016397093422710896,
      "learning_rate": 0.21248853252586372,
      "loss": 0.2333,
      "num_input_tokens_seen": 27946192,
      "step": 14530
    },
    {
      "epoch": 2.3711558854718984,
      "grad_norm": 0.004915071185678244,
      "learning_rate": 0.21243497762976774,
      "loss": 0.2284,
      "num_input_tokens_seen": 27954336,
      "step": 14535
    },
    {
      "epoch": 2.3719716126927155,
      "grad_norm": 0.0029590153135359287,
      "learning_rate": 0.21238141310541356,
      "loss": 0.2336,
      "num_input_tokens_seen": 27963840,
      "step": 14540
    },
    {
      "epoch": 2.372787339913533,
      "grad_norm": 0.0013464963994920254,
      "learning_rate": 0.21232783896106153,
      "loss": 0.2262,
      "num_input_tokens_seen": 27973344,
      "step": 14545
    },
    {
      "epoch": 2.37360306713435,
      "grad_norm": 0.0022322116419672966,
      "learning_rate": 0.21227425520497345,
      "loss": 0.2527,
      "num_input_tokens_seen": 27983200,
      "step": 14550
    },
    {
      "epoch": 2.374418794355168,
      "grad_norm": 0.0029330113902688026,
      "learning_rate": 0.2122206618454127,
      "loss": 0.2322,
      "num_input_tokens_seen": 27992576,
      "step": 14555
    },
    {
      "epoch": 2.375234521575985,
      "grad_norm": 0.002545047551393509,
      "learning_rate": 0.2121670588906439,
      "loss": 0.217,
      "num_input_tokens_seen": 28002128,
      "step": 14560
    },
    {
      "epoch": 2.3760502487968025,
      "grad_norm": 0.0030833950731903315,
      "learning_rate": 0.21211344634893345,
      "loss": 0.2363,
      "num_input_tokens_seen": 28012160,
      "step": 14565
    },
    {
      "epoch": 2.3768659760176196,
      "grad_norm": 0.0026532637421041727,
      "learning_rate": 0.21205982422854897,
      "loss": 0.2176,
      "num_input_tokens_seen": 28021264,
      "step": 14570
    },
    {
      "epoch": 2.377681703238437,
      "grad_norm": 0.0009251474402844906,
      "learning_rate": 0.21200619253775974,
      "loss": 0.2204,
      "num_input_tokens_seen": 28031472,
      "step": 14575
    },
    {
      "epoch": 2.3784974304592543,
      "grad_norm": 0.0013684025034308434,
      "learning_rate": 0.21195255128483637,
      "loss": 0.2205,
      "num_input_tokens_seen": 28041232,
      "step": 14580
    },
    {
      "epoch": 2.379313157680072,
      "grad_norm": 0.0018614499131217599,
      "learning_rate": 0.21189890047805102,
      "loss": 0.2346,
      "num_input_tokens_seen": 28051600,
      "step": 14585
    },
    {
      "epoch": 2.380128884900889,
      "grad_norm": 0.0032809313852339983,
      "learning_rate": 0.21184524012567735,
      "loss": 0.2308,
      "num_input_tokens_seen": 28061888,
      "step": 14590
    },
    {
      "epoch": 2.3809446121217066,
      "grad_norm": 0.00159501435700804,
      "learning_rate": 0.2117915702359905,
      "loss": 0.222,
      "num_input_tokens_seen": 28072160,
      "step": 14595
    },
    {
      "epoch": 2.3817603393425237,
      "grad_norm": 0.0010947897098958492,
      "learning_rate": 0.211737890817267,
      "loss": 0.2083,
      "num_input_tokens_seen": 28082208,
      "step": 14600
    },
    {
      "epoch": 2.3817603393425237,
      "eval_loss": 0.2768711745738983,
      "eval_runtime": 68.0269,
      "eval_samples_per_second": 40.058,
      "eval_steps_per_second": 20.036,
      "num_input_tokens_seen": 28082208,
      "step": 14600
    },
    {
      "epoch": 2.3825760665633413,
      "grad_norm": 0.004892489407211542,
      "learning_rate": 0.21168420187778483,
      "loss": 0.2577,
      "num_input_tokens_seen": 28091728,
      "step": 14605
    },
    {
      "epoch": 2.3833917937841584,
      "grad_norm": 0.001473828568123281,
      "learning_rate": 0.21163050342582362,
      "loss": 0.2479,
      "num_input_tokens_seen": 28100656,
      "step": 14610
    },
    {
      "epoch": 2.384207521004976,
      "grad_norm": 0.0014238801086321473,
      "learning_rate": 0.21157679546966426,
      "loss": 0.2128,
      "num_input_tokens_seen": 28109216,
      "step": 14615
    },
    {
      "epoch": 2.385023248225793,
      "grad_norm": 0.0021940255537629128,
      "learning_rate": 0.2115230780175892,
      "loss": 0.2528,
      "num_input_tokens_seen": 28118752,
      "step": 14620
    },
    {
      "epoch": 2.3858389754466107,
      "grad_norm": 0.002117507392540574,
      "learning_rate": 0.21146935107788237,
      "loss": 0.2268,
      "num_input_tokens_seen": 28128528,
      "step": 14625
    },
    {
      "epoch": 2.386654702667428,
      "grad_norm": 0.002106891479343176,
      "learning_rate": 0.21141561465882916,
      "loss": 0.1963,
      "num_input_tokens_seen": 28139280,
      "step": 14630
    },
    {
      "epoch": 2.3874704298882454,
      "grad_norm": 0.0018905055476352572,
      "learning_rate": 0.21136186876871635,
      "loss": 0.2072,
      "num_input_tokens_seen": 28148384,
      "step": 14635
    },
    {
      "epoch": 2.3882861571090626,
      "grad_norm": 0.0030745016410946846,
      "learning_rate": 0.21130811341583225,
      "loss": 0.2072,
      "num_input_tokens_seen": 28157968,
      "step": 14640
    },
    {
      "epoch": 2.38910188432988,
      "grad_norm": 0.0022775668185204268,
      "learning_rate": 0.21125434860846667,
      "loss": 0.2119,
      "num_input_tokens_seen": 28167712,
      "step": 14645
    },
    {
      "epoch": 2.3899176115506973,
      "grad_norm": 0.003543699625879526,
      "learning_rate": 0.2112005743549107,
      "loss": 0.2538,
      "num_input_tokens_seen": 28177680,
      "step": 14650
    },
    {
      "epoch": 2.390733338771515,
      "grad_norm": 0.0027866377495229244,
      "learning_rate": 0.21114679066345707,
      "loss": 0.2533,
      "num_input_tokens_seen": 28187664,
      "step": 14655
    },
    {
      "epoch": 2.391549065992332,
      "grad_norm": 0.002840483095496893,
      "learning_rate": 0.21109299754239993,
      "loss": 0.2394,
      "num_input_tokens_seen": 28197968,
      "step": 14660
    },
    {
      "epoch": 2.3923647932131495,
      "grad_norm": 0.0014683161862194538,
      "learning_rate": 0.21103919500003482,
      "loss": 0.2074,
      "num_input_tokens_seen": 28207216,
      "step": 14665
    },
    {
      "epoch": 2.393180520433967,
      "grad_norm": 0.0019144181860610843,
      "learning_rate": 0.21098538304465872,
      "loss": 0.2529,
      "num_input_tokens_seen": 28218320,
      "step": 14670
    },
    {
      "epoch": 2.3939962476547842,
      "grad_norm": 0.001393920392729342,
      "learning_rate": 0.2109315616845702,
      "loss": 0.1822,
      "num_input_tokens_seen": 28228512,
      "step": 14675
    },
    {
      "epoch": 2.3948119748756014,
      "grad_norm": 0.001119649619795382,
      "learning_rate": 0.21087773092806925,
      "loss": 0.2381,
      "num_input_tokens_seen": 28237408,
      "step": 14680
    },
    {
      "epoch": 2.395627702096419,
      "grad_norm": 0.0007567527936771512,
      "learning_rate": 0.21082389078345704,
      "loss": 0.2174,
      "num_input_tokens_seen": 28248288,
      "step": 14685
    },
    {
      "epoch": 2.3964434293172365,
      "grad_norm": 0.0018220890779048204,
      "learning_rate": 0.2107700412590365,
      "loss": 0.2168,
      "num_input_tokens_seen": 28257600,
      "step": 14690
    },
    {
      "epoch": 2.3972591565380537,
      "grad_norm": 0.0014705673092976213,
      "learning_rate": 0.210716182363112,
      "loss": 0.2121,
      "num_input_tokens_seen": 28267840,
      "step": 14695
    },
    {
      "epoch": 2.398074883758871,
      "grad_norm": 0.001483533182181418,
      "learning_rate": 0.2106623141039891,
      "loss": 0.1983,
      "num_input_tokens_seen": 28276880,
      "step": 14700
    },
    {
      "epoch": 2.3988906109796884,
      "grad_norm": 0.0017101966077461839,
      "learning_rate": 0.21060843648997507,
      "loss": 0.2535,
      "num_input_tokens_seen": 28285776,
      "step": 14705
    },
    {
      "epoch": 2.399706338200506,
      "grad_norm": 0.0030136157292872667,
      "learning_rate": 0.21055454952937844,
      "loss": 0.1819,
      "num_input_tokens_seen": 28296048,
      "step": 14710
    },
    {
      "epoch": 2.400522065421323,
      "grad_norm": 0.0014969446929171681,
      "learning_rate": 0.21050065323050937,
      "loss": 0.2171,
      "num_input_tokens_seen": 28305616,
      "step": 14715
    },
    {
      "epoch": 2.4013377926421406,
      "grad_norm": 0.002640812424942851,
      "learning_rate": 0.21044674760167928,
      "loss": 0.2333,
      "num_input_tokens_seen": 28315056,
      "step": 14720
    },
    {
      "epoch": 2.4021535198629578,
      "grad_norm": 0.0015966537175700068,
      "learning_rate": 0.210392832651201,
      "loss": 0.2337,
      "num_input_tokens_seen": 28325104,
      "step": 14725
    },
    {
      "epoch": 2.4029692470837754,
      "grad_norm": 0.0011842731619253755,
      "learning_rate": 0.210338908387389,
      "loss": 0.2233,
      "num_input_tokens_seen": 28335040,
      "step": 14730
    },
    {
      "epoch": 2.4037849743045925,
      "grad_norm": 0.0011623589089140296,
      "learning_rate": 0.21028497481855912,
      "loss": 0.2045,
      "num_input_tokens_seen": 28344352,
      "step": 14735
    },
    {
      "epoch": 2.40460070152541,
      "grad_norm": 0.0013648956082761288,
      "learning_rate": 0.21023103195302847,
      "loss": 0.2016,
      "num_input_tokens_seen": 28354064,
      "step": 14740
    },
    {
      "epoch": 2.405416428746227,
      "grad_norm": 0.0023262640461325645,
      "learning_rate": 0.21017707979911582,
      "loss": 0.1767,
      "num_input_tokens_seen": 28363168,
      "step": 14745
    },
    {
      "epoch": 2.4062321559670448,
      "grad_norm": 0.002498546615242958,
      "learning_rate": 0.21012311836514122,
      "loss": 0.2184,
      "num_input_tokens_seen": 28372544,
      "step": 14750
    },
    {
      "epoch": 2.407047883187862,
      "grad_norm": 0.014295702800154686,
      "learning_rate": 0.21006914765942622,
      "loss": 0.217,
      "num_input_tokens_seen": 28383152,
      "step": 14755
    },
    {
      "epoch": 2.4078636104086795,
      "grad_norm": 0.0016608444275334477,
      "learning_rate": 0.2100151676902938,
      "loss": 0.1943,
      "num_input_tokens_seen": 28391712,
      "step": 14760
    },
    {
      "epoch": 2.4086793376294966,
      "grad_norm": 0.00219251262024045,
      "learning_rate": 0.2099611784660683,
      "loss": 0.2195,
      "num_input_tokens_seen": 28401168,
      "step": 14765
    },
    {
      "epoch": 2.409495064850314,
      "grad_norm": 0.002322717336937785,
      "learning_rate": 0.20990717999507552,
      "loss": 0.2619,
      "num_input_tokens_seen": 28410304,
      "step": 14770
    },
    {
      "epoch": 2.4103107920711313,
      "grad_norm": 0.0033512774389237165,
      "learning_rate": 0.20985317228564276,
      "loss": 0.2042,
      "num_input_tokens_seen": 28420192,
      "step": 14775
    },
    {
      "epoch": 2.411126519291949,
      "grad_norm": 0.0032078248914331198,
      "learning_rate": 0.20979915534609872,
      "loss": 0.1961,
      "num_input_tokens_seen": 28430896,
      "step": 14780
    },
    {
      "epoch": 2.411942246512766,
      "grad_norm": 0.0042601078748703,
      "learning_rate": 0.20974512918477342,
      "loss": 0.3114,
      "num_input_tokens_seen": 28441104,
      "step": 14785
    },
    {
      "epoch": 2.4127579737335836,
      "grad_norm": 0.0011604842729866505,
      "learning_rate": 0.2096910938099984,
      "loss": 0.2084,
      "num_input_tokens_seen": 28450000,
      "step": 14790
    },
    {
      "epoch": 2.4135737009544007,
      "grad_norm": 0.001667626784183085,
      "learning_rate": 0.2096370492301066,
      "loss": 0.2159,
      "num_input_tokens_seen": 28458672,
      "step": 14795
    },
    {
      "epoch": 2.4143894281752183,
      "grad_norm": 0.0039048960898071527,
      "learning_rate": 0.2095829954534323,
      "loss": 0.245,
      "num_input_tokens_seen": 28468576,
      "step": 14800
    },
    {
      "epoch": 2.4143894281752183,
      "eval_loss": 0.21673360466957092,
      "eval_runtime": 67.9654,
      "eval_samples_per_second": 40.094,
      "eval_steps_per_second": 20.054,
      "num_input_tokens_seen": 28468576,
      "step": 14800
    },
    {
      "epoch": 2.4152051553960354,
      "grad_norm": 0.002476586727425456,
      "learning_rate": 0.2095289324883114,
      "loss": 0.1868,
      "num_input_tokens_seen": 28477920,
      "step": 14805
    },
    {
      "epoch": 2.416020882616853,
      "grad_norm": 0.00264598336070776,
      "learning_rate": 0.20947486034308097,
      "loss": 0.2262,
      "num_input_tokens_seen": 28486848,
      "step": 14810
    },
    {
      "epoch": 2.41683660983767,
      "grad_norm": 0.0029637205880135298,
      "learning_rate": 0.2094207790260797,
      "loss": 0.2153,
      "num_input_tokens_seen": 28496624,
      "step": 14815
    },
    {
      "epoch": 2.4176523370584877,
      "grad_norm": 0.0012670777505263686,
      "learning_rate": 0.20936668854564758,
      "loss": 0.2429,
      "num_input_tokens_seen": 28506032,
      "step": 14820
    },
    {
      "epoch": 2.418468064279305,
      "grad_norm": 0.0027500237338244915,
      "learning_rate": 0.20931258891012602,
      "loss": 0.2095,
      "num_input_tokens_seen": 28515936,
      "step": 14825
    },
    {
      "epoch": 2.4192837915001224,
      "grad_norm": 0.0025631035678088665,
      "learning_rate": 0.20925848012785792,
      "loss": 0.2194,
      "num_input_tokens_seen": 28523696,
      "step": 14830
    },
    {
      "epoch": 2.4200995187209395,
      "grad_norm": 0.0023437077179551125,
      "learning_rate": 0.20920436220718747,
      "loss": 0.2445,
      "num_input_tokens_seen": 28534384,
      "step": 14835
    },
    {
      "epoch": 2.420915245941757,
      "grad_norm": 0.0019245194271206856,
      "learning_rate": 0.20915023515646033,
      "loss": 0.2134,
      "num_input_tokens_seen": 28544432,
      "step": 14840
    },
    {
      "epoch": 2.4217309731625742,
      "grad_norm": 0.004946418572217226,
      "learning_rate": 0.20909609898402368,
      "loss": 0.2173,
      "num_input_tokens_seen": 28553872,
      "step": 14845
    },
    {
      "epoch": 2.422546700383392,
      "grad_norm": 0.0013332648668438196,
      "learning_rate": 0.2090419536982258,
      "loss": 0.2091,
      "num_input_tokens_seen": 28562848,
      "step": 14850
    },
    {
      "epoch": 2.423362427604209,
      "grad_norm": 0.001380637288093567,
      "learning_rate": 0.2089877993074168,
      "loss": 0.228,
      "num_input_tokens_seen": 28571840,
      "step": 14855
    },
    {
      "epoch": 2.4241781548250265,
      "grad_norm": 0.0029684214387089014,
      "learning_rate": 0.20893363581994784,
      "loss": 0.2383,
      "num_input_tokens_seen": 28580336,
      "step": 14860
    },
    {
      "epoch": 2.424993882045844,
      "grad_norm": 0.0017067838925868273,
      "learning_rate": 0.2088794632441716,
      "loss": 0.2174,
      "num_input_tokens_seen": 28591280,
      "step": 14865
    },
    {
      "epoch": 2.4258096092666612,
      "grad_norm": 0.0016586438287049532,
      "learning_rate": 0.20882528158844219,
      "loss": 0.2029,
      "num_input_tokens_seen": 28600768,
      "step": 14870
    },
    {
      "epoch": 2.4266253364874784,
      "grad_norm": 0.0016672947676852345,
      "learning_rate": 0.20877109086111514,
      "loss": 0.2166,
      "num_input_tokens_seen": 28610544,
      "step": 14875
    },
    {
      "epoch": 2.427441063708296,
      "grad_norm": 0.0015163823263719678,
      "learning_rate": 0.2087168910705473,
      "loss": 0.2762,
      "num_input_tokens_seen": 28621136,
      "step": 14880
    },
    {
      "epoch": 2.4282567909291135,
      "grad_norm": 0.0022685073781758547,
      "learning_rate": 0.208662682225097,
      "loss": 0.1769,
      "num_input_tokens_seen": 28632544,
      "step": 14885
    },
    {
      "epoch": 2.4290725181499306,
      "grad_norm": 0.0019774152897298336,
      "learning_rate": 0.2086084643331239,
      "loss": 0.2699,
      "num_input_tokens_seen": 28642384,
      "step": 14890
    },
    {
      "epoch": 2.429888245370748,
      "grad_norm": 0.002460555639117956,
      "learning_rate": 0.20855423740298906,
      "loss": 0.1982,
      "num_input_tokens_seen": 28651632,
      "step": 14895
    },
    {
      "epoch": 2.4307039725915653,
      "grad_norm": 0.0034677223302423954,
      "learning_rate": 0.208500001443055,
      "loss": 0.2036,
      "num_input_tokens_seen": 28660800,
      "step": 14900
    },
    {
      "epoch": 2.431519699812383,
      "grad_norm": 0.002565046539530158,
      "learning_rate": 0.20844575646168553,
      "loss": 0.2157,
      "num_input_tokens_seen": 28669904,
      "step": 14905
    },
    {
      "epoch": 2.4323354270332,
      "grad_norm": 0.001801421632990241,
      "learning_rate": 0.20839150246724594,
      "loss": 0.2218,
      "num_input_tokens_seen": 28681392,
      "step": 14910
    },
    {
      "epoch": 2.4331511542540176,
      "grad_norm": 0.002848920412361622,
      "learning_rate": 0.20833723946810287,
      "loss": 0.2176,
      "num_input_tokens_seen": 28691824,
      "step": 14915
    },
    {
      "epoch": 2.4339668814748348,
      "grad_norm": 0.0010959430364891887,
      "learning_rate": 0.20828296747262437,
      "loss": 0.2198,
      "num_input_tokens_seen": 28701680,
      "step": 14920
    },
    {
      "epoch": 2.4347826086956523,
      "grad_norm": 0.0017154717352241278,
      "learning_rate": 0.20822868648917986,
      "loss": 0.245,
      "num_input_tokens_seen": 28710688,
      "step": 14925
    },
    {
      "epoch": 2.4355983359164695,
      "grad_norm": 0.0018365397118031979,
      "learning_rate": 0.20817439652614017,
      "loss": 0.2272,
      "num_input_tokens_seen": 28720976,
      "step": 14930
    },
    {
      "epoch": 2.436414063137287,
      "grad_norm": 0.002253451384603977,
      "learning_rate": 0.20812009759187744,
      "loss": 0.2231,
      "num_input_tokens_seen": 28731488,
      "step": 14935
    },
    {
      "epoch": 2.437229790358104,
      "grad_norm": 0.003363601164892316,
      "learning_rate": 0.2080657896947653,
      "loss": 0.1919,
      "num_input_tokens_seen": 28740064,
      "step": 14940
    },
    {
      "epoch": 2.4380455175789217,
      "grad_norm": 0.0021696053445339203,
      "learning_rate": 0.2080114728431787,
      "loss": 0.2309,
      "num_input_tokens_seen": 28748752,
      "step": 14945
    },
    {
      "epoch": 2.438861244799739,
      "grad_norm": 0.0010060009080916643,
      "learning_rate": 0.20795714704549392,
      "loss": 0.1702,
      "num_input_tokens_seen": 28758384,
      "step": 14950
    },
    {
      "epoch": 2.4396769720205564,
      "grad_norm": 0.0016515719471499324,
      "learning_rate": 0.20790281231008875,
      "loss": 0.205,
      "num_input_tokens_seen": 28768704,
      "step": 14955
    },
    {
      "epoch": 2.4404926992413736,
      "grad_norm": 0.0019450797699391842,
      "learning_rate": 0.20784846864534226,
      "loss": 0.2064,
      "num_input_tokens_seen": 28777408,
      "step": 14960
    },
    {
      "epoch": 2.441308426462191,
      "grad_norm": 0.0018305450212210417,
      "learning_rate": 0.20779411605963496,
      "loss": 0.2525,
      "num_input_tokens_seen": 28787456,
      "step": 14965
    },
    {
      "epoch": 2.4421241536830083,
      "grad_norm": 0.001796612050384283,
      "learning_rate": 0.2077397545613487,
      "loss": 0.1814,
      "num_input_tokens_seen": 28797504,
      "step": 14970
    },
    {
      "epoch": 2.442939880903826,
      "grad_norm": 0.0020491883624345064,
      "learning_rate": 0.20768538415886661,
      "loss": 0.2089,
      "num_input_tokens_seen": 28807824,
      "step": 14975
    },
    {
      "epoch": 2.443755608124643,
      "grad_norm": 0.0018592039123177528,
      "learning_rate": 0.20763100486057343,
      "loss": 0.2203,
      "num_input_tokens_seen": 28817456,
      "step": 14980
    },
    {
      "epoch": 2.4445713353454606,
      "grad_norm": 0.003460644045844674,
      "learning_rate": 0.20757661667485502,
      "loss": 0.2599,
      "num_input_tokens_seen": 28827680,
      "step": 14985
    },
    {
      "epoch": 2.4453870625662777,
      "grad_norm": 0.0015889555215835571,
      "learning_rate": 0.2075222196100988,
      "loss": 0.1766,
      "num_input_tokens_seen": 28837456,
      "step": 14990
    },
    {
      "epoch": 2.4462027897870953,
      "grad_norm": 0.001363363116979599,
      "learning_rate": 0.20746781367469344,
      "loss": 0.1799,
      "num_input_tokens_seen": 28847136,
      "step": 14995
    },
    {
      "epoch": 2.4470185170079124,
      "grad_norm": 0.0010361768072471023,
      "learning_rate": 0.207413398877029,
      "loss": 0.2304,
      "num_input_tokens_seen": 28856272,
      "step": 15000
    },
    {
      "epoch": 2.4470185170079124,
      "eval_loss": 0.2214488983154297,
      "eval_runtime": 67.8877,
      "eval_samples_per_second": 40.14,
      "eval_steps_per_second": 20.077,
      "num_input_tokens_seen": 28856272,
      "step": 15000
    },
    {
      "epoch": 2.44783424422873,
      "grad_norm": 0.0018944991752505302,
      "learning_rate": 0.20735897522549698,
      "loss": 0.1997,
      "num_input_tokens_seen": 28864672,
      "step": 15005
    },
    {
      "epoch": 2.448649971449547,
      "grad_norm": 0.0009016687981784344,
      "learning_rate": 0.2073045427284902,
      "loss": 0.1976,
      "num_input_tokens_seen": 28872784,
      "step": 15010
    },
    {
      "epoch": 2.4494656986703647,
      "grad_norm": 0.002656173426657915,
      "learning_rate": 0.2072501013944027,
      "loss": 0.285,
      "num_input_tokens_seen": 28882320,
      "step": 15015
    },
    {
      "epoch": 2.450281425891182,
      "grad_norm": 0.0011795035097748041,
      "learning_rate": 0.20719565123163017,
      "loss": 0.2286,
      "num_input_tokens_seen": 28891296,
      "step": 15020
    },
    {
      "epoch": 2.4510971531119994,
      "grad_norm": 0.0010948567651212215,
      "learning_rate": 0.20714119224856944,
      "loss": 0.2113,
      "num_input_tokens_seen": 28900304,
      "step": 15025
    },
    {
      "epoch": 2.4519128803328165,
      "grad_norm": 0.001975812716409564,
      "learning_rate": 0.2070867244536188,
      "loss": 0.2387,
      "num_input_tokens_seen": 28909440,
      "step": 15030
    },
    {
      "epoch": 2.452728607553634,
      "grad_norm": 0.0026336621958762407,
      "learning_rate": 0.20703224785517785,
      "loss": 0.1867,
      "num_input_tokens_seen": 28918928,
      "step": 15035
    },
    {
      "epoch": 2.4535443347744517,
      "grad_norm": 0.001836908282712102,
      "learning_rate": 0.20697776246164754,
      "loss": 0.2116,
      "num_input_tokens_seen": 28928480,
      "step": 15040
    },
    {
      "epoch": 2.454360061995269,
      "grad_norm": 0.000762640091124922,
      "learning_rate": 0.2069232682814303,
      "loss": 0.245,
      "num_input_tokens_seen": 28937312,
      "step": 15045
    },
    {
      "epoch": 2.455175789216086,
      "grad_norm": 0.0022109567653387785,
      "learning_rate": 0.20686876532292972,
      "loss": 0.2336,
      "num_input_tokens_seen": 28947760,
      "step": 15050
    },
    {
      "epoch": 2.4559915164369035,
      "grad_norm": 0.0011212484678253531,
      "learning_rate": 0.20681425359455083,
      "loss": 0.199,
      "num_input_tokens_seen": 28957920,
      "step": 15055
    },
    {
      "epoch": 2.456807243657721,
      "grad_norm": 0.0025775234680622816,
      "learning_rate": 0.20675973310470008,
      "loss": 0.2536,
      "num_input_tokens_seen": 28967136,
      "step": 15060
    },
    {
      "epoch": 2.457622970878538,
      "grad_norm": 0.002215788932517171,
      "learning_rate": 0.2067052038617852,
      "loss": 0.2509,
      "num_input_tokens_seen": 28976208,
      "step": 15065
    },
    {
      "epoch": 2.4584386980993553,
      "grad_norm": 0.001126080402173102,
      "learning_rate": 0.2066506658742153,
      "loss": 0.1893,
      "num_input_tokens_seen": 28986656,
      "step": 15070
    },
    {
      "epoch": 2.459254425320173,
      "grad_norm": 0.0015971055254340172,
      "learning_rate": 0.20659611915040077,
      "loss": 0.2401,
      "num_input_tokens_seen": 28996928,
      "step": 15075
    },
    {
      "epoch": 2.4600701525409905,
      "grad_norm": 0.0012199608609080315,
      "learning_rate": 0.20654156369875348,
      "loss": 0.1975,
      "num_input_tokens_seen": 29007024,
      "step": 15080
    },
    {
      "epoch": 2.4608858797618076,
      "grad_norm": 0.002442452125251293,
      "learning_rate": 0.20648699952768648,
      "loss": 0.2213,
      "num_input_tokens_seen": 29016480,
      "step": 15085
    },
    {
      "epoch": 2.461701606982625,
      "grad_norm": 0.0022415886633098125,
      "learning_rate": 0.20643242664561437,
      "loss": 0.1917,
      "num_input_tokens_seen": 29024832,
      "step": 15090
    },
    {
      "epoch": 2.4625173342034423,
      "grad_norm": 0.0018164013745263219,
      "learning_rate": 0.20637784506095277,
      "loss": 0.2242,
      "num_input_tokens_seen": 29034064,
      "step": 15095
    },
    {
      "epoch": 2.46333306142426,
      "grad_norm": 0.001870092237368226,
      "learning_rate": 0.20632325478211908,
      "loss": 0.2416,
      "num_input_tokens_seen": 29044064,
      "step": 15100
    },
    {
      "epoch": 2.464148788645077,
      "grad_norm": 0.0022788401693105698,
      "learning_rate": 0.20626865581753165,
      "loss": 0.2211,
      "num_input_tokens_seen": 29053968,
      "step": 15105
    },
    {
      "epoch": 2.4649645158658946,
      "grad_norm": 0.0016074571758508682,
      "learning_rate": 0.2062140481756104,
      "loss": 0.2285,
      "num_input_tokens_seen": 29062144,
      "step": 15110
    },
    {
      "epoch": 2.4657802430867117,
      "grad_norm": 0.0015062851598486304,
      "learning_rate": 0.20615943186477648,
      "loss": 0.2117,
      "num_input_tokens_seen": 29070672,
      "step": 15115
    },
    {
      "epoch": 2.4665959703075293,
      "grad_norm": 0.0019230862380936742,
      "learning_rate": 0.20610480689345242,
      "loss": 0.197,
      "num_input_tokens_seen": 29081072,
      "step": 15120
    },
    {
      "epoch": 2.4674116975283464,
      "grad_norm": 0.002617903985083103,
      "learning_rate": 0.2060501732700621,
      "loss": 0.251,
      "num_input_tokens_seen": 29090192,
      "step": 15125
    },
    {
      "epoch": 2.468227424749164,
      "grad_norm": 0.0015183733776211739,
      "learning_rate": 0.20599553100303067,
      "loss": 0.2168,
      "num_input_tokens_seen": 29100416,
      "step": 15130
    },
    {
      "epoch": 2.469043151969981,
      "grad_norm": 0.0014266237849369645,
      "learning_rate": 0.20594088010078465,
      "loss": 0.2224,
      "num_input_tokens_seen": 29109328,
      "step": 15135
    },
    {
      "epoch": 2.4698588791907987,
      "grad_norm": 0.0018374704523012042,
      "learning_rate": 0.20588622057175196,
      "loss": 0.2314,
      "num_input_tokens_seen": 29119552,
      "step": 15140
    },
    {
      "epoch": 2.470674606411616,
      "grad_norm": 0.0008809151477180421,
      "learning_rate": 0.20583155242436177,
      "loss": 0.2386,
      "num_input_tokens_seen": 29128656,
      "step": 15145
    },
    {
      "epoch": 2.4714903336324334,
      "grad_norm": 0.0018078292487189174,
      "learning_rate": 0.20577687566704453,
      "loss": 0.2099,
      "num_input_tokens_seen": 29137760,
      "step": 15150
    },
    {
      "epoch": 2.4723060608532506,
      "grad_norm": 0.0012448858469724655,
      "learning_rate": 0.20572219030823213,
      "loss": 0.242,
      "num_input_tokens_seen": 29146064,
      "step": 15155
    },
    {
      "epoch": 2.473121788074068,
      "grad_norm": 0.002385142259299755,
      "learning_rate": 0.20566749635635775,
      "loss": 0.2155,
      "num_input_tokens_seen": 29156528,
      "step": 15160
    },
    {
      "epoch": 2.4739375152948853,
      "grad_norm": 0.002195552922785282,
      "learning_rate": 0.20561279381985587,
      "loss": 0.1915,
      "num_input_tokens_seen": 29165936,
      "step": 15165
    },
    {
      "epoch": 2.474753242515703,
      "grad_norm": 0.002246246673166752,
      "learning_rate": 0.2055580827071623,
      "loss": 0.1962,
      "num_input_tokens_seen": 29175536,
      "step": 15170
    },
    {
      "epoch": 2.47556896973652,
      "grad_norm": 0.002587690018117428,
      "learning_rate": 0.20550336302671418,
      "loss": 0.2627,
      "num_input_tokens_seen": 29184928,
      "step": 15175
    },
    {
      "epoch": 2.4763846969573375,
      "grad_norm": 0.00259316130541265,
      "learning_rate": 0.20544863478695,
      "loss": 0.2424,
      "num_input_tokens_seen": 29195232,
      "step": 15180
    },
    {
      "epoch": 2.4772004241781547,
      "grad_norm": 0.0016334699466824532,
      "learning_rate": 0.20539389799630953,
      "loss": 0.2412,
      "num_input_tokens_seen": 29205248,
      "step": 15185
    },
    {
      "epoch": 2.4780161513989722,
      "grad_norm": 0.0017267196672037244,
      "learning_rate": 0.20533915266323388,
      "loss": 0.2192,
      "num_input_tokens_seen": 29215808,
      "step": 15190
    },
    {
      "epoch": 2.4788318786197894,
      "grad_norm": 0.0014366982504725456,
      "learning_rate": 0.20528439879616542,
      "loss": 0.2101,
      "num_input_tokens_seen": 29225952,
      "step": 15195
    },
    {
      "epoch": 2.479647605840607,
      "grad_norm": 0.0011822644155472517,
      "learning_rate": 0.20522963640354794,
      "loss": 0.2417,
      "num_input_tokens_seen": 29234704,
      "step": 15200
    },
    {
      "epoch": 2.479647605840607,
      "eval_loss": 0.21495427191257477,
      "eval_runtime": 68.1125,
      "eval_samples_per_second": 40.007,
      "eval_steps_per_second": 20.011,
      "num_input_tokens_seen": 29234704,
      "step": 15200
    },
    {
      "epoch": 2.480463333061424,
      "grad_norm": 0.0016335638938471675,
      "learning_rate": 0.20517486549382644,
      "loss": 0.2346,
      "num_input_tokens_seen": 29244176,
      "step": 15205
    },
    {
      "epoch": 2.4812790602822417,
      "grad_norm": 0.0019151713931933045,
      "learning_rate": 0.20512008607544735,
      "loss": 0.2182,
      "num_input_tokens_seen": 29253280,
      "step": 15210
    },
    {
      "epoch": 2.482094787503059,
      "grad_norm": 0.0010849789250642061,
      "learning_rate": 0.20506529815685826,
      "loss": 0.2456,
      "num_input_tokens_seen": 29263024,
      "step": 15215
    },
    {
      "epoch": 2.4829105147238764,
      "grad_norm": 0.0016740799183025956,
      "learning_rate": 0.2050105017465082,
      "loss": 0.2566,
      "num_input_tokens_seen": 29272528,
      "step": 15220
    },
    {
      "epoch": 2.4837262419446935,
      "grad_norm": 0.0020126905292272568,
      "learning_rate": 0.20495569685284754,
      "loss": 0.2137,
      "num_input_tokens_seen": 29282192,
      "step": 15225
    },
    {
      "epoch": 2.484541969165511,
      "grad_norm": 0.001489166752435267,
      "learning_rate": 0.20490088348432778,
      "loss": 0.2142,
      "num_input_tokens_seen": 29291504,
      "step": 15230
    },
    {
      "epoch": 2.4853576963863286,
      "grad_norm": 0.002574008423835039,
      "learning_rate": 0.2048460616494018,
      "loss": 0.211,
      "num_input_tokens_seen": 29299424,
      "step": 15235
    },
    {
      "epoch": 2.4861734236071458,
      "grad_norm": 0.00443600257858634,
      "learning_rate": 0.2047912313565239,
      "loss": 0.2015,
      "num_input_tokens_seen": 29309376,
      "step": 15240
    },
    {
      "epoch": 2.486989150827963,
      "grad_norm": 0.0009303492843173444,
      "learning_rate": 0.20473639261414958,
      "loss": 0.2273,
      "num_input_tokens_seen": 29318544,
      "step": 15245
    },
    {
      "epoch": 2.4878048780487805,
      "grad_norm": 0.0007457711617462337,
      "learning_rate": 0.2046815454307357,
      "loss": 0.1983,
      "num_input_tokens_seen": 29328448,
      "step": 15250
    },
    {
      "epoch": 2.488620605269598,
      "grad_norm": 0.0018550468375906348,
      "learning_rate": 0.20462668981474028,
      "loss": 0.2439,
      "num_input_tokens_seen": 29337936,
      "step": 15255
    },
    {
      "epoch": 2.489436332490415,
      "grad_norm": 0.0009728007134981453,
      "learning_rate": 0.20457182577462288,
      "loss": 0.1925,
      "num_input_tokens_seen": 29347120,
      "step": 15260
    },
    {
      "epoch": 2.4902520597112328,
      "grad_norm": 0.0016056502936407924,
      "learning_rate": 0.2045169533188441,
      "loss": 0.1939,
      "num_input_tokens_seen": 29356464,
      "step": 15265
    },
    {
      "epoch": 2.49106778693205,
      "grad_norm": 0.0016835081623867154,
      "learning_rate": 0.20446207245586603,
      "loss": 0.2281,
      "num_input_tokens_seen": 29365584,
      "step": 15270
    },
    {
      "epoch": 2.4918835141528675,
      "grad_norm": 0.0015961843309924006,
      "learning_rate": 0.20440718319415196,
      "loss": 0.2402,
      "num_input_tokens_seen": 29375424,
      "step": 15275
    },
    {
      "epoch": 2.4926992413736846,
      "grad_norm": 0.0014916285872459412,
      "learning_rate": 0.20435228554216653,
      "loss": 0.2167,
      "num_input_tokens_seen": 29385056,
      "step": 15280
    },
    {
      "epoch": 2.493514968594502,
      "grad_norm": 0.0023265003692358732,
      "learning_rate": 0.20429737950837565,
      "loss": 0.2161,
      "num_input_tokens_seen": 29395760,
      "step": 15285
    },
    {
      "epoch": 2.4943306958153193,
      "grad_norm": 0.0015760788228362799,
      "learning_rate": 0.20424246510124647,
      "loss": 0.2225,
      "num_input_tokens_seen": 29405472,
      "step": 15290
    },
    {
      "epoch": 2.495146423036137,
      "grad_norm": 0.002076050965115428,
      "learning_rate": 0.20418754232924755,
      "loss": 0.2276,
      "num_input_tokens_seen": 29414656,
      "step": 15295
    },
    {
      "epoch": 2.495962150256954,
      "grad_norm": 0.00242139445617795,
      "learning_rate": 0.20413261120084863,
      "loss": 0.2164,
      "num_input_tokens_seen": 29424240,
      "step": 15300
    },
    {
      "epoch": 2.4967778774777716,
      "grad_norm": 0.002226895885542035,
      "learning_rate": 0.2040776717245208,
      "loss": 0.2744,
      "num_input_tokens_seen": 29433440,
      "step": 15305
    },
    {
      "epoch": 2.4975936046985887,
      "grad_norm": 0.0012048408389091492,
      "learning_rate": 0.2040227239087364,
      "loss": 0.2146,
      "num_input_tokens_seen": 29442416,
      "step": 15310
    },
    {
      "epoch": 2.4984093319194063,
      "grad_norm": 0.0026003208477050066,
      "learning_rate": 0.20396776776196904,
      "loss": 0.2511,
      "num_input_tokens_seen": 29452944,
      "step": 15315
    },
    {
      "epoch": 2.4992250591402234,
      "grad_norm": 0.0025522003415971994,
      "learning_rate": 0.20391280329269373,
      "loss": 0.2403,
      "num_input_tokens_seen": 29461920,
      "step": 15320
    },
    {
      "epoch": 2.500040786361041,
      "grad_norm": 0.0019515780732035637,
      "learning_rate": 0.20385783050938663,
      "loss": 0.1803,
      "num_input_tokens_seen": 29471520,
      "step": 15325
    },
    {
      "epoch": 2.500856513581858,
      "grad_norm": 0.001000081654638052,
      "learning_rate": 0.20380284942052526,
      "loss": 0.164,
      "num_input_tokens_seen": 29481088,
      "step": 15330
    },
    {
      "epoch": 2.5016722408026757,
      "grad_norm": 0.002108782297000289,
      "learning_rate": 0.2037478600345884,
      "loss": 0.253,
      "num_input_tokens_seen": 29491904,
      "step": 15335
    },
    {
      "epoch": 2.502487968023493,
      "grad_norm": 0.0017249800730496645,
      "learning_rate": 0.20369286236005604,
      "loss": 0.2099,
      "num_input_tokens_seen": 29501776,
      "step": 15340
    },
    {
      "epoch": 2.5033036952443104,
      "grad_norm": 0.0018102729227393866,
      "learning_rate": 0.20363785640540957,
      "loss": 0.2261,
      "num_input_tokens_seen": 29512016,
      "step": 15345
    },
    {
      "epoch": 2.5041194224651275,
      "grad_norm": 0.001206591259688139,
      "learning_rate": 0.2035828421791316,
      "loss": 0.2243,
      "num_input_tokens_seen": 29522064,
      "step": 15350
    },
    {
      "epoch": 2.504935149685945,
      "grad_norm": 0.0025987832341343164,
      "learning_rate": 0.20352781968970599,
      "loss": 0.233,
      "num_input_tokens_seen": 29531440,
      "step": 15355
    },
    {
      "epoch": 2.5057508769067622,
      "grad_norm": 0.001353272469714284,
      "learning_rate": 0.2034727889456179,
      "loss": 0.2221,
      "num_input_tokens_seen": 29541024,
      "step": 15360
    },
    {
      "epoch": 2.50656660412758,
      "grad_norm": 0.003267630934715271,
      "learning_rate": 0.2034177499553538,
      "loss": 0.2674,
      "num_input_tokens_seen": 29552128,
      "step": 15365
    },
    {
      "epoch": 2.507382331348397,
      "grad_norm": 0.002863501198589802,
      "learning_rate": 0.2033627027274014,
      "loss": 0.2111,
      "num_input_tokens_seen": 29561248,
      "step": 15370
    },
    {
      "epoch": 2.5081980585692145,
      "grad_norm": 0.0013531867880374193,
      "learning_rate": 0.20330764727024955,
      "loss": 0.2507,
      "num_input_tokens_seen": 29571152,
      "step": 15375
    },
    {
      "epoch": 2.5090137857900316,
      "grad_norm": 0.001375403837300837,
      "learning_rate": 0.20325258359238868,
      "loss": 0.2067,
      "num_input_tokens_seen": 29581248,
      "step": 15380
    },
    {
      "epoch": 2.5098295130108492,
      "grad_norm": 0.000982609111815691,
      "learning_rate": 0.20319751170231018,
      "loss": 0.2178,
      "num_input_tokens_seen": 29590896,
      "step": 15385
    },
    {
      "epoch": 2.510645240231667,
      "grad_norm": 0.0018687976989895105,
      "learning_rate": 0.2031424316085068,
      "loss": 0.2181,
      "num_input_tokens_seen": 29600064,
      "step": 15390
    },
    {
      "epoch": 2.511460967452484,
      "grad_norm": 0.0011950316838920116,
      "learning_rate": 0.20308734331947265,
      "loss": 0.2074,
      "num_input_tokens_seen": 29608448,
      "step": 15395
    },
    {
      "epoch": 2.512276694673301,
      "grad_norm": 0.002266897587105632,
      "learning_rate": 0.20303224684370305,
      "loss": 0.2322,
      "num_input_tokens_seen": 29617728,
      "step": 15400
    },
    {
      "epoch": 2.512276694673301,
      "eval_loss": 0.21487316489219666,
      "eval_runtime": 67.9806,
      "eval_samples_per_second": 40.085,
      "eval_steps_per_second": 20.05,
      "num_input_tokens_seen": 29617728,
      "step": 15400
    },
    {
      "epoch": 2.5130924218941186,
      "grad_norm": 0.0014858933864161372,
      "learning_rate": 0.20297714218969456,
      "loss": 0.2086,
      "num_input_tokens_seen": 29625632,
      "step": 15405
    },
    {
      "epoch": 2.513908149114936,
      "grad_norm": 0.0022721891291439533,
      "learning_rate": 0.20292202936594497,
      "loss": 0.2219,
      "num_input_tokens_seen": 29635936,
      "step": 15410
    },
    {
      "epoch": 2.5147238763357533,
      "grad_norm": 0.0014115930534899235,
      "learning_rate": 0.2028669083809534,
      "loss": 0.1828,
      "num_input_tokens_seen": 29645408,
      "step": 15415
    },
    {
      "epoch": 2.5155396035565705,
      "grad_norm": 0.0027232610154896975,
      "learning_rate": 0.20281177924322016,
      "loss": 0.2293,
      "num_input_tokens_seen": 29656480,
      "step": 15420
    },
    {
      "epoch": 2.516355330777388,
      "grad_norm": 0.002789836609736085,
      "learning_rate": 0.2027566419612469,
      "loss": 0.2193,
      "num_input_tokens_seen": 29667024,
      "step": 15425
    },
    {
      "epoch": 2.5171710579982056,
      "grad_norm": 0.002186058321967721,
      "learning_rate": 0.20270149654353647,
      "loss": 0.2004,
      "num_input_tokens_seen": 29676944,
      "step": 15430
    },
    {
      "epoch": 2.5179867852190227,
      "grad_norm": 0.0009815297089517117,
      "learning_rate": 0.202646342998593,
      "loss": 0.2424,
      "num_input_tokens_seen": 29686128,
      "step": 15435
    },
    {
      "epoch": 2.51880251243984,
      "grad_norm": 0.0022530059795826674,
      "learning_rate": 0.20259118133492185,
      "loss": 0.1894,
      "num_input_tokens_seen": 29695536,
      "step": 15440
    },
    {
      "epoch": 2.5196182396606575,
      "grad_norm": 0.002197913359850645,
      "learning_rate": 0.20253601156102966,
      "loss": 0.2044,
      "num_input_tokens_seen": 29704928,
      "step": 15445
    },
    {
      "epoch": 2.520433966881475,
      "grad_norm": 0.0026819093618541956,
      "learning_rate": 0.20248083368542422,
      "loss": 0.2155,
      "num_input_tokens_seen": 29715296,
      "step": 15450
    },
    {
      "epoch": 2.521249694102292,
      "grad_norm": 0.0021358467638492584,
      "learning_rate": 0.2024256477166147,
      "loss": 0.2939,
      "num_input_tokens_seen": 29725072,
      "step": 15455
    },
    {
      "epoch": 2.5220654213231093,
      "grad_norm": 0.0026275552809238434,
      "learning_rate": 0.2023704536631115,
      "loss": 0.2489,
      "num_input_tokens_seen": 29734592,
      "step": 15460
    },
    {
      "epoch": 2.522881148543927,
      "grad_norm": 0.0016531444853171706,
      "learning_rate": 0.20231525153342625,
      "loss": 0.2202,
      "num_input_tokens_seen": 29742800,
      "step": 15465
    },
    {
      "epoch": 2.5236968757647444,
      "grad_norm": 0.0017179761780425906,
      "learning_rate": 0.20226004133607173,
      "loss": 0.2217,
      "num_input_tokens_seen": 29752368,
      "step": 15470
    },
    {
      "epoch": 2.5245126029855616,
      "grad_norm": 0.002822363516315818,
      "learning_rate": 0.20220482307956214,
      "loss": 0.2442,
      "num_input_tokens_seen": 29762176,
      "step": 15475
    },
    {
      "epoch": 2.525328330206379,
      "grad_norm": 0.0013605817221105099,
      "learning_rate": 0.20214959677241276,
      "loss": 0.1898,
      "num_input_tokens_seen": 29771920,
      "step": 15480
    },
    {
      "epoch": 2.5261440574271963,
      "grad_norm": 0.0010586156276986003,
      "learning_rate": 0.20209436242314022,
      "loss": 0.2374,
      "num_input_tokens_seen": 29781792,
      "step": 15485
    },
    {
      "epoch": 2.526959784648014,
      "grad_norm": 0.002409190172329545,
      "learning_rate": 0.2020391200402623,
      "loss": 0.2348,
      "num_input_tokens_seen": 29791328,
      "step": 15490
    },
    {
      "epoch": 2.527775511868831,
      "grad_norm": 0.0022349320352077484,
      "learning_rate": 0.2019838696322981,
      "loss": 0.2205,
      "num_input_tokens_seen": 29801472,
      "step": 15495
    },
    {
      "epoch": 2.5285912390896486,
      "grad_norm": 0.0017882795073091984,
      "learning_rate": 0.20192861120776798,
      "loss": 0.2305,
      "num_input_tokens_seen": 29810944,
      "step": 15500
    },
    {
      "epoch": 2.5294069663104657,
      "grad_norm": 0.0039141494780778885,
      "learning_rate": 0.20187334477519345,
      "loss": 0.2394,
      "num_input_tokens_seen": 29821152,
      "step": 15505
    },
    {
      "epoch": 2.5302226935312833,
      "grad_norm": 0.0017564523732289672,
      "learning_rate": 0.20181807034309726,
      "loss": 0.2096,
      "num_input_tokens_seen": 29830880,
      "step": 15510
    },
    {
      "epoch": 2.5310384207521004,
      "grad_norm": 0.001825224608182907,
      "learning_rate": 0.2017627879200034,
      "loss": 0.2062,
      "num_input_tokens_seen": 29840144,
      "step": 15515
    },
    {
      "epoch": 2.531854147972918,
      "grad_norm": 0.0023906398564577103,
      "learning_rate": 0.2017074975144372,
      "loss": 0.2268,
      "num_input_tokens_seen": 29849712,
      "step": 15520
    },
    {
      "epoch": 2.532669875193735,
      "grad_norm": 0.0019331368384882808,
      "learning_rate": 0.20165219913492508,
      "loss": 0.2026,
      "num_input_tokens_seen": 29859520,
      "step": 15525
    },
    {
      "epoch": 2.5334856024145527,
      "grad_norm": 0.001438801409676671,
      "learning_rate": 0.20159689278999468,
      "loss": 0.2089,
      "num_input_tokens_seen": 29870016,
      "step": 15530
    },
    {
      "epoch": 2.53430132963537,
      "grad_norm": 0.0026874924078583717,
      "learning_rate": 0.20154157848817508,
      "loss": 0.2338,
      "num_input_tokens_seen": 29879136,
      "step": 15535
    },
    {
      "epoch": 2.5351170568561874,
      "grad_norm": 0.001351037179119885,
      "learning_rate": 0.20148625623799632,
      "loss": 0.1993,
      "num_input_tokens_seen": 29888928,
      "step": 15540
    },
    {
      "epoch": 2.5359327840770045,
      "grad_norm": 0.0031871898099780083,
      "learning_rate": 0.20143092604798984,
      "loss": 0.211,
      "num_input_tokens_seen": 29898336,
      "step": 15545
    },
    {
      "epoch": 2.536748511297822,
      "grad_norm": 0.0008909789030440152,
      "learning_rate": 0.2013755879266883,
      "loss": 0.2303,
      "num_input_tokens_seen": 29908912,
      "step": 15550
    },
    {
      "epoch": 2.537564238518639,
      "grad_norm": 0.0012774515198543668,
      "learning_rate": 0.20132024188262543,
      "loss": 0.2259,
      "num_input_tokens_seen": 29917584,
      "step": 15555
    },
    {
      "epoch": 2.538379965739457,
      "grad_norm": 0.0011598503915593028,
      "learning_rate": 0.2012648879243363,
      "loss": 0.2305,
      "num_input_tokens_seen": 29925920,
      "step": 15560
    },
    {
      "epoch": 2.539195692960274,
      "grad_norm": 0.0016489699482917786,
      "learning_rate": 0.20120952606035725,
      "loss": 0.2174,
      "num_input_tokens_seen": 29935520,
      "step": 15565
    },
    {
      "epoch": 2.5400114201810915,
      "grad_norm": 0.0015134454006329179,
      "learning_rate": 0.20115415629922576,
      "loss": 0.2094,
      "num_input_tokens_seen": 29946112,
      "step": 15570
    },
    {
      "epoch": 2.5408271474019086,
      "grad_norm": 0.002504665870219469,
      "learning_rate": 0.20109877864948048,
      "loss": 0.2267,
      "num_input_tokens_seen": 29955664,
      "step": 15575
    },
    {
      "epoch": 2.541642874622726,
      "grad_norm": 0.0012667399132624269,
      "learning_rate": 0.20104339311966138,
      "loss": 0.2152,
      "num_input_tokens_seen": 29966224,
      "step": 15580
    },
    {
      "epoch": 2.5424586018435438,
      "grad_norm": 0.001077238586731255,
      "learning_rate": 0.2009879997183097,
      "loss": 0.2289,
      "num_input_tokens_seen": 29976560,
      "step": 15585
    },
    {
      "epoch": 2.543274329064361,
      "grad_norm": 0.0022681839764118195,
      "learning_rate": 0.20093259845396763,
      "loss": 0.1908,
      "num_input_tokens_seen": 29985536,
      "step": 15590
    },
    {
      "epoch": 2.544090056285178,
      "grad_norm": 0.0016950791468843818,
      "learning_rate": 0.20087718933517884,
      "loss": 0.2154,
      "num_input_tokens_seen": 29993808,
      "step": 15595
    },
    {
      "epoch": 2.5449057835059956,
      "grad_norm": 0.0014068526215851307,
      "learning_rate": 0.20082177237048807,
      "loss": 0.2143,
      "num_input_tokens_seen": 30004032,
      "step": 15600
    },
    {
      "epoch": 2.5449057835059956,
      "eval_loss": 0.2237497717142105,
      "eval_runtime": 67.9857,
      "eval_samples_per_second": 40.082,
      "eval_steps_per_second": 20.048,
      "num_input_tokens_seen": 30004032,
      "step": 15600
    },
    {
      "epoch": 2.545721510726813,
      "grad_norm": 0.0027376613579690456,
      "learning_rate": 0.20076634756844133,
      "loss": 0.2543,
      "num_input_tokens_seen": 30012768,
      "step": 15605
    },
    {
      "epoch": 2.5465372379476303,
      "grad_norm": 0.0018913238309323788,
      "learning_rate": 0.20071091493758586,
      "loss": 0.2171,
      "num_input_tokens_seen": 30023216,
      "step": 15610
    },
    {
      "epoch": 2.5473529651684474,
      "grad_norm": 0.0025410158559679985,
      "learning_rate": 0.20065547448647003,
      "loss": 0.2565,
      "num_input_tokens_seen": 30033232,
      "step": 15615
    },
    {
      "epoch": 2.548168692389265,
      "grad_norm": 0.0016417099395766854,
      "learning_rate": 0.20060002622364348,
      "loss": 0.1985,
      "num_input_tokens_seen": 30042224,
      "step": 15620
    },
    {
      "epoch": 2.5489844196100826,
      "grad_norm": 0.0012426257599145174,
      "learning_rate": 0.20054457015765695,
      "loss": 0.1858,
      "num_input_tokens_seen": 30051232,
      "step": 15625
    },
    {
      "epoch": 2.5498001468308997,
      "grad_norm": 0.002617274411022663,
      "learning_rate": 0.20048910629706254,
      "loss": 0.2208,
      "num_input_tokens_seen": 30059520,
      "step": 15630
    },
    {
      "epoch": 2.550615874051717,
      "grad_norm": 0.0015417563263326883,
      "learning_rate": 0.20043363465041347,
      "loss": 0.2339,
      "num_input_tokens_seen": 30067648,
      "step": 15635
    },
    {
      "epoch": 2.5514316012725344,
      "grad_norm": 0.0008653326658532023,
      "learning_rate": 0.2003781552262641,
      "loss": 0.1913,
      "num_input_tokens_seen": 30077760,
      "step": 15640
    },
    {
      "epoch": 2.552247328493352,
      "grad_norm": 0.0017398049822077155,
      "learning_rate": 0.20032266803317014,
      "loss": 0.2114,
      "num_input_tokens_seen": 30087328,
      "step": 15645
    },
    {
      "epoch": 2.553063055714169,
      "grad_norm": 0.002262260764837265,
      "learning_rate": 0.2002671730796884,
      "loss": 0.2582,
      "num_input_tokens_seen": 30096720,
      "step": 15650
    },
    {
      "epoch": 2.5538787829349863,
      "grad_norm": 0.0012426218017935753,
      "learning_rate": 0.20021167037437684,
      "loss": 0.2371,
      "num_input_tokens_seen": 30106864,
      "step": 15655
    },
    {
      "epoch": 2.554694510155804,
      "grad_norm": 0.00355734140612185,
      "learning_rate": 0.20015615992579472,
      "loss": 0.2372,
      "num_input_tokens_seen": 30116464,
      "step": 15660
    },
    {
      "epoch": 2.5555102373766214,
      "grad_norm": 0.0015998557209968567,
      "learning_rate": 0.20010064174250244,
      "loss": 0.2208,
      "num_input_tokens_seen": 30127328,
      "step": 15665
    },
    {
      "epoch": 2.5563259645974385,
      "grad_norm": 0.001337096793577075,
      "learning_rate": 0.2000451158330616,
      "loss": 0.2331,
      "num_input_tokens_seen": 30136512,
      "step": 15670
    },
    {
      "epoch": 2.557141691818256,
      "grad_norm": 0.0012996571604162455,
      "learning_rate": 0.199989582206035,
      "loss": 0.22,
      "num_input_tokens_seen": 30145360,
      "step": 15675
    },
    {
      "epoch": 2.5579574190390733,
      "grad_norm": 0.0009348857565782964,
      "learning_rate": 0.1999340408699866,
      "loss": 0.2435,
      "num_input_tokens_seen": 30155120,
      "step": 15680
    },
    {
      "epoch": 2.558773146259891,
      "grad_norm": 0.0010707969777286053,
      "learning_rate": 0.19987849183348155,
      "loss": 0.2172,
      "num_input_tokens_seen": 30163712,
      "step": 15685
    },
    {
      "epoch": 2.559588873480708,
      "grad_norm": 0.0009421168942935765,
      "learning_rate": 0.19982293510508628,
      "loss": 0.213,
      "num_input_tokens_seen": 30173936,
      "step": 15690
    },
    {
      "epoch": 2.5604046007015255,
      "grad_norm": 0.00115866563282907,
      "learning_rate": 0.19976737069336833,
      "loss": 0.2282,
      "num_input_tokens_seen": 30183376,
      "step": 15695
    },
    {
      "epoch": 2.5612203279223427,
      "grad_norm": 0.0010530363069847226,
      "learning_rate": 0.1997117986068964,
      "loss": 0.2329,
      "num_input_tokens_seen": 30192768,
      "step": 15700
    },
    {
      "epoch": 2.5620360551431602,
      "grad_norm": 0.002256223699077964,
      "learning_rate": 0.19965621885424037,
      "loss": 0.1941,
      "num_input_tokens_seen": 30202448,
      "step": 15705
    },
    {
      "epoch": 2.5628517823639774,
      "grad_norm": 0.0022258807439357042,
      "learning_rate": 0.19960063144397142,
      "loss": 0.2003,
      "num_input_tokens_seen": 30213664,
      "step": 15710
    },
    {
      "epoch": 2.563667509584795,
      "grad_norm": 0.0019627127330750227,
      "learning_rate": 0.19954503638466176,
      "loss": 0.2197,
      "num_input_tokens_seen": 30223712,
      "step": 15715
    },
    {
      "epoch": 2.564483236805612,
      "grad_norm": 0.0029075979255139828,
      "learning_rate": 0.1994894336848848,
      "loss": 0.2416,
      "num_input_tokens_seen": 30232640,
      "step": 15720
    },
    {
      "epoch": 2.5652989640264297,
      "grad_norm": 0.0019561343360692263,
      "learning_rate": 0.1994338233532153,
      "loss": 0.2685,
      "num_input_tokens_seen": 30242048,
      "step": 15725
    },
    {
      "epoch": 2.566114691247247,
      "grad_norm": 0.002084343694150448,
      "learning_rate": 0.19937820539822904,
      "loss": 0.225,
      "num_input_tokens_seen": 30251008,
      "step": 15730
    },
    {
      "epoch": 2.5669304184680644,
      "grad_norm": 0.0022956104949116707,
      "learning_rate": 0.199322579828503,
      "loss": 0.1933,
      "num_input_tokens_seen": 30260160,
      "step": 15735
    },
    {
      "epoch": 2.5677461456888815,
      "grad_norm": 0.001491433009505272,
      "learning_rate": 0.19926694665261527,
      "loss": 0.2263,
      "num_input_tokens_seen": 30268144,
      "step": 15740
    },
    {
      "epoch": 2.568561872909699,
      "grad_norm": 0.0019516702741384506,
      "learning_rate": 0.19921130587914526,
      "loss": 0.2301,
      "num_input_tokens_seen": 30277392,
      "step": 15745
    },
    {
      "epoch": 2.569377600130516,
      "grad_norm": 0.0016313963569700718,
      "learning_rate": 0.19915565751667344,
      "loss": 0.2144,
      "num_input_tokens_seen": 30286848,
      "step": 15750
    },
    {
      "epoch": 2.5701933273513338,
      "grad_norm": 0.0030580649618059397,
      "learning_rate": 0.19910000157378152,
      "loss": 0.2071,
      "num_input_tokens_seen": 30295616,
      "step": 15755
    },
    {
      "epoch": 2.5710090545721513,
      "grad_norm": 0.0015175885055214167,
      "learning_rate": 0.1990443380590523,
      "loss": 0.2248,
      "num_input_tokens_seen": 30305456,
      "step": 15760
    },
    {
      "epoch": 2.5718247817929685,
      "grad_norm": 0.002426989609375596,
      "learning_rate": 0.19898866698106984,
      "loss": 0.2266,
      "num_input_tokens_seen": 30315504,
      "step": 15765
    },
    {
      "epoch": 2.5726405090137856,
      "grad_norm": 0.0014837330672889948,
      "learning_rate": 0.19893298834841933,
      "loss": 0.2525,
      "num_input_tokens_seen": 30324720,
      "step": 15770
    },
    {
      "epoch": 2.573456236234603,
      "grad_norm": 0.0012985612265765667,
      "learning_rate": 0.19887730216968705,
      "loss": 0.1909,
      "num_input_tokens_seen": 30334160,
      "step": 15775
    },
    {
      "epoch": 2.5742719634554208,
      "grad_norm": 0.0016186272259801626,
      "learning_rate": 0.19882160845346053,
      "loss": 0.2178,
      "num_input_tokens_seen": 30344848,
      "step": 15780
    },
    {
      "epoch": 2.575087690676238,
      "grad_norm": 0.001961708767339587,
      "learning_rate": 0.1987659072083285,
      "loss": 0.227,
      "num_input_tokens_seen": 30356032,
      "step": 15785
    },
    {
      "epoch": 2.575903417897055,
      "grad_norm": 0.0014041742542758584,
      "learning_rate": 0.1987101984428807,
      "loss": 0.2134,
      "num_input_tokens_seen": 30366736,
      "step": 15790
    },
    {
      "epoch": 2.5767191451178726,
      "grad_norm": 0.0018375064246356487,
      "learning_rate": 0.19865448216570822,
      "loss": 0.2131,
      "num_input_tokens_seen": 30375952,
      "step": 15795
    },
    {
      "epoch": 2.57753487233869,
      "grad_norm": 0.0012718678917735815,
      "learning_rate": 0.19859875838540317,
      "loss": 0.2134,
      "num_input_tokens_seen": 30386752,
      "step": 15800
    },
    {
      "epoch": 2.57753487233869,
      "eval_loss": 0.2149352729320526,
      "eval_runtime": 68.0054,
      "eval_samples_per_second": 40.07,
      "eval_steps_per_second": 20.043,
      "num_input_tokens_seen": 30386752,
      "step": 15800
    },
    {
      "epoch": 2.5783505995595073,
      "grad_norm": 0.0029731064569205046,
      "learning_rate": 0.1985430271105588,
      "loss": 0.2463,
      "num_input_tokens_seen": 30395568,
      "step": 15805
    },
    {
      "epoch": 2.5791663267803244,
      "grad_norm": 0.0013004217762500048,
      "learning_rate": 0.19848728834976961,
      "loss": 0.2389,
      "num_input_tokens_seen": 30405344,
      "step": 15810
    },
    {
      "epoch": 2.579982054001142,
      "grad_norm": 0.0015243194065988064,
      "learning_rate": 0.19843154211163128,
      "loss": 0.2229,
      "num_input_tokens_seen": 30414576,
      "step": 15815
    },
    {
      "epoch": 2.5807977812219596,
      "grad_norm": 0.0013050712877884507,
      "learning_rate": 0.1983757884047405,
      "loss": 0.1692,
      "num_input_tokens_seen": 30424080,
      "step": 15820
    },
    {
      "epoch": 2.5816135084427767,
      "grad_norm": 0.0021705778781324625,
      "learning_rate": 0.1983200272376952,
      "loss": 0.2271,
      "num_input_tokens_seen": 30433680,
      "step": 15825
    },
    {
      "epoch": 2.582429235663594,
      "grad_norm": 0.002492618979886174,
      "learning_rate": 0.1982642586190945,
      "loss": 0.2324,
      "num_input_tokens_seen": 30442704,
      "step": 15830
    },
    {
      "epoch": 2.5832449628844114,
      "grad_norm": 0.0025220559909939766,
      "learning_rate": 0.1982084825575386,
      "loss": 0.2175,
      "num_input_tokens_seen": 30452656,
      "step": 15835
    },
    {
      "epoch": 2.584060690105229,
      "grad_norm": 0.0016430397517979145,
      "learning_rate": 0.19815269906162883,
      "loss": 0.2072,
      "num_input_tokens_seen": 30462448,
      "step": 15840
    },
    {
      "epoch": 2.584876417326046,
      "grad_norm": 0.0015117151197046041,
      "learning_rate": 0.19809690813996775,
      "loss": 0.1903,
      "num_input_tokens_seen": 30472192,
      "step": 15845
    },
    {
      "epoch": 2.5856921445468637,
      "grad_norm": 0.0013689221814274788,
      "learning_rate": 0.19804110980115905,
      "loss": 0.1946,
      "num_input_tokens_seen": 30481824,
      "step": 15850
    },
    {
      "epoch": 2.586507871767681,
      "grad_norm": 0.002836219733580947,
      "learning_rate": 0.19798530405380746,
      "loss": 0.1814,
      "num_input_tokens_seen": 30491248,
      "step": 15855
    },
    {
      "epoch": 2.5873235989884984,
      "grad_norm": 0.0018414699006825686,
      "learning_rate": 0.19792949090651893,
      "loss": 0.2208,
      "num_input_tokens_seen": 30500448,
      "step": 15860
    },
    {
      "epoch": 2.5881393262093155,
      "grad_norm": 0.015239886939525604,
      "learning_rate": 0.19787367036790066,
      "loss": 0.2103,
      "num_input_tokens_seen": 30509936,
      "step": 15865
    },
    {
      "epoch": 2.588955053430133,
      "grad_norm": 0.0014677244471386075,
      "learning_rate": 0.19781784244656075,
      "loss": 0.1734,
      "num_input_tokens_seen": 30519440,
      "step": 15870
    },
    {
      "epoch": 2.5897707806509502,
      "grad_norm": 0.0013508659321814775,
      "learning_rate": 0.19776200715110864,
      "loss": 0.1743,
      "num_input_tokens_seen": 30529040,
      "step": 15875
    },
    {
      "epoch": 2.590586507871768,
      "grad_norm": 0.0039116861298680305,
      "learning_rate": 0.1977061644901548,
      "loss": 0.2088,
      "num_input_tokens_seen": 30539344,
      "step": 15880
    },
    {
      "epoch": 2.591402235092585,
      "grad_norm": 0.003258773824200034,
      "learning_rate": 0.1976503144723109,
      "loss": 0.2174,
      "num_input_tokens_seen": 30547440,
      "step": 15885
    },
    {
      "epoch": 2.5922179623134025,
      "grad_norm": 0.0022673436906188726,
      "learning_rate": 0.19759445710618967,
      "loss": 0.2001,
      "num_input_tokens_seen": 30556288,
      "step": 15890
    },
    {
      "epoch": 2.5930336895342196,
      "grad_norm": 0.0017375163733959198,
      "learning_rate": 0.19753859240040508,
      "loss": 0.2297,
      "num_input_tokens_seen": 30565840,
      "step": 15895
    },
    {
      "epoch": 2.593849416755037,
      "grad_norm": 0.002176872221753001,
      "learning_rate": 0.1974827203635721,
      "loss": 0.258,
      "num_input_tokens_seen": 30575216,
      "step": 15900
    },
    {
      "epoch": 2.5946651439758543,
      "grad_norm": 0.0018550172680988908,
      "learning_rate": 0.19742684100430694,
      "loss": 0.2463,
      "num_input_tokens_seen": 30585424,
      "step": 15905
    },
    {
      "epoch": 2.595480871196672,
      "grad_norm": 0.0011709029786288738,
      "learning_rate": 0.19737095433122692,
      "loss": 0.2191,
      "num_input_tokens_seen": 30593824,
      "step": 15910
    },
    {
      "epoch": 2.596296598417489,
      "grad_norm": 0.0025421935133635998,
      "learning_rate": 0.19731506035295046,
      "loss": 0.2002,
      "num_input_tokens_seen": 30604176,
      "step": 15915
    },
    {
      "epoch": 2.5971123256383066,
      "grad_norm": 0.002133615082129836,
      "learning_rate": 0.19725915907809702,
      "loss": 0.2381,
      "num_input_tokens_seen": 30613904,
      "step": 15920
    },
    {
      "epoch": 2.5979280528591238,
      "grad_norm": 0.0012623764341697097,
      "learning_rate": 0.1972032505152874,
      "loss": 0.2353,
      "num_input_tokens_seen": 30624080,
      "step": 15925
    },
    {
      "epoch": 2.5987437800799413,
      "grad_norm": 0.0015752629842609167,
      "learning_rate": 0.19714733467314338,
      "loss": 0.2109,
      "num_input_tokens_seen": 30633952,
      "step": 15930
    },
    {
      "epoch": 2.5995595073007585,
      "grad_norm": 0.00128090544603765,
      "learning_rate": 0.19709141156028784,
      "loss": 0.1987,
      "num_input_tokens_seen": 30643872,
      "step": 15935
    },
    {
      "epoch": 2.600375234521576,
      "grad_norm": 0.0013335243565961719,
      "learning_rate": 0.1970354811853448,
      "loss": 0.2491,
      "num_input_tokens_seen": 30655024,
      "step": 15940
    },
    {
      "epoch": 2.601190961742393,
      "grad_norm": 0.0011131390929222107,
      "learning_rate": 0.19697954355693953,
      "loss": 0.2293,
      "num_input_tokens_seen": 30664448,
      "step": 15945
    },
    {
      "epoch": 2.6020066889632107,
      "grad_norm": 0.0026444869581609964,
      "learning_rate": 0.19692359868369827,
      "loss": 0.1885,
      "num_input_tokens_seen": 30674976,
      "step": 15950
    },
    {
      "epoch": 2.6028224161840283,
      "grad_norm": 0.002746515441685915,
      "learning_rate": 0.1968676465742484,
      "loss": 0.1922,
      "num_input_tokens_seen": 30684400,
      "step": 15955
    },
    {
      "epoch": 2.6036381434048455,
      "grad_norm": 0.0016234742943197489,
      "learning_rate": 0.19681168723721845,
      "loss": 0.216,
      "num_input_tokens_seen": 30695216,
      "step": 15960
    },
    {
      "epoch": 2.6044538706256626,
      "grad_norm": 0.0013635570649057627,
      "learning_rate": 0.19675572068123803,
      "loss": 0.2557,
      "num_input_tokens_seen": 30705520,
      "step": 15965
    },
    {
      "epoch": 2.60526959784648,
      "grad_norm": 0.0026287375949323177,
      "learning_rate": 0.19669974691493794,
      "loss": 0.2298,
      "num_input_tokens_seen": 30716304,
      "step": 15970
    },
    {
      "epoch": 2.6060853250672977,
      "grad_norm": 0.0014553345972672105,
      "learning_rate": 0.19664376594695002,
      "loss": 0.227,
      "num_input_tokens_seen": 30726000,
      "step": 15975
    },
    {
      "epoch": 2.606901052288115,
      "grad_norm": 0.0012213204754516482,
      "learning_rate": 0.19658777778590722,
      "loss": 0.2488,
      "num_input_tokens_seen": 30735712,
      "step": 15980
    },
    {
      "epoch": 2.607716779508932,
      "grad_norm": 0.0028174116741865873,
      "learning_rate": 0.19653178244044364,
      "loss": 0.2134,
      "num_input_tokens_seen": 30746224,
      "step": 15985
    },
    {
      "epoch": 2.6085325067297496,
      "grad_norm": 0.001493224292062223,
      "learning_rate": 0.19647577991919443,
      "loss": 0.2235,
      "num_input_tokens_seen": 30755248,
      "step": 15990
    },
    {
      "epoch": 2.609348233950567,
      "grad_norm": 0.0013514773454517126,
      "learning_rate": 0.1964197702307959,
      "loss": 0.2339,
      "num_input_tokens_seen": 30765472,
      "step": 15995
    },
    {
      "epoch": 2.6101639611713843,
      "grad_norm": 0.002522384747862816,
      "learning_rate": 0.19636375338388545,
      "loss": 0.2007,
      "num_input_tokens_seen": 30774224,
      "step": 16000
    },
    {
      "epoch": 2.6101639611713843,
      "eval_loss": 0.2146008163690567,
      "eval_runtime": 68.0575,
      "eval_samples_per_second": 40.04,
      "eval_steps_per_second": 20.027,
      "num_input_tokens_seen": 30774224,
      "step": 16000
    },
    {
      "epoch": 2.6109796883922014,
      "grad_norm": 0.0013140550581738353,
      "learning_rate": 0.1963077293871016,
      "loss": 0.2205,
      "num_input_tokens_seen": 30783648,
      "step": 16005
    },
    {
      "epoch": 2.611795415613019,
      "grad_norm": 0.0009910432854667306,
      "learning_rate": 0.19625169824908395,
      "loss": 0.2202,
      "num_input_tokens_seen": 30793296,
      "step": 16010
    },
    {
      "epoch": 2.6126111428338366,
      "grad_norm": 0.00309326546266675,
      "learning_rate": 0.19619565997847319,
      "loss": 0.2401,
      "num_input_tokens_seen": 30803392,
      "step": 16015
    },
    {
      "epoch": 2.6134268700546537,
      "grad_norm": 0.001308513106778264,
      "learning_rate": 0.19613961458391113,
      "loss": 0.1995,
      "num_input_tokens_seen": 30814528,
      "step": 16020
    },
    {
      "epoch": 2.614242597275471,
      "grad_norm": 0.0032572238706052303,
      "learning_rate": 0.19608356207404065,
      "loss": 0.2329,
      "num_input_tokens_seen": 30824832,
      "step": 16025
    },
    {
      "epoch": 2.6150583244962884,
      "grad_norm": 0.002769637852907181,
      "learning_rate": 0.1960275024575058,
      "loss": 0.2115,
      "num_input_tokens_seen": 30834064,
      "step": 16030
    },
    {
      "epoch": 2.615874051717106,
      "grad_norm": 0.0023053165059536695,
      "learning_rate": 0.19597143574295164,
      "loss": 0.2565,
      "num_input_tokens_seen": 30844048,
      "step": 16035
    },
    {
      "epoch": 2.616689778937923,
      "grad_norm": 0.0014554557856172323,
      "learning_rate": 0.1959153619390244,
      "loss": 0.2166,
      "num_input_tokens_seen": 30854256,
      "step": 16040
    },
    {
      "epoch": 2.6175055061587407,
      "grad_norm": 0.0015379722462967038,
      "learning_rate": 0.1958592810543713,
      "loss": 0.2171,
      "num_input_tokens_seen": 30862752,
      "step": 16045
    },
    {
      "epoch": 2.618321233379558,
      "grad_norm": 0.0011374774621799588,
      "learning_rate": 0.19580319309764077,
      "loss": 0.2019,
      "num_input_tokens_seen": 30871536,
      "step": 16050
    },
    {
      "epoch": 2.6191369606003754,
      "grad_norm": 0.0015881088329479098,
      "learning_rate": 0.1957470980774823,
      "loss": 0.2279,
      "num_input_tokens_seen": 30881200,
      "step": 16055
    },
    {
      "epoch": 2.6199526878211925,
      "grad_norm": 0.000891564879566431,
      "learning_rate": 0.19569099600254639,
      "loss": 0.1993,
      "num_input_tokens_seen": 30889488,
      "step": 16060
    },
    {
      "epoch": 2.62076841504201,
      "grad_norm": 0.0019485808443278074,
      "learning_rate": 0.1956348868814847,
      "loss": 0.1954,
      "num_input_tokens_seen": 30898416,
      "step": 16065
    },
    {
      "epoch": 2.621584142262827,
      "grad_norm": 0.0018783595878630877,
      "learning_rate": 0.19557877072295,
      "loss": 0.1943,
      "num_input_tokens_seen": 30908176,
      "step": 16070
    },
    {
      "epoch": 2.622399869483645,
      "grad_norm": 0.0017469668528065085,
      "learning_rate": 0.19552264753559603,
      "loss": 0.1675,
      "num_input_tokens_seen": 30917920,
      "step": 16075
    },
    {
      "epoch": 2.623215596704462,
      "grad_norm": 0.0030839673709124327,
      "learning_rate": 0.19546651732807774,
      "loss": 0.272,
      "num_input_tokens_seen": 30927856,
      "step": 16080
    },
    {
      "epoch": 2.6240313239252795,
      "grad_norm": 0.003418233245611191,
      "learning_rate": 0.19541038010905112,
      "loss": 0.2234,
      "num_input_tokens_seen": 30937488,
      "step": 16085
    },
    {
      "epoch": 2.6248470511460966,
      "grad_norm": 0.001350239384919405,
      "learning_rate": 0.19535423588717324,
      "loss": 0.2192,
      "num_input_tokens_seen": 30947712,
      "step": 16090
    },
    {
      "epoch": 2.625662778366914,
      "grad_norm": 0.002467974554747343,
      "learning_rate": 0.19529808467110224,
      "loss": 0.1971,
      "num_input_tokens_seen": 30957008,
      "step": 16095
    },
    {
      "epoch": 2.6264785055877313,
      "grad_norm": 0.004199891816824675,
      "learning_rate": 0.19524192646949734,
      "loss": 0.2878,
      "num_input_tokens_seen": 30967408,
      "step": 16100
    },
    {
      "epoch": 2.627294232808549,
      "grad_norm": 0.0012990537798032165,
      "learning_rate": 0.19518576129101878,
      "loss": 0.2334,
      "num_input_tokens_seen": 30976256,
      "step": 16105
    },
    {
      "epoch": 2.628109960029366,
      "grad_norm": 0.003214568132534623,
      "learning_rate": 0.19512958914432804,
      "loss": 0.2113,
      "num_input_tokens_seen": 30984976,
      "step": 16110
    },
    {
      "epoch": 2.6289256872501836,
      "grad_norm": 0.0007988462457433343,
      "learning_rate": 0.1950734100380875,
      "loss": 0.1802,
      "num_input_tokens_seen": 30993232,
      "step": 16115
    },
    {
      "epoch": 2.6297414144710007,
      "grad_norm": 0.0013003014028072357,
      "learning_rate": 0.19501722398096066,
      "loss": 0.2245,
      "num_input_tokens_seen": 31003632,
      "step": 16120
    },
    {
      "epoch": 2.6305571416918183,
      "grad_norm": 0.0018916501430794597,
      "learning_rate": 0.1949610309816122,
      "loss": 0.2487,
      "num_input_tokens_seen": 31013664,
      "step": 16125
    },
    {
      "epoch": 2.631372868912636,
      "grad_norm": 0.0014455352211371064,
      "learning_rate": 0.1949048310487078,
      "loss": 0.2241,
      "num_input_tokens_seen": 31024400,
      "step": 16130
    },
    {
      "epoch": 2.632188596133453,
      "grad_norm": 0.0009079247247427702,
      "learning_rate": 0.19484862419091406,
      "loss": 0.2426,
      "num_input_tokens_seen": 31033984,
      "step": 16135
    },
    {
      "epoch": 2.63300432335427,
      "grad_norm": 0.002070232992991805,
      "learning_rate": 0.19479241041689893,
      "loss": 0.2171,
      "num_input_tokens_seen": 31042848,
      "step": 16140
    },
    {
      "epoch": 2.6338200505750877,
      "grad_norm": 0.0012435808312147856,
      "learning_rate": 0.19473618973533116,
      "loss": 0.2127,
      "num_input_tokens_seen": 31054128,
      "step": 16145
    },
    {
      "epoch": 2.6346357777959053,
      "grad_norm": 0.0019414426060393453,
      "learning_rate": 0.19467996215488076,
      "loss": 0.2271,
      "num_input_tokens_seen": 31064128,
      "step": 16150
    },
    {
      "epoch": 2.6354515050167224,
      "grad_norm": 0.0017472000326961279,
      "learning_rate": 0.1946237276842187,
      "loss": 0.2291,
      "num_input_tokens_seen": 31074240,
      "step": 16155
    },
    {
      "epoch": 2.6362672322375396,
      "grad_norm": 0.0032992043998092413,
      "learning_rate": 0.19456748633201712,
      "loss": 0.2501,
      "num_input_tokens_seen": 31084000,
      "step": 16160
    },
    {
      "epoch": 2.637082959458357,
      "grad_norm": 0.004029198549687862,
      "learning_rate": 0.194511238106949,
      "loss": 0.2211,
      "num_input_tokens_seen": 31093552,
      "step": 16165
    },
    {
      "epoch": 2.6378986866791747,
      "grad_norm": 0.002335679018869996,
      "learning_rate": 0.19445498301768863,
      "loss": 0.1963,
      "num_input_tokens_seen": 31103360,
      "step": 16170
    },
    {
      "epoch": 2.638714413899992,
      "grad_norm": 0.0014392270240932703,
      "learning_rate": 0.19439872107291126,
      "loss": 0.1967,
      "num_input_tokens_seen": 31113136,
      "step": 16175
    },
    {
      "epoch": 2.639530141120809,
      "grad_norm": 0.001048493548296392,
      "learning_rate": 0.1943424522812931,
      "loss": 0.2085,
      "num_input_tokens_seen": 31123280,
      "step": 16180
    },
    {
      "epoch": 2.6403458683416265,
      "grad_norm": 0.003224955638870597,
      "learning_rate": 0.19428617665151157,
      "loss": 0.2689,
      "num_input_tokens_seen": 31133056,
      "step": 16185
    },
    {
      "epoch": 2.641161595562444,
      "grad_norm": 0.001778500503860414,
      "learning_rate": 0.19422989419224507,
      "loss": 0.2004,
      "num_input_tokens_seen": 31143168,
      "step": 16190
    },
    {
      "epoch": 2.6419773227832613,
      "grad_norm": 0.0009610646520741284,
      "learning_rate": 0.19417360491217303,
      "loss": 0.2042,
      "num_input_tokens_seen": 31153328,
      "step": 16195
    },
    {
      "epoch": 2.6427930500040784,
      "grad_norm": 0.002945119049400091,
      "learning_rate": 0.19411730881997605,
      "loss": 0.2347,
      "num_input_tokens_seen": 31164304,
      "step": 16200
    },
    {
      "epoch": 2.6427930500040784,
      "eval_loss": 0.2233179807662964,
      "eval_runtime": 68.007,
      "eval_samples_per_second": 40.069,
      "eval_steps_per_second": 20.042,
      "num_input_tokens_seen": 31164304,
      "step": 16200
    },
    {
      "epoch": 2.643608777224896,
      "grad_norm": 0.004659686703234911,
      "learning_rate": 0.1940610059243356,
      "loss": 0.2353,
      "num_input_tokens_seen": 31174192,
      "step": 16205
    },
    {
      "epoch": 2.6444245044457135,
      "grad_norm": 0.0011662262259051204,
      "learning_rate": 0.19400469623393435,
      "loss": 0.1929,
      "num_input_tokens_seen": 31184064,
      "step": 16210
    },
    {
      "epoch": 2.6452402316665307,
      "grad_norm": 0.0010684190783649683,
      "learning_rate": 0.1939483797574559,
      "loss": 0.2006,
      "num_input_tokens_seen": 31193760,
      "step": 16215
    },
    {
      "epoch": 2.6460559588873482,
      "grad_norm": 0.0032252296805381775,
      "learning_rate": 0.19389205650358504,
      "loss": 0.2289,
      "num_input_tokens_seen": 31203232,
      "step": 16220
    },
    {
      "epoch": 2.6468716861081654,
      "grad_norm": 0.0016875730361789465,
      "learning_rate": 0.19383572648100747,
      "loss": 0.2241,
      "num_input_tokens_seen": 31213088,
      "step": 16225
    },
    {
      "epoch": 2.647687413328983,
      "grad_norm": 0.0014784393133595586,
      "learning_rate": 0.19377938969841,
      "loss": 0.2001,
      "num_input_tokens_seen": 31222496,
      "step": 16230
    },
    {
      "epoch": 2.6485031405498,
      "grad_norm": 0.0011400411603972316,
      "learning_rate": 0.1937230461644805,
      "loss": 0.2343,
      "num_input_tokens_seen": 31231584,
      "step": 16235
    },
    {
      "epoch": 2.6493188677706176,
      "grad_norm": 0.0017621275037527084,
      "learning_rate": 0.19366669588790777,
      "loss": 0.2403,
      "num_input_tokens_seen": 31240416,
      "step": 16240
    },
    {
      "epoch": 2.6501345949914348,
      "grad_norm": 0.001234759809449315,
      "learning_rate": 0.19361033887738185,
      "loss": 0.2312,
      "num_input_tokens_seen": 31250608,
      "step": 16245
    },
    {
      "epoch": 2.6509503222122524,
      "grad_norm": 0.0024107457138597965,
      "learning_rate": 0.19355397514159361,
      "loss": 0.2051,
      "num_input_tokens_seen": 31260016,
      "step": 16250
    },
    {
      "epoch": 2.6517660494330695,
      "grad_norm": 0.003629158716648817,
      "learning_rate": 0.19349760468923508,
      "loss": 0.2529,
      "num_input_tokens_seen": 31270576,
      "step": 16255
    },
    {
      "epoch": 2.652581776653887,
      "grad_norm": 0.008504754863679409,
      "learning_rate": 0.19344122752899925,
      "loss": 0.272,
      "num_input_tokens_seen": 31279200,
      "step": 16260
    },
    {
      "epoch": 2.653397503874704,
      "grad_norm": 0.003955257590860128,
      "learning_rate": 0.1933848436695802,
      "loss": 0.2656,
      "num_input_tokens_seen": 31289232,
      "step": 16265
    },
    {
      "epoch": 2.6542132310955218,
      "grad_norm": 0.001463937689550221,
      "learning_rate": 0.1933284531196731,
      "loss": 0.2177,
      "num_input_tokens_seen": 31299328,
      "step": 16270
    },
    {
      "epoch": 2.655028958316339,
      "grad_norm": 0.0009589198161847889,
      "learning_rate": 0.19327205588797403,
      "loss": 0.1931,
      "num_input_tokens_seen": 31309280,
      "step": 16275
    },
    {
      "epoch": 2.6558446855371565,
      "grad_norm": 0.002553339349105954,
      "learning_rate": 0.19321565198318014,
      "loss": 0.2373,
      "num_input_tokens_seen": 31318784,
      "step": 16280
    },
    {
      "epoch": 2.6566604127579736,
      "grad_norm": 0.0013882851926609874,
      "learning_rate": 0.1931592414139896,
      "loss": 0.195,
      "num_input_tokens_seen": 31328608,
      "step": 16285
    },
    {
      "epoch": 2.657476139978791,
      "grad_norm": 0.0017179888673126698,
      "learning_rate": 0.19310282418910169,
      "loss": 0.2307,
      "num_input_tokens_seen": 31338192,
      "step": 16290
    },
    {
      "epoch": 2.6582918671996083,
      "grad_norm": 0.002085472224280238,
      "learning_rate": 0.1930464003172166,
      "loss": 0.2224,
      "num_input_tokens_seen": 31348416,
      "step": 16295
    },
    {
      "epoch": 2.659107594420426,
      "grad_norm": 0.002935906872153282,
      "learning_rate": 0.19298996980703567,
      "loss": 0.2304,
      "num_input_tokens_seen": 31358128,
      "step": 16300
    },
    {
      "epoch": 2.659923321641243,
      "grad_norm": 0.002016285667195916,
      "learning_rate": 0.19293353266726113,
      "loss": 0.2323,
      "num_input_tokens_seen": 31366144,
      "step": 16305
    },
    {
      "epoch": 2.6607390488620606,
      "grad_norm": 0.002392769791185856,
      "learning_rate": 0.19287708890659633,
      "loss": 0.2423,
      "num_input_tokens_seen": 31375456,
      "step": 16310
    },
    {
      "epoch": 2.6615547760828777,
      "grad_norm": 0.0022560067009180784,
      "learning_rate": 0.19282063853374556,
      "loss": 0.2197,
      "num_input_tokens_seen": 31384432,
      "step": 16315
    },
    {
      "epoch": 2.6623705033036953,
      "grad_norm": 0.002060480648651719,
      "learning_rate": 0.19276418155741423,
      "loss": 0.1997,
      "num_input_tokens_seen": 31395216,
      "step": 16320
    },
    {
      "epoch": 2.663186230524513,
      "grad_norm": 0.0017401641234755516,
      "learning_rate": 0.19270771798630867,
      "loss": 0.222,
      "num_input_tokens_seen": 31403744,
      "step": 16325
    },
    {
      "epoch": 2.66400195774533,
      "grad_norm": 0.000839407613966614,
      "learning_rate": 0.1926512478291363,
      "loss": 0.2297,
      "num_input_tokens_seen": 31414032,
      "step": 16330
    },
    {
      "epoch": 2.664817684966147,
      "grad_norm": 0.0012854994274675846,
      "learning_rate": 0.19259477109460557,
      "loss": 0.2713,
      "num_input_tokens_seen": 31422464,
      "step": 16335
    },
    {
      "epoch": 2.6656334121869647,
      "grad_norm": 0.001782553386874497,
      "learning_rate": 0.19253828779142584,
      "loss": 0.1937,
      "num_input_tokens_seen": 31431984,
      "step": 16340
    },
    {
      "epoch": 2.6664491394077823,
      "grad_norm": 0.001683527952991426,
      "learning_rate": 0.19248179792830755,
      "loss": 0.2216,
      "num_input_tokens_seen": 31442480,
      "step": 16345
    },
    {
      "epoch": 2.6672648666285994,
      "grad_norm": 0.0030651630368083715,
      "learning_rate": 0.19242530151396217,
      "loss": 0.219,
      "num_input_tokens_seen": 31453088,
      "step": 16350
    },
    {
      "epoch": 2.6680805938494165,
      "grad_norm": 0.0012733993353322148,
      "learning_rate": 0.19236879855710215,
      "loss": 0.2311,
      "num_input_tokens_seen": 31461728,
      "step": 16355
    },
    {
      "epoch": 2.668896321070234,
      "grad_norm": 0.0014802207006141543,
      "learning_rate": 0.19231228906644096,
      "loss": 0.2331,
      "num_input_tokens_seen": 31471312,
      "step": 16360
    },
    {
      "epoch": 2.6697120482910517,
      "grad_norm": 0.002696947893127799,
      "learning_rate": 0.19225577305069302,
      "loss": 0.2454,
      "num_input_tokens_seen": 31482096,
      "step": 16365
    },
    {
      "epoch": 2.670527775511869,
      "grad_norm": 0.002205491531640291,
      "learning_rate": 0.1921992505185739,
      "loss": 0.2337,
      "num_input_tokens_seen": 31491632,
      "step": 16370
    },
    {
      "epoch": 2.671343502732686,
      "grad_norm": 0.002048164140433073,
      "learning_rate": 0.19214272147880004,
      "loss": 0.2286,
      "num_input_tokens_seen": 31500608,
      "step": 16375
    },
    {
      "epoch": 2.6721592299535035,
      "grad_norm": 0.0011319813784211874,
      "learning_rate": 0.19208618594008892,
      "loss": 0.2255,
      "num_input_tokens_seen": 31510192,
      "step": 16380
    },
    {
      "epoch": 2.672974957174321,
      "grad_norm": 0.002243302995339036,
      "learning_rate": 0.19202964391115904,
      "loss": 0.2393,
      "num_input_tokens_seen": 31520352,
      "step": 16385
    },
    {
      "epoch": 2.6737906843951382,
      "grad_norm": 0.0021424186415970325,
      "learning_rate": 0.1919730954007299,
      "loss": 0.2535,
      "num_input_tokens_seen": 31530560,
      "step": 16390
    },
    {
      "epoch": 2.6746064116159554,
      "grad_norm": 0.0017501877155154943,
      "learning_rate": 0.19191654041752199,
      "loss": 0.2019,
      "num_input_tokens_seen": 31540560,
      "step": 16395
    },
    {
      "epoch": 2.675422138836773,
      "grad_norm": 0.001361907459795475,
      "learning_rate": 0.19185997897025678,
      "loss": 0.1845,
      "num_input_tokens_seen": 31548832,
      "step": 16400
    },
    {
      "epoch": 2.675422138836773,
      "eval_loss": 0.21533682942390442,
      "eval_runtime": 68.0812,
      "eval_samples_per_second": 40.026,
      "eval_steps_per_second": 20.02,
      "num_input_tokens_seen": 31548832,
      "step": 16400
    },
    {
      "epoch": 2.6762378660575905,
      "grad_norm": 0.00241487892344594,
      "learning_rate": 0.19180341106765672,
      "loss": 0.2106,
      "num_input_tokens_seen": 31559040,
      "step": 16405
    },
    {
      "epoch": 2.6770535932784076,
      "grad_norm": 0.00181131053250283,
      "learning_rate": 0.19174683671844536,
      "loss": 0.2529,
      "num_input_tokens_seen": 31569376,
      "step": 16410
    },
    {
      "epoch": 2.677869320499225,
      "grad_norm": 0.002220016671344638,
      "learning_rate": 0.19169025593134717,
      "loss": 0.2447,
      "num_input_tokens_seen": 31578896,
      "step": 16415
    },
    {
      "epoch": 2.6786850477200423,
      "grad_norm": 0.0012551560066640377,
      "learning_rate": 0.19163366871508764,
      "loss": 0.2414,
      "num_input_tokens_seen": 31588560,
      "step": 16420
    },
    {
      "epoch": 2.67950077494086,
      "grad_norm": 0.002652783878147602,
      "learning_rate": 0.19157707507839317,
      "loss": 0.2276,
      "num_input_tokens_seen": 31598592,
      "step": 16425
    },
    {
      "epoch": 2.680316502161677,
      "grad_norm": 0.0016141677042469382,
      "learning_rate": 0.19152047502999123,
      "loss": 0.2301,
      "num_input_tokens_seen": 31608800,
      "step": 16430
    },
    {
      "epoch": 2.6811322293824946,
      "grad_norm": 0.001308415550738573,
      "learning_rate": 0.19146386857861025,
      "loss": 0.2018,
      "num_input_tokens_seen": 31618656,
      "step": 16435
    },
    {
      "epoch": 2.6819479566033118,
      "grad_norm": 0.0008357168990187347,
      "learning_rate": 0.19140725573297968,
      "loss": 0.255,
      "num_input_tokens_seen": 31628944,
      "step": 16440
    },
    {
      "epoch": 2.6827636838241293,
      "grad_norm": 0.0008629423682577908,
      "learning_rate": 0.19135063650182987,
      "loss": 0.2011,
      "num_input_tokens_seen": 31638384,
      "step": 16445
    },
    {
      "epoch": 2.6835794110449465,
      "grad_norm": 0.002099673030897975,
      "learning_rate": 0.19129401089389234,
      "loss": 0.2113,
      "num_input_tokens_seen": 31648016,
      "step": 16450
    },
    {
      "epoch": 2.684395138265764,
      "grad_norm": 0.001272105611860752,
      "learning_rate": 0.19123737891789938,
      "loss": 0.2134,
      "num_input_tokens_seen": 31658640,
      "step": 16455
    },
    {
      "epoch": 2.685210865486581,
      "grad_norm": 0.0012761253165081143,
      "learning_rate": 0.19118074058258439,
      "loss": 0.2109,
      "num_input_tokens_seen": 31668736,
      "step": 16460
    },
    {
      "epoch": 2.6860265927073987,
      "grad_norm": 0.001838603289797902,
      "learning_rate": 0.1911240958966816,
      "loss": 0.2053,
      "num_input_tokens_seen": 31679168,
      "step": 16465
    },
    {
      "epoch": 2.686842319928216,
      "grad_norm": 0.001208108733408153,
      "learning_rate": 0.19106744486892652,
      "loss": 0.1911,
      "num_input_tokens_seen": 31689216,
      "step": 16470
    },
    {
      "epoch": 2.6876580471490334,
      "grad_norm": 0.0011926498264074326,
      "learning_rate": 0.1910107875080553,
      "loss": 0.1575,
      "num_input_tokens_seen": 31699600,
      "step": 16475
    },
    {
      "epoch": 2.6884737743698506,
      "grad_norm": 0.0016753117088228464,
      "learning_rate": 0.19095412382280533,
      "loss": 0.2256,
      "num_input_tokens_seen": 31707920,
      "step": 16480
    },
    {
      "epoch": 2.689289501590668,
      "grad_norm": 0.0010358338477090001,
      "learning_rate": 0.19089745382191473,
      "loss": 0.1832,
      "num_input_tokens_seen": 31717648,
      "step": 16485
    },
    {
      "epoch": 2.6901052288114853,
      "grad_norm": 0.0009569067042320967,
      "learning_rate": 0.19084077751412284,
      "loss": 0.2232,
      "num_input_tokens_seen": 31727008,
      "step": 16490
    },
    {
      "epoch": 2.690920956032303,
      "grad_norm": 0.0027289213612675667,
      "learning_rate": 0.19078409490816986,
      "loss": 0.2693,
      "num_input_tokens_seen": 31737200,
      "step": 16495
    },
    {
      "epoch": 2.6917366832531204,
      "grad_norm": 0.000978661235421896,
      "learning_rate": 0.19072740601279686,
      "loss": 0.2002,
      "num_input_tokens_seen": 31746896,
      "step": 16500
    },
    {
      "epoch": 2.6925524104739376,
      "grad_norm": 0.0033880111295729876,
      "learning_rate": 0.19067071083674605,
      "loss": 0.2394,
      "num_input_tokens_seen": 31757776,
      "step": 16505
    },
    {
      "epoch": 2.6933681376947547,
      "grad_norm": 0.0019389826338738203,
      "learning_rate": 0.19061400938876052,
      "loss": 0.2317,
      "num_input_tokens_seen": 31768720,
      "step": 16510
    },
    {
      "epoch": 2.6941838649155723,
      "grad_norm": 0.004079551436007023,
      "learning_rate": 0.1905573016775844,
      "loss": 0.2704,
      "num_input_tokens_seen": 31777712,
      "step": 16515
    },
    {
      "epoch": 2.69499959213639,
      "grad_norm": 0.001384124392643571,
      "learning_rate": 0.19050058771196263,
      "loss": 0.2673,
      "num_input_tokens_seen": 31786944,
      "step": 16520
    },
    {
      "epoch": 2.695815319357207,
      "grad_norm": 0.0014360258355736732,
      "learning_rate": 0.19044386750064132,
      "loss": 0.2028,
      "num_input_tokens_seen": 31796976,
      "step": 16525
    },
    {
      "epoch": 2.696631046578024,
      "grad_norm": 0.0022671164479106665,
      "learning_rate": 0.19038714105236737,
      "loss": 0.2545,
      "num_input_tokens_seen": 31807088,
      "step": 16530
    },
    {
      "epoch": 2.6974467737988417,
      "grad_norm": 0.0008910384494811296,
      "learning_rate": 0.19033040837588874,
      "loss": 0.2313,
      "num_input_tokens_seen": 31817264,
      "step": 16535
    },
    {
      "epoch": 2.6982625010196593,
      "grad_norm": 0.000966460385825485,
      "learning_rate": 0.1902736694799543,
      "loss": 0.1774,
      "num_input_tokens_seen": 31826416,
      "step": 16540
    },
    {
      "epoch": 2.6990782282404764,
      "grad_norm": 0.0008443151018582284,
      "learning_rate": 0.19021692437331392,
      "loss": 0.2302,
      "num_input_tokens_seen": 31834832,
      "step": 16545
    },
    {
      "epoch": 2.6998939554612935,
      "grad_norm": 0.0019898416940122843,
      "learning_rate": 0.1901601730647184,
      "loss": 0.2345,
      "num_input_tokens_seen": 31845072,
      "step": 16550
    },
    {
      "epoch": 2.700709682682111,
      "grad_norm": 0.0010057628387585282,
      "learning_rate": 0.19010341556291954,
      "loss": 0.1972,
      "num_input_tokens_seen": 31855920,
      "step": 16555
    },
    {
      "epoch": 2.7015254099029287,
      "grad_norm": 0.0020752192940562963,
      "learning_rate": 0.19004665187667,
      "loss": 0.2005,
      "num_input_tokens_seen": 31866320,
      "step": 16560
    },
    {
      "epoch": 2.702341137123746,
      "grad_norm": 0.0011480085086077452,
      "learning_rate": 0.1899898820147235,
      "loss": 0.2095,
      "num_input_tokens_seen": 31876384,
      "step": 16565
    },
    {
      "epoch": 2.703156864344563,
      "grad_norm": 0.0013833856210112572,
      "learning_rate": 0.18993310598583465,
      "loss": 0.2273,
      "num_input_tokens_seen": 31885936,
      "step": 16570
    },
    {
      "epoch": 2.7039725915653805,
      "grad_norm": 0.0029668828938156366,
      "learning_rate": 0.18987632379875904,
      "loss": 0.2393,
      "num_input_tokens_seen": 31896592,
      "step": 16575
    },
    {
      "epoch": 2.704788318786198,
      "grad_norm": 0.0019663674756884575,
      "learning_rate": 0.18981953546225314,
      "loss": 0.2347,
      "num_input_tokens_seen": 31907296,
      "step": 16580
    },
    {
      "epoch": 2.705604046007015,
      "grad_norm": 0.0007518271449953318,
      "learning_rate": 0.18976274098507445,
      "loss": 0.1974,
      "num_input_tokens_seen": 31917024,
      "step": 16585
    },
    {
      "epoch": 2.7064197732278323,
      "grad_norm": 0.0011033546179533005,
      "learning_rate": 0.18970594037598146,
      "loss": 0.2589,
      "num_input_tokens_seen": 31925408,
      "step": 16590
    },
    {
      "epoch": 2.70723550044865,
      "grad_norm": 0.0026772613637149334,
      "learning_rate": 0.1896491336437335,
      "loss": 0.2177,
      "num_input_tokens_seen": 31933904,
      "step": 16595
    },
    {
      "epoch": 2.7080512276694675,
      "grad_norm": 0.0031158204656094313,
      "learning_rate": 0.18959232079709085,
      "loss": 0.2257,
      "num_input_tokens_seen": 31943568,
      "step": 16600
    },
    {
      "epoch": 2.7080512276694675,
      "eval_loss": 0.21825097501277924,
      "eval_runtime": 68.0531,
      "eval_samples_per_second": 40.042,
      "eval_steps_per_second": 20.028,
      "num_input_tokens_seen": 31943568,
      "step": 16600
    },
    {
      "epoch": 2.7088669548902846,
      "grad_norm": 0.0008224762277677655,
      "learning_rate": 0.18953550184481477,
      "loss": 0.2302,
      "num_input_tokens_seen": 31953312,
      "step": 16605
    },
    {
      "epoch": 2.709682682111102,
      "grad_norm": 0.001393232843838632,
      "learning_rate": 0.18947867679566752,
      "loss": 0.2062,
      "num_input_tokens_seen": 31963616,
      "step": 16610
    },
    {
      "epoch": 2.7104984093319193,
      "grad_norm": 0.0015450629871338606,
      "learning_rate": 0.18942184565841216,
      "loss": 0.2105,
      "num_input_tokens_seen": 31973632,
      "step": 16615
    },
    {
      "epoch": 2.711314136552737,
      "grad_norm": 0.00287983869202435,
      "learning_rate": 0.18936500844181278,
      "loss": 0.2174,
      "num_input_tokens_seen": 31983904,
      "step": 16620
    },
    {
      "epoch": 2.712129863773554,
      "grad_norm": 0.0012277800124138594,
      "learning_rate": 0.18930816515463436,
      "loss": 0.2475,
      "num_input_tokens_seen": 31994096,
      "step": 16625
    },
    {
      "epoch": 2.7129455909943716,
      "grad_norm": 0.0010654822690412402,
      "learning_rate": 0.18925131580564297,
      "loss": 0.1864,
      "num_input_tokens_seen": 32003600,
      "step": 16630
    },
    {
      "epoch": 2.7137613182151887,
      "grad_norm": 0.0021285361144691706,
      "learning_rate": 0.1891944604036054,
      "loss": 0.2215,
      "num_input_tokens_seen": 32013168,
      "step": 16635
    },
    {
      "epoch": 2.7145770454360063,
      "grad_norm": 0.0011375986505299807,
      "learning_rate": 0.1891375989572895,
      "loss": 0.1962,
      "num_input_tokens_seen": 32023008,
      "step": 16640
    },
    {
      "epoch": 2.7153927726568234,
      "grad_norm": 0.0010283844312652946,
      "learning_rate": 0.18908073147546398,
      "loss": 0.2247,
      "num_input_tokens_seen": 32031840,
      "step": 16645
    },
    {
      "epoch": 2.716208499877641,
      "grad_norm": 0.0029386423993855715,
      "learning_rate": 0.18902385796689858,
      "loss": 0.232,
      "num_input_tokens_seen": 32041936,
      "step": 16650
    },
    {
      "epoch": 2.717024227098458,
      "grad_norm": 0.0020910801831632853,
      "learning_rate": 0.18896697844036384,
      "loss": 0.2094,
      "num_input_tokens_seen": 32050896,
      "step": 16655
    },
    {
      "epoch": 2.7178399543192757,
      "grad_norm": 0.0013091759756207466,
      "learning_rate": 0.18891009290463137,
      "loss": 0.2536,
      "num_input_tokens_seen": 32060736,
      "step": 16660
    },
    {
      "epoch": 2.718655681540093,
      "grad_norm": 0.0010858180467039347,
      "learning_rate": 0.18885320136847353,
      "loss": 0.1883,
      "num_input_tokens_seen": 32069792,
      "step": 16665
    },
    {
      "epoch": 2.7194714087609104,
      "grad_norm": 0.0012653330340981483,
      "learning_rate": 0.1887963038406639,
      "loss": 0.1915,
      "num_input_tokens_seen": 32079344,
      "step": 16670
    },
    {
      "epoch": 2.7202871359817276,
      "grad_norm": 0.0012883958406746387,
      "learning_rate": 0.18873940032997658,
      "loss": 0.2554,
      "num_input_tokens_seen": 32087984,
      "step": 16675
    },
    {
      "epoch": 2.721102863202545,
      "grad_norm": 0.0011236111167818308,
      "learning_rate": 0.18868249084518693,
      "loss": 0.227,
      "num_input_tokens_seen": 32097648,
      "step": 16680
    },
    {
      "epoch": 2.7219185904233623,
      "grad_norm": 0.00264880177564919,
      "learning_rate": 0.18862557539507102,
      "loss": 0.1962,
      "num_input_tokens_seen": 32107232,
      "step": 16685
    },
    {
      "epoch": 2.72273431764418,
      "grad_norm": 0.0014563024742528796,
      "learning_rate": 0.18856865398840605,
      "loss": 0.2035,
      "num_input_tokens_seen": 32116496,
      "step": 16690
    },
    {
      "epoch": 2.7235500448649974,
      "grad_norm": 0.0018841633573174477,
      "learning_rate": 0.18851172663396995,
      "loss": 0.1855,
      "num_input_tokens_seen": 32124512,
      "step": 16695
    },
    {
      "epoch": 2.7243657720858145,
      "grad_norm": 0.0013661598786711693,
      "learning_rate": 0.1884547933405416,
      "loss": 0.2097,
      "num_input_tokens_seen": 32134704,
      "step": 16700
    },
    {
      "epoch": 2.7251814993066317,
      "grad_norm": 0.0017124457517638803,
      "learning_rate": 0.1883978541169009,
      "loss": 0.2216,
      "num_input_tokens_seen": 32144496,
      "step": 16705
    },
    {
      "epoch": 2.7259972265274492,
      "grad_norm": 0.003798469202592969,
      "learning_rate": 0.18834090897182854,
      "loss": 0.2382,
      "num_input_tokens_seen": 32153664,
      "step": 16710
    },
    {
      "epoch": 2.726812953748267,
      "grad_norm": 0.0026143984869122505,
      "learning_rate": 0.1882839579141062,
      "loss": 0.2268,
      "num_input_tokens_seen": 32162400,
      "step": 16715
    },
    {
      "epoch": 2.727628680969084,
      "grad_norm": 0.0020178365521132946,
      "learning_rate": 0.18822700095251646,
      "loss": 0.2199,
      "num_input_tokens_seen": 32172544,
      "step": 16720
    },
    {
      "epoch": 2.728444408189901,
      "grad_norm": 0.0012766310246661305,
      "learning_rate": 0.18817003809584273,
      "loss": 0.2103,
      "num_input_tokens_seen": 32181792,
      "step": 16725
    },
    {
      "epoch": 2.7292601354107187,
      "grad_norm": 0.0019669707398861647,
      "learning_rate": 0.1881130693528695,
      "loss": 0.2563,
      "num_input_tokens_seen": 32191408,
      "step": 16730
    },
    {
      "epoch": 2.7300758626315362,
      "grad_norm": 0.0024338068906217813,
      "learning_rate": 0.18805609473238197,
      "loss": 0.2304,
      "num_input_tokens_seen": 32200432,
      "step": 16735
    },
    {
      "epoch": 2.7308915898523534,
      "grad_norm": 0.001517699914984405,
      "learning_rate": 0.18799911424316643,
      "loss": 0.2026,
      "num_input_tokens_seen": 32211424,
      "step": 16740
    },
    {
      "epoch": 2.7317073170731705,
      "grad_norm": 0.0017068764427676797,
      "learning_rate": 0.18794212789400994,
      "loss": 0.1752,
      "num_input_tokens_seen": 32221408,
      "step": 16745
    },
    {
      "epoch": 2.732523044293988,
      "grad_norm": 0.0017904369160532951,
      "learning_rate": 0.18788513569370052,
      "loss": 0.2134,
      "num_input_tokens_seen": 32230480,
      "step": 16750
    },
    {
      "epoch": 2.7333387715148056,
      "grad_norm": 0.0012227774132043123,
      "learning_rate": 0.1878281376510271,
      "loss": 0.2401,
      "num_input_tokens_seen": 32239664,
      "step": 16755
    },
    {
      "epoch": 2.7341544987356228,
      "grad_norm": 0.00320668937638402,
      "learning_rate": 0.18777113377477941,
      "loss": 0.2381,
      "num_input_tokens_seen": 32249072,
      "step": 16760
    },
    {
      "epoch": 2.73497022595644,
      "grad_norm": 0.003646585624665022,
      "learning_rate": 0.1877141240737483,
      "loss": 0.2517,
      "num_input_tokens_seen": 32260960,
      "step": 16765
    },
    {
      "epoch": 2.7357859531772575,
      "grad_norm": 0.0014419177314266562,
      "learning_rate": 0.18765710855672527,
      "loss": 0.2207,
      "num_input_tokens_seen": 32271696,
      "step": 16770
    },
    {
      "epoch": 2.736601680398075,
      "grad_norm": 0.0015894502867013216,
      "learning_rate": 0.18760008723250288,
      "loss": 0.223,
      "num_input_tokens_seen": 32281328,
      "step": 16775
    },
    {
      "epoch": 2.737417407618892,
      "grad_norm": 0.0014865216799080372,
      "learning_rate": 0.18754306010987457,
      "loss": 0.242,
      "num_input_tokens_seen": 32291984,
      "step": 16780
    },
    {
      "epoch": 2.7382331348397098,
      "grad_norm": 0.0011485576396808028,
      "learning_rate": 0.18748602719763457,
      "loss": 0.2355,
      "num_input_tokens_seen": 32300128,
      "step": 16785
    },
    {
      "epoch": 2.739048862060527,
      "grad_norm": 0.0015888113994151354,
      "learning_rate": 0.18742898850457804,
      "loss": 0.2158,
      "num_input_tokens_seen": 32310656,
      "step": 16790
    },
    {
      "epoch": 2.7398645892813445,
      "grad_norm": 0.0014442091342061758,
      "learning_rate": 0.1873719440395012,
      "loss": 0.206,
      "num_input_tokens_seen": 32319248,
      "step": 16795
    },
    {
      "epoch": 2.7406803165021616,
      "grad_norm": 0.0024915735702961683,
      "learning_rate": 0.1873148938112009,
      "loss": 0.2572,
      "num_input_tokens_seen": 32327088,
      "step": 16800
    },
    {
      "epoch": 2.7406803165021616,
      "eval_loss": 0.21645726263523102,
      "eval_runtime": 68.0604,
      "eval_samples_per_second": 40.038,
      "eval_steps_per_second": 20.026,
      "num_input_tokens_seen": 32327088,
      "step": 16800
    },
    {
      "epoch": 2.741496043722979,
      "grad_norm": 0.0008707664674147964,
      "learning_rate": 0.18725783782847508,
      "loss": 0.2087,
      "num_input_tokens_seen": 32337040,
      "step": 16805
    },
    {
      "epoch": 2.7423117709437963,
      "grad_norm": 0.0014663152396678925,
      "learning_rate": 0.1872007761001224,
      "loss": 0.2232,
      "num_input_tokens_seen": 32346800,
      "step": 16810
    },
    {
      "epoch": 2.743127498164614,
      "grad_norm": 0.001914837397634983,
      "learning_rate": 0.1871437086349426,
      "loss": 0.2406,
      "num_input_tokens_seen": 32355728,
      "step": 16815
    },
    {
      "epoch": 2.743943225385431,
      "grad_norm": 0.0023309935349971056,
      "learning_rate": 0.18708663544173615,
      "loss": 0.2185,
      "num_input_tokens_seen": 32364528,
      "step": 16820
    },
    {
      "epoch": 2.7447589526062486,
      "grad_norm": 0.0023990969639271498,
      "learning_rate": 0.18702955652930442,
      "loss": 0.2168,
      "num_input_tokens_seen": 32374448,
      "step": 16825
    },
    {
      "epoch": 2.7455746798270657,
      "grad_norm": 0.0012189536355435848,
      "learning_rate": 0.18697247190644972,
      "loss": 0.2223,
      "num_input_tokens_seen": 32383760,
      "step": 16830
    },
    {
      "epoch": 2.7463904070478833,
      "grad_norm": 0.0009733401238918304,
      "learning_rate": 0.18691538158197527,
      "loss": 0.1914,
      "num_input_tokens_seen": 32393632,
      "step": 16835
    },
    {
      "epoch": 2.7472061342687004,
      "grad_norm": 0.0014027627184987068,
      "learning_rate": 0.1868582855646851,
      "loss": 0.2257,
      "num_input_tokens_seen": 32403456,
      "step": 16840
    },
    {
      "epoch": 2.748021861489518,
      "grad_norm": 0.0011895861243829131,
      "learning_rate": 0.18680118386338404,
      "loss": 0.208,
      "num_input_tokens_seen": 32412656,
      "step": 16845
    },
    {
      "epoch": 2.748837588710335,
      "grad_norm": 0.0017163784941658378,
      "learning_rate": 0.18674407648687794,
      "loss": 0.2452,
      "num_input_tokens_seen": 32422992,
      "step": 16850
    },
    {
      "epoch": 2.7496533159311527,
      "grad_norm": 0.0017204126343131065,
      "learning_rate": 0.1866869634439736,
      "loss": 0.2046,
      "num_input_tokens_seen": 32432816,
      "step": 16855
    },
    {
      "epoch": 2.75046904315197,
      "grad_norm": 0.0017097815871238708,
      "learning_rate": 0.18662984474347838,
      "loss": 0.1968,
      "num_input_tokens_seen": 32442528,
      "step": 16860
    },
    {
      "epoch": 2.7512847703727874,
      "grad_norm": 0.00141466676723212,
      "learning_rate": 0.1865727203942008,
      "loss": 0.2545,
      "num_input_tokens_seen": 32450272,
      "step": 16865
    },
    {
      "epoch": 2.7521004975936045,
      "grad_norm": 0.001585735590197146,
      "learning_rate": 0.1865155904049501,
      "loss": 0.2375,
      "num_input_tokens_seen": 32461152,
      "step": 16870
    },
    {
      "epoch": 2.752916224814422,
      "grad_norm": 0.0020492838229984045,
      "learning_rate": 0.1864584547845365,
      "loss": 0.227,
      "num_input_tokens_seen": 32470832,
      "step": 16875
    },
    {
      "epoch": 2.7537319520352392,
      "grad_norm": 0.0010750783840194345,
      "learning_rate": 0.186401313541771,
      "loss": 0.1992,
      "num_input_tokens_seen": 32480048,
      "step": 16880
    },
    {
      "epoch": 2.754547679256057,
      "grad_norm": 0.0015481255250051618,
      "learning_rate": 0.18634416668546552,
      "loss": 0.238,
      "num_input_tokens_seen": 32488848,
      "step": 16885
    },
    {
      "epoch": 2.7553634064768744,
      "grad_norm": 0.0013866720255464315,
      "learning_rate": 0.1862870142244328,
      "loss": 0.2236,
      "num_input_tokens_seen": 32497328,
      "step": 16890
    },
    {
      "epoch": 2.7561791336976915,
      "grad_norm": 0.001306916237808764,
      "learning_rate": 0.1862298561674865,
      "loss": 0.1978,
      "num_input_tokens_seen": 32506832,
      "step": 16895
    },
    {
      "epoch": 2.7569948609185086,
      "grad_norm": 0.0018112406833097339,
      "learning_rate": 0.18617269252344104,
      "loss": 0.1374,
      "num_input_tokens_seen": 32517664,
      "step": 16900
    },
    {
      "epoch": 2.7578105881393262,
      "grad_norm": 0.003160190535709262,
      "learning_rate": 0.18611552330111186,
      "loss": 0.2802,
      "num_input_tokens_seen": 32527504,
      "step": 16905
    },
    {
      "epoch": 2.758626315360144,
      "grad_norm": 0.0010038440814241767,
      "learning_rate": 0.18605834850931507,
      "loss": 0.1796,
      "num_input_tokens_seen": 32536928,
      "step": 16910
    },
    {
      "epoch": 2.759442042580961,
      "grad_norm": 0.0014553109649568796,
      "learning_rate": 0.18600116815686787,
      "loss": 0.2528,
      "num_input_tokens_seen": 32547232,
      "step": 16915
    },
    {
      "epoch": 2.760257769801778,
      "grad_norm": 0.0011776386527344584,
      "learning_rate": 0.1859439822525881,
      "loss": 0.218,
      "num_input_tokens_seen": 32557536,
      "step": 16920
    },
    {
      "epoch": 2.7610734970225956,
      "grad_norm": 0.001150056254118681,
      "learning_rate": 0.18588679080529455,
      "loss": 0.2281,
      "num_input_tokens_seen": 32567632,
      "step": 16925
    },
    {
      "epoch": 2.761889224243413,
      "grad_norm": 0.0018866684986278415,
      "learning_rate": 0.1858295938238069,
      "loss": 0.2669,
      "num_input_tokens_seen": 32577760,
      "step": 16930
    },
    {
      "epoch": 2.7627049514642303,
      "grad_norm": 0.0018141299951821566,
      "learning_rate": 0.18577239131694562,
      "loss": 0.2354,
      "num_input_tokens_seen": 32586704,
      "step": 16935
    },
    {
      "epoch": 2.7635206786850475,
      "grad_norm": 0.0019449605606496334,
      "learning_rate": 0.18571518329353204,
      "loss": 0.2369,
      "num_input_tokens_seen": 32596448,
      "step": 16940
    },
    {
      "epoch": 2.764336405905865,
      "grad_norm": 0.0019390449160709977,
      "learning_rate": 0.18565796976238838,
      "loss": 0.1904,
      "num_input_tokens_seen": 32606464,
      "step": 16945
    },
    {
      "epoch": 2.7651521331266826,
      "grad_norm": 0.0015003228327259421,
      "learning_rate": 0.18560075073233764,
      "loss": 0.2433,
      "num_input_tokens_seen": 32615792,
      "step": 16950
    },
    {
      "epoch": 2.7659678603474998,
      "grad_norm": 0.0013448350364342332,
      "learning_rate": 0.18554352621220377,
      "loss": 0.2271,
      "num_input_tokens_seen": 32625168,
      "step": 16955
    },
    {
      "epoch": 2.766783587568317,
      "grad_norm": 0.001250170054845512,
      "learning_rate": 0.18548629621081153,
      "loss": 0.2429,
      "num_input_tokens_seen": 32634976,
      "step": 16960
    },
    {
      "epoch": 2.7675993147891345,
      "grad_norm": 0.002084009815007448,
      "learning_rate": 0.18542906073698645,
      "loss": 0.2011,
      "num_input_tokens_seen": 32644800,
      "step": 16965
    },
    {
      "epoch": 2.768415042009952,
      "grad_norm": 0.0017677575815469027,
      "learning_rate": 0.18537181979955494,
      "loss": 0.206,
      "num_input_tokens_seen": 32653856,
      "step": 16970
    },
    {
      "epoch": 2.769230769230769,
      "grad_norm": 0.002144760452210903,
      "learning_rate": 0.18531457340734434,
      "loss": 0.2325,
      "num_input_tokens_seen": 32664192,
      "step": 16975
    },
    {
      "epoch": 2.7700464964515867,
      "grad_norm": 0.0022150976583361626,
      "learning_rate": 0.1852573215691827,
      "loss": 0.1876,
      "num_input_tokens_seen": 32674224,
      "step": 16980
    },
    {
      "epoch": 2.770862223672404,
      "grad_norm": 0.0026116748340427876,
      "learning_rate": 0.18520006429389904,
      "loss": 0.2023,
      "num_input_tokens_seen": 32683088,
      "step": 16985
    },
    {
      "epoch": 2.7716779508932214,
      "grad_norm": 0.001799207297153771,
      "learning_rate": 0.1851428015903231,
      "loss": 0.2175,
      "num_input_tokens_seen": 32692800,
      "step": 16990
    },
    {
      "epoch": 2.7724936781140386,
      "grad_norm": 0.002230857266113162,
      "learning_rate": 0.1850855334672855,
      "loss": 0.2039,
      "num_input_tokens_seen": 32703104,
      "step": 16995
    },
    {
      "epoch": 2.773309405334856,
      "grad_norm": 0.0021873631048947573,
      "learning_rate": 0.1850282599336178,
      "loss": 0.2384,
      "num_input_tokens_seen": 32713728,
      "step": 17000
    },
    {
      "epoch": 2.773309405334856,
      "eval_loss": 0.21259798109531403,
      "eval_runtime": 68.1147,
      "eval_samples_per_second": 40.006,
      "eval_steps_per_second": 20.01,
      "num_input_tokens_seen": 32713728,
      "step": 17000
    },
    {
      "epoch": 2.7741251325556733,
      "grad_norm": 0.0018597505986690521,
      "learning_rate": 0.18497098099815215,
      "loss": 0.2523,
      "num_input_tokens_seen": 32723696,
      "step": 17005
    },
    {
      "epoch": 2.774940859776491,
      "grad_norm": 0.001677878899499774,
      "learning_rate": 0.18491369666972174,
      "loss": 0.2057,
      "num_input_tokens_seen": 32733120,
      "step": 17010
    },
    {
      "epoch": 2.775756586997308,
      "grad_norm": 0.0013932186411693692,
      "learning_rate": 0.1848564069571606,
      "loss": 0.206,
      "num_input_tokens_seen": 32741472,
      "step": 17015
    },
    {
      "epoch": 2.7765723142181256,
      "grad_norm": 0.002581206848844886,
      "learning_rate": 0.18479911186930348,
      "loss": 0.2165,
      "num_input_tokens_seen": 32750832,
      "step": 17020
    },
    {
      "epoch": 2.7773880414389427,
      "grad_norm": 0.0025405073538422585,
      "learning_rate": 0.18474181141498597,
      "loss": 0.1889,
      "num_input_tokens_seen": 32760256,
      "step": 17025
    },
    {
      "epoch": 2.7782037686597603,
      "grad_norm": 0.0026740028988569975,
      "learning_rate": 0.18468450560304453,
      "loss": 0.2395,
      "num_input_tokens_seen": 32770256,
      "step": 17030
    },
    {
      "epoch": 2.7790194958805774,
      "grad_norm": 0.002465682104229927,
      "learning_rate": 0.1846271944423165,
      "loss": 0.1861,
      "num_input_tokens_seen": 32781088,
      "step": 17035
    },
    {
      "epoch": 2.779835223101395,
      "grad_norm": 0.0017491680337116122,
      "learning_rate": 0.18456987794163993,
      "loss": 0.2138,
      "num_input_tokens_seen": 32790400,
      "step": 17040
    },
    {
      "epoch": 2.780650950322212,
      "grad_norm": 0.0028783874586224556,
      "learning_rate": 0.18451255610985373,
      "loss": 0.2432,
      "num_input_tokens_seen": 32800464,
      "step": 17045
    },
    {
      "epoch": 2.7814666775430297,
      "grad_norm": 0.001083904760889709,
      "learning_rate": 0.18445522895579766,
      "loss": 0.2031,
      "num_input_tokens_seen": 32810272,
      "step": 17050
    },
    {
      "epoch": 2.782282404763847,
      "grad_norm": 0.0012796137016266584,
      "learning_rate": 0.1843978964883123,
      "loss": 0.1819,
      "num_input_tokens_seen": 32820768,
      "step": 17055
    },
    {
      "epoch": 2.7830981319846644,
      "grad_norm": 0.0012556766159832478,
      "learning_rate": 0.18434055871623906,
      "loss": 0.2743,
      "num_input_tokens_seen": 32831200,
      "step": 17060
    },
    {
      "epoch": 2.783913859205482,
      "grad_norm": 0.0012569797690957785,
      "learning_rate": 0.18428321564842007,
      "loss": 0.2157,
      "num_input_tokens_seen": 32839216,
      "step": 17065
    },
    {
      "epoch": 2.784729586426299,
      "grad_norm": 0.0025090461131185293,
      "learning_rate": 0.18422586729369841,
      "loss": 0.2166,
      "num_input_tokens_seen": 32848688,
      "step": 17070
    },
    {
      "epoch": 2.785545313647116,
      "grad_norm": 0.003728437703102827,
      "learning_rate": 0.1841685136609179,
      "loss": 0.2421,
      "num_input_tokens_seen": 32857152,
      "step": 17075
    },
    {
      "epoch": 2.786361040867934,
      "grad_norm": 0.0025184843689203262,
      "learning_rate": 0.18411115475892326,
      "loss": 0.2157,
      "num_input_tokens_seen": 32866496,
      "step": 17080
    },
    {
      "epoch": 2.7871767680887514,
      "grad_norm": 0.002286541974171996,
      "learning_rate": 0.18405379059655982,
      "loss": 0.2167,
      "num_input_tokens_seen": 32875088,
      "step": 17085
    },
    {
      "epoch": 2.7879924953095685,
      "grad_norm": 0.0020246021449565887,
      "learning_rate": 0.1839964211826739,
      "loss": 0.2059,
      "num_input_tokens_seen": 32885136,
      "step": 17090
    },
    {
      "epoch": 2.7888082225303856,
      "grad_norm": 0.0021384856663644314,
      "learning_rate": 0.18393904652611265,
      "loss": 0.1969,
      "num_input_tokens_seen": 32894896,
      "step": 17095
    },
    {
      "epoch": 2.789623949751203,
      "grad_norm": 0.002357268473133445,
      "learning_rate": 0.18388166663572392,
      "loss": 0.2375,
      "num_input_tokens_seen": 32903440,
      "step": 17100
    },
    {
      "epoch": 2.7904396769720208,
      "grad_norm": 0.001872071879915893,
      "learning_rate": 0.18382428152035643,
      "loss": 0.236,
      "num_input_tokens_seen": 32912624,
      "step": 17105
    },
    {
      "epoch": 2.791255404192838,
      "grad_norm": 0.0014203505124896765,
      "learning_rate": 0.1837668911888596,
      "loss": 0.2134,
      "num_input_tokens_seen": 32921216,
      "step": 17110
    },
    {
      "epoch": 2.792071131413655,
      "grad_norm": 0.0014285912038758397,
      "learning_rate": 0.18370949565008388,
      "loss": 0.305,
      "num_input_tokens_seen": 32931136,
      "step": 17115
    },
    {
      "epoch": 2.7928868586344726,
      "grad_norm": 0.004123109392821789,
      "learning_rate": 0.1836520949128803,
      "loss": 0.2651,
      "num_input_tokens_seen": 32941056,
      "step": 17120
    },
    {
      "epoch": 2.79370258585529,
      "grad_norm": 0.0015568409580737352,
      "learning_rate": 0.18359468898610076,
      "loss": 0.2135,
      "num_input_tokens_seen": 32951008,
      "step": 17125
    },
    {
      "epoch": 2.7945183130761073,
      "grad_norm": 0.001914727734401822,
      "learning_rate": 0.18353727787859797,
      "loss": 0.2066,
      "num_input_tokens_seen": 32958864,
      "step": 17130
    },
    {
      "epoch": 2.7953340402969244,
      "grad_norm": 0.0014461737591773272,
      "learning_rate": 0.18347986159922552,
      "loss": 0.1878,
      "num_input_tokens_seen": 32967984,
      "step": 17135
    },
    {
      "epoch": 2.796149767517742,
      "grad_norm": 0.002717912197113037,
      "learning_rate": 0.1834224401568377,
      "loss": 0.2466,
      "num_input_tokens_seen": 32978512,
      "step": 17140
    },
    {
      "epoch": 2.7969654947385596,
      "grad_norm": 0.0012782647972926497,
      "learning_rate": 0.1833650135602896,
      "loss": 0.2486,
      "num_input_tokens_seen": 32987952,
      "step": 17145
    },
    {
      "epoch": 2.7977812219593767,
      "grad_norm": 0.0009460340952500701,
      "learning_rate": 0.18330758181843707,
      "loss": 0.2064,
      "num_input_tokens_seen": 32997104,
      "step": 17150
    },
    {
      "epoch": 2.7985969491801943,
      "grad_norm": 0.0014384982641786337,
      "learning_rate": 0.18325014494013686,
      "loss": 0.2479,
      "num_input_tokens_seen": 33007216,
      "step": 17155
    },
    {
      "epoch": 2.7994126764010114,
      "grad_norm": 0.001311219995841384,
      "learning_rate": 0.18319270293424647,
      "loss": 0.1847,
      "num_input_tokens_seen": 33017280,
      "step": 17160
    },
    {
      "epoch": 2.800228403621829,
      "grad_norm": 0.002410293323919177,
      "learning_rate": 0.18313525580962417,
      "loss": 0.2218,
      "num_input_tokens_seen": 33026752,
      "step": 17165
    },
    {
      "epoch": 2.801044130842646,
      "grad_norm": 0.0011970439227297902,
      "learning_rate": 0.18307780357512896,
      "loss": 0.2224,
      "num_input_tokens_seen": 33034672,
      "step": 17170
    },
    {
      "epoch": 2.8018598580634637,
      "grad_norm": 0.0032715527340769768,
      "learning_rate": 0.1830203462396208,
      "loss": 0.2018,
      "num_input_tokens_seen": 33043616,
      "step": 17175
    },
    {
      "epoch": 2.802675585284281,
      "grad_norm": 0.0045157382264733315,
      "learning_rate": 0.18296288381196033,
      "loss": 0.2685,
      "num_input_tokens_seen": 33053808,
      "step": 17180
    },
    {
      "epoch": 2.8034913125050984,
      "grad_norm": 0.001312423963099718,
      "learning_rate": 0.1829054163010089,
      "loss": 0.2151,
      "num_input_tokens_seen": 33063968,
      "step": 17185
    },
    {
      "epoch": 2.8043070397259156,
      "grad_norm": 0.0025378847494721413,
      "learning_rate": 0.18284794371562874,
      "loss": 0.1881,
      "num_input_tokens_seen": 33074368,
      "step": 17190
    },
    {
      "epoch": 2.805122766946733,
      "grad_norm": 0.0015990277752280235,
      "learning_rate": 0.18279046606468288,
      "loss": 0.2087,
      "num_input_tokens_seen": 33084160,
      "step": 17195
    },
    {
      "epoch": 2.8059384941675503,
      "grad_norm": 0.0020378620829433203,
      "learning_rate": 0.1827329833570351,
      "loss": 0.2622,
      "num_input_tokens_seen": 33093744,
      "step": 17200
    },
    {
      "epoch": 2.8059384941675503,
      "eval_loss": 0.21470265090465546,
      "eval_runtime": 67.9695,
      "eval_samples_per_second": 40.092,
      "eval_steps_per_second": 20.053,
      "num_input_tokens_seen": 33093744,
      "step": 17200
    },
    {
      "epoch": 2.806754221388368,
      "grad_norm": 0.0015019119018688798,
      "learning_rate": 0.18267549560154991,
      "loss": 0.2342,
      "num_input_tokens_seen": 33103712,
      "step": 17205
    },
    {
      "epoch": 2.807569948609185,
      "grad_norm": 0.001360650989226997,
      "learning_rate": 0.18261800280709267,
      "loss": 0.2237,
      "num_input_tokens_seen": 33113968,
      "step": 17210
    },
    {
      "epoch": 2.8083856758300025,
      "grad_norm": 0.000981933786533773,
      "learning_rate": 0.18256050498252957,
      "loss": 0.2154,
      "num_input_tokens_seen": 33123200,
      "step": 17215
    },
    {
      "epoch": 2.8092014030508197,
      "grad_norm": 0.0013070376589894295,
      "learning_rate": 0.18250300213672735,
      "loss": 0.2095,
      "num_input_tokens_seen": 33131696,
      "step": 17220
    },
    {
      "epoch": 2.8100171302716372,
      "grad_norm": 0.0020347055979073048,
      "learning_rate": 0.18244549427855378,
      "loss": 0.2097,
      "num_input_tokens_seen": 33141872,
      "step": 17225
    },
    {
      "epoch": 2.8108328574924544,
      "grad_norm": 0.002593195764347911,
      "learning_rate": 0.1823879814168772,
      "loss": 0.2157,
      "num_input_tokens_seen": 33149808,
      "step": 17230
    },
    {
      "epoch": 2.811648584713272,
      "grad_norm": 0.002613932592794299,
      "learning_rate": 0.18233046356056692,
      "loss": 0.2041,
      "num_input_tokens_seen": 33158304,
      "step": 17235
    },
    {
      "epoch": 2.812464311934089,
      "grad_norm": 0.0013588924193754792,
      "learning_rate": 0.18227294071849284,
      "loss": 0.2428,
      "num_input_tokens_seen": 33168144,
      "step": 17240
    },
    {
      "epoch": 2.8132800391549067,
      "grad_norm": 0.003212355077266693,
      "learning_rate": 0.18221541289952578,
      "loss": 0.2485,
      "num_input_tokens_seen": 33179152,
      "step": 17245
    },
    {
      "epoch": 2.814095766375724,
      "grad_norm": 0.0011835177429020405,
      "learning_rate": 0.18215788011253717,
      "loss": 0.22,
      "num_input_tokens_seen": 33190320,
      "step": 17250
    },
    {
      "epoch": 2.8149114935965414,
      "grad_norm": 0.002414737595245242,
      "learning_rate": 0.18210034236639935,
      "loss": 0.2308,
      "num_input_tokens_seen": 33199936,
      "step": 17255
    },
    {
      "epoch": 2.815727220817359,
      "grad_norm": 0.0010666660964488983,
      "learning_rate": 0.1820427996699853,
      "loss": 0.217,
      "num_input_tokens_seen": 33210000,
      "step": 17260
    },
    {
      "epoch": 2.816542948038176,
      "grad_norm": 0.0013518300838768482,
      "learning_rate": 0.1819852520321689,
      "loss": 0.2102,
      "num_input_tokens_seen": 33220112,
      "step": 17265
    },
    {
      "epoch": 2.817358675258993,
      "grad_norm": 0.002244547475129366,
      "learning_rate": 0.18192769946182466,
      "loss": 0.216,
      "num_input_tokens_seen": 33230464,
      "step": 17270
    },
    {
      "epoch": 2.8181744024798108,
      "grad_norm": 0.0021519013680517673,
      "learning_rate": 0.18187014196782794,
      "loss": 0.2179,
      "num_input_tokens_seen": 33240112,
      "step": 17275
    },
    {
      "epoch": 2.8189901297006283,
      "grad_norm": 0.0030847741290926933,
      "learning_rate": 0.18181257955905486,
      "loss": 0.2049,
      "num_input_tokens_seen": 33250224,
      "step": 17280
    },
    {
      "epoch": 2.8198058569214455,
      "grad_norm": 0.0032954595517367125,
      "learning_rate": 0.18175501224438217,
      "loss": 0.1953,
      "num_input_tokens_seen": 33260448,
      "step": 17285
    },
    {
      "epoch": 2.8206215841422626,
      "grad_norm": 0.0022184087429195642,
      "learning_rate": 0.18169744003268756,
      "loss": 0.1968,
      "num_input_tokens_seen": 33270784,
      "step": 17290
    },
    {
      "epoch": 2.82143731136308,
      "grad_norm": 0.0028919887263327837,
      "learning_rate": 0.18163986293284937,
      "loss": 0.2159,
      "num_input_tokens_seen": 33281168,
      "step": 17295
    },
    {
      "epoch": 2.8222530385838978,
      "grad_norm": 0.001810177112929523,
      "learning_rate": 0.18158228095374673,
      "loss": 0.2021,
      "num_input_tokens_seen": 33290816,
      "step": 17300
    },
    {
      "epoch": 2.823068765804715,
      "grad_norm": 0.000926488486584276,
      "learning_rate": 0.18152469410425945,
      "loss": 0.25,
      "num_input_tokens_seen": 33301488,
      "step": 17305
    },
    {
      "epoch": 2.823884493025532,
      "grad_norm": 0.0018490925431251526,
      "learning_rate": 0.18146710239326813,
      "loss": 0.2315,
      "num_input_tokens_seen": 33310592,
      "step": 17310
    },
    {
      "epoch": 2.8247002202463496,
      "grad_norm": 0.0031681868713349104,
      "learning_rate": 0.18140950582965423,
      "loss": 0.2152,
      "num_input_tokens_seen": 33319808,
      "step": 17315
    },
    {
      "epoch": 2.825515947467167,
      "grad_norm": 0.002980048768222332,
      "learning_rate": 0.1813519044222998,
      "loss": 0.254,
      "num_input_tokens_seen": 33330192,
      "step": 17320
    },
    {
      "epoch": 2.8263316746879843,
      "grad_norm": 0.0013149766018614173,
      "learning_rate": 0.18129429818008772,
      "loss": 0.2219,
      "num_input_tokens_seen": 33339952,
      "step": 17325
    },
    {
      "epoch": 2.8271474019088014,
      "grad_norm": 0.0026584304869174957,
      "learning_rate": 0.18123668711190163,
      "loss": 0.2415,
      "num_input_tokens_seen": 33350752,
      "step": 17330
    },
    {
      "epoch": 2.827963129129619,
      "grad_norm": 0.0018115343991667032,
      "learning_rate": 0.18117907122662583,
      "loss": 0.2323,
      "num_input_tokens_seen": 33361392,
      "step": 17335
    },
    {
      "epoch": 2.8287788563504366,
      "grad_norm": 0.001348971389234066,
      "learning_rate": 0.1811214505331454,
      "loss": 0.2146,
      "num_input_tokens_seen": 33370400,
      "step": 17340
    },
    {
      "epoch": 2.8295945835712537,
      "grad_norm": 0.002248800126835704,
      "learning_rate": 0.1810638250403462,
      "loss": 0.2113,
      "num_input_tokens_seen": 33379760,
      "step": 17345
    },
    {
      "epoch": 2.8304103107920713,
      "grad_norm": 0.0017923032864928246,
      "learning_rate": 0.1810061947571148,
      "loss": 0.1974,
      "num_input_tokens_seen": 33389856,
      "step": 17350
    },
    {
      "epoch": 2.8312260380128884,
      "grad_norm": 0.001509751658886671,
      "learning_rate": 0.1809485596923385,
      "loss": 0.2358,
      "num_input_tokens_seen": 33398416,
      "step": 17355
    },
    {
      "epoch": 2.832041765233706,
      "grad_norm": 0.001389391254633665,
      "learning_rate": 0.18089091985490546,
      "loss": 0.1738,
      "num_input_tokens_seen": 33407632,
      "step": 17360
    },
    {
      "epoch": 2.832857492454523,
      "grad_norm": 0.002152335597202182,
      "learning_rate": 0.18083327525370432,
      "loss": 0.2394,
      "num_input_tokens_seen": 33416672,
      "step": 17365
    },
    {
      "epoch": 2.8336732196753407,
      "grad_norm": 0.0010743675520643592,
      "learning_rate": 0.18077562589762464,
      "loss": 0.1838,
      "num_input_tokens_seen": 33426608,
      "step": 17370
    },
    {
      "epoch": 2.834488946896158,
      "grad_norm": 0.0007242867141030729,
      "learning_rate": 0.1807179717955567,
      "loss": 0.1805,
      "num_input_tokens_seen": 33435072,
      "step": 17375
    },
    {
      "epoch": 2.8353046741169754,
      "grad_norm": 0.0014209318906068802,
      "learning_rate": 0.1806603129563915,
      "loss": 0.2516,
      "num_input_tokens_seen": 33444512,
      "step": 17380
    },
    {
      "epoch": 2.8361204013377925,
      "grad_norm": 0.002244531875476241,
      "learning_rate": 0.1806026493890208,
      "loss": 0.2072,
      "num_input_tokens_seen": 33454256,
      "step": 17385
    },
    {
      "epoch": 2.83693612855861,
      "grad_norm": 0.0017084302380681038,
      "learning_rate": 0.18054498110233688,
      "loss": 0.2201,
      "num_input_tokens_seen": 33464208,
      "step": 17390
    },
    {
      "epoch": 2.8377518557794272,
      "grad_norm": 0.0013917917385697365,
      "learning_rate": 0.1804873081052331,
      "loss": 0.2345,
      "num_input_tokens_seen": 33474544,
      "step": 17395
    },
    {
      "epoch": 2.838567583000245,
      "grad_norm": 0.001905798795633018,
      "learning_rate": 0.18042963040660326,
      "loss": 0.2166,
      "num_input_tokens_seen": 33484336,
      "step": 17400
    },
    {
      "epoch": 2.838567583000245,
      "eval_loss": 0.21387672424316406,
      "eval_runtime": 68.0721,
      "eval_samples_per_second": 40.031,
      "eval_steps_per_second": 20.023,
      "num_input_tokens_seen": 33484336,
      "step": 17400
    },
    {
      "epoch": 2.839383310221062,
      "grad_norm": 0.0009047597995959222,
      "learning_rate": 0.180371948015342,
      "loss": 0.2382,
      "num_input_tokens_seen": 33494864,
      "step": 17405
    },
    {
      "epoch": 2.8401990374418795,
      "grad_norm": 0.0018410299671813846,
      "learning_rate": 0.18031426094034472,
      "loss": 0.236,
      "num_input_tokens_seen": 33504224,
      "step": 17410
    },
    {
      "epoch": 2.8410147646626966,
      "grad_norm": 0.002076416974887252,
      "learning_rate": 0.18025656919050737,
      "loss": 0.2466,
      "num_input_tokens_seen": 33513568,
      "step": 17415
    },
    {
      "epoch": 2.841830491883514,
      "grad_norm": 0.0024521765299141407,
      "learning_rate": 0.18019887277472688,
      "loss": 0.235,
      "num_input_tokens_seen": 33522272,
      "step": 17420
    },
    {
      "epoch": 2.8426462191043314,
      "grad_norm": 0.0018232178408652544,
      "learning_rate": 0.18014117170190067,
      "loss": 0.215,
      "num_input_tokens_seen": 33532016,
      "step": 17425
    },
    {
      "epoch": 2.843461946325149,
      "grad_norm": 0.0008960649720393121,
      "learning_rate": 0.18008346598092703,
      "loss": 0.2204,
      "num_input_tokens_seen": 33541984,
      "step": 17430
    },
    {
      "epoch": 2.8442776735459665,
      "grad_norm": 0.0014091355260461569,
      "learning_rate": 0.18002575562070489,
      "loss": 0.2169,
      "num_input_tokens_seen": 33552320,
      "step": 17435
    },
    {
      "epoch": 2.8450934007667836,
      "grad_norm": 0.0013988767750561237,
      "learning_rate": 0.1799680406301339,
      "loss": 0.1983,
      "num_input_tokens_seen": 33562336,
      "step": 17440
    },
    {
      "epoch": 2.8459091279876008,
      "grad_norm": 0.001848592539317906,
      "learning_rate": 0.17991032101811447,
      "loss": 0.2504,
      "num_input_tokens_seen": 33571952,
      "step": 17445
    },
    {
      "epoch": 2.8467248552084183,
      "grad_norm": 0.0013286349130794406,
      "learning_rate": 0.1798525967935476,
      "loss": 0.1843,
      "num_input_tokens_seen": 33581872,
      "step": 17450
    },
    {
      "epoch": 2.847540582429236,
      "grad_norm": 0.0010512815788388252,
      "learning_rate": 0.17979486796533517,
      "loss": 0.2646,
      "num_input_tokens_seen": 33591840,
      "step": 17455
    },
    {
      "epoch": 2.848356309650053,
      "grad_norm": 0.004535474348813295,
      "learning_rate": 0.1797371345423797,
      "loss": 0.2693,
      "num_input_tokens_seen": 33601824,
      "step": 17460
    },
    {
      "epoch": 2.84917203687087,
      "grad_norm": 0.0021609370596706867,
      "learning_rate": 0.17967939653358436,
      "loss": 0.219,
      "num_input_tokens_seen": 33611584,
      "step": 17465
    },
    {
      "epoch": 2.8499877640916877,
      "grad_norm": 0.0016322190640494227,
      "learning_rate": 0.17962165394785315,
      "loss": 0.2186,
      "num_input_tokens_seen": 33620448,
      "step": 17470
    },
    {
      "epoch": 2.8508034913125053,
      "grad_norm": 0.0008504090365022421,
      "learning_rate": 0.17956390679409057,
      "loss": 0.2157,
      "num_input_tokens_seen": 33629776,
      "step": 17475
    },
    {
      "epoch": 2.8516192185333225,
      "grad_norm": 0.001917766872793436,
      "learning_rate": 0.1795061550812021,
      "loss": 0.228,
      "num_input_tokens_seen": 33639344,
      "step": 17480
    },
    {
      "epoch": 2.8524349457541396,
      "grad_norm": 0.0013088869163766503,
      "learning_rate": 0.1794483988180937,
      "loss": 0.2242,
      "num_input_tokens_seen": 33649632,
      "step": 17485
    },
    {
      "epoch": 2.853250672974957,
      "grad_norm": 0.0016689611366018653,
      "learning_rate": 0.17939063801367214,
      "loss": 0.1971,
      "num_input_tokens_seen": 33659280,
      "step": 17490
    },
    {
      "epoch": 2.8540664001957747,
      "grad_norm": 0.0017883493565022945,
      "learning_rate": 0.17933287267684483,
      "loss": 0.2075,
      "num_input_tokens_seen": 33668192,
      "step": 17495
    },
    {
      "epoch": 2.854882127416592,
      "grad_norm": 0.0011917833471670747,
      "learning_rate": 0.17927510281651995,
      "loss": 0.192,
      "num_input_tokens_seen": 33678032,
      "step": 17500
    },
    {
      "epoch": 2.855697854637409,
      "grad_norm": 0.0008299138862639666,
      "learning_rate": 0.17921732844160634,
      "loss": 0.1693,
      "num_input_tokens_seen": 33687776,
      "step": 17505
    },
    {
      "epoch": 2.8565135818582266,
      "grad_norm": 0.0011558729456737638,
      "learning_rate": 0.17915954956101351,
      "loss": 0.2397,
      "num_input_tokens_seen": 33696704,
      "step": 17510
    },
    {
      "epoch": 2.857329309079044,
      "grad_norm": 0.0019781470764428377,
      "learning_rate": 0.17910176618365165,
      "loss": 0.2549,
      "num_input_tokens_seen": 33705600,
      "step": 17515
    },
    {
      "epoch": 2.8581450362998613,
      "grad_norm": 0.0014659094158560038,
      "learning_rate": 0.17904397831843177,
      "loss": 0.2452,
      "num_input_tokens_seen": 33716528,
      "step": 17520
    },
    {
      "epoch": 2.858960763520679,
      "grad_norm": 0.0022996561601758003,
      "learning_rate": 0.17898618597426547,
      "loss": 0.2091,
      "num_input_tokens_seen": 33726928,
      "step": 17525
    },
    {
      "epoch": 2.859776490741496,
      "grad_norm": 0.0021069073118269444,
      "learning_rate": 0.17892838916006495,
      "loss": 0.224,
      "num_input_tokens_seen": 33737280,
      "step": 17530
    },
    {
      "epoch": 2.8605922179623136,
      "grad_norm": 0.0028044108767062426,
      "learning_rate": 0.17887058788474333,
      "loss": 0.2223,
      "num_input_tokens_seen": 33746688,
      "step": 17535
    },
    {
      "epoch": 2.8614079451831307,
      "grad_norm": 0.0015556400176137686,
      "learning_rate": 0.17881278215721427,
      "loss": 0.1936,
      "num_input_tokens_seen": 33757568,
      "step": 17540
    },
    {
      "epoch": 2.8622236724039483,
      "grad_norm": 0.001547330291941762,
      "learning_rate": 0.1787549719863921,
      "loss": 0.2017,
      "num_input_tokens_seen": 33766480,
      "step": 17545
    },
    {
      "epoch": 2.8630393996247654,
      "grad_norm": 0.00223367428407073,
      "learning_rate": 0.17869715738119188,
      "loss": 0.2292,
      "num_input_tokens_seen": 33776240,
      "step": 17550
    },
    {
      "epoch": 2.863855126845583,
      "grad_norm": 0.002743834862485528,
      "learning_rate": 0.17863933835052936,
      "loss": 0.1959,
      "num_input_tokens_seen": 33786016,
      "step": 17555
    },
    {
      "epoch": 2.8646708540664,
      "grad_norm": 0.0026975320652127266,
      "learning_rate": 0.17858151490332097,
      "loss": 0.2412,
      "num_input_tokens_seen": 33795088,
      "step": 17560
    },
    {
      "epoch": 2.8654865812872177,
      "grad_norm": 0.0015489913057535887,
      "learning_rate": 0.17852368704848381,
      "loss": 0.1941,
      "num_input_tokens_seen": 33804704,
      "step": 17565
    },
    {
      "epoch": 2.866302308508035,
      "grad_norm": 0.002795763313770294,
      "learning_rate": 0.17846585479493565,
      "loss": 0.211,
      "num_input_tokens_seen": 33815136,
      "step": 17570
    },
    {
      "epoch": 2.8671180357288524,
      "grad_norm": 0.004379333928227425,
      "learning_rate": 0.178408018151595,
      "loss": 0.2484,
      "num_input_tokens_seen": 33825088,
      "step": 17575
    },
    {
      "epoch": 2.8679337629496695,
      "grad_norm": 0.001192976487800479,
      "learning_rate": 0.17835017712738085,
      "loss": 0.206,
      "num_input_tokens_seen": 33835120,
      "step": 17580
    },
    {
      "epoch": 2.868749490170487,
      "grad_norm": 0.0025148247368633747,
      "learning_rate": 0.17829233173121323,
      "loss": 0.2201,
      "num_input_tokens_seen": 33844704,
      "step": 17585
    },
    {
      "epoch": 2.869565217391304,
      "grad_norm": 0.0011573330266401172,
      "learning_rate": 0.17823448197201244,
      "loss": 0.2453,
      "num_input_tokens_seen": 33854896,
      "step": 17590
    },
    {
      "epoch": 2.870380944612122,
      "grad_norm": 0.0014933132333680987,
      "learning_rate": 0.1781766278586997,
      "loss": 0.2025,
      "num_input_tokens_seen": 33864400,
      "step": 17595
    },
    {
      "epoch": 2.871196671832939,
      "grad_norm": 0.0017392766894772649,
      "learning_rate": 0.1781187694001969,
      "loss": 0.2299,
      "num_input_tokens_seen": 33875072,
      "step": 17600
    },
    {
      "epoch": 2.871196671832939,
      "eval_loss": 0.21700520813465118,
      "eval_runtime": 68.0379,
      "eval_samples_per_second": 40.051,
      "eval_steps_per_second": 20.033,
      "num_input_tokens_seen": 33875072,
      "step": 17600
    },
    {
      "epoch": 2.8720123990537565,
      "grad_norm": 0.005784987937659025,
      "learning_rate": 0.1780609066054265,
      "loss": 0.2157,
      "num_input_tokens_seen": 33885840,
      "step": 17605
    },
    {
      "epoch": 2.8728281262745736,
      "grad_norm": 0.0017423508688807487,
      "learning_rate": 0.17800303948331164,
      "loss": 0.2193,
      "num_input_tokens_seen": 33895728,
      "step": 17610
    },
    {
      "epoch": 2.873643853495391,
      "grad_norm": 0.004144952166825533,
      "learning_rate": 0.1779451680427762,
      "loss": 0.2496,
      "num_input_tokens_seen": 33905344,
      "step": 17615
    },
    {
      "epoch": 2.8744595807162083,
      "grad_norm": 0.0017825474496930838,
      "learning_rate": 0.17788729229274464,
      "loss": 0.2177,
      "num_input_tokens_seen": 33916208,
      "step": 17620
    },
    {
      "epoch": 2.875275307937026,
      "grad_norm": 0.0019591732416301966,
      "learning_rate": 0.17782941224214222,
      "loss": 0.2188,
      "num_input_tokens_seen": 33926096,
      "step": 17625
    },
    {
      "epoch": 2.8760910351578435,
      "grad_norm": 0.0013025728985667229,
      "learning_rate": 0.17777152789989464,
      "loss": 0.2115,
      "num_input_tokens_seen": 33935088,
      "step": 17630
    },
    {
      "epoch": 2.8769067623786606,
      "grad_norm": 0.0011246590875089169,
      "learning_rate": 0.17771363927492845,
      "loss": 0.1983,
      "num_input_tokens_seen": 33945120,
      "step": 17635
    },
    {
      "epoch": 2.8777224895994777,
      "grad_norm": 0.002226547570899129,
      "learning_rate": 0.17765574637617085,
      "loss": 0.1923,
      "num_input_tokens_seen": 33953952,
      "step": 17640
    },
    {
      "epoch": 2.8785382168202953,
      "grad_norm": 0.0043559083715081215,
      "learning_rate": 0.17759784921254962,
      "loss": 0.2282,
      "num_input_tokens_seen": 33963072,
      "step": 17645
    },
    {
      "epoch": 2.879353944041113,
      "grad_norm": 0.003200943348929286,
      "learning_rate": 0.1775399477929932,
      "loss": 0.2235,
      "num_input_tokens_seen": 33971104,
      "step": 17650
    },
    {
      "epoch": 2.88016967126193,
      "grad_norm": 0.0016368796350434422,
      "learning_rate": 0.17748204212643076,
      "loss": 0.2433,
      "num_input_tokens_seen": 33980672,
      "step": 17655
    },
    {
      "epoch": 2.880985398482747,
      "grad_norm": 0.0019670617766678333,
      "learning_rate": 0.17742413222179204,
      "loss": 0.2155,
      "num_input_tokens_seen": 33989696,
      "step": 17660
    },
    {
      "epoch": 2.8818011257035647,
      "grad_norm": 0.002042461186647415,
      "learning_rate": 0.17736621808800754,
      "loss": 0.2426,
      "num_input_tokens_seen": 33999120,
      "step": 17665
    },
    {
      "epoch": 2.8826168529243823,
      "grad_norm": 0.001523864921182394,
      "learning_rate": 0.17730829973400827,
      "loss": 0.245,
      "num_input_tokens_seen": 34009328,
      "step": 17670
    },
    {
      "epoch": 2.8834325801451994,
      "grad_norm": 0.002158194547519088,
      "learning_rate": 0.17725037716872602,
      "loss": 0.2219,
      "num_input_tokens_seen": 34019056,
      "step": 17675
    },
    {
      "epoch": 2.8842483073660166,
      "grad_norm": 0.0016399750020354986,
      "learning_rate": 0.17719245040109313,
      "loss": 0.2007,
      "num_input_tokens_seen": 34029472,
      "step": 17680
    },
    {
      "epoch": 2.885064034586834,
      "grad_norm": 0.0008848065626807511,
      "learning_rate": 0.17713451944004271,
      "loss": 0.1845,
      "num_input_tokens_seen": 34038032,
      "step": 17685
    },
    {
      "epoch": 2.8858797618076517,
      "grad_norm": 0.0020648958161473274,
      "learning_rate": 0.17707658429450843,
      "loss": 0.1628,
      "num_input_tokens_seen": 34047968,
      "step": 17690
    },
    {
      "epoch": 2.886695489028469,
      "grad_norm": 0.00106985738966614,
      "learning_rate": 0.1770186449734245,
      "loss": 0.188,
      "num_input_tokens_seen": 34058624,
      "step": 17695
    },
    {
      "epoch": 2.887511216249286,
      "grad_norm": 0.0013809178490191698,
      "learning_rate": 0.17696070148572599,
      "loss": 0.2693,
      "num_input_tokens_seen": 34067248,
      "step": 17700
    },
    {
      "epoch": 2.8883269434701035,
      "grad_norm": 0.001614537206478417,
      "learning_rate": 0.17690275384034856,
      "loss": 0.2536,
      "num_input_tokens_seen": 34077568,
      "step": 17705
    },
    {
      "epoch": 2.889142670690921,
      "grad_norm": 0.001015102374367416,
      "learning_rate": 0.17684480204622835,
      "loss": 0.1889,
      "num_input_tokens_seen": 34087296,
      "step": 17710
    },
    {
      "epoch": 2.8899583979117383,
      "grad_norm": 0.0017767122481018305,
      "learning_rate": 0.1767868461123023,
      "loss": 0.2256,
      "num_input_tokens_seen": 34096576,
      "step": 17715
    },
    {
      "epoch": 2.890774125132556,
      "grad_norm": 0.00490607600659132,
      "learning_rate": 0.176728886047508,
      "loss": 0.2061,
      "num_input_tokens_seen": 34106160,
      "step": 17720
    },
    {
      "epoch": 2.891589852353373,
      "grad_norm": 0.002020158339291811,
      "learning_rate": 0.17667092186078362,
      "loss": 0.2011,
      "num_input_tokens_seen": 34116608,
      "step": 17725
    },
    {
      "epoch": 2.8924055795741905,
      "grad_norm": 0.0016028835671022534,
      "learning_rate": 0.17661295356106785,
      "loss": 0.2506,
      "num_input_tokens_seen": 34127808,
      "step": 17730
    },
    {
      "epoch": 2.8932213067950077,
      "grad_norm": 0.0015758846420794725,
      "learning_rate": 0.1765549811573002,
      "loss": 0.2159,
      "num_input_tokens_seen": 34137280,
      "step": 17735
    },
    {
      "epoch": 2.8940370340158252,
      "grad_norm": 0.0023136024829000235,
      "learning_rate": 0.17649700465842078,
      "loss": 0.1678,
      "num_input_tokens_seen": 34146848,
      "step": 17740
    },
    {
      "epoch": 2.8948527612366424,
      "grad_norm": 0.0010505297686904669,
      "learning_rate": 0.17643902407337023,
      "loss": 0.1875,
      "num_input_tokens_seen": 34155920,
      "step": 17745
    },
    {
      "epoch": 2.89566848845746,
      "grad_norm": 0.0014328764518722892,
      "learning_rate": 0.17638103941108993,
      "loss": 0.2437,
      "num_input_tokens_seen": 34166160,
      "step": 17750
    },
    {
      "epoch": 2.896484215678277,
      "grad_norm": 0.001594556262716651,
      "learning_rate": 0.1763230506805218,
      "loss": 0.231,
      "num_input_tokens_seen": 34175424,
      "step": 17755
    },
    {
      "epoch": 2.8972999428990946,
      "grad_norm": 0.000934412470087409,
      "learning_rate": 0.1762650578906085,
      "loss": 0.1738,
      "num_input_tokens_seen": 34185904,
      "step": 17760
    },
    {
      "epoch": 2.898115670119912,
      "grad_norm": 0.002419042866677046,
      "learning_rate": 0.1762070610502932,
      "loss": 0.2082,
      "num_input_tokens_seen": 34194528,
      "step": 17765
    },
    {
      "epoch": 2.8989313973407294,
      "grad_norm": 0.0014810984721407294,
      "learning_rate": 0.17614906016851975,
      "loss": 0.1812,
      "num_input_tokens_seen": 34204048,
      "step": 17770
    },
    {
      "epoch": 2.8997471245615465,
      "grad_norm": 0.015243827365338802,
      "learning_rate": 0.17609105525423258,
      "loss": 0.2688,
      "num_input_tokens_seen": 34214432,
      "step": 17775
    },
    {
      "epoch": 2.900562851782364,
      "grad_norm": 0.001541021279990673,
      "learning_rate": 0.1760330463163768,
      "loss": 0.1992,
      "num_input_tokens_seen": 34223936,
      "step": 17780
    },
    {
      "epoch": 2.901378579003181,
      "grad_norm": 0.007006360683590174,
      "learning_rate": 0.17597503336389816,
      "loss": 0.2376,
      "num_input_tokens_seen": 34235152,
      "step": 17785
    },
    {
      "epoch": 2.9021943062239988,
      "grad_norm": 0.0011135831009596586,
      "learning_rate": 0.17591701640574298,
      "loss": 0.2595,
      "num_input_tokens_seen": 34244880,
      "step": 17790
    },
    {
      "epoch": 2.903010033444816,
      "grad_norm": 0.00159752881154418,
      "learning_rate": 0.17585899545085815,
      "loss": 0.2149,
      "num_input_tokens_seen": 34255312,
      "step": 17795
    },
    {
      "epoch": 2.9038257606656335,
      "grad_norm": 0.001980178989470005,
      "learning_rate": 0.17580097050819124,
      "loss": 0.2312,
      "num_input_tokens_seen": 34264832,
      "step": 17800
    },
    {
      "epoch": 2.9038257606656335,
      "eval_loss": 0.21594054996967316,
      "eval_runtime": 68.0397,
      "eval_samples_per_second": 40.05,
      "eval_steps_per_second": 20.032,
      "num_input_tokens_seen": 34264832,
      "step": 17800
    },
    {
      "epoch": 2.904641487886451,
      "grad_norm": 0.0015113059198483825,
      "learning_rate": 0.17574294158669046,
      "loss": 0.2105,
      "num_input_tokens_seen": 34273728,
      "step": 17805
    },
    {
      "epoch": 2.905457215107268,
      "grad_norm": 0.002669821958988905,
      "learning_rate": 0.17568490869530456,
      "loss": 0.213,
      "num_input_tokens_seen": 34283168,
      "step": 17810
    },
    {
      "epoch": 2.9062729423280853,
      "grad_norm": 0.003615923458710313,
      "learning_rate": 0.17562687184298295,
      "loss": 0.2625,
      "num_input_tokens_seen": 34293200,
      "step": 17815
    },
    {
      "epoch": 2.907088669548903,
      "grad_norm": 0.001957744127139449,
      "learning_rate": 0.1755688310386757,
      "loss": 0.2344,
      "num_input_tokens_seen": 34302832,
      "step": 17820
    },
    {
      "epoch": 2.9079043967697205,
      "grad_norm": 0.0011203805916011333,
      "learning_rate": 0.17551078629133335,
      "loss": 0.2726,
      "num_input_tokens_seen": 34312304,
      "step": 17825
    },
    {
      "epoch": 2.9087201239905376,
      "grad_norm": 0.001988765550777316,
      "learning_rate": 0.17545273760990718,
      "loss": 0.2212,
      "num_input_tokens_seen": 34322144,
      "step": 17830
    },
    {
      "epoch": 2.9095358512113547,
      "grad_norm": 0.0016566721023991704,
      "learning_rate": 0.17539468500334904,
      "loss": 0.2101,
      "num_input_tokens_seen": 34333024,
      "step": 17835
    },
    {
      "epoch": 2.9103515784321723,
      "grad_norm": 0.0014853376196697354,
      "learning_rate": 0.17533662848061132,
      "loss": 0.2371,
      "num_input_tokens_seen": 34342208,
      "step": 17840
    },
    {
      "epoch": 2.91116730565299,
      "grad_norm": 0.0022609110455960035,
      "learning_rate": 0.1752785680506471,
      "loss": 0.2276,
      "num_input_tokens_seen": 34350992,
      "step": 17845
    },
    {
      "epoch": 2.911983032873807,
      "grad_norm": 0.0016414955025538802,
      "learning_rate": 0.17522050372241,
      "loss": 0.177,
      "num_input_tokens_seen": 34360096,
      "step": 17850
    },
    {
      "epoch": 2.912798760094624,
      "grad_norm": 0.00231834314763546,
      "learning_rate": 0.17516243550485425,
      "loss": 0.2458,
      "num_input_tokens_seen": 34369312,
      "step": 17855
    },
    {
      "epoch": 2.9136144873154417,
      "grad_norm": 0.00283940602093935,
      "learning_rate": 0.17510436340693478,
      "loss": 0.2101,
      "num_input_tokens_seen": 34380880,
      "step": 17860
    },
    {
      "epoch": 2.9144302145362593,
      "grad_norm": 0.001694033737294376,
      "learning_rate": 0.175046287437607,
      "loss": 0.2308,
      "num_input_tokens_seen": 34389408,
      "step": 17865
    },
    {
      "epoch": 2.9152459417570764,
      "grad_norm": 0.0019988229032605886,
      "learning_rate": 0.17498820760582695,
      "loss": 0.2454,
      "num_input_tokens_seen": 34398192,
      "step": 17870
    },
    {
      "epoch": 2.9160616689778935,
      "grad_norm": 0.00234279315918684,
      "learning_rate": 0.1749301239205512,
      "loss": 0.2118,
      "num_input_tokens_seen": 34406608,
      "step": 17875
    },
    {
      "epoch": 2.916877396198711,
      "grad_norm": 0.0009591405978426337,
      "learning_rate": 0.1748720363907371,
      "loss": 0.2153,
      "num_input_tokens_seen": 34415808,
      "step": 17880
    },
    {
      "epoch": 2.9176931234195287,
      "grad_norm": 0.0017712964909151196,
      "learning_rate": 0.17481394502534242,
      "loss": 0.2013,
      "num_input_tokens_seen": 34425088,
      "step": 17885
    },
    {
      "epoch": 2.918508850640346,
      "grad_norm": 0.0045410748571157455,
      "learning_rate": 0.17475584983332562,
      "loss": 0.2522,
      "num_input_tokens_seen": 34434656,
      "step": 17890
    },
    {
      "epoch": 2.919324577861163,
      "grad_norm": 0.0008547070319764316,
      "learning_rate": 0.17469775082364558,
      "loss": 0.2079,
      "num_input_tokens_seen": 34443808,
      "step": 17895
    },
    {
      "epoch": 2.9201403050819805,
      "grad_norm": 0.001075355219654739,
      "learning_rate": 0.17463964800526205,
      "loss": 0.2484,
      "num_input_tokens_seen": 34453296,
      "step": 17900
    },
    {
      "epoch": 2.920956032302798,
      "grad_norm": 0.002146500162780285,
      "learning_rate": 0.17458154138713522,
      "loss": 0.2461,
      "num_input_tokens_seen": 34463312,
      "step": 17905
    },
    {
      "epoch": 2.9217717595236152,
      "grad_norm": 0.0022852199617773294,
      "learning_rate": 0.17452343097822576,
      "loss": 0.2395,
      "num_input_tokens_seen": 34473600,
      "step": 17910
    },
    {
      "epoch": 2.922587486744433,
      "grad_norm": 0.001544229919090867,
      "learning_rate": 0.17446531678749497,
      "loss": 0.2293,
      "num_input_tokens_seen": 34482960,
      "step": 17915
    },
    {
      "epoch": 2.92340321396525,
      "grad_norm": 0.001317153568379581,
      "learning_rate": 0.17440719882390496,
      "loss": 0.2156,
      "num_input_tokens_seen": 34493680,
      "step": 17920
    },
    {
      "epoch": 2.9242189411860675,
      "grad_norm": 0.0009925996419042349,
      "learning_rate": 0.17434907709641814,
      "loss": 0.2203,
      "num_input_tokens_seen": 34502592,
      "step": 17925
    },
    {
      "epoch": 2.9250346684068846,
      "grad_norm": 0.0008598590502515435,
      "learning_rate": 0.17429095161399769,
      "loss": 0.2359,
      "num_input_tokens_seen": 34512176,
      "step": 17930
    },
    {
      "epoch": 2.925850395627702,
      "grad_norm": 0.0018892547814175487,
      "learning_rate": 0.1742328223856072,
      "loss": 0.2177,
      "num_input_tokens_seen": 34522304,
      "step": 17935
    },
    {
      "epoch": 2.9266661228485193,
      "grad_norm": 0.0015164113137871027,
      "learning_rate": 0.174174689420211,
      "loss": 0.2497,
      "num_input_tokens_seen": 34532560,
      "step": 17940
    },
    {
      "epoch": 2.927481850069337,
      "grad_norm": 0.0008817164925858378,
      "learning_rate": 0.1741165527267739,
      "loss": 0.2293,
      "num_input_tokens_seen": 34542880,
      "step": 17945
    },
    {
      "epoch": 2.928297577290154,
      "grad_norm": 0.002548381220549345,
      "learning_rate": 0.17405841231426125,
      "loss": 0.2335,
      "num_input_tokens_seen": 34551792,
      "step": 17950
    },
    {
      "epoch": 2.9291133045109716,
      "grad_norm": 0.0015332413604483008,
      "learning_rate": 0.1740002681916391,
      "loss": 0.2342,
      "num_input_tokens_seen": 34561872,
      "step": 17955
    },
    {
      "epoch": 2.9299290317317888,
      "grad_norm": 0.0007482761866413057,
      "learning_rate": 0.17394212036787401,
      "loss": 0.2035,
      "num_input_tokens_seen": 34572032,
      "step": 17960
    },
    {
      "epoch": 2.9307447589526063,
      "grad_norm": 0.00227640219964087,
      "learning_rate": 0.1738839688519331,
      "loss": 0.2387,
      "num_input_tokens_seen": 34582640,
      "step": 17965
    },
    {
      "epoch": 2.9315604861734235,
      "grad_norm": 0.0028979768976569176,
      "learning_rate": 0.17382581365278402,
      "loss": 0.2264,
      "num_input_tokens_seen": 34592400,
      "step": 17970
    },
    {
      "epoch": 2.932376213394241,
      "grad_norm": 0.0011254095006734133,
      "learning_rate": 0.17376765477939507,
      "loss": 0.2169,
      "num_input_tokens_seen": 34601664,
      "step": 17975
    },
    {
      "epoch": 2.933191940615058,
      "grad_norm": 0.00113213574513793,
      "learning_rate": 0.1737094922407351,
      "loss": 0.2267,
      "num_input_tokens_seen": 34612064,
      "step": 17980
    },
    {
      "epoch": 2.9340076678358757,
      "grad_norm": 0.0008797074551694095,
      "learning_rate": 0.1736513260457734,
      "loss": 0.2338,
      "num_input_tokens_seen": 34621056,
      "step": 17985
    },
    {
      "epoch": 2.934823395056693,
      "grad_norm": 0.0012715220218524337,
      "learning_rate": 0.17359315620348006,
      "loss": 0.2064,
      "num_input_tokens_seen": 34631168,
      "step": 17990
    },
    {
      "epoch": 2.9356391222775104,
      "grad_norm": 0.002177479211241007,
      "learning_rate": 0.17353498272282547,
      "loss": 0.1876,
      "num_input_tokens_seen": 34641488,
      "step": 17995
    },
    {
      "epoch": 2.936454849498328,
      "grad_norm": 0.0011328832479193807,
      "learning_rate": 0.17347680561278087,
      "loss": 0.2082,
      "num_input_tokens_seen": 34652800,
      "step": 18000
    },
    {
      "epoch": 2.936454849498328,
      "eval_loss": 0.21417401731014252,
      "eval_runtime": 68.0109,
      "eval_samples_per_second": 40.067,
      "eval_steps_per_second": 20.041,
      "num_input_tokens_seen": 34652800,
      "step": 18000
    },
    {
      "epoch": 2.937270576719145,
      "grad_norm": 0.0009778564563021064,
      "learning_rate": 0.1734186248823178,
      "loss": 0.1996,
      "num_input_tokens_seen": 34663040,
      "step": 18005
    },
    {
      "epoch": 2.9380863039399623,
      "grad_norm": 0.0013153342297300696,
      "learning_rate": 0.17336044054040844,
      "loss": 0.2225,
      "num_input_tokens_seen": 34672704,
      "step": 18010
    },
    {
      "epoch": 2.93890203116078,
      "grad_norm": 0.001730925403535366,
      "learning_rate": 0.1733022525960256,
      "loss": 0.1656,
      "num_input_tokens_seen": 34682384,
      "step": 18015
    },
    {
      "epoch": 2.9397177583815974,
      "grad_norm": 0.002214886713773012,
      "learning_rate": 0.1732440610581426,
      "loss": 0.2079,
      "num_input_tokens_seen": 34692208,
      "step": 18020
    },
    {
      "epoch": 2.9405334856024146,
      "grad_norm": 0.003442208981141448,
      "learning_rate": 0.17318586593573326,
      "loss": 0.1903,
      "num_input_tokens_seen": 34702368,
      "step": 18025
    },
    {
      "epoch": 2.9413492128232317,
      "grad_norm": 0.001615069922991097,
      "learning_rate": 0.17312766723777204,
      "loss": 0.1838,
      "num_input_tokens_seen": 34711744,
      "step": 18030
    },
    {
      "epoch": 2.9421649400440493,
      "grad_norm": 0.001428495510481298,
      "learning_rate": 0.1730694649732339,
      "loss": 0.2194,
      "num_input_tokens_seen": 34719840,
      "step": 18035
    },
    {
      "epoch": 2.942980667264867,
      "grad_norm": 0.0029137199744582176,
      "learning_rate": 0.17301125915109428,
      "loss": 0.1962,
      "num_input_tokens_seen": 34729680,
      "step": 18040
    },
    {
      "epoch": 2.943796394485684,
      "grad_norm": 0.001517718774266541,
      "learning_rate": 0.17295304978032938,
      "loss": 0.18,
      "num_input_tokens_seen": 34740384,
      "step": 18045
    },
    {
      "epoch": 2.944612121706501,
      "grad_norm": 0.002863585017621517,
      "learning_rate": 0.17289483686991577,
      "loss": 0.213,
      "num_input_tokens_seen": 34750256,
      "step": 18050
    },
    {
      "epoch": 2.9454278489273187,
      "grad_norm": 0.00231438921764493,
      "learning_rate": 0.1728366204288306,
      "loss": 0.1992,
      "num_input_tokens_seen": 34760688,
      "step": 18055
    },
    {
      "epoch": 2.9462435761481363,
      "grad_norm": 0.001642895513214171,
      "learning_rate": 0.17277840046605153,
      "loss": 0.2329,
      "num_input_tokens_seen": 34771152,
      "step": 18060
    },
    {
      "epoch": 2.9470593033689534,
      "grad_norm": 0.0033554094843566418,
      "learning_rate": 0.17272017699055686,
      "loss": 0.2132,
      "num_input_tokens_seen": 34781488,
      "step": 18065
    },
    {
      "epoch": 2.9478750305897705,
      "grad_norm": 0.0014043573755770922,
      "learning_rate": 0.17266195001132542,
      "loss": 0.2132,
      "num_input_tokens_seen": 34790976,
      "step": 18070
    },
    {
      "epoch": 2.948690757810588,
      "grad_norm": 0.0022943285293877125,
      "learning_rate": 0.17260371953733647,
      "loss": 0.2479,
      "num_input_tokens_seen": 34801056,
      "step": 18075
    },
    {
      "epoch": 2.9495064850314057,
      "grad_norm": 0.001395546249113977,
      "learning_rate": 0.1725454855775699,
      "loss": 0.216,
      "num_input_tokens_seen": 34810800,
      "step": 18080
    },
    {
      "epoch": 2.950322212252223,
      "grad_norm": 0.0030391262844204903,
      "learning_rate": 0.17248724814100616,
      "loss": 0.1958,
      "num_input_tokens_seen": 34818880,
      "step": 18085
    },
    {
      "epoch": 2.9511379394730404,
      "grad_norm": 0.0018077908316627145,
      "learning_rate": 0.17242900723662619,
      "loss": 0.1945,
      "num_input_tokens_seen": 34828448,
      "step": 18090
    },
    {
      "epoch": 2.9519536666938575,
      "grad_norm": 0.0011413735337555408,
      "learning_rate": 0.1723707628734114,
      "loss": 0.2233,
      "num_input_tokens_seen": 34835952,
      "step": 18095
    },
    {
      "epoch": 2.952769393914675,
      "grad_norm": 0.0007793944678269327,
      "learning_rate": 0.1723125150603438,
      "loss": 0.2014,
      "num_input_tokens_seen": 34845408,
      "step": 18100
    },
    {
      "epoch": 2.953585121135492,
      "grad_norm": 0.0026246062479913235,
      "learning_rate": 0.1722542638064061,
      "loss": 0.2179,
      "num_input_tokens_seen": 34854768,
      "step": 18105
    },
    {
      "epoch": 2.95440084835631,
      "grad_norm": 0.0019701530691236258,
      "learning_rate": 0.17219600912058117,
      "loss": 0.1854,
      "num_input_tokens_seen": 34864880,
      "step": 18110
    },
    {
      "epoch": 2.955216575577127,
      "grad_norm": 0.002788576064631343,
      "learning_rate": 0.17213775101185272,
      "loss": 0.254,
      "num_input_tokens_seen": 34874096,
      "step": 18115
    },
    {
      "epoch": 2.9560323027979445,
      "grad_norm": 0.003105682088062167,
      "learning_rate": 0.17207948948920485,
      "loss": 0.2257,
      "num_input_tokens_seen": 34883536,
      "step": 18120
    },
    {
      "epoch": 2.9568480300187616,
      "grad_norm": 0.0023685041815042496,
      "learning_rate": 0.17202122456162228,
      "loss": 0.2405,
      "num_input_tokens_seen": 34893984,
      "step": 18125
    },
    {
      "epoch": 2.957663757239579,
      "grad_norm": 0.0009303182596340775,
      "learning_rate": 0.17196295623809013,
      "loss": 0.1883,
      "num_input_tokens_seen": 34902848,
      "step": 18130
    },
    {
      "epoch": 2.9584794844603963,
      "grad_norm": 0.0023828139528632164,
      "learning_rate": 0.1719046845275941,
      "loss": 0.2278,
      "num_input_tokens_seen": 34913040,
      "step": 18135
    },
    {
      "epoch": 2.959295211681214,
      "grad_norm": 0.001210379647091031,
      "learning_rate": 0.17184640943912044,
      "loss": 0.1942,
      "num_input_tokens_seen": 34921600,
      "step": 18140
    },
    {
      "epoch": 2.960110938902031,
      "grad_norm": 0.0022665977012366056,
      "learning_rate": 0.1717881309816559,
      "loss": 0.189,
      "num_input_tokens_seen": 34932160,
      "step": 18145
    },
    {
      "epoch": 2.9609266661228486,
      "grad_norm": 0.0015165311051532626,
      "learning_rate": 0.1717298491641878,
      "loss": 0.2266,
      "num_input_tokens_seen": 34943040,
      "step": 18150
    },
    {
      "epoch": 2.9617423933436657,
      "grad_norm": 0.0015564830973744392,
      "learning_rate": 0.17167156399570385,
      "loss": 0.2263,
      "num_input_tokens_seen": 34952864,
      "step": 18155
    },
    {
      "epoch": 2.9625581205644833,
      "grad_norm": 0.0021294381003826857,
      "learning_rate": 0.17161327548519242,
      "loss": 0.217,
      "num_input_tokens_seen": 34961840,
      "step": 18160
    },
    {
      "epoch": 2.9633738477853004,
      "grad_norm": 0.0023048652801662683,
      "learning_rate": 0.1715549836416423,
      "loss": 0.2202,
      "num_input_tokens_seen": 34971952,
      "step": 18165
    },
    {
      "epoch": 2.964189575006118,
      "grad_norm": 0.004999556113034487,
      "learning_rate": 0.17149668847404279,
      "loss": 0.2323,
      "num_input_tokens_seen": 34982800,
      "step": 18170
    },
    {
      "epoch": 2.965005302226935,
      "grad_norm": 0.0024575081188231707,
      "learning_rate": 0.1714383899913838,
      "loss": 0.2123,
      "num_input_tokens_seen": 34992464,
      "step": 18175
    },
    {
      "epoch": 2.9658210294477527,
      "grad_norm": 0.0020925290882587433,
      "learning_rate": 0.17138008820265563,
      "loss": 0.2476,
      "num_input_tokens_seen": 35001360,
      "step": 18180
    },
    {
      "epoch": 2.96663675666857,
      "grad_norm": 0.0027569704689085484,
      "learning_rate": 0.17132178311684917,
      "loss": 0.2023,
      "num_input_tokens_seen": 35010880,
      "step": 18185
    },
    {
      "epoch": 2.9674524838893874,
      "grad_norm": 0.0026005934923887253,
      "learning_rate": 0.1712634747429559,
      "loss": 0.2287,
      "num_input_tokens_seen": 35019360,
      "step": 18190
    },
    {
      "epoch": 2.968268211110205,
      "grad_norm": 0.001154293306171894,
      "learning_rate": 0.17120516308996753,
      "loss": 0.201,
      "num_input_tokens_seen": 35027104,
      "step": 18195
    },
    {
      "epoch": 2.969083938331022,
      "grad_norm": 0.002345547080039978,
      "learning_rate": 0.17114684816687653,
      "loss": 0.2204,
      "num_input_tokens_seen": 35036144,
      "step": 18200
    },
    {
      "epoch": 2.969083938331022,
      "eval_loss": 0.21233884990215302,
      "eval_runtime": 68.1381,
      "eval_samples_per_second": 39.992,
      "eval_steps_per_second": 20.003,
      "num_input_tokens_seen": 35036144,
      "step": 18200
    },
    {
      "epoch": 2.9698996655518393,
      "grad_norm": 0.002278957050293684,
      "learning_rate": 0.17108852998267585,
      "loss": 0.1959,
      "num_input_tokens_seen": 35044944,
      "step": 18205
    },
    {
      "epoch": 2.970715392772657,
      "grad_norm": 0.0017530155600979924,
      "learning_rate": 0.17103020854635878,
      "loss": 0.2135,
      "num_input_tokens_seen": 35053936,
      "step": 18210
    },
    {
      "epoch": 2.9715311199934744,
      "grad_norm": 0.0026137311942875385,
      "learning_rate": 0.1709718838669193,
      "loss": 0.1816,
      "num_input_tokens_seen": 35064032,
      "step": 18215
    },
    {
      "epoch": 2.9723468472142915,
      "grad_norm": 0.002297322265803814,
      "learning_rate": 0.17091355595335173,
      "loss": 0.2553,
      "num_input_tokens_seen": 35072832,
      "step": 18220
    },
    {
      "epoch": 2.9731625744351087,
      "grad_norm": 0.0016818257281556726,
      "learning_rate": 0.17085522481465107,
      "loss": 0.1914,
      "num_input_tokens_seen": 35082672,
      "step": 18225
    },
    {
      "epoch": 2.9739783016559262,
      "grad_norm": 0.0020088586024940014,
      "learning_rate": 0.17079689045981264,
      "loss": 0.2506,
      "num_input_tokens_seen": 35090848,
      "step": 18230
    },
    {
      "epoch": 2.974794028876744,
      "grad_norm": 0.002252831356599927,
      "learning_rate": 0.17073855289783238,
      "loss": 0.2453,
      "num_input_tokens_seen": 35099712,
      "step": 18235
    },
    {
      "epoch": 2.975609756097561,
      "grad_norm": 0.0011405445402488112,
      "learning_rate": 0.1706802121377066,
      "loss": 0.2368,
      "num_input_tokens_seen": 35108560,
      "step": 18240
    },
    {
      "epoch": 2.976425483318378,
      "grad_norm": 0.002062432933598757,
      "learning_rate": 0.17062186818843225,
      "loss": 0.1849,
      "num_input_tokens_seen": 35117696,
      "step": 18245
    },
    {
      "epoch": 2.9772412105391957,
      "grad_norm": 0.0013027222594246268,
      "learning_rate": 0.17056352105900668,
      "loss": 0.2413,
      "num_input_tokens_seen": 35127344,
      "step": 18250
    },
    {
      "epoch": 2.9780569377600132,
      "grad_norm": 0.00266990321688354,
      "learning_rate": 0.17050517075842772,
      "loss": 0.2305,
      "num_input_tokens_seen": 35135600,
      "step": 18255
    },
    {
      "epoch": 2.9788726649808304,
      "grad_norm": 0.002025169087573886,
      "learning_rate": 0.17044681729569375,
      "loss": 0.2295,
      "num_input_tokens_seen": 35145632,
      "step": 18260
    },
    {
      "epoch": 2.9796883922016475,
      "grad_norm": 0.0007915790774859488,
      "learning_rate": 0.17038846067980365,
      "loss": 0.19,
      "num_input_tokens_seen": 35154512,
      "step": 18265
    },
    {
      "epoch": 2.980504119422465,
      "grad_norm": 0.001415162580087781,
      "learning_rate": 0.17033010091975664,
      "loss": 0.26,
      "num_input_tokens_seen": 35163952,
      "step": 18270
    },
    {
      "epoch": 2.9813198466432826,
      "grad_norm": 0.0034869983792304993,
      "learning_rate": 0.17027173802455262,
      "loss": 0.205,
      "num_input_tokens_seen": 35173440,
      "step": 18275
    },
    {
      "epoch": 2.9821355738640998,
      "grad_norm": 0.0018413116922602057,
      "learning_rate": 0.1702133720031918,
      "loss": 0.2485,
      "num_input_tokens_seen": 35182272,
      "step": 18280
    },
    {
      "epoch": 2.9829513010849174,
      "grad_norm": 0.0009104040800593793,
      "learning_rate": 0.17015500286467503,
      "loss": 0.243,
      "num_input_tokens_seen": 35191200,
      "step": 18285
    },
    {
      "epoch": 2.9837670283057345,
      "grad_norm": 0.002897723577916622,
      "learning_rate": 0.17009663061800354,
      "loss": 0.2452,
      "num_input_tokens_seen": 35201024,
      "step": 18290
    },
    {
      "epoch": 2.984582755526552,
      "grad_norm": 0.001640206202864647,
      "learning_rate": 0.17003825527217903,
      "loss": 0.1887,
      "num_input_tokens_seen": 35212336,
      "step": 18295
    },
    {
      "epoch": 2.985398482747369,
      "grad_norm": 0.0007858994067646563,
      "learning_rate": 0.16997987683620377,
      "loss": 0.2254,
      "num_input_tokens_seen": 35221408,
      "step": 18300
    },
    {
      "epoch": 2.9862142099681868,
      "grad_norm": 0.000638350669760257,
      "learning_rate": 0.16992149531908043,
      "loss": 0.2616,
      "num_input_tokens_seen": 35230096,
      "step": 18305
    },
    {
      "epoch": 2.987029937189004,
      "grad_norm": 0.0022356400731951,
      "learning_rate": 0.16986311072981214,
      "loss": 0.2167,
      "num_input_tokens_seen": 35240832,
      "step": 18310
    },
    {
      "epoch": 2.9878456644098215,
      "grad_norm": 0.002188502112403512,
      "learning_rate": 0.16980472307740255,
      "loss": 0.2439,
      "num_input_tokens_seen": 35250544,
      "step": 18315
    },
    {
      "epoch": 2.9886613916306386,
      "grad_norm": 0.0018137480365112424,
      "learning_rate": 0.1697463323708558,
      "loss": 0.2058,
      "num_input_tokens_seen": 35258048,
      "step": 18320
    },
    {
      "epoch": 2.989477118851456,
      "grad_norm": 0.0027347758878022432,
      "learning_rate": 0.16968793861917641,
      "loss": 0.2685,
      "num_input_tokens_seen": 35269008,
      "step": 18325
    },
    {
      "epoch": 2.9902928460722733,
      "grad_norm": 0.001125926966778934,
      "learning_rate": 0.16962954183136952,
      "loss": 0.2242,
      "num_input_tokens_seen": 35279616,
      "step": 18330
    },
    {
      "epoch": 2.991108573293091,
      "grad_norm": 0.001696562860161066,
      "learning_rate": 0.16957114201644058,
      "loss": 0.2206,
      "num_input_tokens_seen": 35288976,
      "step": 18335
    },
    {
      "epoch": 2.991924300513908,
      "grad_norm": 0.001997327897697687,
      "learning_rate": 0.16951273918339563,
      "loss": 0.2111,
      "num_input_tokens_seen": 35298976,
      "step": 18340
    },
    {
      "epoch": 2.9927400277347256,
      "grad_norm": 0.002779858885332942,
      "learning_rate": 0.16945433334124105,
      "loss": 0.2093,
      "num_input_tokens_seen": 35308832,
      "step": 18345
    },
    {
      "epoch": 2.9935557549555427,
      "grad_norm": 0.0009963704505935311,
      "learning_rate": 0.1693959244989838,
      "loss": 0.2266,
      "num_input_tokens_seen": 35318320,
      "step": 18350
    },
    {
      "epoch": 2.9943714821763603,
      "grad_norm": 0.0013815296115353703,
      "learning_rate": 0.16933751266563127,
      "loss": 0.2344,
      "num_input_tokens_seen": 35328256,
      "step": 18355
    },
    {
      "epoch": 2.9951872093971774,
      "grad_norm": 0.0016113277524709702,
      "learning_rate": 0.16927909785019118,
      "loss": 0.2199,
      "num_input_tokens_seen": 35337984,
      "step": 18360
    },
    {
      "epoch": 2.996002936617995,
      "grad_norm": 0.0016359577421098948,
      "learning_rate": 0.169220680061672,
      "loss": 0.2358,
      "num_input_tokens_seen": 35347136,
      "step": 18365
    },
    {
      "epoch": 2.9968186638388126,
      "grad_norm": 0.0012055765837430954,
      "learning_rate": 0.16916225930908244,
      "loss": 0.2022,
      "num_input_tokens_seen": 35357040,
      "step": 18370
    },
    {
      "epoch": 2.9976343910596297,
      "grad_norm": 0.00296212756074965,
      "learning_rate": 0.16910383560143163,
      "loss": 0.2255,
      "num_input_tokens_seen": 35366304,
      "step": 18375
    },
    {
      "epoch": 2.998450118280447,
      "grad_norm": 0.0016604192787781358,
      "learning_rate": 0.16904540894772935,
      "loss": 0.2275,
      "num_input_tokens_seen": 35374736,
      "step": 18380
    },
    {
      "epoch": 2.9992658455012644,
      "grad_norm": 0.0009313392802141607,
      "learning_rate": 0.16898697935698562,
      "loss": 0.2167,
      "num_input_tokens_seen": 35382944,
      "step": 18385
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.0016325052129104733,
      "learning_rate": 0.1689285468382111,
      "loss": 0.2218,
      "num_input_tokens_seen": 35390256,
      "step": 18390
    },
    {
      "epoch": 3.0008157272208176,
      "grad_norm": 0.0025163975078612566,
      "learning_rate": 0.16887011140041677,
      "loss": 0.2395,
      "num_input_tokens_seen": 35399136,
      "step": 18395
    },
    {
      "epoch": 3.0016314544416347,
      "grad_norm": 0.0013137286296114326,
      "learning_rate": 0.1688116730526141,
      "loss": 0.1839,
      "num_input_tokens_seen": 35410304,
      "step": 18400
    },
    {
      "epoch": 3.0016314544416347,
      "eval_loss": 0.2114049643278122,
      "eval_runtime": 67.9697,
      "eval_samples_per_second": 40.091,
      "eval_steps_per_second": 20.053,
      "num_input_tokens_seen": 35410304,
      "step": 18400
    },
    {
      "epoch": 3.0024471816624523,
      "grad_norm": 0.0013544406974688172,
      "learning_rate": 0.1687532318038151,
      "loss": 0.2342,
      "num_input_tokens_seen": 35420608,
      "step": 18405
    },
    {
      "epoch": 3.0032629088832694,
      "grad_norm": 0.0010983782121911645,
      "learning_rate": 0.16869478766303206,
      "loss": 0.188,
      "num_input_tokens_seen": 35429248,
      "step": 18410
    },
    {
      "epoch": 3.004078636104087,
      "grad_norm": 0.0014278579037636518,
      "learning_rate": 0.16863634063927788,
      "loss": 0.2159,
      "num_input_tokens_seen": 35439968,
      "step": 18415
    },
    {
      "epoch": 3.004894363324904,
      "grad_norm": 0.0022915739100426435,
      "learning_rate": 0.16857789074156568,
      "loss": 0.2,
      "num_input_tokens_seen": 35448848,
      "step": 18420
    },
    {
      "epoch": 3.0057100905457217,
      "grad_norm": 0.0036480913404375315,
      "learning_rate": 0.16851943797890928,
      "loss": 0.2294,
      "num_input_tokens_seen": 35459168,
      "step": 18425
    },
    {
      "epoch": 3.006525817766539,
      "grad_norm": 0.0026118860114365816,
      "learning_rate": 0.16846098236032284,
      "loss": 0.2274,
      "num_input_tokens_seen": 35468368,
      "step": 18430
    },
    {
      "epoch": 3.0073415449873564,
      "grad_norm": 0.002719941781833768,
      "learning_rate": 0.16840252389482097,
      "loss": 0.1828,
      "num_input_tokens_seen": 35478416,
      "step": 18435
    },
    {
      "epoch": 3.0081572722081735,
      "grad_norm": 0.0010591184254735708,
      "learning_rate": 0.16834406259141857,
      "loss": 0.2337,
      "num_input_tokens_seen": 35488432,
      "step": 18440
    },
    {
      "epoch": 3.008972999428991,
      "grad_norm": 0.0014664205955341458,
      "learning_rate": 0.16828559845913124,
      "loss": 0.2104,
      "num_input_tokens_seen": 35498592,
      "step": 18445
    },
    {
      "epoch": 3.0097887266498082,
      "grad_norm": 0.0012851733481511474,
      "learning_rate": 0.16822713150697488,
      "loss": 0.202,
      "num_input_tokens_seen": 35508768,
      "step": 18450
    },
    {
      "epoch": 3.010604453870626,
      "grad_norm": 0.001711056218482554,
      "learning_rate": 0.16816866174396575,
      "loss": 0.2467,
      "num_input_tokens_seen": 35519824,
      "step": 18455
    },
    {
      "epoch": 3.011420181091443,
      "grad_norm": 0.0011756721651181579,
      "learning_rate": 0.16811018917912057,
      "loss": 0.1999,
      "num_input_tokens_seen": 35530592,
      "step": 18460
    },
    {
      "epoch": 3.0122359083122605,
      "grad_norm": 0.0013330234214663506,
      "learning_rate": 0.16805171382145673,
      "loss": 0.2422,
      "num_input_tokens_seen": 35540416,
      "step": 18465
    },
    {
      "epoch": 3.0130516355330776,
      "grad_norm": 0.0018445245223119855,
      "learning_rate": 0.16799323567999175,
      "loss": 0.1864,
      "num_input_tokens_seen": 35550224,
      "step": 18470
    },
    {
      "epoch": 3.013867362753895,
      "grad_norm": 0.0017956820083782077,
      "learning_rate": 0.16793475476374367,
      "loss": 0.2345,
      "num_input_tokens_seen": 35559344,
      "step": 18475
    },
    {
      "epoch": 3.0146830899747123,
      "grad_norm": 0.0033257193863391876,
      "learning_rate": 0.1678762710817311,
      "loss": 0.2111,
      "num_input_tokens_seen": 35569280,
      "step": 18480
    },
    {
      "epoch": 3.01549881719553,
      "grad_norm": 0.0015847304603084922,
      "learning_rate": 0.1678177846429728,
      "loss": 0.2257,
      "num_input_tokens_seen": 35577616,
      "step": 18485
    },
    {
      "epoch": 3.016314544416347,
      "grad_norm": 0.0015450530918315053,
      "learning_rate": 0.16775929545648827,
      "loss": 0.2021,
      "num_input_tokens_seen": 35586976,
      "step": 18490
    },
    {
      "epoch": 3.0171302716371646,
      "grad_norm": 0.0014156096149235964,
      "learning_rate": 0.16770080353129715,
      "loss": 0.1925,
      "num_input_tokens_seen": 35596928,
      "step": 18495
    },
    {
      "epoch": 3.0179459988579818,
      "grad_norm": 0.001713569276034832,
      "learning_rate": 0.16764230887641968,
      "loss": 0.2102,
      "num_input_tokens_seen": 35607296,
      "step": 18500
    },
    {
      "epoch": 3.0187617260787993,
      "grad_norm": 0.0012271995656192303,
      "learning_rate": 0.1675838115008765,
      "loss": 0.2185,
      "num_input_tokens_seen": 35617248,
      "step": 18505
    },
    {
      "epoch": 3.0195774532996165,
      "grad_norm": 0.0011134024243801832,
      "learning_rate": 0.1675253114136886,
      "loss": 0.1783,
      "num_input_tokens_seen": 35628432,
      "step": 18510
    },
    {
      "epoch": 3.020393180520434,
      "grad_norm": 0.0014027685392647982,
      "learning_rate": 0.16746680862387747,
      "loss": 0.2381,
      "num_input_tokens_seen": 35638784,
      "step": 18515
    },
    {
      "epoch": 3.021208907741251,
      "grad_norm": 0.0013070008717477322,
      "learning_rate": 0.16740830314046493,
      "loss": 0.2,
      "num_input_tokens_seen": 35648496,
      "step": 18520
    },
    {
      "epoch": 3.0220246349620687,
      "grad_norm": 0.0015153932617977262,
      "learning_rate": 0.1673497949724733,
      "loss": 0.2147,
      "num_input_tokens_seen": 35658784,
      "step": 18525
    },
    {
      "epoch": 3.022840362182886,
      "grad_norm": 0.0016198281664401293,
      "learning_rate": 0.16729128412892522,
      "loss": 0.1885,
      "num_input_tokens_seen": 35669328,
      "step": 18530
    },
    {
      "epoch": 3.0236560894037035,
      "grad_norm": 0.002441622782498598,
      "learning_rate": 0.16723277061884384,
      "loss": 0.2041,
      "num_input_tokens_seen": 35679200,
      "step": 18535
    },
    {
      "epoch": 3.0244718166245206,
      "grad_norm": 0.0015798413660377264,
      "learning_rate": 0.16717425445125267,
      "loss": 0.2343,
      "num_input_tokens_seen": 35689296,
      "step": 18540
    },
    {
      "epoch": 3.025287543845338,
      "grad_norm": 0.003838025499135256,
      "learning_rate": 0.16711573563517565,
      "loss": 0.2036,
      "num_input_tokens_seen": 35699776,
      "step": 18545
    },
    {
      "epoch": 3.0261032710661553,
      "grad_norm": 0.002844107337296009,
      "learning_rate": 0.1670572141796371,
      "loss": 0.2562,
      "num_input_tokens_seen": 35708800,
      "step": 18550
    },
    {
      "epoch": 3.026918998286973,
      "grad_norm": 0.0017648838693276048,
      "learning_rate": 0.16699869009366175,
      "loss": 0.2436,
      "num_input_tokens_seen": 35718688,
      "step": 18555
    },
    {
      "epoch": 3.02773472550779,
      "grad_norm": 0.002906572073698044,
      "learning_rate": 0.1669401633862748,
      "loss": 0.2083,
      "num_input_tokens_seen": 35727376,
      "step": 18560
    },
    {
      "epoch": 3.0285504527286076,
      "grad_norm": 0.0015821568667888641,
      "learning_rate": 0.16688163406650178,
      "loss": 0.2295,
      "num_input_tokens_seen": 35735664,
      "step": 18565
    },
    {
      "epoch": 3.0293661799494247,
      "grad_norm": 0.0016026005614548922,
      "learning_rate": 0.1668231021433686,
      "loss": 0.1694,
      "num_input_tokens_seen": 35746832,
      "step": 18570
    },
    {
      "epoch": 3.0301819071702423,
      "grad_norm": 0.0019342470914125443,
      "learning_rate": 0.1667645676259017,
      "loss": 0.2048,
      "num_input_tokens_seen": 35757200,
      "step": 18575
    },
    {
      "epoch": 3.03099763439106,
      "grad_norm": 0.0017013148171827197,
      "learning_rate": 0.1667060305231277,
      "loss": 0.1936,
      "num_input_tokens_seen": 35767264,
      "step": 18580
    },
    {
      "epoch": 3.031813361611877,
      "grad_norm": 0.002223425544798374,
      "learning_rate": 0.16664749084407396,
      "loss": 0.2597,
      "num_input_tokens_seen": 35778000,
      "step": 18585
    },
    {
      "epoch": 3.0326290888326946,
      "grad_norm": 0.0015051604714244604,
      "learning_rate": 0.16658894859776788,
      "loss": 0.1493,
      "num_input_tokens_seen": 35788080,
      "step": 18590
    },
    {
      "epoch": 3.0334448160535117,
      "grad_norm": 0.0027134916745126247,
      "learning_rate": 0.16653040379323752,
      "loss": 0.241,
      "num_input_tokens_seen": 35797264,
      "step": 18595
    },
    {
      "epoch": 3.0342605432743293,
      "grad_norm": 0.0032450438011437654,
      "learning_rate": 0.16647185643951107,
      "loss": 0.2317,
      "num_input_tokens_seen": 35808688,
      "step": 18600
    },
    {
      "epoch": 3.0342605432743293,
      "eval_loss": 0.21090947091579437,
      "eval_runtime": 68.0019,
      "eval_samples_per_second": 40.072,
      "eval_steps_per_second": 20.044,
      "num_input_tokens_seen": 35808688,
      "step": 18600
    },
    {
      "epoch": 3.0350762704951464,
      "grad_norm": 0.002695841481909156,
      "learning_rate": 0.1664133065456174,
      "loss": 0.2266,
      "num_input_tokens_seen": 35818112,
      "step": 18605
    },
    {
      "epoch": 3.035891997715964,
      "grad_norm": 0.0013907661195844412,
      "learning_rate": 0.1663547541205856,
      "loss": 0.198,
      "num_input_tokens_seen": 35828208,
      "step": 18610
    },
    {
      "epoch": 3.036707724936781,
      "grad_norm": 0.0021777236834168434,
      "learning_rate": 0.16629619917344518,
      "loss": 0.2173,
      "num_input_tokens_seen": 35837984,
      "step": 18615
    },
    {
      "epoch": 3.0375234521575987,
      "grad_norm": 0.001914534135721624,
      "learning_rate": 0.16623764171322605,
      "loss": 0.2137,
      "num_input_tokens_seen": 35848416,
      "step": 18620
    },
    {
      "epoch": 3.038339179378416,
      "grad_norm": 0.0022773179225623608,
      "learning_rate": 0.1661790817489585,
      "loss": 0.262,
      "num_input_tokens_seen": 35857680,
      "step": 18625
    },
    {
      "epoch": 3.0391549065992334,
      "grad_norm": 0.0013707446632906795,
      "learning_rate": 0.16612051928967328,
      "loss": 0.2458,
      "num_input_tokens_seen": 35867248,
      "step": 18630
    },
    {
      "epoch": 3.0399706338200505,
      "grad_norm": 0.0011089510517194867,
      "learning_rate": 0.16606195434440138,
      "loss": 0.1893,
      "num_input_tokens_seen": 35877440,
      "step": 18635
    },
    {
      "epoch": 3.040786361040868,
      "grad_norm": 0.002139231888577342,
      "learning_rate": 0.16600338692217426,
      "loss": 0.2365,
      "num_input_tokens_seen": 35886640,
      "step": 18640
    },
    {
      "epoch": 3.041602088261685,
      "grad_norm": 0.0023586789611727,
      "learning_rate": 0.16594481703202374,
      "loss": 0.1997,
      "num_input_tokens_seen": 35896096,
      "step": 18645
    },
    {
      "epoch": 3.042417815482503,
      "grad_norm": 0.0012967456132173538,
      "learning_rate": 0.1658862446829821,
      "loss": 0.1828,
      "num_input_tokens_seen": 35907264,
      "step": 18650
    },
    {
      "epoch": 3.04323354270332,
      "grad_norm": 0.001718170358799398,
      "learning_rate": 0.16582766988408187,
      "loss": 0.1907,
      "num_input_tokens_seen": 35916656,
      "step": 18655
    },
    {
      "epoch": 3.0440492699241375,
      "grad_norm": 0.0016045435331761837,
      "learning_rate": 0.16576909264435608,
      "loss": 0.2332,
      "num_input_tokens_seen": 35926448,
      "step": 18660
    },
    {
      "epoch": 3.0448649971449546,
      "grad_norm": 0.0009217344922944903,
      "learning_rate": 0.16571051297283798,
      "loss": 0.2123,
      "num_input_tokens_seen": 35935488,
      "step": 18665
    },
    {
      "epoch": 3.045680724365772,
      "grad_norm": 0.001000239746645093,
      "learning_rate": 0.16565193087856137,
      "loss": 0.2016,
      "num_input_tokens_seen": 35944400,
      "step": 18670
    },
    {
      "epoch": 3.0464964515865893,
      "grad_norm": 0.0013463415671139956,
      "learning_rate": 0.16559334637056033,
      "loss": 0.194,
      "num_input_tokens_seen": 35955536,
      "step": 18675
    },
    {
      "epoch": 3.047312178807407,
      "grad_norm": 0.0015671865548938513,
      "learning_rate": 0.16553475945786933,
      "loss": 0.1934,
      "num_input_tokens_seen": 35965744,
      "step": 18680
    },
    {
      "epoch": 3.048127906028224,
      "grad_norm": 0.0009147902019321918,
      "learning_rate": 0.16547617014952318,
      "loss": 0.2059,
      "num_input_tokens_seen": 35976560,
      "step": 18685
    },
    {
      "epoch": 3.0489436332490416,
      "grad_norm": 0.0007877497700974345,
      "learning_rate": 0.1654175784545571,
      "loss": 0.2114,
      "num_input_tokens_seen": 35987312,
      "step": 18690
    },
    {
      "epoch": 3.0497593604698587,
      "grad_norm": 0.0013469418045133352,
      "learning_rate": 0.1653589843820067,
      "loss": 0.1702,
      "num_input_tokens_seen": 35996928,
      "step": 18695
    },
    {
      "epoch": 3.0505750876906763,
      "grad_norm": 0.0023176574613898993,
      "learning_rate": 0.1653003879409079,
      "loss": 0.2322,
      "num_input_tokens_seen": 36005888,
      "step": 18700
    },
    {
      "epoch": 3.0513908149114934,
      "grad_norm": 0.0024234780576080084,
      "learning_rate": 0.165241789140297,
      "loss": 0.2396,
      "num_input_tokens_seen": 36016432,
      "step": 18705
    },
    {
      "epoch": 3.052206542132311,
      "grad_norm": 0.002107165986672044,
      "learning_rate": 0.16518318798921064,
      "loss": 0.2779,
      "num_input_tokens_seen": 36026160,
      "step": 18710
    },
    {
      "epoch": 3.053022269353128,
      "grad_norm": 0.0016503770602867007,
      "learning_rate": 0.16512458449668593,
      "loss": 0.2203,
      "num_input_tokens_seen": 36035360,
      "step": 18715
    },
    {
      "epoch": 3.0538379965739457,
      "grad_norm": 0.0019219809910282493,
      "learning_rate": 0.1650659786717602,
      "loss": 0.215,
      "num_input_tokens_seen": 36046400,
      "step": 18720
    },
    {
      "epoch": 3.054653723794763,
      "grad_norm": 0.0011962801218032837,
      "learning_rate": 0.1650073705234712,
      "loss": 0.2076,
      "num_input_tokens_seen": 36054640,
      "step": 18725
    },
    {
      "epoch": 3.0554694510155804,
      "grad_norm": 0.0033327313140034676,
      "learning_rate": 0.16494876006085712,
      "loss": 0.1883,
      "num_input_tokens_seen": 36064480,
      "step": 18730
    },
    {
      "epoch": 3.0562851782363976,
      "grad_norm": 0.003052078653126955,
      "learning_rate": 0.16489014729295634,
      "loss": 0.1939,
      "num_input_tokens_seen": 36073936,
      "step": 18735
    },
    {
      "epoch": 3.057100905457215,
      "grad_norm": 0.0014414074830710888,
      "learning_rate": 0.16483153222880775,
      "loss": 0.224,
      "num_input_tokens_seen": 36084240,
      "step": 18740
    },
    {
      "epoch": 3.0579166326780323,
      "grad_norm": 0.001367082237266004,
      "learning_rate": 0.16477291487745052,
      "loss": 0.2052,
      "num_input_tokens_seen": 36093728,
      "step": 18745
    },
    {
      "epoch": 3.05873235989885,
      "grad_norm": 0.0017730274703353643,
      "learning_rate": 0.16471429524792416,
      "loss": 0.2291,
      "num_input_tokens_seen": 36103456,
      "step": 18750
    },
    {
      "epoch": 3.059548087119667,
      "grad_norm": 0.0020593712106347084,
      "learning_rate": 0.16465567334926856,
      "loss": 0.1998,
      "num_input_tokens_seen": 36113440,
      "step": 18755
    },
    {
      "epoch": 3.0603638143404845,
      "grad_norm": 0.0017459584632888436,
      "learning_rate": 0.16459704919052395,
      "loss": 0.2026,
      "num_input_tokens_seen": 36123728,
      "step": 18760
    },
    {
      "epoch": 3.0611795415613017,
      "grad_norm": 0.0014882755931466818,
      "learning_rate": 0.16453842278073086,
      "loss": 0.2034,
      "num_input_tokens_seen": 36133328,
      "step": 18765
    },
    {
      "epoch": 3.0619952687821193,
      "grad_norm": 0.0015442105941474438,
      "learning_rate": 0.16447979412893038,
      "loss": 0.2584,
      "num_input_tokens_seen": 36141952,
      "step": 18770
    },
    {
      "epoch": 3.062810996002937,
      "grad_norm": 0.0014373993035405874,
      "learning_rate": 0.16442116324416367,
      "loss": 0.2056,
      "num_input_tokens_seen": 36152192,
      "step": 18775
    },
    {
      "epoch": 3.063626723223754,
      "grad_norm": 0.002540124347433448,
      "learning_rate": 0.1643625301354723,
      "loss": 0.2076,
      "num_input_tokens_seen": 36160992,
      "step": 18780
    },
    {
      "epoch": 3.0644424504445715,
      "grad_norm": 0.0013089417479932308,
      "learning_rate": 0.16430389481189828,
      "loss": 0.2237,
      "num_input_tokens_seen": 36171088,
      "step": 18785
    },
    {
      "epoch": 3.0652581776653887,
      "grad_norm": 0.0018977487925440073,
      "learning_rate": 0.164245257282484,
      "loss": 0.204,
      "num_input_tokens_seen": 36180192,
      "step": 18790
    },
    {
      "epoch": 3.0660739048862062,
      "grad_norm": 0.0014828291023150086,
      "learning_rate": 0.16418661755627195,
      "loss": 0.2652,
      "num_input_tokens_seen": 36190672,
      "step": 18795
    },
    {
      "epoch": 3.0668896321070234,
      "grad_norm": 0.0032480843365192413,
      "learning_rate": 0.16412797564230527,
      "loss": 0.2439,
      "num_input_tokens_seen": 36200720,
      "step": 18800
    },
    {
      "epoch": 3.0668896321070234,
      "eval_loss": 0.22333426773548126,
      "eval_runtime": 67.9864,
      "eval_samples_per_second": 40.082,
      "eval_steps_per_second": 20.048,
      "num_input_tokens_seen": 36200720,
      "step": 18800
    },
    {
      "epoch": 3.067705359327841,
      "grad_norm": 0.001720596686936915,
      "learning_rate": 0.16406933154962713,
      "loss": 0.2135,
      "num_input_tokens_seen": 36210592,
      "step": 18805
    },
    {
      "epoch": 3.068521086548658,
      "grad_norm": 0.0014865194680169225,
      "learning_rate": 0.16401068528728133,
      "loss": 0.2095,
      "num_input_tokens_seen": 36219040,
      "step": 18810
    },
    {
      "epoch": 3.0693368137694756,
      "grad_norm": 0.0014198004500940442,
      "learning_rate": 0.16395203686431173,
      "loss": 0.2637,
      "num_input_tokens_seen": 36228224,
      "step": 18815
    },
    {
      "epoch": 3.0701525409902928,
      "grad_norm": 0.0014659457374364138,
      "learning_rate": 0.16389338628976277,
      "loss": 0.2324,
      "num_input_tokens_seen": 36238160,
      "step": 18820
    },
    {
      "epoch": 3.0709682682111104,
      "grad_norm": 0.001970804063603282,
      "learning_rate": 0.163834733572679,
      "loss": 0.2292,
      "num_input_tokens_seen": 36247760,
      "step": 18825
    },
    {
      "epoch": 3.0717839954319275,
      "grad_norm": 0.0007974112522788346,
      "learning_rate": 0.16377607872210545,
      "loss": 0.2242,
      "num_input_tokens_seen": 36255584,
      "step": 18830
    },
    {
      "epoch": 3.072599722652745,
      "grad_norm": 0.0017302477499470115,
      "learning_rate": 0.16371742174708748,
      "loss": 0.2115,
      "num_input_tokens_seen": 36264112,
      "step": 18835
    },
    {
      "epoch": 3.073415449873562,
      "grad_norm": 0.0016615563072264194,
      "learning_rate": 0.16365876265667065,
      "loss": 0.198,
      "num_input_tokens_seen": 36273664,
      "step": 18840
    },
    {
      "epoch": 3.0742311770943798,
      "grad_norm": 0.001537351869046688,
      "learning_rate": 0.163600101459901,
      "loss": 0.2339,
      "num_input_tokens_seen": 36284384,
      "step": 18845
    },
    {
      "epoch": 3.075046904315197,
      "grad_norm": 0.0009147286764346063,
      "learning_rate": 0.16354143816582484,
      "loss": 0.2093,
      "num_input_tokens_seen": 36294464,
      "step": 18850
    },
    {
      "epoch": 3.0758626315360145,
      "grad_norm": 0.0012454780517145991,
      "learning_rate": 0.1634827727834887,
      "loss": 0.2007,
      "num_input_tokens_seen": 36303424,
      "step": 18855
    },
    {
      "epoch": 3.0766783587568316,
      "grad_norm": 0.0022274162620306015,
      "learning_rate": 0.16342410532193954,
      "loss": 0.1991,
      "num_input_tokens_seen": 36312240,
      "step": 18860
    },
    {
      "epoch": 3.077494085977649,
      "grad_norm": 0.002462883247062564,
      "learning_rate": 0.16336543579022464,
      "loss": 0.194,
      "num_input_tokens_seen": 36321456,
      "step": 18865
    },
    {
      "epoch": 3.0783098131984663,
      "grad_norm": 0.0022676687221974134,
      "learning_rate": 0.16330676419739157,
      "loss": 0.1873,
      "num_input_tokens_seen": 36331296,
      "step": 18870
    },
    {
      "epoch": 3.079125540419284,
      "grad_norm": 0.001768338494002819,
      "learning_rate": 0.1632480905524883,
      "loss": 0.1876,
      "num_input_tokens_seen": 36341344,
      "step": 18875
    },
    {
      "epoch": 3.079941267640101,
      "grad_norm": 0.001967884600162506,
      "learning_rate": 0.16318941486456293,
      "loss": 0.1906,
      "num_input_tokens_seen": 36349552,
      "step": 18880
    },
    {
      "epoch": 3.0807569948609186,
      "grad_norm": 0.0024701727088540792,
      "learning_rate": 0.16313073714266405,
      "loss": 0.1607,
      "num_input_tokens_seen": 36358880,
      "step": 18885
    },
    {
      "epoch": 3.0815727220817357,
      "grad_norm": 0.0019925010856240988,
      "learning_rate": 0.16307205739584052,
      "loss": 0.1838,
      "num_input_tokens_seen": 36368576,
      "step": 18890
    },
    {
      "epoch": 3.0823884493025533,
      "grad_norm": 0.003696218365803361,
      "learning_rate": 0.16301337563314144,
      "loss": 0.2449,
      "num_input_tokens_seen": 36376832,
      "step": 18895
    },
    {
      "epoch": 3.0832041765233704,
      "grad_norm": 0.0014140476705506444,
      "learning_rate": 0.1629546918636163,
      "loss": 0.2404,
      "num_input_tokens_seen": 36387712,
      "step": 18900
    },
    {
      "epoch": 3.084019903744188,
      "grad_norm": 0.002034059725701809,
      "learning_rate": 0.16289600609631485,
      "loss": 0.2032,
      "num_input_tokens_seen": 36397456,
      "step": 18905
    },
    {
      "epoch": 3.084835630965005,
      "grad_norm": 0.00260501098819077,
      "learning_rate": 0.16283731834028722,
      "loss": 0.2179,
      "num_input_tokens_seen": 36406944,
      "step": 18910
    },
    {
      "epoch": 3.0856513581858227,
      "grad_norm": 0.0015970285749062896,
      "learning_rate": 0.16277862860458378,
      "loss": 0.2109,
      "num_input_tokens_seen": 36416864,
      "step": 18915
    },
    {
      "epoch": 3.08646708540664,
      "grad_norm": 0.0025856385473161936,
      "learning_rate": 0.16271993689825526,
      "loss": 0.2114,
      "num_input_tokens_seen": 36426880,
      "step": 18920
    },
    {
      "epoch": 3.0872828126274574,
      "grad_norm": 0.003033213084563613,
      "learning_rate": 0.1626612432303526,
      "loss": 0.2619,
      "num_input_tokens_seen": 36435840,
      "step": 18925
    },
    {
      "epoch": 3.0880985398482745,
      "grad_norm": 0.0022944696247577667,
      "learning_rate": 0.1626025476099271,
      "loss": 0.193,
      "num_input_tokens_seen": 36444672,
      "step": 18930
    },
    {
      "epoch": 3.088914267069092,
      "grad_norm": 0.0027386427391320467,
      "learning_rate": 0.1625438500460304,
      "loss": 0.2417,
      "num_input_tokens_seen": 36453744,
      "step": 18935
    },
    {
      "epoch": 3.0897299942899092,
      "grad_norm": 0.0022174471523612738,
      "learning_rate": 0.16248515054771442,
      "loss": 0.2349,
      "num_input_tokens_seen": 36464016,
      "step": 18940
    },
    {
      "epoch": 3.090545721510727,
      "grad_norm": 0.0011402794625610113,
      "learning_rate": 0.16242644912403123,
      "loss": 0.2498,
      "num_input_tokens_seen": 36472800,
      "step": 18945
    },
    {
      "epoch": 3.0913614487315444,
      "grad_norm": 0.0011020752135664225,
      "learning_rate": 0.1623677457840335,
      "loss": 0.2072,
      "num_input_tokens_seen": 36482672,
      "step": 18950
    },
    {
      "epoch": 3.0921771759523615,
      "grad_norm": 0.0028815839905291796,
      "learning_rate": 0.16230904053677397,
      "loss": 0.2196,
      "num_input_tokens_seen": 36493216,
      "step": 18955
    },
    {
      "epoch": 3.092992903173179,
      "grad_norm": 0.0023103856947273016,
      "learning_rate": 0.16225033339130568,
      "loss": 0.2698,
      "num_input_tokens_seen": 36501840,
      "step": 18960
    },
    {
      "epoch": 3.0938086303939962,
      "grad_norm": 0.0022042449563741684,
      "learning_rate": 0.16219162435668197,
      "loss": 0.1905,
      "num_input_tokens_seen": 36511456,
      "step": 18965
    },
    {
      "epoch": 3.094624357614814,
      "grad_norm": 0.0009444722672924399,
      "learning_rate": 0.16213291344195666,
      "loss": 0.1753,
      "num_input_tokens_seen": 36520640,
      "step": 18970
    },
    {
      "epoch": 3.095440084835631,
      "grad_norm": 0.0011191093362867832,
      "learning_rate": 0.16207420065618358,
      "loss": 0.2497,
      "num_input_tokens_seen": 36530608,
      "step": 18975
    },
    {
      "epoch": 3.0962558120564485,
      "grad_norm": 0.00260208104737103,
      "learning_rate": 0.16201548600841706,
      "loss": 0.2274,
      "num_input_tokens_seen": 36541360,
      "step": 18980
    },
    {
      "epoch": 3.0970715392772656,
      "grad_norm": 0.001327218604274094,
      "learning_rate": 0.16195676950771154,
      "loss": 0.2315,
      "num_input_tokens_seen": 36550992,
      "step": 18985
    },
    {
      "epoch": 3.097887266498083,
      "grad_norm": 0.0014053519116714597,
      "learning_rate": 0.16189805116312198,
      "loss": 0.1945,
      "num_input_tokens_seen": 36560656,
      "step": 18990
    },
    {
      "epoch": 3.0987029937189003,
      "grad_norm": 0.001720819273032248,
      "learning_rate": 0.16183933098370337,
      "loss": 0.2662,
      "num_input_tokens_seen": 36570736,
      "step": 18995
    },
    {
      "epoch": 3.099518720939718,
      "grad_norm": 0.0012807623716071248,
      "learning_rate": 0.16178060897851115,
      "loss": 0.2262,
      "num_input_tokens_seen": 36580112,
      "step": 19000
    },
    {
      "epoch": 3.099518720939718,
      "eval_loss": 0.22773709893226624,
      "eval_runtime": 68.0504,
      "eval_samples_per_second": 40.044,
      "eval_steps_per_second": 20.029,
      "num_input_tokens_seen": 36580112,
      "step": 19000
    },
    {
      "epoch": 3.100334448160535,
      "grad_norm": 0.0013605847489088774,
      "learning_rate": 0.16172188515660096,
      "loss": 0.195,
      "num_input_tokens_seen": 36590096,
      "step": 19005
    },
    {
      "epoch": 3.1011501753813526,
      "grad_norm": 0.0018952673999592662,
      "learning_rate": 0.16166315952702878,
      "loss": 0.2551,
      "num_input_tokens_seen": 36599936,
      "step": 19010
    },
    {
      "epoch": 3.1019659026021698,
      "grad_norm": 0.0006726362626068294,
      "learning_rate": 0.16160443209885084,
      "loss": 0.2351,
      "num_input_tokens_seen": 36609280,
      "step": 19015
    },
    {
      "epoch": 3.1027816298229873,
      "grad_norm": 0.0012013331288471818,
      "learning_rate": 0.16154570288112363,
      "loss": 0.1994,
      "num_input_tokens_seen": 36618704,
      "step": 19020
    },
    {
      "epoch": 3.1035973570438045,
      "grad_norm": 0.0019523447845131159,
      "learning_rate": 0.16148697188290395,
      "loss": 0.1837,
      "num_input_tokens_seen": 36628144,
      "step": 19025
    },
    {
      "epoch": 3.104413084264622,
      "grad_norm": 0.0017585901077836752,
      "learning_rate": 0.16142823911324888,
      "loss": 0.1796,
      "num_input_tokens_seen": 36638704,
      "step": 19030
    },
    {
      "epoch": 3.105228811485439,
      "grad_norm": 0.0010224205907434225,
      "learning_rate": 0.16136950458121568,
      "loss": 0.1928,
      "num_input_tokens_seen": 36647680,
      "step": 19035
    },
    {
      "epoch": 3.1060445387062567,
      "grad_norm": 0.0025815069675445557,
      "learning_rate": 0.16131076829586205,
      "loss": 0.2042,
      "num_input_tokens_seen": 36659360,
      "step": 19040
    },
    {
      "epoch": 3.106860265927074,
      "grad_norm": 0.0016992364544421434,
      "learning_rate": 0.1612520302662457,
      "loss": 0.1862,
      "num_input_tokens_seen": 36668752,
      "step": 19045
    },
    {
      "epoch": 3.1076759931478914,
      "grad_norm": 0.0009512357064522803,
      "learning_rate": 0.16119329050142497,
      "loss": 0.2102,
      "num_input_tokens_seen": 36678480,
      "step": 19050
    },
    {
      "epoch": 3.1084917203687086,
      "grad_norm": 0.0033373055048286915,
      "learning_rate": 0.16113454901045818,
      "loss": 0.2265,
      "num_input_tokens_seen": 36686720,
      "step": 19055
    },
    {
      "epoch": 3.109307447589526,
      "grad_norm": 0.001617291709408164,
      "learning_rate": 0.16107580580240397,
      "loss": 0.2247,
      "num_input_tokens_seen": 36695920,
      "step": 19060
    },
    {
      "epoch": 3.1101231748103433,
      "grad_norm": 0.0020774949807673693,
      "learning_rate": 0.16101706088632134,
      "loss": 0.2739,
      "num_input_tokens_seen": 36706064,
      "step": 19065
    },
    {
      "epoch": 3.110938902031161,
      "grad_norm": 0.0016694347141310573,
      "learning_rate": 0.16095831427126947,
      "loss": 0.2311,
      "num_input_tokens_seen": 36714176,
      "step": 19070
    },
    {
      "epoch": 3.111754629251978,
      "grad_norm": 0.001985488459467888,
      "learning_rate": 0.16089956596630783,
      "loss": 0.1702,
      "num_input_tokens_seen": 36724848,
      "step": 19075
    },
    {
      "epoch": 3.1125703564727956,
      "grad_norm": 0.0015100226737558842,
      "learning_rate": 0.16084081598049618,
      "loss": 0.2578,
      "num_input_tokens_seen": 36733984,
      "step": 19080
    },
    {
      "epoch": 3.1133860836936127,
      "grad_norm": 0.0008078489918261766,
      "learning_rate": 0.1607820643228944,
      "loss": 0.2264,
      "num_input_tokens_seen": 36742496,
      "step": 19085
    },
    {
      "epoch": 3.1142018109144303,
      "grad_norm": 0.0017966253217309713,
      "learning_rate": 0.16072331100256285,
      "loss": 0.1683,
      "num_input_tokens_seen": 36752368,
      "step": 19090
    },
    {
      "epoch": 3.1150175381352474,
      "grad_norm": 0.0029999776743352413,
      "learning_rate": 0.16066455602856197,
      "loss": 0.2302,
      "num_input_tokens_seen": 36763344,
      "step": 19095
    },
    {
      "epoch": 3.115833265356065,
      "grad_norm": 0.0018429687479510903,
      "learning_rate": 0.16060579940995257,
      "loss": 0.1858,
      "num_input_tokens_seen": 36771520,
      "step": 19100
    },
    {
      "epoch": 3.116648992576882,
      "grad_norm": 0.0015178713947534561,
      "learning_rate": 0.16054704115579557,
      "loss": 0.1992,
      "num_input_tokens_seen": 36781312,
      "step": 19105
    },
    {
      "epoch": 3.1174647197976997,
      "grad_norm": 0.0015341660473495722,
      "learning_rate": 0.1604882812751523,
      "loss": 0.1982,
      "num_input_tokens_seen": 36789984,
      "step": 19110
    },
    {
      "epoch": 3.118280447018517,
      "grad_norm": 0.002091852715238929,
      "learning_rate": 0.16042951977708425,
      "loss": 0.2102,
      "num_input_tokens_seen": 36799520,
      "step": 19115
    },
    {
      "epoch": 3.1190961742393344,
      "grad_norm": 0.0009678913047537208,
      "learning_rate": 0.16037075667065318,
      "loss": 0.1895,
      "num_input_tokens_seen": 36809568,
      "step": 19120
    },
    {
      "epoch": 3.1199119014601515,
      "grad_norm": 0.0020883555989712477,
      "learning_rate": 0.1603119919649211,
      "loss": 0.2022,
      "num_input_tokens_seen": 36819632,
      "step": 19125
    },
    {
      "epoch": 3.120727628680969,
      "grad_norm": 0.0012708815047517419,
      "learning_rate": 0.16025322566895028,
      "loss": 0.1949,
      "num_input_tokens_seen": 36829568,
      "step": 19130
    },
    {
      "epoch": 3.121543355901786,
      "grad_norm": 0.0009403364383615553,
      "learning_rate": 0.16019445779180322,
      "loss": 0.2027,
      "num_input_tokens_seen": 36839328,
      "step": 19135
    },
    {
      "epoch": 3.122359083122604,
      "grad_norm": 0.0027840237598866224,
      "learning_rate": 0.16013568834254271,
      "loss": 0.2096,
      "num_input_tokens_seen": 36848672,
      "step": 19140
    },
    {
      "epoch": 3.1231748103434214,
      "grad_norm": 0.0016453778371214867,
      "learning_rate": 0.1600769173302316,
      "loss": 0.2411,
      "num_input_tokens_seen": 36858624,
      "step": 19145
    },
    {
      "epoch": 3.1239905375642385,
      "grad_norm": 0.0026903969701379538,
      "learning_rate": 0.16001814476393322,
      "loss": 0.2819,
      "num_input_tokens_seen": 36867728,
      "step": 19150
    },
    {
      "epoch": 3.124806264785056,
      "grad_norm": 0.0012022304581478238,
      "learning_rate": 0.15995937065271104,
      "loss": 0.2306,
      "num_input_tokens_seen": 36876832,
      "step": 19155
    },
    {
      "epoch": 3.125621992005873,
      "grad_norm": 0.001213929965160787,
      "learning_rate": 0.15990059500562873,
      "loss": 0.1945,
      "num_input_tokens_seen": 36886944,
      "step": 19160
    },
    {
      "epoch": 3.126437719226691,
      "grad_norm": 0.0012446569744497538,
      "learning_rate": 0.15984181783175025,
      "loss": 0.1999,
      "num_input_tokens_seen": 36895472,
      "step": 19165
    },
    {
      "epoch": 3.127253446447508,
      "grad_norm": 0.0011956588132306933,
      "learning_rate": 0.1597830391401398,
      "loss": 0.1924,
      "num_input_tokens_seen": 36904912,
      "step": 19170
    },
    {
      "epoch": 3.1280691736683255,
      "grad_norm": 0.0029981585685163736,
      "learning_rate": 0.15972425893986178,
      "loss": 0.2343,
      "num_input_tokens_seen": 36915424,
      "step": 19175
    },
    {
      "epoch": 3.1288849008891426,
      "grad_norm": 0.0016796898562461138,
      "learning_rate": 0.15966547723998084,
      "loss": 0.2102,
      "num_input_tokens_seen": 36924240,
      "step": 19180
    },
    {
      "epoch": 3.12970062810996,
      "grad_norm": 0.0014058140804991126,
      "learning_rate": 0.15960669404956176,
      "loss": 0.2437,
      "num_input_tokens_seen": 36934448,
      "step": 19185
    },
    {
      "epoch": 3.1305163553307773,
      "grad_norm": 0.001517944852821529,
      "learning_rate": 0.1595479093776698,
      "loss": 0.1965,
      "num_input_tokens_seen": 36943664,
      "step": 19190
    },
    {
      "epoch": 3.131332082551595,
      "grad_norm": 0.0021051696967333555,
      "learning_rate": 0.15948912323337022,
      "loss": 0.2269,
      "num_input_tokens_seen": 36954064,
      "step": 19195
    },
    {
      "epoch": 3.132147809772412,
      "grad_norm": 0.000929211382754147,
      "learning_rate": 0.1594303356257286,
      "loss": 0.2039,
      "num_input_tokens_seen": 36961872,
      "step": 19200
    },
    {
      "epoch": 3.132147809772412,
      "eval_loss": 0.2111775279045105,
      "eval_runtime": 68.0599,
      "eval_samples_per_second": 40.038,
      "eval_steps_per_second": 20.026,
      "num_input_tokens_seen": 36961872,
      "step": 19200
    },
    {
      "epoch": 3.1329635369932296,
      "grad_norm": 0.0011451981263235211,
      "learning_rate": 0.15937154656381072,
      "loss": 0.235,
      "num_input_tokens_seen": 36971328,
      "step": 19205
    },
    {
      "epoch": 3.1337792642140467,
      "grad_norm": 0.0026138005778193474,
      "learning_rate": 0.15931275605668258,
      "loss": 0.2243,
      "num_input_tokens_seen": 36980224,
      "step": 19210
    },
    {
      "epoch": 3.1345949914348643,
      "grad_norm": 0.0013999691000208259,
      "learning_rate": 0.1592539641134104,
      "loss": 0.2031,
      "num_input_tokens_seen": 36989024,
      "step": 19215
    },
    {
      "epoch": 3.1354107186556814,
      "grad_norm": 0.002257116138935089,
      "learning_rate": 0.1591951707430607,
      "loss": 0.1853,
      "num_input_tokens_seen": 36997968,
      "step": 19220
    },
    {
      "epoch": 3.136226445876499,
      "grad_norm": 0.0013350935187190771,
      "learning_rate": 0.15913637595470007,
      "loss": 0.2055,
      "num_input_tokens_seen": 37007136,
      "step": 19225
    },
    {
      "epoch": 3.137042173097316,
      "grad_norm": 0.0011454278137534857,
      "learning_rate": 0.15907757975739548,
      "loss": 0.2571,
      "num_input_tokens_seen": 37017264,
      "step": 19230
    },
    {
      "epoch": 3.1378579003181337,
      "grad_norm": 0.0012841176940128207,
      "learning_rate": 0.159018782160214,
      "loss": 0.2124,
      "num_input_tokens_seen": 37026880,
      "step": 19235
    },
    {
      "epoch": 3.138673627538951,
      "grad_norm": 0.0010296845575794578,
      "learning_rate": 0.158959983172223,
      "loss": 0.2115,
      "num_input_tokens_seen": 37035984,
      "step": 19240
    },
    {
      "epoch": 3.1394893547597684,
      "grad_norm": 0.0023049383889883757,
      "learning_rate": 0.15890118280249,
      "loss": 0.2141,
      "num_input_tokens_seen": 37047760,
      "step": 19245
    },
    {
      "epoch": 3.1403050819805856,
      "grad_norm": 0.0016069597331807017,
      "learning_rate": 0.15884238106008275,
      "loss": 0.2231,
      "num_input_tokens_seen": 37058144,
      "step": 19250
    },
    {
      "epoch": 3.141120809201403,
      "grad_norm": 0.00391291081905365,
      "learning_rate": 0.15878357795406922,
      "loss": 0.2145,
      "num_input_tokens_seen": 37067760,
      "step": 19255
    },
    {
      "epoch": 3.1419365364222203,
      "grad_norm": 0.0017060709651559591,
      "learning_rate": 0.15872477349351757,
      "loss": 0.2114,
      "num_input_tokens_seen": 37076384,
      "step": 19260
    },
    {
      "epoch": 3.142752263643038,
      "grad_norm": 0.0018249345012009144,
      "learning_rate": 0.15866596768749622,
      "loss": 0.1987,
      "num_input_tokens_seen": 37085968,
      "step": 19265
    },
    {
      "epoch": 3.143567990863855,
      "grad_norm": 0.0014775737654417753,
      "learning_rate": 0.15860716054507373,
      "loss": 0.2344,
      "num_input_tokens_seen": 37096128,
      "step": 19270
    },
    {
      "epoch": 3.1443837180846725,
      "grad_norm": 0.0034423270262777805,
      "learning_rate": 0.1585483520753189,
      "loss": 0.2325,
      "num_input_tokens_seen": 37105664,
      "step": 19275
    },
    {
      "epoch": 3.1451994453054897,
      "grad_norm": 0.0017568813636898994,
      "learning_rate": 0.1584895422873008,
      "loss": 0.1735,
      "num_input_tokens_seen": 37114688,
      "step": 19280
    },
    {
      "epoch": 3.1460151725263072,
      "grad_norm": 0.0019378842553123832,
      "learning_rate": 0.1584307311900886,
      "loss": 0.2388,
      "num_input_tokens_seen": 37124416,
      "step": 19285
    },
    {
      "epoch": 3.1468308997471244,
      "grad_norm": 0.0010423099156469107,
      "learning_rate": 0.1583719187927517,
      "loss": 0.1879,
      "num_input_tokens_seen": 37133680,
      "step": 19290
    },
    {
      "epoch": 3.147646626967942,
      "grad_norm": 0.002301889006048441,
      "learning_rate": 0.15831310510435967,
      "loss": 0.2178,
      "num_input_tokens_seen": 37144144,
      "step": 19295
    },
    {
      "epoch": 3.148462354188759,
      "grad_norm": 0.0011316945310682058,
      "learning_rate": 0.15825429013398243,
      "loss": 0.1605,
      "num_input_tokens_seen": 37154560,
      "step": 19300
    },
    {
      "epoch": 3.1492780814095767,
      "grad_norm": 0.0016140806255862117,
      "learning_rate": 0.15819547389068986,
      "loss": 0.2121,
      "num_input_tokens_seen": 37164848,
      "step": 19305
    },
    {
      "epoch": 3.150093808630394,
      "grad_norm": 0.0008479186217300594,
      "learning_rate": 0.1581366563835522,
      "loss": 0.1859,
      "num_input_tokens_seen": 37176000,
      "step": 19310
    },
    {
      "epoch": 3.1509095358512114,
      "grad_norm": 0.0015130023239180446,
      "learning_rate": 0.15807783762163993,
      "loss": 0.204,
      "num_input_tokens_seen": 37185680,
      "step": 19315
    },
    {
      "epoch": 3.151725263072029,
      "grad_norm": 0.0010766852647066116,
      "learning_rate": 0.15801901761402365,
      "loss": 0.2153,
      "num_input_tokens_seen": 37195600,
      "step": 19320
    },
    {
      "epoch": 3.152540990292846,
      "grad_norm": 0.0034756700042635202,
      "learning_rate": 0.157960196369774,
      "loss": 0.2495,
      "num_input_tokens_seen": 37204464,
      "step": 19325
    },
    {
      "epoch": 3.153356717513663,
      "grad_norm": 0.0017927419394254684,
      "learning_rate": 0.157901373897962,
      "loss": 0.2224,
      "num_input_tokens_seen": 37213360,
      "step": 19330
    },
    {
      "epoch": 3.1541724447344808,
      "grad_norm": 0.0013738729758188128,
      "learning_rate": 0.15784255020765892,
      "loss": 0.2145,
      "num_input_tokens_seen": 37222608,
      "step": 19335
    },
    {
      "epoch": 3.1549881719552983,
      "grad_norm": 0.0016474603908136487,
      "learning_rate": 0.157783725307936,
      "loss": 0.2338,
      "num_input_tokens_seen": 37233072,
      "step": 19340
    },
    {
      "epoch": 3.1558038991761155,
      "grad_norm": 0.0025015694554895163,
      "learning_rate": 0.15772489920786484,
      "loss": 0.2223,
      "num_input_tokens_seen": 37241984,
      "step": 19345
    },
    {
      "epoch": 3.156619626396933,
      "grad_norm": 0.0013316357508301735,
      "learning_rate": 0.15766607191651713,
      "loss": 0.1749,
      "num_input_tokens_seen": 37250832,
      "step": 19350
    },
    {
      "epoch": 3.15743535361775,
      "grad_norm": 0.001680137007497251,
      "learning_rate": 0.1576072434429648,
      "loss": 0.239,
      "num_input_tokens_seen": 37260608,
      "step": 19355
    },
    {
      "epoch": 3.1582510808385678,
      "grad_norm": 0.0018525540363043547,
      "learning_rate": 0.15754841379627998,
      "loss": 0.2094,
      "num_input_tokens_seen": 37271056,
      "step": 19360
    },
    {
      "epoch": 3.159066808059385,
      "grad_norm": 0.0008110274211503565,
      "learning_rate": 0.15748958298553484,
      "loss": 0.1889,
      "num_input_tokens_seen": 37280864,
      "step": 19365
    },
    {
      "epoch": 3.1598825352802025,
      "grad_norm": 0.0013752977829426527,
      "learning_rate": 0.1574307510198019,
      "loss": 0.2672,
      "num_input_tokens_seen": 37290592,
      "step": 19370
    },
    {
      "epoch": 3.1606982625010196,
      "grad_norm": 0.0008746447274461389,
      "learning_rate": 0.15737191790815375,
      "loss": 0.2035,
      "num_input_tokens_seen": 37299136,
      "step": 19375
    },
    {
      "epoch": 3.161513989721837,
      "grad_norm": 0.001190099399536848,
      "learning_rate": 0.15731308365966323,
      "loss": 0.2679,
      "num_input_tokens_seen": 37308272,
      "step": 19380
    },
    {
      "epoch": 3.1623297169426543,
      "grad_norm": 0.002366551198065281,
      "learning_rate": 0.15725424828340331,
      "loss": 0.2478,
      "num_input_tokens_seen": 37317696,
      "step": 19385
    },
    {
      "epoch": 3.163145444163472,
      "grad_norm": 0.001023756223730743,
      "learning_rate": 0.15719541178844715,
      "loss": 0.2129,
      "num_input_tokens_seen": 37327312,
      "step": 19390
    },
    {
      "epoch": 3.163961171384289,
      "grad_norm": 0.003129407297819853,
      "learning_rate": 0.15713657418386806,
      "loss": 0.2187,
      "num_input_tokens_seen": 37336384,
      "step": 19395
    },
    {
      "epoch": 3.1647768986051066,
      "grad_norm": 0.000990825006738305,
      "learning_rate": 0.15707773547873957,
      "loss": 0.2283,
      "num_input_tokens_seen": 37345136,
      "step": 19400
    },
    {
      "epoch": 3.1647768986051066,
      "eval_loss": 0.2210145890712738,
      "eval_runtime": 68.1132,
      "eval_samples_per_second": 40.007,
      "eval_steps_per_second": 20.011,
      "num_input_tokens_seen": 37345136,
      "step": 19400
    },
    {
      "epoch": 3.1655926258259237,
      "grad_norm": 0.0021987631916999817,
      "learning_rate": 0.1570188956821353,
      "loss": 0.2095,
      "num_input_tokens_seen": 37354864,
      "step": 19405
    },
    {
      "epoch": 3.1664083530467413,
      "grad_norm": 0.0021960113663226366,
      "learning_rate": 0.1569600548031291,
      "loss": 0.1886,
      "num_input_tokens_seen": 37365408,
      "step": 19410
    },
    {
      "epoch": 3.1672240802675584,
      "grad_norm": 0.00216636061668396,
      "learning_rate": 0.156901212850795,
      "loss": 0.2039,
      "num_input_tokens_seen": 37375952,
      "step": 19415
    },
    {
      "epoch": 3.168039807488376,
      "grad_norm": 0.00153292715549469,
      "learning_rate": 0.15684236983420716,
      "loss": 0.2274,
      "num_input_tokens_seen": 37385808,
      "step": 19420
    },
    {
      "epoch": 3.168855534709193,
      "grad_norm": 0.0008734974544495344,
      "learning_rate": 0.1567835257624399,
      "loss": 0.1708,
      "num_input_tokens_seen": 37395808,
      "step": 19425
    },
    {
      "epoch": 3.1696712619300107,
      "grad_norm": 0.0009743308764882386,
      "learning_rate": 0.1567246806445677,
      "loss": 0.2182,
      "num_input_tokens_seen": 37404464,
      "step": 19430
    },
    {
      "epoch": 3.170486989150828,
      "grad_norm": 0.0014398063067346811,
      "learning_rate": 0.15666583448966526,
      "loss": 0.2003,
      "num_input_tokens_seen": 37414128,
      "step": 19435
    },
    {
      "epoch": 3.1713027163716454,
      "grad_norm": 0.0017300012987107038,
      "learning_rate": 0.1566069873068074,
      "loss": 0.2416,
      "num_input_tokens_seen": 37424240,
      "step": 19440
    },
    {
      "epoch": 3.1721184435924625,
      "grad_norm": 0.0013004770735278726,
      "learning_rate": 0.156548139105069,
      "loss": 0.2005,
      "num_input_tokens_seen": 37433696,
      "step": 19445
    },
    {
      "epoch": 3.17293417081328,
      "grad_norm": 0.0011301711201667786,
      "learning_rate": 0.15648928989352529,
      "loss": 0.1822,
      "num_input_tokens_seen": 37443392,
      "step": 19450
    },
    {
      "epoch": 3.1737498980340972,
      "grad_norm": 0.0012441580183804035,
      "learning_rate": 0.15643043968125156,
      "loss": 0.2667,
      "num_input_tokens_seen": 37453968,
      "step": 19455
    },
    {
      "epoch": 3.174565625254915,
      "grad_norm": 0.0013890027767047286,
      "learning_rate": 0.15637158847732316,
      "loss": 0.2725,
      "num_input_tokens_seen": 37462336,
      "step": 19460
    },
    {
      "epoch": 3.175381352475732,
      "grad_norm": 0.001593310502357781,
      "learning_rate": 0.15631273629081582,
      "loss": 0.1971,
      "num_input_tokens_seen": 37472944,
      "step": 19465
    },
    {
      "epoch": 3.1761970796965495,
      "grad_norm": 0.00110615452285856,
      "learning_rate": 0.15625388313080518,
      "loss": 0.1907,
      "num_input_tokens_seen": 37482736,
      "step": 19470
    },
    {
      "epoch": 3.1770128069173666,
      "grad_norm": 0.0010562891839072108,
      "learning_rate": 0.15619502900636714,
      "loss": 0.2057,
      "num_input_tokens_seen": 37491872,
      "step": 19475
    },
    {
      "epoch": 3.1778285341381842,
      "grad_norm": 0.002713182708248496,
      "learning_rate": 0.15613617392657783,
      "loss": 0.2397,
      "num_input_tokens_seen": 37502096,
      "step": 19480
    },
    {
      "epoch": 3.1786442613590014,
      "grad_norm": 0.0018152502598240972,
      "learning_rate": 0.15607731790051335,
      "loss": 0.2283,
      "num_input_tokens_seen": 37512032,
      "step": 19485
    },
    {
      "epoch": 3.179459988579819,
      "grad_norm": 0.0009757653460837901,
      "learning_rate": 0.15601846093725008,
      "loss": 0.1984,
      "num_input_tokens_seen": 37522656,
      "step": 19490
    },
    {
      "epoch": 3.180275715800636,
      "grad_norm": 0.0017714952118694782,
      "learning_rate": 0.1559596030458645,
      "loss": 0.2117,
      "num_input_tokens_seen": 37531840,
      "step": 19495
    },
    {
      "epoch": 3.1810914430214536,
      "grad_norm": 0.0010082261869683862,
      "learning_rate": 0.1559007442354333,
      "loss": 0.1752,
      "num_input_tokens_seen": 37540736,
      "step": 19500
    },
    {
      "epoch": 3.1819071702422708,
      "grad_norm": 0.0011613998794928193,
      "learning_rate": 0.15584188451503314,
      "loss": 0.2377,
      "num_input_tokens_seen": 37549248,
      "step": 19505
    },
    {
      "epoch": 3.1827228974630883,
      "grad_norm": 0.0009230644791387022,
      "learning_rate": 0.15578302389374094,
      "loss": 0.2084,
      "num_input_tokens_seen": 37558640,
      "step": 19510
    },
    {
      "epoch": 3.183538624683906,
      "grad_norm": 0.0013183675473555923,
      "learning_rate": 0.1557241623806338,
      "loss": 0.2275,
      "num_input_tokens_seen": 37568416,
      "step": 19515
    },
    {
      "epoch": 3.184354351904723,
      "grad_norm": 0.001616958063095808,
      "learning_rate": 0.15566529998478887,
      "loss": 0.1991,
      "num_input_tokens_seen": 37577808,
      "step": 19520
    },
    {
      "epoch": 3.1851700791255406,
      "grad_norm": 0.0012913800310343504,
      "learning_rate": 0.15560643671528354,
      "loss": 0.1871,
      "num_input_tokens_seen": 37588768,
      "step": 19525
    },
    {
      "epoch": 3.1859858063463578,
      "grad_norm": 0.002204278716817498,
      "learning_rate": 0.15554757258119514,
      "loss": 0.2271,
      "num_input_tokens_seen": 37598560,
      "step": 19530
    },
    {
      "epoch": 3.1868015335671753,
      "grad_norm": 0.0016794799594208598,
      "learning_rate": 0.1554887075916014,
      "loss": 0.2451,
      "num_input_tokens_seen": 37608992,
      "step": 19535
    },
    {
      "epoch": 3.1876172607879925,
      "grad_norm": 0.0024394665379077196,
      "learning_rate": 0.15542984175558,
      "loss": 0.1913,
      "num_input_tokens_seen": 37619648,
      "step": 19540
    },
    {
      "epoch": 3.18843298800881,
      "grad_norm": 0.0014528570463880897,
      "learning_rate": 0.1553709750822087,
      "loss": 0.209,
      "num_input_tokens_seen": 37630352,
      "step": 19545
    },
    {
      "epoch": 3.189248715229627,
      "grad_norm": 0.002244244795292616,
      "learning_rate": 0.15531210758056554,
      "loss": 0.1943,
      "num_input_tokens_seen": 37639440,
      "step": 19550
    },
    {
      "epoch": 3.1900644424504447,
      "grad_norm": 0.0022622805554419756,
      "learning_rate": 0.15525323925972867,
      "loss": 0.2032,
      "num_input_tokens_seen": 37649872,
      "step": 19555
    },
    {
      "epoch": 3.190880169671262,
      "grad_norm": 0.0019277931423857808,
      "learning_rate": 0.15519437012877627,
      "loss": 0.207,
      "num_input_tokens_seen": 37659088,
      "step": 19560
    },
    {
      "epoch": 3.1916958968920794,
      "grad_norm": 0.0014765936648473144,
      "learning_rate": 0.15513550019678676,
      "loss": 0.2416,
      "num_input_tokens_seen": 37668080,
      "step": 19565
    },
    {
      "epoch": 3.1925116241128966,
      "grad_norm": 0.004002448171377182,
      "learning_rate": 0.15507662947283854,
      "loss": 0.2342,
      "num_input_tokens_seen": 37678032,
      "step": 19570
    },
    {
      "epoch": 3.193327351333714,
      "grad_norm": 0.0021076530683785677,
      "learning_rate": 0.15501775796601028,
      "loss": 0.2137,
      "num_input_tokens_seen": 37687728,
      "step": 19575
    },
    {
      "epoch": 3.1941430785545313,
      "grad_norm": 0.0015540332533419132,
      "learning_rate": 0.15495888568538066,
      "loss": 0.1707,
      "num_input_tokens_seen": 37697104,
      "step": 19580
    },
    {
      "epoch": 3.194958805775349,
      "grad_norm": 0.0017472004983574152,
      "learning_rate": 0.1549000126400286,
      "loss": 0.2515,
      "num_input_tokens_seen": 37705536,
      "step": 19585
    },
    {
      "epoch": 3.195774532996166,
      "grad_norm": 0.0008607034105807543,
      "learning_rate": 0.15484113883903294,
      "loss": 0.1893,
      "num_input_tokens_seen": 37714944,
      "step": 19590
    },
    {
      "epoch": 3.1965902602169836,
      "grad_norm": 0.0014295667642727494,
      "learning_rate": 0.15478226429147288,
      "loss": 0.1699,
      "num_input_tokens_seen": 37724512,
      "step": 19595
    },
    {
      "epoch": 3.1974059874378007,
      "grad_norm": 0.0013379165902733803,
      "learning_rate": 0.15472338900642757,
      "loss": 0.1909,
      "num_input_tokens_seen": 37732992,
      "step": 19600
    },
    {
      "epoch": 3.1974059874378007,
      "eval_loss": 0.209974005818367,
      "eval_runtime": 68.023,
      "eval_samples_per_second": 40.06,
      "eval_steps_per_second": 20.037,
      "num_input_tokens_seen": 37732992,
      "step": 19600
    },
    {
      "epoch": 3.1982217146586183,
      "grad_norm": 0.000978569034487009,
      "learning_rate": 0.15466451299297632,
      "loss": 0.1541,
      "num_input_tokens_seen": 37743120,
      "step": 19605
    },
    {
      "epoch": 3.1990374418794354,
      "grad_norm": 0.001615013345144689,
      "learning_rate": 0.15460563626019852,
      "loss": 0.1622,
      "num_input_tokens_seen": 37750736,
      "step": 19610
    },
    {
      "epoch": 3.199853169100253,
      "grad_norm": 0.0012680906802415848,
      "learning_rate": 0.15454675881717375,
      "loss": 0.2664,
      "num_input_tokens_seen": 37761600,
      "step": 19615
    },
    {
      "epoch": 3.20066889632107,
      "grad_norm": 0.0015257054474204779,
      "learning_rate": 0.1544878806729816,
      "loss": 0.2315,
      "num_input_tokens_seen": 37771328,
      "step": 19620
    },
    {
      "epoch": 3.2014846235418877,
      "grad_norm": 0.0013908121036365628,
      "learning_rate": 0.1544290018367019,
      "loss": 0.1983,
      "num_input_tokens_seen": 37781104,
      "step": 19625
    },
    {
      "epoch": 3.202300350762705,
      "grad_norm": 0.0019112771842628717,
      "learning_rate": 0.15437012231741445,
      "loss": 0.2662,
      "num_input_tokens_seen": 37791424,
      "step": 19630
    },
    {
      "epoch": 3.2031160779835224,
      "grad_norm": 0.0034713330678641796,
      "learning_rate": 0.1543112421241992,
      "loss": 0.2411,
      "num_input_tokens_seen": 37800160,
      "step": 19635
    },
    {
      "epoch": 3.2039318052043395,
      "grad_norm": 0.001728718401864171,
      "learning_rate": 0.15425236126613626,
      "loss": 0.2236,
      "num_input_tokens_seen": 37810064,
      "step": 19640
    },
    {
      "epoch": 3.204747532425157,
      "grad_norm": 0.0014902795664966106,
      "learning_rate": 0.15419347975230577,
      "loss": 0.2033,
      "num_input_tokens_seen": 37818368,
      "step": 19645
    },
    {
      "epoch": 3.205563259645974,
      "grad_norm": 0.00143267004750669,
      "learning_rate": 0.154134597591788,
      "loss": 0.2375,
      "num_input_tokens_seen": 37828192,
      "step": 19650
    },
    {
      "epoch": 3.206378986866792,
      "grad_norm": 0.0012845293385908008,
      "learning_rate": 0.1540757147936633,
      "loss": 0.199,
      "num_input_tokens_seen": 37836416,
      "step": 19655
    },
    {
      "epoch": 3.207194714087609,
      "grad_norm": 0.0019915865268558264,
      "learning_rate": 0.1540168313670122,
      "loss": 0.2037,
      "num_input_tokens_seen": 37846736,
      "step": 19660
    },
    {
      "epoch": 3.2080104413084265,
      "grad_norm": 0.0010099589126184583,
      "learning_rate": 0.1539579473209152,
      "loss": 0.1869,
      "num_input_tokens_seen": 37855920,
      "step": 19665
    },
    {
      "epoch": 3.2088261685292436,
      "grad_norm": 0.0019241663394495845,
      "learning_rate": 0.15389906266445294,
      "loss": 0.2141,
      "num_input_tokens_seen": 37865632,
      "step": 19670
    },
    {
      "epoch": 3.209641895750061,
      "grad_norm": 0.001613806700333953,
      "learning_rate": 0.15384017740670627,
      "loss": 0.2127,
      "num_input_tokens_seen": 37874336,
      "step": 19675
    },
    {
      "epoch": 3.2104576229708783,
      "grad_norm": 0.002813864266499877,
      "learning_rate": 0.15378129155675602,
      "loss": 0.2825,
      "num_input_tokens_seen": 37885264,
      "step": 19680
    },
    {
      "epoch": 3.211273350191696,
      "grad_norm": 0.0025438619777560234,
      "learning_rate": 0.15372240512368307,
      "loss": 0.2445,
      "num_input_tokens_seen": 37894240,
      "step": 19685
    },
    {
      "epoch": 3.2120890774125135,
      "grad_norm": 0.002131744986400008,
      "learning_rate": 0.1536635181165684,
      "loss": 0.2332,
      "num_input_tokens_seen": 37905600,
      "step": 19690
    },
    {
      "epoch": 3.2129048046333306,
      "grad_norm": 0.002107860054820776,
      "learning_rate": 0.15360463054449328,
      "loss": 0.2001,
      "num_input_tokens_seen": 37914800,
      "step": 19695
    },
    {
      "epoch": 3.2137205318541477,
      "grad_norm": 0.0009695342741906643,
      "learning_rate": 0.1535457424165388,
      "loss": 0.2113,
      "num_input_tokens_seen": 37923520,
      "step": 19700
    },
    {
      "epoch": 3.2145362590749653,
      "grad_norm": 0.0020159673877060413,
      "learning_rate": 0.15348685374178628,
      "loss": 0.2087,
      "num_input_tokens_seen": 37933072,
      "step": 19705
    },
    {
      "epoch": 3.215351986295783,
      "grad_norm": 0.00153746095020324,
      "learning_rate": 0.1534279645293171,
      "loss": 0.2595,
      "num_input_tokens_seen": 37943440,
      "step": 19710
    },
    {
      "epoch": 3.2161677135166,
      "grad_norm": 0.0011682750191539526,
      "learning_rate": 0.1533690747882127,
      "loss": 0.2149,
      "num_input_tokens_seen": 37952928,
      "step": 19715
    },
    {
      "epoch": 3.2169834407374176,
      "grad_norm": 0.0016794831026345491,
      "learning_rate": 0.15331018452755465,
      "loss": 0.2113,
      "num_input_tokens_seen": 37963232,
      "step": 19720
    },
    {
      "epoch": 3.2177991679582347,
      "grad_norm": 0.0016613155603408813,
      "learning_rate": 0.15325129375642457,
      "loss": 0.1943,
      "num_input_tokens_seen": 37972864,
      "step": 19725
    },
    {
      "epoch": 3.2186148951790523,
      "grad_norm": 0.0010838160524144769,
      "learning_rate": 0.15319240248390406,
      "loss": 0.1939,
      "num_input_tokens_seen": 37982896,
      "step": 19730
    },
    {
      "epoch": 3.2194306223998694,
      "grad_norm": 0.0013180329697206616,
      "learning_rate": 0.153133510719075,
      "loss": 0.1965,
      "num_input_tokens_seen": 37992720,
      "step": 19735
    },
    {
      "epoch": 3.220246349620687,
      "grad_norm": 0.0016124889953061938,
      "learning_rate": 0.15307461847101922,
      "loss": 0.1923,
      "num_input_tokens_seen": 38003840,
      "step": 19740
    },
    {
      "epoch": 3.221062076841504,
      "grad_norm": 0.0014730901457369328,
      "learning_rate": 0.15301572574881864,
      "loss": 0.2133,
      "num_input_tokens_seen": 38014336,
      "step": 19745
    },
    {
      "epoch": 3.2218778040623217,
      "grad_norm": 0.0029623019509017467,
      "learning_rate": 0.15295683256155523,
      "loss": 0.2145,
      "num_input_tokens_seen": 38024048,
      "step": 19750
    },
    {
      "epoch": 3.222693531283139,
      "grad_norm": 0.0013423438649624586,
      "learning_rate": 0.15289793891831113,
      "loss": 0.2155,
      "num_input_tokens_seen": 38033440,
      "step": 19755
    },
    {
      "epoch": 3.2235092585039564,
      "grad_norm": 0.001830191700719297,
      "learning_rate": 0.15283904482816837,
      "loss": 0.2053,
      "num_input_tokens_seen": 38042592,
      "step": 19760
    },
    {
      "epoch": 3.2243249857247736,
      "grad_norm": 0.002725588157773018,
      "learning_rate": 0.15278015030020928,
      "loss": 0.2057,
      "num_input_tokens_seen": 38051200,
      "step": 19765
    },
    {
      "epoch": 3.225140712945591,
      "grad_norm": 0.0010763065656647086,
      "learning_rate": 0.152721255343516,
      "loss": 0.2316,
      "num_input_tokens_seen": 38059200,
      "step": 19770
    },
    {
      "epoch": 3.2259564401664083,
      "grad_norm": 0.0007833443814888597,
      "learning_rate": 0.15266235996717098,
      "loss": 0.1563,
      "num_input_tokens_seen": 38068848,
      "step": 19775
    },
    {
      "epoch": 3.226772167387226,
      "grad_norm": 0.002639548387378454,
      "learning_rate": 0.15260346418025664,
      "loss": 0.2039,
      "num_input_tokens_seen": 38079632,
      "step": 19780
    },
    {
      "epoch": 3.227587894608043,
      "grad_norm": 0.0016184259438887239,
      "learning_rate": 0.15254456799185537,
      "loss": 0.2199,
      "num_input_tokens_seen": 38089344,
      "step": 19785
    },
    {
      "epoch": 3.2284036218288605,
      "grad_norm": 0.0015563126653432846,
      "learning_rate": 0.15248567141104974,
      "loss": 0.2504,
      "num_input_tokens_seen": 38099200,
      "step": 19790
    },
    {
      "epoch": 3.2292193490496777,
      "grad_norm": 0.0014778685290366411,
      "learning_rate": 0.15242677444692232,
      "loss": 0.228,
      "num_input_tokens_seen": 38108160,
      "step": 19795
    },
    {
      "epoch": 3.2300350762704952,
      "grad_norm": 0.0014542903518304229,
      "learning_rate": 0.15236787710855584,
      "loss": 0.2207,
      "num_input_tokens_seen": 38118784,
      "step": 19800
    },
    {
      "epoch": 3.2300350762704952,
      "eval_loss": 0.21590077877044678,
      "eval_runtime": 67.8969,
      "eval_samples_per_second": 40.134,
      "eval_steps_per_second": 20.075,
      "num_input_tokens_seen": 38118784,
      "step": 19800
    },
    {
      "epoch": 3.2308508034913124,
      "grad_norm": 0.0008913842611946166,
      "learning_rate": 0.1523089794050329,
      "loss": 0.1907,
      "num_input_tokens_seen": 38127328,
      "step": 19805
    },
    {
      "epoch": 3.23166653071213,
      "grad_norm": 0.003312137210741639,
      "learning_rate": 0.15225008134543633,
      "loss": 0.2488,
      "num_input_tokens_seen": 38136864,
      "step": 19810
    },
    {
      "epoch": 3.232482257932947,
      "grad_norm": 0.0020457576029002666,
      "learning_rate": 0.15219118293884895,
      "loss": 0.211,
      "num_input_tokens_seen": 38146304,
      "step": 19815
    },
    {
      "epoch": 3.2332979851537647,
      "grad_norm": 0.0011189701035618782,
      "learning_rate": 0.15213228419435362,
      "loss": 0.1988,
      "num_input_tokens_seen": 38155520,
      "step": 19820
    },
    {
      "epoch": 3.234113712374582,
      "grad_norm": 0.0033267084509134293,
      "learning_rate": 0.15207338512103327,
      "loss": 0.2375,
      "num_input_tokens_seen": 38164896,
      "step": 19825
    },
    {
      "epoch": 3.2349294395953994,
      "grad_norm": 0.0015624924562871456,
      "learning_rate": 0.1520144857279709,
      "loss": 0.2165,
      "num_input_tokens_seen": 38174736,
      "step": 19830
    },
    {
      "epoch": 3.2357451668162165,
      "grad_norm": 0.001455163350328803,
      "learning_rate": 0.1519555860242495,
      "loss": 0.1908,
      "num_input_tokens_seen": 38184240,
      "step": 19835
    },
    {
      "epoch": 3.236560894037034,
      "grad_norm": 0.0022895997390151024,
      "learning_rate": 0.15189668601895218,
      "loss": 0.1962,
      "num_input_tokens_seen": 38194560,
      "step": 19840
    },
    {
      "epoch": 3.237376621257851,
      "grad_norm": 0.0012195646995678544,
      "learning_rate": 0.151837785721162,
      "loss": 0.2299,
      "num_input_tokens_seen": 38204752,
      "step": 19845
    },
    {
      "epoch": 3.2381923484786688,
      "grad_norm": 0.0014265088830143213,
      "learning_rate": 0.15177888513996218,
      "loss": 0.2541,
      "num_input_tokens_seen": 38214880,
      "step": 19850
    },
    {
      "epoch": 3.239008075699486,
      "grad_norm": 0.0009489099029451609,
      "learning_rate": 0.15171998428443592,
      "loss": 0.2145,
      "num_input_tokens_seen": 38224352,
      "step": 19855
    },
    {
      "epoch": 3.2398238029203035,
      "grad_norm": 0.0031456893775612116,
      "learning_rate": 0.1516610831636665,
      "loss": 0.2405,
      "num_input_tokens_seen": 38232208,
      "step": 19860
    },
    {
      "epoch": 3.2406395301411206,
      "grad_norm": 0.0033136457204818726,
      "learning_rate": 0.15160218178673715,
      "loss": 0.251,
      "num_input_tokens_seen": 38242160,
      "step": 19865
    },
    {
      "epoch": 3.241455257361938,
      "grad_norm": 0.0028898166492581367,
      "learning_rate": 0.15154328016273122,
      "loss": 0.1903,
      "num_input_tokens_seen": 38251952,
      "step": 19870
    },
    {
      "epoch": 3.2422709845827553,
      "grad_norm": 0.001919145230203867,
      "learning_rate": 0.1514843783007321,
      "loss": 0.2632,
      "num_input_tokens_seen": 38262272,
      "step": 19875
    },
    {
      "epoch": 3.243086711803573,
      "grad_norm": 0.0019197749206796288,
      "learning_rate": 0.15142547620982322,
      "loss": 0.256,
      "num_input_tokens_seen": 38271952,
      "step": 19880
    },
    {
      "epoch": 3.2439024390243905,
      "grad_norm": 0.004640647675842047,
      "learning_rate": 0.15136657389908797,
      "loss": 0.2442,
      "num_input_tokens_seen": 38282528,
      "step": 19885
    },
    {
      "epoch": 3.2447181662452076,
      "grad_norm": 0.0013262736611068249,
      "learning_rate": 0.15130767137760986,
      "loss": 0.2038,
      "num_input_tokens_seen": 38291952,
      "step": 19890
    },
    {
      "epoch": 3.245533893466025,
      "grad_norm": 0.0011430532904341817,
      "learning_rate": 0.15124876865447243,
      "loss": 0.2107,
      "num_input_tokens_seen": 38301920,
      "step": 19895
    },
    {
      "epoch": 3.2463496206868423,
      "grad_norm": 0.001186672947369516,
      "learning_rate": 0.15118986573875912,
      "loss": 0.1789,
      "num_input_tokens_seen": 38312128,
      "step": 19900
    },
    {
      "epoch": 3.24716534790766,
      "grad_norm": 0.0013389021623879671,
      "learning_rate": 0.15113096263955358,
      "loss": 0.1883,
      "num_input_tokens_seen": 38321024,
      "step": 19905
    },
    {
      "epoch": 3.247981075128477,
      "grad_norm": 0.0009083229233510792,
      "learning_rate": 0.1510720593659394,
      "loss": 0.2268,
      "num_input_tokens_seen": 38331520,
      "step": 19910
    },
    {
      "epoch": 3.2487968023492946,
      "grad_norm": 0.0014070391189306974,
      "learning_rate": 0.15101315592700015,
      "loss": 0.1849,
      "num_input_tokens_seen": 38341008,
      "step": 19915
    },
    {
      "epoch": 3.2496125295701117,
      "grad_norm": 0.0009928965009748936,
      "learning_rate": 0.15095425233181956,
      "loss": 0.1647,
      "num_input_tokens_seen": 38350880,
      "step": 19920
    },
    {
      "epoch": 3.2504282567909293,
      "grad_norm": 0.0024373086635023355,
      "learning_rate": 0.15089534858948128,
      "loss": 0.2217,
      "num_input_tokens_seen": 38359920,
      "step": 19925
    },
    {
      "epoch": 3.2512439840117464,
      "grad_norm": 0.0007840582984499633,
      "learning_rate": 0.15083644470906898,
      "loss": 0.21,
      "num_input_tokens_seen": 38369840,
      "step": 19930
    },
    {
      "epoch": 3.252059711232564,
      "grad_norm": 0.0011496063088998199,
      "learning_rate": 0.1507775406996664,
      "loss": 0.2086,
      "num_input_tokens_seen": 38380224,
      "step": 19935
    },
    {
      "epoch": 3.252875438453381,
      "grad_norm": 0.0019440617179498076,
      "learning_rate": 0.15071863657035725,
      "loss": 0.2639,
      "num_input_tokens_seen": 38389904,
      "step": 19940
    },
    {
      "epoch": 3.2536911656741987,
      "grad_norm": 0.0011665847850963473,
      "learning_rate": 0.15065973233022534,
      "loss": 0.2107,
      "num_input_tokens_seen": 38400032,
      "step": 19945
    },
    {
      "epoch": 3.254506892895016,
      "grad_norm": 0.0018217676552012563,
      "learning_rate": 0.15060082798835442,
      "loss": 0.2195,
      "num_input_tokens_seen": 38408544,
      "step": 19950
    },
    {
      "epoch": 3.2553226201158334,
      "grad_norm": 0.001477102399803698,
      "learning_rate": 0.15054192355382823,
      "loss": 0.2143,
      "num_input_tokens_seen": 38417312,
      "step": 19955
    },
    {
      "epoch": 3.2561383473366505,
      "grad_norm": 0.0012907409109175205,
      "learning_rate": 0.15048301903573066,
      "loss": 0.2095,
      "num_input_tokens_seen": 38426416,
      "step": 19960
    },
    {
      "epoch": 3.256954074557468,
      "grad_norm": 0.0011342071229591966,
      "learning_rate": 0.15042411444314546,
      "loss": 0.1876,
      "num_input_tokens_seen": 38437264,
      "step": 19965
    },
    {
      "epoch": 3.2577698017782852,
      "grad_norm": 0.0008117184042930603,
      "learning_rate": 0.1503652097851565,
      "loss": 0.2105,
      "num_input_tokens_seen": 38446576,
      "step": 19970
    },
    {
      "epoch": 3.258585528999103,
      "grad_norm": 0.0021116090938448906,
      "learning_rate": 0.15030630507084758,
      "loss": 0.1818,
      "num_input_tokens_seen": 38457056,
      "step": 19975
    },
    {
      "epoch": 3.25940125621992,
      "grad_norm": 0.0020407114643603563,
      "learning_rate": 0.1502474003093026,
      "loss": 0.1974,
      "num_input_tokens_seen": 38465808,
      "step": 19980
    },
    {
      "epoch": 3.2602169834407375,
      "grad_norm": 0.0011818078346550465,
      "learning_rate": 0.15018849550960536,
      "loss": 0.1961,
      "num_input_tokens_seen": 38474592,
      "step": 19985
    },
    {
      "epoch": 3.2610327106615546,
      "grad_norm": 0.0017825535032898188,
      "learning_rate": 0.15012959068083975,
      "loss": 0.2325,
      "num_input_tokens_seen": 38485120,
      "step": 19990
    },
    {
      "epoch": 3.261848437882372,
      "grad_norm": 0.0021377599332481623,
      "learning_rate": 0.1500706858320896,
      "loss": 0.1882,
      "num_input_tokens_seen": 38493408,
      "step": 19995
    },
    {
      "epoch": 3.2626641651031894,
      "grad_norm": 0.001869553467258811,
      "learning_rate": 0.15001178097243886,
      "loss": 0.1914,
      "num_input_tokens_seen": 38503392,
      "step": 20000
    },
    {
      "epoch": 3.2626641651031894,
      "eval_loss": 0.2145588994026184,
      "eval_runtime": 68.1471,
      "eval_samples_per_second": 39.987,
      "eval_steps_per_second": 20.001,
      "num_input_tokens_seen": 38503392,
      "step": 20000
    },
    {
      "epoch": 3.263479892324007,
      "grad_norm": 0.0009797109523788095,
      "learning_rate": 0.1499528761109713,
      "loss": 0.2379,
      "num_input_tokens_seen": 38512656,
      "step": 20005
    },
    {
      "epoch": 3.264295619544824,
      "grad_norm": 0.003197483019903302,
      "learning_rate": 0.14989397125677087,
      "loss": 0.2558,
      "num_input_tokens_seen": 38521952,
      "step": 20010
    },
    {
      "epoch": 3.2651113467656416,
      "grad_norm": 0.005157782230526209,
      "learning_rate": 0.14983506641892141,
      "loss": 0.2632,
      "num_input_tokens_seen": 38532384,
      "step": 20015
    },
    {
      "epoch": 3.2659270739864588,
      "grad_norm": 0.002413134789094329,
      "learning_rate": 0.14977616160650672,
      "loss": 0.2182,
      "num_input_tokens_seen": 38540800,
      "step": 20020
    },
    {
      "epoch": 3.2667428012072763,
      "grad_norm": 0.0015259676147252321,
      "learning_rate": 0.14971725682861076,
      "loss": 0.176,
      "num_input_tokens_seen": 38550448,
      "step": 20025
    },
    {
      "epoch": 3.2675585284280935,
      "grad_norm": 0.0034024110063910484,
      "learning_rate": 0.14965835209431738,
      "loss": 0.2369,
      "num_input_tokens_seen": 38560144,
      "step": 20030
    },
    {
      "epoch": 3.268374255648911,
      "grad_norm": 0.0015005870955064893,
      "learning_rate": 0.14959944741271036,
      "loss": 0.268,
      "num_input_tokens_seen": 38569728,
      "step": 20035
    },
    {
      "epoch": 3.269189982869728,
      "grad_norm": 0.0030342591926455498,
      "learning_rate": 0.14954054279287363,
      "loss": 0.2468,
      "num_input_tokens_seen": 38580544,
      "step": 20040
    },
    {
      "epoch": 3.2700057100905457,
      "grad_norm": 0.0027570463716983795,
      "learning_rate": 0.14948163824389094,
      "loss": 0.2173,
      "num_input_tokens_seen": 38590880,
      "step": 20045
    },
    {
      "epoch": 3.270821437311363,
      "grad_norm": 0.0019282234134152532,
      "learning_rate": 0.14942273377484613,
      "loss": 0.225,
      "num_input_tokens_seen": 38599328,
      "step": 20050
    },
    {
      "epoch": 3.2716371645321805,
      "grad_norm": 0.0011167324846610427,
      "learning_rate": 0.1493638293948231,
      "loss": 0.216,
      "num_input_tokens_seen": 38607440,
      "step": 20055
    },
    {
      "epoch": 3.272452891752998,
      "grad_norm": 0.0019578796345740557,
      "learning_rate": 0.14930492511290547,
      "loss": 0.2086,
      "num_input_tokens_seen": 38617024,
      "step": 20060
    },
    {
      "epoch": 3.273268618973815,
      "grad_norm": 0.0009435929823666811,
      "learning_rate": 0.14924602093817715,
      "loss": 0.1668,
      "num_input_tokens_seen": 38626576,
      "step": 20065
    },
    {
      "epoch": 3.2740843461946323,
      "grad_norm": 0.0013185183051973581,
      "learning_rate": 0.14918711687972194,
      "loss": 0.1915,
      "num_input_tokens_seen": 38635760,
      "step": 20070
    },
    {
      "epoch": 3.27490007341545,
      "grad_norm": 0.001264616148546338,
      "learning_rate": 0.14912821294662346,
      "loss": 0.2226,
      "num_input_tokens_seen": 38643952,
      "step": 20075
    },
    {
      "epoch": 3.2757158006362674,
      "grad_norm": 0.0014034671476110816,
      "learning_rate": 0.14906930914796554,
      "loss": 0.2423,
      "num_input_tokens_seen": 38654272,
      "step": 20080
    },
    {
      "epoch": 3.2765315278570846,
      "grad_norm": 0.0011086933081969619,
      "learning_rate": 0.14901040549283182,
      "loss": 0.2436,
      "num_input_tokens_seen": 38664336,
      "step": 20085
    },
    {
      "epoch": 3.2773472550779017,
      "grad_norm": 0.0015303074615076184,
      "learning_rate": 0.148951501990306,
      "loss": 0.2021,
      "num_input_tokens_seen": 38674128,
      "step": 20090
    },
    {
      "epoch": 3.2781629822987193,
      "grad_norm": 0.0017426467966288328,
      "learning_rate": 0.14889259864947177,
      "loss": 0.1998,
      "num_input_tokens_seen": 38684768,
      "step": 20095
    },
    {
      "epoch": 3.278978709519537,
      "grad_norm": 0.001469665439799428,
      "learning_rate": 0.14883369547941272,
      "loss": 0.1938,
      "num_input_tokens_seen": 38693904,
      "step": 20100
    },
    {
      "epoch": 3.279794436740354,
      "grad_norm": 0.0029561244882643223,
      "learning_rate": 0.14877479248921247,
      "loss": 0.2457,
      "num_input_tokens_seen": 38704752,
      "step": 20105
    },
    {
      "epoch": 3.2806101639611716,
      "grad_norm": 0.002162884222343564,
      "learning_rate": 0.14871588968795468,
      "loss": 0.2428,
      "num_input_tokens_seen": 38714064,
      "step": 20110
    },
    {
      "epoch": 3.2814258911819887,
      "grad_norm": 0.00273935217410326,
      "learning_rate": 0.1486569870847228,
      "loss": 0.2051,
      "num_input_tokens_seen": 38724768,
      "step": 20115
    },
    {
      "epoch": 3.2822416184028063,
      "grad_norm": 0.0009281979291699827,
      "learning_rate": 0.1485980846886004,
      "loss": 0.1872,
      "num_input_tokens_seen": 38733680,
      "step": 20120
    },
    {
      "epoch": 3.2830573456236234,
      "grad_norm": 0.0010530782165005803,
      "learning_rate": 0.14853918250867096,
      "loss": 0.2139,
      "num_input_tokens_seen": 38743424,
      "step": 20125
    },
    {
      "epoch": 3.283873072844441,
      "grad_norm": 0.0013675456866621971,
      "learning_rate": 0.1484802805540179,
      "loss": 0.1735,
      "num_input_tokens_seen": 38753760,
      "step": 20130
    },
    {
      "epoch": 3.284688800065258,
      "grad_norm": 0.0015133850974962115,
      "learning_rate": 0.14842137883372472,
      "loss": 0.258,
      "num_input_tokens_seen": 38762528,
      "step": 20135
    },
    {
      "epoch": 3.2855045272860757,
      "grad_norm": 0.0016412172699347138,
      "learning_rate": 0.14836247735687474,
      "loss": 0.1442,
      "num_input_tokens_seen": 38771472,
      "step": 20140
    },
    {
      "epoch": 3.286320254506893,
      "grad_norm": 0.0030982375610619783,
      "learning_rate": 0.14830357613255132,
      "loss": 0.1918,
      "num_input_tokens_seen": 38781104,
      "step": 20145
    },
    {
      "epoch": 3.2871359817277104,
      "grad_norm": 0.002007411327213049,
      "learning_rate": 0.1482446751698378,
      "loss": 0.1843,
      "num_input_tokens_seen": 38790208,
      "step": 20150
    },
    {
      "epoch": 3.2879517089485275,
      "grad_norm": 0.0021287978161126375,
      "learning_rate": 0.14818577447781744,
      "loss": 0.2327,
      "num_input_tokens_seen": 38799312,
      "step": 20155
    },
    {
      "epoch": 3.288767436169345,
      "grad_norm": 0.0025673636700958014,
      "learning_rate": 0.14812687406557346,
      "loss": 0.2708,
      "num_input_tokens_seen": 38808656,
      "step": 20160
    },
    {
      "epoch": 3.289583163390162,
      "grad_norm": 0.003854782786220312,
      "learning_rate": 0.14806797394218899,
      "loss": 0.2812,
      "num_input_tokens_seen": 38817584,
      "step": 20165
    },
    {
      "epoch": 3.29039889061098,
      "grad_norm": 0.0015170053811743855,
      "learning_rate": 0.1480090741167472,
      "loss": 0.2358,
      "num_input_tokens_seen": 38827248,
      "step": 20170
    },
    {
      "epoch": 3.291214617831797,
      "grad_norm": 0.001664910465478897,
      "learning_rate": 0.1479501745983313,
      "loss": 0.1991,
      "num_input_tokens_seen": 38836720,
      "step": 20175
    },
    {
      "epoch": 3.2920303450526145,
      "grad_norm": 0.0018999830354005098,
      "learning_rate": 0.14789127539602415,
      "loss": 0.2238,
      "num_input_tokens_seen": 38845888,
      "step": 20180
    },
    {
      "epoch": 3.2928460722734316,
      "grad_norm": 0.000849398726131767,
      "learning_rate": 0.14783237651890885,
      "loss": 0.2322,
      "num_input_tokens_seen": 38855408,
      "step": 20185
    },
    {
      "epoch": 3.293661799494249,
      "grad_norm": 0.0013343425234779716,
      "learning_rate": 0.14777347797606838,
      "loss": 0.2236,
      "num_input_tokens_seen": 38865744,
      "step": 20190
    },
    {
      "epoch": 3.2944775267150663,
      "grad_norm": 0.001810527639463544,
      "learning_rate": 0.14771457977658553,
      "loss": 0.222,
      "num_input_tokens_seen": 38876688,
      "step": 20195
    },
    {
      "epoch": 3.295293253935884,
      "grad_norm": 0.0017383379163220525,
      "learning_rate": 0.14765568192954326,
      "loss": 0.2317,
      "num_input_tokens_seen": 38885696,
      "step": 20200
    },
    {
      "epoch": 3.295293253935884,
      "eval_loss": 0.21616487205028534,
      "eval_runtime": 67.9556,
      "eval_samples_per_second": 40.1,
      "eval_steps_per_second": 20.057,
      "num_input_tokens_seen": 38885696,
      "step": 20200
    },
    {
      "epoch": 3.296108981156701,
      "grad_norm": 0.0018505596090108156,
      "learning_rate": 0.14759678444402421,
      "loss": 0.1925,
      "num_input_tokens_seen": 38895952,
      "step": 20205
    },
    {
      "epoch": 3.2969247083775186,
      "grad_norm": 0.0012932870304211974,
      "learning_rate": 0.14753788732911122,
      "loss": 0.2046,
      "num_input_tokens_seen": 38905152,
      "step": 20210
    },
    {
      "epoch": 3.2977404355983357,
      "grad_norm": 0.0009280751110054553,
      "learning_rate": 0.147478990593887,
      "loss": 0.1913,
      "num_input_tokens_seen": 38915424,
      "step": 20215
    },
    {
      "epoch": 3.2985561628191533,
      "grad_norm": 0.0009553747368045151,
      "learning_rate": 0.14742009424743405,
      "loss": 0.1687,
      "num_input_tokens_seen": 38923648,
      "step": 20220
    },
    {
      "epoch": 3.2993718900399704,
      "grad_norm": 0.0027516901027411222,
      "learning_rate": 0.14736119829883504,
      "loss": 0.2259,
      "num_input_tokens_seen": 38933072,
      "step": 20225
    },
    {
      "epoch": 3.300187617260788,
      "grad_norm": 0.001784766442142427,
      "learning_rate": 0.14730230275717243,
      "loss": 0.1683,
      "num_input_tokens_seen": 38943952,
      "step": 20230
    },
    {
      "epoch": 3.3010033444816056,
      "grad_norm": 0.0030431626364588737,
      "learning_rate": 0.14724340763152854,
      "loss": 0.256,
      "num_input_tokens_seen": 38954800,
      "step": 20235
    },
    {
      "epoch": 3.3018190717024227,
      "grad_norm": 0.0016664270078763366,
      "learning_rate": 0.14718451293098594,
      "loss": 0.2181,
      "num_input_tokens_seen": 38964880,
      "step": 20240
    },
    {
      "epoch": 3.30263479892324,
      "grad_norm": 0.0025675352662801743,
      "learning_rate": 0.14712561866462676,
      "loss": 0.2126,
      "num_input_tokens_seen": 38974064,
      "step": 20245
    },
    {
      "epoch": 3.3034505261440574,
      "grad_norm": 0.0011110317427664995,
      "learning_rate": 0.1470667248415333,
      "loss": 0.2109,
      "num_input_tokens_seen": 38984752,
      "step": 20250
    },
    {
      "epoch": 3.304266253364875,
      "grad_norm": 0.0016159146325662732,
      "learning_rate": 0.1470078314707878,
      "loss": 0.1877,
      "num_input_tokens_seen": 38995456,
      "step": 20255
    },
    {
      "epoch": 3.305081980585692,
      "grad_norm": 0.0011680523166432977,
      "learning_rate": 0.14694893856147223,
      "loss": 0.1931,
      "num_input_tokens_seen": 39003808,
      "step": 20260
    },
    {
      "epoch": 3.3058977078065093,
      "grad_norm": 0.003060820046812296,
      "learning_rate": 0.14689004612266868,
      "loss": 0.2369,
      "num_input_tokens_seen": 39012992,
      "step": 20265
    },
    {
      "epoch": 3.306713435027327,
      "grad_norm": 0.005580793134868145,
      "learning_rate": 0.14683115416345913,
      "loss": 0.2593,
      "num_input_tokens_seen": 39022240,
      "step": 20270
    },
    {
      "epoch": 3.3075291622481444,
      "grad_norm": 0.0020790689159184694,
      "learning_rate": 0.1467722626929254,
      "loss": 0.2685,
      "num_input_tokens_seen": 39031920,
      "step": 20275
    },
    {
      "epoch": 3.3083448894689615,
      "grad_norm": 0.0023597991093993187,
      "learning_rate": 0.14671337172014937,
      "loss": 0.2161,
      "num_input_tokens_seen": 39041632,
      "step": 20280
    },
    {
      "epoch": 3.309160616689779,
      "grad_norm": 0.0019502969225868583,
      "learning_rate": 0.14665448125421265,
      "loss": 0.1973,
      "num_input_tokens_seen": 39050848,
      "step": 20285
    },
    {
      "epoch": 3.3099763439105963,
      "grad_norm": 0.002185826189815998,
      "learning_rate": 0.146595591304197,
      "loss": 0.247,
      "num_input_tokens_seen": 39061184,
      "step": 20290
    },
    {
      "epoch": 3.310792071131414,
      "grad_norm": 0.0010557285277172923,
      "learning_rate": 0.14653670187918397,
      "loss": 0.2067,
      "num_input_tokens_seen": 39070816,
      "step": 20295
    },
    {
      "epoch": 3.311607798352231,
      "grad_norm": 0.0023275811690837145,
      "learning_rate": 0.14647781298825502,
      "loss": 0.1708,
      "num_input_tokens_seen": 39081312,
      "step": 20300
    },
    {
      "epoch": 3.3124235255730485,
      "grad_norm": 0.001955032115802169,
      "learning_rate": 0.14641892464049153,
      "loss": 0.167,
      "num_input_tokens_seen": 39090192,
      "step": 20305
    },
    {
      "epoch": 3.3132392527938657,
      "grad_norm": 0.0014442156534641981,
      "learning_rate": 0.14636003684497495,
      "loss": 0.2044,
      "num_input_tokens_seen": 39099488,
      "step": 20310
    },
    {
      "epoch": 3.3140549800146832,
      "grad_norm": 0.002397510688751936,
      "learning_rate": 0.14630114961078636,
      "loss": 0.2112,
      "num_input_tokens_seen": 39109104,
      "step": 20315
    },
    {
      "epoch": 3.3148707072355004,
      "grad_norm": 0.0018667291151359677,
      "learning_rate": 0.14624226294700704,
      "loss": 0.2191,
      "num_input_tokens_seen": 39119088,
      "step": 20320
    },
    {
      "epoch": 3.315686434456318,
      "grad_norm": 0.0012344606220722198,
      "learning_rate": 0.14618337686271793,
      "loss": 0.2022,
      "num_input_tokens_seen": 39129040,
      "step": 20325
    },
    {
      "epoch": 3.316502161677135,
      "grad_norm": 0.0016047105891630054,
      "learning_rate": 0.1461244913670001,
      "loss": 0.2536,
      "num_input_tokens_seen": 39138416,
      "step": 20330
    },
    {
      "epoch": 3.3173178888979526,
      "grad_norm": 0.001388898235745728,
      "learning_rate": 0.1460656064689344,
      "loss": 0.2158,
      "num_input_tokens_seen": 39148384,
      "step": 20335
    },
    {
      "epoch": 3.31813361611877,
      "grad_norm": 0.0016030961414799094,
      "learning_rate": 0.14600672217760163,
      "loss": 0.2154,
      "num_input_tokens_seen": 39157840,
      "step": 20340
    },
    {
      "epoch": 3.3189493433395874,
      "grad_norm": 0.0023413654416799545,
      "learning_rate": 0.14594783850208248,
      "loss": 0.2001,
      "num_input_tokens_seen": 39167728,
      "step": 20345
    },
    {
      "epoch": 3.3197650705604045,
      "grad_norm": 0.002159525640308857,
      "learning_rate": 0.14588895545145758,
      "loss": 0.2184,
      "num_input_tokens_seen": 39176752,
      "step": 20350
    },
    {
      "epoch": 3.320580797781222,
      "grad_norm": 0.0024054618552327156,
      "learning_rate": 0.14583007303480738,
      "loss": 0.2426,
      "num_input_tokens_seen": 39186480,
      "step": 20355
    },
    {
      "epoch": 3.321396525002039,
      "grad_norm": 0.0014907200820744038,
      "learning_rate": 0.14577119126121235,
      "loss": 0.2338,
      "num_input_tokens_seen": 39196512,
      "step": 20360
    },
    {
      "epoch": 3.3222122522228568,
      "grad_norm": 0.002923280466347933,
      "learning_rate": 0.14571231013975272,
      "loss": 0.1833,
      "num_input_tokens_seen": 39205312,
      "step": 20365
    },
    {
      "epoch": 3.323027979443674,
      "grad_norm": 0.001633778098039329,
      "learning_rate": 0.1456534296795088,
      "loss": 0.1844,
      "num_input_tokens_seen": 39214880,
      "step": 20370
    },
    {
      "epoch": 3.3238437066644915,
      "grad_norm": 0.002465428551658988,
      "learning_rate": 0.14559454988956066,
      "loss": 0.2278,
      "num_input_tokens_seen": 39225520,
      "step": 20375
    },
    {
      "epoch": 3.3246594338853086,
      "grad_norm": 0.0015268062707036734,
      "learning_rate": 0.1455356707789882,
      "loss": 0.1546,
      "num_input_tokens_seen": 39235072,
      "step": 20380
    },
    {
      "epoch": 3.325475161106126,
      "grad_norm": 0.0028700430411845446,
      "learning_rate": 0.14547679235687147,
      "loss": 0.2566,
      "num_input_tokens_seen": 39244592,
      "step": 20385
    },
    {
      "epoch": 3.3262908883269433,
      "grad_norm": 0.0016832125838845968,
      "learning_rate": 0.14541791463229023,
      "loss": 0.215,
      "num_input_tokens_seen": 39252432,
      "step": 20390
    },
    {
      "epoch": 3.327106615547761,
      "grad_norm": 0.003026845632120967,
      "learning_rate": 0.14535903761432406,
      "loss": 0.2173,
      "num_input_tokens_seen": 39261520,
      "step": 20395
    },
    {
      "epoch": 3.327922342768578,
      "grad_norm": 0.002173985354602337,
      "learning_rate": 0.1453001613120527,
      "loss": 0.212,
      "num_input_tokens_seen": 39270320,
      "step": 20400
    },
    {
      "epoch": 3.327922342768578,
      "eval_loss": 0.21317175030708313,
      "eval_runtime": 67.992,
      "eval_samples_per_second": 40.078,
      "eval_steps_per_second": 20.046,
      "num_input_tokens_seen": 39270320,
      "step": 20400
    },
    {
      "epoch": 3.3287380699893956,
      "grad_norm": 0.0014866990968585014,
      "learning_rate": 0.14524128573455547,
      "loss": 0.2353,
      "num_input_tokens_seen": 39280288,
      "step": 20405
    },
    {
      "epoch": 3.3295537972102127,
      "grad_norm": 0.0021927906200289726,
      "learning_rate": 0.14518241089091177,
      "loss": 0.2129,
      "num_input_tokens_seen": 39290240,
      "step": 20410
    },
    {
      "epoch": 3.3303695244310303,
      "grad_norm": 0.003966664895415306,
      "learning_rate": 0.1451235367902009,
      "loss": 0.218,
      "num_input_tokens_seen": 39299616,
      "step": 20415
    },
    {
      "epoch": 3.3311852516518474,
      "grad_norm": 0.0019204745767638087,
      "learning_rate": 0.1450646634415019,
      "loss": 0.2511,
      "num_input_tokens_seen": 39309920,
      "step": 20420
    },
    {
      "epoch": 3.332000978872665,
      "grad_norm": 0.002019170206040144,
      "learning_rate": 0.1450057908538938,
      "loss": 0.1632,
      "num_input_tokens_seen": 39319824,
      "step": 20425
    },
    {
      "epoch": 3.3328167060934826,
      "grad_norm": 0.006474389228969812,
      "learning_rate": 0.14494691903645557,
      "loss": 0.2036,
      "num_input_tokens_seen": 39330032,
      "step": 20430
    },
    {
      "epoch": 3.3336324333142997,
      "grad_norm": 0.001106209121644497,
      "learning_rate": 0.14488804799826588,
      "loss": 0.2028,
      "num_input_tokens_seen": 39339408,
      "step": 20435
    },
    {
      "epoch": 3.334448160535117,
      "grad_norm": 0.0009817141108214855,
      "learning_rate": 0.14482917774840348,
      "loss": 0.2528,
      "num_input_tokens_seen": 39349616,
      "step": 20440
    },
    {
      "epoch": 3.3352638877559344,
      "grad_norm": 0.0020636653061956167,
      "learning_rate": 0.14477030829594684,
      "loss": 0.1857,
      "num_input_tokens_seen": 39359312,
      "step": 20445
    },
    {
      "epoch": 3.336079614976752,
      "grad_norm": 0.0009358407114632428,
      "learning_rate": 0.14471143964997432,
      "loss": 0.2205,
      "num_input_tokens_seen": 39368080,
      "step": 20450
    },
    {
      "epoch": 3.336895342197569,
      "grad_norm": 0.0015105264028534293,
      "learning_rate": 0.14465257181956434,
      "loss": 0.1946,
      "num_input_tokens_seen": 39377952,
      "step": 20455
    },
    {
      "epoch": 3.3377110694183862,
      "grad_norm": 0.0017298227176070213,
      "learning_rate": 0.1445937048137949,
      "loss": 0.2055,
      "num_input_tokens_seen": 39388064,
      "step": 20460
    },
    {
      "epoch": 3.338526796639204,
      "grad_norm": 0.002533981343731284,
      "learning_rate": 0.14453483864174416,
      "loss": 0.2543,
      "num_input_tokens_seen": 39398704,
      "step": 20465
    },
    {
      "epoch": 3.3393425238600214,
      "grad_norm": 0.0019737055990844965,
      "learning_rate": 0.14447597331249,
      "loss": 0.2255,
      "num_input_tokens_seen": 39407776,
      "step": 20470
    },
    {
      "epoch": 3.3401582510808385,
      "grad_norm": 0.0017014432232826948,
      "learning_rate": 0.1444171088351102,
      "loss": 0.2103,
      "num_input_tokens_seen": 39417776,
      "step": 20475
    },
    {
      "epoch": 3.340973978301656,
      "grad_norm": 0.0012758182128891349,
      "learning_rate": 0.14435824521868235,
      "loss": 0.2119,
      "num_input_tokens_seen": 39427776,
      "step": 20480
    },
    {
      "epoch": 3.3417897055224732,
      "grad_norm": 0.0012250237632542849,
      "learning_rate": 0.14429938247228397,
      "loss": 0.2028,
      "num_input_tokens_seen": 39437584,
      "step": 20485
    },
    {
      "epoch": 3.342605432743291,
      "grad_norm": 0.0025720687117427588,
      "learning_rate": 0.14424052060499243,
      "loss": 0.2255,
      "num_input_tokens_seen": 39448064,
      "step": 20490
    },
    {
      "epoch": 3.343421159964108,
      "grad_norm": 0.0013082048390060663,
      "learning_rate": 0.14418165962588506,
      "loss": 0.234,
      "num_input_tokens_seen": 39457712,
      "step": 20495
    },
    {
      "epoch": 3.3442368871849255,
      "grad_norm": 0.0013657130766659975,
      "learning_rate": 0.1441227995440388,
      "loss": 0.2634,
      "num_input_tokens_seen": 39467744,
      "step": 20500
    },
    {
      "epoch": 3.3450526144057426,
      "grad_norm": 0.002204444957897067,
      "learning_rate": 0.14406394036853082,
      "loss": 0.2301,
      "num_input_tokens_seen": 39477152,
      "step": 20505
    },
    {
      "epoch": 3.34586834162656,
      "grad_norm": 0.0017370838904753327,
      "learning_rate": 0.14400508210843774,
      "loss": 0.1881,
      "num_input_tokens_seen": 39488576,
      "step": 20510
    },
    {
      "epoch": 3.3466840688473773,
      "grad_norm": 0.001749026938341558,
      "learning_rate": 0.1439462247728364,
      "loss": 0.208,
      "num_input_tokens_seen": 39498416,
      "step": 20515
    },
    {
      "epoch": 3.347499796068195,
      "grad_norm": 0.0012960792519152164,
      "learning_rate": 0.14388736837080326,
      "loss": 0.2275,
      "num_input_tokens_seen": 39507712,
      "step": 20520
    },
    {
      "epoch": 3.348315523289012,
      "grad_norm": 0.001068644574843347,
      "learning_rate": 0.14382851291141469,
      "loss": 0.1826,
      "num_input_tokens_seen": 39517184,
      "step": 20525
    },
    {
      "epoch": 3.3491312505098296,
      "grad_norm": 0.001451171818189323,
      "learning_rate": 0.14376965840374697,
      "loss": 0.2161,
      "num_input_tokens_seen": 39526576,
      "step": 20530
    },
    {
      "epoch": 3.3499469777306468,
      "grad_norm": 0.0010216811206191778,
      "learning_rate": 0.14371080485687632,
      "loss": 0.223,
      "num_input_tokens_seen": 39536288,
      "step": 20535
    },
    {
      "epoch": 3.3507627049514643,
      "grad_norm": 0.0018927832134068012,
      "learning_rate": 0.1436519522798785,
      "loss": 0.2226,
      "num_input_tokens_seen": 39546880,
      "step": 20540
    },
    {
      "epoch": 3.3515784321722815,
      "grad_norm": 0.002653246745467186,
      "learning_rate": 0.14359310068182948,
      "loss": 0.2294,
      "num_input_tokens_seen": 39556448,
      "step": 20545
    },
    {
      "epoch": 3.352394159393099,
      "grad_norm": 0.001515139127150178,
      "learning_rate": 0.14353425007180484,
      "loss": 0.2018,
      "num_input_tokens_seen": 39566736,
      "step": 20550
    },
    {
      "epoch": 3.353209886613916,
      "grad_norm": 0.0019062028732150793,
      "learning_rate": 0.14347540045888005,
      "loss": 0.2107,
      "num_input_tokens_seen": 39577472,
      "step": 20555
    },
    {
      "epoch": 3.3540256138347337,
      "grad_norm": 0.0008905883296392858,
      "learning_rate": 0.14341655185213056,
      "loss": 0.2117,
      "num_input_tokens_seen": 39586736,
      "step": 20560
    },
    {
      "epoch": 3.354841341055551,
      "grad_norm": 0.0012107411166653037,
      "learning_rate": 0.14335770426063144,
      "loss": 0.2476,
      "num_input_tokens_seen": 39596640,
      "step": 20565
    },
    {
      "epoch": 3.3556570682763684,
      "grad_norm": 0.0012510470114648342,
      "learning_rate": 0.1432988576934578,
      "loss": 0.181,
      "num_input_tokens_seen": 39606800,
      "step": 20570
    },
    {
      "epoch": 3.3564727954971856,
      "grad_norm": 0.0016612590989097953,
      "learning_rate": 0.14324001215968457,
      "loss": 0.1969,
      "num_input_tokens_seen": 39615808,
      "step": 20575
    },
    {
      "epoch": 3.357288522718003,
      "grad_norm": 0.0015927022323012352,
      "learning_rate": 0.14318116766838637,
      "loss": 0.2033,
      "num_input_tokens_seen": 39625984,
      "step": 20580
    },
    {
      "epoch": 3.3581042499388203,
      "grad_norm": 0.0011147981276735663,
      "learning_rate": 0.14312232422863788,
      "loss": 0.2147,
      "num_input_tokens_seen": 39636832,
      "step": 20585
    },
    {
      "epoch": 3.358919977159638,
      "grad_norm": 0.002596965990960598,
      "learning_rate": 0.14306348184951334,
      "loss": 0.1949,
      "num_input_tokens_seen": 39646416,
      "step": 20590
    },
    {
      "epoch": 3.359735704380455,
      "grad_norm": 0.0016109931748360395,
      "learning_rate": 0.1430046405400871,
      "loss": 0.22,
      "num_input_tokens_seen": 39656128,
      "step": 20595
    },
    {
      "epoch": 3.3605514316012726,
      "grad_norm": 0.0013272883370518684,
      "learning_rate": 0.14294580030943324,
      "loss": 0.2109,
      "num_input_tokens_seen": 39665472,
      "step": 20600
    },
    {
      "epoch": 3.3605514316012726,
      "eval_loss": 0.20963165163993835,
      "eval_runtime": 67.943,
      "eval_samples_per_second": 40.107,
      "eval_steps_per_second": 20.061,
      "num_input_tokens_seen": 39665472,
      "step": 20600
    },
    {
      "epoch": 3.36136715882209,
      "grad_norm": 0.0015523872571066022,
      "learning_rate": 0.14288696116662553,
      "loss": 0.2784,
      "num_input_tokens_seen": 39674736,
      "step": 20605
    },
    {
      "epoch": 3.3621828860429073,
      "grad_norm": 0.0018290667794644833,
      "learning_rate": 0.1428281231207378,
      "loss": 0.2156,
      "num_input_tokens_seen": 39684928,
      "step": 20610
    },
    {
      "epoch": 3.3629986132637244,
      "grad_norm": 0.0022728254552930593,
      "learning_rate": 0.1427692861808437,
      "loss": 0.1588,
      "num_input_tokens_seen": 39693504,
      "step": 20615
    },
    {
      "epoch": 3.363814340484542,
      "grad_norm": 0.001490477123297751,
      "learning_rate": 0.1427104503560165,
      "loss": 0.2043,
      "num_input_tokens_seen": 39703632,
      "step": 20620
    },
    {
      "epoch": 3.3646300677053596,
      "grad_norm": 0.0012730370508506894,
      "learning_rate": 0.14265161565532947,
      "loss": 0.1848,
      "num_input_tokens_seen": 39714736,
      "step": 20625
    },
    {
      "epoch": 3.3654457949261767,
      "grad_norm": 0.0016741007566452026,
      "learning_rate": 0.14259278208785564,
      "loss": 0.2336,
      "num_input_tokens_seen": 39723840,
      "step": 20630
    },
    {
      "epoch": 3.366261522146994,
      "grad_norm": 0.0017121577402576804,
      "learning_rate": 0.14253394966266789,
      "loss": 0.1966,
      "num_input_tokens_seen": 39733888,
      "step": 20635
    },
    {
      "epoch": 3.3670772493678114,
      "grad_norm": 0.0014556347159668803,
      "learning_rate": 0.14247511838883894,
      "loss": 0.1823,
      "num_input_tokens_seen": 39744208,
      "step": 20640
    },
    {
      "epoch": 3.367892976588629,
      "grad_norm": 0.001198640326038003,
      "learning_rate": 0.14241628827544126,
      "loss": 0.1757,
      "num_input_tokens_seen": 39752192,
      "step": 20645
    },
    {
      "epoch": 3.368708703809446,
      "grad_norm": 0.0008618306019343436,
      "learning_rate": 0.14235745933154723,
      "loss": 0.1963,
      "num_input_tokens_seen": 39762288,
      "step": 20650
    },
    {
      "epoch": 3.3695244310302637,
      "grad_norm": 0.002663149032741785,
      "learning_rate": 0.14229863156622907,
      "loss": 0.2058,
      "num_input_tokens_seen": 39772256,
      "step": 20655
    },
    {
      "epoch": 3.370340158251081,
      "grad_norm": 0.002474699867889285,
      "learning_rate": 0.14223980498855868,
      "loss": 0.2036,
      "num_input_tokens_seen": 39782128,
      "step": 20660
    },
    {
      "epoch": 3.3711558854718984,
      "grad_norm": 0.0016077117761597037,
      "learning_rate": 0.14218097960760792,
      "loss": 0.198,
      "num_input_tokens_seen": 39791696,
      "step": 20665
    },
    {
      "epoch": 3.3719716126927155,
      "grad_norm": 0.0023816905450075865,
      "learning_rate": 0.1421221554324483,
      "loss": 0.1609,
      "num_input_tokens_seen": 39801280,
      "step": 20670
    },
    {
      "epoch": 3.372787339913533,
      "grad_norm": 0.005034681875258684,
      "learning_rate": 0.1420633324721513,
      "loss": 0.2425,
      "num_input_tokens_seen": 39811456,
      "step": 20675
    },
    {
      "epoch": 3.37360306713435,
      "grad_norm": 0.0016021475894376636,
      "learning_rate": 0.14200451073578824,
      "loss": 0.2127,
      "num_input_tokens_seen": 39820880,
      "step": 20680
    },
    {
      "epoch": 3.374418794355168,
      "grad_norm": 0.0014216398121789098,
      "learning_rate": 0.14194569023243003,
      "loss": 0.1581,
      "num_input_tokens_seen": 39831504,
      "step": 20685
    },
    {
      "epoch": 3.375234521575985,
      "grad_norm": 0.0015526637434959412,
      "learning_rate": 0.14188687097114766,
      "loss": 0.1813,
      "num_input_tokens_seen": 39842032,
      "step": 20690
    },
    {
      "epoch": 3.3760502487968025,
      "grad_norm": 0.001010137959383428,
      "learning_rate": 0.14182805296101172,
      "loss": 0.1903,
      "num_input_tokens_seen": 39852848,
      "step": 20695
    },
    {
      "epoch": 3.3768659760176196,
      "grad_norm": 0.00223541259765625,
      "learning_rate": 0.14176923621109272,
      "loss": 0.2086,
      "num_input_tokens_seen": 39862288,
      "step": 20700
    },
    {
      "epoch": 3.377681703238437,
      "grad_norm": 0.002321352483704686,
      "learning_rate": 0.14171042073046097,
      "loss": 0.2404,
      "num_input_tokens_seen": 39872208,
      "step": 20705
    },
    {
      "epoch": 3.3784974304592543,
      "grad_norm": 0.0021686828695237637,
      "learning_rate": 0.14165160652818642,
      "loss": 0.264,
      "num_input_tokens_seen": 39881936,
      "step": 20710
    },
    {
      "epoch": 3.379313157680072,
      "grad_norm": 0.0013377639697864652,
      "learning_rate": 0.14159279361333907,
      "loss": 0.1886,
      "num_input_tokens_seen": 39891376,
      "step": 20715
    },
    {
      "epoch": 3.380128884900889,
      "grad_norm": 0.002349739894270897,
      "learning_rate": 0.14153398199498868,
      "loss": 0.2332,
      "num_input_tokens_seen": 39899920,
      "step": 20720
    },
    {
      "epoch": 3.3809446121217066,
      "grad_norm": 0.002581463661044836,
      "learning_rate": 0.14147517168220458,
      "loss": 0.2166,
      "num_input_tokens_seen": 39909840,
      "step": 20725
    },
    {
      "epoch": 3.3817603393425237,
      "grad_norm": 0.001727169961668551,
      "learning_rate": 0.14141636268405616,
      "loss": 0.2173,
      "num_input_tokens_seen": 39920352,
      "step": 20730
    },
    {
      "epoch": 3.3825760665633413,
      "grad_norm": 0.0018946370109915733,
      "learning_rate": 0.14135755500961253,
      "loss": 0.1847,
      "num_input_tokens_seen": 39930032,
      "step": 20735
    },
    {
      "epoch": 3.3833917937841584,
      "grad_norm": 0.0033697548788040876,
      "learning_rate": 0.14129874866794245,
      "loss": 0.2762,
      "num_input_tokens_seen": 39939104,
      "step": 20740
    },
    {
      "epoch": 3.384207521004976,
      "grad_norm": 0.0017499455716460943,
      "learning_rate": 0.14123994366811476,
      "loss": 0.1644,
      "num_input_tokens_seen": 39947920,
      "step": 20745
    },
    {
      "epoch": 3.385023248225793,
      "grad_norm": 0.001179737620986998,
      "learning_rate": 0.14118114001919774,
      "loss": 0.1888,
      "num_input_tokens_seen": 39956640,
      "step": 20750
    },
    {
      "epoch": 3.3858389754466107,
      "grad_norm": 0.0015946178464218974,
      "learning_rate": 0.14112233773025978,
      "loss": 0.156,
      "num_input_tokens_seen": 39967072,
      "step": 20755
    },
    {
      "epoch": 3.386654702667428,
      "grad_norm": 0.001370242447592318,
      "learning_rate": 0.14106353681036896,
      "loss": 0.2164,
      "num_input_tokens_seen": 39976832,
      "step": 20760
    },
    {
      "epoch": 3.3874704298882454,
      "grad_norm": 0.0015148210804909468,
      "learning_rate": 0.14100473726859303,
      "loss": 0.2107,
      "num_input_tokens_seen": 39986816,
      "step": 20765
    },
    {
      "epoch": 3.3882861571090626,
      "grad_norm": 0.001113488688133657,
      "learning_rate": 0.14094593911399964,
      "loss": 0.1847,
      "num_input_tokens_seen": 39994800,
      "step": 20770
    },
    {
      "epoch": 3.38910188432988,
      "grad_norm": 0.0022006663493812084,
      "learning_rate": 0.14088714235565625,
      "loss": 0.2298,
      "num_input_tokens_seen": 40003968,
      "step": 20775
    },
    {
      "epoch": 3.3899176115506973,
      "grad_norm": 0.0012916033156216145,
      "learning_rate": 0.14082834700263,
      "loss": 0.1601,
      "num_input_tokens_seen": 40012352,
      "step": 20780
    },
    {
      "epoch": 3.390733338771515,
      "grad_norm": 0.001466162852011621,
      "learning_rate": 0.14076955306398795,
      "loss": 0.2584,
      "num_input_tokens_seen": 40020928,
      "step": 20785
    },
    {
      "epoch": 3.391549065992332,
      "grad_norm": 0.0019833578262478113,
      "learning_rate": 0.14071076054879675,
      "loss": 0.2583,
      "num_input_tokens_seen": 40031184,
      "step": 20790
    },
    {
      "epoch": 3.3923647932131495,
      "grad_norm": 0.0023399842903017998,
      "learning_rate": 0.14065196946612302,
      "loss": 0.2392,
      "num_input_tokens_seen": 40040784,
      "step": 20795
    },
    {
      "epoch": 3.393180520433967,
      "grad_norm": 0.0014113421784713864,
      "learning_rate": 0.1405931798250331,
      "loss": 0.2342,
      "num_input_tokens_seen": 40049680,
      "step": 20800
    },
    {
      "epoch": 3.393180520433967,
      "eval_loss": 0.21923308074474335,
      "eval_runtime": 68.1049,
      "eval_samples_per_second": 40.012,
      "eval_steps_per_second": 20.013,
      "num_input_tokens_seen": 40049680,
      "step": 20800
    },
    {
      "epoch": 3.3939962476547842,
      "grad_norm": 0.001350864302366972,
      "learning_rate": 0.14053439163459308,
      "loss": 0.2028,
      "num_input_tokens_seen": 40058896,
      "step": 20805
    },
    {
      "epoch": 3.3948119748756014,
      "grad_norm": 0.002964758314192295,
      "learning_rate": 0.14047560490386876,
      "loss": 0.2274,
      "num_input_tokens_seen": 40068944,
      "step": 20810
    },
    {
      "epoch": 3.395627702096419,
      "grad_norm": 0.0012672763550654054,
      "learning_rate": 0.14041681964192593,
      "loss": 0.2036,
      "num_input_tokens_seen": 40078144,
      "step": 20815
    },
    {
      "epoch": 3.3964434293172365,
      "grad_norm": 0.0032804920338094234,
      "learning_rate": 0.14035803585782988,
      "loss": 0.2506,
      "num_input_tokens_seen": 40087888,
      "step": 20820
    },
    {
      "epoch": 3.3972591565380537,
      "grad_norm": 0.0016031537670642138,
      "learning_rate": 0.14029925356064593,
      "loss": 0.2352,
      "num_input_tokens_seen": 40098096,
      "step": 20825
    },
    {
      "epoch": 3.398074883758871,
      "grad_norm": 0.00157178845256567,
      "learning_rate": 0.1402404727594389,
      "loss": 0.2091,
      "num_input_tokens_seen": 40108816,
      "step": 20830
    },
    {
      "epoch": 3.3988906109796884,
      "grad_norm": 0.001343485084362328,
      "learning_rate": 0.1401816934632737,
      "loss": 0.2248,
      "num_input_tokens_seen": 40117888,
      "step": 20835
    },
    {
      "epoch": 3.399706338200506,
      "grad_norm": 0.0007966895354911685,
      "learning_rate": 0.1401229156812147,
      "loss": 0.1935,
      "num_input_tokens_seen": 40126368,
      "step": 20840
    },
    {
      "epoch": 3.400522065421323,
      "grad_norm": 0.0013800611486658454,
      "learning_rate": 0.14006413942232626,
      "loss": 0.1716,
      "num_input_tokens_seen": 40135728,
      "step": 20845
    },
    {
      "epoch": 3.4013377926421406,
      "grad_norm": 0.0012727576540783048,
      "learning_rate": 0.14000536469567235,
      "loss": 0.2349,
      "num_input_tokens_seen": 40145504,
      "step": 20850
    },
    {
      "epoch": 3.4021535198629578,
      "grad_norm": 0.0011790699791163206,
      "learning_rate": 0.13994659151031685,
      "loss": 0.2148,
      "num_input_tokens_seen": 40155312,
      "step": 20855
    },
    {
      "epoch": 3.4029692470837754,
      "grad_norm": 0.002126689301803708,
      "learning_rate": 0.13988781987532323,
      "loss": 0.1876,
      "num_input_tokens_seen": 40163840,
      "step": 20860
    },
    {
      "epoch": 3.4037849743045925,
      "grad_norm": 0.0026540583930909634,
      "learning_rate": 0.1398290497997549,
      "loss": 0.2255,
      "num_input_tokens_seen": 40174128,
      "step": 20865
    },
    {
      "epoch": 3.40460070152541,
      "grad_norm": 0.001623008050955832,
      "learning_rate": 0.13977028129267488,
      "loss": 0.2324,
      "num_input_tokens_seen": 40182416,
      "step": 20870
    },
    {
      "epoch": 3.405416428746227,
      "grad_norm": 0.0017906021093949676,
      "learning_rate": 0.13971151436314605,
      "loss": 0.2071,
      "num_input_tokens_seen": 40192848,
      "step": 20875
    },
    {
      "epoch": 3.4062321559670448,
      "grad_norm": 0.0024284280370920897,
      "learning_rate": 0.13965274902023103,
      "loss": 0.2027,
      "num_input_tokens_seen": 40202240,
      "step": 20880
    },
    {
      "epoch": 3.407047883187862,
      "grad_norm": 0.003068220801651478,
      "learning_rate": 0.13959398527299208,
      "loss": 0.2156,
      "num_input_tokens_seen": 40211760,
      "step": 20885
    },
    {
      "epoch": 3.4078636104086795,
      "grad_norm": 0.0014857770875096321,
      "learning_rate": 0.13953522313049138,
      "loss": 0.2255,
      "num_input_tokens_seen": 40221808,
      "step": 20890
    },
    {
      "epoch": 3.4086793376294966,
      "grad_norm": 0.0012710074661299586,
      "learning_rate": 0.13947646260179083,
      "loss": 0.1603,
      "num_input_tokens_seen": 40231152,
      "step": 20895
    },
    {
      "epoch": 3.409495064850314,
      "grad_norm": 0.0025674947537481785,
      "learning_rate": 0.13941770369595194,
      "loss": 0.2516,
      "num_input_tokens_seen": 40242512,
      "step": 20900
    },
    {
      "epoch": 3.4103107920711313,
      "grad_norm": 0.001622311188839376,
      "learning_rate": 0.1393589464220362,
      "loss": 0.2108,
      "num_input_tokens_seen": 40252160,
      "step": 20905
    },
    {
      "epoch": 3.411126519291949,
      "grad_norm": 0.003071011044085026,
      "learning_rate": 0.13930019078910455,
      "loss": 0.2189,
      "num_input_tokens_seen": 40261264,
      "step": 20910
    },
    {
      "epoch": 3.411942246512766,
      "grad_norm": 0.0012011094950139523,
      "learning_rate": 0.139241436806218,
      "loss": 0.2059,
      "num_input_tokens_seen": 40271472,
      "step": 20915
    },
    {
      "epoch": 3.4127579737335836,
      "grad_norm": 0.0024135077837854624,
      "learning_rate": 0.13918268448243712,
      "loss": 0.2159,
      "num_input_tokens_seen": 40282208,
      "step": 20920
    },
    {
      "epoch": 3.4135737009544007,
      "grad_norm": 0.002403420628979802,
      "learning_rate": 0.13912393382682217,
      "loss": 0.2243,
      "num_input_tokens_seen": 40291008,
      "step": 20925
    },
    {
      "epoch": 3.4143894281752183,
      "grad_norm": 0.0014786372194066644,
      "learning_rate": 0.1390651848484333,
      "loss": 0.2173,
      "num_input_tokens_seen": 40301440,
      "step": 20930
    },
    {
      "epoch": 3.4152051553960354,
      "grad_norm": 0.0013326499611139297,
      "learning_rate": 0.1390064375563304,
      "loss": 0.2203,
      "num_input_tokens_seen": 40311328,
      "step": 20935
    },
    {
      "epoch": 3.416020882616853,
      "grad_norm": 0.0015091546811163425,
      "learning_rate": 0.13894769195957293,
      "loss": 0.2321,
      "num_input_tokens_seen": 40320208,
      "step": 20940
    },
    {
      "epoch": 3.41683660983767,
      "grad_norm": 0.0020349607802927494,
      "learning_rate": 0.13888894806722032,
      "loss": 0.2166,
      "num_input_tokens_seen": 40329808,
      "step": 20945
    },
    {
      "epoch": 3.4176523370584877,
      "grad_norm": 0.0013954488094896078,
      "learning_rate": 0.1388302058883315,
      "loss": 0.2073,
      "num_input_tokens_seen": 40338720,
      "step": 20950
    },
    {
      "epoch": 3.418468064279305,
      "grad_norm": 0.0020899013616144657,
      "learning_rate": 0.13877146543196528,
      "loss": 0.2352,
      "num_input_tokens_seen": 40348832,
      "step": 20955
    },
    {
      "epoch": 3.4192837915001224,
      "grad_norm": 0.0009406546014361084,
      "learning_rate": 0.13871272670718027,
      "loss": 0.1809,
      "num_input_tokens_seen": 40357872,
      "step": 20960
    },
    {
      "epoch": 3.4200995187209395,
      "grad_norm": 0.002793113235384226,
      "learning_rate": 0.13865398972303455,
      "loss": 0.2051,
      "num_input_tokens_seen": 40367584,
      "step": 20965
    },
    {
      "epoch": 3.420915245941757,
      "grad_norm": 0.0018134219571948051,
      "learning_rate": 0.13859525448858623,
      "loss": 0.2121,
      "num_input_tokens_seen": 40378064,
      "step": 20970
    },
    {
      "epoch": 3.4217309731625742,
      "grad_norm": 0.0010874677682295442,
      "learning_rate": 0.13853652101289304,
      "loss": 0.2323,
      "num_input_tokens_seen": 40387760,
      "step": 20975
    },
    {
      "epoch": 3.422546700383392,
      "grad_norm": 0.0017438187496736646,
      "learning_rate": 0.13847778930501234,
      "loss": 0.2053,
      "num_input_tokens_seen": 40397824,
      "step": 20980
    },
    {
      "epoch": 3.423362427604209,
      "grad_norm": 0.0017965314909815788,
      "learning_rate": 0.1384190593740013,
      "loss": 0.1926,
      "num_input_tokens_seen": 40408608,
      "step": 20985
    },
    {
      "epoch": 3.4241781548250265,
      "grad_norm": 0.001195087330415845,
      "learning_rate": 0.13836033122891686,
      "loss": 0.1824,
      "num_input_tokens_seen": 40417936,
      "step": 20990
    },
    {
      "epoch": 3.424993882045844,
      "grad_norm": 0.001595670124515891,
      "learning_rate": 0.1383016048788156,
      "loss": 0.2071,
      "num_input_tokens_seen": 40427600,
      "step": 20995
    },
    {
      "epoch": 3.4258096092666612,
      "grad_norm": 0.001559269381687045,
      "learning_rate": 0.13824288033275392,
      "loss": 0.2229,
      "num_input_tokens_seen": 40436560,
      "step": 21000
    },
    {
      "epoch": 3.4258096092666612,
      "eval_loss": 0.2065121978521347,
      "eval_runtime": 68.0265,
      "eval_samples_per_second": 40.058,
      "eval_steps_per_second": 20.036,
      "num_input_tokens_seen": 40436560,
      "step": 21000
    },
    {
      "epoch": 3.4266253364874784,
      "grad_norm": 0.001678924891166389,
      "learning_rate": 0.1381841575997878,
      "loss": 0.1928,
      "num_input_tokens_seen": 40446528,
      "step": 21005
    },
    {
      "epoch": 3.427441063708296,
      "grad_norm": 0.0026199377607554197,
      "learning_rate": 0.13812543668897306,
      "loss": 0.1796,
      "num_input_tokens_seen": 40455296,
      "step": 21010
    },
    {
      "epoch": 3.4282567909291135,
      "grad_norm": 0.0022121097426861525,
      "learning_rate": 0.13806671760936526,
      "loss": 0.2382,
      "num_input_tokens_seen": 40465712,
      "step": 21015
    },
    {
      "epoch": 3.4290725181499306,
      "grad_norm": 0.00207288539968431,
      "learning_rate": 0.13800800037001956,
      "loss": 0.2378,
      "num_input_tokens_seen": 40474880,
      "step": 21020
    },
    {
      "epoch": 3.429888245370748,
      "grad_norm": 0.0017545655136927962,
      "learning_rate": 0.13794928497999087,
      "loss": 0.2348,
      "num_input_tokens_seen": 40483392,
      "step": 21025
    },
    {
      "epoch": 3.4307039725915653,
      "grad_norm": 0.0015912012895569205,
      "learning_rate": 0.1378905714483339,
      "loss": 0.1677,
      "num_input_tokens_seen": 40492400,
      "step": 21030
    },
    {
      "epoch": 3.431519699812383,
      "grad_norm": 0.003029563231393695,
      "learning_rate": 0.13783185978410295,
      "loss": 0.2464,
      "num_input_tokens_seen": 40502704,
      "step": 21035
    },
    {
      "epoch": 3.4323354270332,
      "grad_norm": 0.0032945831771939993,
      "learning_rate": 0.13777314999635218,
      "loss": 0.2121,
      "num_input_tokens_seen": 40511152,
      "step": 21040
    },
    {
      "epoch": 3.4331511542540176,
      "grad_norm": 0.002014985540881753,
      "learning_rate": 0.1377144420941353,
      "loss": 0.1664,
      "num_input_tokens_seen": 40521648,
      "step": 21045
    },
    {
      "epoch": 3.4339668814748348,
      "grad_norm": 0.001438884879462421,
      "learning_rate": 0.13765573608650586,
      "loss": 0.1969,
      "num_input_tokens_seen": 40531360,
      "step": 21050
    },
    {
      "epoch": 3.4347826086956523,
      "grad_norm": 0.002133224857971072,
      "learning_rate": 0.13759703198251702,
      "loss": 0.1804,
      "num_input_tokens_seen": 40541200,
      "step": 21055
    },
    {
      "epoch": 3.4355983359164695,
      "grad_norm": 0.0029228234197944403,
      "learning_rate": 0.13753832979122174,
      "loss": 0.1729,
      "num_input_tokens_seen": 40551104,
      "step": 21060
    },
    {
      "epoch": 3.436414063137287,
      "grad_norm": 0.003622530261054635,
      "learning_rate": 0.13747962952167264,
      "loss": 0.2327,
      "num_input_tokens_seen": 40560608,
      "step": 21065
    },
    {
      "epoch": 3.437229790358104,
      "grad_norm": 0.00176885852124542,
      "learning_rate": 0.13742093118292192,
      "loss": 0.2209,
      "num_input_tokens_seen": 40570208,
      "step": 21070
    },
    {
      "epoch": 3.4380455175789217,
      "grad_norm": 0.00262393313460052,
      "learning_rate": 0.13736223478402174,
      "loss": 0.2008,
      "num_input_tokens_seen": 40578160,
      "step": 21075
    },
    {
      "epoch": 3.438861244799739,
      "grad_norm": 0.0031368513591587543,
      "learning_rate": 0.1373035403340238,
      "loss": 0.2203,
      "num_input_tokens_seen": 40587792,
      "step": 21080
    },
    {
      "epoch": 3.4396769720205564,
      "grad_norm": 0.0019009726820513606,
      "learning_rate": 0.13724484784197943,
      "loss": 0.222,
      "num_input_tokens_seen": 40596800,
      "step": 21085
    },
    {
      "epoch": 3.4404926992413736,
      "grad_norm": 0.0033658372703939676,
      "learning_rate": 0.13718615731693987,
      "loss": 0.2044,
      "num_input_tokens_seen": 40605952,
      "step": 21090
    },
    {
      "epoch": 3.441308426462191,
      "grad_norm": 0.0021232282742857933,
      "learning_rate": 0.13712746876795587,
      "loss": 0.2812,
      "num_input_tokens_seen": 40615456,
      "step": 21095
    },
    {
      "epoch": 3.4421241536830083,
      "grad_norm": 0.0022811663802713156,
      "learning_rate": 0.13706878220407792,
      "loss": 0.2459,
      "num_input_tokens_seen": 40625536,
      "step": 21100
    },
    {
      "epoch": 3.442939880903826,
      "grad_norm": 0.0020147417671978474,
      "learning_rate": 0.13701009763435631,
      "loss": 0.2056,
      "num_input_tokens_seen": 40636000,
      "step": 21105
    },
    {
      "epoch": 3.443755608124643,
      "grad_norm": 0.003143276320770383,
      "learning_rate": 0.13695141506784084,
      "loss": 0.2204,
      "num_input_tokens_seen": 40646032,
      "step": 21110
    },
    {
      "epoch": 3.4445713353454606,
      "grad_norm": 0.0014706070069223642,
      "learning_rate": 0.13689273451358114,
      "loss": 0.2542,
      "num_input_tokens_seen": 40656016,
      "step": 21115
    },
    {
      "epoch": 3.4453870625662777,
      "grad_norm": 0.0012634925078600645,
      "learning_rate": 0.13683405598062653,
      "loss": 0.2149,
      "num_input_tokens_seen": 40667584,
      "step": 21120
    },
    {
      "epoch": 3.4462027897870953,
      "grad_norm": 0.0014563316944986582,
      "learning_rate": 0.1367753794780259,
      "loss": 0.2159,
      "num_input_tokens_seen": 40678176,
      "step": 21125
    },
    {
      "epoch": 3.4470185170079124,
      "grad_norm": 0.0015230984427034855,
      "learning_rate": 0.13671670501482802,
      "loss": 0.2125,
      "num_input_tokens_seen": 40686480,
      "step": 21130
    },
    {
      "epoch": 3.44783424422873,
      "grad_norm": 0.0012798880925402045,
      "learning_rate": 0.1366580326000811,
      "loss": 0.2004,
      "num_input_tokens_seen": 40696816,
      "step": 21135
    },
    {
      "epoch": 3.448649971449547,
      "grad_norm": 0.0009738304652273655,
      "learning_rate": 0.1365993622428332,
      "loss": 0.1977,
      "num_input_tokens_seen": 40706736,
      "step": 21140
    },
    {
      "epoch": 3.4494656986703647,
      "grad_norm": 0.0026154478546231985,
      "learning_rate": 0.13654069395213211,
      "loss": 0.2168,
      "num_input_tokens_seen": 40716192,
      "step": 21145
    },
    {
      "epoch": 3.450281425891182,
      "grad_norm": 0.0016924089286476374,
      "learning_rate": 0.13648202773702509,
      "loss": 0.2262,
      "num_input_tokens_seen": 40725520,
      "step": 21150
    },
    {
      "epoch": 3.4510971531119994,
      "grad_norm": 0.002528631826862693,
      "learning_rate": 0.13642336360655927,
      "loss": 0.2225,
      "num_input_tokens_seen": 40734800,
      "step": 21155
    },
    {
      "epoch": 3.4519128803328165,
      "grad_norm": 0.0014836465707048774,
      "learning_rate": 0.13636470156978145,
      "loss": 0.2144,
      "num_input_tokens_seen": 40744208,
      "step": 21160
    },
    {
      "epoch": 3.452728607553634,
      "grad_norm": 0.0009587031090632081,
      "learning_rate": 0.13630604163573798,
      "loss": 0.2124,
      "num_input_tokens_seen": 40753936,
      "step": 21165
    },
    {
      "epoch": 3.4535443347744517,
      "grad_norm": 0.0015643320512026548,
      "learning_rate": 0.13624738381347495,
      "loss": 0.241,
      "num_input_tokens_seen": 40764208,
      "step": 21170
    },
    {
      "epoch": 3.454360061995269,
      "grad_norm": 0.0011680681491270661,
      "learning_rate": 0.1361887281120382,
      "loss": 0.2143,
      "num_input_tokens_seen": 40774272,
      "step": 21175
    },
    {
      "epoch": 3.455175789216086,
      "grad_norm": 0.0018410227494314313,
      "learning_rate": 0.13613007454047307,
      "loss": 0.2306,
      "num_input_tokens_seen": 40783696,
      "step": 21180
    },
    {
      "epoch": 3.4559915164369035,
      "grad_norm": 0.0016529904678463936,
      "learning_rate": 0.13607142310782486,
      "loss": 0.2281,
      "num_input_tokens_seen": 40793376,
      "step": 21185
    },
    {
      "epoch": 3.456807243657721,
      "grad_norm": 0.001524485182017088,
      "learning_rate": 0.13601277382313814,
      "loss": 0.1871,
      "num_input_tokens_seen": 40801920,
      "step": 21190
    },
    {
      "epoch": 3.457622970878538,
      "grad_norm": 0.0017961693229153752,
      "learning_rate": 0.1359541266954575,
      "loss": 0.2583,
      "num_input_tokens_seen": 40810880,
      "step": 21195
    },
    {
      "epoch": 3.4584386980993553,
      "grad_norm": 0.0009107418009079993,
      "learning_rate": 0.13589548173382707,
      "loss": 0.2153,
      "num_input_tokens_seen": 40820704,
      "step": 21200
    },
    {
      "epoch": 3.4584386980993553,
      "eval_loss": 0.2072586864233017,
      "eval_runtime": 67.9662,
      "eval_samples_per_second": 40.093,
      "eval_steps_per_second": 20.054,
      "num_input_tokens_seen": 40820704,
      "step": 21200
    },
    {
      "epoch": 3.459254425320173,
      "grad_norm": 0.0013041668571531773,
      "learning_rate": 0.1358368389472906,
      "loss": 0.2244,
      "num_input_tokens_seen": 40830592,
      "step": 21205
    },
    {
      "epoch": 3.4600701525409905,
      "grad_norm": 0.0017525319708511233,
      "learning_rate": 0.13577819834489155,
      "loss": 0.2457,
      "num_input_tokens_seen": 40840160,
      "step": 21210
    },
    {
      "epoch": 3.4608858797618076,
      "grad_norm": 0.0013486194657161832,
      "learning_rate": 0.135719559935673,
      "loss": 0.232,
      "num_input_tokens_seen": 40849248,
      "step": 21215
    },
    {
      "epoch": 3.461701606982625,
      "grad_norm": 0.0017811545403674245,
      "learning_rate": 0.13566092372867775,
      "loss": 0.1813,
      "num_input_tokens_seen": 40859648,
      "step": 21220
    },
    {
      "epoch": 3.4625173342034423,
      "grad_norm": 0.0016152458265423775,
      "learning_rate": 0.13560228973294833,
      "loss": 0.2286,
      "num_input_tokens_seen": 40868672,
      "step": 21225
    },
    {
      "epoch": 3.46333306142426,
      "grad_norm": 0.002086964901536703,
      "learning_rate": 0.13554365795752668,
      "loss": 0.2012,
      "num_input_tokens_seen": 40877552,
      "step": 21230
    },
    {
      "epoch": 3.464148788645077,
      "grad_norm": 0.001509175868704915,
      "learning_rate": 0.1354850284114547,
      "loss": 0.2206,
      "num_input_tokens_seen": 40886224,
      "step": 21235
    },
    {
      "epoch": 3.4649645158658946,
      "grad_norm": 0.003005284583196044,
      "learning_rate": 0.13542640110377374,
      "loss": 0.2441,
      "num_input_tokens_seen": 40894464,
      "step": 21240
    },
    {
      "epoch": 3.4657802430867117,
      "grad_norm": 0.0011401427909731865,
      "learning_rate": 0.13536777604352487,
      "loss": 0.2383,
      "num_input_tokens_seen": 40904688,
      "step": 21245
    },
    {
      "epoch": 3.4665959703075293,
      "grad_norm": 0.0018015699461102486,
      "learning_rate": 0.13530915323974887,
      "loss": 0.1681,
      "num_input_tokens_seen": 40914304,
      "step": 21250
    },
    {
      "epoch": 3.4674116975283464,
      "grad_norm": 0.0013143123360350728,
      "learning_rate": 0.13525053270148596,
      "loss": 0.2332,
      "num_input_tokens_seen": 40923504,
      "step": 21255
    },
    {
      "epoch": 3.468227424749164,
      "grad_norm": 0.0012885869946330786,
      "learning_rate": 0.13519191443777628,
      "loss": 0.1834,
      "num_input_tokens_seen": 40933536,
      "step": 21260
    },
    {
      "epoch": 3.469043151969981,
      "grad_norm": 0.0022673553321510553,
      "learning_rate": 0.13513329845765953,
      "loss": 0.3008,
      "num_input_tokens_seen": 40942336,
      "step": 21265
    },
    {
      "epoch": 3.4698588791907987,
      "grad_norm": 0.0018656650790944695,
      "learning_rate": 0.13507468477017495,
      "loss": 0.218,
      "num_input_tokens_seen": 40952096,
      "step": 21270
    },
    {
      "epoch": 3.470674606411616,
      "grad_norm": 0.0019863329362124205,
      "learning_rate": 0.13501607338436153,
      "loss": 0.2369,
      "num_input_tokens_seen": 40962416,
      "step": 21275
    },
    {
      "epoch": 3.4714903336324334,
      "grad_norm": 0.0016584547702223063,
      "learning_rate": 0.13495746430925798,
      "loss": 0.23,
      "num_input_tokens_seen": 40971680,
      "step": 21280
    },
    {
      "epoch": 3.4723060608532506,
      "grad_norm": 0.002066980116069317,
      "learning_rate": 0.13489885755390238,
      "loss": 0.2313,
      "num_input_tokens_seen": 40982656,
      "step": 21285
    },
    {
      "epoch": 3.473121788074068,
      "grad_norm": 0.0014260491589084268,
      "learning_rate": 0.13484025312733275,
      "loss": 0.1471,
      "num_input_tokens_seen": 40990560,
      "step": 21290
    },
    {
      "epoch": 3.4739375152948853,
      "grad_norm": 0.0023722141049802303,
      "learning_rate": 0.13478165103858658,
      "loss": 0.2034,
      "num_input_tokens_seen": 41000416,
      "step": 21295
    },
    {
      "epoch": 3.474753242515703,
      "grad_norm": 0.002852557459846139,
      "learning_rate": 0.13472305129670106,
      "loss": 0.2281,
      "num_input_tokens_seen": 41010144,
      "step": 21300
    },
    {
      "epoch": 3.47556896973652,
      "grad_norm": 0.001216256059706211,
      "learning_rate": 0.13466445391071305,
      "loss": 0.2485,
      "num_input_tokens_seen": 41020288,
      "step": 21305
    },
    {
      "epoch": 3.4763846969573375,
      "grad_norm": 0.0023828940466046333,
      "learning_rate": 0.13460585888965895,
      "loss": 0.198,
      "num_input_tokens_seen": 41029040,
      "step": 21310
    },
    {
      "epoch": 3.4772004241781547,
      "grad_norm": 0.0025763658341020346,
      "learning_rate": 0.13454726624257482,
      "loss": 0.1909,
      "num_input_tokens_seen": 41039728,
      "step": 21315
    },
    {
      "epoch": 3.4780161513989722,
      "grad_norm": 0.001814188901335001,
      "learning_rate": 0.1344886759784965,
      "loss": 0.1928,
      "num_input_tokens_seen": 41048736,
      "step": 21320
    },
    {
      "epoch": 3.4788318786197894,
      "grad_norm": 0.0018277913331985474,
      "learning_rate": 0.13443008810645923,
      "loss": 0.2142,
      "num_input_tokens_seen": 41057824,
      "step": 21325
    },
    {
      "epoch": 3.479647605840607,
      "grad_norm": 0.0008466334547847509,
      "learning_rate": 0.13437150263549807,
      "loss": 0.215,
      "num_input_tokens_seen": 41067824,
      "step": 21330
    },
    {
      "epoch": 3.480463333061424,
      "grad_norm": 0.0014994360972195864,
      "learning_rate": 0.13431291957464755,
      "loss": 0.2073,
      "num_input_tokens_seen": 41078144,
      "step": 21335
    },
    {
      "epoch": 3.4812790602822417,
      "grad_norm": 0.0016469020629301667,
      "learning_rate": 0.13425433893294197,
      "loss": 0.1508,
      "num_input_tokens_seen": 41086896,
      "step": 21340
    },
    {
      "epoch": 3.482094787503059,
      "grad_norm": 0.001558731310069561,
      "learning_rate": 0.13419576071941525,
      "loss": 0.2494,
      "num_input_tokens_seen": 41096944,
      "step": 21345
    },
    {
      "epoch": 3.4829105147238764,
      "grad_norm": 0.0021288509014993906,
      "learning_rate": 0.1341371849431008,
      "loss": 0.2237,
      "num_input_tokens_seen": 41107904,
      "step": 21350
    },
    {
      "epoch": 3.4837262419446935,
      "grad_norm": 0.002309065079316497,
      "learning_rate": 0.13407861161303178,
      "loss": 0.2319,
      "num_input_tokens_seen": 41117920,
      "step": 21355
    },
    {
      "epoch": 3.484541969165511,
      "grad_norm": 0.0022132352460175753,
      "learning_rate": 0.13402004073824098,
      "loss": 0.2522,
      "num_input_tokens_seen": 41126784,
      "step": 21360
    },
    {
      "epoch": 3.4853576963863286,
      "grad_norm": 0.0024218715261667967,
      "learning_rate": 0.13396147232776062,
      "loss": 0.231,
      "num_input_tokens_seen": 41135792,
      "step": 21365
    },
    {
      "epoch": 3.4861734236071458,
      "grad_norm": 0.0021219614427536726,
      "learning_rate": 0.13390290639062288,
      "loss": 0.1991,
      "num_input_tokens_seen": 41145840,
      "step": 21370
    },
    {
      "epoch": 3.486989150827963,
      "grad_norm": 0.002286199713125825,
      "learning_rate": 0.13384434293585917,
      "loss": 0.2437,
      "num_input_tokens_seen": 41156224,
      "step": 21375
    },
    {
      "epoch": 3.4878048780487805,
      "grad_norm": 0.0028409655205905437,
      "learning_rate": 0.13378578197250088,
      "loss": 0.2184,
      "num_input_tokens_seen": 41165776,
      "step": 21380
    },
    {
      "epoch": 3.488620605269598,
      "grad_norm": 0.0019951460417360067,
      "learning_rate": 0.13372722350957872,
      "loss": 0.2237,
      "num_input_tokens_seen": 41175216,
      "step": 21385
    },
    {
      "epoch": 3.489436332490415,
      "grad_norm": 0.0015969609376043081,
      "learning_rate": 0.13366866755612322,
      "loss": 0.19,
      "num_input_tokens_seen": 41184656,
      "step": 21390
    },
    {
      "epoch": 3.4902520597112328,
      "grad_norm": 0.001362510840408504,
      "learning_rate": 0.13361011412116436,
      "loss": 0.2185,
      "num_input_tokens_seen": 41193888,
      "step": 21395
    },
    {
      "epoch": 3.49106778693205,
      "grad_norm": 0.0022827035281807184,
      "learning_rate": 0.13355156321373196,
      "loss": 0.2711,
      "num_input_tokens_seen": 41202080,
      "step": 21400
    },
    {
      "epoch": 3.49106778693205,
      "eval_loss": 0.20800067484378815,
      "eval_runtime": 68.0542,
      "eval_samples_per_second": 40.042,
      "eval_steps_per_second": 20.028,
      "num_input_tokens_seen": 41202080,
      "step": 21400
    },
    {
      "epoch": 3.4918835141528675,
      "grad_norm": 0.0023350915871560574,
      "learning_rate": 0.13349301484285514,
      "loss": 0.2085,
      "num_input_tokens_seen": 41211888,
      "step": 21405
    },
    {
      "epoch": 3.4926992413736846,
      "grad_norm": 0.001769774709828198,
      "learning_rate": 0.13343446901756295,
      "loss": 0.1886,
      "num_input_tokens_seen": 41220832,
      "step": 21410
    },
    {
      "epoch": 3.493514968594502,
      "grad_norm": 0.0018773607444018126,
      "learning_rate": 0.13337592574688376,
      "loss": 0.1858,
      "num_input_tokens_seen": 41229600,
      "step": 21415
    },
    {
      "epoch": 3.4943306958153193,
      "grad_norm": 0.0030846905428916216,
      "learning_rate": 0.13331738503984572,
      "loss": 0.2458,
      "num_input_tokens_seen": 41238416,
      "step": 21420
    },
    {
      "epoch": 3.495146423036137,
      "grad_norm": 0.003059802809730172,
      "learning_rate": 0.1332588469054766,
      "loss": 0.2811,
      "num_input_tokens_seen": 41248816,
      "step": 21425
    },
    {
      "epoch": 3.495962150256954,
      "grad_norm": 0.0017869886942207813,
      "learning_rate": 0.1332003113528036,
      "loss": 0.1954,
      "num_input_tokens_seen": 41257872,
      "step": 21430
    },
    {
      "epoch": 3.4967778774777716,
      "grad_norm": 0.0012442775769159198,
      "learning_rate": 0.13314177839085373,
      "loss": 0.2005,
      "num_input_tokens_seen": 41267920,
      "step": 21435
    },
    {
      "epoch": 3.4975936046985887,
      "grad_norm": 0.0020600815769284964,
      "learning_rate": 0.13308324802865354,
      "loss": 0.2058,
      "num_input_tokens_seen": 41278000,
      "step": 21440
    },
    {
      "epoch": 3.4984093319194063,
      "grad_norm": 0.0023206237237900496,
      "learning_rate": 0.13302472027522905,
      "loss": 0.2102,
      "num_input_tokens_seen": 41287872,
      "step": 21445
    },
    {
      "epoch": 3.4992250591402234,
      "grad_norm": 0.0018581870244815946,
      "learning_rate": 0.13296619513960606,
      "loss": 0.2012,
      "num_input_tokens_seen": 41297840,
      "step": 21450
    },
    {
      "epoch": 3.500040786361041,
      "grad_norm": 0.001142972381785512,
      "learning_rate": 0.1329076726308098,
      "loss": 0.2282,
      "num_input_tokens_seen": 41307440,
      "step": 21455
    },
    {
      "epoch": 3.500856513581858,
      "grad_norm": 0.001704719732515514,
      "learning_rate": 0.13284915275786519,
      "loss": 0.2601,
      "num_input_tokens_seen": 41315840,
      "step": 21460
    },
    {
      "epoch": 3.5016722408026757,
      "grad_norm": 0.002557921689003706,
      "learning_rate": 0.1327906355297968,
      "loss": 0.2427,
      "num_input_tokens_seen": 41325216,
      "step": 21465
    },
    {
      "epoch": 3.502487968023493,
      "grad_norm": 0.0017411899752914906,
      "learning_rate": 0.13273212095562867,
      "loss": 0.2324,
      "num_input_tokens_seen": 41334768,
      "step": 21470
    },
    {
      "epoch": 3.5033036952443104,
      "grad_norm": 0.001761924591846764,
      "learning_rate": 0.13267360904438444,
      "loss": 0.1767,
      "num_input_tokens_seen": 41344432,
      "step": 21475
    },
    {
      "epoch": 3.5041194224651275,
      "grad_norm": 0.002473600674420595,
      "learning_rate": 0.1326150998050875,
      "loss": 0.2386,
      "num_input_tokens_seen": 41353408,
      "step": 21480
    },
    {
      "epoch": 3.504935149685945,
      "grad_norm": 0.0014492771588265896,
      "learning_rate": 0.1325565932467606,
      "loss": 0.2319,
      "num_input_tokens_seen": 41362272,
      "step": 21485
    },
    {
      "epoch": 3.5057508769067622,
      "grad_norm": 0.0013756667030975223,
      "learning_rate": 0.13249808937842628,
      "loss": 0.2152,
      "num_input_tokens_seen": 41372336,
      "step": 21490
    },
    {
      "epoch": 3.50656660412758,
      "grad_norm": 0.0028855002019554377,
      "learning_rate": 0.1324395882091065,
      "loss": 0.2343,
      "num_input_tokens_seen": 41382368,
      "step": 21495
    },
    {
      "epoch": 3.507382331348397,
      "grad_norm": 0.002534457016736269,
      "learning_rate": 0.13238108974782284,
      "loss": 0.2177,
      "num_input_tokens_seen": 41392672,
      "step": 21500
    },
    {
      "epoch": 3.5081980585692145,
      "grad_norm": 0.0012938572326675057,
      "learning_rate": 0.13232259400359664,
      "loss": 0.1695,
      "num_input_tokens_seen": 41402992,
      "step": 21505
    },
    {
      "epoch": 3.5090137857900316,
      "grad_norm": 0.0015945605700835586,
      "learning_rate": 0.13226410098544852,
      "loss": 0.2127,
      "num_input_tokens_seen": 41413312,
      "step": 21510
    },
    {
      "epoch": 3.5098295130108492,
      "grad_norm": 0.0022132634185254574,
      "learning_rate": 0.13220561070239892,
      "loss": 0.2755,
      "num_input_tokens_seen": 41422448,
      "step": 21515
    },
    {
      "epoch": 3.510645240231667,
      "grad_norm": 0.002346470020711422,
      "learning_rate": 0.13214712316346783,
      "loss": 0.2023,
      "num_input_tokens_seen": 41432224,
      "step": 21520
    },
    {
      "epoch": 3.511460967452484,
      "grad_norm": 0.0024731350131332874,
      "learning_rate": 0.13208863837767465,
      "loss": 0.2155,
      "num_input_tokens_seen": 41441712,
      "step": 21525
    },
    {
      "epoch": 3.512276694673301,
      "grad_norm": 0.0009279125370085239,
      "learning_rate": 0.13203015635403856,
      "loss": 0.1796,
      "num_input_tokens_seen": 41451232,
      "step": 21530
    },
    {
      "epoch": 3.5130924218941186,
      "grad_norm": 0.002182003576308489,
      "learning_rate": 0.13197167710157817,
      "loss": 0.1817,
      "num_input_tokens_seen": 41460960,
      "step": 21535
    },
    {
      "epoch": 3.513908149114936,
      "grad_norm": 0.0034190050791949034,
      "learning_rate": 0.13191320062931167,
      "loss": 0.2239,
      "num_input_tokens_seen": 41471072,
      "step": 21540
    },
    {
      "epoch": 3.5147238763357533,
      "grad_norm": 0.0013116071932017803,
      "learning_rate": 0.13185472694625702,
      "loss": 0.2226,
      "num_input_tokens_seen": 41481680,
      "step": 21545
    },
    {
      "epoch": 3.5155396035565705,
      "grad_norm": 0.003525950014591217,
      "learning_rate": 0.13179625606143142,
      "loss": 0.2524,
      "num_input_tokens_seen": 41491216,
      "step": 21550
    },
    {
      "epoch": 3.516355330777388,
      "grad_norm": 0.001201279112137854,
      "learning_rate": 0.13173778798385188,
      "loss": 0.2039,
      "num_input_tokens_seen": 41501232,
      "step": 21555
    },
    {
      "epoch": 3.5171710579982056,
      "grad_norm": 0.0018421000568196177,
      "learning_rate": 0.13167932272253505,
      "loss": 0.2243,
      "num_input_tokens_seen": 41510656,
      "step": 21560
    },
    {
      "epoch": 3.5179867852190227,
      "grad_norm": 0.0030847513116896152,
      "learning_rate": 0.1316208602864968,
      "loss": 0.2223,
      "num_input_tokens_seen": 41520864,
      "step": 21565
    },
    {
      "epoch": 3.51880251243984,
      "grad_norm": 0.001574751571752131,
      "learning_rate": 0.13156240068475292,
      "loss": 0.2348,
      "num_input_tokens_seen": 41529264,
      "step": 21570
    },
    {
      "epoch": 3.5196182396606575,
      "grad_norm": 0.003951877821236849,
      "learning_rate": 0.1315039439263185,
      "loss": 0.2285,
      "num_input_tokens_seen": 41540144,
      "step": 21575
    },
    {
      "epoch": 3.520433966881475,
      "grad_norm": 0.0009781618136912584,
      "learning_rate": 0.13144549002020833,
      "loss": 0.2141,
      "num_input_tokens_seen": 41549808,
      "step": 21580
    },
    {
      "epoch": 3.521249694102292,
      "grad_norm": 0.0017319602193310857,
      "learning_rate": 0.13138703897543688,
      "loss": 0.191,
      "num_input_tokens_seen": 41558192,
      "step": 21585
    },
    {
      "epoch": 3.5220654213231093,
      "grad_norm": 0.0023397887125611305,
      "learning_rate": 0.1313285908010178,
      "loss": 0.2059,
      "num_input_tokens_seen": 41568624,
      "step": 21590
    },
    {
      "epoch": 3.522881148543927,
      "grad_norm": 0.0024058434646576643,
      "learning_rate": 0.13127014550596475,
      "loss": 0.2454,
      "num_input_tokens_seen": 41578848,
      "step": 21595
    },
    {
      "epoch": 3.5236968757647444,
      "grad_norm": 0.002029478782787919,
      "learning_rate": 0.1312117030992906,
      "loss": 0.2189,
      "num_input_tokens_seen": 41588560,
      "step": 21600
    },
    {
      "epoch": 3.5236968757647444,
      "eval_loss": 0.21699471771717072,
      "eval_runtime": 67.9223,
      "eval_samples_per_second": 40.119,
      "eval_steps_per_second": 20.067,
      "num_input_tokens_seen": 41588560,
      "step": 21600
    },
    {
      "epoch": 3.5245126029855616,
      "grad_norm": 0.0010947518749162555,
      "learning_rate": 0.13115326359000795,
      "loss": 0.2158,
      "num_input_tokens_seen": 41597536,
      "step": 21605
    },
    {
      "epoch": 3.525328330206379,
      "grad_norm": 0.0024325596168637276,
      "learning_rate": 0.13109482698712896,
      "loss": 0.2038,
      "num_input_tokens_seen": 41606080,
      "step": 21610
    },
    {
      "epoch": 3.5261440574271963,
      "grad_norm": 0.003602794371545315,
      "learning_rate": 0.1310363932996651,
      "loss": 0.214,
      "num_input_tokens_seen": 41615952,
      "step": 21615
    },
    {
      "epoch": 3.526959784648014,
      "grad_norm": 0.0021106689237058163,
      "learning_rate": 0.13097796253662775,
      "loss": 0.1901,
      "num_input_tokens_seen": 41624480,
      "step": 21620
    },
    {
      "epoch": 3.527775511868831,
      "grad_norm": 0.0019189348677173257,
      "learning_rate": 0.1309195347070277,
      "loss": 0.2185,
      "num_input_tokens_seen": 41633904,
      "step": 21625
    },
    {
      "epoch": 3.5285912390896486,
      "grad_norm": 0.001119087217375636,
      "learning_rate": 0.13086110981987506,
      "loss": 0.2027,
      "num_input_tokens_seen": 41644000,
      "step": 21630
    },
    {
      "epoch": 3.5294069663104657,
      "grad_norm": 0.002469309838488698,
      "learning_rate": 0.13080268788417987,
      "loss": 0.1957,
      "num_input_tokens_seen": 41652720,
      "step": 21635
    },
    {
      "epoch": 3.5302226935312833,
      "grad_norm": 0.002545478055253625,
      "learning_rate": 0.1307442689089515,
      "loss": 0.2271,
      "num_input_tokens_seen": 41662144,
      "step": 21640
    },
    {
      "epoch": 3.5310384207521004,
      "grad_norm": 0.001433705911040306,
      "learning_rate": 0.13068585290319873,
      "loss": 0.1636,
      "num_input_tokens_seen": 41671984,
      "step": 21645
    },
    {
      "epoch": 3.531854147972918,
      "grad_norm": 0.0023532703053206205,
      "learning_rate": 0.13062743987593026,
      "loss": 0.237,
      "num_input_tokens_seen": 41681312,
      "step": 21650
    },
    {
      "epoch": 3.532669875193735,
      "grad_norm": 0.00220701121725142,
      "learning_rate": 0.13056902983615395,
      "loss": 0.1996,
      "num_input_tokens_seen": 41691056,
      "step": 21655
    },
    {
      "epoch": 3.5334856024145527,
      "grad_norm": 0.0018589134560897946,
      "learning_rate": 0.13051062279287742,
      "loss": 0.1804,
      "num_input_tokens_seen": 41701520,
      "step": 21660
    },
    {
      "epoch": 3.53430132963537,
      "grad_norm": 0.003665987169370055,
      "learning_rate": 0.13045221875510782,
      "loss": 0.2256,
      "num_input_tokens_seen": 41710752,
      "step": 21665
    },
    {
      "epoch": 3.5351170568561874,
      "grad_norm": 0.002664187690243125,
      "learning_rate": 0.13039381773185174,
      "loss": 0.1958,
      "num_input_tokens_seen": 41720752,
      "step": 21670
    },
    {
      "epoch": 3.5359327840770045,
      "grad_norm": 0.0032664253376424313,
      "learning_rate": 0.1303354197321153,
      "loss": 0.2272,
      "num_input_tokens_seen": 41731504,
      "step": 21675
    },
    {
      "epoch": 3.536748511297822,
      "grad_norm": 0.002979992190375924,
      "learning_rate": 0.13027702476490433,
      "loss": 0.1941,
      "num_input_tokens_seen": 41741936,
      "step": 21680
    },
    {
      "epoch": 3.537564238518639,
      "grad_norm": 0.001933246268890798,
      "learning_rate": 0.1302186328392239,
      "loss": 0.1992,
      "num_input_tokens_seen": 41751968,
      "step": 21685
    },
    {
      "epoch": 3.538379965739457,
      "grad_norm": 0.0021457255352288485,
      "learning_rate": 0.130160243964079,
      "loss": 0.2051,
      "num_input_tokens_seen": 41761808,
      "step": 21690
    },
    {
      "epoch": 3.539195692960274,
      "grad_norm": 0.0019667514134198427,
      "learning_rate": 0.13010185814847372,
      "loss": 0.2271,
      "num_input_tokens_seen": 41771504,
      "step": 21695
    },
    {
      "epoch": 3.5400114201810915,
      "grad_norm": 0.0016818305011838675,
      "learning_rate": 0.13004347540141192,
      "loss": 0.3037,
      "num_input_tokens_seen": 41782160,
      "step": 21700
    },
    {
      "epoch": 3.5408271474019086,
      "grad_norm": 0.0008446758147329092,
      "learning_rate": 0.12998509573189712,
      "loss": 0.2628,
      "num_input_tokens_seen": 41791664,
      "step": 21705
    },
    {
      "epoch": 3.541642874622726,
      "grad_norm": 0.0016846999060362577,
      "learning_rate": 0.12992671914893203,
      "loss": 0.2173,
      "num_input_tokens_seen": 41800896,
      "step": 21710
    },
    {
      "epoch": 3.5424586018435438,
      "grad_norm": 0.0016833175905048847,
      "learning_rate": 0.12986834566151909,
      "loss": 0.2351,
      "num_input_tokens_seen": 41811424,
      "step": 21715
    },
    {
      "epoch": 3.543274329064361,
      "grad_norm": 0.0012696267804130912,
      "learning_rate": 0.12980997527866028,
      "loss": 0.1689,
      "num_input_tokens_seen": 41821440,
      "step": 21720
    },
    {
      "epoch": 3.544090056285178,
      "grad_norm": 0.001239925972186029,
      "learning_rate": 0.12975160800935692,
      "loss": 0.2047,
      "num_input_tokens_seen": 41829776,
      "step": 21725
    },
    {
      "epoch": 3.5449057835059956,
      "grad_norm": 0.0023794735316187143,
      "learning_rate": 0.12969324386261016,
      "loss": 0.1936,
      "num_input_tokens_seen": 41840512,
      "step": 21730
    },
    {
      "epoch": 3.545721510726813,
      "grad_norm": 0.002077195793390274,
      "learning_rate": 0.12963488284742034,
      "loss": 0.2452,
      "num_input_tokens_seen": 41849952,
      "step": 21735
    },
    {
      "epoch": 3.5465372379476303,
      "grad_norm": 0.002639466430991888,
      "learning_rate": 0.12957652497278752,
      "loss": 0.1988,
      "num_input_tokens_seen": 41858944,
      "step": 21740
    },
    {
      "epoch": 3.5473529651684474,
      "grad_norm": 0.0016393024707213044,
      "learning_rate": 0.12951817024771117,
      "loss": 0.209,
      "num_input_tokens_seen": 41869760,
      "step": 21745
    },
    {
      "epoch": 3.548168692389265,
      "grad_norm": 0.0016547669656574726,
      "learning_rate": 0.12945981868119041,
      "loss": 0.195,
      "num_input_tokens_seen": 41879072,
      "step": 21750
    },
    {
      "epoch": 3.5489844196100826,
      "grad_norm": 0.0016072764992713928,
      "learning_rate": 0.12940147028222376,
      "loss": 0.1859,
      "num_input_tokens_seen": 41888912,
      "step": 21755
    },
    {
      "epoch": 3.5498001468308997,
      "grad_norm": 0.0023862062953412533,
      "learning_rate": 0.12934312505980916,
      "loss": 0.1853,
      "num_input_tokens_seen": 41898176,
      "step": 21760
    },
    {
      "epoch": 3.550615874051717,
      "grad_norm": 0.002306748181581497,
      "learning_rate": 0.1292847830229443,
      "loss": 0.2489,
      "num_input_tokens_seen": 41906336,
      "step": 21765
    },
    {
      "epoch": 3.5514316012725344,
      "grad_norm": 0.0012143809581175447,
      "learning_rate": 0.12922644418062626,
      "loss": 0.1863,
      "num_input_tokens_seen": 41916640,
      "step": 21770
    },
    {
      "epoch": 3.552247328493352,
      "grad_norm": 0.0011597422417253256,
      "learning_rate": 0.1291681085418515,
      "loss": 0.1719,
      "num_input_tokens_seen": 41926880,
      "step": 21775
    },
    {
      "epoch": 3.553063055714169,
      "grad_norm": 0.001270358799956739,
      "learning_rate": 0.12910977611561628,
      "loss": 0.2314,
      "num_input_tokens_seen": 41936240,
      "step": 21780
    },
    {
      "epoch": 3.5538787829349863,
      "grad_norm": 0.0026025495026260614,
      "learning_rate": 0.1290514469109161,
      "loss": 0.1828,
      "num_input_tokens_seen": 41947136,
      "step": 21785
    },
    {
      "epoch": 3.554694510155804,
      "grad_norm": 0.002070537069812417,
      "learning_rate": 0.128993120936746,
      "loss": 0.2238,
      "num_input_tokens_seen": 41957440,
      "step": 21790
    },
    {
      "epoch": 3.5555102373766214,
      "grad_norm": 0.0018212427385151386,
      "learning_rate": 0.12893479820210071,
      "loss": 0.2027,
      "num_input_tokens_seen": 41968000,
      "step": 21795
    },
    {
      "epoch": 3.5563259645974385,
      "grad_norm": 0.001976245315745473,
      "learning_rate": 0.1288764787159742,
      "loss": 0.1831,
      "num_input_tokens_seen": 41977888,
      "step": 21800
    },
    {
      "epoch": 3.5563259645974385,
      "eval_loss": 0.20622321963310242,
      "eval_runtime": 67.9399,
      "eval_samples_per_second": 40.109,
      "eval_steps_per_second": 20.062,
      "num_input_tokens_seen": 41977888,
      "step": 21800
    },
    {
      "epoch": 3.557141691818256,
      "grad_norm": 0.0014154976233839989,
      "learning_rate": 0.1288181624873601,
      "loss": 0.203,
      "num_input_tokens_seen": 41988144,
      "step": 21805
    },
    {
      "epoch": 3.5579574190390733,
      "grad_norm": 0.0018804159481078386,
      "learning_rate": 0.12875984952525163,
      "loss": 0.2446,
      "num_input_tokens_seen": 41997824,
      "step": 21810
    },
    {
      "epoch": 3.558773146259891,
      "grad_norm": 0.0030733547173440456,
      "learning_rate": 0.12870153983864122,
      "loss": 0.2399,
      "num_input_tokens_seen": 42007808,
      "step": 21815
    },
    {
      "epoch": 3.559588873480708,
      "grad_norm": 0.002309761242941022,
      "learning_rate": 0.12864323343652104,
      "loss": 0.1813,
      "num_input_tokens_seen": 42016816,
      "step": 21820
    },
    {
      "epoch": 3.5604046007015255,
      "grad_norm": 0.002960807178169489,
      "learning_rate": 0.12858493032788268,
      "loss": 0.1754,
      "num_input_tokens_seen": 42027264,
      "step": 21825
    },
    {
      "epoch": 3.5612203279223427,
      "grad_norm": 0.0022521745413541794,
      "learning_rate": 0.12852663052171714,
      "loss": 0.2455,
      "num_input_tokens_seen": 42037312,
      "step": 21830
    },
    {
      "epoch": 3.5620360551431602,
      "grad_norm": 0.0031844591721892357,
      "learning_rate": 0.12846833402701507,
      "loss": 0.211,
      "num_input_tokens_seen": 42047008,
      "step": 21835
    },
    {
      "epoch": 3.5628517823639774,
      "grad_norm": 0.001446034642867744,
      "learning_rate": 0.12841004085276642,
      "loss": 0.2211,
      "num_input_tokens_seen": 42057056,
      "step": 21840
    },
    {
      "epoch": 3.563667509584795,
      "grad_norm": 0.0016173465410247445,
      "learning_rate": 0.12835175100796076,
      "loss": 0.222,
      "num_input_tokens_seen": 42066512,
      "step": 21845
    },
    {
      "epoch": 3.564483236805612,
      "grad_norm": 0.001446888200007379,
      "learning_rate": 0.12829346450158724,
      "loss": 0.1645,
      "num_input_tokens_seen": 42075280,
      "step": 21850
    },
    {
      "epoch": 3.5652989640264297,
      "grad_norm": 0.002003604080528021,
      "learning_rate": 0.12823518134263423,
      "loss": 0.1539,
      "num_input_tokens_seen": 42085392,
      "step": 21855
    },
    {
      "epoch": 3.566114691247247,
      "grad_norm": 0.004047142807394266,
      "learning_rate": 0.12817690154008973,
      "loss": 0.1991,
      "num_input_tokens_seen": 42095504,
      "step": 21860
    },
    {
      "epoch": 3.5669304184680644,
      "grad_norm": 0.0018017595866695046,
      "learning_rate": 0.12811862510294134,
      "loss": 0.2365,
      "num_input_tokens_seen": 42104912,
      "step": 21865
    },
    {
      "epoch": 3.5677461456888815,
      "grad_norm": 0.003027226310223341,
      "learning_rate": 0.12806035204017585,
      "loss": 0.2183,
      "num_input_tokens_seen": 42114736,
      "step": 21870
    },
    {
      "epoch": 3.568561872909699,
      "grad_norm": 0.00549274729564786,
      "learning_rate": 0.12800208236077987,
      "loss": 0.1796,
      "num_input_tokens_seen": 42123760,
      "step": 21875
    },
    {
      "epoch": 3.569377600130516,
      "grad_norm": 0.002113444497808814,
      "learning_rate": 0.12794381607373917,
      "loss": 0.1882,
      "num_input_tokens_seen": 42134752,
      "step": 21880
    },
    {
      "epoch": 3.5701933273513338,
      "grad_norm": 0.004274734761565924,
      "learning_rate": 0.12788555318803924,
      "loss": 0.2411,
      "num_input_tokens_seen": 42144592,
      "step": 21885
    },
    {
      "epoch": 3.5710090545721513,
      "grad_norm": 0.001714960322715342,
      "learning_rate": 0.1278272937126649,
      "loss": 0.1676,
      "num_input_tokens_seen": 42153296,
      "step": 21890
    },
    {
      "epoch": 3.5718247817929685,
      "grad_norm": 0.0022627527359873056,
      "learning_rate": 0.1277690376566005,
      "loss": 0.2026,
      "num_input_tokens_seen": 42163904,
      "step": 21895
    },
    {
      "epoch": 3.5726405090137856,
      "grad_norm": 0.001804314786568284,
      "learning_rate": 0.12771078502882985,
      "loss": 0.2121,
      "num_input_tokens_seen": 42173488,
      "step": 21900
    },
    {
      "epoch": 3.573456236234603,
      "grad_norm": 0.0033099811989814043,
      "learning_rate": 0.12765253583833633,
      "loss": 0.2307,
      "num_input_tokens_seen": 42182976,
      "step": 21905
    },
    {
      "epoch": 3.5742719634554208,
      "grad_norm": 0.002910110866650939,
      "learning_rate": 0.12759429009410256,
      "loss": 0.2388,
      "num_input_tokens_seen": 42192240,
      "step": 21910
    },
    {
      "epoch": 3.575087690676238,
      "grad_norm": 0.0011226222850382328,
      "learning_rate": 0.12753604780511085,
      "loss": 0.18,
      "num_input_tokens_seen": 42200720,
      "step": 21915
    },
    {
      "epoch": 3.575903417897055,
      "grad_norm": 0.001978253945708275,
      "learning_rate": 0.12747780898034283,
      "loss": 0.2032,
      "num_input_tokens_seen": 42210544,
      "step": 21920
    },
    {
      "epoch": 3.5767191451178726,
      "grad_norm": 0.003611690830439329,
      "learning_rate": 0.12741957362877973,
      "loss": 0.2193,
      "num_input_tokens_seen": 42220128,
      "step": 21925
    },
    {
      "epoch": 3.57753487233869,
      "grad_norm": 0.0021881472785025835,
      "learning_rate": 0.12736134175940214,
      "loss": 0.2126,
      "num_input_tokens_seen": 42229072,
      "step": 21930
    },
    {
      "epoch": 3.5783505995595073,
      "grad_norm": 0.0020262482576072216,
      "learning_rate": 0.12730311338119016,
      "loss": 0.1848,
      "num_input_tokens_seen": 42239088,
      "step": 21935
    },
    {
      "epoch": 3.5791663267803244,
      "grad_norm": 0.0011286152293905616,
      "learning_rate": 0.12724488850312327,
      "loss": 0.1997,
      "num_input_tokens_seen": 42248864,
      "step": 21940
    },
    {
      "epoch": 3.579982054001142,
      "grad_norm": 0.0027327993884682655,
      "learning_rate": 0.1271866671341806,
      "loss": 0.1922,
      "num_input_tokens_seen": 42256960,
      "step": 21945
    },
    {
      "epoch": 3.5807977812219596,
      "grad_norm": 0.0016593114705756307,
      "learning_rate": 0.12712844928334047,
      "loss": 0.2458,
      "num_input_tokens_seen": 42266208,
      "step": 21950
    },
    {
      "epoch": 3.5816135084427767,
      "grad_norm": 0.002060920000076294,
      "learning_rate": 0.12707023495958095,
      "loss": 0.1865,
      "num_input_tokens_seen": 42276016,
      "step": 21955
    },
    {
      "epoch": 3.582429235663594,
      "grad_norm": 0.001509983791038394,
      "learning_rate": 0.12701202417187932,
      "loss": 0.2286,
      "num_input_tokens_seen": 42286128,
      "step": 21960
    },
    {
      "epoch": 3.5832449628844114,
      "grad_norm": 0.001674409955739975,
      "learning_rate": 0.12695381692921243,
      "loss": 0.1938,
      "num_input_tokens_seen": 42295120,
      "step": 21965
    },
    {
      "epoch": 3.584060690105229,
      "grad_norm": 0.0017733045388013124,
      "learning_rate": 0.12689561324055665,
      "loss": 0.2235,
      "num_input_tokens_seen": 42305456,
      "step": 21970
    },
    {
      "epoch": 3.584876417326046,
      "grad_norm": 0.0023686056956648827,
      "learning_rate": 0.12683741311488758,
      "loss": 0.2037,
      "num_input_tokens_seen": 42314688,
      "step": 21975
    },
    {
      "epoch": 3.5856921445468637,
      "grad_norm": 0.0022202315740287304,
      "learning_rate": 0.1267792165611805,
      "loss": 0.1583,
      "num_input_tokens_seen": 42323936,
      "step": 21980
    },
    {
      "epoch": 3.586507871767681,
      "grad_norm": 0.001944542396813631,
      "learning_rate": 0.1267210235884101,
      "loss": 0.2133,
      "num_input_tokens_seen": 42332496,
      "step": 21985
    },
    {
      "epoch": 3.5873235989884984,
      "grad_norm": 0.002953647868707776,
      "learning_rate": 0.12666283420555033,
      "loss": 0.2364,
      "num_input_tokens_seen": 42342432,
      "step": 21990
    },
    {
      "epoch": 3.5881393262093155,
      "grad_norm": 0.0029624905437231064,
      "learning_rate": 0.12660464842157487,
      "loss": 0.2408,
      "num_input_tokens_seen": 42351584,
      "step": 21995
    },
    {
      "epoch": 3.588955053430133,
      "grad_norm": 0.003025593003258109,
      "learning_rate": 0.1265464662454566,
      "loss": 0.2423,
      "num_input_tokens_seen": 42361392,
      "step": 22000
    },
    {
      "epoch": 3.588955053430133,
      "eval_loss": 0.20464308559894562,
      "eval_runtime": 68.0102,
      "eval_samples_per_second": 40.067,
      "eval_steps_per_second": 20.041,
      "num_input_tokens_seen": 42361392,
      "step": 22000
    },
    {
      "epoch": 3.5897707806509502,
      "grad_norm": 0.0021402970887720585,
      "learning_rate": 0.12648828768616793,
      "loss": 0.2421,
      "num_input_tokens_seen": 42371904,
      "step": 22005
    },
    {
      "epoch": 3.590586507871768,
      "grad_norm": 0.002380963647738099,
      "learning_rate": 0.12643011275268085,
      "loss": 0.1837,
      "num_input_tokens_seen": 42380688,
      "step": 22010
    },
    {
      "epoch": 3.591402235092585,
      "grad_norm": 0.0021503036841750145,
      "learning_rate": 0.1263719414539665,
      "loss": 0.2274,
      "num_input_tokens_seen": 42390864,
      "step": 22015
    },
    {
      "epoch": 3.5922179623134025,
      "grad_norm": 0.0012828329345211387,
      "learning_rate": 0.1263137737989957,
      "loss": 0.2101,
      "num_input_tokens_seen": 42399696,
      "step": 22020
    },
    {
      "epoch": 3.5930336895342196,
      "grad_norm": 0.0015616017626598477,
      "learning_rate": 0.1262556097967387,
      "loss": 0.2604,
      "num_input_tokens_seen": 42409360,
      "step": 22025
    },
    {
      "epoch": 3.593849416755037,
      "grad_norm": 0.002690204419195652,
      "learning_rate": 0.126197449456165,
      "loss": 0.1791,
      "num_input_tokens_seen": 42419168,
      "step": 22030
    },
    {
      "epoch": 3.5946651439758543,
      "grad_norm": 0.0021047997288405895,
      "learning_rate": 0.12613929278624378,
      "loss": 0.2065,
      "num_input_tokens_seen": 42429456,
      "step": 22035
    },
    {
      "epoch": 3.595480871196672,
      "grad_norm": 0.0022914065048098564,
      "learning_rate": 0.12608113979594343,
      "loss": 0.1968,
      "num_input_tokens_seen": 42440112,
      "step": 22040
    },
    {
      "epoch": 3.596296598417489,
      "grad_norm": 0.0019908526446670294,
      "learning_rate": 0.1260229904942319,
      "loss": 0.2155,
      "num_input_tokens_seen": 42450576,
      "step": 22045
    },
    {
      "epoch": 3.5971123256383066,
      "grad_norm": 0.0017862542299553752,
      "learning_rate": 0.12596484489007662,
      "loss": 0.2418,
      "num_input_tokens_seen": 42459488,
      "step": 22050
    },
    {
      "epoch": 3.5979280528591238,
      "grad_norm": 0.003216444980353117,
      "learning_rate": 0.1259067029924442,
      "loss": 0.2169,
      "num_input_tokens_seen": 42467280,
      "step": 22055
    },
    {
      "epoch": 3.5987437800799413,
      "grad_norm": 0.004069799091666937,
      "learning_rate": 0.12584856481030096,
      "loss": 0.2696,
      "num_input_tokens_seen": 42477104,
      "step": 22060
    },
    {
      "epoch": 3.5995595073007585,
      "grad_norm": 0.001592431915923953,
      "learning_rate": 0.12579043035261261,
      "loss": 0.1677,
      "num_input_tokens_seen": 42486960,
      "step": 22065
    },
    {
      "epoch": 3.600375234521576,
      "grad_norm": 0.0022479414474219084,
      "learning_rate": 0.1257322996283441,
      "loss": 0.1609,
      "num_input_tokens_seen": 42495104,
      "step": 22070
    },
    {
      "epoch": 3.601190961742393,
      "grad_norm": 0.0020900280214846134,
      "learning_rate": 0.12567417264645994,
      "loss": 0.2378,
      "num_input_tokens_seen": 42503664,
      "step": 22075
    },
    {
      "epoch": 3.6020066889632107,
      "grad_norm": 0.003166846465319395,
      "learning_rate": 0.12561604941592408,
      "loss": 0.2007,
      "num_input_tokens_seen": 42513584,
      "step": 22080
    },
    {
      "epoch": 3.6028224161840283,
      "grad_norm": 0.002819610061123967,
      "learning_rate": 0.12555792994569978,
      "loss": 0.1742,
      "num_input_tokens_seen": 42524240,
      "step": 22085
    },
    {
      "epoch": 3.6036381434048455,
      "grad_norm": 0.0014522504061460495,
      "learning_rate": 0.1254998142447499,
      "loss": 0.2265,
      "num_input_tokens_seen": 42533680,
      "step": 22090
    },
    {
      "epoch": 3.6044538706256626,
      "grad_norm": 0.001503033796325326,
      "learning_rate": 0.1254417023220365,
      "loss": 0.2002,
      "num_input_tokens_seen": 42543904,
      "step": 22095
    },
    {
      "epoch": 3.60526959784648,
      "grad_norm": 0.0017492480110377073,
      "learning_rate": 0.12538359418652126,
      "loss": 0.2408,
      "num_input_tokens_seen": 42553344,
      "step": 22100
    },
    {
      "epoch": 3.6060853250672977,
      "grad_norm": 0.0030785980634391308,
      "learning_rate": 0.12532548984716513,
      "loss": 0.2138,
      "num_input_tokens_seen": 42562256,
      "step": 22105
    },
    {
      "epoch": 3.606901052288115,
      "grad_norm": 0.0022696356754750013,
      "learning_rate": 0.12526738931292855,
      "loss": 0.2401,
      "num_input_tokens_seen": 42571504,
      "step": 22110
    },
    {
      "epoch": 3.607716779508932,
      "grad_norm": 0.0033308102283626795,
      "learning_rate": 0.1252092925927714,
      "loss": 0.2368,
      "num_input_tokens_seen": 42581712,
      "step": 22115
    },
    {
      "epoch": 3.6085325067297496,
      "grad_norm": 0.0017383340746164322,
      "learning_rate": 0.12515119969565278,
      "loss": 0.1968,
      "num_input_tokens_seen": 42591440,
      "step": 22120
    },
    {
      "epoch": 3.609348233950567,
      "grad_norm": 0.0030722611118108034,
      "learning_rate": 0.12509311063053144,
      "loss": 0.2017,
      "num_input_tokens_seen": 42600720,
      "step": 22125
    },
    {
      "epoch": 3.6101639611713843,
      "grad_norm": 0.0023290442768484354,
      "learning_rate": 0.1250350254063655,
      "loss": 0.1926,
      "num_input_tokens_seen": 42611248,
      "step": 22130
    },
    {
      "epoch": 3.6109796883922014,
      "grad_norm": 0.0016598365036770701,
      "learning_rate": 0.1249769440321123,
      "loss": 0.2145,
      "num_input_tokens_seen": 42620304,
      "step": 22135
    },
    {
      "epoch": 3.611795415613019,
      "grad_norm": 0.002829186385497451,
      "learning_rate": 0.12491886651672884,
      "loss": 0.2221,
      "num_input_tokens_seen": 42629936,
      "step": 22140
    },
    {
      "epoch": 3.6126111428338366,
      "grad_norm": 0.003285913495346904,
      "learning_rate": 0.12486079286917139,
      "loss": 0.2453,
      "num_input_tokens_seen": 42639968,
      "step": 22145
    },
    {
      "epoch": 3.6134268700546537,
      "grad_norm": 0.0013465272495523095,
      "learning_rate": 0.12480272309839553,
      "loss": 0.1788,
      "num_input_tokens_seen": 42650176,
      "step": 22150
    },
    {
      "epoch": 3.614242597275471,
      "grad_norm": 0.002078782767057419,
      "learning_rate": 0.12474465721335648,
      "loss": 0.204,
      "num_input_tokens_seen": 42660368,
      "step": 22155
    },
    {
      "epoch": 3.6150583244962884,
      "grad_norm": 0.001750097842887044,
      "learning_rate": 0.12468659522300861,
      "loss": 0.2073,
      "num_input_tokens_seen": 42670736,
      "step": 22160
    },
    {
      "epoch": 3.615874051717106,
      "grad_norm": 0.0015771714970469475,
      "learning_rate": 0.12462853713630584,
      "loss": 0.2534,
      "num_input_tokens_seen": 42679952,
      "step": 22165
    },
    {
      "epoch": 3.616689778937923,
      "grad_norm": 0.0016930144047364593,
      "learning_rate": 0.12457048296220156,
      "loss": 0.1957,
      "num_input_tokens_seen": 42689216,
      "step": 22170
    },
    {
      "epoch": 3.6175055061587407,
      "grad_norm": 0.0019640368409454823,
      "learning_rate": 0.12451243270964832,
      "loss": 0.2701,
      "num_input_tokens_seen": 42699392,
      "step": 22175
    },
    {
      "epoch": 3.618321233379558,
      "grad_norm": 0.0013568198774009943,
      "learning_rate": 0.12445438638759827,
      "loss": 0.2209,
      "num_input_tokens_seen": 42709920,
      "step": 22180
    },
    {
      "epoch": 3.6191369606003754,
      "grad_norm": 0.001862777047790587,
      "learning_rate": 0.1243963440050029,
      "loss": 0.2176,
      "num_input_tokens_seen": 42719616,
      "step": 22185
    },
    {
      "epoch": 3.6199526878211925,
      "grad_norm": 0.0024721783120185137,
      "learning_rate": 0.12433830557081298,
      "loss": 0.2065,
      "num_input_tokens_seen": 42728624,
      "step": 22190
    },
    {
      "epoch": 3.62076841504201,
      "grad_norm": 0.001635366934351623,
      "learning_rate": 0.12428027109397889,
      "loss": 0.1879,
      "num_input_tokens_seen": 42737840,
      "step": 22195
    },
    {
      "epoch": 3.621584142262827,
      "grad_norm": 0.0014725008513778448,
      "learning_rate": 0.12422224058345015,
      "loss": 0.1997,
      "num_input_tokens_seen": 42746416,
      "step": 22200
    },
    {
      "epoch": 3.621584142262827,
      "eval_loss": 0.2065480798482895,
      "eval_runtime": 68.0306,
      "eval_samples_per_second": 40.056,
      "eval_steps_per_second": 20.035,
      "num_input_tokens_seen": 42746416,
      "step": 22200
    },
    {
      "epoch": 3.622399869483645,
      "grad_norm": 0.0017906873254105449,
      "learning_rate": 0.12416421404817583,
      "loss": 0.2043,
      "num_input_tokens_seen": 42756288,
      "step": 22205
    },
    {
      "epoch": 3.623215596704462,
      "grad_norm": 0.002456771209836006,
      "learning_rate": 0.12410619149710447,
      "loss": 0.2398,
      "num_input_tokens_seen": 42766688,
      "step": 22210
    },
    {
      "epoch": 3.6240313239252795,
      "grad_norm": 0.0015449408674612641,
      "learning_rate": 0.12404817293918374,
      "loss": 0.1946,
      "num_input_tokens_seen": 42775520,
      "step": 22215
    },
    {
      "epoch": 3.6248470511460966,
      "grad_norm": 0.0016432419652119279,
      "learning_rate": 0.12399015838336086,
      "loss": 0.161,
      "num_input_tokens_seen": 42785360,
      "step": 22220
    },
    {
      "epoch": 3.625662778366914,
      "grad_norm": 0.0016242179553955793,
      "learning_rate": 0.12393214783858246,
      "loss": 0.1342,
      "num_input_tokens_seen": 42794432,
      "step": 22225
    },
    {
      "epoch": 3.6264785055877313,
      "grad_norm": 0.0010997327044606209,
      "learning_rate": 0.1238741413137944,
      "loss": 0.2162,
      "num_input_tokens_seen": 42804272,
      "step": 22230
    },
    {
      "epoch": 3.627294232808549,
      "grad_norm": 0.0019047742243856192,
      "learning_rate": 0.12381613881794212,
      "loss": 0.2701,
      "num_input_tokens_seen": 42814544,
      "step": 22235
    },
    {
      "epoch": 3.628109960029366,
      "grad_norm": 0.0016189503949135542,
      "learning_rate": 0.12375814035997022,
      "loss": 0.2016,
      "num_input_tokens_seen": 42824496,
      "step": 22240
    },
    {
      "epoch": 3.6289256872501836,
      "grad_norm": 0.0018792490009218454,
      "learning_rate": 0.12370014594882285,
      "loss": 0.2071,
      "num_input_tokens_seen": 42834032,
      "step": 22245
    },
    {
      "epoch": 3.6297414144710007,
      "grad_norm": 0.004691870417445898,
      "learning_rate": 0.12364215559344356,
      "loss": 0.2002,
      "num_input_tokens_seen": 42841344,
      "step": 22250
    },
    {
      "epoch": 3.6305571416918183,
      "grad_norm": 0.003623911179602146,
      "learning_rate": 0.12358416930277506,
      "loss": 0.1981,
      "num_input_tokens_seen": 42849776,
      "step": 22255
    },
    {
      "epoch": 3.631372868912636,
      "grad_norm": 0.0012929994845762849,
      "learning_rate": 0.1235261870857596,
      "loss": 0.1688,
      "num_input_tokens_seen": 42859280,
      "step": 22260
    },
    {
      "epoch": 3.632188596133453,
      "grad_norm": 0.0018207760294899344,
      "learning_rate": 0.12346820895133884,
      "loss": 0.1883,
      "num_input_tokens_seen": 42868800,
      "step": 22265
    },
    {
      "epoch": 3.63300432335427,
      "grad_norm": 0.0019018712919205427,
      "learning_rate": 0.12341023490845361,
      "loss": 0.2105,
      "num_input_tokens_seen": 42878848,
      "step": 22270
    },
    {
      "epoch": 3.6338200505750877,
      "grad_norm": 0.0036074745003134012,
      "learning_rate": 0.12335226496604437,
      "loss": 0.2015,
      "num_input_tokens_seen": 42887648,
      "step": 22275
    },
    {
      "epoch": 3.6346357777959053,
      "grad_norm": 0.003114050719887018,
      "learning_rate": 0.12329429913305069,
      "loss": 0.2171,
      "num_input_tokens_seen": 42897392,
      "step": 22280
    },
    {
      "epoch": 3.6354515050167224,
      "grad_norm": 0.0020107293967157602,
      "learning_rate": 0.12323633741841171,
      "loss": 0.2004,
      "num_input_tokens_seen": 42906384,
      "step": 22285
    },
    {
      "epoch": 3.6362672322375396,
      "grad_norm": 0.004176556132733822,
      "learning_rate": 0.12317837983106583,
      "loss": 0.1917,
      "num_input_tokens_seen": 42915072,
      "step": 22290
    },
    {
      "epoch": 3.637082959458357,
      "grad_norm": 0.004162255208939314,
      "learning_rate": 0.12312042637995087,
      "loss": 0.2255,
      "num_input_tokens_seen": 42924032,
      "step": 22295
    },
    {
      "epoch": 3.6378986866791747,
      "grad_norm": 0.0024906834587454796,
      "learning_rate": 0.12306247707400389,
      "loss": 0.2369,
      "num_input_tokens_seen": 42933168,
      "step": 22300
    },
    {
      "epoch": 3.638714413899992,
      "grad_norm": 0.0042428490705788136,
      "learning_rate": 0.12300453192216154,
      "loss": 0.1999,
      "num_input_tokens_seen": 42942944,
      "step": 22305
    },
    {
      "epoch": 3.639530141120809,
      "grad_norm": 0.001932006562128663,
      "learning_rate": 0.12294659093335956,
      "loss": 0.2362,
      "num_input_tokens_seen": 42952464,
      "step": 22310
    },
    {
      "epoch": 3.6403458683416265,
      "grad_norm": 0.004223518539220095,
      "learning_rate": 0.12288865411653327,
      "loss": 0.2403,
      "num_input_tokens_seen": 42962880,
      "step": 22315
    },
    {
      "epoch": 3.641161595562444,
      "grad_norm": 0.002722426550462842,
      "learning_rate": 0.12283072148061717,
      "loss": 0.2269,
      "num_input_tokens_seen": 42972976,
      "step": 22320
    },
    {
      "epoch": 3.6419773227832613,
      "grad_norm": 0.0010852169943973422,
      "learning_rate": 0.12277279303454529,
      "loss": 0.1999,
      "num_input_tokens_seen": 42983024,
      "step": 22325
    },
    {
      "epoch": 3.6427930500040784,
      "grad_norm": 0.0027851250488311052,
      "learning_rate": 0.12271486878725091,
      "loss": 0.2442,
      "num_input_tokens_seen": 42993472,
      "step": 22330
    },
    {
      "epoch": 3.643608777224896,
      "grad_norm": 0.0033840073738247156,
      "learning_rate": 0.12265694874766658,
      "loss": 0.2237,
      "num_input_tokens_seen": 43003344,
      "step": 22335
    },
    {
      "epoch": 3.6444245044457135,
      "grad_norm": 0.0029955576173961163,
      "learning_rate": 0.12259903292472435,
      "loss": 0.2517,
      "num_input_tokens_seen": 43012080,
      "step": 22340
    },
    {
      "epoch": 3.6452402316665307,
      "grad_norm": 0.004131867550313473,
      "learning_rate": 0.12254112132735567,
      "loss": 0.2563,
      "num_input_tokens_seen": 43020128,
      "step": 22345
    },
    {
      "epoch": 3.6460559588873482,
      "grad_norm": 0.0018883533775806427,
      "learning_rate": 0.12248321396449108,
      "loss": 0.2146,
      "num_input_tokens_seen": 43029712,
      "step": 22350
    },
    {
      "epoch": 3.6468716861081654,
      "grad_norm": 0.0023023150861263275,
      "learning_rate": 0.12242531084506075,
      "loss": 0.2424,
      "num_input_tokens_seen": 43039024,
      "step": 22355
    },
    {
      "epoch": 3.647687413328983,
      "grad_norm": 0.002549126511439681,
      "learning_rate": 0.122367411977994,
      "loss": 0.1939,
      "num_input_tokens_seen": 43048816,
      "step": 22360
    },
    {
      "epoch": 3.6485031405498,
      "grad_norm": 0.004212372470647097,
      "learning_rate": 0.12230951737221954,
      "loss": 0.2569,
      "num_input_tokens_seen": 43059136,
      "step": 22365
    },
    {
      "epoch": 3.6493188677706176,
      "grad_norm": 0.0030689698178321123,
      "learning_rate": 0.12225162703666555,
      "loss": 0.201,
      "num_input_tokens_seen": 43067904,
      "step": 22370
    },
    {
      "epoch": 3.6501345949914348,
      "grad_norm": 0.0018879581475630403,
      "learning_rate": 0.1221937409802593,
      "loss": 0.2093,
      "num_input_tokens_seen": 43077776,
      "step": 22375
    },
    {
      "epoch": 3.6509503222122524,
      "grad_norm": 0.0018847516039386392,
      "learning_rate": 0.12213585921192768,
      "loss": 0.1956,
      "num_input_tokens_seen": 43087824,
      "step": 22380
    },
    {
      "epoch": 3.6517660494330695,
      "grad_norm": 0.0029871517326682806,
      "learning_rate": 0.1220779817405967,
      "loss": 0.2207,
      "num_input_tokens_seen": 43097760,
      "step": 22385
    },
    {
      "epoch": 3.652581776653887,
      "grad_norm": 0.004608910530805588,
      "learning_rate": 0.12202010857519181,
      "loss": 0.1975,
      "num_input_tokens_seen": 43107264,
      "step": 22390
    },
    {
      "epoch": 3.653397503874704,
      "grad_norm": 0.00244755856692791,
      "learning_rate": 0.12196223972463785,
      "loss": 0.1833,
      "num_input_tokens_seen": 43115984,
      "step": 22395
    },
    {
      "epoch": 3.6542132310955218,
      "grad_norm": 0.002986618084833026,
      "learning_rate": 0.12190437519785885,
      "loss": 0.1787,
      "num_input_tokens_seen": 43126400,
      "step": 22400
    },
    {
      "epoch": 3.6542132310955218,
      "eval_loss": 0.2105056792497635,
      "eval_runtime": 68.1115,
      "eval_samples_per_second": 40.008,
      "eval_steps_per_second": 20.011,
      "num_input_tokens_seen": 43126400,
      "step": 22400
    },
    {
      "epoch": 3.655028958316339,
      "grad_norm": 0.0041323513723909855,
      "learning_rate": 0.12184651500377823,
      "loss": 0.2264,
      "num_input_tokens_seen": 43136672,
      "step": 22405
    },
    {
      "epoch": 3.6558446855371565,
      "grad_norm": 0.0028404248878359795,
      "learning_rate": 0.12178865915131885,
      "loss": 0.1978,
      "num_input_tokens_seen": 43145216,
      "step": 22410
    },
    {
      "epoch": 3.6566604127579736,
      "grad_norm": 0.0032824755180627108,
      "learning_rate": 0.1217308076494027,
      "loss": 0.2244,
      "num_input_tokens_seen": 43154480,
      "step": 22415
    },
    {
      "epoch": 3.657476139978791,
      "grad_norm": 0.00539375888183713,
      "learning_rate": 0.12167296050695134,
      "loss": 0.215,
      "num_input_tokens_seen": 43164496,
      "step": 22420
    },
    {
      "epoch": 3.6582918671996083,
      "grad_norm": 0.0010028217220678926,
      "learning_rate": 0.12161511773288536,
      "loss": 0.2347,
      "num_input_tokens_seen": 43174192,
      "step": 22425
    },
    {
      "epoch": 3.659107594420426,
      "grad_norm": 0.003911934327334166,
      "learning_rate": 0.121557279336125,
      "loss": 0.2483,
      "num_input_tokens_seen": 43184688,
      "step": 22430
    },
    {
      "epoch": 3.659923321641243,
      "grad_norm": 0.0028183271642774343,
      "learning_rate": 0.12149944532558957,
      "loss": 0.2583,
      "num_input_tokens_seen": 43194448,
      "step": 22435
    },
    {
      "epoch": 3.6607390488620606,
      "grad_norm": 0.0016880633775144815,
      "learning_rate": 0.12144161571019785,
      "loss": 0.1731,
      "num_input_tokens_seen": 43205184,
      "step": 22440
    },
    {
      "epoch": 3.6615547760828777,
      "grad_norm": 0.0020049125887453556,
      "learning_rate": 0.12138379049886781,
      "loss": 0.1939,
      "num_input_tokens_seen": 43214528,
      "step": 22445
    },
    {
      "epoch": 3.6623705033036953,
      "grad_norm": 0.001611386309377849,
      "learning_rate": 0.12132596970051697,
      "loss": 0.198,
      "num_input_tokens_seen": 43223808,
      "step": 22450
    },
    {
      "epoch": 3.663186230524513,
      "grad_norm": 0.0030861208215355873,
      "learning_rate": 0.12126815332406189,
      "loss": 0.2012,
      "num_input_tokens_seen": 43233712,
      "step": 22455
    },
    {
      "epoch": 3.66400195774533,
      "grad_norm": 0.0022075059823691845,
      "learning_rate": 0.12121034137841868,
      "loss": 0.1728,
      "num_input_tokens_seen": 43243008,
      "step": 22460
    },
    {
      "epoch": 3.664817684966147,
      "grad_norm": 0.0028370690997689962,
      "learning_rate": 0.12115253387250258,
      "loss": 0.237,
      "num_input_tokens_seen": 43253344,
      "step": 22465
    },
    {
      "epoch": 3.6656334121869647,
      "grad_norm": 0.003137683728709817,
      "learning_rate": 0.12109473081522831,
      "loss": 0.1718,
      "num_input_tokens_seen": 43261904,
      "step": 22470
    },
    {
      "epoch": 3.6664491394077823,
      "grad_norm": 0.0022642435505986214,
      "learning_rate": 0.12103693221550982,
      "loss": 0.2394,
      "num_input_tokens_seen": 43271040,
      "step": 22475
    },
    {
      "epoch": 3.6672648666285994,
      "grad_norm": 0.002576550468802452,
      "learning_rate": 0.12097913808226027,
      "loss": 0.1842,
      "num_input_tokens_seen": 43281248,
      "step": 22480
    },
    {
      "epoch": 3.6680805938494165,
      "grad_norm": 0.0029811791609972715,
      "learning_rate": 0.12092134842439234,
      "loss": 0.2133,
      "num_input_tokens_seen": 43289312,
      "step": 22485
    },
    {
      "epoch": 3.668896321070234,
      "grad_norm": 0.002421395853161812,
      "learning_rate": 0.12086356325081798,
      "loss": 0.184,
      "num_input_tokens_seen": 43300016,
      "step": 22490
    },
    {
      "epoch": 3.6697120482910517,
      "grad_norm": 0.0010737570701166987,
      "learning_rate": 0.12080578257044824,
      "loss": 0.1736,
      "num_input_tokens_seen": 43310560,
      "step": 22495
    },
    {
      "epoch": 3.670527775511869,
      "grad_norm": 0.0015484158648177981,
      "learning_rate": 0.12074800639219378,
      "loss": 0.2131,
      "num_input_tokens_seen": 43320272,
      "step": 22500
    },
    {
      "epoch": 3.671343502732686,
      "grad_norm": 0.004547641146928072,
      "learning_rate": 0.12069023472496428,
      "loss": 0.232,
      "num_input_tokens_seen": 43329264,
      "step": 22505
    },
    {
      "epoch": 3.6721592299535035,
      "grad_norm": 0.00301565183326602,
      "learning_rate": 0.12063246757766893,
      "loss": 0.2563,
      "num_input_tokens_seen": 43339136,
      "step": 22510
    },
    {
      "epoch": 3.672974957174321,
      "grad_norm": 0.0018137884326279163,
      "learning_rate": 0.12057470495921618,
      "loss": 0.2378,
      "num_input_tokens_seen": 43348160,
      "step": 22515
    },
    {
      "epoch": 3.6737906843951382,
      "grad_norm": 0.0010737121338024735,
      "learning_rate": 0.12051694687851364,
      "loss": 0.2159,
      "num_input_tokens_seen": 43356848,
      "step": 22520
    },
    {
      "epoch": 3.6746064116159554,
      "grad_norm": 0.0021582848858088255,
      "learning_rate": 0.12045919334446839,
      "loss": 0.1731,
      "num_input_tokens_seen": 43366944,
      "step": 22525
    },
    {
      "epoch": 3.675422138836773,
      "grad_norm": 0.002060161903500557,
      "learning_rate": 0.12040144436598683,
      "loss": 0.2251,
      "num_input_tokens_seen": 43377024,
      "step": 22530
    },
    {
      "epoch": 3.6762378660575905,
      "grad_norm": 0.0030297094490379095,
      "learning_rate": 0.12034369995197444,
      "loss": 0.2019,
      "num_input_tokens_seen": 43386976,
      "step": 22535
    },
    {
      "epoch": 3.6770535932784076,
      "grad_norm": 0.0025611252058297396,
      "learning_rate": 0.12028596011133627,
      "loss": 0.2231,
      "num_input_tokens_seen": 43396496,
      "step": 22540
    },
    {
      "epoch": 3.677869320499225,
      "grad_norm": 0.0013605961576104164,
      "learning_rate": 0.12022822485297643,
      "loss": 0.1772,
      "num_input_tokens_seen": 43406592,
      "step": 22545
    },
    {
      "epoch": 3.6786850477200423,
      "grad_norm": 0.0026898598298430443,
      "learning_rate": 0.12017049418579843,
      "loss": 0.2713,
      "num_input_tokens_seen": 43415984,
      "step": 22550
    },
    {
      "epoch": 3.67950077494086,
      "grad_norm": 0.002310642972588539,
      "learning_rate": 0.12011276811870514,
      "loss": 0.2126,
      "num_input_tokens_seen": 43423888,
      "step": 22555
    },
    {
      "epoch": 3.680316502161677,
      "grad_norm": 0.001598566072061658,
      "learning_rate": 0.12005504666059852,
      "loss": 0.168,
      "num_input_tokens_seen": 43434592,
      "step": 22560
    },
    {
      "epoch": 3.6811322293824946,
      "grad_norm": 0.0021115003619343042,
      "learning_rate": 0.11999732982038003,
      "loss": 0.1851,
      "num_input_tokens_seen": 43443968,
      "step": 22565
    },
    {
      "epoch": 3.6819479566033118,
      "grad_norm": 0.003055148059502244,
      "learning_rate": 0.11993961760695038,
      "loss": 0.1881,
      "num_input_tokens_seen": 43453488,
      "step": 22570
    },
    {
      "epoch": 3.6827636838241293,
      "grad_norm": 0.0028201413806527853,
      "learning_rate": 0.11988191002920942,
      "loss": 0.2227,
      "num_input_tokens_seen": 43462304,
      "step": 22575
    },
    {
      "epoch": 3.6835794110449465,
      "grad_norm": 0.0020261569879949093,
      "learning_rate": 0.11982420709605641,
      "loss": 0.244,
      "num_input_tokens_seen": 43473376,
      "step": 22580
    },
    {
      "epoch": 3.684395138265764,
      "grad_norm": 0.0022943990770727396,
      "learning_rate": 0.11976650881638991,
      "loss": 0.2107,
      "num_input_tokens_seen": 43483776,
      "step": 22585
    },
    {
      "epoch": 3.685210865486581,
      "grad_norm": 0.0015912852250039577,
      "learning_rate": 0.11970881519910764,
      "loss": 0.213,
      "num_input_tokens_seen": 43494208,
      "step": 22590
    },
    {
      "epoch": 3.6860265927073987,
      "grad_norm": 0.0019346155459061265,
      "learning_rate": 0.1196511262531068,
      "loss": 0.1943,
      "num_input_tokens_seen": 43504736,
      "step": 22595
    },
    {
      "epoch": 3.686842319928216,
      "grad_norm": 0.002819799119606614,
      "learning_rate": 0.11959344198728361,
      "loss": 0.245,
      "num_input_tokens_seen": 43513248,
      "step": 22600
    },
    {
      "epoch": 3.686842319928216,
      "eval_loss": 0.20583947002887726,
      "eval_runtime": 67.9195,
      "eval_samples_per_second": 40.121,
      "eval_steps_per_second": 20.068,
      "num_input_tokens_seen": 43513248,
      "step": 22600
    },
    {
      "epoch": 3.6876580471490334,
      "grad_norm": 0.00214870716445148,
      "learning_rate": 0.11953576241053378,
      "loss": 0.1958,
      "num_input_tokens_seen": 43521248,
      "step": 22605
    },
    {
      "epoch": 3.6884737743698506,
      "grad_norm": 0.002381507307291031,
      "learning_rate": 0.11947808753175228,
      "loss": 0.2392,
      "num_input_tokens_seen": 43532192,
      "step": 22610
    },
    {
      "epoch": 3.689289501590668,
      "grad_norm": 0.0020526284351944923,
      "learning_rate": 0.1194204173598332,
      "loss": 0.2088,
      "num_input_tokens_seen": 43541424,
      "step": 22615
    },
    {
      "epoch": 3.6901052288114853,
      "grad_norm": 0.0021180484909564257,
      "learning_rate": 0.11936275190367007,
      "loss": 0.2249,
      "num_input_tokens_seen": 43552416,
      "step": 22620
    },
    {
      "epoch": 3.690920956032303,
      "grad_norm": 0.0015344124985858798,
      "learning_rate": 0.11930509117215563,
      "loss": 0.1843,
      "num_input_tokens_seen": 43562176,
      "step": 22625
    },
    {
      "epoch": 3.6917366832531204,
      "grad_norm": 0.002506963210180402,
      "learning_rate": 0.11924743517418179,
      "loss": 0.2596,
      "num_input_tokens_seen": 43572944,
      "step": 22630
    },
    {
      "epoch": 3.6925524104739376,
      "grad_norm": 0.002236006548628211,
      "learning_rate": 0.11918978391864,
      "loss": 0.2252,
      "num_input_tokens_seen": 43581360,
      "step": 22635
    },
    {
      "epoch": 3.6933681376947547,
      "grad_norm": 0.0022351485677063465,
      "learning_rate": 0.11913213741442065,
      "loss": 0.1966,
      "num_input_tokens_seen": 43592064,
      "step": 22640
    },
    {
      "epoch": 3.6941838649155723,
      "grad_norm": 0.0019778397399932146,
      "learning_rate": 0.11907449567041364,
      "loss": 0.2089,
      "num_input_tokens_seen": 43602064,
      "step": 22645
    },
    {
      "epoch": 3.69499959213639,
      "grad_norm": 0.002340830396860838,
      "learning_rate": 0.11901685869550803,
      "loss": 0.2133,
      "num_input_tokens_seen": 43610592,
      "step": 22650
    },
    {
      "epoch": 3.695815319357207,
      "grad_norm": 0.0033031480852514505,
      "learning_rate": 0.1189592264985922,
      "loss": 0.1817,
      "num_input_tokens_seen": 43619968,
      "step": 22655
    },
    {
      "epoch": 3.696631046578024,
      "grad_norm": 0.002258048625662923,
      "learning_rate": 0.11890159908855373,
      "loss": 0.2714,
      "num_input_tokens_seen": 43629616,
      "step": 22660
    },
    {
      "epoch": 3.6974467737988417,
      "grad_norm": 0.0015909827779978514,
      "learning_rate": 0.11884397647427941,
      "loss": 0.1924,
      "num_input_tokens_seen": 43639600,
      "step": 22665
    },
    {
      "epoch": 3.6982625010196593,
      "grad_norm": 0.0019708070904016495,
      "learning_rate": 0.11878635866465546,
      "loss": 0.2127,
      "num_input_tokens_seen": 43649568,
      "step": 22670
    },
    {
      "epoch": 3.6990782282404764,
      "grad_norm": 0.0025619345251470804,
      "learning_rate": 0.11872874566856734,
      "loss": 0.1975,
      "num_input_tokens_seen": 43658368,
      "step": 22675
    },
    {
      "epoch": 3.6998939554612935,
      "grad_norm": 0.0030529266223311424,
      "learning_rate": 0.11867113749489955,
      "loss": 0.2277,
      "num_input_tokens_seen": 43667888,
      "step": 22680
    },
    {
      "epoch": 3.700709682682111,
      "grad_norm": 0.0026918707881122828,
      "learning_rate": 0.11861353415253607,
      "loss": 0.2181,
      "num_input_tokens_seen": 43677776,
      "step": 22685
    },
    {
      "epoch": 3.7015254099029287,
      "grad_norm": 0.0021123315673321486,
      "learning_rate": 0.11855593565036011,
      "loss": 0.1896,
      "num_input_tokens_seen": 43686608,
      "step": 22690
    },
    {
      "epoch": 3.702341137123746,
      "grad_norm": 0.0017825588583946228,
      "learning_rate": 0.11849834199725394,
      "loss": 0.1953,
      "num_input_tokens_seen": 43696880,
      "step": 22695
    },
    {
      "epoch": 3.703156864344563,
      "grad_norm": 0.0015061493031680584,
      "learning_rate": 0.1184407532020994,
      "loss": 0.2413,
      "num_input_tokens_seen": 43707040,
      "step": 22700
    },
    {
      "epoch": 3.7039725915653805,
      "grad_norm": 0.0027471818029880524,
      "learning_rate": 0.11838316927377723,
      "loss": 0.2902,
      "num_input_tokens_seen": 43717568,
      "step": 22705
    },
    {
      "epoch": 3.704788318786198,
      "grad_norm": 0.00202369736507535,
      "learning_rate": 0.11832559022116766,
      "loss": 0.2175,
      "num_input_tokens_seen": 43726800,
      "step": 22710
    },
    {
      "epoch": 3.705604046007015,
      "grad_norm": 0.0015844492008909583,
      "learning_rate": 0.11826801605315022,
      "loss": 0.2089,
      "num_input_tokens_seen": 43736496,
      "step": 22715
    },
    {
      "epoch": 3.7064197732278323,
      "grad_norm": 0.0009686527773737907,
      "learning_rate": 0.1182104467786034,
      "loss": 0.2003,
      "num_input_tokens_seen": 43746816,
      "step": 22720
    },
    {
      "epoch": 3.70723550044865,
      "grad_norm": 0.0019678049720823765,
      "learning_rate": 0.1181528824064052,
      "loss": 0.2558,
      "num_input_tokens_seen": 43755696,
      "step": 22725
    },
    {
      "epoch": 3.7080512276694675,
      "grad_norm": 0.0010851540137082338,
      "learning_rate": 0.11809532294543279,
      "loss": 0.2433,
      "num_input_tokens_seen": 43765600,
      "step": 22730
    },
    {
      "epoch": 3.7088669548902846,
      "grad_norm": 0.001823533559218049,
      "learning_rate": 0.11803776840456245,
      "loss": 0.1592,
      "num_input_tokens_seen": 43774832,
      "step": 22735
    },
    {
      "epoch": 3.709682682111102,
      "grad_norm": 0.002029124414548278,
      "learning_rate": 0.11798021879266997,
      "loss": 0.18,
      "num_input_tokens_seen": 43784912,
      "step": 22740
    },
    {
      "epoch": 3.7104984093319193,
      "grad_norm": 0.002553381258621812,
      "learning_rate": 0.11792267411863006,
      "loss": 0.2487,
      "num_input_tokens_seen": 43793024,
      "step": 22745
    },
    {
      "epoch": 3.711314136552737,
      "grad_norm": 0.0021216492168605328,
      "learning_rate": 0.1178651343913169,
      "loss": 0.2014,
      "num_input_tokens_seen": 43802896,
      "step": 22750
    },
    {
      "epoch": 3.712129863773554,
      "grad_norm": 0.001156610087491572,
      "learning_rate": 0.11780759961960392,
      "loss": 0.173,
      "num_input_tokens_seen": 43812032,
      "step": 22755
    },
    {
      "epoch": 3.7129455909943716,
      "grad_norm": 0.0014575818786397576,
      "learning_rate": 0.1177500698123636,
      "loss": 0.2421,
      "num_input_tokens_seen": 43820592,
      "step": 22760
    },
    {
      "epoch": 3.7137613182151887,
      "grad_norm": 0.0021210506092756987,
      "learning_rate": 0.11769254497846778,
      "loss": 0.1825,
      "num_input_tokens_seen": 43829648,
      "step": 22765
    },
    {
      "epoch": 3.7145770454360063,
      "grad_norm": 0.0022633918561041355,
      "learning_rate": 0.11763502512678758,
      "loss": 0.2142,
      "num_input_tokens_seen": 43839184,
      "step": 22770
    },
    {
      "epoch": 3.7153927726568234,
      "grad_norm": 0.002571829129010439,
      "learning_rate": 0.11757751026619315,
      "loss": 0.2302,
      "num_input_tokens_seen": 43849488,
      "step": 22775
    },
    {
      "epoch": 3.716208499877641,
      "grad_norm": 0.002778562717139721,
      "learning_rate": 0.11752000040555416,
      "loss": 0.1795,
      "num_input_tokens_seen": 43858128,
      "step": 22780
    },
    {
      "epoch": 3.717024227098458,
      "grad_norm": 0.002942524617537856,
      "learning_rate": 0.11746249555373921,
      "loss": 0.2107,
      "num_input_tokens_seen": 43869136,
      "step": 22785
    },
    {
      "epoch": 3.7178399543192757,
      "grad_norm": 0.00232516648247838,
      "learning_rate": 0.11740499571961638,
      "loss": 0.2137,
      "num_input_tokens_seen": 43877680,
      "step": 22790
    },
    {
      "epoch": 3.718655681540093,
      "grad_norm": 0.001733990153297782,
      "learning_rate": 0.11734750091205279,
      "loss": 0.2077,
      "num_input_tokens_seen": 43886960,
      "step": 22795
    },
    {
      "epoch": 3.7194714087609104,
      "grad_norm": 0.0029144510626792908,
      "learning_rate": 0.11729001113991493,
      "loss": 0.1915,
      "num_input_tokens_seen": 43896720,
      "step": 22800
    },
    {
      "epoch": 3.7194714087609104,
      "eval_loss": 0.20064640045166016,
      "eval_runtime": 67.9922,
      "eval_samples_per_second": 40.078,
      "eval_steps_per_second": 20.046,
      "num_input_tokens_seen": 43896720,
      "step": 22800
    },
    {
      "epoch": 3.7202871359817276,
      "grad_norm": 0.002969583263620734,
      "learning_rate": 0.11723252641206837,
      "loss": 0.2042,
      "num_input_tokens_seen": 43906848,
      "step": 22805
    },
    {
      "epoch": 3.721102863202545,
      "grad_norm": 0.0021824485156685114,
      "learning_rate": 0.11717504673737808,
      "loss": 0.2337,
      "num_input_tokens_seen": 43916304,
      "step": 22810
    },
    {
      "epoch": 3.7219185904233623,
      "grad_norm": 0.001364784431643784,
      "learning_rate": 0.11711757212470802,
      "loss": 0.1801,
      "num_input_tokens_seen": 43925808,
      "step": 22815
    },
    {
      "epoch": 3.72273431764418,
      "grad_norm": 0.001592146814800799,
      "learning_rate": 0.11706010258292165,
      "loss": 0.1663,
      "num_input_tokens_seen": 43935408,
      "step": 22820
    },
    {
      "epoch": 3.7235500448649974,
      "grad_norm": 0.0017167309997603297,
      "learning_rate": 0.11700263812088131,
      "loss": 0.1893,
      "num_input_tokens_seen": 43945712,
      "step": 22825
    },
    {
      "epoch": 3.7243657720858145,
      "grad_norm": 0.0023188935592770576,
      "learning_rate": 0.11694517874744892,
      "loss": 0.1929,
      "num_input_tokens_seen": 43956832,
      "step": 22830
    },
    {
      "epoch": 3.7251814993066317,
      "grad_norm": 0.003072374500334263,
      "learning_rate": 0.11688772447148532,
      "loss": 0.2642,
      "num_input_tokens_seen": 43966512,
      "step": 22835
    },
    {
      "epoch": 3.7259972265274492,
      "grad_norm": 0.0019330541836097836,
      "learning_rate": 0.11683027530185074,
      "loss": 0.2058,
      "num_input_tokens_seen": 43977152,
      "step": 22840
    },
    {
      "epoch": 3.726812953748267,
      "grad_norm": 0.0016101913060992956,
      "learning_rate": 0.11677283124740451,
      "loss": 0.2051,
      "num_input_tokens_seen": 43986592,
      "step": 22845
    },
    {
      "epoch": 3.727628680969084,
      "grad_norm": 0.0018239894416183233,
      "learning_rate": 0.11671539231700531,
      "loss": 0.196,
      "num_input_tokens_seen": 43996080,
      "step": 22850
    },
    {
      "epoch": 3.728444408189901,
      "grad_norm": 0.004823924973607063,
      "learning_rate": 0.11665795851951084,
      "loss": 0.2025,
      "num_input_tokens_seen": 44005232,
      "step": 22855
    },
    {
      "epoch": 3.7292601354107187,
      "grad_norm": 0.0031462896149605513,
      "learning_rate": 0.11660052986377825,
      "loss": 0.189,
      "num_input_tokens_seen": 44014784,
      "step": 22860
    },
    {
      "epoch": 3.7300758626315362,
      "grad_norm": 0.0028913854621350765,
      "learning_rate": 0.1165431063586636,
      "loss": 0.2669,
      "num_input_tokens_seen": 44025248,
      "step": 22865
    },
    {
      "epoch": 3.7308915898523534,
      "grad_norm": 0.002427018480375409,
      "learning_rate": 0.11648568801302245,
      "loss": 0.2119,
      "num_input_tokens_seen": 44035328,
      "step": 22870
    },
    {
      "epoch": 3.7317073170731705,
      "grad_norm": 0.0016814148984849453,
      "learning_rate": 0.11642827483570937,
      "loss": 0.1679,
      "num_input_tokens_seen": 44044192,
      "step": 22875
    },
    {
      "epoch": 3.732523044293988,
      "grad_norm": 0.003886628430336714,
      "learning_rate": 0.11637086683557815,
      "loss": 0.1754,
      "num_input_tokens_seen": 44053712,
      "step": 22880
    },
    {
      "epoch": 3.7333387715148056,
      "grad_norm": 0.0020027768332511187,
      "learning_rate": 0.11631346402148188,
      "loss": 0.2024,
      "num_input_tokens_seen": 44063280,
      "step": 22885
    },
    {
      "epoch": 3.7341544987356228,
      "grad_norm": 0.002560259308665991,
      "learning_rate": 0.11625606640227285,
      "loss": 0.2046,
      "num_input_tokens_seen": 44072320,
      "step": 22890
    },
    {
      "epoch": 3.73497022595644,
      "grad_norm": 0.0034341595601290464,
      "learning_rate": 0.11619867398680238,
      "loss": 0.277,
      "num_input_tokens_seen": 44080848,
      "step": 22895
    },
    {
      "epoch": 3.7357859531772575,
      "grad_norm": 0.0031780125573277473,
      "learning_rate": 0.11614128678392119,
      "loss": 0.2574,
      "num_input_tokens_seen": 44090080,
      "step": 22900
    },
    {
      "epoch": 3.736601680398075,
      "grad_norm": 0.0026825512759387493,
      "learning_rate": 0.11608390480247906,
      "loss": 0.1933,
      "num_input_tokens_seen": 44099408,
      "step": 22905
    },
    {
      "epoch": 3.737417407618892,
      "grad_norm": 0.00241719838231802,
      "learning_rate": 0.11602652805132499,
      "loss": 0.2499,
      "num_input_tokens_seen": 44108256,
      "step": 22910
    },
    {
      "epoch": 3.7382331348397098,
      "grad_norm": 0.0018764849519357085,
      "learning_rate": 0.11596915653930731,
      "loss": 0.1928,
      "num_input_tokens_seen": 44117200,
      "step": 22915
    },
    {
      "epoch": 3.739048862060527,
      "grad_norm": 0.001717445789836347,
      "learning_rate": 0.11591179027527328,
      "loss": 0.1952,
      "num_input_tokens_seen": 44127408,
      "step": 22920
    },
    {
      "epoch": 3.7398645892813445,
      "grad_norm": 0.0013635021168738604,
      "learning_rate": 0.11585442926806956,
      "loss": 0.2523,
      "num_input_tokens_seen": 44137136,
      "step": 22925
    },
    {
      "epoch": 3.7406803165021616,
      "grad_norm": 0.002651103539392352,
      "learning_rate": 0.11579707352654202,
      "loss": 0.2711,
      "num_input_tokens_seen": 44144944,
      "step": 22930
    },
    {
      "epoch": 3.741496043722979,
      "grad_norm": 0.0011960016563534737,
      "learning_rate": 0.11573972305953548,
      "loss": 0.2241,
      "num_input_tokens_seen": 44154656,
      "step": 22935
    },
    {
      "epoch": 3.7423117709437963,
      "grad_norm": 0.002589890733361244,
      "learning_rate": 0.11568237787589426,
      "loss": 0.2025,
      "num_input_tokens_seen": 44164416,
      "step": 22940
    },
    {
      "epoch": 3.743127498164614,
      "grad_norm": 0.0014851903542876244,
      "learning_rate": 0.11562503798446161,
      "loss": 0.2,
      "num_input_tokens_seen": 44173232,
      "step": 22945
    },
    {
      "epoch": 3.743943225385431,
      "grad_norm": 0.002042003907263279,
      "learning_rate": 0.11556770339408005,
      "loss": 0.2099,
      "num_input_tokens_seen": 44182592,
      "step": 22950
    },
    {
      "epoch": 3.7447589526062486,
      "grad_norm": 0.0016370001249015331,
      "learning_rate": 0.1155103741135914,
      "loss": 0.2003,
      "num_input_tokens_seen": 44191712,
      "step": 22955
    },
    {
      "epoch": 3.7455746798270657,
      "grad_norm": 0.002170998603105545,
      "learning_rate": 0.1154530501518364,
      "loss": 0.2247,
      "num_input_tokens_seen": 44201168,
      "step": 22960
    },
    {
      "epoch": 3.7463904070478833,
      "grad_norm": 0.0027612720150500536,
      "learning_rate": 0.11539573151765523,
      "loss": 0.1955,
      "num_input_tokens_seen": 44211632,
      "step": 22965
    },
    {
      "epoch": 3.7472061342687004,
      "grad_norm": 0.002381841652095318,
      "learning_rate": 0.11533841821988719,
      "loss": 0.2205,
      "num_input_tokens_seen": 44221024,
      "step": 22970
    },
    {
      "epoch": 3.748021861489518,
      "grad_norm": 0.0027051432989537716,
      "learning_rate": 0.11528111026737059,
      "loss": 0.2176,
      "num_input_tokens_seen": 44231280,
      "step": 22975
    },
    {
      "epoch": 3.748837588710335,
      "grad_norm": 0.0022651581093668938,
      "learning_rate": 0.11522380766894312,
      "loss": 0.2271,
      "num_input_tokens_seen": 44240960,
      "step": 22980
    },
    {
      "epoch": 3.7496533159311527,
      "grad_norm": 0.0022279289551079273,
      "learning_rate": 0.11516651043344152,
      "loss": 0.2005,
      "num_input_tokens_seen": 44249936,
      "step": 22985
    },
    {
      "epoch": 3.75046904315197,
      "grad_norm": 0.001698785345070064,
      "learning_rate": 0.11510921856970172,
      "loss": 0.2039,
      "num_input_tokens_seen": 44259296,
      "step": 22990
    },
    {
      "epoch": 3.7512847703727874,
      "grad_norm": 0.002010393887758255,
      "learning_rate": 0.11505193208655895,
      "loss": 0.1991,
      "num_input_tokens_seen": 44268800,
      "step": 22995
    },
    {
      "epoch": 3.7521004975936045,
      "grad_norm": 0.002053352538496256,
      "learning_rate": 0.11499465099284738,
      "loss": 0.1758,
      "num_input_tokens_seen": 44278640,
      "step": 23000
    },
    {
      "epoch": 3.7521004975936045,
      "eval_loss": 0.20381800830364227,
      "eval_runtime": 67.951,
      "eval_samples_per_second": 40.102,
      "eval_steps_per_second": 20.059,
      "num_input_tokens_seen": 44278640,
      "step": 23000
    },
    {
      "epoch": 3.752916224814422,
      "grad_norm": 0.0026297387667000294,
      "learning_rate": 0.1149373752974006,
      "loss": 0.188,
      "num_input_tokens_seen": 44287952,
      "step": 23005
    },
    {
      "epoch": 3.7537319520352392,
      "grad_norm": 0.0033000537659972906,
      "learning_rate": 0.11488010500905109,
      "loss": 0.1957,
      "num_input_tokens_seen": 44298192,
      "step": 23010
    },
    {
      "epoch": 3.754547679256057,
      "grad_norm": 0.003483405103906989,
      "learning_rate": 0.11482284013663077,
      "loss": 0.216,
      "num_input_tokens_seen": 44308640,
      "step": 23015
    },
    {
      "epoch": 3.7553634064768744,
      "grad_norm": 0.001985593233257532,
      "learning_rate": 0.11476558068897061,
      "loss": 0.1786,
      "num_input_tokens_seen": 44319088,
      "step": 23020
    },
    {
      "epoch": 3.7561791336976915,
      "grad_norm": 0.002463524229824543,
      "learning_rate": 0.11470832667490061,
      "loss": 0.2754,
      "num_input_tokens_seen": 44328336,
      "step": 23025
    },
    {
      "epoch": 3.7569948609185086,
      "grad_norm": 0.002269395627081394,
      "learning_rate": 0.11465107810325013,
      "loss": 0.2263,
      "num_input_tokens_seen": 44338176,
      "step": 23030
    },
    {
      "epoch": 3.7578105881393262,
      "grad_norm": 0.0022509261034429073,
      "learning_rate": 0.11459383498284771,
      "loss": 0.1723,
      "num_input_tokens_seen": 44347552,
      "step": 23035
    },
    {
      "epoch": 3.758626315360144,
      "grad_norm": 0.0018453672528266907,
      "learning_rate": 0.11453659732252082,
      "loss": 0.2328,
      "num_input_tokens_seen": 44356560,
      "step": 23040
    },
    {
      "epoch": 3.759442042580961,
      "grad_norm": 0.0029500634409487247,
      "learning_rate": 0.11447936513109633,
      "loss": 0.2087,
      "num_input_tokens_seen": 44366560,
      "step": 23045
    },
    {
      "epoch": 3.760257769801778,
      "grad_norm": 0.0014816727489233017,
      "learning_rate": 0.11442213841740011,
      "loss": 0.1747,
      "num_input_tokens_seen": 44375264,
      "step": 23050
    },
    {
      "epoch": 3.7610734970225956,
      "grad_norm": 0.002326234942302108,
      "learning_rate": 0.1143649171902572,
      "loss": 0.2393,
      "num_input_tokens_seen": 44385968,
      "step": 23055
    },
    {
      "epoch": 3.761889224243413,
      "grad_norm": 0.0009322265395894647,
      "learning_rate": 0.11430770145849194,
      "loss": 0.1661,
      "num_input_tokens_seen": 44394208,
      "step": 23060
    },
    {
      "epoch": 3.7627049514642303,
      "grad_norm": 0.002645517000928521,
      "learning_rate": 0.11425049123092756,
      "loss": 0.2089,
      "num_input_tokens_seen": 44403808,
      "step": 23065
    },
    {
      "epoch": 3.7635206786850475,
      "grad_norm": 0.001464575994759798,
      "learning_rate": 0.11419328651638674,
      "loss": 0.2129,
      "num_input_tokens_seen": 44413728,
      "step": 23070
    },
    {
      "epoch": 3.764336405905865,
      "grad_norm": 0.00196349760517478,
      "learning_rate": 0.11413608732369115,
      "loss": 0.1857,
      "num_input_tokens_seen": 44424256,
      "step": 23075
    },
    {
      "epoch": 3.7651521331266826,
      "grad_norm": 0.004454497713595629,
      "learning_rate": 0.11407889366166153,
      "loss": 0.2223,
      "num_input_tokens_seen": 44434592,
      "step": 23080
    },
    {
      "epoch": 3.7659678603474998,
      "grad_norm": 0.003249505767598748,
      "learning_rate": 0.11402170553911797,
      "loss": 0.163,
      "num_input_tokens_seen": 44443824,
      "step": 23085
    },
    {
      "epoch": 3.766783587568317,
      "grad_norm": 0.0019876956939697266,
      "learning_rate": 0.11396452296487955,
      "loss": 0.2453,
      "num_input_tokens_seen": 44453616,
      "step": 23090
    },
    {
      "epoch": 3.7675993147891345,
      "grad_norm": 0.004828844219446182,
      "learning_rate": 0.11390734594776449,
      "loss": 0.2305,
      "num_input_tokens_seen": 44462544,
      "step": 23095
    },
    {
      "epoch": 3.768415042009952,
      "grad_norm": 0.0021916034165769815,
      "learning_rate": 0.11385017449659031,
      "loss": 0.1863,
      "num_input_tokens_seen": 44473872,
      "step": 23100
    },
    {
      "epoch": 3.769230769230769,
      "grad_norm": 0.002371345879510045,
      "learning_rate": 0.11379300862017344,
      "loss": 0.2333,
      "num_input_tokens_seen": 44483664,
      "step": 23105
    },
    {
      "epoch": 3.7700464964515867,
      "grad_norm": 0.002667879918590188,
      "learning_rate": 0.11373584832732966,
      "loss": 0.2324,
      "num_input_tokens_seen": 44493120,
      "step": 23110
    },
    {
      "epoch": 3.770862223672404,
      "grad_norm": 0.002561059780418873,
      "learning_rate": 0.11367869362687386,
      "loss": 0.1962,
      "num_input_tokens_seen": 44502048,
      "step": 23115
    },
    {
      "epoch": 3.7716779508932214,
      "grad_norm": 0.0024562457110732794,
      "learning_rate": 0.11362154452761988,
      "loss": 0.2144,
      "num_input_tokens_seen": 44511920,
      "step": 23120
    },
    {
      "epoch": 3.7724936781140386,
      "grad_norm": 0.0025652642361819744,
      "learning_rate": 0.11356440103838095,
      "loss": 0.2247,
      "num_input_tokens_seen": 44521936,
      "step": 23125
    },
    {
      "epoch": 3.773309405334856,
      "grad_norm": 0.002075925935059786,
      "learning_rate": 0.11350726316796922,
      "loss": 0.1809,
      "num_input_tokens_seen": 44531840,
      "step": 23130
    },
    {
      "epoch": 3.7741251325556733,
      "grad_norm": 0.0008815979235805571,
      "learning_rate": 0.11345013092519607,
      "loss": 0.1428,
      "num_input_tokens_seen": 44540224,
      "step": 23135
    },
    {
      "epoch": 3.774940859776491,
      "grad_norm": 0.0019798032008111477,
      "learning_rate": 0.11339300431887213,
      "loss": 0.2418,
      "num_input_tokens_seen": 44549312,
      "step": 23140
    },
    {
      "epoch": 3.775756586997308,
      "grad_norm": 0.0033301906660199165,
      "learning_rate": 0.11333588335780687,
      "loss": 0.2172,
      "num_input_tokens_seen": 44559664,
      "step": 23145
    },
    {
      "epoch": 3.7765723142181256,
      "grad_norm": 0.0025622951798141003,
      "learning_rate": 0.11327876805080916,
      "loss": 0.1833,
      "num_input_tokens_seen": 44568800,
      "step": 23150
    },
    {
      "epoch": 3.7773880414389427,
      "grad_norm": 0.0018835733644664288,
      "learning_rate": 0.11322165840668696,
      "loss": 0.2416,
      "num_input_tokens_seen": 44578896,
      "step": 23155
    },
    {
      "epoch": 3.7782037686597603,
      "grad_norm": 0.0014745219377800822,
      "learning_rate": 0.11316455443424717,
      "loss": 0.2083,
      "num_input_tokens_seen": 44589344,
      "step": 23160
    },
    {
      "epoch": 3.7790194958805774,
      "grad_norm": 0.002538602566346526,
      "learning_rate": 0.11310745614229603,
      "loss": 0.213,
      "num_input_tokens_seen": 44599744,
      "step": 23165
    },
    {
      "epoch": 3.779835223101395,
      "grad_norm": 0.0021089562214910984,
      "learning_rate": 0.1130503635396387,
      "loss": 0.202,
      "num_input_tokens_seen": 44609600,
      "step": 23170
    },
    {
      "epoch": 3.780650950322212,
      "grad_norm": 0.003076982917264104,
      "learning_rate": 0.11299327663507966,
      "loss": 0.2225,
      "num_input_tokens_seen": 44619296,
      "step": 23175
    },
    {
      "epoch": 3.7814666775430297,
      "grad_norm": 0.0020871362648904324,
      "learning_rate": 0.11293619543742246,
      "loss": 0.2211,
      "num_input_tokens_seen": 44628928,
      "step": 23180
    },
    {
      "epoch": 3.782282404763847,
      "grad_norm": 0.002601550193503499,
      "learning_rate": 0.11287911995546965,
      "loss": 0.2217,
      "num_input_tokens_seen": 44638016,
      "step": 23185
    },
    {
      "epoch": 3.7830981319846644,
      "grad_norm": 0.0018041940638795495,
      "learning_rate": 0.11282205019802308,
      "loss": 0.2334,
      "num_input_tokens_seen": 44647520,
      "step": 23190
    },
    {
      "epoch": 3.783913859205482,
      "grad_norm": 0.0022442282643169165,
      "learning_rate": 0.11276498617388354,
      "loss": 0.2244,
      "num_input_tokens_seen": 44656784,
      "step": 23195
    },
    {
      "epoch": 3.784729586426299,
      "grad_norm": 0.001586667844094336,
      "learning_rate": 0.11270792789185109,
      "loss": 0.1641,
      "num_input_tokens_seen": 44666464,
      "step": 23200
    },
    {
      "epoch": 3.784729586426299,
      "eval_loss": 0.20502206683158875,
      "eval_runtime": 68.0933,
      "eval_samples_per_second": 40.019,
      "eval_steps_per_second": 20.017,
      "num_input_tokens_seen": 44666464,
      "step": 23200
    },
    {
      "epoch": 3.785545313647116,
      "grad_norm": 0.002349213697016239,
      "learning_rate": 0.11265087536072482,
      "loss": 0.1996,
      "num_input_tokens_seen": 44676944,
      "step": 23205
    },
    {
      "epoch": 3.786361040867934,
      "grad_norm": 0.0012286119163036346,
      "learning_rate": 0.11259382858930288,
      "loss": 0.1955,
      "num_input_tokens_seen": 44686912,
      "step": 23210
    },
    {
      "epoch": 3.7871767680887514,
      "grad_norm": 0.0024398076348006725,
      "learning_rate": 0.11253678758638262,
      "loss": 0.2421,
      "num_input_tokens_seen": 44696080,
      "step": 23215
    },
    {
      "epoch": 3.7879924953095685,
      "grad_norm": 0.004050595685839653,
      "learning_rate": 0.11247975236076059,
      "loss": 0.253,
      "num_input_tokens_seen": 44705408,
      "step": 23220
    },
    {
      "epoch": 3.7888082225303856,
      "grad_norm": 0.0027047928888350725,
      "learning_rate": 0.11242272292123218,
      "loss": 0.209,
      "num_input_tokens_seen": 44715424,
      "step": 23225
    },
    {
      "epoch": 3.789623949751203,
      "grad_norm": 0.0023379377089440823,
      "learning_rate": 0.11236569927659217,
      "loss": 0.2235,
      "num_input_tokens_seen": 44725536,
      "step": 23230
    },
    {
      "epoch": 3.7904396769720208,
      "grad_norm": 0.0026703623589128256,
      "learning_rate": 0.11230868143563429,
      "loss": 0.207,
      "num_input_tokens_seen": 44735840,
      "step": 23235
    },
    {
      "epoch": 3.791255404192838,
      "grad_norm": 0.002434083726257086,
      "learning_rate": 0.11225166940715131,
      "loss": 0.2226,
      "num_input_tokens_seen": 44745792,
      "step": 23240
    },
    {
      "epoch": 3.792071131413655,
      "grad_norm": 0.004179303999990225,
      "learning_rate": 0.11219466319993537,
      "loss": 0.2109,
      "num_input_tokens_seen": 44755424,
      "step": 23245
    },
    {
      "epoch": 3.7928868586344726,
      "grad_norm": 0.0019705977756530046,
      "learning_rate": 0.11213766282277739,
      "loss": 0.1903,
      "num_input_tokens_seen": 44765280,
      "step": 23250
    },
    {
      "epoch": 3.79370258585529,
      "grad_norm": 0.002305697649717331,
      "learning_rate": 0.11208066828446761,
      "loss": 0.2041,
      "num_input_tokens_seen": 44773760,
      "step": 23255
    },
    {
      "epoch": 3.7945183130761073,
      "grad_norm": 0.0025427339132875204,
      "learning_rate": 0.11202367959379537,
      "loss": 0.202,
      "num_input_tokens_seen": 44783280,
      "step": 23260
    },
    {
      "epoch": 3.7953340402969244,
      "grad_norm": 0.003717358224093914,
      "learning_rate": 0.11196669675954894,
      "loss": 0.2016,
      "num_input_tokens_seen": 44793040,
      "step": 23265
    },
    {
      "epoch": 3.796149767517742,
      "grad_norm": 0.001677213003858924,
      "learning_rate": 0.1119097197905158,
      "loss": 0.1349,
      "num_input_tokens_seen": 44803072,
      "step": 23270
    },
    {
      "epoch": 3.7969654947385596,
      "grad_norm": 0.003785987151786685,
      "learning_rate": 0.11185274869548259,
      "loss": 0.2251,
      "num_input_tokens_seen": 44812848,
      "step": 23275
    },
    {
      "epoch": 3.7977812219593767,
      "grad_norm": 0.0021089378278702497,
      "learning_rate": 0.11179578348323486,
      "loss": 0.1673,
      "num_input_tokens_seen": 44821664,
      "step": 23280
    },
    {
      "epoch": 3.7985969491801943,
      "grad_norm": 0.002736347494646907,
      "learning_rate": 0.1117388241625575,
      "loss": 0.2073,
      "num_input_tokens_seen": 44831312,
      "step": 23285
    },
    {
      "epoch": 3.7994126764010114,
      "grad_norm": 0.0019294072408229113,
      "learning_rate": 0.11168187074223421,
      "loss": 0.2516,
      "num_input_tokens_seen": 44841072,
      "step": 23290
    },
    {
      "epoch": 3.800228403621829,
      "grad_norm": 0.002455860609188676,
      "learning_rate": 0.11162492323104796,
      "loss": 0.2486,
      "num_input_tokens_seen": 44850896,
      "step": 23295
    },
    {
      "epoch": 3.801044130842646,
      "grad_norm": 0.005034377798438072,
      "learning_rate": 0.11156798163778091,
      "loss": 0.1824,
      "num_input_tokens_seen": 44859344,
      "step": 23300
    },
    {
      "epoch": 3.8018598580634637,
      "grad_norm": 0.001008868683129549,
      "learning_rate": 0.11151104597121399,
      "loss": 0.2356,
      "num_input_tokens_seen": 44868992,
      "step": 23305
    },
    {
      "epoch": 3.802675585284281,
      "grad_norm": 0.002748217200860381,
      "learning_rate": 0.11145411624012742,
      "loss": 0.2136,
      "num_input_tokens_seen": 44880224,
      "step": 23310
    },
    {
      "epoch": 3.8034913125050984,
      "grad_norm": 0.0030752497259527445,
      "learning_rate": 0.11139719245330063,
      "loss": 0.2572,
      "num_input_tokens_seen": 44888976,
      "step": 23315
    },
    {
      "epoch": 3.8043070397259156,
      "grad_norm": 0.0024853842332959175,
      "learning_rate": 0.11134027461951179,
      "loss": 0.2263,
      "num_input_tokens_seen": 44898064,
      "step": 23320
    },
    {
      "epoch": 3.805122766946733,
      "grad_norm": 0.00263150199316442,
      "learning_rate": 0.11128336274753849,
      "loss": 0.2268,
      "num_input_tokens_seen": 44908000,
      "step": 23325
    },
    {
      "epoch": 3.8059384941675503,
      "grad_norm": 0.0032011643052101135,
      "learning_rate": 0.11122645684615715,
      "loss": 0.2038,
      "num_input_tokens_seen": 44916768,
      "step": 23330
    },
    {
      "epoch": 3.806754221388368,
      "grad_norm": 0.0013167993165552616,
      "learning_rate": 0.11116955692414345,
      "loss": 0.2063,
      "num_input_tokens_seen": 44926192,
      "step": 23335
    },
    {
      "epoch": 3.807569948609185,
      "grad_norm": 0.002124513266608119,
      "learning_rate": 0.11111266299027203,
      "loss": 0.195,
      "num_input_tokens_seen": 44935472,
      "step": 23340
    },
    {
      "epoch": 3.8083856758300025,
      "grad_norm": 0.00285601569339633,
      "learning_rate": 0.11105577505331668,
      "loss": 0.2123,
      "num_input_tokens_seen": 44944560,
      "step": 23345
    },
    {
      "epoch": 3.8092014030508197,
      "grad_norm": 0.002322019310668111,
      "learning_rate": 0.11099889312205018,
      "loss": 0.2272,
      "num_input_tokens_seen": 44952576,
      "step": 23350
    },
    {
      "epoch": 3.8100171302716372,
      "grad_norm": 0.0034681304823607206,
      "learning_rate": 0.11094201720524455,
      "loss": 0.2732,
      "num_input_tokens_seen": 44962432,
      "step": 23355
    },
    {
      "epoch": 3.8108328574924544,
      "grad_norm": 0.004574883263558149,
      "learning_rate": 0.11088514731167064,
      "loss": 0.2048,
      "num_input_tokens_seen": 44972480,
      "step": 23360
    },
    {
      "epoch": 3.811648584713272,
      "grad_norm": 0.0027780767995864153,
      "learning_rate": 0.11082828345009862,
      "loss": 0.1824,
      "num_input_tokens_seen": 44983040,
      "step": 23365
    },
    {
      "epoch": 3.812464311934089,
      "grad_norm": 0.002631571376696229,
      "learning_rate": 0.11077142562929748,
      "loss": 0.2309,
      "num_input_tokens_seen": 44992304,
      "step": 23370
    },
    {
      "epoch": 3.8132800391549067,
      "grad_norm": 0.0023117312230169773,
      "learning_rate": 0.11071457385803554,
      "loss": 0.2119,
      "num_input_tokens_seen": 45001552,
      "step": 23375
    },
    {
      "epoch": 3.814095766375724,
      "grad_norm": 0.0031315977685153484,
      "learning_rate": 0.11065772814508001,
      "loss": 0.1616,
      "num_input_tokens_seen": 45011152,
      "step": 23380
    },
    {
      "epoch": 3.8149114935965414,
      "grad_norm": 0.0016411020187661052,
      "learning_rate": 0.11060088849919715,
      "loss": 0.2048,
      "num_input_tokens_seen": 45020496,
      "step": 23385
    },
    {
      "epoch": 3.815727220817359,
      "grad_norm": 0.0013082092627882957,
      "learning_rate": 0.11054405492915244,
      "loss": 0.2163,
      "num_input_tokens_seen": 45029632,
      "step": 23390
    },
    {
      "epoch": 3.816542948038176,
      "grad_norm": 0.0025091315619647503,
      "learning_rate": 0.11048722744371031,
      "loss": 0.185,
      "num_input_tokens_seen": 45038288,
      "step": 23395
    },
    {
      "epoch": 3.817358675258993,
      "grad_norm": 0.0018256010953336954,
      "learning_rate": 0.1104304060516342,
      "loss": 0.2257,
      "num_input_tokens_seen": 45047360,
      "step": 23400
    },
    {
      "epoch": 3.817358675258993,
      "eval_loss": 0.20431892573833466,
      "eval_runtime": 67.9877,
      "eval_samples_per_second": 40.081,
      "eval_steps_per_second": 20.048,
      "num_input_tokens_seen": 45047360,
      "step": 23400
    },
    {
      "epoch": 3.8181744024798108,
      "grad_norm": 0.002217255998402834,
      "learning_rate": 0.11037359076168682,
      "loss": 0.1578,
      "num_input_tokens_seen": 45056672,
      "step": 23405
    },
    {
      "epoch": 3.8189901297006283,
      "grad_norm": 0.003013692097738385,
      "learning_rate": 0.11031678158262966,
      "loss": 0.2047,
      "num_input_tokens_seen": 45065712,
      "step": 23410
    },
    {
      "epoch": 3.8198058569214455,
      "grad_norm": 0.004243901930749416,
      "learning_rate": 0.11025997852322349,
      "loss": 0.1973,
      "num_input_tokens_seen": 45074112,
      "step": 23415
    },
    {
      "epoch": 3.8206215841422626,
      "grad_norm": 0.0016062975628301501,
      "learning_rate": 0.11020318159222807,
      "loss": 0.2058,
      "num_input_tokens_seen": 45081824,
      "step": 23420
    },
    {
      "epoch": 3.82143731136308,
      "grad_norm": 0.0024816179648041725,
      "learning_rate": 0.1101463907984021,
      "loss": 0.224,
      "num_input_tokens_seen": 45091424,
      "step": 23425
    },
    {
      "epoch": 3.8222530385838978,
      "grad_norm": 0.0014017985668033361,
      "learning_rate": 0.11008960615050352,
      "loss": 0.1974,
      "num_input_tokens_seen": 45101664,
      "step": 23430
    },
    {
      "epoch": 3.823068765804715,
      "grad_norm": 0.002146300161257386,
      "learning_rate": 0.11003282765728925,
      "loss": 0.2585,
      "num_input_tokens_seen": 45112192,
      "step": 23435
    },
    {
      "epoch": 3.823884493025532,
      "grad_norm": 0.002794125583022833,
      "learning_rate": 0.10997605532751518,
      "loss": 0.2055,
      "num_input_tokens_seen": 45123392,
      "step": 23440
    },
    {
      "epoch": 3.8247002202463496,
      "grad_norm": 0.0016358771827071905,
      "learning_rate": 0.1099192891699364,
      "loss": 0.2276,
      "num_input_tokens_seen": 45133584,
      "step": 23445
    },
    {
      "epoch": 3.825515947467167,
      "grad_norm": 0.0028199029620736837,
      "learning_rate": 0.10986252919330687,
      "loss": 0.2655,
      "num_input_tokens_seen": 45143552,
      "step": 23450
    },
    {
      "epoch": 3.8263316746879843,
      "grad_norm": 0.0035014834720641375,
      "learning_rate": 0.10980577540637973,
      "loss": 0.265,
      "num_input_tokens_seen": 45152832,
      "step": 23455
    },
    {
      "epoch": 3.8271474019088014,
      "grad_norm": 0.00202010665088892,
      "learning_rate": 0.10974902781790719,
      "loss": 0.2091,
      "num_input_tokens_seen": 45161776,
      "step": 23460
    },
    {
      "epoch": 3.827963129129619,
      "grad_norm": 0.0022104138042777777,
      "learning_rate": 0.10969228643664032,
      "loss": 0.1926,
      "num_input_tokens_seen": 45172352,
      "step": 23465
    },
    {
      "epoch": 3.8287788563504366,
      "grad_norm": 0.0019596475176513195,
      "learning_rate": 0.10963555127132942,
      "loss": 0.2045,
      "num_input_tokens_seen": 45182288,
      "step": 23470
    },
    {
      "epoch": 3.8295945835712537,
      "grad_norm": 0.0013887628447264433,
      "learning_rate": 0.10957882233072382,
      "loss": 0.188,
      "num_input_tokens_seen": 45191664,
      "step": 23475
    },
    {
      "epoch": 3.8304103107920713,
      "grad_norm": 0.0029235580004751682,
      "learning_rate": 0.10952209962357176,
      "loss": 0.2557,
      "num_input_tokens_seen": 45200240,
      "step": 23480
    },
    {
      "epoch": 3.8312260380128884,
      "grad_norm": 0.0013753839302808046,
      "learning_rate": 0.10946538315862062,
      "loss": 0.1986,
      "num_input_tokens_seen": 45209968,
      "step": 23485
    },
    {
      "epoch": 3.832041765233706,
      "grad_norm": 0.0024062003940343857,
      "learning_rate": 0.10940867294461679,
      "loss": 0.1751,
      "num_input_tokens_seen": 45219520,
      "step": 23490
    },
    {
      "epoch": 3.832857492454523,
      "grad_norm": 0.003055289387702942,
      "learning_rate": 0.10935196899030565,
      "loss": 0.1867,
      "num_input_tokens_seen": 45228832,
      "step": 23495
    },
    {
      "epoch": 3.8336732196753407,
      "grad_norm": 0.0024055244866758585,
      "learning_rate": 0.10929527130443177,
      "loss": 0.2364,
      "num_input_tokens_seen": 45238704,
      "step": 23500
    },
    {
      "epoch": 3.834488946896158,
      "grad_norm": 0.0027316943742334843,
      "learning_rate": 0.1092385798957385,
      "loss": 0.2249,
      "num_input_tokens_seen": 45248208,
      "step": 23505
    },
    {
      "epoch": 3.8353046741169754,
      "grad_norm": 0.0008723723003640771,
      "learning_rate": 0.10918189477296848,
      "loss": 0.2138,
      "num_input_tokens_seen": 45257296,
      "step": 23510
    },
    {
      "epoch": 3.8361204013377925,
      "grad_norm": 0.0016440956387668848,
      "learning_rate": 0.1091252159448633,
      "loss": 0.203,
      "num_input_tokens_seen": 45266672,
      "step": 23515
    },
    {
      "epoch": 3.83693612855861,
      "grad_norm": 0.001288123894482851,
      "learning_rate": 0.10906854342016345,
      "loss": 0.1857,
      "num_input_tokens_seen": 45276528,
      "step": 23520
    },
    {
      "epoch": 3.8377518557794272,
      "grad_norm": 0.0013888924149796367,
      "learning_rate": 0.10901187720760858,
      "loss": 0.211,
      "num_input_tokens_seen": 45284384,
      "step": 23525
    },
    {
      "epoch": 3.838567583000245,
      "grad_norm": 0.0014833855675533414,
      "learning_rate": 0.10895521731593734,
      "loss": 0.2032,
      "num_input_tokens_seen": 45292896,
      "step": 23530
    },
    {
      "epoch": 3.839383310221062,
      "grad_norm": 0.0032075392082333565,
      "learning_rate": 0.10889856375388733,
      "loss": 0.1986,
      "num_input_tokens_seen": 45303168,
      "step": 23535
    },
    {
      "epoch": 3.8401990374418795,
      "grad_norm": 0.0011110648047178984,
      "learning_rate": 0.1088419165301954,
      "loss": 0.1682,
      "num_input_tokens_seen": 45312992,
      "step": 23540
    },
    {
      "epoch": 3.8410147646626966,
      "grad_norm": 0.0016220222460106015,
      "learning_rate": 0.1087852756535971,
      "loss": 0.2152,
      "num_input_tokens_seen": 45323296,
      "step": 23545
    },
    {
      "epoch": 3.841830491883514,
      "grad_norm": 0.0016012676060199738,
      "learning_rate": 0.10872864113282725,
      "loss": 0.1824,
      "num_input_tokens_seen": 45331424,
      "step": 23550
    },
    {
      "epoch": 3.8426462191043314,
      "grad_norm": 0.0014699248131364584,
      "learning_rate": 0.10867201297661958,
      "loss": 0.2387,
      "num_input_tokens_seen": 45339904,
      "step": 23555
    },
    {
      "epoch": 3.843461946325149,
      "grad_norm": 0.0012889039935544133,
      "learning_rate": 0.10861539119370689,
      "loss": 0.2312,
      "num_input_tokens_seen": 45348704,
      "step": 23560
    },
    {
      "epoch": 3.8442776735459665,
      "grad_norm": 0.0025727502070367336,
      "learning_rate": 0.10855877579282096,
      "loss": 0.1905,
      "num_input_tokens_seen": 45357840,
      "step": 23565
    },
    {
      "epoch": 3.8450934007667836,
      "grad_norm": 0.0016402757028117776,
      "learning_rate": 0.10850216678269252,
      "loss": 0.2399,
      "num_input_tokens_seen": 45367216,
      "step": 23570
    },
    {
      "epoch": 3.8459091279876008,
      "grad_norm": 0.0017882250249385834,
      "learning_rate": 0.10844556417205146,
      "loss": 0.1777,
      "num_input_tokens_seen": 45375296,
      "step": 23575
    },
    {
      "epoch": 3.8467248552084183,
      "grad_norm": 0.0023915409110486507,
      "learning_rate": 0.10838896796962669,
      "loss": 0.1737,
      "num_input_tokens_seen": 45385632,
      "step": 23580
    },
    {
      "epoch": 3.847540582429236,
      "grad_norm": 0.0039905705489218235,
      "learning_rate": 0.1083323781841459,
      "loss": 0.229,
      "num_input_tokens_seen": 45395536,
      "step": 23585
    },
    {
      "epoch": 3.848356309650053,
      "grad_norm": 0.0013860516482964158,
      "learning_rate": 0.10827579482433607,
      "loss": 0.1841,
      "num_input_tokens_seen": 45405072,
      "step": 23590
    },
    {
      "epoch": 3.84917203687087,
      "grad_norm": 0.0022627480793744326,
      "learning_rate": 0.10821921789892304,
      "loss": 0.1696,
      "num_input_tokens_seen": 45416080,
      "step": 23595
    },
    {
      "epoch": 3.8499877640916877,
      "grad_norm": 0.0024531790986657143,
      "learning_rate": 0.10816264741663158,
      "loss": 0.2563,
      "num_input_tokens_seen": 45426496,
      "step": 23600
    },
    {
      "epoch": 3.8499877640916877,
      "eval_loss": 0.1992967575788498,
      "eval_runtime": 67.9833,
      "eval_samples_per_second": 40.083,
      "eval_steps_per_second": 20.049,
      "num_input_tokens_seen": 45426496,
      "step": 23600
    },
    {
      "epoch": 3.8508034913125053,
      "grad_norm": 0.0025683266576379538,
      "learning_rate": 0.10810608338618573,
      "loss": 0.241,
      "num_input_tokens_seen": 45435712,
      "step": 23605
    },
    {
      "epoch": 3.8516192185333225,
      "grad_norm": 0.0037264544516801834,
      "learning_rate": 0.10804952581630821,
      "loss": 0.2123,
      "num_input_tokens_seen": 45446432,
      "step": 23610
    },
    {
      "epoch": 3.8524349457541396,
      "grad_norm": 0.002171426312997937,
      "learning_rate": 0.10799297471572102,
      "loss": 0.2405,
      "num_input_tokens_seen": 45455760,
      "step": 23615
    },
    {
      "epoch": 3.853250672974957,
      "grad_norm": 0.0029582602437585592,
      "learning_rate": 0.10793643009314507,
      "loss": 0.2333,
      "num_input_tokens_seen": 45466304,
      "step": 23620
    },
    {
      "epoch": 3.8540664001957747,
      "grad_norm": 0.0017168777994811535,
      "learning_rate": 0.10787989195730015,
      "loss": 0.2152,
      "num_input_tokens_seen": 45474544,
      "step": 23625
    },
    {
      "epoch": 3.854882127416592,
      "grad_norm": 0.0023438388016074896,
      "learning_rate": 0.10782336031690525,
      "loss": 0.254,
      "num_input_tokens_seen": 45484736,
      "step": 23630
    },
    {
      "epoch": 3.855697854637409,
      "grad_norm": 0.00192731071729213,
      "learning_rate": 0.10776683518067821,
      "loss": 0.1923,
      "num_input_tokens_seen": 45494064,
      "step": 23635
    },
    {
      "epoch": 3.8565135818582266,
      "grad_norm": 0.001387129188515246,
      "learning_rate": 0.10771031655733587,
      "loss": 0.1874,
      "num_input_tokens_seen": 45504016,
      "step": 23640
    },
    {
      "epoch": 3.857329309079044,
      "grad_norm": 0.0017088403692469,
      "learning_rate": 0.10765380445559422,
      "loss": 0.217,
      "num_input_tokens_seen": 45512720,
      "step": 23645
    },
    {
      "epoch": 3.8581450362998613,
      "grad_norm": 0.0011545693268999457,
      "learning_rate": 0.10759729888416801,
      "loss": 0.2214,
      "num_input_tokens_seen": 45523056,
      "step": 23650
    },
    {
      "epoch": 3.858960763520679,
      "grad_norm": 0.002202635630965233,
      "learning_rate": 0.10754079985177119,
      "loss": 0.1888,
      "num_input_tokens_seen": 45532672,
      "step": 23655
    },
    {
      "epoch": 3.859776490741496,
      "grad_norm": 0.0013410153333097696,
      "learning_rate": 0.10748430736711667,
      "loss": 0.1791,
      "num_input_tokens_seen": 45541632,
      "step": 23660
    },
    {
      "epoch": 3.8605922179623136,
      "grad_norm": 0.0020184831228107214,
      "learning_rate": 0.10742782143891623,
      "loss": 0.2057,
      "num_input_tokens_seen": 45552016,
      "step": 23665
    },
    {
      "epoch": 3.8614079451831307,
      "grad_norm": 0.0016878900351002812,
      "learning_rate": 0.10737134207588069,
      "loss": 0.1882,
      "num_input_tokens_seen": 45562048,
      "step": 23670
    },
    {
      "epoch": 3.8622236724039483,
      "grad_norm": 0.0024882331490516663,
      "learning_rate": 0.10731486928671992,
      "loss": 0.2231,
      "num_input_tokens_seen": 45572544,
      "step": 23675
    },
    {
      "epoch": 3.8630393996247654,
      "grad_norm": 0.0022286316379904747,
      "learning_rate": 0.10725840308014269,
      "loss": 0.1724,
      "num_input_tokens_seen": 45582672,
      "step": 23680
    },
    {
      "epoch": 3.863855126845583,
      "grad_norm": 0.0028467460069805384,
      "learning_rate": 0.10720194346485688,
      "loss": 0.2254,
      "num_input_tokens_seen": 45592464,
      "step": 23685
    },
    {
      "epoch": 3.8646708540664,
      "grad_norm": 0.004320256412029266,
      "learning_rate": 0.10714549044956918,
      "loss": 0.2577,
      "num_input_tokens_seen": 45602160,
      "step": 23690
    },
    {
      "epoch": 3.8654865812872177,
      "grad_norm": 0.0032870243303477764,
      "learning_rate": 0.10708904404298542,
      "loss": 0.2268,
      "num_input_tokens_seen": 45611504,
      "step": 23695
    },
    {
      "epoch": 3.866302308508035,
      "grad_norm": 0.0017756287707015872,
      "learning_rate": 0.1070326042538103,
      "loss": 0.1947,
      "num_input_tokens_seen": 45621232,
      "step": 23700
    },
    {
      "epoch": 3.8671180357288524,
      "grad_norm": 0.002194675849750638,
      "learning_rate": 0.10697617109074758,
      "loss": 0.1658,
      "num_input_tokens_seen": 45630576,
      "step": 23705
    },
    {
      "epoch": 3.8679337629496695,
      "grad_norm": 0.0028315498493611813,
      "learning_rate": 0.10691974456249999,
      "loss": 0.2373,
      "num_input_tokens_seen": 45640240,
      "step": 23710
    },
    {
      "epoch": 3.868749490170487,
      "grad_norm": 0.0017591473879292607,
      "learning_rate": 0.10686332467776909,
      "loss": 0.2198,
      "num_input_tokens_seen": 45649680,
      "step": 23715
    },
    {
      "epoch": 3.869565217391304,
      "grad_norm": 0.0018899370916187763,
      "learning_rate": 0.10680691144525563,
      "loss": 0.2291,
      "num_input_tokens_seen": 45658832,
      "step": 23720
    },
    {
      "epoch": 3.870380944612122,
      "grad_norm": 0.0017948533641174436,
      "learning_rate": 0.10675050487365928,
      "loss": 0.2237,
      "num_input_tokens_seen": 45667744,
      "step": 23725
    },
    {
      "epoch": 3.871196671832939,
      "grad_norm": 0.0029742789920419455,
      "learning_rate": 0.10669410497167851,
      "loss": 0.2555,
      "num_input_tokens_seen": 45677760,
      "step": 23730
    },
    {
      "epoch": 3.8720123990537565,
      "grad_norm": 0.0016827172366902232,
      "learning_rate": 0.10663771174801102,
      "loss": 0.186,
      "num_input_tokens_seen": 45687088,
      "step": 23735
    },
    {
      "epoch": 3.8728281262745736,
      "grad_norm": 0.0018423489527776837,
      "learning_rate": 0.10658132521135329,
      "loss": 0.221,
      "num_input_tokens_seen": 45696816,
      "step": 23740
    },
    {
      "epoch": 3.873643853495391,
      "grad_norm": 0.001835763338021934,
      "learning_rate": 0.10652494537040084,
      "loss": 0.2057,
      "num_input_tokens_seen": 45706544,
      "step": 23745
    },
    {
      "epoch": 3.8744595807162083,
      "grad_norm": 0.002407278399914503,
      "learning_rate": 0.1064685722338482,
      "loss": 0.2119,
      "num_input_tokens_seen": 45716320,
      "step": 23750
    },
    {
      "epoch": 3.875275307937026,
      "grad_norm": 0.0029854122549295425,
      "learning_rate": 0.10641220581038871,
      "loss": 0.1907,
      "num_input_tokens_seen": 45726992,
      "step": 23755
    },
    {
      "epoch": 3.8760910351578435,
      "grad_norm": 0.0018504793988540769,
      "learning_rate": 0.10635584610871483,
      "loss": 0.181,
      "num_input_tokens_seen": 45736160,
      "step": 23760
    },
    {
      "epoch": 3.8769067623786606,
      "grad_norm": 0.0021403608843684196,
      "learning_rate": 0.10629949313751803,
      "loss": 0.2141,
      "num_input_tokens_seen": 45744976,
      "step": 23765
    },
    {
      "epoch": 3.8777224895994777,
      "grad_norm": 0.0025614576879888773,
      "learning_rate": 0.10624314690548849,
      "loss": 0.1947,
      "num_input_tokens_seen": 45754912,
      "step": 23770
    },
    {
      "epoch": 3.8785382168202953,
      "grad_norm": 0.0018087916541844606,
      "learning_rate": 0.1061868074213156,
      "loss": 0.1764,
      "num_input_tokens_seen": 45765200,
      "step": 23775
    },
    {
      "epoch": 3.879353944041113,
      "grad_norm": 0.0018525675404816866,
      "learning_rate": 0.10613047469368765,
      "loss": 0.2052,
      "num_input_tokens_seen": 45775280,
      "step": 23780
    },
    {
      "epoch": 3.88016967126193,
      "grad_norm": 0.0015468454221263528,
      "learning_rate": 0.10607414873129171,
      "loss": 0.1981,
      "num_input_tokens_seen": 45784832,
      "step": 23785
    },
    {
      "epoch": 3.880985398482747,
      "grad_norm": 0.0015303255058825016,
      "learning_rate": 0.10601782954281413,
      "loss": 0.1818,
      "num_input_tokens_seen": 45794032,
      "step": 23790
    },
    {
      "epoch": 3.8818011257035647,
      "grad_norm": 0.0016246216837316751,
      "learning_rate": 0.1059615171369399,
      "loss": 0.1998,
      "num_input_tokens_seen": 45803360,
      "step": 23795
    },
    {
      "epoch": 3.8826168529243823,
      "grad_norm": 0.0016629662131890655,
      "learning_rate": 0.10590521152235312,
      "loss": 0.2483,
      "num_input_tokens_seen": 45813536,
      "step": 23800
    },
    {
      "epoch": 3.8826168529243823,
      "eval_loss": 0.19949309527873993,
      "eval_runtime": 68.0214,
      "eval_samples_per_second": 40.061,
      "eval_steps_per_second": 20.038,
      "num_input_tokens_seen": 45813536,
      "step": 23800
    },
    {
      "epoch": 3.8834325801451994,
      "grad_norm": 0.002871827455237508,
      "learning_rate": 0.1058489127077369,
      "loss": 0.2307,
      "num_input_tokens_seen": 45823008,
      "step": 23805
    },
    {
      "epoch": 3.8842483073660166,
      "grad_norm": 0.0030733365565538406,
      "learning_rate": 0.1057926207017732,
      "loss": 0.1704,
      "num_input_tokens_seen": 45830208,
      "step": 23810
    },
    {
      "epoch": 3.885064034586834,
      "grad_norm": 0.002008663257583976,
      "learning_rate": 0.10573633551314285,
      "loss": 0.2879,
      "num_input_tokens_seen": 45840928,
      "step": 23815
    },
    {
      "epoch": 3.8858797618076517,
      "grad_norm": 0.0026702075265347958,
      "learning_rate": 0.1056800571505259,
      "loss": 0.1907,
      "num_input_tokens_seen": 45850992,
      "step": 23820
    },
    {
      "epoch": 3.886695489028469,
      "grad_norm": 0.0013478012988343835,
      "learning_rate": 0.10562378562260105,
      "loss": 0.193,
      "num_input_tokens_seen": 45860320,
      "step": 23825
    },
    {
      "epoch": 3.887511216249286,
      "grad_norm": 0.0025534811429679394,
      "learning_rate": 0.10556752093804615,
      "loss": 0.2661,
      "num_input_tokens_seen": 45870368,
      "step": 23830
    },
    {
      "epoch": 3.8883269434701035,
      "grad_norm": 0.0016981707885861397,
      "learning_rate": 0.10551126310553786,
      "loss": 0.2125,
      "num_input_tokens_seen": 45879744,
      "step": 23835
    },
    {
      "epoch": 3.889142670690921,
      "grad_norm": 0.0016575169283896685,
      "learning_rate": 0.10545501213375187,
      "loss": 0.1652,
      "num_input_tokens_seen": 45889424,
      "step": 23840
    },
    {
      "epoch": 3.8899583979117383,
      "grad_norm": 0.001680668443441391,
      "learning_rate": 0.10539876803136287,
      "loss": 0.2173,
      "num_input_tokens_seen": 45899344,
      "step": 23845
    },
    {
      "epoch": 3.890774125132556,
      "grad_norm": 0.0016675192164257169,
      "learning_rate": 0.10534253080704428,
      "loss": 0.1343,
      "num_input_tokens_seen": 45907088,
      "step": 23850
    },
    {
      "epoch": 3.891589852353373,
      "grad_norm": 0.0038611937779933214,
      "learning_rate": 0.10528630046946862,
      "loss": 0.1714,
      "num_input_tokens_seen": 45915600,
      "step": 23855
    },
    {
      "epoch": 3.8924055795741905,
      "grad_norm": 0.001770879840478301,
      "learning_rate": 0.1052300770273074,
      "loss": 0.2183,
      "num_input_tokens_seen": 45925856,
      "step": 23860
    },
    {
      "epoch": 3.8932213067950077,
      "grad_norm": 0.0037389337085187435,
      "learning_rate": 0.10517386048923086,
      "loss": 0.2253,
      "num_input_tokens_seen": 45935328,
      "step": 23865
    },
    {
      "epoch": 3.8940370340158252,
      "grad_norm": 0.0036713569425046444,
      "learning_rate": 0.10511765086390841,
      "loss": 0.2034,
      "num_input_tokens_seen": 45944832,
      "step": 23870
    },
    {
      "epoch": 3.8948527612366424,
      "grad_norm": 0.0017755554290488362,
      "learning_rate": 0.10506144816000816,
      "loss": 0.1791,
      "num_input_tokens_seen": 45955264,
      "step": 23875
    },
    {
      "epoch": 3.89566848845746,
      "grad_norm": 0.0029274632688611746,
      "learning_rate": 0.10500525238619736,
      "loss": 0.2299,
      "num_input_tokens_seen": 45963312,
      "step": 23880
    },
    {
      "epoch": 3.896484215678277,
      "grad_norm": 0.0024853518698364496,
      "learning_rate": 0.10494906355114209,
      "loss": 0.2387,
      "num_input_tokens_seen": 45973056,
      "step": 23885
    },
    {
      "epoch": 3.8972999428990946,
      "grad_norm": 0.002034431090578437,
      "learning_rate": 0.10489288166350737,
      "loss": 0.1809,
      "num_input_tokens_seen": 45982416,
      "step": 23890
    },
    {
      "epoch": 3.898115670119912,
      "grad_norm": 0.0019382641185075045,
      "learning_rate": 0.10483670673195711,
      "loss": 0.2254,
      "num_input_tokens_seen": 45993104,
      "step": 23895
    },
    {
      "epoch": 3.8989313973407294,
      "grad_norm": 0.0016676423838362098,
      "learning_rate": 0.10478053876515431,
      "loss": 0.2242,
      "num_input_tokens_seen": 46002528,
      "step": 23900
    },
    {
      "epoch": 3.8997471245615465,
      "grad_norm": 0.001780638238415122,
      "learning_rate": 0.10472437777176061,
      "loss": 0.2136,
      "num_input_tokens_seen": 46011408,
      "step": 23905
    },
    {
      "epoch": 3.900562851782364,
      "grad_norm": 0.0022160648368299007,
      "learning_rate": 0.1046682237604369,
      "loss": 0.2281,
      "num_input_tokens_seen": 46019280,
      "step": 23910
    },
    {
      "epoch": 3.901378579003181,
      "grad_norm": 0.003552102716639638,
      "learning_rate": 0.1046120767398427,
      "loss": 0.25,
      "num_input_tokens_seen": 46028720,
      "step": 23915
    },
    {
      "epoch": 3.9021943062239988,
      "grad_norm": 0.0020967249292880297,
      "learning_rate": 0.10455593671863667,
      "loss": 0.1872,
      "num_input_tokens_seen": 46038192,
      "step": 23920
    },
    {
      "epoch": 3.903010033444816,
      "grad_norm": 0.0020500565879046917,
      "learning_rate": 0.1044998037054763,
      "loss": 0.1847,
      "num_input_tokens_seen": 46047120,
      "step": 23925
    },
    {
      "epoch": 3.9038257606656335,
      "grad_norm": 0.003815381322056055,
      "learning_rate": 0.10444367770901794,
      "loss": 0.2287,
      "num_input_tokens_seen": 46057008,
      "step": 23930
    },
    {
      "epoch": 3.904641487886451,
      "grad_norm": 0.0016337692504748702,
      "learning_rate": 0.10438755873791698,
      "loss": 0.2043,
      "num_input_tokens_seen": 46066176,
      "step": 23935
    },
    {
      "epoch": 3.905457215107268,
      "grad_norm": 0.002837869804352522,
      "learning_rate": 0.10433144680082775,
      "loss": 0.2271,
      "num_input_tokens_seen": 46076464,
      "step": 23940
    },
    {
      "epoch": 3.9062729423280853,
      "grad_norm": 0.0030731786973774433,
      "learning_rate": 0.10427534190640322,
      "loss": 0.2272,
      "num_input_tokens_seen": 46086656,
      "step": 23945
    },
    {
      "epoch": 3.907088669548903,
      "grad_norm": 0.002731954213231802,
      "learning_rate": 0.10421924406329568,
      "loss": 0.195,
      "num_input_tokens_seen": 46095648,
      "step": 23950
    },
    {
      "epoch": 3.9079043967697205,
      "grad_norm": 0.002623209496960044,
      "learning_rate": 0.10416315328015598,
      "loss": 0.2006,
      "num_input_tokens_seen": 46104960,
      "step": 23955
    },
    {
      "epoch": 3.9087201239905376,
      "grad_norm": 0.00236231810413301,
      "learning_rate": 0.10410706956563402,
      "loss": 0.2306,
      "num_input_tokens_seen": 46114992,
      "step": 23960
    },
    {
      "epoch": 3.9095358512113547,
      "grad_norm": 0.0018114980775862932,
      "learning_rate": 0.10405099292837874,
      "loss": 0.2088,
      "num_input_tokens_seen": 46123728,
      "step": 23965
    },
    {
      "epoch": 3.9103515784321723,
      "grad_norm": 0.0014486616710200906,
      "learning_rate": 0.10399492337703771,
      "loss": 0.2401,
      "num_input_tokens_seen": 46134016,
      "step": 23970
    },
    {
      "epoch": 3.91116730565299,
      "grad_norm": 0.0017445554258301854,
      "learning_rate": 0.10393886092025764,
      "loss": 0.1929,
      "num_input_tokens_seen": 46145088,
      "step": 23975
    },
    {
      "epoch": 3.911983032873807,
      "grad_norm": 0.0017799382330849767,
      "learning_rate": 0.10388280556668412,
      "loss": 0.2212,
      "num_input_tokens_seen": 46155728,
      "step": 23980
    },
    {
      "epoch": 3.912798760094624,
      "grad_norm": 0.002516968408599496,
      "learning_rate": 0.10382675732496145,
      "loss": 0.2328,
      "num_input_tokens_seen": 46165072,
      "step": 23985
    },
    {
      "epoch": 3.9136144873154417,
      "grad_norm": 0.003051685867831111,
      "learning_rate": 0.10377071620373311,
      "loss": 0.1789,
      "num_input_tokens_seen": 46174416,
      "step": 23990
    },
    {
      "epoch": 3.9144302145362593,
      "grad_norm": 0.00359928491525352,
      "learning_rate": 0.10371468221164128,
      "loss": 0.2263,
      "num_input_tokens_seen": 46182912,
      "step": 23995
    },
    {
      "epoch": 3.9152459417570764,
      "grad_norm": 0.003409268334507942,
      "learning_rate": 0.10365865535732706,
      "loss": 0.2581,
      "num_input_tokens_seen": 46192656,
      "step": 24000
    },
    {
      "epoch": 3.9152459417570764,
      "eval_loss": 0.2037983536720276,
      "eval_runtime": 68.0679,
      "eval_samples_per_second": 40.034,
      "eval_steps_per_second": 20.024,
      "num_input_tokens_seen": 46192656,
      "step": 24000
    },
    {
      "epoch": 3.9160616689778935,
      "grad_norm": 0.005137925501912832,
      "learning_rate": 0.10360263564943062,
      "loss": 0.1724,
      "num_input_tokens_seen": 46201376,
      "step": 24005
    },
    {
      "epoch": 3.916877396198711,
      "grad_norm": 0.002781091956421733,
      "learning_rate": 0.10354662309659075,
      "loss": 0.2156,
      "num_input_tokens_seen": 46211728,
      "step": 24010
    },
    {
      "epoch": 3.9176931234195287,
      "grad_norm": 0.002886237343773246,
      "learning_rate": 0.10349061770744537,
      "loss": 0.2208,
      "num_input_tokens_seen": 46219136,
      "step": 24015
    },
    {
      "epoch": 3.918508850640346,
      "grad_norm": 0.0038962778635323048,
      "learning_rate": 0.10343461949063128,
      "loss": 0.2148,
      "num_input_tokens_seen": 46228384,
      "step": 24020
    },
    {
      "epoch": 3.919324577861163,
      "grad_norm": 0.0017872232710942626,
      "learning_rate": 0.103378628454784,
      "loss": 0.2203,
      "num_input_tokens_seen": 46238288,
      "step": 24025
    },
    {
      "epoch": 3.9201403050819805,
      "grad_norm": 0.0020939381793141365,
      "learning_rate": 0.10332264460853811,
      "loss": 0.1725,
      "num_input_tokens_seen": 46248096,
      "step": 24030
    },
    {
      "epoch": 3.920956032302798,
      "grad_norm": 0.0020723093766719103,
      "learning_rate": 0.10326666796052701,
      "loss": 0.2306,
      "num_input_tokens_seen": 46257248,
      "step": 24035
    },
    {
      "epoch": 3.9217717595236152,
      "grad_norm": 0.0032699734438210726,
      "learning_rate": 0.10321069851938296,
      "loss": 0.1849,
      "num_input_tokens_seen": 46267296,
      "step": 24040
    },
    {
      "epoch": 3.922587486744433,
      "grad_norm": 0.002659143880009651,
      "learning_rate": 0.10315473629373724,
      "loss": 0.1856,
      "num_input_tokens_seen": 46276000,
      "step": 24045
    },
    {
      "epoch": 3.92340321396525,
      "grad_norm": 0.0014252729015424848,
      "learning_rate": 0.10309878129221982,
      "loss": 0.2428,
      "num_input_tokens_seen": 46286080,
      "step": 24050
    },
    {
      "epoch": 3.9242189411860675,
      "grad_norm": 0.001124174683354795,
      "learning_rate": 0.10304283352345973,
      "loss": 0.2116,
      "num_input_tokens_seen": 46295824,
      "step": 24055
    },
    {
      "epoch": 3.9250346684068846,
      "grad_norm": 0.0021714314352720976,
      "learning_rate": 0.10298689299608486,
      "loss": 0.1662,
      "num_input_tokens_seen": 46305376,
      "step": 24060
    },
    {
      "epoch": 3.925850395627702,
      "grad_norm": 0.0018478414276614785,
      "learning_rate": 0.10293095971872188,
      "loss": 0.1494,
      "num_input_tokens_seen": 46315104,
      "step": 24065
    },
    {
      "epoch": 3.9266661228485193,
      "grad_norm": 0.002425743965432048,
      "learning_rate": 0.10287503369999645,
      "loss": 0.1939,
      "num_input_tokens_seen": 46324624,
      "step": 24070
    },
    {
      "epoch": 3.927481850069337,
      "grad_norm": 0.0021441553253680468,
      "learning_rate": 0.10281911494853295,
      "loss": 0.2178,
      "num_input_tokens_seen": 46334080,
      "step": 24075
    },
    {
      "epoch": 3.928297577290154,
      "grad_norm": 0.003407212905585766,
      "learning_rate": 0.10276320347295485,
      "loss": 0.1861,
      "num_input_tokens_seen": 46344608,
      "step": 24080
    },
    {
      "epoch": 3.9291133045109716,
      "grad_norm": 0.0019376622512936592,
      "learning_rate": 0.10270729928188446,
      "loss": 0.1626,
      "num_input_tokens_seen": 46354288,
      "step": 24085
    },
    {
      "epoch": 3.9299290317317888,
      "grad_norm": 0.0025898050516843796,
      "learning_rate": 0.10265140238394276,
      "loss": 0.1651,
      "num_input_tokens_seen": 46363984,
      "step": 24090
    },
    {
      "epoch": 3.9307447589526063,
      "grad_norm": 0.002686057472601533,
      "learning_rate": 0.10259551278774988,
      "loss": 0.2124,
      "num_input_tokens_seen": 46374208,
      "step": 24095
    },
    {
      "epoch": 3.9315604861734235,
      "grad_norm": 0.0022835012059658766,
      "learning_rate": 0.10253963050192462,
      "loss": 0.2276,
      "num_input_tokens_seen": 46384544,
      "step": 24100
    },
    {
      "epoch": 3.932376213394241,
      "grad_norm": 0.005996425170451403,
      "learning_rate": 0.10248375553508478,
      "loss": 0.1957,
      "num_input_tokens_seen": 46394592,
      "step": 24105
    },
    {
      "epoch": 3.933191940615058,
      "grad_norm": 0.0024363549891859293,
      "learning_rate": 0.102427887895847,
      "loss": 0.178,
      "num_input_tokens_seen": 46404864,
      "step": 24110
    },
    {
      "epoch": 3.9340076678358757,
      "grad_norm": 0.002734558191150427,
      "learning_rate": 0.10237202759282668,
      "loss": 0.241,
      "num_input_tokens_seen": 46415312,
      "step": 24115
    },
    {
      "epoch": 3.934823395056693,
      "grad_norm": 0.0016643339768052101,
      "learning_rate": 0.10231617463463821,
      "loss": 0.1941,
      "num_input_tokens_seen": 46425360,
      "step": 24120
    },
    {
      "epoch": 3.9356391222775104,
      "grad_norm": 0.0038419936317950487,
      "learning_rate": 0.10226032902989492,
      "loss": 0.1982,
      "num_input_tokens_seen": 46434960,
      "step": 24125
    },
    {
      "epoch": 3.936454849498328,
      "grad_norm": 0.0009539120364934206,
      "learning_rate": 0.10220449078720877,
      "loss": 0.1973,
      "num_input_tokens_seen": 46445424,
      "step": 24130
    },
    {
      "epoch": 3.937270576719145,
      "grad_norm": 0.002415733179077506,
      "learning_rate": 0.1021486599151908,
      "loss": 0.194,
      "num_input_tokens_seen": 46455504,
      "step": 24135
    },
    {
      "epoch": 3.9380863039399623,
      "grad_norm": 0.0015449980273842812,
      "learning_rate": 0.10209283642245084,
      "loss": 0.2,
      "num_input_tokens_seen": 46463696,
      "step": 24140
    },
    {
      "epoch": 3.93890203116078,
      "grad_norm": 0.0013518253108486533,
      "learning_rate": 0.10203702031759748,
      "loss": 0.19,
      "num_input_tokens_seen": 46473424,
      "step": 24145
    },
    {
      "epoch": 3.9397177583815974,
      "grad_norm": 0.0025128866545856,
      "learning_rate": 0.1019812116092384,
      "loss": 0.1564,
      "num_input_tokens_seen": 46482768,
      "step": 24150
    },
    {
      "epoch": 3.9405334856024146,
      "grad_norm": 0.0033863321878015995,
      "learning_rate": 0.10192541030597986,
      "loss": 0.2144,
      "num_input_tokens_seen": 46491216,
      "step": 24155
    },
    {
      "epoch": 3.9413492128232317,
      "grad_norm": 0.003338270355015993,
      "learning_rate": 0.1018696164164272,
      "loss": 0.2637,
      "num_input_tokens_seen": 46501120,
      "step": 24160
    },
    {
      "epoch": 3.9421649400440493,
      "grad_norm": 0.001527066808193922,
      "learning_rate": 0.10181382994918459,
      "loss": 0.1867,
      "num_input_tokens_seen": 46512080,
      "step": 24165
    },
    {
      "epoch": 3.942980667264867,
      "grad_norm": 0.0025218185037374496,
      "learning_rate": 0.10175805091285492,
      "loss": 0.1966,
      "num_input_tokens_seen": 46520960,
      "step": 24170
    },
    {
      "epoch": 3.943796394485684,
      "grad_norm": 0.001801245380192995,
      "learning_rate": 0.10170227931603999,
      "loss": 0.1808,
      "num_input_tokens_seen": 46531184,
      "step": 24175
    },
    {
      "epoch": 3.944612121706501,
      "grad_norm": 0.0036101029254496098,
      "learning_rate": 0.10164651516734062,
      "loss": 0.1783,
      "num_input_tokens_seen": 46541008,
      "step": 24180
    },
    {
      "epoch": 3.9454278489273187,
      "grad_norm": 0.002025010297074914,
      "learning_rate": 0.1015907584753562,
      "loss": 0.2273,
      "num_input_tokens_seen": 46550832,
      "step": 24185
    },
    {
      "epoch": 3.9462435761481363,
      "grad_norm": 0.001863646786659956,
      "learning_rate": 0.10153500924868523,
      "loss": 0.2532,
      "num_input_tokens_seen": 46560576,
      "step": 24190
    },
    {
      "epoch": 3.9470593033689534,
      "grad_norm": 0.003649457823485136,
      "learning_rate": 0.10147926749592483,
      "loss": 0.1754,
      "num_input_tokens_seen": 46569312,
      "step": 24195
    },
    {
      "epoch": 3.9478750305897705,
      "grad_norm": 0.001085809082724154,
      "learning_rate": 0.10142353322567112,
      "loss": 0.1759,
      "num_input_tokens_seen": 46576928,
      "step": 24200
    },
    {
      "epoch": 3.9478750305897705,
      "eval_loss": 0.20152698457241058,
      "eval_runtime": 67.9725,
      "eval_samples_per_second": 40.09,
      "eval_steps_per_second": 20.052,
      "num_input_tokens_seen": 46576928,
      "step": 24200
    },
    {
      "epoch": 3.948690757810588,
      "grad_norm": 0.0030411379411816597,
      "learning_rate": 0.1013678064465191,
      "loss": 0.1979,
      "num_input_tokens_seen": 46588144,
      "step": 24205
    },
    {
      "epoch": 3.9495064850314057,
      "grad_norm": 0.001242815051227808,
      "learning_rate": 0.10131208716706244,
      "loss": 0.1978,
      "num_input_tokens_seen": 46597680,
      "step": 24210
    },
    {
      "epoch": 3.950322212252223,
      "grad_norm": 0.0014989019837230444,
      "learning_rate": 0.10125637539589379,
      "loss": 0.2132,
      "num_input_tokens_seen": 46608272,
      "step": 24215
    },
    {
      "epoch": 3.9511379394730404,
      "grad_norm": 0.0029587061144411564,
      "learning_rate": 0.10120067114160464,
      "loss": 0.2266,
      "num_input_tokens_seen": 46618416,
      "step": 24220
    },
    {
      "epoch": 3.9519536666938575,
      "grad_norm": 0.0024672767613083124,
      "learning_rate": 0.10114497441278517,
      "loss": 0.1916,
      "num_input_tokens_seen": 46627824,
      "step": 24225
    },
    {
      "epoch": 3.952769393914675,
      "grad_norm": 0.0018656172323971987,
      "learning_rate": 0.10108928521802468,
      "loss": 0.2111,
      "num_input_tokens_seen": 46637680,
      "step": 24230
    },
    {
      "epoch": 3.953585121135492,
      "grad_norm": 0.0019540097564458847,
      "learning_rate": 0.101033603565911,
      "loss": 0.133,
      "num_input_tokens_seen": 46647056,
      "step": 24235
    },
    {
      "epoch": 3.95440084835631,
      "grad_norm": 0.004186154343187809,
      "learning_rate": 0.10097792946503102,
      "loss": 0.2097,
      "num_input_tokens_seen": 46656672,
      "step": 24240
    },
    {
      "epoch": 3.955216575577127,
      "grad_norm": 0.0028378926217556,
      "learning_rate": 0.10092226292397039,
      "loss": 0.2783,
      "num_input_tokens_seen": 46666848,
      "step": 24245
    },
    {
      "epoch": 3.9560323027979445,
      "grad_norm": 0.002474416745826602,
      "learning_rate": 0.10086660395131354,
      "loss": 0.2255,
      "num_input_tokens_seen": 46676192,
      "step": 24250
    },
    {
      "epoch": 3.9568480300187616,
      "grad_norm": 0.0022616267669945955,
      "learning_rate": 0.10081095255564385,
      "loss": 0.1504,
      "num_input_tokens_seen": 46685216,
      "step": 24255
    },
    {
      "epoch": 3.957663757239579,
      "grad_norm": 0.0012698391219601035,
      "learning_rate": 0.10075530874554335,
      "loss": 0.1669,
      "num_input_tokens_seen": 46694352,
      "step": 24260
    },
    {
      "epoch": 3.9584794844603963,
      "grad_norm": 0.003011077642440796,
      "learning_rate": 0.10069967252959311,
      "loss": 0.1858,
      "num_input_tokens_seen": 46704416,
      "step": 24265
    },
    {
      "epoch": 3.959295211681214,
      "grad_norm": 0.0035729387309402227,
      "learning_rate": 0.10064404391637297,
      "loss": 0.1897,
      "num_input_tokens_seen": 46713744,
      "step": 24270
    },
    {
      "epoch": 3.960110938902031,
      "grad_norm": 0.0033979169093072414,
      "learning_rate": 0.10058842291446145,
      "loss": 0.1845,
      "num_input_tokens_seen": 46722896,
      "step": 24275
    },
    {
      "epoch": 3.9609266661228486,
      "grad_norm": 0.0039266059175133705,
      "learning_rate": 0.10053280953243608,
      "loss": 0.2565,
      "num_input_tokens_seen": 46732320,
      "step": 24280
    },
    {
      "epoch": 3.9617423933436657,
      "grad_norm": 0.002265565562993288,
      "learning_rate": 0.10047720377887315,
      "loss": 0.1887,
      "num_input_tokens_seen": 46741504,
      "step": 24285
    },
    {
      "epoch": 3.9625581205644833,
      "grad_norm": 0.00336926500312984,
      "learning_rate": 0.10042160566234767,
      "loss": 0.2775,
      "num_input_tokens_seen": 46750656,
      "step": 24290
    },
    {
      "epoch": 3.9633738477853004,
      "grad_norm": 0.0019238792592659593,
      "learning_rate": 0.10036601519143372,
      "loss": 0.1952,
      "num_input_tokens_seen": 46760608,
      "step": 24295
    },
    {
      "epoch": 3.964189575006118,
      "grad_norm": 0.0018382557900622487,
      "learning_rate": 0.1003104323747039,
      "loss": 0.1568,
      "num_input_tokens_seen": 46770288,
      "step": 24300
    },
    {
      "epoch": 3.965005302226935,
      "grad_norm": 0.00214407779276371,
      "learning_rate": 0.10025485722072984,
      "loss": 0.2253,
      "num_input_tokens_seen": 46779024,
      "step": 24305
    },
    {
      "epoch": 3.9658210294477527,
      "grad_norm": 0.0023201541043817997,
      "learning_rate": 0.10019928973808201,
      "loss": 0.2183,
      "num_input_tokens_seen": 46789648,
      "step": 24310
    },
    {
      "epoch": 3.96663675666857,
      "grad_norm": 0.002651042537763715,
      "learning_rate": 0.10014372993532945,
      "loss": 0.201,
      "num_input_tokens_seen": 46800224,
      "step": 24315
    },
    {
      "epoch": 3.9674524838893874,
      "grad_norm": 0.003302166936919093,
      "learning_rate": 0.1000881778210403,
      "loss": 0.2042,
      "num_input_tokens_seen": 46809264,
      "step": 24320
    },
    {
      "epoch": 3.968268211110205,
      "grad_norm": 0.0022262781858444214,
      "learning_rate": 0.10003263340378142,
      "loss": 0.1934,
      "num_input_tokens_seen": 46816656,
      "step": 24325
    },
    {
      "epoch": 3.969083938331022,
      "grad_norm": 0.0021145138889551163,
      "learning_rate": 0.09997709669211834,
      "loss": 0.1976,
      "num_input_tokens_seen": 46828224,
      "step": 24330
    },
    {
      "epoch": 3.9698996655518393,
      "grad_norm": 0.00186373689211905,
      "learning_rate": 0.0999215676946156,
      "loss": 0.1864,
      "num_input_tokens_seen": 46838032,
      "step": 24335
    },
    {
      "epoch": 3.970715392772657,
      "grad_norm": 0.0030393176712095737,
      "learning_rate": 0.0998660464198364,
      "loss": 0.2286,
      "num_input_tokens_seen": 46847696,
      "step": 24340
    },
    {
      "epoch": 3.9715311199934744,
      "grad_norm": 0.00226524006575346,
      "learning_rate": 0.09981053287634288,
      "loss": 0.192,
      "num_input_tokens_seen": 46857232,
      "step": 24345
    },
    {
      "epoch": 3.9723468472142915,
      "grad_norm": 0.0011894418857991695,
      "learning_rate": 0.09975502707269596,
      "loss": 0.1986,
      "num_input_tokens_seen": 46867040,
      "step": 24350
    },
    {
      "epoch": 3.9731625744351087,
      "grad_norm": 0.0021734866313636303,
      "learning_rate": 0.09969952901745524,
      "loss": 0.1868,
      "num_input_tokens_seen": 46876704,
      "step": 24355
    },
    {
      "epoch": 3.9739783016559262,
      "grad_norm": 0.0013716669054701924,
      "learning_rate": 0.09964403871917925,
      "loss": 0.2549,
      "num_input_tokens_seen": 46887216,
      "step": 24360
    },
    {
      "epoch": 3.974794028876744,
      "grad_norm": 0.0019068596884608269,
      "learning_rate": 0.09958855618642536,
      "loss": 0.1925,
      "num_input_tokens_seen": 46896288,
      "step": 24365
    },
    {
      "epoch": 3.975609756097561,
      "grad_norm": 0.0024982222821563482,
      "learning_rate": 0.09953308142774955,
      "loss": 0.2279,
      "num_input_tokens_seen": 46906176,
      "step": 24370
    },
    {
      "epoch": 3.976425483318378,
      "grad_norm": 0.0022947806864976883,
      "learning_rate": 0.09947761445170686,
      "loss": 0.2414,
      "num_input_tokens_seen": 46916768,
      "step": 24375
    },
    {
      "epoch": 3.9772412105391957,
      "grad_norm": 0.0016765529289841652,
      "learning_rate": 0.09942215526685086,
      "loss": 0.2164,
      "num_input_tokens_seen": 46926224,
      "step": 24380
    },
    {
      "epoch": 3.9780569377600132,
      "grad_norm": 0.0024430681951344013,
      "learning_rate": 0.09936670388173414,
      "loss": 0.1594,
      "num_input_tokens_seen": 46935280,
      "step": 24385
    },
    {
      "epoch": 3.9788726649808304,
      "grad_norm": 0.0015544280176982284,
      "learning_rate": 0.09931126030490799,
      "loss": 0.21,
      "num_input_tokens_seen": 46945616,
      "step": 24390
    },
    {
      "epoch": 3.9796883922016475,
      "grad_norm": 0.0010785548947751522,
      "learning_rate": 0.0992558245449225,
      "loss": 0.1863,
      "num_input_tokens_seen": 46955920,
      "step": 24395
    },
    {
      "epoch": 3.980504119422465,
      "grad_norm": 0.0011158707784488797,
      "learning_rate": 0.09920039661032651,
      "loss": 0.1628,
      "num_input_tokens_seen": 46965120,
      "step": 24400
    },
    {
      "epoch": 3.980504119422465,
      "eval_loss": 0.20062856376171112,
      "eval_runtime": 68.0815,
      "eval_samples_per_second": 40.026,
      "eval_steps_per_second": 20.02,
      "num_input_tokens_seen": 46965120,
      "step": 24400
    },
    {
      "epoch": 3.9813198466432826,
      "grad_norm": 0.0024260482750833035,
      "learning_rate": 0.09914497650966782,
      "loss": 0.2315,
      "num_input_tokens_seen": 46976032,
      "step": 24405
    },
    {
      "epoch": 3.9821355738640998,
      "grad_norm": 0.002165354322642088,
      "learning_rate": 0.09908956425149276,
      "loss": 0.1941,
      "num_input_tokens_seen": 46984528,
      "step": 24410
    },
    {
      "epoch": 3.9829513010849174,
      "grad_norm": 0.0016029521357268095,
      "learning_rate": 0.09903415984434677,
      "loss": 0.1855,
      "num_input_tokens_seen": 46993936,
      "step": 24415
    },
    {
      "epoch": 3.9837670283057345,
      "grad_norm": 0.0028175662737339735,
      "learning_rate": 0.09897876329677373,
      "loss": 0.1966,
      "num_input_tokens_seen": 47003776,
      "step": 24420
    },
    {
      "epoch": 3.984582755526552,
      "grad_norm": 0.0031644657719880342,
      "learning_rate": 0.09892337461731658,
      "loss": 0.2362,
      "num_input_tokens_seen": 47013088,
      "step": 24425
    },
    {
      "epoch": 3.985398482747369,
      "grad_norm": 0.0022177803330123425,
      "learning_rate": 0.09886799381451693,
      "loss": 0.2067,
      "num_input_tokens_seen": 47022256,
      "step": 24430
    },
    {
      "epoch": 3.9862142099681868,
      "grad_norm": 0.002059142105281353,
      "learning_rate": 0.09881262089691521,
      "loss": 0.2016,
      "num_input_tokens_seen": 47031648,
      "step": 24435
    },
    {
      "epoch": 3.987029937189004,
      "grad_norm": 0.00157170498277992,
      "learning_rate": 0.09875725587305059,
      "loss": 0.2265,
      "num_input_tokens_seen": 47041984,
      "step": 24440
    },
    {
      "epoch": 3.9878456644098215,
      "grad_norm": 0.0025889836251735687,
      "learning_rate": 0.09870189875146111,
      "loss": 0.1895,
      "num_input_tokens_seen": 47051264,
      "step": 24445
    },
    {
      "epoch": 3.9886613916306386,
      "grad_norm": 0.0018637310713529587,
      "learning_rate": 0.09864654954068346,
      "loss": 0.193,
      "num_input_tokens_seen": 47060240,
      "step": 24450
    },
    {
      "epoch": 3.989477118851456,
      "grad_norm": 0.001867369282990694,
      "learning_rate": 0.09859120824925326,
      "loss": 0.1867,
      "num_input_tokens_seen": 47069392,
      "step": 24455
    },
    {
      "epoch": 3.9902928460722733,
      "grad_norm": 0.0020237017888575792,
      "learning_rate": 0.09853587488570474,
      "loss": 0.1899,
      "num_input_tokens_seen": 47078688,
      "step": 24460
    },
    {
      "epoch": 3.991108573293091,
      "grad_norm": 0.002356564626097679,
      "learning_rate": 0.09848054945857107,
      "loss": 0.1874,
      "num_input_tokens_seen": 47088400,
      "step": 24465
    },
    {
      "epoch": 3.991924300513908,
      "grad_norm": 0.0025364537723362446,
      "learning_rate": 0.09842523197638416,
      "loss": 0.2283,
      "num_input_tokens_seen": 47098304,
      "step": 24470
    },
    {
      "epoch": 3.9927400277347256,
      "grad_norm": 0.00221051718108356,
      "learning_rate": 0.09836992244767452,
      "loss": 0.2104,
      "num_input_tokens_seen": 47107872,
      "step": 24475
    },
    {
      "epoch": 3.9935557549555427,
      "grad_norm": 0.0013471308629959822,
      "learning_rate": 0.09831462088097168,
      "loss": 0.1444,
      "num_input_tokens_seen": 47117664,
      "step": 24480
    },
    {
      "epoch": 3.9943714821763603,
      "grad_norm": 0.004233002662658691,
      "learning_rate": 0.09825932728480385,
      "loss": 0.2121,
      "num_input_tokens_seen": 47126080,
      "step": 24485
    },
    {
      "epoch": 3.9951872093971774,
      "grad_norm": 0.002062931889668107,
      "learning_rate": 0.09820404166769794,
      "loss": 0.1583,
      "num_input_tokens_seen": 47135600,
      "step": 24490
    },
    {
      "epoch": 3.996002936617995,
      "grad_norm": 0.002604201901704073,
      "learning_rate": 0.09814876403817978,
      "loss": 0.1979,
      "num_input_tokens_seen": 47145696,
      "step": 24495
    },
    {
      "epoch": 3.9968186638388126,
      "grad_norm": 0.0027880410198122263,
      "learning_rate": 0.09809349440477376,
      "loss": 0.2293,
      "num_input_tokens_seen": 47156032,
      "step": 24500
    },
    {
      "epoch": 3.9976343910596297,
      "grad_norm": 0.0023816481698304415,
      "learning_rate": 0.09803823277600317,
      "loss": 0.1985,
      "num_input_tokens_seen": 47165424,
      "step": 24505
    },
    {
      "epoch": 3.998450118280447,
      "grad_norm": 0.0019407873041927814,
      "learning_rate": 0.09798297916039014,
      "loss": 0.2106,
      "num_input_tokens_seen": 47174544,
      "step": 24510
    },
    {
      "epoch": 3.9992658455012644,
      "grad_norm": 0.0029249282088130713,
      "learning_rate": 0.09792773356645534,
      "loss": 0.1884,
      "num_input_tokens_seen": 47184320,
      "step": 24515
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.002250638324767351,
      "learning_rate": 0.09787249600271843,
      "loss": 0.1907,
      "num_input_tokens_seen": 47193024,
      "step": 24520
    },
    {
      "epoch": 4.000815727220817,
      "grad_norm": 0.0017905719578266144,
      "learning_rate": 0.09781726647769776,
      "loss": 0.219,
      "num_input_tokens_seen": 47203712,
      "step": 24525
    },
    {
      "epoch": 4.001631454441635,
      "grad_norm": 0.0024005586747080088,
      "learning_rate": 0.0977620449999103,
      "loss": 0.207,
      "num_input_tokens_seen": 47214032,
      "step": 24530
    },
    {
      "epoch": 4.002447181662452,
      "grad_norm": 0.0014588871272280812,
      "learning_rate": 0.09770683157787204,
      "loss": 0.2078,
      "num_input_tokens_seen": 47221840,
      "step": 24535
    },
    {
      "epoch": 4.003262908883269,
      "grad_norm": 0.002782401628792286,
      "learning_rate": 0.09765162622009745,
      "loss": 0.1863,
      "num_input_tokens_seen": 47229776,
      "step": 24540
    },
    {
      "epoch": 4.0040786361040865,
      "grad_norm": 0.0015626177191734314,
      "learning_rate": 0.09759642893509995,
      "loss": 0.2182,
      "num_input_tokens_seen": 47240128,
      "step": 24545
    },
    {
      "epoch": 4.004894363324905,
      "grad_norm": 0.0012331680627539754,
      "learning_rate": 0.09754123973139169,
      "loss": 0.1931,
      "num_input_tokens_seen": 47250112,
      "step": 24550
    },
    {
      "epoch": 4.005710090545722,
      "grad_norm": 0.002305392874404788,
      "learning_rate": 0.09748605861748345,
      "loss": 0.191,
      "num_input_tokens_seen": 47259568,
      "step": 24555
    },
    {
      "epoch": 4.006525817766539,
      "grad_norm": 0.00367857632227242,
      "learning_rate": 0.0974308856018849,
      "loss": 0.1918,
      "num_input_tokens_seen": 47268240,
      "step": 24560
    },
    {
      "epoch": 4.007341544987356,
      "grad_norm": 0.0018360043177381158,
      "learning_rate": 0.09737572069310449,
      "loss": 0.2065,
      "num_input_tokens_seen": 47277920,
      "step": 24565
    },
    {
      "epoch": 4.008157272208174,
      "grad_norm": 0.0022312358487397432,
      "learning_rate": 0.09732056389964922,
      "loss": 0.2144,
      "num_input_tokens_seen": 47288704,
      "step": 24570
    },
    {
      "epoch": 4.008972999428991,
      "grad_norm": 0.002056217286735773,
      "learning_rate": 0.097265415230025,
      "loss": 0.203,
      "num_input_tokens_seen": 47299136,
      "step": 24575
    },
    {
      "epoch": 4.009788726649808,
      "grad_norm": 0.0014626536285504699,
      "learning_rate": 0.09721027469273648,
      "loss": 0.1671,
      "num_input_tokens_seen": 47308784,
      "step": 24580
    },
    {
      "epoch": 4.010604453870625,
      "grad_norm": 0.0010801558382809162,
      "learning_rate": 0.09715514229628695,
      "loss": 0.1957,
      "num_input_tokens_seen": 47318864,
      "step": 24585
    },
    {
      "epoch": 4.011420181091443,
      "grad_norm": 0.0027532908134162426,
      "learning_rate": 0.09710001804917864,
      "loss": 0.207,
      "num_input_tokens_seen": 47327840,
      "step": 24590
    },
    {
      "epoch": 4.0122359083122605,
      "grad_norm": 0.0022887804079800844,
      "learning_rate": 0.09704490195991226,
      "loss": 0.2221,
      "num_input_tokens_seen": 47337280,
      "step": 24595
    },
    {
      "epoch": 4.013051635533078,
      "grad_norm": 0.003204970620572567,
      "learning_rate": 0.09698979403698753,
      "loss": 0.2243,
      "num_input_tokens_seen": 47347920,
      "step": 24600
    },
    {
      "epoch": 4.013051635533078,
      "eval_loss": 0.20168572664260864,
      "eval_runtime": 68.046,
      "eval_samples_per_second": 40.046,
      "eval_steps_per_second": 20.031,
      "num_input_tokens_seen": 47347920,
      "step": 24600
    },
    {
      "epoch": 4.013867362753895,
      "grad_norm": 0.0028559935744851828,
      "learning_rate": 0.0969346942889027,
      "loss": 0.2238,
      "num_input_tokens_seen": 47358416,
      "step": 24605
    },
    {
      "epoch": 4.014683089974713,
      "grad_norm": 0.006196482107043266,
      "learning_rate": 0.09687960272415487,
      "loss": 0.2152,
      "num_input_tokens_seen": 47367632,
      "step": 24610
    },
    {
      "epoch": 4.01549881719553,
      "grad_norm": 0.0019362734165042639,
      "learning_rate": 0.0968245193512399,
      "loss": 0.1416,
      "num_input_tokens_seen": 47377488,
      "step": 24615
    },
    {
      "epoch": 4.016314544416347,
      "grad_norm": 0.0015306496061384678,
      "learning_rate": 0.09676944417865221,
      "loss": 0.1595,
      "num_input_tokens_seen": 47386464,
      "step": 24620
    },
    {
      "epoch": 4.017130271637164,
      "grad_norm": 0.002425111597403884,
      "learning_rate": 0.09671437721488517,
      "loss": 0.1971,
      "num_input_tokens_seen": 47395904,
      "step": 24625
    },
    {
      "epoch": 4.017945998857982,
      "grad_norm": 0.001529867178760469,
      "learning_rate": 0.09665931846843086,
      "loss": 0.1845,
      "num_input_tokens_seen": 47405760,
      "step": 24630
    },
    {
      "epoch": 4.018761726078799,
      "grad_norm": 0.0014269892126321793,
      "learning_rate": 0.0966042679477799,
      "loss": 0.1941,
      "num_input_tokens_seen": 47416688,
      "step": 24635
    },
    {
      "epoch": 4.0195774532996165,
      "grad_norm": 0.0023558444809168577,
      "learning_rate": 0.09654922566142186,
      "loss": 0.2303,
      "num_input_tokens_seen": 47427024,
      "step": 24640
    },
    {
      "epoch": 4.020393180520434,
      "grad_norm": 0.0011752257123589516,
      "learning_rate": 0.09649419161784498,
      "loss": 0.2167,
      "num_input_tokens_seen": 47436880,
      "step": 24645
    },
    {
      "epoch": 4.021208907741252,
      "grad_norm": 0.002374724019318819,
      "learning_rate": 0.09643916582553606,
      "loss": 0.1829,
      "num_input_tokens_seen": 47445840,
      "step": 24650
    },
    {
      "epoch": 4.022024634962069,
      "grad_norm": 0.0027346659917384386,
      "learning_rate": 0.09638414829298093,
      "loss": 0.2937,
      "num_input_tokens_seen": 47454832,
      "step": 24655
    },
    {
      "epoch": 4.022840362182886,
      "grad_norm": 0.0014644275652244687,
      "learning_rate": 0.09632913902866386,
      "loss": 0.1544,
      "num_input_tokens_seen": 47464496,
      "step": 24660
    },
    {
      "epoch": 4.023656089403703,
      "grad_norm": 0.00152347341645509,
      "learning_rate": 0.096274138041068,
      "loss": 0.166,
      "num_input_tokens_seen": 47474512,
      "step": 24665
    },
    {
      "epoch": 4.024471816624521,
      "grad_norm": 0.0019381894962862134,
      "learning_rate": 0.09621914533867527,
      "loss": 0.2108,
      "num_input_tokens_seen": 47484752,
      "step": 24670
    },
    {
      "epoch": 4.025287543845338,
      "grad_norm": 0.0014447914436459541,
      "learning_rate": 0.09616416092996616,
      "loss": 0.1613,
      "num_input_tokens_seen": 47494592,
      "step": 24675
    },
    {
      "epoch": 4.026103271066155,
      "grad_norm": 0.0021899817511439323,
      "learning_rate": 0.09610918482342,
      "loss": 0.1607,
      "num_input_tokens_seen": 47504128,
      "step": 24680
    },
    {
      "epoch": 4.026918998286972,
      "grad_norm": 0.002384913619607687,
      "learning_rate": 0.09605421702751478,
      "loss": 0.2068,
      "num_input_tokens_seen": 47514896,
      "step": 24685
    },
    {
      "epoch": 4.02773472550779,
      "grad_norm": 0.0025493446737527847,
      "learning_rate": 0.09599925755072718,
      "loss": 0.209,
      "num_input_tokens_seen": 47526160,
      "step": 24690
    },
    {
      "epoch": 4.028550452728608,
      "grad_norm": 0.0024333226028829813,
      "learning_rate": 0.09594430640153273,
      "loss": 0.1667,
      "num_input_tokens_seen": 47535744,
      "step": 24695
    },
    {
      "epoch": 4.029366179949425,
      "grad_norm": 0.002148900181055069,
      "learning_rate": 0.09588936358840547,
      "loss": 0.2046,
      "num_input_tokens_seen": 47545840,
      "step": 24700
    },
    {
      "epoch": 4.030181907170242,
      "grad_norm": 0.0029530466999858618,
      "learning_rate": 0.09583442911981836,
      "loss": 0.2508,
      "num_input_tokens_seen": 47555680,
      "step": 24705
    },
    {
      "epoch": 4.03099763439106,
      "grad_norm": 0.0020403650123625994,
      "learning_rate": 0.09577950300424302,
      "loss": 0.1915,
      "num_input_tokens_seen": 47565200,
      "step": 24710
    },
    {
      "epoch": 4.031813361611877,
      "grad_norm": 0.001635260763578117,
      "learning_rate": 0.09572458525014967,
      "loss": 0.1224,
      "num_input_tokens_seen": 47574112,
      "step": 24715
    },
    {
      "epoch": 4.032629088832694,
      "grad_norm": 0.0024884240701794624,
      "learning_rate": 0.0956696758660073,
      "loss": 0.1841,
      "num_input_tokens_seen": 47583664,
      "step": 24720
    },
    {
      "epoch": 4.033444816053512,
      "grad_norm": 0.003614294109866023,
      "learning_rate": 0.09561477486028373,
      "loss": 0.2176,
      "num_input_tokens_seen": 47592880,
      "step": 24725
    },
    {
      "epoch": 4.034260543274329,
      "grad_norm": 0.0021982891485095024,
      "learning_rate": 0.09555988224144528,
      "loss": 0.2077,
      "num_input_tokens_seen": 47601744,
      "step": 24730
    },
    {
      "epoch": 4.035076270495146,
      "grad_norm": 0.0021012318320572376,
      "learning_rate": 0.09550499801795717,
      "loss": 0.1746,
      "num_input_tokens_seen": 47612464,
      "step": 24735
    },
    {
      "epoch": 4.0358919977159635,
      "grad_norm": 0.0013432647101581097,
      "learning_rate": 0.09545012219828314,
      "loss": 0.1946,
      "num_input_tokens_seen": 47622368,
      "step": 24740
    },
    {
      "epoch": 4.0367077249367815,
      "grad_norm": 0.0019237027736380696,
      "learning_rate": 0.09539525479088577,
      "loss": 0.1938,
      "num_input_tokens_seen": 47631488,
      "step": 24745
    },
    {
      "epoch": 4.037523452157599,
      "grad_norm": 0.0024907165206968784,
      "learning_rate": 0.0953403958042264,
      "loss": 0.1908,
      "num_input_tokens_seen": 47642400,
      "step": 24750
    },
    {
      "epoch": 4.038339179378416,
      "grad_norm": 0.0020885413978248835,
      "learning_rate": 0.09528554524676484,
      "loss": 0.2088,
      "num_input_tokens_seen": 47650224,
      "step": 24755
    },
    {
      "epoch": 4.039154906599233,
      "grad_norm": 0.002295637736096978,
      "learning_rate": 0.09523070312695978,
      "loss": 0.2009,
      "num_input_tokens_seen": 47661280,
      "step": 24760
    },
    {
      "epoch": 4.039970633820051,
      "grad_norm": 0.002123464597389102,
      "learning_rate": 0.09517586945326863,
      "loss": 0.17,
      "num_input_tokens_seen": 47671152,
      "step": 24765
    },
    {
      "epoch": 4.040786361040868,
      "grad_norm": 0.0033887845929712057,
      "learning_rate": 0.0951210442341473,
      "loss": 0.2178,
      "num_input_tokens_seen": 47682336,
      "step": 24770
    },
    {
      "epoch": 4.041602088261685,
      "grad_norm": 0.004596555605530739,
      "learning_rate": 0.09506622747805066,
      "loss": 0.2082,
      "num_input_tokens_seen": 47691424,
      "step": 24775
    },
    {
      "epoch": 4.042417815482502,
      "grad_norm": 0.0024988993536680937,
      "learning_rate": 0.09501141919343203,
      "loss": 0.2031,
      "num_input_tokens_seen": 47700608,
      "step": 24780
    },
    {
      "epoch": 4.04323354270332,
      "grad_norm": 0.0024745562113821507,
      "learning_rate": 0.09495661938874361,
      "loss": 0.1828,
      "num_input_tokens_seen": 47710288,
      "step": 24785
    },
    {
      "epoch": 4.0440492699241375,
      "grad_norm": 0.0020676699932664633,
      "learning_rate": 0.0949018280724362,
      "loss": 0.186,
      "num_input_tokens_seen": 47720528,
      "step": 24790
    },
    {
      "epoch": 4.044864997144955,
      "grad_norm": 0.0021891165524721146,
      "learning_rate": 0.09484704525295934,
      "loss": 0.1864,
      "num_input_tokens_seen": 47730912,
      "step": 24795
    },
    {
      "epoch": 4.045680724365772,
      "grad_norm": 0.0035899158101528883,
      "learning_rate": 0.09479227093876112,
      "loss": 0.1911,
      "num_input_tokens_seen": 47741360,
      "step": 24800
    },
    {
      "epoch": 4.045680724365772,
      "eval_loss": 0.1990923434495926,
      "eval_runtime": 68.0204,
      "eval_samples_per_second": 40.061,
      "eval_steps_per_second": 20.038,
      "num_input_tokens_seen": 47741360,
      "step": 24800
    },
    {
      "epoch": 4.04649645158659,
      "grad_norm": 0.002711962442845106,
      "learning_rate": 0.0947375051382886,
      "loss": 0.2304,
      "num_input_tokens_seen": 47750032,
      "step": 24805
    },
    {
      "epoch": 4.047312178807407,
      "grad_norm": 0.001679121283814311,
      "learning_rate": 0.09468274785998718,
      "loss": 0.1467,
      "num_input_tokens_seen": 47759648,
      "step": 24810
    },
    {
      "epoch": 4.048127906028224,
      "grad_norm": 0.0015922648599371314,
      "learning_rate": 0.09462799911230127,
      "loss": 0.2028,
      "num_input_tokens_seen": 47769680,
      "step": 24815
    },
    {
      "epoch": 4.048943633249041,
      "grad_norm": 0.0027133366093039513,
      "learning_rate": 0.0945732589036737,
      "loss": 0.1969,
      "num_input_tokens_seen": 47780448,
      "step": 24820
    },
    {
      "epoch": 4.049759360469859,
      "grad_norm": 0.002467147074639797,
      "learning_rate": 0.09451852724254614,
      "loss": 0.161,
      "num_input_tokens_seen": 47788704,
      "step": 24825
    },
    {
      "epoch": 4.050575087690676,
      "grad_norm": 0.0012010714272037148,
      "learning_rate": 0.09446380413735894,
      "loss": 0.2364,
      "num_input_tokens_seen": 47799168,
      "step": 24830
    },
    {
      "epoch": 4.051390814911493,
      "grad_norm": 0.004214515443891287,
      "learning_rate": 0.09440908959655099,
      "loss": 0.2369,
      "num_input_tokens_seen": 47810176,
      "step": 24835
    },
    {
      "epoch": 4.052206542132311,
      "grad_norm": 0.0015532250981777906,
      "learning_rate": 0.09435438362856004,
      "loss": 0.2858,
      "num_input_tokens_seen": 47819728,
      "step": 24840
    },
    {
      "epoch": 4.053022269353129,
      "grad_norm": 0.0017211964586749673,
      "learning_rate": 0.0942996862418225,
      "loss": 0.1825,
      "num_input_tokens_seen": 47829264,
      "step": 24845
    },
    {
      "epoch": 4.053837996573946,
      "grad_norm": 0.0042256200686097145,
      "learning_rate": 0.09424499744477322,
      "loss": 0.2053,
      "num_input_tokens_seen": 47838288,
      "step": 24850
    },
    {
      "epoch": 4.054653723794763,
      "grad_norm": 0.0021816191729158163,
      "learning_rate": 0.09419031724584608,
      "loss": 0.1635,
      "num_input_tokens_seen": 47848000,
      "step": 24855
    },
    {
      "epoch": 4.05546945101558,
      "grad_norm": 0.002575488993898034,
      "learning_rate": 0.09413564565347331,
      "loss": 0.2349,
      "num_input_tokens_seen": 47858656,
      "step": 24860
    },
    {
      "epoch": 4.056285178236398,
      "grad_norm": 0.0024998129811137915,
      "learning_rate": 0.094080982676086,
      "loss": 0.1735,
      "num_input_tokens_seen": 47868592,
      "step": 24865
    },
    {
      "epoch": 4.057100905457215,
      "grad_norm": 0.0033183873165398836,
      "learning_rate": 0.09402632832211395,
      "loss": 0.2228,
      "num_input_tokens_seen": 47878112,
      "step": 24870
    },
    {
      "epoch": 4.057916632678032,
      "grad_norm": 0.002196670277044177,
      "learning_rate": 0.09397168259998541,
      "loss": 0.2042,
      "num_input_tokens_seen": 47888736,
      "step": 24875
    },
    {
      "epoch": 4.058732359898849,
      "grad_norm": 0.0015584141947329044,
      "learning_rate": 0.09391704551812759,
      "loss": 0.1632,
      "num_input_tokens_seen": 47897968,
      "step": 24880
    },
    {
      "epoch": 4.059548087119667,
      "grad_norm": 0.0019739733543246984,
      "learning_rate": 0.09386241708496605,
      "loss": 0.1743,
      "num_input_tokens_seen": 47906912,
      "step": 24885
    },
    {
      "epoch": 4.0603638143404845,
      "grad_norm": 0.0019530728459358215,
      "learning_rate": 0.09380779730892527,
      "loss": 0.2786,
      "num_input_tokens_seen": 47916656,
      "step": 24890
    },
    {
      "epoch": 4.061179541561302,
      "grad_norm": 0.0026267932262271643,
      "learning_rate": 0.09375318619842836,
      "loss": 0.2055,
      "num_input_tokens_seen": 47926688,
      "step": 24895
    },
    {
      "epoch": 4.06199526878212,
      "grad_norm": 0.0016827770741656423,
      "learning_rate": 0.09369858376189696,
      "loss": 0.2106,
      "num_input_tokens_seen": 47936656,
      "step": 24900
    },
    {
      "epoch": 4.062810996002937,
      "grad_norm": 0.0022855568677186966,
      "learning_rate": 0.09364399000775143,
      "loss": 0.1712,
      "num_input_tokens_seen": 47945760,
      "step": 24905
    },
    {
      "epoch": 4.063626723223754,
      "grad_norm": 0.0018978435546159744,
      "learning_rate": 0.09358940494441093,
      "loss": 0.1934,
      "num_input_tokens_seen": 47955328,
      "step": 24910
    },
    {
      "epoch": 4.064442450444571,
      "grad_norm": 0.0018302879761904478,
      "learning_rate": 0.09353482858029301,
      "loss": 0.2173,
      "num_input_tokens_seen": 47964512,
      "step": 24915
    },
    {
      "epoch": 4.065258177665389,
      "grad_norm": 0.0019397009164094925,
      "learning_rate": 0.09348026092381419,
      "loss": 0.2465,
      "num_input_tokens_seen": 47973600,
      "step": 24920
    },
    {
      "epoch": 4.066073904886206,
      "grad_norm": 0.0014116567326709628,
      "learning_rate": 0.09342570198338931,
      "loss": 0.1261,
      "num_input_tokens_seen": 47982256,
      "step": 24925
    },
    {
      "epoch": 4.066889632107023,
      "grad_norm": 0.004083470907062292,
      "learning_rate": 0.0933711517674322,
      "loss": 0.1806,
      "num_input_tokens_seen": 47992736,
      "step": 24930
    },
    {
      "epoch": 4.0677053593278405,
      "grad_norm": 0.002204306423664093,
      "learning_rate": 0.09331661028435513,
      "loss": 0.2049,
      "num_input_tokens_seen": 48002816,
      "step": 24935
    },
    {
      "epoch": 4.0685210865486585,
      "grad_norm": 0.00550994835793972,
      "learning_rate": 0.09326207754256909,
      "loss": 0.2121,
      "num_input_tokens_seen": 48011792,
      "step": 24940
    },
    {
      "epoch": 4.069336813769476,
      "grad_norm": 0.0026851457078009844,
      "learning_rate": 0.09320755355048366,
      "loss": 0.19,
      "num_input_tokens_seen": 48021296,
      "step": 24945
    },
    {
      "epoch": 4.070152540990293,
      "grad_norm": 0.003787572728469968,
      "learning_rate": 0.09315303831650722,
      "loss": 0.1603,
      "num_input_tokens_seen": 48030528,
      "step": 24950
    },
    {
      "epoch": 4.07096826821111,
      "grad_norm": 0.0017604457680135965,
      "learning_rate": 0.09309853184904661,
      "loss": 0.2441,
      "num_input_tokens_seen": 48041312,
      "step": 24955
    },
    {
      "epoch": 4.071783995431928,
      "grad_norm": 0.0031238216906785965,
      "learning_rate": 0.09304403415650753,
      "loss": 0.2016,
      "num_input_tokens_seen": 48051056,
      "step": 24960
    },
    {
      "epoch": 4.072599722652745,
      "grad_norm": 0.0017595182871446013,
      "learning_rate": 0.09298954524729405,
      "loss": 0.1916,
      "num_input_tokens_seen": 48060560,
      "step": 24965
    },
    {
      "epoch": 4.073415449873562,
      "grad_norm": 0.0029286888893693686,
      "learning_rate": 0.09293506512980916,
      "loss": 0.1649,
      "num_input_tokens_seen": 48070368,
      "step": 24970
    },
    {
      "epoch": 4.074231177094379,
      "grad_norm": 0.003794437739998102,
      "learning_rate": 0.0928805938124544,
      "loss": 0.2203,
      "num_input_tokens_seen": 48079968,
      "step": 24975
    },
    {
      "epoch": 4.075046904315197,
      "grad_norm": 0.0011127856560051441,
      "learning_rate": 0.09282613130362982,
      "loss": 0.1876,
      "num_input_tokens_seen": 48090000,
      "step": 24980
    },
    {
      "epoch": 4.0758626315360145,
      "grad_norm": 0.0027586265932768583,
      "learning_rate": 0.09277167761173427,
      "loss": 0.175,
      "num_input_tokens_seen": 48100672,
      "step": 24985
    },
    {
      "epoch": 4.076678358756832,
      "grad_norm": 0.0015858662081882358,
      "learning_rate": 0.0927172327451653,
      "loss": 0.1545,
      "num_input_tokens_seen": 48111088,
      "step": 24990
    },
    {
      "epoch": 4.077494085977649,
      "grad_norm": 0.0034362305887043476,
      "learning_rate": 0.09266279671231882,
      "loss": 0.17,
      "num_input_tokens_seen": 48121360,
      "step": 24995
    },
    {
      "epoch": 4.078309813198467,
      "grad_norm": 0.0018418926047161222,
      "learning_rate": 0.09260836952158967,
      "loss": 0.1695,
      "num_input_tokens_seen": 48131120,
      "step": 25000
    },
    {
      "epoch": 4.078309813198467,
      "eval_loss": 0.2028859257698059,
      "eval_runtime": 68.0268,
      "eval_samples_per_second": 40.058,
      "eval_steps_per_second": 20.036,
      "num_input_tokens_seen": 48131120,
      "step": 25000
    },
    {
      "epoch": 4.079125540419284,
      "grad_norm": 0.003845767118036747,
      "learning_rate": 0.09255395118137114,
      "loss": 0.1763,
      "num_input_tokens_seen": 48140752,
      "step": 25005
    },
    {
      "epoch": 4.079941267640101,
      "grad_norm": 0.0020014597102999687,
      "learning_rate": 0.09249954170005527,
      "loss": 0.1361,
      "num_input_tokens_seen": 48150208,
      "step": 25010
    },
    {
      "epoch": 4.080756994860918,
      "grad_norm": 0.004980428609997034,
      "learning_rate": 0.0924451410860327,
      "loss": 0.1567,
      "num_input_tokens_seen": 48160288,
      "step": 25015
    },
    {
      "epoch": 4.081572722081736,
      "grad_norm": 0.0029600381385535,
      "learning_rate": 0.09239074934769258,
      "loss": 0.1554,
      "num_input_tokens_seen": 48169744,
      "step": 25020
    },
    {
      "epoch": 4.082388449302553,
      "grad_norm": 0.003933183383196592,
      "learning_rate": 0.09233636649342288,
      "loss": 0.2011,
      "num_input_tokens_seen": 48178528,
      "step": 25025
    },
    {
      "epoch": 4.08320417652337,
      "grad_norm": 0.0026957793161273003,
      "learning_rate": 0.09228199253161017,
      "loss": 0.2469,
      "num_input_tokens_seen": 48187376,
      "step": 25030
    },
    {
      "epoch": 4.0840199037441876,
      "grad_norm": 0.0030625658109784126,
      "learning_rate": 0.09222762747063949,
      "loss": 0.1947,
      "num_input_tokens_seen": 48198176,
      "step": 25035
    },
    {
      "epoch": 4.084835630965006,
      "grad_norm": 0.004474468529224396,
      "learning_rate": 0.09217327131889473,
      "loss": 0.228,
      "num_input_tokens_seen": 48207760,
      "step": 25040
    },
    {
      "epoch": 4.085651358185823,
      "grad_norm": 0.003066872013732791,
      "learning_rate": 0.09211892408475818,
      "loss": 0.2313,
      "num_input_tokens_seen": 48217376,
      "step": 25045
    },
    {
      "epoch": 4.08646708540664,
      "grad_norm": 0.0021206068340688944,
      "learning_rate": 0.09206458577661089,
      "loss": 0.2065,
      "num_input_tokens_seen": 48226304,
      "step": 25050
    },
    {
      "epoch": 4.087282812627457,
      "grad_norm": 0.00220806198194623,
      "learning_rate": 0.09201025640283263,
      "loss": 0.168,
      "num_input_tokens_seen": 48235712,
      "step": 25055
    },
    {
      "epoch": 4.088098539848275,
      "grad_norm": 0.0026159328408539295,
      "learning_rate": 0.09195593597180148,
      "loss": 0.1719,
      "num_input_tokens_seen": 48245328,
      "step": 25060
    },
    {
      "epoch": 4.088914267069092,
      "grad_norm": 0.0027828277088701725,
      "learning_rate": 0.09190162449189444,
      "loss": 0.1663,
      "num_input_tokens_seen": 48255680,
      "step": 25065
    },
    {
      "epoch": 4.089729994289909,
      "grad_norm": 0.003734783036634326,
      "learning_rate": 0.09184732197148705,
      "loss": 0.2324,
      "num_input_tokens_seen": 48264272,
      "step": 25070
    },
    {
      "epoch": 4.090545721510727,
      "grad_norm": 0.0027920410502701998,
      "learning_rate": 0.09179302841895343,
      "loss": 0.1657,
      "num_input_tokens_seen": 48273600,
      "step": 25075
    },
    {
      "epoch": 4.091361448731544,
      "grad_norm": 0.0017669580411165953,
      "learning_rate": 0.09173874384266625,
      "loss": 0.1956,
      "num_input_tokens_seen": 48285040,
      "step": 25080
    },
    {
      "epoch": 4.0921771759523615,
      "grad_norm": 0.002457278547808528,
      "learning_rate": 0.09168446825099695,
      "loss": 0.1431,
      "num_input_tokens_seen": 48294784,
      "step": 25085
    },
    {
      "epoch": 4.092992903173179,
      "grad_norm": 0.002798590110614896,
      "learning_rate": 0.09163020165231545,
      "loss": 0.1819,
      "num_input_tokens_seen": 48304640,
      "step": 25090
    },
    {
      "epoch": 4.093808630393997,
      "grad_norm": 0.002106459578499198,
      "learning_rate": 0.09157594405499044,
      "loss": 0.2224,
      "num_input_tokens_seen": 48314976,
      "step": 25095
    },
    {
      "epoch": 4.094624357614814,
      "grad_norm": 0.0026728822849690914,
      "learning_rate": 0.09152169546738899,
      "loss": 0.1851,
      "num_input_tokens_seen": 48323808,
      "step": 25100
    },
    {
      "epoch": 4.095440084835631,
      "grad_norm": 0.0021521635353565216,
      "learning_rate": 0.09146745589787698,
      "loss": 0.2207,
      "num_input_tokens_seen": 48332832,
      "step": 25105
    },
    {
      "epoch": 4.096255812056448,
      "grad_norm": 0.002153879962861538,
      "learning_rate": 0.09141322535481891,
      "loss": 0.1656,
      "num_input_tokens_seen": 48341696,
      "step": 25110
    },
    {
      "epoch": 4.097071539277266,
      "grad_norm": 0.0030051975045353174,
      "learning_rate": 0.0913590038465777,
      "loss": 0.2354,
      "num_input_tokens_seen": 48350400,
      "step": 25115
    },
    {
      "epoch": 4.097887266498083,
      "grad_norm": 0.0029201407451182604,
      "learning_rate": 0.09130479138151505,
      "loss": 0.1335,
      "num_input_tokens_seen": 48361424,
      "step": 25120
    },
    {
      "epoch": 4.0987029937189,
      "grad_norm": 0.0028951659332960844,
      "learning_rate": 0.09125058796799114,
      "loss": 0.2213,
      "num_input_tokens_seen": 48370064,
      "step": 25125
    },
    {
      "epoch": 4.0995187209397175,
      "grad_norm": 0.004096047021448612,
      "learning_rate": 0.09119639361436485,
      "loss": 0.2419,
      "num_input_tokens_seen": 48380784,
      "step": 25130
    },
    {
      "epoch": 4.1003344481605355,
      "grad_norm": 0.002829260192811489,
      "learning_rate": 0.09114220832899368,
      "loss": 0.2462,
      "num_input_tokens_seen": 48390480,
      "step": 25135
    },
    {
      "epoch": 4.101150175381353,
      "grad_norm": 0.002605781424790621,
      "learning_rate": 0.0910880321202336,
      "loss": 0.1775,
      "num_input_tokens_seen": 48400720,
      "step": 25140
    },
    {
      "epoch": 4.10196590260217,
      "grad_norm": 0.0032927775755524635,
      "learning_rate": 0.09103386499643933,
      "loss": 0.2052,
      "num_input_tokens_seen": 48410208,
      "step": 25145
    },
    {
      "epoch": 4.102781629822987,
      "grad_norm": 0.0031777459662407637,
      "learning_rate": 0.09097970696596407,
      "loss": 0.1798,
      "num_input_tokens_seen": 48419776,
      "step": 25150
    },
    {
      "epoch": 4.103597357043805,
      "grad_norm": 0.002858187071979046,
      "learning_rate": 0.09092555803715971,
      "loss": 0.1934,
      "num_input_tokens_seen": 48428624,
      "step": 25155
    },
    {
      "epoch": 4.104413084264622,
      "grad_norm": 0.0017563881119713187,
      "learning_rate": 0.0908714182183767,
      "loss": 0.1505,
      "num_input_tokens_seen": 48437680,
      "step": 25160
    },
    {
      "epoch": 4.105228811485439,
      "grad_norm": 0.003686486044898629,
      "learning_rate": 0.090817287517964,
      "loss": 0.2098,
      "num_input_tokens_seen": 48446560,
      "step": 25165
    },
    {
      "epoch": 4.106044538706256,
      "grad_norm": 0.001638572895899415,
      "learning_rate": 0.09076316594426931,
      "loss": 0.1149,
      "num_input_tokens_seen": 48455328,
      "step": 25170
    },
    {
      "epoch": 4.106860265927074,
      "grad_norm": 0.004211008083075285,
      "learning_rate": 0.09070905350563888,
      "loss": 0.2646,
      "num_input_tokens_seen": 48464176,
      "step": 25175
    },
    {
      "epoch": 4.1076759931478914,
      "grad_norm": 0.0018048033816739917,
      "learning_rate": 0.09065495021041745,
      "loss": 0.2246,
      "num_input_tokens_seen": 48474096,
      "step": 25180
    },
    {
      "epoch": 4.108491720368709,
      "grad_norm": 0.0014765316154807806,
      "learning_rate": 0.09060085606694851,
      "loss": 0.1439,
      "num_input_tokens_seen": 48484144,
      "step": 25185
    },
    {
      "epoch": 4.109307447589526,
      "grad_norm": 0.00256058550439775,
      "learning_rate": 0.09054677108357405,
      "loss": 0.1749,
      "num_input_tokens_seen": 48493200,
      "step": 25190
    },
    {
      "epoch": 4.110123174810344,
      "grad_norm": 0.00124354159925133,
      "learning_rate": 0.09049269526863457,
      "loss": 0.2479,
      "num_input_tokens_seen": 48503104,
      "step": 25195
    },
    {
      "epoch": 4.110938902031161,
      "grad_norm": 0.0020795150194317102,
      "learning_rate": 0.09043862863046935,
      "loss": 0.1807,
      "num_input_tokens_seen": 48513200,
      "step": 25200
    },
    {
      "epoch": 4.110938902031161,
      "eval_loss": 0.19893963634967804,
      "eval_runtime": 67.8715,
      "eval_samples_per_second": 40.149,
      "eval_steps_per_second": 20.082,
      "num_input_tokens_seen": 48513200,
      "step": 25200
    },
    {
      "epoch": 4.111754629251978,
      "grad_norm": 0.0016844397177919745,
      "learning_rate": 0.09038457117741602,
      "loss": 0.1763,
      "num_input_tokens_seen": 48521232,
      "step": 25205
    },
    {
      "epoch": 4.112570356472795,
      "grad_norm": 0.003133587772026658,
      "learning_rate": 0.09033052291781099,
      "loss": 0.266,
      "num_input_tokens_seen": 48531776,
      "step": 25210
    },
    {
      "epoch": 4.113386083693613,
      "grad_norm": 0.001832063077017665,
      "learning_rate": 0.09027648385998926,
      "loss": 0.2147,
      "num_input_tokens_seen": 48539856,
      "step": 25215
    },
    {
      "epoch": 4.11420181091443,
      "grad_norm": 0.0024593290872871876,
      "learning_rate": 0.09022245401228417,
      "loss": 0.1704,
      "num_input_tokens_seen": 48550928,
      "step": 25220
    },
    {
      "epoch": 4.115017538135247,
      "grad_norm": 0.003960874862968922,
      "learning_rate": 0.09016843338302792,
      "loss": 0.1808,
      "num_input_tokens_seen": 48560000,
      "step": 25225
    },
    {
      "epoch": 4.1158332653560645,
      "grad_norm": 0.0026070980820804834,
      "learning_rate": 0.09011442198055115,
      "loss": 0.244,
      "num_input_tokens_seen": 48569328,
      "step": 25230
    },
    {
      "epoch": 4.1166489925768825,
      "grad_norm": 0.001429410302080214,
      "learning_rate": 0.09006041981318305,
      "loss": 0.1815,
      "num_input_tokens_seen": 48579296,
      "step": 25235
    },
    {
      "epoch": 4.1174647197977,
      "grad_norm": 0.002551940968260169,
      "learning_rate": 0.09000642688925149,
      "loss": 0.1508,
      "num_input_tokens_seen": 48589536,
      "step": 25240
    },
    {
      "epoch": 4.118280447018517,
      "grad_norm": 0.003235632088035345,
      "learning_rate": 0.0899524432170828,
      "loss": 0.1701,
      "num_input_tokens_seen": 48599552,
      "step": 25245
    },
    {
      "epoch": 4.119096174239334,
      "grad_norm": 0.0032508436124771833,
      "learning_rate": 0.08989846880500196,
      "loss": 0.2078,
      "num_input_tokens_seen": 48609776,
      "step": 25250
    },
    {
      "epoch": 4.119911901460152,
      "grad_norm": 0.0014922473346814513,
      "learning_rate": 0.08984450366133256,
      "loss": 0.1558,
      "num_input_tokens_seen": 48620016,
      "step": 25255
    },
    {
      "epoch": 4.120727628680969,
      "grad_norm": 0.0020729557145386934,
      "learning_rate": 0.08979054779439664,
      "loss": 0.1733,
      "num_input_tokens_seen": 48630576,
      "step": 25260
    },
    {
      "epoch": 4.121543355901786,
      "grad_norm": 0.0027053661178797483,
      "learning_rate": 0.08973660121251485,
      "loss": 0.2242,
      "num_input_tokens_seen": 48638896,
      "step": 25265
    },
    {
      "epoch": 4.122359083122603,
      "grad_norm": 0.0030172436963766813,
      "learning_rate": 0.08968266392400655,
      "loss": 0.1824,
      "num_input_tokens_seen": 48649088,
      "step": 25270
    },
    {
      "epoch": 4.123174810343421,
      "grad_norm": 0.0027123510371893644,
      "learning_rate": 0.0896287359371894,
      "loss": 0.1758,
      "num_input_tokens_seen": 48658000,
      "step": 25275
    },
    {
      "epoch": 4.1239905375642385,
      "grad_norm": 0.005584419704973698,
      "learning_rate": 0.08957481726037989,
      "loss": 0.1967,
      "num_input_tokens_seen": 48667216,
      "step": 25280
    },
    {
      "epoch": 4.124806264785056,
      "grad_norm": 0.0034338817931711674,
      "learning_rate": 0.08952090790189286,
      "loss": 0.2022,
      "num_input_tokens_seen": 48676240,
      "step": 25285
    },
    {
      "epoch": 4.125621992005874,
      "grad_norm": 0.004028587602078915,
      "learning_rate": 0.08946700787004187,
      "loss": 0.2086,
      "num_input_tokens_seen": 48685488,
      "step": 25290
    },
    {
      "epoch": 4.126437719226691,
      "grad_norm": 0.0029976097866892815,
      "learning_rate": 0.08941311717313899,
      "loss": 0.1696,
      "num_input_tokens_seen": 48695808,
      "step": 25295
    },
    {
      "epoch": 4.127253446447508,
      "grad_norm": 0.0028232818003743887,
      "learning_rate": 0.08935923581949483,
      "loss": 0.2151,
      "num_input_tokens_seen": 48705840,
      "step": 25300
    },
    {
      "epoch": 4.128069173668325,
      "grad_norm": 0.0010788291692733765,
      "learning_rate": 0.0893053638174185,
      "loss": 0.1745,
      "num_input_tokens_seen": 48716000,
      "step": 25305
    },
    {
      "epoch": 4.128884900889143,
      "grad_norm": 0.005161326844245195,
      "learning_rate": 0.0892515011752179,
      "loss": 0.2785,
      "num_input_tokens_seen": 48723920,
      "step": 25310
    },
    {
      "epoch": 4.12970062810996,
      "grad_norm": 0.002192999003455043,
      "learning_rate": 0.08919764790119918,
      "loss": 0.2379,
      "num_input_tokens_seen": 48733248,
      "step": 25315
    },
    {
      "epoch": 4.130516355330777,
      "grad_norm": 0.0017179918941110373,
      "learning_rate": 0.08914380400366727,
      "loss": 0.2105,
      "num_input_tokens_seen": 48743296,
      "step": 25320
    },
    {
      "epoch": 4.1313320825515945,
      "grad_norm": 0.003359486348927021,
      "learning_rate": 0.08908996949092551,
      "loss": 0.1895,
      "num_input_tokens_seen": 48752224,
      "step": 25325
    },
    {
      "epoch": 4.1321478097724125,
      "grad_norm": 0.0022404694464057684,
      "learning_rate": 0.08903614437127592,
      "loss": 0.2052,
      "num_input_tokens_seen": 48761152,
      "step": 25330
    },
    {
      "epoch": 4.13296353699323,
      "grad_norm": 0.002907851245254278,
      "learning_rate": 0.088982328653019,
      "loss": 0.1819,
      "num_input_tokens_seen": 48770768,
      "step": 25335
    },
    {
      "epoch": 4.133779264214047,
      "grad_norm": 0.0015616297023370862,
      "learning_rate": 0.0889285223444538,
      "loss": 0.2136,
      "num_input_tokens_seen": 48780496,
      "step": 25340
    },
    {
      "epoch": 4.134594991434864,
      "grad_norm": 0.002585160080343485,
      "learning_rate": 0.08887472545387787,
      "loss": 0.2052,
      "num_input_tokens_seen": 48789808,
      "step": 25345
    },
    {
      "epoch": 4.135410718655682,
      "grad_norm": 0.002316290745511651,
      "learning_rate": 0.08882093798958751,
      "loss": 0.1626,
      "num_input_tokens_seen": 48798976,
      "step": 25350
    },
    {
      "epoch": 4.136226445876499,
      "grad_norm": 0.001409118645824492,
      "learning_rate": 0.08876715995987726,
      "loss": 0.223,
      "num_input_tokens_seen": 48807808,
      "step": 25355
    },
    {
      "epoch": 4.137042173097316,
      "grad_norm": 0.003247019601985812,
      "learning_rate": 0.08871339137304052,
      "loss": 0.2154,
      "num_input_tokens_seen": 48816672,
      "step": 25360
    },
    {
      "epoch": 4.137857900318133,
      "grad_norm": 0.0046682958491146564,
      "learning_rate": 0.0886596322373689,
      "loss": 0.2175,
      "num_input_tokens_seen": 48826720,
      "step": 25365
    },
    {
      "epoch": 4.138673627538951,
      "grad_norm": 0.0026971285697072744,
      "learning_rate": 0.08860588256115293,
      "loss": 0.2703,
      "num_input_tokens_seen": 48836464,
      "step": 25370
    },
    {
      "epoch": 4.139489354759768,
      "grad_norm": 0.0028000683523714542,
      "learning_rate": 0.0885521423526814,
      "loss": 0.1741,
      "num_input_tokens_seen": 48845808,
      "step": 25375
    },
    {
      "epoch": 4.1403050819805856,
      "grad_norm": 0.002924466971307993,
      "learning_rate": 0.08849841162024165,
      "loss": 0.2043,
      "num_input_tokens_seen": 48855328,
      "step": 25380
    },
    {
      "epoch": 4.141120809201403,
      "grad_norm": 0.0013486172538250685,
      "learning_rate": 0.08844469037211973,
      "loss": 0.2065,
      "num_input_tokens_seen": 48865504,
      "step": 25385
    },
    {
      "epoch": 4.141936536422221,
      "grad_norm": 0.0027224079240113497,
      "learning_rate": 0.08839097861660014,
      "loss": 0.1637,
      "num_input_tokens_seen": 48874016,
      "step": 25390
    },
    {
      "epoch": 4.142752263643038,
      "grad_norm": 0.002869487041607499,
      "learning_rate": 0.08833727636196585,
      "loss": 0.224,
      "num_input_tokens_seen": 48884336,
      "step": 25395
    },
    {
      "epoch": 4.143567990863855,
      "grad_norm": 0.002383334329351783,
      "learning_rate": 0.08828358361649848,
      "loss": 0.1917,
      "num_input_tokens_seen": 48894496,
      "step": 25400
    },
    {
      "epoch": 4.143567990863855,
      "eval_loss": 0.20018358528614044,
      "eval_runtime": 68.0192,
      "eval_samples_per_second": 40.062,
      "eval_steps_per_second": 20.038,
      "num_input_tokens_seen": 48894496,
      "step": 25400
    },
    {
      "epoch": 4.144383718084672,
      "grad_norm": 0.0035938264336436987,
      "learning_rate": 0.08822990038847807,
      "loss": 0.194,
      "num_input_tokens_seen": 48903408,
      "step": 25405
    },
    {
      "epoch": 4.14519944530549,
      "grad_norm": 0.0016714254161342978,
      "learning_rate": 0.08817622668618325,
      "loss": 0.2136,
      "num_input_tokens_seen": 48912624,
      "step": 25410
    },
    {
      "epoch": 4.146015172526307,
      "grad_norm": 0.002180726733058691,
      "learning_rate": 0.08812256251789125,
      "loss": 0.1829,
      "num_input_tokens_seen": 48922544,
      "step": 25415
    },
    {
      "epoch": 4.146830899747124,
      "grad_norm": 0.0024493879172950983,
      "learning_rate": 0.08806890789187766,
      "loss": 0.1831,
      "num_input_tokens_seen": 48931984,
      "step": 25420
    },
    {
      "epoch": 4.1476466269679415,
      "grad_norm": 0.002116970019415021,
      "learning_rate": 0.08801526281641672,
      "loss": 0.2384,
      "num_input_tokens_seen": 48941872,
      "step": 25425
    },
    {
      "epoch": 4.1484623541887595,
      "grad_norm": 0.00137610943056643,
      "learning_rate": 0.0879616272997813,
      "loss": 0.1929,
      "num_input_tokens_seen": 48952576,
      "step": 25430
    },
    {
      "epoch": 4.149278081409577,
      "grad_norm": 0.002392690861597657,
      "learning_rate": 0.08790800135024247,
      "loss": 0.2316,
      "num_input_tokens_seen": 48962560,
      "step": 25435
    },
    {
      "epoch": 4.150093808630394,
      "grad_norm": 0.002643081359565258,
      "learning_rate": 0.08785438497607023,
      "loss": 0.1955,
      "num_input_tokens_seen": 48971136,
      "step": 25440
    },
    {
      "epoch": 4.150909535851211,
      "grad_norm": 0.0015728241996839643,
      "learning_rate": 0.08780077818553277,
      "loss": 0.163,
      "num_input_tokens_seen": 48980144,
      "step": 25445
    },
    {
      "epoch": 4.151725263072029,
      "grad_norm": 0.002204844495281577,
      "learning_rate": 0.0877471809868969,
      "loss": 0.1365,
      "num_input_tokens_seen": 48989856,
      "step": 25450
    },
    {
      "epoch": 4.152540990292846,
      "grad_norm": 0.003251529298722744,
      "learning_rate": 0.08769359338842811,
      "loss": 0.1835,
      "num_input_tokens_seen": 49000528,
      "step": 25455
    },
    {
      "epoch": 4.153356717513663,
      "grad_norm": 0.0017681351164355874,
      "learning_rate": 0.08764001539839016,
      "loss": 0.1619,
      "num_input_tokens_seen": 49009744,
      "step": 25460
    },
    {
      "epoch": 4.154172444734481,
      "grad_norm": 0.00238128867931664,
      "learning_rate": 0.08758644702504548,
      "loss": 0.201,
      "num_input_tokens_seen": 49018816,
      "step": 25465
    },
    {
      "epoch": 4.154988171955298,
      "grad_norm": 0.002155103487893939,
      "learning_rate": 0.0875328882766551,
      "loss": 0.2174,
      "num_input_tokens_seen": 49028912,
      "step": 25470
    },
    {
      "epoch": 4.1558038991761155,
      "grad_norm": 0.0014268583618104458,
      "learning_rate": 0.08747933916147828,
      "loss": 0.1487,
      "num_input_tokens_seen": 49039440,
      "step": 25475
    },
    {
      "epoch": 4.156619626396933,
      "grad_norm": 0.006805421318858862,
      "learning_rate": 0.0874257996877731,
      "loss": 0.1947,
      "num_input_tokens_seen": 49050176,
      "step": 25480
    },
    {
      "epoch": 4.157435353617751,
      "grad_norm": 0.002409805078059435,
      "learning_rate": 0.08737226986379593,
      "loss": 0.1914,
      "num_input_tokens_seen": 49059504,
      "step": 25485
    },
    {
      "epoch": 4.158251080838568,
      "grad_norm": 0.0023296799045056105,
      "learning_rate": 0.08731874969780173,
      "loss": 0.15,
      "num_input_tokens_seen": 49069280,
      "step": 25490
    },
    {
      "epoch": 4.159066808059385,
      "grad_norm": 0.001486390014179051,
      "learning_rate": 0.08726523919804412,
      "loss": 0.2097,
      "num_input_tokens_seen": 49077360,
      "step": 25495
    },
    {
      "epoch": 4.159882535280202,
      "grad_norm": 0.0023055088240653276,
      "learning_rate": 0.08721173837277492,
      "loss": 0.2035,
      "num_input_tokens_seen": 49087312,
      "step": 25500
    },
    {
      "epoch": 4.16069826250102,
      "grad_norm": 0.00410457281395793,
      "learning_rate": 0.08715824723024479,
      "loss": 0.2388,
      "num_input_tokens_seen": 49097776,
      "step": 25505
    },
    {
      "epoch": 4.161513989721837,
      "grad_norm": 0.007399382069706917,
      "learning_rate": 0.08710476577870258,
      "loss": 0.1932,
      "num_input_tokens_seen": 49107024,
      "step": 25510
    },
    {
      "epoch": 4.162329716942654,
      "grad_norm": 0.0023819447960704565,
      "learning_rate": 0.08705129402639587,
      "loss": 0.2315,
      "num_input_tokens_seen": 49116400,
      "step": 25515
    },
    {
      "epoch": 4.163145444163471,
      "grad_norm": 0.0034252943005412817,
      "learning_rate": 0.08699783198157078,
      "loss": 0.2251,
      "num_input_tokens_seen": 49126048,
      "step": 25520
    },
    {
      "epoch": 4.1639611713842895,
      "grad_norm": 0.004749044310301542,
      "learning_rate": 0.08694437965247163,
      "loss": 0.1744,
      "num_input_tokens_seen": 49136176,
      "step": 25525
    },
    {
      "epoch": 4.164776898605107,
      "grad_norm": 0.002856671577319503,
      "learning_rate": 0.08689093704734165,
      "loss": 0.219,
      "num_input_tokens_seen": 49145952,
      "step": 25530
    },
    {
      "epoch": 4.165592625825924,
      "grad_norm": 0.0025042290799319744,
      "learning_rate": 0.08683750417442222,
      "loss": 0.1542,
      "num_input_tokens_seen": 49155536,
      "step": 25535
    },
    {
      "epoch": 4.166408353046741,
      "grad_norm": 0.002557438565418124,
      "learning_rate": 0.08678408104195334,
      "loss": 0.2176,
      "num_input_tokens_seen": 49164624,
      "step": 25540
    },
    {
      "epoch": 4.167224080267559,
      "grad_norm": 0.0030342955142259598,
      "learning_rate": 0.08673066765817365,
      "loss": 0.2485,
      "num_input_tokens_seen": 49173552,
      "step": 25545
    },
    {
      "epoch": 4.168039807488376,
      "grad_norm": 0.004042220767587423,
      "learning_rate": 0.08667726403132005,
      "loss": 0.211,
      "num_input_tokens_seen": 49182240,
      "step": 25550
    },
    {
      "epoch": 4.168855534709193,
      "grad_norm": 0.0016034366562962532,
      "learning_rate": 0.0866238701696281,
      "loss": 0.186,
      "num_input_tokens_seen": 49192784,
      "step": 25555
    },
    {
      "epoch": 4.16967126193001,
      "grad_norm": 0.0025161292869597673,
      "learning_rate": 0.08657048608133185,
      "loss": 0.2532,
      "num_input_tokens_seen": 49202992,
      "step": 25560
    },
    {
      "epoch": 4.170486989150828,
      "grad_norm": 0.0024002804420888424,
      "learning_rate": 0.08651711177466369,
      "loss": 0.1984,
      "num_input_tokens_seen": 49213648,
      "step": 25565
    },
    {
      "epoch": 4.171302716371645,
      "grad_norm": 0.002284002723172307,
      "learning_rate": 0.08646374725785466,
      "loss": 0.1638,
      "num_input_tokens_seen": 49221568,
      "step": 25570
    },
    {
      "epoch": 4.1721184435924625,
      "grad_norm": 0.0025535679887980223,
      "learning_rate": 0.08641039253913434,
      "loss": 0.1973,
      "num_input_tokens_seen": 49231568,
      "step": 25575
    },
    {
      "epoch": 4.17293417081328,
      "grad_norm": 0.004579131491482258,
      "learning_rate": 0.08635704762673052,
      "loss": 0.2409,
      "num_input_tokens_seen": 49241472,
      "step": 25580
    },
    {
      "epoch": 4.173749898034098,
      "grad_norm": 0.0012289007427170873,
      "learning_rate": 0.08630371252886981,
      "loss": 0.2321,
      "num_input_tokens_seen": 49250816,
      "step": 25585
    },
    {
      "epoch": 4.174565625254915,
      "grad_norm": 0.0033077674452215433,
      "learning_rate": 0.08625038725377704,
      "loss": 0.2016,
      "num_input_tokens_seen": 49259792,
      "step": 25590
    },
    {
      "epoch": 4.175381352475732,
      "grad_norm": 0.002768861362710595,
      "learning_rate": 0.08619707180967566,
      "loss": 0.1872,
      "num_input_tokens_seen": 49270624,
      "step": 25595
    },
    {
      "epoch": 4.176197079696549,
      "grad_norm": 0.0023425426334142685,
      "learning_rate": 0.08614376620478768,
      "loss": 0.2566,
      "num_input_tokens_seen": 49280736,
      "step": 25600
    },
    {
      "epoch": 4.176197079696549,
      "eval_loss": 0.1973174810409546,
      "eval_runtime": 68.0778,
      "eval_samples_per_second": 40.028,
      "eval_steps_per_second": 20.021,
      "num_input_tokens_seen": 49280736,
      "step": 25600
    },
    {
      "epoch": 4.177012806917367,
      "grad_norm": 0.0028788906056433916,
      "learning_rate": 0.08609047044733344,
      "loss": 0.2064,
      "num_input_tokens_seen": 49289696,
      "step": 25605
    },
    {
      "epoch": 4.177828534138184,
      "grad_norm": 0.0023402171209454536,
      "learning_rate": 0.08603718454553168,
      "loss": 0.1778,
      "num_input_tokens_seen": 49299648,
      "step": 25610
    },
    {
      "epoch": 4.178644261359001,
      "grad_norm": 0.0021668123081326485,
      "learning_rate": 0.08598390850759997,
      "loss": 0.1817,
      "num_input_tokens_seen": 49309136,
      "step": 25615
    },
    {
      "epoch": 4.1794599885798185,
      "grad_norm": 0.002242762129753828,
      "learning_rate": 0.08593064234175397,
      "loss": 0.1884,
      "num_input_tokens_seen": 49318960,
      "step": 25620
    },
    {
      "epoch": 4.1802757158006365,
      "grad_norm": 0.0020654357504099607,
      "learning_rate": 0.08587738605620815,
      "loss": 0.1677,
      "num_input_tokens_seen": 49328624,
      "step": 25625
    },
    {
      "epoch": 4.181091443021454,
      "grad_norm": 0.001705216709524393,
      "learning_rate": 0.08582413965917512,
      "loss": 0.2013,
      "num_input_tokens_seen": 49339472,
      "step": 25630
    },
    {
      "epoch": 4.181907170242271,
      "grad_norm": 0.00155417260248214,
      "learning_rate": 0.08577090315886628,
      "loss": 0.2142,
      "num_input_tokens_seen": 49348704,
      "step": 25635
    },
    {
      "epoch": 4.182722897463089,
      "grad_norm": 0.0014842953532934189,
      "learning_rate": 0.08571767656349136,
      "loss": 0.1738,
      "num_input_tokens_seen": 49357600,
      "step": 25640
    },
    {
      "epoch": 4.183538624683906,
      "grad_norm": 0.002243249909952283,
      "learning_rate": 0.08566445988125847,
      "loss": 0.2323,
      "num_input_tokens_seen": 49366896,
      "step": 25645
    },
    {
      "epoch": 4.184354351904723,
      "grad_norm": 0.0022929986007511616,
      "learning_rate": 0.08561125312037436,
      "loss": 0.1855,
      "num_input_tokens_seen": 49375792,
      "step": 25650
    },
    {
      "epoch": 4.18517007912554,
      "grad_norm": 0.0019015047000721097,
      "learning_rate": 0.08555805628904424,
      "loss": 0.1789,
      "num_input_tokens_seen": 49385328,
      "step": 25655
    },
    {
      "epoch": 4.185985806346358,
      "grad_norm": 0.0017852748278528452,
      "learning_rate": 0.08550486939547161,
      "loss": 0.2276,
      "num_input_tokens_seen": 49394640,
      "step": 25660
    },
    {
      "epoch": 4.186801533567175,
      "grad_norm": 0.0024790980387479067,
      "learning_rate": 0.08545169244785869,
      "loss": 0.1979,
      "num_input_tokens_seen": 49404000,
      "step": 25665
    },
    {
      "epoch": 4.1876172607879925,
      "grad_norm": 0.0016837375005707145,
      "learning_rate": 0.08539852545440589,
      "loss": 0.2072,
      "num_input_tokens_seen": 49412400,
      "step": 25670
    },
    {
      "epoch": 4.18843298800881,
      "grad_norm": 0.0023725146893411875,
      "learning_rate": 0.08534536842331235,
      "loss": 0.1987,
      "num_input_tokens_seen": 49423504,
      "step": 25675
    },
    {
      "epoch": 4.189248715229628,
      "grad_norm": 0.0019754243548959494,
      "learning_rate": 0.08529222136277545,
      "loss": 0.246,
      "num_input_tokens_seen": 49432816,
      "step": 25680
    },
    {
      "epoch": 4.190064442450445,
      "grad_norm": 0.0025336153339594603,
      "learning_rate": 0.08523908428099125,
      "loss": 0.2175,
      "num_input_tokens_seen": 49443440,
      "step": 25685
    },
    {
      "epoch": 4.190880169671262,
      "grad_norm": 0.0020823560189455748,
      "learning_rate": 0.08518595718615402,
      "loss": 0.2079,
      "num_input_tokens_seen": 49452928,
      "step": 25690
    },
    {
      "epoch": 4.191695896892079,
      "grad_norm": 0.0018845928134396672,
      "learning_rate": 0.08513284008645675,
      "loss": 0.1563,
      "num_input_tokens_seen": 49462480,
      "step": 25695
    },
    {
      "epoch": 4.192511624112897,
      "grad_norm": 0.0030282435473054647,
      "learning_rate": 0.08507973299009065,
      "loss": 0.1794,
      "num_input_tokens_seen": 49471344,
      "step": 25700
    },
    {
      "epoch": 4.193327351333714,
      "grad_norm": 0.0021418470423668623,
      "learning_rate": 0.08502663590524563,
      "loss": 0.1695,
      "num_input_tokens_seen": 49480880,
      "step": 25705
    },
    {
      "epoch": 4.194143078554531,
      "grad_norm": 0.002749928506091237,
      "learning_rate": 0.08497354884010981,
      "loss": 0.1816,
      "num_input_tokens_seen": 49491520,
      "step": 25710
    },
    {
      "epoch": 4.194958805775348,
      "grad_norm": 0.0014837962808087468,
      "learning_rate": 0.0849204718028699,
      "loss": 0.2132,
      "num_input_tokens_seen": 49501584,
      "step": 25715
    },
    {
      "epoch": 4.195774532996166,
      "grad_norm": 0.00141610624268651,
      "learning_rate": 0.08486740480171118,
      "loss": 0.2459,
      "num_input_tokens_seen": 49510944,
      "step": 25720
    },
    {
      "epoch": 4.196590260216984,
      "grad_norm": 0.0021439676638692617,
      "learning_rate": 0.08481434784481706,
      "loss": 0.1682,
      "num_input_tokens_seen": 49520272,
      "step": 25725
    },
    {
      "epoch": 4.197405987437801,
      "grad_norm": 0.0024567402433604,
      "learning_rate": 0.08476130094036968,
      "loss": 0.1429,
      "num_input_tokens_seen": 49530544,
      "step": 25730
    },
    {
      "epoch": 4.198221714658618,
      "grad_norm": 0.0033111644443124533,
      "learning_rate": 0.08470826409654961,
      "loss": 0.2046,
      "num_input_tokens_seen": 49540560,
      "step": 25735
    },
    {
      "epoch": 4.199037441879436,
      "grad_norm": 0.001836466253735125,
      "learning_rate": 0.08465523732153564,
      "loss": 0.2011,
      "num_input_tokens_seen": 49548928,
      "step": 25740
    },
    {
      "epoch": 4.199853169100253,
      "grad_norm": 0.0015630976995453238,
      "learning_rate": 0.08460222062350532,
      "loss": 0.1484,
      "num_input_tokens_seen": 49557744,
      "step": 25745
    },
    {
      "epoch": 4.20066889632107,
      "grad_norm": 0.0019023733912035823,
      "learning_rate": 0.08454921401063442,
      "loss": 0.204,
      "num_input_tokens_seen": 49567376,
      "step": 25750
    },
    {
      "epoch": 4.201484623541887,
      "grad_norm": 0.0021403769496828318,
      "learning_rate": 0.08449621749109716,
      "loss": 0.2254,
      "num_input_tokens_seen": 49577632,
      "step": 25755
    },
    {
      "epoch": 4.202300350762705,
      "grad_norm": 0.0013994730543345213,
      "learning_rate": 0.08444323107306641,
      "loss": 0.1628,
      "num_input_tokens_seen": 49587600,
      "step": 25760
    },
    {
      "epoch": 4.203116077983522,
      "grad_norm": 0.0018012751825153828,
      "learning_rate": 0.0843902547647132,
      "loss": 0.1808,
      "num_input_tokens_seen": 49596480,
      "step": 25765
    },
    {
      "epoch": 4.2039318052043395,
      "grad_norm": 0.0027270615100860596,
      "learning_rate": 0.0843372885742072,
      "loss": 0.2457,
      "num_input_tokens_seen": 49604896,
      "step": 25770
    },
    {
      "epoch": 4.204747532425157,
      "grad_norm": 0.002796435495838523,
      "learning_rate": 0.08428433250971652,
      "loss": 0.1785,
      "num_input_tokens_seen": 49615248,
      "step": 25775
    },
    {
      "epoch": 4.205563259645975,
      "grad_norm": 0.0017297384329140186,
      "learning_rate": 0.08423138657940757,
      "loss": 0.2093,
      "num_input_tokens_seen": 49625136,
      "step": 25780
    },
    {
      "epoch": 4.206378986866792,
      "grad_norm": 0.0041849855333566666,
      "learning_rate": 0.08417845079144536,
      "loss": 0.2495,
      "num_input_tokens_seen": 49634240,
      "step": 25785
    },
    {
      "epoch": 4.207194714087609,
      "grad_norm": 0.0015346648870036006,
      "learning_rate": 0.08412552515399314,
      "loss": 0.1941,
      "num_input_tokens_seen": 49643456,
      "step": 25790
    },
    {
      "epoch": 4.208010441308426,
      "grad_norm": 0.003405725583434105,
      "learning_rate": 0.08407260967521278,
      "loss": 0.2065,
      "num_input_tokens_seen": 49652896,
      "step": 25795
    },
    {
      "epoch": 4.208826168529244,
      "grad_norm": 0.0021098540164530277,
      "learning_rate": 0.08401970436326454,
      "loss": 0.2054,
      "num_input_tokens_seen": 49662304,
      "step": 25800
    },
    {
      "epoch": 4.208826168529244,
      "eval_loss": 0.19668760895729065,
      "eval_runtime": 67.9565,
      "eval_samples_per_second": 40.099,
      "eval_steps_per_second": 20.057,
      "num_input_tokens_seen": 49662304,
      "step": 25800
    },
    {
      "epoch": 4.209641895750061,
      "grad_norm": 0.0029238995630294085,
      "learning_rate": 0.08396680922630702,
      "loss": 0.2414,
      "num_input_tokens_seen": 49671504,
      "step": 25805
    },
    {
      "epoch": 4.210457622970878,
      "grad_norm": 0.0017967912135645747,
      "learning_rate": 0.08391392427249732,
      "loss": 0.2105,
      "num_input_tokens_seen": 49681504,
      "step": 25810
    },
    {
      "epoch": 4.211273350191696,
      "grad_norm": 0.003671867772936821,
      "learning_rate": 0.08386104950999107,
      "loss": 0.2065,
      "num_input_tokens_seen": 49691616,
      "step": 25815
    },
    {
      "epoch": 4.2120890774125135,
      "grad_norm": 0.0034918540623039007,
      "learning_rate": 0.0838081849469421,
      "loss": 0.1928,
      "num_input_tokens_seen": 49700848,
      "step": 25820
    },
    {
      "epoch": 4.212904804633331,
      "grad_norm": 0.002881404710933566,
      "learning_rate": 0.08375533059150281,
      "loss": 0.1771,
      "num_input_tokens_seen": 49711344,
      "step": 25825
    },
    {
      "epoch": 4.213720531854148,
      "grad_norm": 0.002007128903642297,
      "learning_rate": 0.08370248645182406,
      "loss": 0.1886,
      "num_input_tokens_seen": 49719952,
      "step": 25830
    },
    {
      "epoch": 4.214536259074965,
      "grad_norm": 0.004515788517892361,
      "learning_rate": 0.083649652536055,
      "loss": 0.2128,
      "num_input_tokens_seen": 49728832,
      "step": 25835
    },
    {
      "epoch": 4.215351986295783,
      "grad_norm": 0.0013072005240246654,
      "learning_rate": 0.08359682885234339,
      "loss": 0.1982,
      "num_input_tokens_seen": 49739264,
      "step": 25840
    },
    {
      "epoch": 4.2161677135166,
      "grad_norm": 0.0031948410905897617,
      "learning_rate": 0.08354401540883516,
      "loss": 0.1781,
      "num_input_tokens_seen": 49748720,
      "step": 25845
    },
    {
      "epoch": 4.216983440737417,
      "grad_norm": 0.0023730716202408075,
      "learning_rate": 0.0834912122136749,
      "loss": 0.1997,
      "num_input_tokens_seen": 49757920,
      "step": 25850
    },
    {
      "epoch": 4.217799167958235,
      "grad_norm": 0.0017642844468355179,
      "learning_rate": 0.0834384192750056,
      "loss": 0.1665,
      "num_input_tokens_seen": 49767808,
      "step": 25855
    },
    {
      "epoch": 4.218614895179052,
      "grad_norm": 0.0015760231763124466,
      "learning_rate": 0.08338563660096844,
      "loss": 0.2138,
      "num_input_tokens_seen": 49777952,
      "step": 25860
    },
    {
      "epoch": 4.219430622399869,
      "grad_norm": 0.0019371015951037407,
      "learning_rate": 0.08333286419970329,
      "loss": 0.1685,
      "num_input_tokens_seen": 49788928,
      "step": 25865
    },
    {
      "epoch": 4.220246349620687,
      "grad_norm": 0.004189939238131046,
      "learning_rate": 0.08328010207934824,
      "loss": 0.2119,
      "num_input_tokens_seen": 49799008,
      "step": 25870
    },
    {
      "epoch": 4.221062076841505,
      "grad_norm": 0.0028236745856702328,
      "learning_rate": 0.08322735024803989,
      "loss": 0.1905,
      "num_input_tokens_seen": 49808976,
      "step": 25875
    },
    {
      "epoch": 4.221877804062322,
      "grad_norm": 0.0044397315941751,
      "learning_rate": 0.08317460871391331,
      "loss": 0.1944,
      "num_input_tokens_seen": 49819456,
      "step": 25880
    },
    {
      "epoch": 4.222693531283139,
      "grad_norm": 0.0023059432860463858,
      "learning_rate": 0.08312187748510179,
      "loss": 0.1895,
      "num_input_tokens_seen": 49829152,
      "step": 25885
    },
    {
      "epoch": 4.223509258503956,
      "grad_norm": 0.0019127285340800881,
      "learning_rate": 0.08306915656973726,
      "loss": 0.1961,
      "num_input_tokens_seen": 49838336,
      "step": 25890
    },
    {
      "epoch": 4.224324985724774,
      "grad_norm": 0.004286922514438629,
      "learning_rate": 0.08301644597594988,
      "loss": 0.1964,
      "num_input_tokens_seen": 49848048,
      "step": 25895
    },
    {
      "epoch": 4.225140712945591,
      "grad_norm": 0.0023194162640720606,
      "learning_rate": 0.08296374571186826,
      "loss": 0.2624,
      "num_input_tokens_seen": 49858880,
      "step": 25900
    },
    {
      "epoch": 4.225956440166408,
      "grad_norm": 0.003078353824093938,
      "learning_rate": 0.08291105578561955,
      "loss": 0.1736,
      "num_input_tokens_seen": 49868992,
      "step": 25905
    },
    {
      "epoch": 4.226772167387225,
      "grad_norm": 0.0013524418463930488,
      "learning_rate": 0.08285837620532904,
      "loss": 0.1814,
      "num_input_tokens_seen": 49878032,
      "step": 25910
    },
    {
      "epoch": 4.227587894608043,
      "grad_norm": 0.0029982796404510736,
      "learning_rate": 0.0828057069791207,
      "loss": 0.2314,
      "num_input_tokens_seen": 49887792,
      "step": 25915
    },
    {
      "epoch": 4.2284036218288605,
      "grad_norm": 0.002571947406977415,
      "learning_rate": 0.0827530481151168,
      "loss": 0.1606,
      "num_input_tokens_seen": 49897392,
      "step": 25920
    },
    {
      "epoch": 4.229219349049678,
      "grad_norm": 0.002041211351752281,
      "learning_rate": 0.08270039962143792,
      "loss": 0.2196,
      "num_input_tokens_seen": 49907600,
      "step": 25925
    },
    {
      "epoch": 4.230035076270495,
      "grad_norm": 0.00295381061732769,
      "learning_rate": 0.08264776150620314,
      "loss": 0.2285,
      "num_input_tokens_seen": 49918480,
      "step": 25930
    },
    {
      "epoch": 4.230850803491313,
      "grad_norm": 0.002387156942859292,
      "learning_rate": 0.08259513377753,
      "loss": 0.1805,
      "num_input_tokens_seen": 49928576,
      "step": 25935
    },
    {
      "epoch": 4.23166653071213,
      "grad_norm": 0.002424153033643961,
      "learning_rate": 0.08254251644353423,
      "loss": 0.1833,
      "num_input_tokens_seen": 49937888,
      "step": 25940
    },
    {
      "epoch": 4.232482257932947,
      "grad_norm": 0.004263810347765684,
      "learning_rate": 0.08248990951233022,
      "loss": 0.2146,
      "num_input_tokens_seen": 49946496,
      "step": 25945
    },
    {
      "epoch": 4.233297985153764,
      "grad_norm": 0.002260181587189436,
      "learning_rate": 0.08243731299203048,
      "loss": 0.1814,
      "num_input_tokens_seen": 49956528,
      "step": 25950
    },
    {
      "epoch": 4.234113712374582,
      "grad_norm": 0.002691785106435418,
      "learning_rate": 0.08238472689074612,
      "loss": 0.2078,
      "num_input_tokens_seen": 49966080,
      "step": 25955
    },
    {
      "epoch": 4.234929439595399,
      "grad_norm": 0.001569586107507348,
      "learning_rate": 0.08233215121658666,
      "loss": 0.1525,
      "num_input_tokens_seen": 49975680,
      "step": 25960
    },
    {
      "epoch": 4.2357451668162165,
      "grad_norm": 0.0024308429565280676,
      "learning_rate": 0.08227958597765982,
      "loss": 0.1641,
      "num_input_tokens_seen": 49984560,
      "step": 25965
    },
    {
      "epoch": 4.236560894037034,
      "grad_norm": 0.0027664832305163145,
      "learning_rate": 0.08222703118207181,
      "loss": 0.191,
      "num_input_tokens_seen": 49992448,
      "step": 25970
    },
    {
      "epoch": 4.237376621257852,
      "grad_norm": 0.0023368175607174635,
      "learning_rate": 0.08217448683792734,
      "loss": 0.1674,
      "num_input_tokens_seen": 50003184,
      "step": 25975
    },
    {
      "epoch": 4.238192348478669,
      "grad_norm": 0.002087517874315381,
      "learning_rate": 0.08212195295332926,
      "loss": 0.1908,
      "num_input_tokens_seen": 50013104,
      "step": 25980
    },
    {
      "epoch": 4.239008075699486,
      "grad_norm": 0.0015705206897109747,
      "learning_rate": 0.08206942953637915,
      "loss": 0.1879,
      "num_input_tokens_seen": 50021680,
      "step": 25985
    },
    {
      "epoch": 4.239823802920303,
      "grad_norm": 0.002682347083464265,
      "learning_rate": 0.08201691659517658,
      "loss": 0.1577,
      "num_input_tokens_seen": 50030752,
      "step": 25990
    },
    {
      "epoch": 4.240639530141121,
      "grad_norm": 0.0017623919993638992,
      "learning_rate": 0.08196441413781981,
      "loss": 0.2011,
      "num_input_tokens_seen": 50039664,
      "step": 25995
    },
    {
      "epoch": 4.241455257361938,
      "grad_norm": 0.0029897617641836405,
      "learning_rate": 0.08191192217240544,
      "loss": 0.1702,
      "num_input_tokens_seen": 50049312,
      "step": 26000
    },
    {
      "epoch": 4.241455257361938,
      "eval_loss": 0.19734841585159302,
      "eval_runtime": 67.9498,
      "eval_samples_per_second": 40.103,
      "eval_steps_per_second": 20.059,
      "num_input_tokens_seen": 50049312,
      "step": 26000
    },
    {
      "epoch": 4.242270984582755,
      "grad_norm": 0.002170601626858115,
      "learning_rate": 0.08185944070702823,
      "loss": 0.2029,
      "num_input_tokens_seen": 50058128,
      "step": 26005
    },
    {
      "epoch": 4.243086711803572,
      "grad_norm": 0.002160490956157446,
      "learning_rate": 0.08180696974978159,
      "loss": 0.155,
      "num_input_tokens_seen": 50068784,
      "step": 26010
    },
    {
      "epoch": 4.2439024390243905,
      "grad_norm": 0.0029188385233283043,
      "learning_rate": 0.08175450930875724,
      "loss": 0.2051,
      "num_input_tokens_seen": 50078960,
      "step": 26015
    },
    {
      "epoch": 4.244718166245208,
      "grad_norm": 0.002178508322685957,
      "learning_rate": 0.08170205939204513,
      "loss": 0.1586,
      "num_input_tokens_seen": 50088928,
      "step": 26020
    },
    {
      "epoch": 4.245533893466025,
      "grad_norm": 0.0016722369473427534,
      "learning_rate": 0.08164962000773379,
      "loss": 0.2578,
      "num_input_tokens_seen": 50098240,
      "step": 26025
    },
    {
      "epoch": 4.246349620686843,
      "grad_norm": 0.002695433096960187,
      "learning_rate": 0.08159719116390995,
      "loss": 0.1821,
      "num_input_tokens_seen": 50107440,
      "step": 26030
    },
    {
      "epoch": 4.24716534790766,
      "grad_norm": 0.0022577422205358744,
      "learning_rate": 0.08154477286865887,
      "loss": 0.244,
      "num_input_tokens_seen": 50116448,
      "step": 26035
    },
    {
      "epoch": 4.247981075128477,
      "grad_norm": 0.0015690403524786234,
      "learning_rate": 0.08149236513006404,
      "loss": 0.1585,
      "num_input_tokens_seen": 50126432,
      "step": 26040
    },
    {
      "epoch": 4.248796802349294,
      "grad_norm": 0.0034844211768358946,
      "learning_rate": 0.08143996795620746,
      "loss": 0.1556,
      "num_input_tokens_seen": 50135744,
      "step": 26045
    },
    {
      "epoch": 4.249612529570112,
      "grad_norm": 0.0011278863530606031,
      "learning_rate": 0.08138758135516938,
      "loss": 0.1811,
      "num_input_tokens_seen": 50146608,
      "step": 26050
    },
    {
      "epoch": 4.250428256790929,
      "grad_norm": 0.0015751119935885072,
      "learning_rate": 0.08133520533502851,
      "loss": 0.1551,
      "num_input_tokens_seen": 50156752,
      "step": 26055
    },
    {
      "epoch": 4.251243984011746,
      "grad_norm": 0.005397565197199583,
      "learning_rate": 0.08128283990386184,
      "loss": 0.1609,
      "num_input_tokens_seen": 50166112,
      "step": 26060
    },
    {
      "epoch": 4.2520597112325635,
      "grad_norm": 0.004663103260099888,
      "learning_rate": 0.08123048506974488,
      "loss": 0.1872,
      "num_input_tokens_seen": 50176800,
      "step": 26065
    },
    {
      "epoch": 4.252875438453382,
      "grad_norm": 0.004608467221260071,
      "learning_rate": 0.08117814084075124,
      "loss": 0.2007,
      "num_input_tokens_seen": 50186576,
      "step": 26070
    },
    {
      "epoch": 4.253691165674199,
      "grad_norm": 0.005334425717592239,
      "learning_rate": 0.08112580722495318,
      "loss": 0.2713,
      "num_input_tokens_seen": 50197280,
      "step": 26075
    },
    {
      "epoch": 4.254506892895016,
      "grad_norm": 0.0026245436165481806,
      "learning_rate": 0.08107348423042122,
      "loss": 0.1321,
      "num_input_tokens_seen": 50205840,
      "step": 26080
    },
    {
      "epoch": 4.255322620115833,
      "grad_norm": 0.0018342206021770835,
      "learning_rate": 0.08102117186522413,
      "loss": 0.218,
      "num_input_tokens_seen": 50216320,
      "step": 26085
    },
    {
      "epoch": 4.256138347336651,
      "grad_norm": 0.0018443366279825568,
      "learning_rate": 0.08096887013742916,
      "loss": 0.1914,
      "num_input_tokens_seen": 50225584,
      "step": 26090
    },
    {
      "epoch": 4.256954074557468,
      "grad_norm": 0.0019745067693293095,
      "learning_rate": 0.08091657905510198,
      "loss": 0.2055,
      "num_input_tokens_seen": 50233808,
      "step": 26095
    },
    {
      "epoch": 4.257769801778285,
      "grad_norm": 0.00303484289906919,
      "learning_rate": 0.08086429862630642,
      "loss": 0.1761,
      "num_input_tokens_seen": 50244672,
      "step": 26100
    },
    {
      "epoch": 4.258585528999102,
      "grad_norm": 0.007908706553280354,
      "learning_rate": 0.08081202885910488,
      "loss": 0.1762,
      "num_input_tokens_seen": 50254288,
      "step": 26105
    },
    {
      "epoch": 4.25940125621992,
      "grad_norm": 0.0024151259567588568,
      "learning_rate": 0.08075976976155795,
      "loss": 0.1691,
      "num_input_tokens_seen": 50263760,
      "step": 26110
    },
    {
      "epoch": 4.2602169834407375,
      "grad_norm": 0.001989659620448947,
      "learning_rate": 0.08070752134172461,
      "loss": 0.2087,
      "num_input_tokens_seen": 50273120,
      "step": 26115
    },
    {
      "epoch": 4.261032710661555,
      "grad_norm": 0.002867093775421381,
      "learning_rate": 0.08065528360766229,
      "loss": 0.1664,
      "num_input_tokens_seen": 50281872,
      "step": 26120
    },
    {
      "epoch": 4.261848437882372,
      "grad_norm": 0.005086763761937618,
      "learning_rate": 0.08060305656742664,
      "loss": 0.2714,
      "num_input_tokens_seen": 50290464,
      "step": 26125
    },
    {
      "epoch": 4.26266416510319,
      "grad_norm": 0.002685453277081251,
      "learning_rate": 0.08055084022907182,
      "loss": 0.2396,
      "num_input_tokens_seen": 50299200,
      "step": 26130
    },
    {
      "epoch": 4.263479892324007,
      "grad_norm": 0.0028162605594843626,
      "learning_rate": 0.08049863460065014,
      "loss": 0.1928,
      "num_input_tokens_seen": 50308576,
      "step": 26135
    },
    {
      "epoch": 4.264295619544824,
      "grad_norm": 0.003572865156456828,
      "learning_rate": 0.0804464396902124,
      "loss": 0.1541,
      "num_input_tokens_seen": 50317968,
      "step": 26140
    },
    {
      "epoch": 4.265111346765641,
      "grad_norm": 0.003970385529100895,
      "learning_rate": 0.08039425550580777,
      "loss": 0.2587,
      "num_input_tokens_seen": 50325792,
      "step": 26145
    },
    {
      "epoch": 4.265927073986459,
      "grad_norm": 0.0045647635124623775,
      "learning_rate": 0.08034208205548363,
      "loss": 0.2352,
      "num_input_tokens_seen": 50335984,
      "step": 26150
    },
    {
      "epoch": 4.266742801207276,
      "grad_norm": 0.0040622977539896965,
      "learning_rate": 0.08028991934728581,
      "loss": 0.2054,
      "num_input_tokens_seen": 50346048,
      "step": 26155
    },
    {
      "epoch": 4.2675585284280935,
      "grad_norm": 0.0021862166468054056,
      "learning_rate": 0.0802377673892585,
      "loss": 0.2184,
      "num_input_tokens_seen": 50354400,
      "step": 26160
    },
    {
      "epoch": 4.268374255648911,
      "grad_norm": 0.0022134461905807257,
      "learning_rate": 0.0801856261894441,
      "loss": 0.1719,
      "num_input_tokens_seen": 50365248,
      "step": 26165
    },
    {
      "epoch": 4.269189982869729,
      "grad_norm": 0.003034577937796712,
      "learning_rate": 0.08013349575588354,
      "loss": 0.1814,
      "num_input_tokens_seen": 50376128,
      "step": 26170
    },
    {
      "epoch": 4.270005710090546,
      "grad_norm": 0.003232040209695697,
      "learning_rate": 0.08008137609661586,
      "loss": 0.2365,
      "num_input_tokens_seen": 50385168,
      "step": 26175
    },
    {
      "epoch": 4.270821437311363,
      "grad_norm": 0.002478013513609767,
      "learning_rate": 0.08002926721967872,
      "loss": 0.1917,
      "num_input_tokens_seen": 50394160,
      "step": 26180
    },
    {
      "epoch": 4.27163716453218,
      "grad_norm": 0.0015757351648062468,
      "learning_rate": 0.07997716913310782,
      "loss": 0.1873,
      "num_input_tokens_seen": 50404288,
      "step": 26185
    },
    {
      "epoch": 4.272452891752998,
      "grad_norm": 0.0027831080369651318,
      "learning_rate": 0.07992508184493745,
      "loss": 0.1756,
      "num_input_tokens_seen": 50414160,
      "step": 26190
    },
    {
      "epoch": 4.273268618973815,
      "grad_norm": 0.0035687617491930723,
      "learning_rate": 0.07987300536320001,
      "loss": 0.2468,
      "num_input_tokens_seen": 50423760,
      "step": 26195
    },
    {
      "epoch": 4.274084346194632,
      "grad_norm": 0.0021839637774974108,
      "learning_rate": 0.07982093969592649,
      "loss": 0.2251,
      "num_input_tokens_seen": 50433008,
      "step": 26200
    },
    {
      "epoch": 4.274084346194632,
      "eval_loss": 0.20155489444732666,
      "eval_runtime": 68.0818,
      "eval_samples_per_second": 40.025,
      "eval_steps_per_second": 20.02,
      "num_input_tokens_seen": 50433008,
      "step": 26200
    },
    {
      "epoch": 4.27490007341545,
      "grad_norm": 0.003191782161593437,
      "learning_rate": 0.07976888485114592,
      "loss": 0.2023,
      "num_input_tokens_seen": 50442320,
      "step": 26205
    },
    {
      "epoch": 4.275715800636267,
      "grad_norm": 0.001981667708605528,
      "learning_rate": 0.07971684083688595,
      "loss": 0.1622,
      "num_input_tokens_seen": 50451120,
      "step": 26210
    },
    {
      "epoch": 4.276531527857085,
      "grad_norm": 0.002664444502443075,
      "learning_rate": 0.0796648076611723,
      "loss": 0.1805,
      "num_input_tokens_seen": 50461024,
      "step": 26215
    },
    {
      "epoch": 4.277347255077902,
      "grad_norm": 0.0026474653277546167,
      "learning_rate": 0.07961278533202922,
      "loss": 0.1843,
      "num_input_tokens_seen": 50470672,
      "step": 26220
    },
    {
      "epoch": 4.27816298229872,
      "grad_norm": 0.002274462953209877,
      "learning_rate": 0.07956077385747919,
      "loss": 0.1606,
      "num_input_tokens_seen": 50480304,
      "step": 26225
    },
    {
      "epoch": 4.278978709519537,
      "grad_norm": 0.003321586176753044,
      "learning_rate": 0.079508773245543,
      "loss": 0.2357,
      "num_input_tokens_seen": 50489328,
      "step": 26230
    },
    {
      "epoch": 4.279794436740354,
      "grad_norm": 0.004171502310782671,
      "learning_rate": 0.07945678350423982,
      "loss": 0.1837,
      "num_input_tokens_seen": 50498944,
      "step": 26235
    },
    {
      "epoch": 4.280610163961171,
      "grad_norm": 0.002584881614893675,
      "learning_rate": 0.07940480464158717,
      "loss": 0.1954,
      "num_input_tokens_seen": 50509168,
      "step": 26240
    },
    {
      "epoch": 4.281425891181989,
      "grad_norm": 0.002836412051692605,
      "learning_rate": 0.07935283666560076,
      "loss": 0.2298,
      "num_input_tokens_seen": 50519248,
      "step": 26245
    },
    {
      "epoch": 4.282241618402806,
      "grad_norm": 0.0041968608275055885,
      "learning_rate": 0.07930087958429478,
      "loss": 0.1931,
      "num_input_tokens_seen": 50528816,
      "step": 26250
    },
    {
      "epoch": 4.283057345623623,
      "grad_norm": 0.0017210675869137049,
      "learning_rate": 0.07924893340568159,
      "loss": 0.1668,
      "num_input_tokens_seen": 50538528,
      "step": 26255
    },
    {
      "epoch": 4.2838730728444405,
      "grad_norm": 0.0027473745867609978,
      "learning_rate": 0.07919699813777205,
      "loss": 0.2013,
      "num_input_tokens_seen": 50548608,
      "step": 26260
    },
    {
      "epoch": 4.2846888000652585,
      "grad_norm": 0.0033360382076352835,
      "learning_rate": 0.07914507378857515,
      "loss": 0.2549,
      "num_input_tokens_seen": 50558128,
      "step": 26265
    },
    {
      "epoch": 4.285504527286076,
      "grad_norm": 0.002464616671204567,
      "learning_rate": 0.07909316036609822,
      "loss": 0.1906,
      "num_input_tokens_seen": 50568304,
      "step": 26270
    },
    {
      "epoch": 4.286320254506893,
      "grad_norm": 0.0026593448128551245,
      "learning_rate": 0.07904125787834704,
      "loss": 0.2177,
      "num_input_tokens_seen": 50577232,
      "step": 26275
    },
    {
      "epoch": 4.28713598172771,
      "grad_norm": 0.0013379196170717478,
      "learning_rate": 0.07898936633332569,
      "loss": 0.174,
      "num_input_tokens_seen": 50586448,
      "step": 26280
    },
    {
      "epoch": 4.287951708948528,
      "grad_norm": 0.003288897220045328,
      "learning_rate": 0.07893748573903635,
      "loss": 0.3154,
      "num_input_tokens_seen": 50596640,
      "step": 26285
    },
    {
      "epoch": 4.288767436169345,
      "grad_norm": 0.002361331135034561,
      "learning_rate": 0.0788856161034798,
      "loss": 0.2046,
      "num_input_tokens_seen": 50606784,
      "step": 26290
    },
    {
      "epoch": 4.289583163390162,
      "grad_norm": 0.0032858599442988634,
      "learning_rate": 0.07883375743465487,
      "loss": 0.251,
      "num_input_tokens_seen": 50615696,
      "step": 26295
    },
    {
      "epoch": 4.290398890610979,
      "grad_norm": 0.0031201615929603577,
      "learning_rate": 0.07878190974055888,
      "loss": 0.1897,
      "num_input_tokens_seen": 50624176,
      "step": 26300
    },
    {
      "epoch": 4.291214617831797,
      "grad_norm": 0.003612009808421135,
      "learning_rate": 0.07873007302918746,
      "loss": 0.2007,
      "num_input_tokens_seen": 50634432,
      "step": 26305
    },
    {
      "epoch": 4.2920303450526145,
      "grad_norm": 0.0013805076014250517,
      "learning_rate": 0.07867824730853433,
      "loss": 0.1797,
      "num_input_tokens_seen": 50644256,
      "step": 26310
    },
    {
      "epoch": 4.292846072273432,
      "grad_norm": 0.0030139347072690725,
      "learning_rate": 0.07862643258659176,
      "loss": 0.2017,
      "num_input_tokens_seen": 50653760,
      "step": 26315
    },
    {
      "epoch": 4.293661799494249,
      "grad_norm": 0.0029738678131252527,
      "learning_rate": 0.07857462887135026,
      "loss": 0.2659,
      "num_input_tokens_seen": 50663264,
      "step": 26320
    },
    {
      "epoch": 4.294477526715067,
      "grad_norm": 0.004949885420501232,
      "learning_rate": 0.0785228361707986,
      "loss": 0.2958,
      "num_input_tokens_seen": 50672704,
      "step": 26325
    },
    {
      "epoch": 4.295293253935884,
      "grad_norm": 0.0025054560974240303,
      "learning_rate": 0.07847105449292378,
      "loss": 0.2074,
      "num_input_tokens_seen": 50682160,
      "step": 26330
    },
    {
      "epoch": 4.296108981156701,
      "grad_norm": 0.0024530338123440742,
      "learning_rate": 0.0784192838457113,
      "loss": 0.2042,
      "num_input_tokens_seen": 50690960,
      "step": 26335
    },
    {
      "epoch": 4.296924708377518,
      "grad_norm": 0.0022188264410942793,
      "learning_rate": 0.07836752423714473,
      "loss": 0.1681,
      "num_input_tokens_seen": 50700496,
      "step": 26340
    },
    {
      "epoch": 4.297740435598336,
      "grad_norm": 0.0016393036348745227,
      "learning_rate": 0.07831577567520616,
      "loss": 0.2452,
      "num_input_tokens_seen": 50710480,
      "step": 26345
    },
    {
      "epoch": 4.298556162819153,
      "grad_norm": 0.002733982400968671,
      "learning_rate": 0.07826403816787579,
      "loss": 0.2016,
      "num_input_tokens_seen": 50720528,
      "step": 26350
    },
    {
      "epoch": 4.2993718900399704,
      "grad_norm": 0.0014639557339251041,
      "learning_rate": 0.0782123117231322,
      "loss": 0.2271,
      "num_input_tokens_seen": 50728880,
      "step": 26355
    },
    {
      "epoch": 4.300187617260788,
      "grad_norm": 0.001818807446397841,
      "learning_rate": 0.07816059634895237,
      "loss": 0.2382,
      "num_input_tokens_seen": 50738960,
      "step": 26360
    },
    {
      "epoch": 4.301003344481606,
      "grad_norm": 0.0015768770826980472,
      "learning_rate": 0.0781088920533113,
      "loss": 0.2184,
      "num_input_tokens_seen": 50748496,
      "step": 26365
    },
    {
      "epoch": 4.301819071702423,
      "grad_norm": 0.0030673937872052193,
      "learning_rate": 0.07805719884418257,
      "loss": 0.1849,
      "num_input_tokens_seen": 50757904,
      "step": 26370
    },
    {
      "epoch": 4.30263479892324,
      "grad_norm": 0.0020500938408076763,
      "learning_rate": 0.07800551672953779,
      "loss": 0.2427,
      "num_input_tokens_seen": 50767328,
      "step": 26375
    },
    {
      "epoch": 4.303450526144058,
      "grad_norm": 0.0024069328792393208,
      "learning_rate": 0.07795384571734709,
      "loss": 0.2039,
      "num_input_tokens_seen": 50778288,
      "step": 26380
    },
    {
      "epoch": 4.304266253364875,
      "grad_norm": 0.0021757897920906544,
      "learning_rate": 0.07790218581557883,
      "loss": 0.2089,
      "num_input_tokens_seen": 50787888,
      "step": 26385
    },
    {
      "epoch": 4.305081980585692,
      "grad_norm": 0.004107036627829075,
      "learning_rate": 0.07785053703219949,
      "loss": 0.2206,
      "num_input_tokens_seen": 50796336,
      "step": 26390
    },
    {
      "epoch": 4.305897707806509,
      "grad_norm": 0.003047388978302479,
      "learning_rate": 0.07779889937517409,
      "loss": 0.2003,
      "num_input_tokens_seen": 50805968,
      "step": 26395
    },
    {
      "epoch": 4.306713435027326,
      "grad_norm": 0.0031079137697815895,
      "learning_rate": 0.0777472728524657,
      "loss": 0.2118,
      "num_input_tokens_seen": 50815824,
      "step": 26400
    },
    {
      "epoch": 4.306713435027326,
      "eval_loss": 0.1986146718263626,
      "eval_runtime": 68.0215,
      "eval_samples_per_second": 40.061,
      "eval_steps_per_second": 20.038,
      "num_input_tokens_seen": 50815824,
      "step": 26400
    },
    {
      "epoch": 4.307529162248144,
      "grad_norm": 0.0024784787092357874,
      "learning_rate": 0.07769565747203584,
      "loss": 0.2294,
      "num_input_tokens_seen": 50825488,
      "step": 26405
    },
    {
      "epoch": 4.3083448894689615,
      "grad_norm": 0.0014552678912878036,
      "learning_rate": 0.07764405324184427,
      "loss": 0.2531,
      "num_input_tokens_seen": 50835568,
      "step": 26410
    },
    {
      "epoch": 4.309160616689779,
      "grad_norm": 0.0025655904319137335,
      "learning_rate": 0.07759246016984889,
      "loss": 0.1574,
      "num_input_tokens_seen": 50844688,
      "step": 26415
    },
    {
      "epoch": 4.309976343910597,
      "grad_norm": 0.0022657462395727634,
      "learning_rate": 0.07754087826400609,
      "loss": 0.1885,
      "num_input_tokens_seen": 50853760,
      "step": 26420
    },
    {
      "epoch": 4.310792071131414,
      "grad_norm": 0.0028017177246510983,
      "learning_rate": 0.0774893075322705,
      "loss": 0.2262,
      "num_input_tokens_seen": 50862960,
      "step": 26425
    },
    {
      "epoch": 4.311607798352231,
      "grad_norm": 0.0024249295238405466,
      "learning_rate": 0.07743774798259484,
      "loss": 0.1868,
      "num_input_tokens_seen": 50872288,
      "step": 26430
    },
    {
      "epoch": 4.312423525573048,
      "grad_norm": 0.0035991892218589783,
      "learning_rate": 0.07738619962293032,
      "loss": 0.2192,
      "num_input_tokens_seen": 50882560,
      "step": 26435
    },
    {
      "epoch": 4.313239252793866,
      "grad_norm": 0.002039461163803935,
      "learning_rate": 0.0773346624612264,
      "loss": 0.1964,
      "num_input_tokens_seen": 50891968,
      "step": 26440
    },
    {
      "epoch": 4.314054980014683,
      "grad_norm": 0.0017549187177792192,
      "learning_rate": 0.07728313650543066,
      "loss": 0.2273,
      "num_input_tokens_seen": 50902160,
      "step": 26445
    },
    {
      "epoch": 4.3148707072355,
      "grad_norm": 0.0026035772170871496,
      "learning_rate": 0.07723162176348913,
      "loss": 0.2152,
      "num_input_tokens_seen": 50911872,
      "step": 26450
    },
    {
      "epoch": 4.3156864344563175,
      "grad_norm": 0.0013374134432524443,
      "learning_rate": 0.07718011824334593,
      "loss": 0.1654,
      "num_input_tokens_seen": 50920832,
      "step": 26455
    },
    {
      "epoch": 4.3165021616771355,
      "grad_norm": 0.0019439453026279807,
      "learning_rate": 0.07712862595294363,
      "loss": 0.1996,
      "num_input_tokens_seen": 50930592,
      "step": 26460
    },
    {
      "epoch": 4.317317888897953,
      "grad_norm": 0.0014336203457787633,
      "learning_rate": 0.07707714490022301,
      "loss": 0.1682,
      "num_input_tokens_seen": 50940272,
      "step": 26465
    },
    {
      "epoch": 4.31813361611877,
      "grad_norm": 0.0020367042161524296,
      "learning_rate": 0.07702567509312298,
      "loss": 0.1905,
      "num_input_tokens_seen": 50950016,
      "step": 26470
    },
    {
      "epoch": 4.318949343339587,
      "grad_norm": 0.002051320858299732,
      "learning_rate": 0.07697421653958098,
      "loss": 0.2158,
      "num_input_tokens_seen": 50960336,
      "step": 26475
    },
    {
      "epoch": 4.319765070560405,
      "grad_norm": 0.0019128567073494196,
      "learning_rate": 0.07692276924753247,
      "loss": 0.1687,
      "num_input_tokens_seen": 50969680,
      "step": 26480
    },
    {
      "epoch": 4.320580797781222,
      "grad_norm": 0.003161439672112465,
      "learning_rate": 0.07687133322491124,
      "loss": 0.23,
      "num_input_tokens_seen": 50978800,
      "step": 26485
    },
    {
      "epoch": 4.321396525002039,
      "grad_norm": 0.0029849475249648094,
      "learning_rate": 0.07681990847964948,
      "loss": 0.2187,
      "num_input_tokens_seen": 50988640,
      "step": 26490
    },
    {
      "epoch": 4.322212252222856,
      "grad_norm": 0.0028658562805503607,
      "learning_rate": 0.0767684950196774,
      "loss": 0.1689,
      "num_input_tokens_seen": 50998720,
      "step": 26495
    },
    {
      "epoch": 4.323027979443674,
      "grad_norm": 0.0022734578233212233,
      "learning_rate": 0.0767170928529237,
      "loss": 0.2097,
      "num_input_tokens_seen": 51009456,
      "step": 26500
    },
    {
      "epoch": 4.3238437066644915,
      "grad_norm": 0.0031597083434462547,
      "learning_rate": 0.07666570198731526,
      "loss": 0.2034,
      "num_input_tokens_seen": 51019328,
      "step": 26505
    },
    {
      "epoch": 4.324659433885309,
      "grad_norm": 0.004452461376786232,
      "learning_rate": 0.07661432243077708,
      "loss": 0.1752,
      "num_input_tokens_seen": 51028784,
      "step": 26510
    },
    {
      "epoch": 4.325475161106126,
      "grad_norm": 0.0023237885907292366,
      "learning_rate": 0.0765629541912326,
      "loss": 0.1896,
      "num_input_tokens_seen": 51039696,
      "step": 26515
    },
    {
      "epoch": 4.326290888326944,
      "grad_norm": 0.0011921394616365433,
      "learning_rate": 0.07651159727660352,
      "loss": 0.1298,
      "num_input_tokens_seen": 51048496,
      "step": 26520
    },
    {
      "epoch": 4.327106615547761,
      "grad_norm": 0.0022835079580545425,
      "learning_rate": 0.07646025169480959,
      "loss": 0.1956,
      "num_input_tokens_seen": 51057648,
      "step": 26525
    },
    {
      "epoch": 4.327922342768578,
      "grad_norm": 0.002570393029600382,
      "learning_rate": 0.07640891745376908,
      "loss": 0.1897,
      "num_input_tokens_seen": 51066304,
      "step": 26530
    },
    {
      "epoch": 4.328738069989395,
      "grad_norm": 0.0018186995293945074,
      "learning_rate": 0.07635759456139822,
      "loss": 0.1848,
      "num_input_tokens_seen": 51075360,
      "step": 26535
    },
    {
      "epoch": 4.329553797210213,
      "grad_norm": 0.0014647557400166988,
      "learning_rate": 0.0763062830256118,
      "loss": 0.1398,
      "num_input_tokens_seen": 51085296,
      "step": 26540
    },
    {
      "epoch": 4.33036952443103,
      "grad_norm": 0.0022531074937433004,
      "learning_rate": 0.07625498285432258,
      "loss": 0.1725,
      "num_input_tokens_seen": 51096064,
      "step": 26545
    },
    {
      "epoch": 4.331185251651847,
      "grad_norm": 0.0026272088289260864,
      "learning_rate": 0.07620369405544176,
      "loss": 0.1775,
      "num_input_tokens_seen": 51105392,
      "step": 26550
    },
    {
      "epoch": 4.332000978872665,
      "grad_norm": 0.0016654761275276542,
      "learning_rate": 0.07615241663687868,
      "loss": 0.2072,
      "num_input_tokens_seen": 51115136,
      "step": 26555
    },
    {
      "epoch": 4.332816706093483,
      "grad_norm": 0.0029964165296405554,
      "learning_rate": 0.07610115060654106,
      "loss": 0.1801,
      "num_input_tokens_seen": 51123984,
      "step": 26560
    },
    {
      "epoch": 4.3336324333143,
      "grad_norm": 0.0025888727977871895,
      "learning_rate": 0.07604989597233458,
      "loss": 0.1862,
      "num_input_tokens_seen": 51134208,
      "step": 26565
    },
    {
      "epoch": 4.334448160535117,
      "grad_norm": 0.0023296601139009,
      "learning_rate": 0.07599865274216352,
      "loss": 0.1898,
      "num_input_tokens_seen": 51143504,
      "step": 26570
    },
    {
      "epoch": 4.335263887755934,
      "grad_norm": 0.0032761781476438046,
      "learning_rate": 0.07594742092393013,
      "loss": 0.2043,
      "num_input_tokens_seen": 51152928,
      "step": 26575
    },
    {
      "epoch": 4.336079614976752,
      "grad_norm": 0.0018713145982474089,
      "learning_rate": 0.07589620052553503,
      "loss": 0.1816,
      "num_input_tokens_seen": 51162512,
      "step": 26580
    },
    {
      "epoch": 4.336895342197569,
      "grad_norm": 0.0016206419095396996,
      "learning_rate": 0.0758449915548771,
      "loss": 0.1911,
      "num_input_tokens_seen": 51172704,
      "step": 26585
    },
    {
      "epoch": 4.337711069418386,
      "grad_norm": 0.002060868078842759,
      "learning_rate": 0.07579379401985332,
      "loss": 0.2314,
      "num_input_tokens_seen": 51181488,
      "step": 26590
    },
    {
      "epoch": 4.338526796639204,
      "grad_norm": 0.0021483725868165493,
      "learning_rate": 0.07574260792835905,
      "loss": 0.2204,
      "num_input_tokens_seen": 51191648,
      "step": 26595
    },
    {
      "epoch": 4.339342523860021,
      "grad_norm": 0.004693264607340097,
      "learning_rate": 0.07569143328828784,
      "loss": 0.1826,
      "num_input_tokens_seen": 51200224,
      "step": 26600
    },
    {
      "epoch": 4.339342523860021,
      "eval_loss": 0.2019011378288269,
      "eval_runtime": 68.0753,
      "eval_samples_per_second": 40.029,
      "eval_steps_per_second": 20.022,
      "num_input_tokens_seen": 51200224,
      "step": 26600
    },
    {
      "epoch": 4.3401582510808385,
      "grad_norm": 0.002058642450720072,
      "learning_rate": 0.0756402701075314,
      "loss": 0.2094,
      "num_input_tokens_seen": 51210992,
      "step": 26605
    },
    {
      "epoch": 4.340973978301656,
      "grad_norm": 0.004525191616266966,
      "learning_rate": 0.07558911839397982,
      "loss": 0.1903,
      "num_input_tokens_seen": 51220400,
      "step": 26610
    },
    {
      "epoch": 4.341789705522474,
      "grad_norm": 0.0020976848900318146,
      "learning_rate": 0.07553797815552123,
      "loss": 0.1563,
      "num_input_tokens_seen": 51229600,
      "step": 26615
    },
    {
      "epoch": 4.342605432743291,
      "grad_norm": 0.0030059285927563906,
      "learning_rate": 0.07548684940004222,
      "loss": 0.2437,
      "num_input_tokens_seen": 51239584,
      "step": 26620
    },
    {
      "epoch": 4.343421159964108,
      "grad_norm": 0.004509368911385536,
      "learning_rate": 0.07543573213542744,
      "loss": 0.1733,
      "num_input_tokens_seen": 51250352,
      "step": 26625
    },
    {
      "epoch": 4.344236887184925,
      "grad_norm": 0.002660147612914443,
      "learning_rate": 0.0753846263695597,
      "loss": 0.1481,
      "num_input_tokens_seen": 51261056,
      "step": 26630
    },
    {
      "epoch": 4.345052614405743,
      "grad_norm": 0.0026347902603447437,
      "learning_rate": 0.07533353211032029,
      "loss": 0.2923,
      "num_input_tokens_seen": 51271472,
      "step": 26635
    },
    {
      "epoch": 4.34586834162656,
      "grad_norm": 0.00413455069065094,
      "learning_rate": 0.07528244936558857,
      "loss": 0.2519,
      "num_input_tokens_seen": 51281168,
      "step": 26640
    },
    {
      "epoch": 4.346684068847377,
      "grad_norm": 0.002838372252881527,
      "learning_rate": 0.07523137814324206,
      "loss": 0.2134,
      "num_input_tokens_seen": 51290512,
      "step": 26645
    },
    {
      "epoch": 4.3474997960681945,
      "grad_norm": 0.0026399781927466393,
      "learning_rate": 0.07518031845115672,
      "loss": 0.1551,
      "num_input_tokens_seen": 51300096,
      "step": 26650
    },
    {
      "epoch": 4.3483155232890125,
      "grad_norm": 0.0029045166447758675,
      "learning_rate": 0.07512927029720647,
      "loss": 0.2038,
      "num_input_tokens_seen": 51310592,
      "step": 26655
    },
    {
      "epoch": 4.34913125050983,
      "grad_norm": 0.0026297755539417267,
      "learning_rate": 0.0750782336892636,
      "loss": 0.1804,
      "num_input_tokens_seen": 51320112,
      "step": 26660
    },
    {
      "epoch": 4.349946977730647,
      "grad_norm": 0.004208586644381285,
      "learning_rate": 0.0750272086351987,
      "loss": 0.1812,
      "num_input_tokens_seen": 51329680,
      "step": 26665
    },
    {
      "epoch": 4.350762704951464,
      "grad_norm": 0.0021127318032085896,
      "learning_rate": 0.07497619514288031,
      "loss": 0.192,
      "num_input_tokens_seen": 51339472,
      "step": 26670
    },
    {
      "epoch": 4.351578432172282,
      "grad_norm": 0.0032886408735066652,
      "learning_rate": 0.07492519322017545,
      "loss": 0.2026,
      "num_input_tokens_seen": 51348784,
      "step": 26675
    },
    {
      "epoch": 4.352394159393099,
      "grad_norm": 0.0051469882018864155,
      "learning_rate": 0.0748742028749493,
      "loss": 0.2339,
      "num_input_tokens_seen": 51359104,
      "step": 26680
    },
    {
      "epoch": 4.353209886613916,
      "grad_norm": 0.003003358840942383,
      "learning_rate": 0.0748232241150651,
      "loss": 0.2203,
      "num_input_tokens_seen": 51369360,
      "step": 26685
    },
    {
      "epoch": 4.354025613834733,
      "grad_norm": 0.002010344760492444,
      "learning_rate": 0.07477225694838453,
      "loss": 0.2219,
      "num_input_tokens_seen": 51379184,
      "step": 26690
    },
    {
      "epoch": 4.354841341055551,
      "grad_norm": 0.002116012154147029,
      "learning_rate": 0.07472130138276731,
      "loss": 0.2762,
      "num_input_tokens_seen": 51388240,
      "step": 26695
    },
    {
      "epoch": 4.3556570682763684,
      "grad_norm": 0.0038426066748797894,
      "learning_rate": 0.07467035742607138,
      "loss": 0.1929,
      "num_input_tokens_seen": 51398192,
      "step": 26700
    },
    {
      "epoch": 4.356472795497186,
      "grad_norm": 0.0012764906277880073,
      "learning_rate": 0.07461942508615303,
      "loss": 0.1387,
      "num_input_tokens_seen": 51408064,
      "step": 26705
    },
    {
      "epoch": 4.357288522718003,
      "grad_norm": 0.0041689579375088215,
      "learning_rate": 0.07456850437086657,
      "loss": 0.2057,
      "num_input_tokens_seen": 51417088,
      "step": 26710
    },
    {
      "epoch": 4.358104249938821,
      "grad_norm": 0.003736286424100399,
      "learning_rate": 0.07451759528806468,
      "loss": 0.2262,
      "num_input_tokens_seen": 51426624,
      "step": 26715
    },
    {
      "epoch": 4.358919977159638,
      "grad_norm": 0.0021485784091055393,
      "learning_rate": 0.0744666978455982,
      "loss": 0.1571,
      "num_input_tokens_seen": 51436416,
      "step": 26720
    },
    {
      "epoch": 4.359735704380455,
      "grad_norm": 0.003171271178871393,
      "learning_rate": 0.07441581205131609,
      "loss": 0.2044,
      "num_input_tokens_seen": 51444832,
      "step": 26725
    },
    {
      "epoch": 4.360551431601272,
      "grad_norm": 0.0024184384383261204,
      "learning_rate": 0.07436493791306566,
      "loss": 0.1659,
      "num_input_tokens_seen": 51454624,
      "step": 26730
    },
    {
      "epoch": 4.36136715882209,
      "grad_norm": 0.0035826368257403374,
      "learning_rate": 0.07431407543869223,
      "loss": 0.1394,
      "num_input_tokens_seen": 51463520,
      "step": 26735
    },
    {
      "epoch": 4.362182886042907,
      "grad_norm": 0.005830102134495974,
      "learning_rate": 0.0742632246360395,
      "loss": 0.2094,
      "num_input_tokens_seen": 51473760,
      "step": 26740
    },
    {
      "epoch": 4.362998613263724,
      "grad_norm": 0.0021363170817494392,
      "learning_rate": 0.07421238551294934,
      "loss": 0.2401,
      "num_input_tokens_seen": 51483760,
      "step": 26745
    },
    {
      "epoch": 4.3638143404845415,
      "grad_norm": 0.0026326191145926714,
      "learning_rate": 0.07416155807726171,
      "loss": 0.1616,
      "num_input_tokens_seen": 51492240,
      "step": 26750
    },
    {
      "epoch": 4.3646300677053596,
      "grad_norm": 0.0028636669740080833,
      "learning_rate": 0.07411074233681492,
      "loss": 0.171,
      "num_input_tokens_seen": 51502336,
      "step": 26755
    },
    {
      "epoch": 4.365445794926177,
      "grad_norm": 0.0018405133159831166,
      "learning_rate": 0.07405993829944528,
      "loss": 0.1938,
      "num_input_tokens_seen": 51510960,
      "step": 26760
    },
    {
      "epoch": 4.366261522146994,
      "grad_norm": 0.004411693196743727,
      "learning_rate": 0.07400914597298755,
      "loss": 0.2071,
      "num_input_tokens_seen": 51520352,
      "step": 26765
    },
    {
      "epoch": 4.367077249367812,
      "grad_norm": 0.0021948025096207857,
      "learning_rate": 0.07395836536527445,
      "loss": 0.1536,
      "num_input_tokens_seen": 51530432,
      "step": 26770
    },
    {
      "epoch": 4.367892976588629,
      "grad_norm": 0.003803413128480315,
      "learning_rate": 0.07390759648413696,
      "loss": 0.2406,
      "num_input_tokens_seen": 51539456,
      "step": 26775
    },
    {
      "epoch": 4.368708703809446,
      "grad_norm": 0.0025582972448319197,
      "learning_rate": 0.07385683933740435,
      "loss": 0.1756,
      "num_input_tokens_seen": 51550240,
      "step": 26780
    },
    {
      "epoch": 4.369524431030263,
      "grad_norm": 0.004626155365258455,
      "learning_rate": 0.07380609393290402,
      "loss": 0.2115,
      "num_input_tokens_seen": 51559920,
      "step": 26785
    },
    {
      "epoch": 4.370340158251081,
      "grad_norm": 0.002295738784596324,
      "learning_rate": 0.07375536027846147,
      "loss": 0.159,
      "num_input_tokens_seen": 51568432,
      "step": 26790
    },
    {
      "epoch": 4.371155885471898,
      "grad_norm": 0.0018602850614115596,
      "learning_rate": 0.07370463838190057,
      "loss": 0.1531,
      "num_input_tokens_seen": 51577184,
      "step": 26795
    },
    {
      "epoch": 4.3719716126927155,
      "grad_norm": 0.0035131103359162807,
      "learning_rate": 0.07365392825104317,
      "loss": 0.2334,
      "num_input_tokens_seen": 51585680,
      "step": 26800
    },
    {
      "epoch": 4.3719716126927155,
      "eval_loss": 0.19681048393249512,
      "eval_runtime": 68.1204,
      "eval_samples_per_second": 40.003,
      "eval_steps_per_second": 20.009,
      "num_input_tokens_seen": 51585680,
      "step": 26800
    },
    {
      "epoch": 4.372787339913533,
      "grad_norm": 0.0026303580962121487,
      "learning_rate": 0.07360322989370945,
      "loss": 0.1969,
      "num_input_tokens_seen": 51595104,
      "step": 26805
    },
    {
      "epoch": 4.373603067134351,
      "grad_norm": 0.001882886397652328,
      "learning_rate": 0.07355254331771781,
      "loss": 0.1287,
      "num_input_tokens_seen": 51605376,
      "step": 26810
    },
    {
      "epoch": 4.374418794355168,
      "grad_norm": 0.0036131145898252726,
      "learning_rate": 0.07350186853088461,
      "loss": 0.1694,
      "num_input_tokens_seen": 51614352,
      "step": 26815
    },
    {
      "epoch": 4.375234521575985,
      "grad_norm": 0.0023249902296811342,
      "learning_rate": 0.07345120554102462,
      "loss": 0.1681,
      "num_input_tokens_seen": 51624832,
      "step": 26820
    },
    {
      "epoch": 4.376050248796802,
      "grad_norm": 0.0015980815514922142,
      "learning_rate": 0.07340055435595079,
      "loss": 0.2467,
      "num_input_tokens_seen": 51633616,
      "step": 26825
    },
    {
      "epoch": 4.37686597601762,
      "grad_norm": 0.0037563364021480083,
      "learning_rate": 0.07334991498347401,
      "loss": 0.2051,
      "num_input_tokens_seen": 51643120,
      "step": 26830
    },
    {
      "epoch": 4.377681703238437,
      "grad_norm": 0.002392795169726014,
      "learning_rate": 0.07329928743140365,
      "loss": 0.2068,
      "num_input_tokens_seen": 51653376,
      "step": 26835
    },
    {
      "epoch": 4.378497430459254,
      "grad_norm": 0.0053400639444589615,
      "learning_rate": 0.07324867170754705,
      "loss": 0.2344,
      "num_input_tokens_seen": 51662848,
      "step": 26840
    },
    {
      "epoch": 4.3793131576800715,
      "grad_norm": 0.0032478580251336098,
      "learning_rate": 0.07319806781970974,
      "loss": 0.2383,
      "num_input_tokens_seen": 51673312,
      "step": 26845
    },
    {
      "epoch": 4.3801288849008895,
      "grad_norm": 0.003158979117870331,
      "learning_rate": 0.07314747577569555,
      "loss": 0.2715,
      "num_input_tokens_seen": 51683328,
      "step": 26850
    },
    {
      "epoch": 4.380944612121707,
      "grad_norm": 0.0020195094402879477,
      "learning_rate": 0.07309689558330636,
      "loss": 0.1781,
      "num_input_tokens_seen": 51692432,
      "step": 26855
    },
    {
      "epoch": 4.381760339342524,
      "grad_norm": 0.002522873692214489,
      "learning_rate": 0.0730463272503423,
      "loss": 0.1736,
      "num_input_tokens_seen": 51701760,
      "step": 26860
    },
    {
      "epoch": 4.382576066563341,
      "grad_norm": 0.001217803219333291,
      "learning_rate": 0.07299577078460168,
      "loss": 0.2123,
      "num_input_tokens_seen": 51712000,
      "step": 26865
    },
    {
      "epoch": 4.383391793784159,
      "grad_norm": 0.0031135794706642628,
      "learning_rate": 0.07294522619388083,
      "loss": 0.1973,
      "num_input_tokens_seen": 51721536,
      "step": 26870
    },
    {
      "epoch": 4.384207521004976,
      "grad_norm": 0.0034007993526756763,
      "learning_rate": 0.07289469348597452,
      "loss": 0.1838,
      "num_input_tokens_seen": 51730224,
      "step": 26875
    },
    {
      "epoch": 4.385023248225793,
      "grad_norm": 0.0015086360508576035,
      "learning_rate": 0.07284417266867535,
      "loss": 0.2037,
      "num_input_tokens_seen": 51739376,
      "step": 26880
    },
    {
      "epoch": 4.38583897544661,
      "grad_norm": 0.0027436562813818455,
      "learning_rate": 0.07279366374977439,
      "loss": 0.1889,
      "num_input_tokens_seen": 51749728,
      "step": 26885
    },
    {
      "epoch": 4.386654702667428,
      "grad_norm": 0.003985397983342409,
      "learning_rate": 0.07274316673706074,
      "loss": 0.1607,
      "num_input_tokens_seen": 51759456,
      "step": 26890
    },
    {
      "epoch": 4.387470429888245,
      "grad_norm": 0.0029801749624311924,
      "learning_rate": 0.07269268163832161,
      "loss": 0.1978,
      "num_input_tokens_seen": 51769984,
      "step": 26895
    },
    {
      "epoch": 4.388286157109063,
      "grad_norm": 0.0025570944417268038,
      "learning_rate": 0.07264220846134248,
      "loss": 0.2035,
      "num_input_tokens_seen": 51779344,
      "step": 26900
    },
    {
      "epoch": 4.38910188432988,
      "grad_norm": 0.00250998861156404,
      "learning_rate": 0.07259174721390699,
      "loss": 0.195,
      "num_input_tokens_seen": 51789792,
      "step": 26905
    },
    {
      "epoch": 4.389917611550698,
      "grad_norm": 0.0030877229291945696,
      "learning_rate": 0.07254129790379686,
      "loss": 0.1677,
      "num_input_tokens_seen": 51799136,
      "step": 26910
    },
    {
      "epoch": 4.390733338771515,
      "grad_norm": 0.0027819618117064238,
      "learning_rate": 0.072490860538792,
      "loss": 0.2054,
      "num_input_tokens_seen": 51809168,
      "step": 26915
    },
    {
      "epoch": 4.391549065992332,
      "grad_norm": 0.005148421507328749,
      "learning_rate": 0.07244043512667042,
      "loss": 0.2296,
      "num_input_tokens_seen": 51818160,
      "step": 26920
    },
    {
      "epoch": 4.392364793213149,
      "grad_norm": 0.004157343413680792,
      "learning_rate": 0.07239002167520843,
      "loss": 0.1892,
      "num_input_tokens_seen": 51827280,
      "step": 26925
    },
    {
      "epoch": 4.393180520433967,
      "grad_norm": 0.002947902074083686,
      "learning_rate": 0.07233962019218045,
      "loss": 0.1529,
      "num_input_tokens_seen": 51836048,
      "step": 26930
    },
    {
      "epoch": 4.393996247654784,
      "grad_norm": 0.0019151553278788924,
      "learning_rate": 0.07228923068535892,
      "loss": 0.1464,
      "num_input_tokens_seen": 51845680,
      "step": 26935
    },
    {
      "epoch": 4.394811974875601,
      "grad_norm": 0.0040406351909041405,
      "learning_rate": 0.0722388531625146,
      "loss": 0.1608,
      "num_input_tokens_seen": 51855552,
      "step": 26940
    },
    {
      "epoch": 4.395627702096419,
      "grad_norm": 0.0024288869462907314,
      "learning_rate": 0.07218848763141639,
      "loss": 0.2138,
      "num_input_tokens_seen": 51865856,
      "step": 26945
    },
    {
      "epoch": 4.3964434293172365,
      "grad_norm": 0.005937614478170872,
      "learning_rate": 0.07213813409983118,
      "loss": 0.1872,
      "num_input_tokens_seen": 51874112,
      "step": 26950
    },
    {
      "epoch": 4.397259156538054,
      "grad_norm": 0.002955482807010412,
      "learning_rate": 0.0720877925755242,
      "loss": 0.2718,
      "num_input_tokens_seen": 51883296,
      "step": 26955
    },
    {
      "epoch": 4.398074883758871,
      "grad_norm": 0.003716655308380723,
      "learning_rate": 0.07203746306625866,
      "loss": 0.186,
      "num_input_tokens_seen": 51893776,
      "step": 26960
    },
    {
      "epoch": 4.398890610979688,
      "grad_norm": 0.002717943163588643,
      "learning_rate": 0.07198714557979606,
      "loss": 0.2583,
      "num_input_tokens_seen": 51903776,
      "step": 26965
    },
    {
      "epoch": 4.399706338200506,
      "grad_norm": 0.0031806884799152613,
      "learning_rate": 0.07193684012389602,
      "loss": 0.2095,
      "num_input_tokens_seen": 51911888,
      "step": 26970
    },
    {
      "epoch": 4.400522065421323,
      "grad_norm": 0.0012671489967033267,
      "learning_rate": 0.07188654670631621,
      "loss": 0.1729,
      "num_input_tokens_seen": 51921280,
      "step": 26975
    },
    {
      "epoch": 4.40133779264214,
      "grad_norm": 0.0030327970162034035,
      "learning_rate": 0.07183626533481258,
      "loss": 0.1783,
      "num_input_tokens_seen": 51930896,
      "step": 26980
    },
    {
      "epoch": 4.402153519862958,
      "grad_norm": 0.0036403967533260584,
      "learning_rate": 0.07178599601713909,
      "loss": 0.164,
      "num_input_tokens_seen": 51940560,
      "step": 26985
    },
    {
      "epoch": 4.402969247083775,
      "grad_norm": 0.0036721457727253437,
      "learning_rate": 0.07173573876104786,
      "loss": 0.2011,
      "num_input_tokens_seen": 51950016,
      "step": 26990
    },
    {
      "epoch": 4.4037849743045925,
      "grad_norm": 0.0015828119358047843,
      "learning_rate": 0.0716854935742893,
      "loss": 0.1843,
      "num_input_tokens_seen": 51959568,
      "step": 26995
    },
    {
      "epoch": 4.40460070152541,
      "grad_norm": 0.0027093414682894945,
      "learning_rate": 0.07163526046461174,
      "loss": 0.2112,
      "num_input_tokens_seen": 51969184,
      "step": 27000
    },
    {
      "epoch": 4.40460070152541,
      "eval_loss": 0.19621458649635315,
      "eval_runtime": 68.044,
      "eval_samples_per_second": 40.048,
      "eval_steps_per_second": 20.031,
      "num_input_tokens_seen": 51969184,
      "step": 27000
    },
    {
      "epoch": 4.405416428746228,
      "grad_norm": 0.003032470354810357,
      "learning_rate": 0.07158503943976181,
      "loss": 0.1721,
      "num_input_tokens_seen": 51978944,
      "step": 27005
    },
    {
      "epoch": 4.406232155967045,
      "grad_norm": 0.004094034433364868,
      "learning_rate": 0.07153483050748427,
      "loss": 0.223,
      "num_input_tokens_seen": 51989328,
      "step": 27010
    },
    {
      "epoch": 4.407047883187862,
      "grad_norm": 0.0016443236963823438,
      "learning_rate": 0.07148463367552188,
      "loss": 0.1796,
      "num_input_tokens_seen": 52000128,
      "step": 27015
    },
    {
      "epoch": 4.407863610408679,
      "grad_norm": 0.003043234581127763,
      "learning_rate": 0.07143444895161565,
      "loss": 0.1776,
      "num_input_tokens_seen": 52010432,
      "step": 27020
    },
    {
      "epoch": 4.408679337629497,
      "grad_norm": 0.002135033719241619,
      "learning_rate": 0.07138427634350476,
      "loss": 0.1842,
      "num_input_tokens_seen": 52019696,
      "step": 27025
    },
    {
      "epoch": 4.409495064850314,
      "grad_norm": 0.002003804314881563,
      "learning_rate": 0.07133411585892636,
      "loss": 0.1863,
      "num_input_tokens_seen": 52028384,
      "step": 27030
    },
    {
      "epoch": 4.410310792071131,
      "grad_norm": 0.002182263182476163,
      "learning_rate": 0.07128396750561593,
      "loss": 0.2082,
      "num_input_tokens_seen": 52038720,
      "step": 27035
    },
    {
      "epoch": 4.411126519291948,
      "grad_norm": 0.0020631186198443174,
      "learning_rate": 0.07123383129130685,
      "loss": 0.1323,
      "num_input_tokens_seen": 52047856,
      "step": 27040
    },
    {
      "epoch": 4.4119422465127665,
      "grad_norm": 0.002640196355059743,
      "learning_rate": 0.07118370722373084,
      "loss": 0.1848,
      "num_input_tokens_seen": 52057520,
      "step": 27045
    },
    {
      "epoch": 4.412757973733584,
      "grad_norm": 0.004350035917013884,
      "learning_rate": 0.07113359531061769,
      "loss": 0.1798,
      "num_input_tokens_seen": 52065984,
      "step": 27050
    },
    {
      "epoch": 4.413573700954401,
      "grad_norm": 0.002068519126623869,
      "learning_rate": 0.07108349555969525,
      "loss": 0.1971,
      "num_input_tokens_seen": 52076000,
      "step": 27055
    },
    {
      "epoch": 4.414389428175218,
      "grad_norm": 0.0022528974805027246,
      "learning_rate": 0.07103340797868944,
      "loss": 0.1843,
      "num_input_tokens_seen": 52085424,
      "step": 27060
    },
    {
      "epoch": 4.415205155396036,
      "grad_norm": 0.0014889283338561654,
      "learning_rate": 0.07098333257532453,
      "loss": 0.1317,
      "num_input_tokens_seen": 52094384,
      "step": 27065
    },
    {
      "epoch": 4.416020882616853,
      "grad_norm": 0.004577981308102608,
      "learning_rate": 0.07093326935732269,
      "loss": 0.1682,
      "num_input_tokens_seen": 52104304,
      "step": 27070
    },
    {
      "epoch": 4.41683660983767,
      "grad_norm": 0.003917752765119076,
      "learning_rate": 0.0708832183324044,
      "loss": 0.2453,
      "num_input_tokens_seen": 52114928,
      "step": 27075
    },
    {
      "epoch": 4.417652337058487,
      "grad_norm": 0.0019177765352651477,
      "learning_rate": 0.07083317950828799,
      "loss": 0.2073,
      "num_input_tokens_seen": 52125456,
      "step": 27080
    },
    {
      "epoch": 4.418468064279305,
      "grad_norm": 0.002754881512373686,
      "learning_rate": 0.0707831528926902,
      "loss": 0.1596,
      "num_input_tokens_seen": 52134512,
      "step": 27085
    },
    {
      "epoch": 4.419283791500122,
      "grad_norm": 0.0034852006938308477,
      "learning_rate": 0.07073313849332578,
      "loss": 0.2061,
      "num_input_tokens_seen": 52144816,
      "step": 27090
    },
    {
      "epoch": 4.4200995187209395,
      "grad_norm": 0.00399914151057601,
      "learning_rate": 0.07068313631790749,
      "loss": 0.1881,
      "num_input_tokens_seen": 52153936,
      "step": 27095
    },
    {
      "epoch": 4.420915245941757,
      "grad_norm": 0.003302789991721511,
      "learning_rate": 0.07063314637414632,
      "loss": 0.1783,
      "num_input_tokens_seen": 52164032,
      "step": 27100
    },
    {
      "epoch": 4.421730973162575,
      "grad_norm": 0.0034129186533391476,
      "learning_rate": 0.07058316866975144,
      "loss": 0.1937,
      "num_input_tokens_seen": 52174160,
      "step": 27105
    },
    {
      "epoch": 4.422546700383392,
      "grad_norm": 0.0013630235334858298,
      "learning_rate": 0.0705332032124299,
      "loss": 0.1587,
      "num_input_tokens_seen": 52185008,
      "step": 27110
    },
    {
      "epoch": 4.423362427604209,
      "grad_norm": 0.0022496813908219337,
      "learning_rate": 0.0704832500098871,
      "loss": 0.2034,
      "num_input_tokens_seen": 52194432,
      "step": 27115
    },
    {
      "epoch": 4.424178154825027,
      "grad_norm": 0.002198048634454608,
      "learning_rate": 0.07043330906982641,
      "loss": 0.1876,
      "num_input_tokens_seen": 52204176,
      "step": 27120
    },
    {
      "epoch": 4.424993882045844,
      "grad_norm": 0.0015211635036394,
      "learning_rate": 0.07038338039994936,
      "loss": 0.1793,
      "num_input_tokens_seen": 52214048,
      "step": 27125
    },
    {
      "epoch": 4.425809609266661,
      "grad_norm": 0.0019095534225925803,
      "learning_rate": 0.07033346400795562,
      "loss": 0.2301,
      "num_input_tokens_seen": 52223872,
      "step": 27130
    },
    {
      "epoch": 4.426625336487478,
      "grad_norm": 0.0014045683201402426,
      "learning_rate": 0.07028355990154282,
      "loss": 0.1408,
      "num_input_tokens_seen": 52233376,
      "step": 27135
    },
    {
      "epoch": 4.4274410637082955,
      "grad_norm": 0.0029991702176630497,
      "learning_rate": 0.07023366808840685,
      "loss": 0.2332,
      "num_input_tokens_seen": 52243424,
      "step": 27140
    },
    {
      "epoch": 4.4282567909291135,
      "grad_norm": 0.003328840248286724,
      "learning_rate": 0.07018378857624172,
      "loss": 0.2141,
      "num_input_tokens_seen": 52253408,
      "step": 27145
    },
    {
      "epoch": 4.429072518149931,
      "grad_norm": 0.0029678400605916977,
      "learning_rate": 0.0701339213727394,
      "loss": 0.1539,
      "num_input_tokens_seen": 52261808,
      "step": 27150
    },
    {
      "epoch": 4.429888245370748,
      "grad_norm": 0.002655092626810074,
      "learning_rate": 0.07008406648559008,
      "loss": 0.2306,
      "num_input_tokens_seen": 52271120,
      "step": 27155
    },
    {
      "epoch": 4.430703972591566,
      "grad_norm": 0.0018771998584270477,
      "learning_rate": 0.07003422392248196,
      "loss": 0.144,
      "num_input_tokens_seen": 52280896,
      "step": 27160
    },
    {
      "epoch": 4.431519699812383,
      "grad_norm": 0.00225904886610806,
      "learning_rate": 0.06998439369110142,
      "loss": 0.1241,
      "num_input_tokens_seen": 52289920,
      "step": 27165
    },
    {
      "epoch": 4.4323354270332,
      "grad_norm": 0.0018808071035891771,
      "learning_rate": 0.06993457579913295,
      "loss": 0.1731,
      "num_input_tokens_seen": 52299664,
      "step": 27170
    },
    {
      "epoch": 4.433151154254017,
      "grad_norm": 0.0025173129979521036,
      "learning_rate": 0.06988477025425903,
      "loss": 0.239,
      "num_input_tokens_seen": 52310336,
      "step": 27175
    },
    {
      "epoch": 4.433966881474835,
      "grad_norm": 0.0028262161649763584,
      "learning_rate": 0.06983497706416032,
      "loss": 0.1651,
      "num_input_tokens_seen": 52321552,
      "step": 27180
    },
    {
      "epoch": 4.434782608695652,
      "grad_norm": 0.00340109970420599,
      "learning_rate": 0.0697851962365156,
      "loss": 0.1409,
      "num_input_tokens_seen": 52332352,
      "step": 27185
    },
    {
      "epoch": 4.4355983359164695,
      "grad_norm": 0.003191346535459161,
      "learning_rate": 0.06973542777900163,
      "loss": 0.1716,
      "num_input_tokens_seen": 52342816,
      "step": 27190
    },
    {
      "epoch": 4.436414063137287,
      "grad_norm": 0.0016469741240143776,
      "learning_rate": 0.06968567169929342,
      "loss": 0.1813,
      "num_input_tokens_seen": 52353376,
      "step": 27195
    },
    {
      "epoch": 4.437229790358105,
      "grad_norm": 0.0025168005377054214,
      "learning_rate": 0.06963592800506392,
      "loss": 0.1786,
      "num_input_tokens_seen": 52363216,
      "step": 27200
    },
    {
      "epoch": 4.437229790358105,
      "eval_loss": 0.19892600178718567,
      "eval_runtime": 68.01,
      "eval_samples_per_second": 40.068,
      "eval_steps_per_second": 20.041,
      "num_input_tokens_seen": 52363216,
      "step": 27200
    },
    {
      "epoch": 4.438045517578922,
      "grad_norm": 0.005123560316860676,
      "learning_rate": 0.06958619670398417,
      "loss": 0.2261,
      "num_input_tokens_seen": 52371984,
      "step": 27205
    },
    {
      "epoch": 4.438861244799739,
      "grad_norm": 0.0021122090984135866,
      "learning_rate": 0.0695364778037235,
      "loss": 0.1836,
      "num_input_tokens_seen": 52380368,
      "step": 27210
    },
    {
      "epoch": 4.439676972020556,
      "grad_norm": 0.0027311998419463634,
      "learning_rate": 0.06948677131194907,
      "loss": 0.1639,
      "num_input_tokens_seen": 52389744,
      "step": 27215
    },
    {
      "epoch": 4.440492699241374,
      "grad_norm": 0.003985821269452572,
      "learning_rate": 0.06943707723632629,
      "loss": 0.169,
      "num_input_tokens_seen": 52398784,
      "step": 27220
    },
    {
      "epoch": 4.441308426462191,
      "grad_norm": 0.0028037196025252342,
      "learning_rate": 0.06938739558451867,
      "loss": 0.2054,
      "num_input_tokens_seen": 52409248,
      "step": 27225
    },
    {
      "epoch": 4.442124153683008,
      "grad_norm": 0.0021666765678673983,
      "learning_rate": 0.06933772636418763,
      "loss": 0.1678,
      "num_input_tokens_seen": 52419376,
      "step": 27230
    },
    {
      "epoch": 4.442939880903825,
      "grad_norm": 0.002141084987670183,
      "learning_rate": 0.06928806958299293,
      "loss": 0.1294,
      "num_input_tokens_seen": 52429152,
      "step": 27235
    },
    {
      "epoch": 4.443755608124643,
      "grad_norm": 0.0033340658992528915,
      "learning_rate": 0.06923842524859211,
      "loss": 0.2383,
      "num_input_tokens_seen": 52437760,
      "step": 27240
    },
    {
      "epoch": 4.444571335345461,
      "grad_norm": 0.004702188540250063,
      "learning_rate": 0.06918879336864105,
      "loss": 0.2201,
      "num_input_tokens_seen": 52446736,
      "step": 27245
    },
    {
      "epoch": 4.445387062566278,
      "grad_norm": 0.0036938670091331005,
      "learning_rate": 0.06913917395079362,
      "loss": 0.1916,
      "num_input_tokens_seen": 52456032,
      "step": 27250
    },
    {
      "epoch": 4.446202789787095,
      "grad_norm": 0.002587227150797844,
      "learning_rate": 0.0690895670027017,
      "loss": 0.2191,
      "num_input_tokens_seen": 52465312,
      "step": 27255
    },
    {
      "epoch": 4.447018517007913,
      "grad_norm": 0.005203684791922569,
      "learning_rate": 0.06903997253201531,
      "loss": 0.2182,
      "num_input_tokens_seen": 52474944,
      "step": 27260
    },
    {
      "epoch": 4.44783424422873,
      "grad_norm": 0.004277829546481371,
      "learning_rate": 0.06899039054638263,
      "loss": 0.2341,
      "num_input_tokens_seen": 52484352,
      "step": 27265
    },
    {
      "epoch": 4.448649971449547,
      "grad_norm": 0.002263879869133234,
      "learning_rate": 0.06894082105344976,
      "loss": 0.181,
      "num_input_tokens_seen": 52494464,
      "step": 27270
    },
    {
      "epoch": 4.449465698670364,
      "grad_norm": 0.002688197186216712,
      "learning_rate": 0.06889126406086087,
      "loss": 0.217,
      "num_input_tokens_seen": 52503024,
      "step": 27275
    },
    {
      "epoch": 4.450281425891182,
      "grad_norm": 0.001770448638126254,
      "learning_rate": 0.0688417195762584,
      "loss": 0.1561,
      "num_input_tokens_seen": 52512896,
      "step": 27280
    },
    {
      "epoch": 4.451097153111999,
      "grad_norm": 0.0022799132857471704,
      "learning_rate": 0.06879218760728262,
      "loss": 0.1619,
      "num_input_tokens_seen": 52523072,
      "step": 27285
    },
    {
      "epoch": 4.4519128803328165,
      "grad_norm": 0.0016479695914313197,
      "learning_rate": 0.06874266816157207,
      "loss": 0.2137,
      "num_input_tokens_seen": 52532016,
      "step": 27290
    },
    {
      "epoch": 4.4527286075536345,
      "grad_norm": 0.0023002861998975277,
      "learning_rate": 0.06869316124676321,
      "loss": 0.1857,
      "num_input_tokens_seen": 52541744,
      "step": 27295
    },
    {
      "epoch": 4.453544334774452,
      "grad_norm": 0.003047087462618947,
      "learning_rate": 0.06864366687049062,
      "loss": 0.1912,
      "num_input_tokens_seen": 52551520,
      "step": 27300
    },
    {
      "epoch": 4.454360061995269,
      "grad_norm": 0.0018709116848185658,
      "learning_rate": 0.06859418504038704,
      "loss": 0.1872,
      "num_input_tokens_seen": 52560320,
      "step": 27305
    },
    {
      "epoch": 4.455175789216086,
      "grad_norm": 0.0025100286584347486,
      "learning_rate": 0.06854471576408311,
      "loss": 0.1731,
      "num_input_tokens_seen": 52569408,
      "step": 27310
    },
    {
      "epoch": 4.455991516436903,
      "grad_norm": 0.0012841743882745504,
      "learning_rate": 0.06849525904920767,
      "loss": 0.1716,
      "num_input_tokens_seen": 52579632,
      "step": 27315
    },
    {
      "epoch": 4.456807243657721,
      "grad_norm": 0.002553118159994483,
      "learning_rate": 0.06844581490338748,
      "loss": 0.1524,
      "num_input_tokens_seen": 52588336,
      "step": 27320
    },
    {
      "epoch": 4.457622970878538,
      "grad_norm": 0.002598679857328534,
      "learning_rate": 0.06839638333424752,
      "loss": 0.179,
      "num_input_tokens_seen": 52597840,
      "step": 27325
    },
    {
      "epoch": 4.458438698099355,
      "grad_norm": 0.003241907339543104,
      "learning_rate": 0.06834696434941082,
      "loss": 0.1698,
      "num_input_tokens_seen": 52608624,
      "step": 27330
    },
    {
      "epoch": 4.459254425320173,
      "grad_norm": 0.0021645198576152325,
      "learning_rate": 0.06829755795649824,
      "loss": 0.1544,
      "num_input_tokens_seen": 52617328,
      "step": 27335
    },
    {
      "epoch": 4.4600701525409905,
      "grad_norm": 0.0021733520552515984,
      "learning_rate": 0.06824816416312904,
      "loss": 0.1737,
      "num_input_tokens_seen": 52626912,
      "step": 27340
    },
    {
      "epoch": 4.460885879761808,
      "grad_norm": 0.002069577807560563,
      "learning_rate": 0.06819878297692027,
      "loss": 0.2427,
      "num_input_tokens_seen": 52635184,
      "step": 27345
    },
    {
      "epoch": 4.461701606982625,
      "grad_norm": 0.0036867244634777308,
      "learning_rate": 0.0681494144054871,
      "loss": 0.2442,
      "num_input_tokens_seen": 52643920,
      "step": 27350
    },
    {
      "epoch": 4.462517334203443,
      "grad_norm": 0.002526025753468275,
      "learning_rate": 0.06810005845644286,
      "loss": 0.1969,
      "num_input_tokens_seen": 52652880,
      "step": 27355
    },
    {
      "epoch": 4.46333306142426,
      "grad_norm": 0.0034842730965465307,
      "learning_rate": 0.06805071513739878,
      "loss": 0.1964,
      "num_input_tokens_seen": 52662848,
      "step": 27360
    },
    {
      "epoch": 4.464148788645077,
      "grad_norm": 0.003204033710062504,
      "learning_rate": 0.06800138445596428,
      "loss": 0.1872,
      "num_input_tokens_seen": 52671616,
      "step": 27365
    },
    {
      "epoch": 4.464964515865894,
      "grad_norm": 0.0012397945392876863,
      "learning_rate": 0.06795206641974678,
      "loss": 0.1581,
      "num_input_tokens_seen": 52680992,
      "step": 27370
    },
    {
      "epoch": 4.465780243086712,
      "grad_norm": 0.004164813086390495,
      "learning_rate": 0.06790276103635169,
      "loss": 0.1667,
      "num_input_tokens_seen": 52690128,
      "step": 27375
    },
    {
      "epoch": 4.466595970307529,
      "grad_norm": 0.003166272770613432,
      "learning_rate": 0.0678534683133826,
      "loss": 0.1756,
      "num_input_tokens_seen": 52698144,
      "step": 27380
    },
    {
      "epoch": 4.467411697528346,
      "grad_norm": 0.0008312504505738616,
      "learning_rate": 0.06780418825844095,
      "loss": 0.1257,
      "num_input_tokens_seen": 52706720,
      "step": 27385
    },
    {
      "epoch": 4.468227424749164,
      "grad_norm": 0.002335282741114497,
      "learning_rate": 0.0677549208791264,
      "loss": 0.1835,
      "num_input_tokens_seen": 52717040,
      "step": 27390
    },
    {
      "epoch": 4.469043151969982,
      "grad_norm": 0.0029704091139137745,
      "learning_rate": 0.06770566618303668,
      "loss": 0.1695,
      "num_input_tokens_seen": 52727280,
      "step": 27395
    },
    {
      "epoch": 4.469858879190799,
      "grad_norm": 0.003986184485256672,
      "learning_rate": 0.06765642417776736,
      "loss": 0.2068,
      "num_input_tokens_seen": 52737552,
      "step": 27400
    },
    {
      "epoch": 4.469858879190799,
      "eval_loss": 0.19658809900283813,
      "eval_runtime": 67.9475,
      "eval_samples_per_second": 40.104,
      "eval_steps_per_second": 20.06,
      "num_input_tokens_seen": 52737552,
      "step": 27400
    },
    {
      "epoch": 4.470674606411616,
      "grad_norm": 0.006288312375545502,
      "learning_rate": 0.0676071948709122,
      "loss": 0.3001,
      "num_input_tokens_seen": 52747440,
      "step": 27405
    },
    {
      "epoch": 4.471490333632433,
      "grad_norm": 0.0032432987354695797,
      "learning_rate": 0.06755797827006307,
      "loss": 0.1592,
      "num_input_tokens_seen": 52756400,
      "step": 27410
    },
    {
      "epoch": 4.472306060853251,
      "grad_norm": 0.002131915418431163,
      "learning_rate": 0.06750877438280974,
      "loss": 0.1879,
      "num_input_tokens_seen": 52766176,
      "step": 27415
    },
    {
      "epoch": 4.473121788074068,
      "grad_norm": 0.003941650502383709,
      "learning_rate": 0.06745958321673998,
      "loss": 0.2108,
      "num_input_tokens_seen": 52775952,
      "step": 27420
    },
    {
      "epoch": 4.473937515294885,
      "grad_norm": 0.001844251062721014,
      "learning_rate": 0.0674104047794398,
      "loss": 0.174,
      "num_input_tokens_seen": 52784800,
      "step": 27425
    },
    {
      "epoch": 4.474753242515702,
      "grad_norm": 0.003216993762180209,
      "learning_rate": 0.06736123907849303,
      "loss": 0.2207,
      "num_input_tokens_seen": 52792928,
      "step": 27430
    },
    {
      "epoch": 4.47556896973652,
      "grad_norm": 0.002304649678990245,
      "learning_rate": 0.06731208612148178,
      "loss": 0.2459,
      "num_input_tokens_seen": 52802624,
      "step": 27435
    },
    {
      "epoch": 4.4763846969573375,
      "grad_norm": 0.002960635581985116,
      "learning_rate": 0.0672629459159859,
      "loss": 0.2318,
      "num_input_tokens_seen": 52812832,
      "step": 27440
    },
    {
      "epoch": 4.477200424178155,
      "grad_norm": 0.0032471746671944857,
      "learning_rate": 0.0672138184695835,
      "loss": 0.2332,
      "num_input_tokens_seen": 52821392,
      "step": 27445
    },
    {
      "epoch": 4.478016151398972,
      "grad_norm": 0.002841808134689927,
      "learning_rate": 0.0671647037898507,
      "loss": 0.2249,
      "num_input_tokens_seen": 52830912,
      "step": 27450
    },
    {
      "epoch": 4.47883187861979,
      "grad_norm": 0.004565534647554159,
      "learning_rate": 0.0671156018843615,
      "loss": 0.2184,
      "num_input_tokens_seen": 52840752,
      "step": 27455
    },
    {
      "epoch": 4.479647605840607,
      "grad_norm": 0.001797503442503512,
      "learning_rate": 0.06706651276068812,
      "loss": 0.141,
      "num_input_tokens_seen": 52850720,
      "step": 27460
    },
    {
      "epoch": 4.480463333061424,
      "grad_norm": 0.0021269910503178835,
      "learning_rate": 0.06701743642640064,
      "loss": 0.1616,
      "num_input_tokens_seen": 52860848,
      "step": 27465
    },
    {
      "epoch": 4.481279060282241,
      "grad_norm": 0.0012898050481453538,
      "learning_rate": 0.06696837288906729,
      "loss": 0.1781,
      "num_input_tokens_seen": 52870336,
      "step": 27470
    },
    {
      "epoch": 4.482094787503059,
      "grad_norm": 0.007603366393595934,
      "learning_rate": 0.06691932215625432,
      "loss": 0.1523,
      "num_input_tokens_seen": 52879584,
      "step": 27475
    },
    {
      "epoch": 4.482910514723876,
      "grad_norm": 0.001502207014709711,
      "learning_rate": 0.06687028423552589,
      "loss": 0.1766,
      "num_input_tokens_seen": 52888608,
      "step": 27480
    },
    {
      "epoch": 4.4837262419446935,
      "grad_norm": 0.0033549850340932608,
      "learning_rate": 0.06682125913444435,
      "loss": 0.2184,
      "num_input_tokens_seen": 52896928,
      "step": 27485
    },
    {
      "epoch": 4.484541969165511,
      "grad_norm": 0.001844186568632722,
      "learning_rate": 0.0667722468605699,
      "loss": 0.1937,
      "num_input_tokens_seen": 52905680,
      "step": 27490
    },
    {
      "epoch": 4.485357696386329,
      "grad_norm": 0.002145745325833559,
      "learning_rate": 0.06672324742146094,
      "loss": 0.224,
      "num_input_tokens_seen": 52914448,
      "step": 27495
    },
    {
      "epoch": 4.486173423607146,
      "grad_norm": 0.002960905898362398,
      "learning_rate": 0.06667426082467373,
      "loss": 0.1906,
      "num_input_tokens_seen": 52923904,
      "step": 27500
    },
    {
      "epoch": 4.486989150827963,
      "grad_norm": 0.0014753354480490088,
      "learning_rate": 0.0666252870777626,
      "loss": 0.1973,
      "num_input_tokens_seen": 52932832,
      "step": 27505
    },
    {
      "epoch": 4.487804878048781,
      "grad_norm": 0.0030907178297638893,
      "learning_rate": 0.06657632618827995,
      "loss": 0.2284,
      "num_input_tokens_seen": 52941440,
      "step": 27510
    },
    {
      "epoch": 4.488620605269598,
      "grad_norm": 0.0030846029985696077,
      "learning_rate": 0.06652737816377623,
      "loss": 0.2033,
      "num_input_tokens_seen": 52949488,
      "step": 27515
    },
    {
      "epoch": 4.489436332490415,
      "grad_norm": 0.0025031997356563807,
      "learning_rate": 0.06647844301179971,
      "loss": 0.1504,
      "num_input_tokens_seen": 52959600,
      "step": 27520
    },
    {
      "epoch": 4.490252059711232,
      "grad_norm": 0.001476795645430684,
      "learning_rate": 0.06642952073989689,
      "loss": 0.1413,
      "num_input_tokens_seen": 52970464,
      "step": 27525
    },
    {
      "epoch": 4.49106778693205,
      "grad_norm": 0.002196338027715683,
      "learning_rate": 0.06638061135561223,
      "loss": 0.1883,
      "num_input_tokens_seen": 52979104,
      "step": 27530
    },
    {
      "epoch": 4.4918835141528675,
      "grad_norm": 0.002568236319348216,
      "learning_rate": 0.06633171486648808,
      "loss": 0.1981,
      "num_input_tokens_seen": 52989296,
      "step": 27535
    },
    {
      "epoch": 4.492699241373685,
      "grad_norm": 0.0018965478520840406,
      "learning_rate": 0.06628283128006499,
      "loss": 0.1981,
      "num_input_tokens_seen": 52998352,
      "step": 27540
    },
    {
      "epoch": 4.493514968594502,
      "grad_norm": 0.002292097080498934,
      "learning_rate": 0.0662339606038813,
      "loss": 0.2337,
      "num_input_tokens_seen": 53007632,
      "step": 27545
    },
    {
      "epoch": 4.49433069581532,
      "grad_norm": 0.002278221072629094,
      "learning_rate": 0.06618510284547358,
      "loss": 0.2339,
      "num_input_tokens_seen": 53016992,
      "step": 27550
    },
    {
      "epoch": 4.495146423036137,
      "grad_norm": 0.004853078629821539,
      "learning_rate": 0.06613625801237633,
      "loss": 0.2637,
      "num_input_tokens_seen": 53026032,
      "step": 27555
    },
    {
      "epoch": 4.495962150256954,
      "grad_norm": 0.0018860516138374805,
      "learning_rate": 0.066087426112122,
      "loss": 0.2119,
      "num_input_tokens_seen": 53035296,
      "step": 27560
    },
    {
      "epoch": 4.496777877477771,
      "grad_norm": 0.002954247407615185,
      "learning_rate": 0.06603860715224101,
      "loss": 0.199,
      "num_input_tokens_seen": 53044560,
      "step": 27565
    },
    {
      "epoch": 4.497593604698589,
      "grad_norm": 0.002947040367871523,
      "learning_rate": 0.06598980114026198,
      "loss": 0.1739,
      "num_input_tokens_seen": 53054992,
      "step": 27570
    },
    {
      "epoch": 4.498409331919406,
      "grad_norm": 0.002547762356698513,
      "learning_rate": 0.06594100808371128,
      "loss": 0.2068,
      "num_input_tokens_seen": 53065456,
      "step": 27575
    },
    {
      "epoch": 4.499225059140223,
      "grad_norm": 0.0019913529977202415,
      "learning_rate": 0.06589222799011357,
      "loss": 0.2321,
      "num_input_tokens_seen": 53074800,
      "step": 27580
    },
    {
      "epoch": 4.5000407863610405,
      "grad_norm": 0.003697501728311181,
      "learning_rate": 0.0658434608669912,
      "loss": 0.2359,
      "num_input_tokens_seen": 53084912,
      "step": 27585
    },
    {
      "epoch": 4.500856513581859,
      "grad_norm": 0.0018313712207600474,
      "learning_rate": 0.06579470672186473,
      "loss": 0.163,
      "num_input_tokens_seen": 53094288,
      "step": 27590
    },
    {
      "epoch": 4.501672240802676,
      "grad_norm": 0.002855292521417141,
      "learning_rate": 0.06574596556225275,
      "loss": 0.2133,
      "num_input_tokens_seen": 53103616,
      "step": 27595
    },
    {
      "epoch": 4.502487968023493,
      "grad_norm": 0.0032759010791778564,
      "learning_rate": 0.06569723739567161,
      "loss": 0.1941,
      "num_input_tokens_seen": 53112128,
      "step": 27600
    },
    {
      "epoch": 4.502487968023493,
      "eval_loss": 0.19556502997875214,
      "eval_runtime": 68.1396,
      "eval_samples_per_second": 39.991,
      "eval_steps_per_second": 20.003,
      "num_input_tokens_seen": 53112128,
      "step": 27600
    },
    {
      "epoch": 4.50330369524431,
      "grad_norm": 0.001990116434171796,
      "learning_rate": 0.06564852222963588,
      "loss": 0.1723,
      "num_input_tokens_seen": 53122464,
      "step": 27605
    },
    {
      "epoch": 4.504119422465128,
      "grad_norm": 0.0025777677074074745,
      "learning_rate": 0.06559982007165813,
      "loss": 0.1563,
      "num_input_tokens_seen": 53132464,
      "step": 27610
    },
    {
      "epoch": 4.504935149685945,
      "grad_norm": 0.001247818348929286,
      "learning_rate": 0.06555113092924868,
      "loss": 0.1636,
      "num_input_tokens_seen": 53142384,
      "step": 27615
    },
    {
      "epoch": 4.505750876906762,
      "grad_norm": 0.0012458337005227804,
      "learning_rate": 0.06550245480991615,
      "loss": 0.1455,
      "num_input_tokens_seen": 53151984,
      "step": 27620
    },
    {
      "epoch": 4.506566604127579,
      "grad_norm": 0.0047118342481553555,
      "learning_rate": 0.0654537917211669,
      "loss": 0.2201,
      "num_input_tokens_seen": 53160896,
      "step": 27625
    },
    {
      "epoch": 4.507382331348397,
      "grad_norm": 0.0026039606891572475,
      "learning_rate": 0.0654051416705055,
      "loss": 0.1651,
      "num_input_tokens_seen": 53169792,
      "step": 27630
    },
    {
      "epoch": 4.5081980585692145,
      "grad_norm": 0.001871332060545683,
      "learning_rate": 0.06535650466543427,
      "loss": 0.1789,
      "num_input_tokens_seen": 53178656,
      "step": 27635
    },
    {
      "epoch": 4.509013785790032,
      "grad_norm": 0.0013245994923636317,
      "learning_rate": 0.0653078807134538,
      "loss": 0.1769,
      "num_input_tokens_seen": 53186880,
      "step": 27640
    },
    {
      "epoch": 4.50982951301085,
      "grad_norm": 0.0018105607014149427,
      "learning_rate": 0.06525926982206236,
      "loss": 0.1711,
      "num_input_tokens_seen": 53195440,
      "step": 27645
    },
    {
      "epoch": 4.510645240231667,
      "grad_norm": 0.0031481580808758736,
      "learning_rate": 0.06521067199875648,
      "loss": 0.1665,
      "num_input_tokens_seen": 53205008,
      "step": 27650
    },
    {
      "epoch": 4.511460967452484,
      "grad_norm": 0.0017284691566601396,
      "learning_rate": 0.06516208725103047,
      "loss": 0.185,
      "num_input_tokens_seen": 53213536,
      "step": 27655
    },
    {
      "epoch": 4.512276694673301,
      "grad_norm": 0.002584281377494335,
      "learning_rate": 0.06511351558637678,
      "loss": 0.1926,
      "num_input_tokens_seen": 53222080,
      "step": 27660
    },
    {
      "epoch": 4.513092421894118,
      "grad_norm": 0.004010786302387714,
      "learning_rate": 0.06506495701228569,
      "loss": 0.193,
      "num_input_tokens_seen": 53232000,
      "step": 27665
    },
    {
      "epoch": 4.513908149114936,
      "grad_norm": 0.0020349675323814154,
      "learning_rate": 0.06501641153624559,
      "loss": 0.1493,
      "num_input_tokens_seen": 53240944,
      "step": 27670
    },
    {
      "epoch": 4.514723876335753,
      "grad_norm": 0.0017012242460623384,
      "learning_rate": 0.06496787916574286,
      "loss": 0.2126,
      "num_input_tokens_seen": 53252144,
      "step": 27675
    },
    {
      "epoch": 4.5155396035565705,
      "grad_norm": 0.001324535463936627,
      "learning_rate": 0.06491935990826168,
      "loss": 0.1724,
      "num_input_tokens_seen": 53261360,
      "step": 27680
    },
    {
      "epoch": 4.5163553307773885,
      "grad_norm": 0.0018843024736270308,
      "learning_rate": 0.0648708537712844,
      "loss": 0.1542,
      "num_input_tokens_seen": 53271712,
      "step": 27685
    },
    {
      "epoch": 4.517171057998206,
      "grad_norm": 0.0012087409850209951,
      "learning_rate": 0.06482236076229132,
      "loss": 0.2685,
      "num_input_tokens_seen": 53279984,
      "step": 27690
    },
    {
      "epoch": 4.517986785219023,
      "grad_norm": 0.004038800951093435,
      "learning_rate": 0.06477388088876056,
      "loss": 0.2085,
      "num_input_tokens_seen": 53288880,
      "step": 27695
    },
    {
      "epoch": 4.51880251243984,
      "grad_norm": 0.0026875429321080446,
      "learning_rate": 0.06472541415816846,
      "loss": 0.1899,
      "num_input_tokens_seen": 53298272,
      "step": 27700
    },
    {
      "epoch": 4.519618239660657,
      "grad_norm": 0.0020005092956125736,
      "learning_rate": 0.06467696057798909,
      "loss": 0.2507,
      "num_input_tokens_seen": 53307600,
      "step": 27705
    },
    {
      "epoch": 4.520433966881475,
      "grad_norm": 0.001456179190427065,
      "learning_rate": 0.0646285201556946,
      "loss": 0.1922,
      "num_input_tokens_seen": 53315968,
      "step": 27710
    },
    {
      "epoch": 4.521249694102292,
      "grad_norm": 0.0016788364155218005,
      "learning_rate": 0.06458009289875521,
      "loss": 0.1818,
      "num_input_tokens_seen": 53324864,
      "step": 27715
    },
    {
      "epoch": 4.522065421323109,
      "grad_norm": 0.002673896262422204,
      "learning_rate": 0.0645316788146389,
      "loss": 0.1718,
      "num_input_tokens_seen": 53335360,
      "step": 27720
    },
    {
      "epoch": 4.522881148543927,
      "grad_norm": 0.0025403599720448256,
      "learning_rate": 0.06448327791081175,
      "loss": 0.1726,
      "num_input_tokens_seen": 53344400,
      "step": 27725
    },
    {
      "epoch": 4.523696875764744,
      "grad_norm": 0.0022165291011333466,
      "learning_rate": 0.0644348901947379,
      "loss": 0.1746,
      "num_input_tokens_seen": 53353904,
      "step": 27730
    },
    {
      "epoch": 4.524512602985562,
      "grad_norm": 0.003192533040419221,
      "learning_rate": 0.06438651567387917,
      "loss": 0.1913,
      "num_input_tokens_seen": 53363408,
      "step": 27735
    },
    {
      "epoch": 4.525328330206379,
      "grad_norm": 0.007411435712128878,
      "learning_rate": 0.0643381543556957,
      "loss": 0.2322,
      "num_input_tokens_seen": 53373520,
      "step": 27740
    },
    {
      "epoch": 4.526144057427197,
      "grad_norm": 0.0042326850816607475,
      "learning_rate": 0.06428980624764526,
      "loss": 0.2284,
      "num_input_tokens_seen": 53384256,
      "step": 27745
    },
    {
      "epoch": 4.526959784648014,
      "grad_norm": 0.0023452509194612503,
      "learning_rate": 0.06424147135718378,
      "loss": 0.1828,
      "num_input_tokens_seen": 53393328,
      "step": 27750
    },
    {
      "epoch": 4.527775511868831,
      "grad_norm": 0.0034268798772245646,
      "learning_rate": 0.06419314969176519,
      "loss": 0.198,
      "num_input_tokens_seen": 53402464,
      "step": 27755
    },
    {
      "epoch": 4.528591239089648,
      "grad_norm": 0.0027399430982768536,
      "learning_rate": 0.06414484125884118,
      "loss": 0.1714,
      "num_input_tokens_seen": 53412448,
      "step": 27760
    },
    {
      "epoch": 4.529406966310466,
      "grad_norm": 0.007004948332905769,
      "learning_rate": 0.06409654606586157,
      "loss": 0.1708,
      "num_input_tokens_seen": 53421200,
      "step": 27765
    },
    {
      "epoch": 4.530222693531283,
      "grad_norm": 0.001718056621029973,
      "learning_rate": 0.06404826412027415,
      "loss": 0.1463,
      "num_input_tokens_seen": 53430704,
      "step": 27770
    },
    {
      "epoch": 4.5310384207521,
      "grad_norm": 0.0021670160349458456,
      "learning_rate": 0.06399999542952453,
      "loss": 0.1814,
      "num_input_tokens_seen": 53441264,
      "step": 27775
    },
    {
      "epoch": 4.5318541479729175,
      "grad_norm": 0.002083947416394949,
      "learning_rate": 0.0639517400010563,
      "loss": 0.1973,
      "num_input_tokens_seen": 53451344,
      "step": 27780
    },
    {
      "epoch": 4.5326698751937355,
      "grad_norm": 0.0013165668351575732,
      "learning_rate": 0.06390349784231118,
      "loss": 0.1915,
      "num_input_tokens_seen": 53460576,
      "step": 27785
    },
    {
      "epoch": 4.533485602414553,
      "grad_norm": 0.008207817561924458,
      "learning_rate": 0.06385526896072859,
      "loss": 0.2024,
      "num_input_tokens_seen": 53469968,
      "step": 27790
    },
    {
      "epoch": 4.53430132963537,
      "grad_norm": 0.0026004426181316376,
      "learning_rate": 0.06380705336374613,
      "loss": 0.166,
      "num_input_tokens_seen": 53479840,
      "step": 27795
    },
    {
      "epoch": 4.535117056856187,
      "grad_norm": 0.0026870027650147676,
      "learning_rate": 0.06375885105879918,
      "loss": 0.1346,
      "num_input_tokens_seen": 53489200,
      "step": 27800
    },
    {
      "epoch": 4.535117056856187,
      "eval_loss": 0.19578807055950165,
      "eval_runtime": 67.9877,
      "eval_samples_per_second": 40.081,
      "eval_steps_per_second": 20.048,
      "num_input_tokens_seen": 53489200,
      "step": 27800
    },
    {
      "epoch": 4.535932784077005,
      "grad_norm": 0.0018224819796159863,
      "learning_rate": 0.06371066205332115,
      "loss": 0.142,
      "num_input_tokens_seen": 53498336,
      "step": 27805
    },
    {
      "epoch": 4.536748511297822,
      "grad_norm": 0.0017726437654346228,
      "learning_rate": 0.06366248635474347,
      "loss": 0.2167,
      "num_input_tokens_seen": 53506784,
      "step": 27810
    },
    {
      "epoch": 4.537564238518639,
      "grad_norm": 0.004208351485431194,
      "learning_rate": 0.06361432397049532,
      "loss": 0.2156,
      "num_input_tokens_seen": 53517936,
      "step": 27815
    },
    {
      "epoch": 4.538379965739456,
      "grad_norm": 0.0024457082618027925,
      "learning_rate": 0.06356617490800408,
      "loss": 0.116,
      "num_input_tokens_seen": 53528432,
      "step": 27820
    },
    {
      "epoch": 4.539195692960274,
      "grad_norm": 0.0025565181858837605,
      "learning_rate": 0.06351803917469478,
      "loss": 0.2718,
      "num_input_tokens_seen": 53537824,
      "step": 27825
    },
    {
      "epoch": 4.5400114201810915,
      "grad_norm": 0.002953763585537672,
      "learning_rate": 0.06346991677799067,
      "loss": 0.258,
      "num_input_tokens_seen": 53546608,
      "step": 27830
    },
    {
      "epoch": 4.540827147401909,
      "grad_norm": 0.0023235478438436985,
      "learning_rate": 0.06342180772531283,
      "loss": 0.1612,
      "num_input_tokens_seen": 53556896,
      "step": 27835
    },
    {
      "epoch": 4.541642874622726,
      "grad_norm": 0.0024920981377363205,
      "learning_rate": 0.06337371202408021,
      "loss": 0.2117,
      "num_input_tokens_seen": 53565328,
      "step": 27840
    },
    {
      "epoch": 4.542458601843544,
      "grad_norm": 0.0011947698658332229,
      "learning_rate": 0.06332562968170984,
      "loss": 0.214,
      "num_input_tokens_seen": 53574272,
      "step": 27845
    },
    {
      "epoch": 4.543274329064361,
      "grad_norm": 0.0017072376795113087,
      "learning_rate": 0.06327756070561656,
      "loss": 0.203,
      "num_input_tokens_seen": 53583680,
      "step": 27850
    },
    {
      "epoch": 4.544090056285178,
      "grad_norm": 0.002636371646076441,
      "learning_rate": 0.06322950510321329,
      "loss": 0.181,
      "num_input_tokens_seen": 53593232,
      "step": 27855
    },
    {
      "epoch": 4.544905783505996,
      "grad_norm": 0.0019812965765595436,
      "learning_rate": 0.06318146288191076,
      "loss": 0.1857,
      "num_input_tokens_seen": 53603248,
      "step": 27860
    },
    {
      "epoch": 4.545721510726813,
      "grad_norm": 0.0013595951022580266,
      "learning_rate": 0.06313343404911763,
      "loss": 0.1571,
      "num_input_tokens_seen": 53612656,
      "step": 27865
    },
    {
      "epoch": 4.54653723794763,
      "grad_norm": 0.0018232185393571854,
      "learning_rate": 0.0630854186122406,
      "loss": 0.1838,
      "num_input_tokens_seen": 53622240,
      "step": 27870
    },
    {
      "epoch": 4.5473529651684474,
      "grad_norm": 0.002710913307964802,
      "learning_rate": 0.06303741657868431,
      "loss": 0.2007,
      "num_input_tokens_seen": 53633776,
      "step": 27875
    },
    {
      "epoch": 4.548168692389265,
      "grad_norm": 0.0041793473064899445,
      "learning_rate": 0.06298942795585115,
      "loss": 0.234,
      "num_input_tokens_seen": 53641872,
      "step": 27880
    },
    {
      "epoch": 4.548984419610083,
      "grad_norm": 0.0016623135888949037,
      "learning_rate": 0.06294145275114167,
      "loss": 0.1866,
      "num_input_tokens_seen": 53652256,
      "step": 27885
    },
    {
      "epoch": 4.5498001468309,
      "grad_norm": 0.0022194308694452047,
      "learning_rate": 0.06289349097195428,
      "loss": 0.2079,
      "num_input_tokens_seen": 53661648,
      "step": 27890
    },
    {
      "epoch": 4.550615874051717,
      "grad_norm": 0.0017685136990621686,
      "learning_rate": 0.06284554262568516,
      "loss": 0.2789,
      "num_input_tokens_seen": 53671072,
      "step": 27895
    },
    {
      "epoch": 4.551431601272535,
      "grad_norm": 0.004599795676767826,
      "learning_rate": 0.06279760771972868,
      "loss": 0.1692,
      "num_input_tokens_seen": 53680592,
      "step": 27900
    },
    {
      "epoch": 4.552247328493352,
      "grad_norm": 0.0011049811728298664,
      "learning_rate": 0.06274968626147688,
      "loss": 0.1452,
      "num_input_tokens_seen": 53689568,
      "step": 27905
    },
    {
      "epoch": 4.553063055714169,
      "grad_norm": 0.002779363887384534,
      "learning_rate": 0.06270177825831993,
      "loss": 0.1544,
      "num_input_tokens_seen": 53700096,
      "step": 27910
    },
    {
      "epoch": 4.553878782934986,
      "grad_norm": 0.002466006902977824,
      "learning_rate": 0.06265388371764587,
      "loss": 0.1939,
      "num_input_tokens_seen": 53709776,
      "step": 27915
    },
    {
      "epoch": 4.554694510155803,
      "grad_norm": 0.0029016665648669004,
      "learning_rate": 0.0626060026468406,
      "loss": 0.2338,
      "num_input_tokens_seen": 53718640,
      "step": 27920
    },
    {
      "epoch": 4.555510237376621,
      "grad_norm": 0.001304826932027936,
      "learning_rate": 0.06255813505328794,
      "loss": 0.18,
      "num_input_tokens_seen": 53727536,
      "step": 27925
    },
    {
      "epoch": 4.5563259645974385,
      "grad_norm": 0.0013672936474904418,
      "learning_rate": 0.06251028094436978,
      "loss": 0.1833,
      "num_input_tokens_seen": 53738224,
      "step": 27930
    },
    {
      "epoch": 4.557141691818256,
      "grad_norm": 0.0019871238619089127,
      "learning_rate": 0.06246244032746568,
      "loss": 0.1986,
      "num_input_tokens_seen": 53747168,
      "step": 27935
    },
    {
      "epoch": 4.557957419039074,
      "grad_norm": 0.002139924792572856,
      "learning_rate": 0.06241461320995342,
      "loss": 0.1852,
      "num_input_tokens_seen": 53756656,
      "step": 27940
    },
    {
      "epoch": 4.558773146259891,
      "grad_norm": 0.0027771180029958487,
      "learning_rate": 0.062366799599208426,
      "loss": 0.1979,
      "num_input_tokens_seen": 53764208,
      "step": 27945
    },
    {
      "epoch": 4.559588873480708,
      "grad_norm": 0.00391788175329566,
      "learning_rate": 0.06231899950260418,
      "loss": 0.2509,
      "num_input_tokens_seen": 53773872,
      "step": 27950
    },
    {
      "epoch": 4.560404600701525,
      "grad_norm": 0.0019807873759418726,
      "learning_rate": 0.06227121292751214,
      "loss": 0.1759,
      "num_input_tokens_seen": 53783536,
      "step": 27955
    },
    {
      "epoch": 4.561220327922343,
      "grad_norm": 0.0028622394893318415,
      "learning_rate": 0.062223439881301496,
      "loss": 0.2168,
      "num_input_tokens_seen": 53793872,
      "step": 27960
    },
    {
      "epoch": 4.56203605514316,
      "grad_norm": 0.0024969037622213364,
      "learning_rate": 0.06217568037133948,
      "loss": 0.2093,
      "num_input_tokens_seen": 53803200,
      "step": 27965
    },
    {
      "epoch": 4.562851782363977,
      "grad_norm": 0.0037832939997315407,
      "learning_rate": 0.06212793440499126,
      "loss": 0.2107,
      "num_input_tokens_seen": 53812416,
      "step": 27970
    },
    {
      "epoch": 4.5636675095847945,
      "grad_norm": 0.0019303305307403207,
      "learning_rate": 0.062080201989619783,
      "loss": 0.1974,
      "num_input_tokens_seen": 53822352,
      "step": 27975
    },
    {
      "epoch": 4.5644832368056125,
      "grad_norm": 0.0020366169046610594,
      "learning_rate": 0.062032483132586094,
      "loss": 0.1806,
      "num_input_tokens_seen": 53832448,
      "step": 27980
    },
    {
      "epoch": 4.56529896402643,
      "grad_norm": 0.0008889806922525167,
      "learning_rate": 0.0619847778412489,
      "loss": 0.18,
      "num_input_tokens_seen": 53841264,
      "step": 27985
    },
    {
      "epoch": 4.566114691247247,
      "grad_norm": 0.0025586963165551424,
      "learning_rate": 0.06193708612296509,
      "loss": 0.1868,
      "num_input_tokens_seen": 53850960,
      "step": 27990
    },
    {
      "epoch": 4.566930418468064,
      "grad_norm": 0.003126161638647318,
      "learning_rate": 0.06188940798508923,
      "loss": 0.1928,
      "num_input_tokens_seen": 53860592,
      "step": 27995
    },
    {
      "epoch": 4.567746145688882,
      "grad_norm": 0.007206039968878031,
      "learning_rate": 0.06184174343497397,
      "loss": 0.1824,
      "num_input_tokens_seen": 53870832,
      "step": 28000
    },
    {
      "epoch": 4.567746145688882,
      "eval_loss": 0.1953621208667755,
      "eval_runtime": 67.9764,
      "eval_samples_per_second": 40.087,
      "eval_steps_per_second": 20.051,
      "num_input_tokens_seen": 53870832,
      "step": 28000
    },
    {
      "epoch": 4.568561872909699,
      "grad_norm": 0.001815410447306931,
      "learning_rate": 0.061794092479969726,
      "loss": 0.1628,
      "num_input_tokens_seen": 53879408,
      "step": 28005
    },
    {
      "epoch": 4.569377600130516,
      "grad_norm": 0.0027751876041293144,
      "learning_rate": 0.06174645512742485,
      "loss": 0.1594,
      "num_input_tokens_seen": 53889104,
      "step": 28010
    },
    {
      "epoch": 4.570193327351333,
      "grad_norm": 0.0012544842902570963,
      "learning_rate": 0.06169883138468565,
      "loss": 0.1572,
      "num_input_tokens_seen": 53899424,
      "step": 28015
    },
    {
      "epoch": 4.571009054572151,
      "grad_norm": 0.0022865342907607555,
      "learning_rate": 0.06165122125909637,
      "loss": 0.1822,
      "num_input_tokens_seen": 53909056,
      "step": 28020
    },
    {
      "epoch": 4.5718247817929685,
      "grad_norm": 0.003923430573195219,
      "learning_rate": 0.061603624757998965,
      "loss": 0.1927,
      "num_input_tokens_seen": 53919744,
      "step": 28025
    },
    {
      "epoch": 4.572640509013786,
      "grad_norm": 0.002997263567522168,
      "learning_rate": 0.0615560418887335,
      "loss": 0.2397,
      "num_input_tokens_seen": 53929888,
      "step": 28030
    },
    {
      "epoch": 4.573456236234604,
      "grad_norm": 0.003866825718432665,
      "learning_rate": 0.06150847265863787,
      "loss": 0.1751,
      "num_input_tokens_seen": 53940144,
      "step": 28035
    },
    {
      "epoch": 4.574271963455421,
      "grad_norm": 0.004648393485695124,
      "learning_rate": 0.061460917075047757,
      "loss": 0.2437,
      "num_input_tokens_seen": 53949200,
      "step": 28040
    },
    {
      "epoch": 4.575087690676238,
      "grad_norm": 0.002267837757244706,
      "learning_rate": 0.06141337514529694,
      "loss": 0.2027,
      "num_input_tokens_seen": 53958256,
      "step": 28045
    },
    {
      "epoch": 4.575903417897055,
      "grad_norm": 0.0016653429483994842,
      "learning_rate": 0.06136584687671687,
      "loss": 0.155,
      "num_input_tokens_seen": 53967744,
      "step": 28050
    },
    {
      "epoch": 4.576719145117872,
      "grad_norm": 0.0027707829140126705,
      "learning_rate": 0.061318332276637064,
      "loss": 0.1931,
      "num_input_tokens_seen": 53976096,
      "step": 28055
    },
    {
      "epoch": 4.57753487233869,
      "grad_norm": 0.0035877074114978313,
      "learning_rate": 0.06127083135238491,
      "loss": 0.2351,
      "num_input_tokens_seen": 53986784,
      "step": 28060
    },
    {
      "epoch": 4.578350599559507,
      "grad_norm": 0.0015553053235635161,
      "learning_rate": 0.06122334411128555,
      "loss": 0.2037,
      "num_input_tokens_seen": 53996640,
      "step": 28065
    },
    {
      "epoch": 4.579166326780324,
      "grad_norm": 0.0019098474876955152,
      "learning_rate": 0.06117587056066223,
      "loss": 0.1811,
      "num_input_tokens_seen": 54005072,
      "step": 28070
    },
    {
      "epoch": 4.5799820540011424,
      "grad_norm": 0.002807003678753972,
      "learning_rate": 0.06112841070783589,
      "loss": 0.1993,
      "num_input_tokens_seen": 54015072,
      "step": 28075
    },
    {
      "epoch": 4.58079778122196,
      "grad_norm": 0.0015671143773943186,
      "learning_rate": 0.061080964560125406,
      "loss": 0.2021,
      "num_input_tokens_seen": 54023424,
      "step": 28080
    },
    {
      "epoch": 4.581613508442777,
      "grad_norm": 0.002620474901050329,
      "learning_rate": 0.06103353212484766,
      "loss": 0.2112,
      "num_input_tokens_seen": 54031712,
      "step": 28085
    },
    {
      "epoch": 4.582429235663594,
      "grad_norm": 0.002661012811586261,
      "learning_rate": 0.06098611340931722,
      "loss": 0.185,
      "num_input_tokens_seen": 54040672,
      "step": 28090
    },
    {
      "epoch": 4.583244962884411,
      "grad_norm": 0.004079767968505621,
      "learning_rate": 0.06093870842084672,
      "loss": 0.18,
      "num_input_tokens_seen": 54052528,
      "step": 28095
    },
    {
      "epoch": 4.584060690105229,
      "grad_norm": 0.0017786315875127912,
      "learning_rate": 0.06089131716674666,
      "loss": 0.1869,
      "num_input_tokens_seen": 54061328,
      "step": 28100
    },
    {
      "epoch": 4.584876417326046,
      "grad_norm": 0.0027488127816468477,
      "learning_rate": 0.060843939654325226,
      "loss": 0.1879,
      "num_input_tokens_seen": 54071216,
      "step": 28105
    },
    {
      "epoch": 4.585692144546863,
      "grad_norm": 0.0012947047362104058,
      "learning_rate": 0.06079657589088873,
      "loss": 0.1555,
      "num_input_tokens_seen": 54081216,
      "step": 28110
    },
    {
      "epoch": 4.586507871767681,
      "grad_norm": 0.0029538809321820736,
      "learning_rate": 0.06074922588374126,
      "loss": 0.1999,
      "num_input_tokens_seen": 54092064,
      "step": 28115
    },
    {
      "epoch": 4.587323598988498,
      "grad_norm": 0.003092455444857478,
      "learning_rate": 0.06070188964018472,
      "loss": 0.1943,
      "num_input_tokens_seen": 54102112,
      "step": 28120
    },
    {
      "epoch": 4.5881393262093155,
      "grad_norm": 0.001160339917987585,
      "learning_rate": 0.06065456716751902,
      "loss": 0.1835,
      "num_input_tokens_seen": 54112656,
      "step": 28125
    },
    {
      "epoch": 4.588955053430133,
      "grad_norm": 0.0032197076361626387,
      "learning_rate": 0.06060725847304182,
      "loss": 0.152,
      "num_input_tokens_seen": 54122256,
      "step": 28130
    },
    {
      "epoch": 4.589770780650951,
      "grad_norm": 0.0021951242815703154,
      "learning_rate": 0.06055996356404877,
      "loss": 0.1715,
      "num_input_tokens_seen": 54131984,
      "step": 28135
    },
    {
      "epoch": 4.590586507871768,
      "grad_norm": 0.0023734706919640303,
      "learning_rate": 0.06051268244783327,
      "loss": 0.228,
      "num_input_tokens_seen": 54142000,
      "step": 28140
    },
    {
      "epoch": 4.591402235092585,
      "grad_norm": 0.003333666129037738,
      "learning_rate": 0.06046541513168676,
      "loss": 0.1919,
      "num_input_tokens_seen": 54151120,
      "step": 28145
    },
    {
      "epoch": 4.592217962313402,
      "grad_norm": 0.0020852871239185333,
      "learning_rate": 0.060418161622898356,
      "loss": 0.2009,
      "num_input_tokens_seen": 54161328,
      "step": 28150
    },
    {
      "epoch": 4.59303368953422,
      "grad_norm": 0.003258774522691965,
      "learning_rate": 0.06037092192875521,
      "loss": 0.1934,
      "num_input_tokens_seen": 54171760,
      "step": 28155
    },
    {
      "epoch": 4.593849416755037,
      "grad_norm": 0.003058698959648609,
      "learning_rate": 0.060323696056542225,
      "loss": 0.2022,
      "num_input_tokens_seen": 54181872,
      "step": 28160
    },
    {
      "epoch": 4.594665143975854,
      "grad_norm": 0.0029128682799637318,
      "learning_rate": 0.06027648401354229,
      "loss": 0.1897,
      "num_input_tokens_seen": 54192608,
      "step": 28165
    },
    {
      "epoch": 4.5954808711966715,
      "grad_norm": 0.001234238501638174,
      "learning_rate": 0.06022928580703601,
      "loss": 0.211,
      "num_input_tokens_seen": 54202816,
      "step": 28170
    },
    {
      "epoch": 4.5962965984174895,
      "grad_norm": 0.0021214664448052645,
      "learning_rate": 0.060182101444301986,
      "loss": 0.1327,
      "num_input_tokens_seen": 54212816,
      "step": 28175
    },
    {
      "epoch": 4.597112325638307,
      "grad_norm": 0.010391674004495144,
      "learning_rate": 0.06013493093261669,
      "loss": 0.2645,
      "num_input_tokens_seen": 54222032,
      "step": 28180
    },
    {
      "epoch": 4.597928052859124,
      "grad_norm": 0.0037405153270810843,
      "learning_rate": 0.06008777427925432,
      "loss": 0.2882,
      "num_input_tokens_seen": 54231872,
      "step": 28185
    },
    {
      "epoch": 4.598743780079941,
      "grad_norm": 0.002407464198768139,
      "learning_rate": 0.06004063149148705,
      "loss": 0.1915,
      "num_input_tokens_seen": 54241296,
      "step": 28190
    },
    {
      "epoch": 4.599559507300759,
      "grad_norm": 0.0029563638381659985,
      "learning_rate": 0.05999350257658497,
      "loss": 0.2251,
      "num_input_tokens_seen": 54251024,
      "step": 28195
    },
    {
      "epoch": 4.600375234521576,
      "grad_norm": 0.002273564226925373,
      "learning_rate": 0.05994638754181582,
      "loss": 0.1838,
      "num_input_tokens_seen": 54260848,
      "step": 28200
    },
    {
      "epoch": 4.600375234521576,
      "eval_loss": 0.19505779445171356,
      "eval_runtime": 68.0121,
      "eval_samples_per_second": 40.066,
      "eval_steps_per_second": 20.041,
      "num_input_tokens_seen": 54260848,
      "step": 28200
    },
    {
      "epoch": 4.601190961742393,
      "grad_norm": 0.002530493773519993,
      "learning_rate": 0.059899286394445445,
      "loss": 0.1795,
      "num_input_tokens_seen": 54270624,
      "step": 28205
    },
    {
      "epoch": 4.602006688963211,
      "grad_norm": 0.002403340069577098,
      "learning_rate": 0.059852199141737346,
      "loss": 0.2231,
      "num_input_tokens_seen": 54280048,
      "step": 28210
    },
    {
      "epoch": 4.602822416184028,
      "grad_norm": 0.002773628570139408,
      "learning_rate": 0.05980512579095304,
      "loss": 0.2017,
      "num_input_tokens_seen": 54290096,
      "step": 28215
    },
    {
      "epoch": 4.6036381434048455,
      "grad_norm": 0.0019625944551080465,
      "learning_rate": 0.05975806634935181,
      "loss": 0.1683,
      "num_input_tokens_seen": 54299120,
      "step": 28220
    },
    {
      "epoch": 4.604453870625663,
      "grad_norm": 0.0014749945839866996,
      "learning_rate": 0.05971102082419076,
      "loss": 0.1527,
      "num_input_tokens_seen": 54309840,
      "step": 28225
    },
    {
      "epoch": 4.60526959784648,
      "grad_norm": 0.002730217296630144,
      "learning_rate": 0.05966398922272492,
      "loss": 0.1926,
      "num_input_tokens_seen": 54319200,
      "step": 28230
    },
    {
      "epoch": 4.606085325067298,
      "grad_norm": 0.0014261400792747736,
      "learning_rate": 0.059616971552207236,
      "loss": 0.1817,
      "num_input_tokens_seen": 54327520,
      "step": 28235
    },
    {
      "epoch": 4.606901052288115,
      "grad_norm": 0.001943122362717986,
      "learning_rate": 0.059569967819888305,
      "loss": 0.1598,
      "num_input_tokens_seen": 54337792,
      "step": 28240
    },
    {
      "epoch": 4.607716779508932,
      "grad_norm": 0.0023357777390629053,
      "learning_rate": 0.05952297803301681,
      "loss": 0.1849,
      "num_input_tokens_seen": 54348080,
      "step": 28245
    },
    {
      "epoch": 4.60853250672975,
      "grad_norm": 0.001006350968964398,
      "learning_rate": 0.059476002198839056,
      "loss": 0.17,
      "num_input_tokens_seen": 54357680,
      "step": 28250
    },
    {
      "epoch": 4.609348233950567,
      "grad_norm": 0.0017851609736680984,
      "learning_rate": 0.05942904032459935,
      "loss": 0.1718,
      "num_input_tokens_seen": 54367744,
      "step": 28255
    },
    {
      "epoch": 4.610163961171384,
      "grad_norm": 0.0041318489238619804,
      "learning_rate": 0.05938209241753987,
      "loss": 0.1866,
      "num_input_tokens_seen": 54376320,
      "step": 28260
    },
    {
      "epoch": 4.610979688392201,
      "grad_norm": 0.0025974682066589594,
      "learning_rate": 0.05933515848490046,
      "loss": 0.1373,
      "num_input_tokens_seen": 54385648,
      "step": 28265
    },
    {
      "epoch": 4.6117954156130185,
      "grad_norm": 0.002025559777393937,
      "learning_rate": 0.059288238533918985,
      "loss": 0.1652,
      "num_input_tokens_seen": 54394256,
      "step": 28270
    },
    {
      "epoch": 4.6126111428338366,
      "grad_norm": 0.002882466185837984,
      "learning_rate": 0.05924133257183113,
      "loss": 0.1683,
      "num_input_tokens_seen": 54403248,
      "step": 28275
    },
    {
      "epoch": 4.613426870054654,
      "grad_norm": 0.005441322457045317,
      "learning_rate": 0.059194440605870285,
      "loss": 0.1885,
      "num_input_tokens_seen": 54413248,
      "step": 28280
    },
    {
      "epoch": 4.614242597275471,
      "grad_norm": 0.0043938662856817245,
      "learning_rate": 0.059147562643267884,
      "loss": 0.2475,
      "num_input_tokens_seen": 54424160,
      "step": 28285
    },
    {
      "epoch": 4.615058324496289,
      "grad_norm": 0.0014238612493500113,
      "learning_rate": 0.059100698691253055,
      "loss": 0.2271,
      "num_input_tokens_seen": 54433344,
      "step": 28290
    },
    {
      "epoch": 4.615874051717106,
      "grad_norm": 0.005071574356406927,
      "learning_rate": 0.05905384875705273,
      "loss": 0.2178,
      "num_input_tokens_seen": 54442320,
      "step": 28295
    },
    {
      "epoch": 4.616689778937923,
      "grad_norm": 0.002015179954469204,
      "learning_rate": 0.05900701284789189,
      "loss": 0.157,
      "num_input_tokens_seen": 54451360,
      "step": 28300
    },
    {
      "epoch": 4.61750550615874,
      "grad_norm": 0.0019585497211664915,
      "learning_rate": 0.058960190970993115,
      "loss": 0.156,
      "num_input_tokens_seen": 54461216,
      "step": 28305
    },
    {
      "epoch": 4.618321233379558,
      "grad_norm": 0.00393232237547636,
      "learning_rate": 0.058913383133576955,
      "loss": 0.2777,
      "num_input_tokens_seen": 54471824,
      "step": 28310
    },
    {
      "epoch": 4.619136960600375,
      "grad_norm": 0.002367655746638775,
      "learning_rate": 0.05886658934286185,
      "loss": 0.2083,
      "num_input_tokens_seen": 54481120,
      "step": 28315
    },
    {
      "epoch": 4.6199526878211925,
      "grad_norm": 0.0023171622306108475,
      "learning_rate": 0.058819809606063846,
      "loss": 0.208,
      "num_input_tokens_seen": 54491712,
      "step": 28320
    },
    {
      "epoch": 4.62076841504201,
      "grad_norm": 0.0037362955044955015,
      "learning_rate": 0.05877304393039711,
      "loss": 0.2182,
      "num_input_tokens_seen": 54501488,
      "step": 28325
    },
    {
      "epoch": 4.621584142262828,
      "grad_norm": 0.003007179358974099,
      "learning_rate": 0.05872629232307338,
      "loss": 0.2429,
      "num_input_tokens_seen": 54512544,
      "step": 28330
    },
    {
      "epoch": 4.622399869483645,
      "grad_norm": 0.002673984970897436,
      "learning_rate": 0.05867955479130239,
      "loss": 0.2067,
      "num_input_tokens_seen": 54522240,
      "step": 28335
    },
    {
      "epoch": 4.623215596704462,
      "grad_norm": 0.00383394374512136,
      "learning_rate": 0.058632831342291705,
      "loss": 0.2165,
      "num_input_tokens_seen": 54531872,
      "step": 28340
    },
    {
      "epoch": 4.624031323925279,
      "grad_norm": 0.0030590358655899763,
      "learning_rate": 0.05858612198324655,
      "loss": 0.2121,
      "num_input_tokens_seen": 54541264,
      "step": 28345
    },
    {
      "epoch": 4.624847051146097,
      "grad_norm": 0.0015448490157723427,
      "learning_rate": 0.05853942672137025,
      "loss": 0.1918,
      "num_input_tokens_seen": 54549920,
      "step": 28350
    },
    {
      "epoch": 4.625662778366914,
      "grad_norm": 0.0024710886646062136,
      "learning_rate": 0.05849274556386363,
      "loss": 0.1677,
      "num_input_tokens_seen": 54558832,
      "step": 28355
    },
    {
      "epoch": 4.626478505587731,
      "grad_norm": 0.003954332787543535,
      "learning_rate": 0.05844607851792567,
      "loss": 0.21,
      "num_input_tokens_seen": 54568864,
      "step": 28360
    },
    {
      "epoch": 4.6272942328085485,
      "grad_norm": 0.0018655653111636639,
      "learning_rate": 0.058399425590752924,
      "loss": 0.1604,
      "num_input_tokens_seen": 54578144,
      "step": 28365
    },
    {
      "epoch": 4.6281099600293665,
      "grad_norm": 0.0017605593893676996,
      "learning_rate": 0.05835278678953985,
      "loss": 0.1329,
      "num_input_tokens_seen": 54588400,
      "step": 28370
    },
    {
      "epoch": 4.628925687250184,
      "grad_norm": 0.0036044088192284107,
      "learning_rate": 0.05830616212147874,
      "loss": 0.2208,
      "num_input_tokens_seen": 54598464,
      "step": 28375
    },
    {
      "epoch": 4.629741414471001,
      "grad_norm": 0.0028181427624076605,
      "learning_rate": 0.058259551593759784,
      "loss": 0.1746,
      "num_input_tokens_seen": 54608528,
      "step": 28380
    },
    {
      "epoch": 4.630557141691818,
      "grad_norm": 0.0023191734217107296,
      "learning_rate": 0.058212955213570804,
      "loss": 0.2512,
      "num_input_tokens_seen": 54618032,
      "step": 28385
    },
    {
      "epoch": 4.631372868912636,
      "grad_norm": 0.002106662606820464,
      "learning_rate": 0.0581663729880976,
      "loss": 0.1613,
      "num_input_tokens_seen": 54628608,
      "step": 28390
    },
    {
      "epoch": 4.632188596133453,
      "grad_norm": 0.003630736842751503,
      "learning_rate": 0.05811980492452379,
      "loss": 0.2027,
      "num_input_tokens_seen": 54638432,
      "step": 28395
    },
    {
      "epoch": 4.63300432335427,
      "grad_norm": 0.0022633604239672422,
      "learning_rate": 0.058073251030030644,
      "loss": 0.2061,
      "num_input_tokens_seen": 54647840,
      "step": 28400
    },
    {
      "epoch": 4.63300432335427,
      "eval_loss": 0.19611786305904388,
      "eval_runtime": 68.025,
      "eval_samples_per_second": 40.059,
      "eval_steps_per_second": 20.037,
      "num_input_tokens_seen": 54647840,
      "step": 28400
    },
    {
      "epoch": 4.633820050575087,
      "grad_norm": 0.0038421531207859516,
      "learning_rate": 0.05802671131179747,
      "loss": 0.21,
      "num_input_tokens_seen": 54656816,
      "step": 28405
    },
    {
      "epoch": 4.634635777795905,
      "grad_norm": 0.00454589631408453,
      "learning_rate": 0.057980185777001154,
      "loss": 0.2411,
      "num_input_tokens_seen": 54666112,
      "step": 28410
    },
    {
      "epoch": 4.635451505016722,
      "grad_norm": 0.0014142525615170598,
      "learning_rate": 0.057933674432816606,
      "loss": 0.1606,
      "num_input_tokens_seen": 54675440,
      "step": 28415
    },
    {
      "epoch": 4.63626723223754,
      "grad_norm": 0.002435361035168171,
      "learning_rate": 0.05788717728641648,
      "loss": 0.1833,
      "num_input_tokens_seen": 54685392,
      "step": 28420
    },
    {
      "epoch": 4.637082959458358,
      "grad_norm": 0.003714477177709341,
      "learning_rate": 0.057840694344971126,
      "loss": 0.1995,
      "num_input_tokens_seen": 54696320,
      "step": 28425
    },
    {
      "epoch": 4.637898686679175,
      "grad_norm": 0.0029366069938987494,
      "learning_rate": 0.0577942256156489,
      "loss": 0.1372,
      "num_input_tokens_seen": 54705904,
      "step": 28430
    },
    {
      "epoch": 4.638714413899992,
      "grad_norm": 0.0020169769413769245,
      "learning_rate": 0.057747771105615804,
      "loss": 0.1684,
      "num_input_tokens_seen": 54716016,
      "step": 28435
    },
    {
      "epoch": 4.639530141120809,
      "grad_norm": 0.001829913118854165,
      "learning_rate": 0.05770133082203568,
      "loss": 0.2246,
      "num_input_tokens_seen": 54725968,
      "step": 28440
    },
    {
      "epoch": 4.640345868341626,
      "grad_norm": 0.002441260265186429,
      "learning_rate": 0.0576549047720703,
      "loss": 0.2183,
      "num_input_tokens_seen": 54735152,
      "step": 28445
    },
    {
      "epoch": 4.641161595562444,
      "grad_norm": 0.0011196862906217575,
      "learning_rate": 0.05760849296287902,
      "loss": 0.1851,
      "num_input_tokens_seen": 54744192,
      "step": 28450
    },
    {
      "epoch": 4.641977322783261,
      "grad_norm": 0.0018318374641239643,
      "learning_rate": 0.05756209540161919,
      "loss": 0.2261,
      "num_input_tokens_seen": 54753728,
      "step": 28455
    },
    {
      "epoch": 4.642793050004078,
      "grad_norm": 0.002367221750319004,
      "learning_rate": 0.05751571209544595,
      "loss": 0.2004,
      "num_input_tokens_seen": 54761808,
      "step": 28460
    },
    {
      "epoch": 4.643608777224896,
      "grad_norm": 0.003735046833753586,
      "learning_rate": 0.057469343051512085,
      "loss": 0.1577,
      "num_input_tokens_seen": 54771936,
      "step": 28465
    },
    {
      "epoch": 4.6444245044457135,
      "grad_norm": 0.0035460167564451694,
      "learning_rate": 0.057422988276968324,
      "loss": 0.2082,
      "num_input_tokens_seen": 54782032,
      "step": 28470
    },
    {
      "epoch": 4.645240231666531,
      "grad_norm": 0.0031799371354281902,
      "learning_rate": 0.05737664777896323,
      "loss": 0.1895,
      "num_input_tokens_seen": 54792112,
      "step": 28475
    },
    {
      "epoch": 4.646055958887348,
      "grad_norm": 0.0013360659359022975,
      "learning_rate": 0.057330321564642975,
      "loss": 0.201,
      "num_input_tokens_seen": 54802160,
      "step": 28480
    },
    {
      "epoch": 4.646871686108166,
      "grad_norm": 0.004135674796998501,
      "learning_rate": 0.05728400964115174,
      "loss": 0.1972,
      "num_input_tokens_seen": 54810352,
      "step": 28485
    },
    {
      "epoch": 4.647687413328983,
      "grad_norm": 0.002783099887892604,
      "learning_rate": 0.057237712015631305,
      "loss": 0.2292,
      "num_input_tokens_seen": 54819616,
      "step": 28490
    },
    {
      "epoch": 4.6485031405498,
      "grad_norm": 0.0011773555306717753,
      "learning_rate": 0.057191428695221425,
      "loss": 0.1552,
      "num_input_tokens_seen": 54829600,
      "step": 28495
    },
    {
      "epoch": 4.649318867770617,
      "grad_norm": 0.003182956948876381,
      "learning_rate": 0.05714515968705958,
      "loss": 0.1874,
      "num_input_tokens_seen": 54839872,
      "step": 28500
    },
    {
      "epoch": 4.650134594991435,
      "grad_norm": 0.0013738460838794708,
      "learning_rate": 0.05709890499828099,
      "loss": 0.1744,
      "num_input_tokens_seen": 54849568,
      "step": 28505
    },
    {
      "epoch": 4.650950322212252,
      "grad_norm": 0.0019051054259762168,
      "learning_rate": 0.05705266463601868,
      "loss": 0.1328,
      "num_input_tokens_seen": 54860304,
      "step": 28510
    },
    {
      "epoch": 4.6517660494330695,
      "grad_norm": 0.001969793578609824,
      "learning_rate": 0.057006438607403565,
      "loss": 0.1792,
      "num_input_tokens_seen": 54870512,
      "step": 28515
    },
    {
      "epoch": 4.652581776653887,
      "grad_norm": 0.0029621622525155544,
      "learning_rate": 0.056960226919564205,
      "loss": 0.1415,
      "num_input_tokens_seen": 54879856,
      "step": 28520
    },
    {
      "epoch": 4.653397503874705,
      "grad_norm": 0.00620528357103467,
      "learning_rate": 0.05691402957962713,
      "loss": 0.3,
      "num_input_tokens_seen": 54888720,
      "step": 28525
    },
    {
      "epoch": 4.654213231095522,
      "grad_norm": 0.0022774185054004192,
      "learning_rate": 0.05686784659471642,
      "loss": 0.2438,
      "num_input_tokens_seen": 54898352,
      "step": 28530
    },
    {
      "epoch": 4.655028958316339,
      "grad_norm": 0.0023988576140254736,
      "learning_rate": 0.056821677971954136,
      "loss": 0.2005,
      "num_input_tokens_seen": 54908288,
      "step": 28535
    },
    {
      "epoch": 4.655844685537156,
      "grad_norm": 0.0016234347131103277,
      "learning_rate": 0.05677552371846012,
      "loss": 0.2181,
      "num_input_tokens_seen": 54917840,
      "step": 28540
    },
    {
      "epoch": 4.656660412757974,
      "grad_norm": 0.0010823842603713274,
      "learning_rate": 0.05672938384135182,
      "loss": 0.1557,
      "num_input_tokens_seen": 54927968,
      "step": 28545
    },
    {
      "epoch": 4.657476139978791,
      "grad_norm": 0.0027122378814965487,
      "learning_rate": 0.05668325834774465,
      "loss": 0.2055,
      "num_input_tokens_seen": 54938096,
      "step": 28550
    },
    {
      "epoch": 4.658291867199608,
      "grad_norm": 0.003432312747463584,
      "learning_rate": 0.05663714724475177,
      "loss": 0.2532,
      "num_input_tokens_seen": 54948256,
      "step": 28555
    },
    {
      "epoch": 4.659107594420425,
      "grad_norm": 0.002171639120206237,
      "learning_rate": 0.05659105053948403,
      "loss": 0.2096,
      "num_input_tokens_seen": 54957920,
      "step": 28560
    },
    {
      "epoch": 4.6599233216412435,
      "grad_norm": 0.0031667405273765326,
      "learning_rate": 0.056544968239050176,
      "loss": 0.2397,
      "num_input_tokens_seen": 54967296,
      "step": 28565
    },
    {
      "epoch": 4.660739048862061,
      "grad_norm": 0.0023264118935912848,
      "learning_rate": 0.056498900350556616,
      "loss": 0.181,
      "num_input_tokens_seen": 54977376,
      "step": 28570
    },
    {
      "epoch": 4.661554776082878,
      "grad_norm": 0.0023707901127636433,
      "learning_rate": 0.05645284688110766,
      "loss": 0.1782,
      "num_input_tokens_seen": 54986688,
      "step": 28575
    },
    {
      "epoch": 4.662370503303695,
      "grad_norm": 0.0017324336804449558,
      "learning_rate": 0.05640680783780532,
      "loss": 0.1967,
      "num_input_tokens_seen": 54996000,
      "step": 28580
    },
    {
      "epoch": 4.663186230524513,
      "grad_norm": 0.0013342654565349221,
      "learning_rate": 0.056360783227749324,
      "loss": 0.1839,
      "num_input_tokens_seen": 55006080,
      "step": 28585
    },
    {
      "epoch": 4.66400195774533,
      "grad_norm": 0.0017811943544074893,
      "learning_rate": 0.05631477305803728,
      "loss": 0.1698,
      "num_input_tokens_seen": 55015440,
      "step": 28590
    },
    {
      "epoch": 4.664817684966147,
      "grad_norm": 0.003001485951244831,
      "learning_rate": 0.05626877733576462,
      "loss": 0.1971,
      "num_input_tokens_seen": 55025440,
      "step": 28595
    },
    {
      "epoch": 4.665633412186965,
      "grad_norm": 0.003019707277417183,
      "learning_rate": 0.05622279606802435,
      "loss": 0.2111,
      "num_input_tokens_seen": 55035376,
      "step": 28600
    },
    {
      "epoch": 4.665633412186965,
      "eval_loss": 0.19640907645225525,
      "eval_runtime": 68.0667,
      "eval_samples_per_second": 40.034,
      "eval_steps_per_second": 20.024,
      "num_input_tokens_seen": 55035376,
      "step": 28600
    },
    {
      "epoch": 4.666449139407782,
      "grad_norm": 0.0034801801666617393,
      "learning_rate": 0.05617682926190744,
      "loss": 0.1744,
      "num_input_tokens_seen": 55046064,
      "step": 28605
    },
    {
      "epoch": 4.667264866628599,
      "grad_norm": 0.002019364619627595,
      "learning_rate": 0.05613087692450248,
      "loss": 0.1994,
      "num_input_tokens_seen": 55056288,
      "step": 28610
    },
    {
      "epoch": 4.6680805938494165,
      "grad_norm": 0.0025816387496888638,
      "learning_rate": 0.05608493906289592,
      "loss": 0.2236,
      "num_input_tokens_seen": 55066032,
      "step": 28615
    },
    {
      "epoch": 4.668896321070234,
      "grad_norm": 0.005535957869142294,
      "learning_rate": 0.05603901568417201,
      "loss": 0.2056,
      "num_input_tokens_seen": 55075088,
      "step": 28620
    },
    {
      "epoch": 4.669712048291052,
      "grad_norm": 0.0022905105724930763,
      "learning_rate": 0.055993106795412625,
      "loss": 0.1617,
      "num_input_tokens_seen": 55084896,
      "step": 28625
    },
    {
      "epoch": 4.670527775511869,
      "grad_norm": 0.0019115688046440482,
      "learning_rate": 0.05594721240369759,
      "loss": 0.165,
      "num_input_tokens_seen": 55095040,
      "step": 28630
    },
    {
      "epoch": 4.671343502732686,
      "grad_norm": 0.0011780164204537868,
      "learning_rate": 0.055901332516104296,
      "loss": 0.1837,
      "num_input_tokens_seen": 55104736,
      "step": 28635
    },
    {
      "epoch": 4.672159229953504,
      "grad_norm": 0.0010001210030168295,
      "learning_rate": 0.05585546713970804,
      "loss": 0.1683,
      "num_input_tokens_seen": 55115184,
      "step": 28640
    },
    {
      "epoch": 4.672974957174321,
      "grad_norm": 0.0019020631443709135,
      "learning_rate": 0.05580961628158189,
      "loss": 0.1808,
      "num_input_tokens_seen": 55124112,
      "step": 28645
    },
    {
      "epoch": 4.673790684395138,
      "grad_norm": 0.0027200353797525167,
      "learning_rate": 0.05576377994879659,
      "loss": 0.2195,
      "num_input_tokens_seen": 55133568,
      "step": 28650
    },
    {
      "epoch": 4.674606411615955,
      "grad_norm": 0.0020825457759201527,
      "learning_rate": 0.05571795814842063,
      "loss": 0.176,
      "num_input_tokens_seen": 55142992,
      "step": 28655
    },
    {
      "epoch": 4.6754221388367725,
      "grad_norm": 0.0029243144672363997,
      "learning_rate": 0.05567215088752037,
      "loss": 0.1548,
      "num_input_tokens_seen": 55152400,
      "step": 28660
    },
    {
      "epoch": 4.6762378660575905,
      "grad_norm": 0.004359607119113207,
      "learning_rate": 0.05562635817315981,
      "loss": 0.2956,
      "num_input_tokens_seen": 55162080,
      "step": 28665
    },
    {
      "epoch": 4.677053593278408,
      "grad_norm": 0.00399134773761034,
      "learning_rate": 0.05558058001240083,
      "loss": 0.2437,
      "num_input_tokens_seen": 55171104,
      "step": 28670
    },
    {
      "epoch": 4.677869320499225,
      "grad_norm": 0.0019188608275726438,
      "learning_rate": 0.055534816412302915,
      "loss": 0.1632,
      "num_input_tokens_seen": 55180544,
      "step": 28675
    },
    {
      "epoch": 4.678685047720043,
      "grad_norm": 0.002404669998213649,
      "learning_rate": 0.055489067379923436,
      "loss": 0.2302,
      "num_input_tokens_seen": 55190032,
      "step": 28680
    },
    {
      "epoch": 4.67950077494086,
      "grad_norm": 0.002772595500573516,
      "learning_rate": 0.055443332922317505,
      "loss": 0.2204,
      "num_input_tokens_seen": 55199024,
      "step": 28685
    },
    {
      "epoch": 4.680316502161677,
      "grad_norm": 0.002532385056838393,
      "learning_rate": 0.055397613046537876,
      "loss": 0.2504,
      "num_input_tokens_seen": 55208944,
      "step": 28690
    },
    {
      "epoch": 4.681132229382494,
      "grad_norm": 0.002037086756899953,
      "learning_rate": 0.055351907759635145,
      "loss": 0.1713,
      "num_input_tokens_seen": 55218256,
      "step": 28695
    },
    {
      "epoch": 4.681947956603312,
      "grad_norm": 0.0028093233704566956,
      "learning_rate": 0.05530621706865772,
      "loss": 0.1995,
      "num_input_tokens_seen": 55226736,
      "step": 28700
    },
    {
      "epoch": 4.682763683824129,
      "grad_norm": 0.0029524124693125486,
      "learning_rate": 0.055260540980651564,
      "loss": 0.1981,
      "num_input_tokens_seen": 55234896,
      "step": 28705
    },
    {
      "epoch": 4.6835794110449465,
      "grad_norm": 0.0018920196453109384,
      "learning_rate": 0.05521487950266062,
      "loss": 0.246,
      "num_input_tokens_seen": 55244544,
      "step": 28710
    },
    {
      "epoch": 4.684395138265764,
      "grad_norm": 0.0029650998767465353,
      "learning_rate": 0.055169232641726344,
      "loss": 0.1717,
      "num_input_tokens_seen": 55255136,
      "step": 28715
    },
    {
      "epoch": 4.685210865486582,
      "grad_norm": 0.0024705149699002504,
      "learning_rate": 0.055123600404888166,
      "loss": 0.1838,
      "num_input_tokens_seen": 55264752,
      "step": 28720
    },
    {
      "epoch": 4.686026592707399,
      "grad_norm": 0.002527492819353938,
      "learning_rate": 0.05507798279918309,
      "loss": 0.1588,
      "num_input_tokens_seen": 55274464,
      "step": 28725
    },
    {
      "epoch": 4.686842319928216,
      "grad_norm": 0.002086174674332142,
      "learning_rate": 0.0550323798316459,
      "loss": 0.2275,
      "num_input_tokens_seen": 55284880,
      "step": 28730
    },
    {
      "epoch": 4.687658047149033,
      "grad_norm": 0.001975259743630886,
      "learning_rate": 0.05498679150930916,
      "loss": 0.1967,
      "num_input_tokens_seen": 55294800,
      "step": 28735
    },
    {
      "epoch": 4.688473774369851,
      "grad_norm": 0.0024979915469884872,
      "learning_rate": 0.05494121783920323,
      "loss": 0.2113,
      "num_input_tokens_seen": 55303280,
      "step": 28740
    },
    {
      "epoch": 4.689289501590668,
      "grad_norm": 0.0019046165980398655,
      "learning_rate": 0.05489565882835605,
      "loss": 0.1841,
      "num_input_tokens_seen": 55312624,
      "step": 28745
    },
    {
      "epoch": 4.690105228811485,
      "grad_norm": 0.0012417398393154144,
      "learning_rate": 0.05485011448379348,
      "loss": 0.1784,
      "num_input_tokens_seen": 55322784,
      "step": 28750
    },
    {
      "epoch": 4.690920956032302,
      "grad_norm": 0.00152652058750391,
      "learning_rate": 0.05480458481253893,
      "loss": 0.2119,
      "num_input_tokens_seen": 55332960,
      "step": 28755
    },
    {
      "epoch": 4.69173668325312,
      "grad_norm": 0.0017959225224331021,
      "learning_rate": 0.054759069821613715,
      "loss": 0.2034,
      "num_input_tokens_seen": 55342512,
      "step": 28760
    },
    {
      "epoch": 4.692552410473938,
      "grad_norm": 0.0021552809048444033,
      "learning_rate": 0.05471356951803683,
      "loss": 0.1727,
      "num_input_tokens_seen": 55352016,
      "step": 28765
    },
    {
      "epoch": 4.693368137694755,
      "grad_norm": 0.00098593276925385,
      "learning_rate": 0.054668083908824945,
      "loss": 0.1306,
      "num_input_tokens_seen": 55362016,
      "step": 28770
    },
    {
      "epoch": 4.694183864915573,
      "grad_norm": 0.0030385134741663933,
      "learning_rate": 0.054622613000992526,
      "loss": 0.2603,
      "num_input_tokens_seen": 55372592,
      "step": 28775
    },
    {
      "epoch": 4.69499959213639,
      "grad_norm": 0.003834495320916176,
      "learning_rate": 0.05457715680155182,
      "loss": 0.1653,
      "num_input_tokens_seen": 55382080,
      "step": 28780
    },
    {
      "epoch": 4.695815319357207,
      "grad_norm": 0.0015438413247466087,
      "learning_rate": 0.05453171531751265,
      "loss": 0.1857,
      "num_input_tokens_seen": 55392128,
      "step": 28785
    },
    {
      "epoch": 4.696631046578024,
      "grad_norm": 0.0015402135904878378,
      "learning_rate": 0.05448628855588276,
      "loss": 0.1896,
      "num_input_tokens_seen": 55402176,
      "step": 28790
    },
    {
      "epoch": 4.697446773798841,
      "grad_norm": 0.002586755668744445,
      "learning_rate": 0.05444087652366746,
      "loss": 0.1627,
      "num_input_tokens_seen": 55411200,
      "step": 28795
    },
    {
      "epoch": 4.698262501019659,
      "grad_norm": 0.0029529498424381018,
      "learning_rate": 0.05439547922786984,
      "loss": 0.2662,
      "num_input_tokens_seen": 55421296,
      "step": 28800
    },
    {
      "epoch": 4.698262501019659,
      "eval_loss": 0.19375553727149963,
      "eval_runtime": 67.8924,
      "eval_samples_per_second": 40.137,
      "eval_steps_per_second": 20.076,
      "num_input_tokens_seen": 55421296,
      "step": 28800
    },
    {
      "epoch": 4.699078228240476,
      "grad_norm": 0.0022678417153656483,
      "learning_rate": 0.0543500966754908,
      "loss": 0.1869,
      "num_input_tokens_seen": 55430512,
      "step": 28805
    },
    {
      "epoch": 4.6998939554612935,
      "grad_norm": 0.002505409764125943,
      "learning_rate": 0.05430472887352882,
      "loss": 0.2077,
      "num_input_tokens_seen": 55440848,
      "step": 28810
    },
    {
      "epoch": 4.7007096826821115,
      "grad_norm": 0.0021420118864625692,
      "learning_rate": 0.05425937582898023,
      "loss": 0.2056,
      "num_input_tokens_seen": 55450880,
      "step": 28815
    },
    {
      "epoch": 4.701525409902929,
      "grad_norm": 0.0024644238874316216,
      "learning_rate": 0.054214037548839085,
      "loss": 0.1781,
      "num_input_tokens_seen": 55460880,
      "step": 28820
    },
    {
      "epoch": 4.702341137123746,
      "grad_norm": 0.002497501904144883,
      "learning_rate": 0.05416871404009703,
      "loss": 0.2304,
      "num_input_tokens_seen": 55469952,
      "step": 28825
    },
    {
      "epoch": 4.703156864344563,
      "grad_norm": 0.002756503177806735,
      "learning_rate": 0.054123405309743605,
      "loss": 0.1997,
      "num_input_tokens_seen": 55480256,
      "step": 28830
    },
    {
      "epoch": 4.70397259156538,
      "grad_norm": 0.00135518005117774,
      "learning_rate": 0.0540781113647659,
      "loss": 0.1799,
      "num_input_tokens_seen": 55489584,
      "step": 28835
    },
    {
      "epoch": 4.704788318786198,
      "grad_norm": 0.0032287663780152798,
      "learning_rate": 0.054032832212148836,
      "loss": 0.2048,
      "num_input_tokens_seen": 55499056,
      "step": 28840
    },
    {
      "epoch": 4.705604046007015,
      "grad_norm": 0.0023523534182459116,
      "learning_rate": 0.0539875678588751,
      "loss": 0.206,
      "num_input_tokens_seen": 55509264,
      "step": 28845
    },
    {
      "epoch": 4.706419773227832,
      "grad_norm": 0.00178954575676471,
      "learning_rate": 0.05394231831192492,
      "loss": 0.1697,
      "num_input_tokens_seen": 55518816,
      "step": 28850
    },
    {
      "epoch": 4.70723550044865,
      "grad_norm": 0.0023320401087403297,
      "learning_rate": 0.05389708357827639,
      "loss": 0.2134,
      "num_input_tokens_seen": 55529024,
      "step": 28855
    },
    {
      "epoch": 4.7080512276694675,
      "grad_norm": 0.0025909957475960255,
      "learning_rate": 0.05385186366490533,
      "loss": 0.2456,
      "num_input_tokens_seen": 55538352,
      "step": 28860
    },
    {
      "epoch": 4.708866954890285,
      "grad_norm": 0.0030921560246497393,
      "learning_rate": 0.053806658578785166,
      "loss": 0.2054,
      "num_input_tokens_seen": 55548016,
      "step": 28865
    },
    {
      "epoch": 4.709682682111102,
      "grad_norm": 0.0026995984371751547,
      "learning_rate": 0.05376146832688705,
      "loss": 0.194,
      "num_input_tokens_seen": 55559376,
      "step": 28870
    },
    {
      "epoch": 4.71049840933192,
      "grad_norm": 0.0012210491113364697,
      "learning_rate": 0.053716292916179964,
      "loss": 0.1411,
      "num_input_tokens_seen": 55568864,
      "step": 28875
    },
    {
      "epoch": 4.711314136552737,
      "grad_norm": 0.0023236486595124006,
      "learning_rate": 0.05367113235363045,
      "loss": 0.2239,
      "num_input_tokens_seen": 55579360,
      "step": 28880
    },
    {
      "epoch": 4.712129863773554,
      "grad_norm": 0.0025625363923609257,
      "learning_rate": 0.05362598664620289,
      "loss": 0.2351,
      "num_input_tokens_seen": 55588000,
      "step": 28885
    },
    {
      "epoch": 4.712945590994371,
      "grad_norm": 0.001516298041678965,
      "learning_rate": 0.053580855800859285,
      "loss": 0.1682,
      "num_input_tokens_seen": 55596576,
      "step": 28890
    },
    {
      "epoch": 4.713761318215189,
      "grad_norm": 0.0025875489227473736,
      "learning_rate": 0.05353573982455938,
      "loss": 0.1729,
      "num_input_tokens_seen": 55605856,
      "step": 28895
    },
    {
      "epoch": 4.714577045436006,
      "grad_norm": 0.001574049238115549,
      "learning_rate": 0.053490638724260686,
      "loss": 0.173,
      "num_input_tokens_seen": 55615168,
      "step": 28900
    },
    {
      "epoch": 4.715392772656823,
      "grad_norm": 0.003209901973605156,
      "learning_rate": 0.05344555250691827,
      "loss": 0.175,
      "num_input_tokens_seen": 55624816,
      "step": 28905
    },
    {
      "epoch": 4.716208499877641,
      "grad_norm": 0.0030644237995147705,
      "learning_rate": 0.053400481179485086,
      "loss": 0.258,
      "num_input_tokens_seen": 55633920,
      "step": 28910
    },
    {
      "epoch": 4.717024227098459,
      "grad_norm": 0.0020663149189203978,
      "learning_rate": 0.05335542474891159,
      "loss": 0.2216,
      "num_input_tokens_seen": 55644912,
      "step": 28915
    },
    {
      "epoch": 4.717839954319276,
      "grad_norm": 0.0024967172648757696,
      "learning_rate": 0.053310383222146124,
      "loss": 0.185,
      "num_input_tokens_seen": 55653952,
      "step": 28920
    },
    {
      "epoch": 4.718655681540093,
      "grad_norm": 0.004095849581062794,
      "learning_rate": 0.053265356606134684,
      "loss": 0.2153,
      "num_input_tokens_seen": 55662800,
      "step": 28925
    },
    {
      "epoch": 4.71947140876091,
      "grad_norm": 0.0019755400717258453,
      "learning_rate": 0.053220344907820856,
      "loss": 0.1956,
      "num_input_tokens_seen": 55672544,
      "step": 28930
    },
    {
      "epoch": 4.720287135981728,
      "grad_norm": 0.0029968980234116316,
      "learning_rate": 0.05317534813414608,
      "loss": 0.2226,
      "num_input_tokens_seen": 55680560,
      "step": 28935
    },
    {
      "epoch": 4.721102863202545,
      "grad_norm": 0.003997948952019215,
      "learning_rate": 0.05313036629204942,
      "loss": 0.1788,
      "num_input_tokens_seen": 55689984,
      "step": 28940
    },
    {
      "epoch": 4.721918590423362,
      "grad_norm": 0.0018134384881705046,
      "learning_rate": 0.05308539938846756,
      "loss": 0.2189,
      "num_input_tokens_seen": 55698800,
      "step": 28945
    },
    {
      "epoch": 4.72273431764418,
      "grad_norm": 0.0034580137580633163,
      "learning_rate": 0.05304044743033507,
      "loss": 0.2106,
      "num_input_tokens_seen": 55708512,
      "step": 28950
    },
    {
      "epoch": 4.723550044864997,
      "grad_norm": 0.004395937547087669,
      "learning_rate": 0.05299551042458401,
      "loss": 0.2005,
      "num_input_tokens_seen": 55719424,
      "step": 28955
    },
    {
      "epoch": 4.7243657720858145,
      "grad_norm": 0.002270882250741124,
      "learning_rate": 0.052950588378144266,
      "loss": 0.1866,
      "num_input_tokens_seen": 55729216,
      "step": 28960
    },
    {
      "epoch": 4.725181499306632,
      "grad_norm": 0.00497998483479023,
      "learning_rate": 0.052905681297943465,
      "loss": 0.2015,
      "num_input_tokens_seen": 55738752,
      "step": 28965
    },
    {
      "epoch": 4.725997226527449,
      "grad_norm": 0.0017699227901175618,
      "learning_rate": 0.0528607891909067,
      "loss": 0.1805,
      "num_input_tokens_seen": 55749232,
      "step": 28970
    },
    {
      "epoch": 4.726812953748267,
      "grad_norm": 0.0016231477493420243,
      "learning_rate": 0.05281591206395697,
      "loss": 0.2358,
      "num_input_tokens_seen": 55758816,
      "step": 28975
    },
    {
      "epoch": 4.727628680969084,
      "grad_norm": 0.0024679433554410934,
      "learning_rate": 0.05277104992401496,
      "loss": 0.1822,
      "num_input_tokens_seen": 55767728,
      "step": 28980
    },
    {
      "epoch": 4.728444408189901,
      "grad_norm": 0.002785114338621497,
      "learning_rate": 0.05272620277799884,
      "loss": 0.2487,
      "num_input_tokens_seen": 55778000,
      "step": 28985
    },
    {
      "epoch": 4.729260135410719,
      "grad_norm": 0.0016901287017390132,
      "learning_rate": 0.05268137063282473,
      "loss": 0.1517,
      "num_input_tokens_seen": 55787664,
      "step": 28990
    },
    {
      "epoch": 4.730075862631536,
      "grad_norm": 0.002778394613415003,
      "learning_rate": 0.0526365534954062,
      "loss": 0.1891,
      "num_input_tokens_seen": 55797392,
      "step": 28995
    },
    {
      "epoch": 4.730891589852353,
      "grad_norm": 0.002451346255838871,
      "learning_rate": 0.052591751372654656,
      "loss": 0.2119,
      "num_input_tokens_seen": 55807776,
      "step": 29000
    },
    {
      "epoch": 4.730891589852353,
      "eval_loss": 0.19436302781105042,
      "eval_runtime": 67.962,
      "eval_samples_per_second": 40.096,
      "eval_steps_per_second": 20.055,
      "num_input_tokens_seen": 55807776,
      "step": 29000
    },
    {
      "epoch": 4.7317073170731705,
      "grad_norm": 0.0028512156568467617,
      "learning_rate": 0.05254696427147921,
      "loss": 0.202,
      "num_input_tokens_seen": 55818000,
      "step": 29005
    },
    {
      "epoch": 4.732523044293988,
      "grad_norm": 0.002005903050303459,
      "learning_rate": 0.052502192198786546,
      "loss": 0.157,
      "num_input_tokens_seen": 55827584,
      "step": 29010
    },
    {
      "epoch": 4.733338771514806,
      "grad_norm": 0.0024625523947179317,
      "learning_rate": 0.05245743516148103,
      "loss": 0.2227,
      "num_input_tokens_seen": 55835744,
      "step": 29015
    },
    {
      "epoch": 4.734154498735623,
      "grad_norm": 0.002415379509329796,
      "learning_rate": 0.05241269316646486,
      "loss": 0.1623,
      "num_input_tokens_seen": 55845472,
      "step": 29020
    },
    {
      "epoch": 4.73497022595644,
      "grad_norm": 0.0020225951448082924,
      "learning_rate": 0.052367966220637725,
      "loss": 0.1677,
      "num_input_tokens_seen": 55854848,
      "step": 29025
    },
    {
      "epoch": 4.735785953177258,
      "grad_norm": 0.00275794486515224,
      "learning_rate": 0.05232325433089716,
      "loss": 0.1852,
      "num_input_tokens_seen": 55863984,
      "step": 29030
    },
    {
      "epoch": 4.736601680398075,
      "grad_norm": 0.0023512595798820257,
      "learning_rate": 0.052278557504138214,
      "loss": 0.2794,
      "num_input_tokens_seen": 55872624,
      "step": 29035
    },
    {
      "epoch": 4.737417407618892,
      "grad_norm": 0.0022172010503709316,
      "learning_rate": 0.05223387574725372,
      "loss": 0.1881,
      "num_input_tokens_seen": 55882576,
      "step": 29040
    },
    {
      "epoch": 4.738233134839709,
      "grad_norm": 0.0017391120782122016,
      "learning_rate": 0.05218920906713428,
      "loss": 0.1922,
      "num_input_tokens_seen": 55892496,
      "step": 29045
    },
    {
      "epoch": 4.739048862060527,
      "grad_norm": 0.0017781006172299385,
      "learning_rate": 0.05214455747066789,
      "loss": 0.1752,
      "num_input_tokens_seen": 55900656,
      "step": 29050
    },
    {
      "epoch": 4.7398645892813445,
      "grad_norm": 0.004547936376184225,
      "learning_rate": 0.05209992096474048,
      "loss": 0.2282,
      "num_input_tokens_seen": 55909824,
      "step": 29055
    },
    {
      "epoch": 4.740680316502162,
      "grad_norm": 0.0018304265104234219,
      "learning_rate": 0.05205529955623559,
      "loss": 0.1724,
      "num_input_tokens_seen": 55920576,
      "step": 29060
    },
    {
      "epoch": 4.741496043722979,
      "grad_norm": 0.004366436041891575,
      "learning_rate": 0.052010693252034314,
      "loss": 0.1923,
      "num_input_tokens_seen": 55929408,
      "step": 29065
    },
    {
      "epoch": 4.742311770943797,
      "grad_norm": 0.0017407267587259412,
      "learning_rate": 0.0519661020590156,
      "loss": 0.1683,
      "num_input_tokens_seen": 55938928,
      "step": 29070
    },
    {
      "epoch": 4.743127498164614,
      "grad_norm": 0.0028544224333018064,
      "learning_rate": 0.05192152598405586,
      "loss": 0.1603,
      "num_input_tokens_seen": 55947312,
      "step": 29075
    },
    {
      "epoch": 4.743943225385431,
      "grad_norm": 0.0022298621479421854,
      "learning_rate": 0.05187696503402941,
      "loss": 0.1584,
      "num_input_tokens_seen": 55956912,
      "step": 29080
    },
    {
      "epoch": 4.744758952606248,
      "grad_norm": 0.0019854444544762373,
      "learning_rate": 0.05183241921580798,
      "loss": 0.1754,
      "num_input_tokens_seen": 55966912,
      "step": 29085
    },
    {
      "epoch": 4.745574679827066,
      "grad_norm": 0.003006221726536751,
      "learning_rate": 0.051787888536261206,
      "loss": 0.259,
      "num_input_tokens_seen": 55977264,
      "step": 29090
    },
    {
      "epoch": 4.746390407047883,
      "grad_norm": 0.00332773569971323,
      "learning_rate": 0.051743373002256184,
      "loss": 0.2391,
      "num_input_tokens_seen": 55986000,
      "step": 29095
    },
    {
      "epoch": 4.7472061342687,
      "grad_norm": 0.0029824452940374613,
      "learning_rate": 0.05169887262065787,
      "loss": 0.2662,
      "num_input_tokens_seen": 55995184,
      "step": 29100
    },
    {
      "epoch": 4.7480218614895175,
      "grad_norm": 0.0021548597142100334,
      "learning_rate": 0.051654387398328665,
      "loss": 0.1729,
      "num_input_tokens_seen": 56005712,
      "step": 29105
    },
    {
      "epoch": 4.748837588710336,
      "grad_norm": 0.0018699085339903831,
      "learning_rate": 0.05160991734212888,
      "loss": 0.1842,
      "num_input_tokens_seen": 56014992,
      "step": 29110
    },
    {
      "epoch": 4.749653315931153,
      "grad_norm": 0.0027399733662605286,
      "learning_rate": 0.051565462458916224,
      "loss": 0.1715,
      "num_input_tokens_seen": 56024288,
      "step": 29115
    },
    {
      "epoch": 4.75046904315197,
      "grad_norm": 0.001735443715006113,
      "learning_rate": 0.05152102275554627,
      "loss": 0.1602,
      "num_input_tokens_seen": 56034880,
      "step": 29120
    },
    {
      "epoch": 4.751284770372787,
      "grad_norm": 0.002137997653335333,
      "learning_rate": 0.05147659823887222,
      "loss": 0.1563,
      "num_input_tokens_seen": 56044896,
      "step": 29125
    },
    {
      "epoch": 4.752100497593605,
      "grad_norm": 0.002575586549937725,
      "learning_rate": 0.05143218891574479,
      "loss": 0.1525,
      "num_input_tokens_seen": 56053792,
      "step": 29130
    },
    {
      "epoch": 4.752916224814422,
      "grad_norm": 0.0025288995821028948,
      "learning_rate": 0.0513877947930125,
      "loss": 0.293,
      "num_input_tokens_seen": 56063344,
      "step": 29135
    },
    {
      "epoch": 4.753731952035239,
      "grad_norm": 0.005011097993701696,
      "learning_rate": 0.051343415877521566,
      "loss": 0.2048,
      "num_input_tokens_seen": 56073008,
      "step": 29140
    },
    {
      "epoch": 4.754547679256056,
      "grad_norm": 0.0014807601692155004,
      "learning_rate": 0.051299052176115634,
      "loss": 0.2006,
      "num_input_tokens_seen": 56083952,
      "step": 29145
    },
    {
      "epoch": 4.755363406476874,
      "grad_norm": 0.004539248067885637,
      "learning_rate": 0.051254703695636256,
      "loss": 0.2246,
      "num_input_tokens_seen": 56094928,
      "step": 29150
    },
    {
      "epoch": 4.7561791336976915,
      "grad_norm": 0.0024419152177870274,
      "learning_rate": 0.05121037044292249,
      "loss": 0.1783,
      "num_input_tokens_seen": 56104352,
      "step": 29155
    },
    {
      "epoch": 4.756994860918509,
      "grad_norm": 0.005062132142484188,
      "learning_rate": 0.05116605242481101,
      "loss": 0.2065,
      "num_input_tokens_seen": 56113712,
      "step": 29160
    },
    {
      "epoch": 4.757810588139327,
      "grad_norm": 0.003913912456482649,
      "learning_rate": 0.05112174964813634,
      "loss": 0.1344,
      "num_input_tokens_seen": 56121776,
      "step": 29165
    },
    {
      "epoch": 4.758626315360144,
      "grad_norm": 0.0012227335246279836,
      "learning_rate": 0.05107746211973038,
      "loss": 0.1986,
      "num_input_tokens_seen": 56131744,
      "step": 29170
    },
    {
      "epoch": 4.759442042580961,
      "grad_norm": 0.0020671088714152575,
      "learning_rate": 0.05103318984642291,
      "loss": 0.2246,
      "num_input_tokens_seen": 56142160,
      "step": 29175
    },
    {
      "epoch": 4.760257769801778,
      "grad_norm": 0.0035075126215815544,
      "learning_rate": 0.05098893283504131,
      "loss": 0.2191,
      "num_input_tokens_seen": 56151488,
      "step": 29180
    },
    {
      "epoch": 4.761073497022595,
      "grad_norm": 0.0027788912411779165,
      "learning_rate": 0.050944691092410475,
      "loss": 0.2368,
      "num_input_tokens_seen": 56161376,
      "step": 29185
    },
    {
      "epoch": 4.761889224243413,
      "grad_norm": 0.003181300824508071,
      "learning_rate": 0.05090046462535313,
      "loss": 0.189,
      "num_input_tokens_seen": 56170288,
      "step": 29190
    },
    {
      "epoch": 4.76270495146423,
      "grad_norm": 0.0023953348863869905,
      "learning_rate": 0.050856253440689454,
      "loss": 0.1708,
      "num_input_tokens_seen": 56180048,
      "step": 29195
    },
    {
      "epoch": 4.7635206786850475,
      "grad_norm": 0.0017511354526504874,
      "learning_rate": 0.050812057545237405,
      "loss": 0.1516,
      "num_input_tokens_seen": 56188960,
      "step": 29200
    },
    {
      "epoch": 4.7635206786850475,
      "eval_loss": 0.19300255179405212,
      "eval_runtime": 67.9628,
      "eval_samples_per_second": 40.095,
      "eval_steps_per_second": 20.055,
      "num_input_tokens_seen": 56188960,
      "step": 29200
    },
    {
      "epoch": 4.7643364059058655,
      "grad_norm": 0.002575760241597891,
      "learning_rate": 0.0507678769458126,
      "loss": 0.1502,
      "num_input_tokens_seen": 56198912,
      "step": 29205
    },
    {
      "epoch": 4.765152133126683,
      "grad_norm": 0.0016892666462808847,
      "learning_rate": 0.050723711649228155,
      "loss": 0.211,
      "num_input_tokens_seen": 56207792,
      "step": 29210
    },
    {
      "epoch": 4.7659678603475,
      "grad_norm": 0.002403902355581522,
      "learning_rate": 0.05067956166229496,
      "loss": 0.1824,
      "num_input_tokens_seen": 56216784,
      "step": 29215
    },
    {
      "epoch": 4.766783587568317,
      "grad_norm": 0.0032254885882139206,
      "learning_rate": 0.05063542699182155,
      "loss": 0.1904,
      "num_input_tokens_seen": 56226576,
      "step": 29220
    },
    {
      "epoch": 4.767599314789135,
      "grad_norm": 0.001898628892377019,
      "learning_rate": 0.050591307644613996,
      "loss": 0.2203,
      "num_input_tokens_seen": 56236032,
      "step": 29225
    },
    {
      "epoch": 4.768415042009952,
      "grad_norm": 0.0023935833014547825,
      "learning_rate": 0.05054720362747599,
      "loss": 0.2023,
      "num_input_tokens_seen": 56245152,
      "step": 29230
    },
    {
      "epoch": 4.769230769230769,
      "grad_norm": 0.0017108102329075336,
      "learning_rate": 0.050503114947209035,
      "loss": 0.1879,
      "num_input_tokens_seen": 56255920,
      "step": 29235
    },
    {
      "epoch": 4.770046496451586,
      "grad_norm": 0.001983082387596369,
      "learning_rate": 0.05045904161061207,
      "loss": 0.2191,
      "num_input_tokens_seen": 56264864,
      "step": 29240
    },
    {
      "epoch": 4.770862223672404,
      "grad_norm": 0.0018956559943035245,
      "learning_rate": 0.05041498362448185,
      "loss": 0.176,
      "num_input_tokens_seen": 56274496,
      "step": 29245
    },
    {
      "epoch": 4.771677950893221,
      "grad_norm": 0.002058745827525854,
      "learning_rate": 0.05037094099561256,
      "loss": 0.2238,
      "num_input_tokens_seen": 56283328,
      "step": 29250
    },
    {
      "epoch": 4.772493678114039,
      "grad_norm": 0.0012770540779456496,
      "learning_rate": 0.05032691373079624,
      "loss": 0.1873,
      "num_input_tokens_seen": 56292992,
      "step": 29255
    },
    {
      "epoch": 4.773309405334856,
      "grad_norm": 0.0015830937772989273,
      "learning_rate": 0.05028290183682234,
      "loss": 0.1673,
      "num_input_tokens_seen": 56301648,
      "step": 29260
    },
    {
      "epoch": 4.774125132555674,
      "grad_norm": 0.005406280513852835,
      "learning_rate": 0.050238905320478096,
      "loss": 0.2501,
      "num_input_tokens_seen": 56310560,
      "step": 29265
    },
    {
      "epoch": 4.774940859776491,
      "grad_norm": 0.00230521522462368,
      "learning_rate": 0.05019492418854838,
      "loss": 0.2628,
      "num_input_tokens_seen": 56321632,
      "step": 29270
    },
    {
      "epoch": 4.775756586997308,
      "grad_norm": 0.0014933665515854955,
      "learning_rate": 0.05015095844781554,
      "loss": 0.1782,
      "num_input_tokens_seen": 56331568,
      "step": 29275
    },
    {
      "epoch": 4.776572314218125,
      "grad_norm": 0.0020083922427147627,
      "learning_rate": 0.05010700810505968,
      "loss": 0.1873,
      "num_input_tokens_seen": 56342016,
      "step": 29280
    },
    {
      "epoch": 4.777388041438943,
      "grad_norm": 0.002196091925725341,
      "learning_rate": 0.05006307316705856,
      "loss": 0.1568,
      "num_input_tokens_seen": 56351584,
      "step": 29285
    },
    {
      "epoch": 4.77820376865976,
      "grad_norm": 0.0019324307795614004,
      "learning_rate": 0.0500191536405874,
      "loss": 0.174,
      "num_input_tokens_seen": 56360480,
      "step": 29290
    },
    {
      "epoch": 4.779019495880577,
      "grad_norm": 0.001857086201198399,
      "learning_rate": 0.04997524953241922,
      "loss": 0.1805,
      "num_input_tokens_seen": 56370160,
      "step": 29295
    },
    {
      "epoch": 4.7798352231013945,
      "grad_norm": 0.003704878268763423,
      "learning_rate": 0.049931360849324556,
      "loss": 0.1723,
      "num_input_tokens_seen": 56379488,
      "step": 29300
    },
    {
      "epoch": 4.7806509503222125,
      "grad_norm": 0.002043588552623987,
      "learning_rate": 0.04988748759807155,
      "loss": 0.1408,
      "num_input_tokens_seen": 56388800,
      "step": 29305
    },
    {
      "epoch": 4.78146667754303,
      "grad_norm": 0.0022651120088994503,
      "learning_rate": 0.0498436297854261,
      "loss": 0.1817,
      "num_input_tokens_seen": 56398816,
      "step": 29310
    },
    {
      "epoch": 4.782282404763847,
      "grad_norm": 0.0029596227686852217,
      "learning_rate": 0.04979978741815152,
      "loss": 0.1908,
      "num_input_tokens_seen": 56409152,
      "step": 29315
    },
    {
      "epoch": 4.783098131984664,
      "grad_norm": 0.0016120392829179764,
      "learning_rate": 0.04975596050300891,
      "loss": 0.201,
      "num_input_tokens_seen": 56417632,
      "step": 29320
    },
    {
      "epoch": 4.783913859205482,
      "grad_norm": 0.0036792876198887825,
      "learning_rate": 0.049712149046757005,
      "loss": 0.1917,
      "num_input_tokens_seen": 56426144,
      "step": 29325
    },
    {
      "epoch": 4.784729586426299,
      "grad_norm": 0.002935066819190979,
      "learning_rate": 0.04966835305615194,
      "loss": 0.2013,
      "num_input_tokens_seen": 56436736,
      "step": 29330
    },
    {
      "epoch": 4.785545313647116,
      "grad_norm": 0.002060975879430771,
      "learning_rate": 0.049624572537947755,
      "loss": 0.197,
      "num_input_tokens_seen": 56446208,
      "step": 29335
    },
    {
      "epoch": 4.786361040867934,
      "grad_norm": 0.0024581800680607557,
      "learning_rate": 0.04958080749889582,
      "loss": 0.1963,
      "num_input_tokens_seen": 56457056,
      "step": 29340
    },
    {
      "epoch": 4.787176768088751,
      "grad_norm": 0.002143502701073885,
      "learning_rate": 0.049537057945745304,
      "loss": 0.1787,
      "num_input_tokens_seen": 56468032,
      "step": 29345
    },
    {
      "epoch": 4.7879924953095685,
      "grad_norm": 0.0021200194023549557,
      "learning_rate": 0.049493323885243,
      "loss": 0.1701,
      "num_input_tokens_seen": 56478768,
      "step": 29350
    },
    {
      "epoch": 4.788808222530386,
      "grad_norm": 0.002185186604037881,
      "learning_rate": 0.04944960532413318,
      "loss": 0.1628,
      "num_input_tokens_seen": 56488624,
      "step": 29355
    },
    {
      "epoch": 4.789623949751203,
      "grad_norm": 0.002801814815029502,
      "learning_rate": 0.049405902269157774,
      "loss": 0.1865,
      "num_input_tokens_seen": 56498832,
      "step": 29360
    },
    {
      "epoch": 4.790439676972021,
      "grad_norm": 0.001824424136430025,
      "learning_rate": 0.04936221472705646,
      "loss": 0.1923,
      "num_input_tokens_seen": 56507920,
      "step": 29365
    },
    {
      "epoch": 4.791255404192838,
      "grad_norm": 0.0017813558224588633,
      "learning_rate": 0.04931854270456632,
      "loss": 0.277,
      "num_input_tokens_seen": 56518464,
      "step": 29370
    },
    {
      "epoch": 4.792071131413655,
      "grad_norm": 0.0009312191396020353,
      "learning_rate": 0.049274886208422075,
      "loss": 0.1312,
      "num_input_tokens_seen": 56527456,
      "step": 29375
    },
    {
      "epoch": 4.792886858634473,
      "grad_norm": 0.0025948118418455124,
      "learning_rate": 0.049231245245356235,
      "loss": 0.2499,
      "num_input_tokens_seen": 56537280,
      "step": 29380
    },
    {
      "epoch": 4.79370258585529,
      "grad_norm": 0.002542786067351699,
      "learning_rate": 0.049187619822098655,
      "loss": 0.1966,
      "num_input_tokens_seen": 56547168,
      "step": 29385
    },
    {
      "epoch": 4.794518313076107,
      "grad_norm": 0.0018261231016367674,
      "learning_rate": 0.04914400994537705,
      "loss": 0.1768,
      "num_input_tokens_seen": 56557264,
      "step": 29390
    },
    {
      "epoch": 4.7953340402969244,
      "grad_norm": 0.003396329702809453,
      "learning_rate": 0.049100415621916485,
      "loss": 0.2601,
      "num_input_tokens_seen": 56568656,
      "step": 29395
    },
    {
      "epoch": 4.796149767517742,
      "grad_norm": 0.0019143286626785994,
      "learning_rate": 0.04905683685843981,
      "loss": 0.2201,
      "num_input_tokens_seen": 56576864,
      "step": 29400
    },
    {
      "epoch": 4.796149767517742,
      "eval_loss": 0.19305111467838287,
      "eval_runtime": 67.8823,
      "eval_samples_per_second": 40.143,
      "eval_steps_per_second": 20.079,
      "num_input_tokens_seen": 56576864,
      "step": 29400
    },
    {
      "epoch": 4.79696549473856,
      "grad_norm": 0.0021048833150416613,
      "learning_rate": 0.049013273661667495,
      "loss": 0.1401,
      "num_input_tokens_seen": 56585440,
      "step": 29405
    },
    {
      "epoch": 4.797781221959377,
      "grad_norm": 0.0024040360003709793,
      "learning_rate": 0.048969726038317396,
      "loss": 0.1785,
      "num_input_tokens_seen": 56594992,
      "step": 29410
    },
    {
      "epoch": 4.798596949180194,
      "grad_norm": 0.0015123037155717611,
      "learning_rate": 0.048926193995105206,
      "loss": 0.1888,
      "num_input_tokens_seen": 56603232,
      "step": 29415
    },
    {
      "epoch": 4.799412676401012,
      "grad_norm": 0.002540669171139598,
      "learning_rate": 0.048882677538744035,
      "loss": 0.2035,
      "num_input_tokens_seen": 56613200,
      "step": 29420
    },
    {
      "epoch": 4.800228403621829,
      "grad_norm": 0.002747542690485716,
      "learning_rate": 0.048839176675944715,
      "loss": 0.184,
      "num_input_tokens_seen": 56623504,
      "step": 29425
    },
    {
      "epoch": 4.801044130842646,
      "grad_norm": 0.001594758708961308,
      "learning_rate": 0.04879569141341566,
      "loss": 0.1963,
      "num_input_tokens_seen": 56632480,
      "step": 29430
    },
    {
      "epoch": 4.801859858063463,
      "grad_norm": 0.0021744181867688894,
      "learning_rate": 0.04875222175786274,
      "loss": 0.1528,
      "num_input_tokens_seen": 56640848,
      "step": 29435
    },
    {
      "epoch": 4.802675585284281,
      "grad_norm": 0.002438048832118511,
      "learning_rate": 0.04870876771598966,
      "loss": 0.1612,
      "num_input_tokens_seen": 56650496,
      "step": 29440
    },
    {
      "epoch": 4.803491312505098,
      "grad_norm": 0.0022545508109033108,
      "learning_rate": 0.04866532929449744,
      "loss": 0.1877,
      "num_input_tokens_seen": 56659424,
      "step": 29445
    },
    {
      "epoch": 4.8043070397259156,
      "grad_norm": 0.0017405182588845491,
      "learning_rate": 0.048621906500084945,
      "loss": 0.1265,
      "num_input_tokens_seen": 56669632,
      "step": 29450
    },
    {
      "epoch": 4.805122766946733,
      "grad_norm": 0.0033307308331131935,
      "learning_rate": 0.04857849933944845,
      "loss": 0.2287,
      "num_input_tokens_seen": 56679872,
      "step": 29455
    },
    {
      "epoch": 4.805938494167551,
      "grad_norm": 0.0022298821713775396,
      "learning_rate": 0.048535107819281866,
      "loss": 0.2291,
      "num_input_tokens_seen": 56688912,
      "step": 29460
    },
    {
      "epoch": 4.806754221388368,
      "grad_norm": 0.001992085948586464,
      "learning_rate": 0.04849173194627675,
      "loss": 0.1472,
      "num_input_tokens_seen": 56699184,
      "step": 29465
    },
    {
      "epoch": 4.807569948609185,
      "grad_norm": 0.0034650147426873446,
      "learning_rate": 0.04844837172712223,
      "loss": 0.2044,
      "num_input_tokens_seen": 56710160,
      "step": 29470
    },
    {
      "epoch": 4.808385675830002,
      "grad_norm": 0.002148865256458521,
      "learning_rate": 0.04840502716850494,
      "loss": 0.1928,
      "num_input_tokens_seen": 56719888,
      "step": 29475
    },
    {
      "epoch": 4.80920140305082,
      "grad_norm": 0.0015301761450245976,
      "learning_rate": 0.04836169827710916,
      "loss": 0.1961,
      "num_input_tokens_seen": 56729488,
      "step": 29480
    },
    {
      "epoch": 4.810017130271637,
      "grad_norm": 0.001792157650925219,
      "learning_rate": 0.04831838505961684,
      "loss": 0.1833,
      "num_input_tokens_seen": 56739840,
      "step": 29485
    },
    {
      "epoch": 4.810832857492454,
      "grad_norm": 0.006141630467027426,
      "learning_rate": 0.048275087522707295,
      "loss": 0.2292,
      "num_input_tokens_seen": 56748816,
      "step": 29490
    },
    {
      "epoch": 4.8116485847132715,
      "grad_norm": 0.004269104450941086,
      "learning_rate": 0.04823180567305766,
      "loss": 0.1453,
      "num_input_tokens_seen": 56758736,
      "step": 29495
    },
    {
      "epoch": 4.8124643119340895,
      "grad_norm": 0.0035027191042900085,
      "learning_rate": 0.04818853951734244,
      "loss": 0.1489,
      "num_input_tokens_seen": 56768720,
      "step": 29500
    },
    {
      "epoch": 4.813280039154907,
      "grad_norm": 0.0013475606683641672,
      "learning_rate": 0.04814528906223387,
      "loss": 0.2278,
      "num_input_tokens_seen": 56778208,
      "step": 29505
    },
    {
      "epoch": 4.814095766375724,
      "grad_norm": 0.002546807285398245,
      "learning_rate": 0.04810205431440177,
      "loss": 0.2106,
      "num_input_tokens_seen": 56787744,
      "step": 29510
    },
    {
      "epoch": 4.814911493596542,
      "grad_norm": 0.0032138589303940535,
      "learning_rate": 0.04805883528051341,
      "loss": 0.1904,
      "num_input_tokens_seen": 56798048,
      "step": 29515
    },
    {
      "epoch": 4.815727220817359,
      "grad_norm": 0.0021075482945889235,
      "learning_rate": 0.048015631967233685,
      "loss": 0.19,
      "num_input_tokens_seen": 56806848,
      "step": 29520
    },
    {
      "epoch": 4.816542948038176,
      "grad_norm": 0.003538792720064521,
      "learning_rate": 0.04797244438122517,
      "loss": 0.2572,
      "num_input_tokens_seen": 56816416,
      "step": 29525
    },
    {
      "epoch": 4.817358675258993,
      "grad_norm": 0.0020733214914798737,
      "learning_rate": 0.04792927252914784,
      "loss": 0.2049,
      "num_input_tokens_seen": 56825744,
      "step": 29530
    },
    {
      "epoch": 4.81817440247981,
      "grad_norm": 0.0035724497865885496,
      "learning_rate": 0.04788611641765944,
      "loss": 0.2163,
      "num_input_tokens_seen": 56836384,
      "step": 29535
    },
    {
      "epoch": 4.818990129700628,
      "grad_norm": 0.0020028678700327873,
      "learning_rate": 0.04784297605341508,
      "loss": 0.2149,
      "num_input_tokens_seen": 56846096,
      "step": 29540
    },
    {
      "epoch": 4.8198058569214455,
      "grad_norm": 0.002531951293349266,
      "learning_rate": 0.04779985144306761,
      "loss": 0.1686,
      "num_input_tokens_seen": 56855312,
      "step": 29545
    },
    {
      "epoch": 4.820621584142263,
      "grad_norm": 0.0018352778861299157,
      "learning_rate": 0.047756742593267405,
      "loss": 0.2051,
      "num_input_tokens_seen": 56864416,
      "step": 29550
    },
    {
      "epoch": 4.821437311363081,
      "grad_norm": 0.0019635639619082212,
      "learning_rate": 0.047713649510662315,
      "loss": 0.1872,
      "num_input_tokens_seen": 56873888,
      "step": 29555
    },
    {
      "epoch": 4.822253038583898,
      "grad_norm": 0.00282266759313643,
      "learning_rate": 0.04767057220189789,
      "loss": 0.1806,
      "num_input_tokens_seen": 56883696,
      "step": 29560
    },
    {
      "epoch": 4.823068765804715,
      "grad_norm": 0.0017728005768731236,
      "learning_rate": 0.04762751067361722,
      "loss": 0.191,
      "num_input_tokens_seen": 56893120,
      "step": 29565
    },
    {
      "epoch": 4.823884493025532,
      "grad_norm": 0.002377412747591734,
      "learning_rate": 0.04758446493246086,
      "loss": 0.1963,
      "num_input_tokens_seen": 56901952,
      "step": 29570
    },
    {
      "epoch": 4.824700220246349,
      "grad_norm": 0.0033271515276283026,
      "learning_rate": 0.047541434985067084,
      "loss": 0.2147,
      "num_input_tokens_seen": 56911632,
      "step": 29575
    },
    {
      "epoch": 4.825515947467167,
      "grad_norm": 0.0020758695900440216,
      "learning_rate": 0.047498420838071556,
      "loss": 0.2067,
      "num_input_tokens_seen": 56921072,
      "step": 29580
    },
    {
      "epoch": 4.826331674687984,
      "grad_norm": 0.0028646027203649282,
      "learning_rate": 0.04745542249810772,
      "loss": 0.1821,
      "num_input_tokens_seen": 56930272,
      "step": 29585
    },
    {
      "epoch": 4.827147401908801,
      "grad_norm": 0.003983207978308201,
      "learning_rate": 0.047412439971806324,
      "loss": 0.2205,
      "num_input_tokens_seen": 56940416,
      "step": 29590
    },
    {
      "epoch": 4.8279631291296194,
      "grad_norm": 0.0021004145964980125,
      "learning_rate": 0.04736947326579592,
      "loss": 0.1819,
      "num_input_tokens_seen": 56949584,
      "step": 29595
    },
    {
      "epoch": 4.828778856350437,
      "grad_norm": 0.002944555599242449,
      "learning_rate": 0.04732652238670245,
      "loss": 0.1645,
      "num_input_tokens_seen": 56959888,
      "step": 29600
    },
    {
      "epoch": 4.828778856350437,
      "eval_loss": 0.19367516040802002,
      "eval_runtime": 68.0136,
      "eval_samples_per_second": 40.066,
      "eval_steps_per_second": 20.04,
      "num_input_tokens_seen": 56959888,
      "step": 29600
    },
    {
      "epoch": 4.829594583571254,
      "grad_norm": 0.0027020128909498453,
      "learning_rate": 0.04728358734114952,
      "loss": 0.1607,
      "num_input_tokens_seen": 56968656,
      "step": 29605
    },
    {
      "epoch": 4.830410310792071,
      "grad_norm": 0.0019239396788179874,
      "learning_rate": 0.04724066813575821,
      "loss": 0.1654,
      "num_input_tokens_seen": 56978640,
      "step": 29610
    },
    {
      "epoch": 4.831226038012889,
      "grad_norm": 0.0021236035972833633,
      "learning_rate": 0.04719776477714729,
      "loss": 0.1838,
      "num_input_tokens_seen": 56987712,
      "step": 29615
    },
    {
      "epoch": 4.832041765233706,
      "grad_norm": 0.0027379468083381653,
      "learning_rate": 0.047154877271932856,
      "loss": 0.2241,
      "num_input_tokens_seen": 56997392,
      "step": 29620
    },
    {
      "epoch": 4.832857492454523,
      "grad_norm": 0.00171506660990417,
      "learning_rate": 0.0471120056267288,
      "loss": 0.208,
      "num_input_tokens_seen": 57008048,
      "step": 29625
    },
    {
      "epoch": 4.83367321967534,
      "grad_norm": 0.002336434554308653,
      "learning_rate": 0.047069149848146495,
      "loss": 0.1924,
      "num_input_tokens_seen": 57018064,
      "step": 29630
    },
    {
      "epoch": 4.834488946896158,
      "grad_norm": 0.003613077336922288,
      "learning_rate": 0.04702630994279473,
      "loss": 0.2271,
      "num_input_tokens_seen": 57027248,
      "step": 29635
    },
    {
      "epoch": 4.835304674116975,
      "grad_norm": 0.004081647843122482,
      "learning_rate": 0.046983485917280035,
      "loss": 0.2171,
      "num_input_tokens_seen": 57036032,
      "step": 29640
    },
    {
      "epoch": 4.8361204013377925,
      "grad_norm": 0.002379255834966898,
      "learning_rate": 0.04694067777820644,
      "loss": 0.2034,
      "num_input_tokens_seen": 57046720,
      "step": 29645
    },
    {
      "epoch": 4.83693612855861,
      "grad_norm": 0.0023471922613680363,
      "learning_rate": 0.046897885532175415,
      "loss": 0.2102,
      "num_input_tokens_seen": 57057200,
      "step": 29650
    },
    {
      "epoch": 4.837751855779428,
      "grad_norm": 0.0040240660309791565,
      "learning_rate": 0.04685510918578613,
      "loss": 0.2141,
      "num_input_tokens_seen": 57067168,
      "step": 29655
    },
    {
      "epoch": 4.838567583000245,
      "grad_norm": 0.0019841743633151054,
      "learning_rate": 0.04681234874563519,
      "loss": 0.2019,
      "num_input_tokens_seen": 57075584,
      "step": 29660
    },
    {
      "epoch": 4.839383310221062,
      "grad_norm": 0.0016066549578681588,
      "learning_rate": 0.046769604218316836,
      "loss": 0.1649,
      "num_input_tokens_seen": 57086032,
      "step": 29665
    },
    {
      "epoch": 4.840199037441879,
      "grad_norm": 0.003353222506120801,
      "learning_rate": 0.04672687561042279,
      "loss": 0.205,
      "num_input_tokens_seen": 57095824,
      "step": 29670
    },
    {
      "epoch": 4.841014764662697,
      "grad_norm": 0.0023839022032916546,
      "learning_rate": 0.046684162928542286,
      "loss": 0.1557,
      "num_input_tokens_seen": 57105936,
      "step": 29675
    },
    {
      "epoch": 4.841830491883514,
      "grad_norm": 0.0012020908761769533,
      "learning_rate": 0.04664146617926222,
      "loss": 0.1916,
      "num_input_tokens_seen": 57117280,
      "step": 29680
    },
    {
      "epoch": 4.842646219104331,
      "grad_norm": 0.002177569782361388,
      "learning_rate": 0.046598785369167,
      "loss": 0.1914,
      "num_input_tokens_seen": 57127488,
      "step": 29685
    },
    {
      "epoch": 4.8434619463251485,
      "grad_norm": 0.003263436956331134,
      "learning_rate": 0.046556120504838434,
      "loss": 0.1988,
      "num_input_tokens_seen": 57137760,
      "step": 29690
    },
    {
      "epoch": 4.8442776735459665,
      "grad_norm": 0.0018549712840467691,
      "learning_rate": 0.04651347159285609,
      "loss": 0.1908,
      "num_input_tokens_seen": 57148432,
      "step": 29695
    },
    {
      "epoch": 4.845093400766784,
      "grad_norm": 0.0029882045928388834,
      "learning_rate": 0.04647083863979688,
      "loss": 0.2125,
      "num_input_tokens_seen": 57158352,
      "step": 29700
    },
    {
      "epoch": 4.845909127987601,
      "grad_norm": 0.003195288823917508,
      "learning_rate": 0.04642822165223538,
      "loss": 0.2504,
      "num_input_tokens_seen": 57167968,
      "step": 29705
    },
    {
      "epoch": 4.846724855208418,
      "grad_norm": 0.002638209378346801,
      "learning_rate": 0.046385620636743716,
      "loss": 0.2025,
      "num_input_tokens_seen": 57177536,
      "step": 29710
    },
    {
      "epoch": 4.847540582429236,
      "grad_norm": 0.0017185028409585357,
      "learning_rate": 0.04634303559989141,
      "loss": 0.2248,
      "num_input_tokens_seen": 57187344,
      "step": 29715
    },
    {
      "epoch": 4.848356309650053,
      "grad_norm": 0.002220612019300461,
      "learning_rate": 0.046300466548245635,
      "loss": 0.2129,
      "num_input_tokens_seen": 57196320,
      "step": 29720
    },
    {
      "epoch": 4.84917203687087,
      "grad_norm": 0.0031428395304828882,
      "learning_rate": 0.04625791348837114,
      "loss": 0.1681,
      "num_input_tokens_seen": 57205472,
      "step": 29725
    },
    {
      "epoch": 4.849987764091688,
      "grad_norm": 0.0027125512715429068,
      "learning_rate": 0.046215376426830095,
      "loss": 0.1802,
      "num_input_tokens_seen": 57215984,
      "step": 29730
    },
    {
      "epoch": 4.850803491312505,
      "grad_norm": 0.0028109753038734198,
      "learning_rate": 0.04617285537018219,
      "loss": 0.2443,
      "num_input_tokens_seen": 57224288,
      "step": 29735
    },
    {
      "epoch": 4.8516192185333225,
      "grad_norm": 0.002480078022927046,
      "learning_rate": 0.046130350324984803,
      "loss": 0.1776,
      "num_input_tokens_seen": 57233248,
      "step": 29740
    },
    {
      "epoch": 4.85243494575414,
      "grad_norm": 0.004271018784493208,
      "learning_rate": 0.046087861297792666,
      "loss": 0.1781,
      "num_input_tokens_seen": 57243456,
      "step": 29745
    },
    {
      "epoch": 4.853250672974957,
      "grad_norm": 0.003279057564213872,
      "learning_rate": 0.0460453882951582,
      "loss": 0.2017,
      "num_input_tokens_seen": 57253408,
      "step": 29750
    },
    {
      "epoch": 4.854066400195775,
      "grad_norm": 0.0026694450061768293,
      "learning_rate": 0.04600293132363119,
      "loss": 0.2045,
      "num_input_tokens_seen": 57263072,
      "step": 29755
    },
    {
      "epoch": 4.854882127416592,
      "grad_norm": 0.001401787158101797,
      "learning_rate": 0.045960490389759086,
      "loss": 0.1803,
      "num_input_tokens_seen": 57273552,
      "step": 29760
    },
    {
      "epoch": 4.855697854637409,
      "grad_norm": 0.003475272562354803,
      "learning_rate": 0.04591806550008685,
      "loss": 0.19,
      "num_input_tokens_seen": 57282752,
      "step": 29765
    },
    {
      "epoch": 4.856513581858227,
      "grad_norm": 0.002458826871588826,
      "learning_rate": 0.045875656661156825,
      "loss": 0.2453,
      "num_input_tokens_seen": 57291872,
      "step": 29770
    },
    {
      "epoch": 4.857329309079044,
      "grad_norm": 0.0021904492750763893,
      "learning_rate": 0.04583326387950911,
      "loss": 0.1601,
      "num_input_tokens_seen": 57300496,
      "step": 29775
    },
    {
      "epoch": 4.858145036299861,
      "grad_norm": 0.0025809723883867264,
      "learning_rate": 0.0457908871616811,
      "loss": 0.1845,
      "num_input_tokens_seen": 57309616,
      "step": 29780
    },
    {
      "epoch": 4.858960763520678,
      "grad_norm": 0.002708196872845292,
      "learning_rate": 0.04574852651420786,
      "loss": 0.1788,
      "num_input_tokens_seen": 57318752,
      "step": 29785
    },
    {
      "epoch": 4.859776490741496,
      "grad_norm": 0.004855055827647448,
      "learning_rate": 0.045706181943621985,
      "loss": 0.1951,
      "num_input_tokens_seen": 57328608,
      "step": 29790
    },
    {
      "epoch": 4.8605922179623136,
      "grad_norm": 0.004618712700903416,
      "learning_rate": 0.04566385345645344,
      "loss": 0.2017,
      "num_input_tokens_seen": 57337552,
      "step": 29795
    },
    {
      "epoch": 4.861407945183131,
      "grad_norm": 0.002175053348764777,
      "learning_rate": 0.04562154105922993,
      "loss": 0.1807,
      "num_input_tokens_seen": 57347776,
      "step": 29800
    },
    {
      "epoch": 4.861407945183131,
      "eval_loss": 0.19494879245758057,
      "eval_runtime": 67.9835,
      "eval_samples_per_second": 40.083,
      "eval_steps_per_second": 20.049,
      "num_input_tokens_seen": 57347776,
      "step": 29800
    },
    {
      "epoch": 4.862223672403948,
      "grad_norm": 0.0031141533982008696,
      "learning_rate": 0.04557924475847642,
      "loss": 0.2188,
      "num_input_tokens_seen": 57358224,
      "step": 29805
    },
    {
      "epoch": 4.863039399624766,
      "grad_norm": 0.002001725137233734,
      "learning_rate": 0.04553696456071567,
      "loss": 0.1844,
      "num_input_tokens_seen": 57366768,
      "step": 29810
    },
    {
      "epoch": 4.863855126845583,
      "grad_norm": 0.0014966310700401664,
      "learning_rate": 0.045494700472467724,
      "loss": 0.1643,
      "num_input_tokens_seen": 57377024,
      "step": 29815
    },
    {
      "epoch": 4.8646708540664,
      "grad_norm": 0.0026147556491196156,
      "learning_rate": 0.04545245250025024,
      "loss": 0.1812,
      "num_input_tokens_seen": 57386976,
      "step": 29820
    },
    {
      "epoch": 4.865486581287217,
      "grad_norm": 0.004104554653167725,
      "learning_rate": 0.045410220650578384,
      "loss": 0.1946,
      "num_input_tokens_seen": 57396480,
      "step": 29825
    },
    {
      "epoch": 4.866302308508035,
      "grad_norm": 0.002547402400523424,
      "learning_rate": 0.04536800492996492,
      "loss": 0.1755,
      "num_input_tokens_seen": 57407072,
      "step": 29830
    },
    {
      "epoch": 4.867118035728852,
      "grad_norm": 0.00206821714527905,
      "learning_rate": 0.04532580534491994,
      "loss": 0.1891,
      "num_input_tokens_seen": 57416768,
      "step": 29835
    },
    {
      "epoch": 4.8679337629496695,
      "grad_norm": 0.003382711671292782,
      "learning_rate": 0.045283621901951183,
      "loss": 0.1897,
      "num_input_tokens_seen": 57426256,
      "step": 29840
    },
    {
      "epoch": 4.868749490170487,
      "grad_norm": 0.003316262736916542,
      "learning_rate": 0.04524145460756393,
      "loss": 0.2268,
      "num_input_tokens_seen": 57436480,
      "step": 29845
    },
    {
      "epoch": 4.869565217391305,
      "grad_norm": 0.0035995577927678823,
      "learning_rate": 0.045199303468260794,
      "loss": 0.258,
      "num_input_tokens_seen": 57445200,
      "step": 29850
    },
    {
      "epoch": 4.870380944612122,
      "grad_norm": 0.0018065816257148981,
      "learning_rate": 0.04515716849054214,
      "loss": 0.1687,
      "num_input_tokens_seen": 57454704,
      "step": 29855
    },
    {
      "epoch": 4.871196671832939,
      "grad_norm": 0.003059525042772293,
      "learning_rate": 0.04511504968090558,
      "loss": 0.1988,
      "num_input_tokens_seen": 57464864,
      "step": 29860
    },
    {
      "epoch": 4.872012399053756,
      "grad_norm": 0.003224764484912157,
      "learning_rate": 0.04507294704584644,
      "loss": 0.1739,
      "num_input_tokens_seen": 57474880,
      "step": 29865
    },
    {
      "epoch": 4.872828126274574,
      "grad_norm": 0.0020988613832741976,
      "learning_rate": 0.04503086059185749,
      "loss": 0.1839,
      "num_input_tokens_seen": 57483408,
      "step": 29870
    },
    {
      "epoch": 4.873643853495391,
      "grad_norm": 0.0020632685627788305,
      "learning_rate": 0.04498879032542893,
      "loss": 0.2086,
      "num_input_tokens_seen": 57492800,
      "step": 29875
    },
    {
      "epoch": 4.874459580716208,
      "grad_norm": 0.0029104026034474373,
      "learning_rate": 0.0449467362530486,
      "loss": 0.2099,
      "num_input_tokens_seen": 57502720,
      "step": 29880
    },
    {
      "epoch": 4.8752753079370255,
      "grad_norm": 0.0038117326330393553,
      "learning_rate": 0.04490469838120171,
      "loss": 0.1784,
      "num_input_tokens_seen": 57512272,
      "step": 29885
    },
    {
      "epoch": 4.8760910351578435,
      "grad_norm": 0.005373617634177208,
      "learning_rate": 0.04486267671637101,
      "loss": 0.2214,
      "num_input_tokens_seen": 57520496,
      "step": 29890
    },
    {
      "epoch": 4.876906762378661,
      "grad_norm": 0.003652568906545639,
      "learning_rate": 0.04482067126503683,
      "loss": 0.2376,
      "num_input_tokens_seen": 57528912,
      "step": 29895
    },
    {
      "epoch": 4.877722489599478,
      "grad_norm": 0.0015539133455604315,
      "learning_rate": 0.04477868203367687,
      "loss": 0.2384,
      "num_input_tokens_seen": 57538048,
      "step": 29900
    },
    {
      "epoch": 4.878538216820296,
      "grad_norm": 0.002435368252918124,
      "learning_rate": 0.044736709028766426,
      "loss": 0.1966,
      "num_input_tokens_seen": 57547104,
      "step": 29905
    },
    {
      "epoch": 4.879353944041113,
      "grad_norm": 0.0031884091440588236,
      "learning_rate": 0.04469475225677832,
      "loss": 0.2518,
      "num_input_tokens_seen": 57556352,
      "step": 29910
    },
    {
      "epoch": 4.88016967126193,
      "grad_norm": 0.002583138644695282,
      "learning_rate": 0.04465281172418273,
      "loss": 0.1728,
      "num_input_tokens_seen": 57566032,
      "step": 29915
    },
    {
      "epoch": 4.880985398482747,
      "grad_norm": 0.0018886412726715207,
      "learning_rate": 0.044610887437447476,
      "loss": 0.1722,
      "num_input_tokens_seen": 57576320,
      "step": 29920
    },
    {
      "epoch": 4.881801125703564,
      "grad_norm": 0.0013794496189802885,
      "learning_rate": 0.044568979403037744,
      "loss": 0.1661,
      "num_input_tokens_seen": 57584960,
      "step": 29925
    },
    {
      "epoch": 4.882616852924382,
      "grad_norm": 0.001907300902530551,
      "learning_rate": 0.04452708762741631,
      "loss": 0.1994,
      "num_input_tokens_seen": 57594624,
      "step": 29930
    },
    {
      "epoch": 4.883432580145199,
      "grad_norm": 0.002611299743875861,
      "learning_rate": 0.044485212117043475,
      "loss": 0.2215,
      "num_input_tokens_seen": 57603424,
      "step": 29935
    },
    {
      "epoch": 4.884248307366017,
      "grad_norm": 0.003888528561219573,
      "learning_rate": 0.04444335287837687,
      "loss": 0.2014,
      "num_input_tokens_seen": 57612944,
      "step": 29940
    },
    {
      "epoch": 4.885064034586835,
      "grad_norm": 0.003630121238529682,
      "learning_rate": 0.04440150991787179,
      "loss": 0.2012,
      "num_input_tokens_seen": 57623024,
      "step": 29945
    },
    {
      "epoch": 4.885879761807652,
      "grad_norm": 0.002784923417493701,
      "learning_rate": 0.04435968324198088,
      "loss": 0.1711,
      "num_input_tokens_seen": 57633120,
      "step": 29950
    },
    {
      "epoch": 4.886695489028469,
      "grad_norm": 0.001864221296273172,
      "learning_rate": 0.04431787285715442,
      "loss": 0.2021,
      "num_input_tokens_seen": 57641552,
      "step": 29955
    },
    {
      "epoch": 4.887511216249286,
      "grad_norm": 0.0033413602504879236,
      "learning_rate": 0.04427607876984004,
      "loss": 0.2129,
      "num_input_tokens_seen": 57651888,
      "step": 29960
    },
    {
      "epoch": 4.888326943470103,
      "grad_norm": 0.0026694044936448336,
      "learning_rate": 0.044234300986482886,
      "loss": 0.1796,
      "num_input_tokens_seen": 57660416,
      "step": 29965
    },
    {
      "epoch": 4.889142670690921,
      "grad_norm": 0.0015529574593529105,
      "learning_rate": 0.04419253951352566,
      "loss": 0.1716,
      "num_input_tokens_seen": 57669904,
      "step": 29970
    },
    {
      "epoch": 4.889958397911738,
      "grad_norm": 0.0014416859485208988,
      "learning_rate": 0.044150794357408533,
      "loss": 0.161,
      "num_input_tokens_seen": 57678800,
      "step": 29975
    },
    {
      "epoch": 4.890774125132555,
      "grad_norm": 0.002370762638747692,
      "learning_rate": 0.044109065524569065,
      "loss": 0.1906,
      "num_input_tokens_seen": 57688768,
      "step": 29980
    },
    {
      "epoch": 4.891589852353373,
      "grad_norm": 0.00604906165972352,
      "learning_rate": 0.0440673530214424,
      "loss": 0.2215,
      "num_input_tokens_seen": 57698960,
      "step": 29985
    },
    {
      "epoch": 4.8924055795741905,
      "grad_norm": 0.002386645646765828,
      "learning_rate": 0.04402565685446117,
      "loss": 0.1813,
      "num_input_tokens_seen": 57708896,
      "step": 29990
    },
    {
      "epoch": 4.893221306795008,
      "grad_norm": 0.0033855235669761896,
      "learning_rate": 0.04398397703005536,
      "loss": 0.2204,
      "num_input_tokens_seen": 57718464,
      "step": 29995
    },
    {
      "epoch": 4.894037034015825,
      "grad_norm": 0.0027308452408760786,
      "learning_rate": 0.043942313554652626,
      "loss": 0.2215,
      "num_input_tokens_seen": 57727072,
      "step": 30000
    },
    {
      "epoch": 4.894037034015825,
      "eval_loss": 0.19535082578659058,
      "eval_runtime": 67.9288,
      "eval_samples_per_second": 40.116,
      "eval_steps_per_second": 20.065,
      "num_input_tokens_seen": 57727072,
      "step": 30000
    },
    {
      "epoch": 4.894852761236643,
      "grad_norm": 0.0019610365852713585,
      "learning_rate": 0.0439006664346779,
      "loss": 0.2083,
      "num_input_tokens_seen": 57734992,
      "step": 30005
    },
    {
      "epoch": 4.89566848845746,
      "grad_norm": 0.004191551823168993,
      "learning_rate": 0.043859035676553755,
      "loss": 0.2118,
      "num_input_tokens_seen": 57745408,
      "step": 30010
    },
    {
      "epoch": 4.896484215678277,
      "grad_norm": 0.00214355974458158,
      "learning_rate": 0.043817421286700194,
      "loss": 0.2027,
      "num_input_tokens_seen": 57755888,
      "step": 30015
    },
    {
      "epoch": 4.897299942899094,
      "grad_norm": 0.0018442615400999784,
      "learning_rate": 0.043775823271534585,
      "loss": 0.2245,
      "num_input_tokens_seen": 57766992,
      "step": 30020
    },
    {
      "epoch": 4.898115670119912,
      "grad_norm": 0.0014656624989584088,
      "learning_rate": 0.04373424163747197,
      "loss": 0.1727,
      "num_input_tokens_seen": 57777712,
      "step": 30025
    },
    {
      "epoch": 4.898931397340729,
      "grad_norm": 0.002156872069463134,
      "learning_rate": 0.04369267639092473,
      "loss": 0.2076,
      "num_input_tokens_seen": 57788448,
      "step": 30030
    },
    {
      "epoch": 4.8997471245615465,
      "grad_norm": 0.0016707335598766804,
      "learning_rate": 0.04365112753830268,
      "loss": 0.2211,
      "num_input_tokens_seen": 57798192,
      "step": 30035
    },
    {
      "epoch": 4.900562851782364,
      "grad_norm": 0.002545725554227829,
      "learning_rate": 0.04360959508601327,
      "loss": 0.1478,
      "num_input_tokens_seen": 57808400,
      "step": 30040
    },
    {
      "epoch": 4.901378579003182,
      "grad_norm": 0.005429337732493877,
      "learning_rate": 0.04356807904046123,
      "loss": 0.2116,
      "num_input_tokens_seen": 57818720,
      "step": 30045
    },
    {
      "epoch": 4.902194306223999,
      "grad_norm": 0.0029529372695833445,
      "learning_rate": 0.04352657940804892,
      "loss": 0.1958,
      "num_input_tokens_seen": 57828032,
      "step": 30050
    },
    {
      "epoch": 4.903010033444816,
      "grad_norm": 0.0035152535419911146,
      "learning_rate": 0.04348509619517613,
      "loss": 0.1936,
      "num_input_tokens_seen": 57836992,
      "step": 30055
    },
    {
      "epoch": 4.903825760665633,
      "grad_norm": 0.0023408792912960052,
      "learning_rate": 0.04344362940824002,
      "loss": 0.1243,
      "num_input_tokens_seen": 57846080,
      "step": 30060
    },
    {
      "epoch": 4.904641487886451,
      "grad_norm": 0.0033307080157101154,
      "learning_rate": 0.04340217905363533,
      "loss": 0.1742,
      "num_input_tokens_seen": 57855776,
      "step": 30065
    },
    {
      "epoch": 4.905457215107268,
      "grad_norm": 0.003874883521348238,
      "learning_rate": 0.04336074513775425,
      "loss": 0.2232,
      "num_input_tokens_seen": 57865872,
      "step": 30070
    },
    {
      "epoch": 4.906272942328085,
      "grad_norm": 0.0024282175581902266,
      "learning_rate": 0.04331932766698636,
      "loss": 0.1945,
      "num_input_tokens_seen": 57875728,
      "step": 30075
    },
    {
      "epoch": 4.907088669548903,
      "grad_norm": 0.0018866771133616567,
      "learning_rate": 0.0432779266477188,
      "loss": 0.2386,
      "num_input_tokens_seen": 57884688,
      "step": 30080
    },
    {
      "epoch": 4.9079043967697205,
      "grad_norm": 0.0011807242408394814,
      "learning_rate": 0.04323654208633607,
      "loss": 0.1485,
      "num_input_tokens_seen": 57894944,
      "step": 30085
    },
    {
      "epoch": 4.908720123990538,
      "grad_norm": 0.0020711300894618034,
      "learning_rate": 0.04319517398922024,
      "loss": 0.1961,
      "num_input_tokens_seen": 57904448,
      "step": 30090
    },
    {
      "epoch": 4.909535851211355,
      "grad_norm": 0.002506205579265952,
      "learning_rate": 0.04315382236275079,
      "loss": 0.2272,
      "num_input_tokens_seen": 57914096,
      "step": 30095
    },
    {
      "epoch": 4.910351578432172,
      "grad_norm": 0.0021584173664450645,
      "learning_rate": 0.043112487213304664,
      "loss": 0.1933,
      "num_input_tokens_seen": 57924352,
      "step": 30100
    },
    {
      "epoch": 4.91116730565299,
      "grad_norm": 0.0016976617043837905,
      "learning_rate": 0.04307116854725618,
      "loss": 0.1648,
      "num_input_tokens_seen": 57935024,
      "step": 30105
    },
    {
      "epoch": 4.911983032873807,
      "grad_norm": 0.003229920519515872,
      "learning_rate": 0.043029866370977325,
      "loss": 0.227,
      "num_input_tokens_seen": 57945696,
      "step": 30110
    },
    {
      "epoch": 4.912798760094624,
      "grad_norm": 0.0018728821305558085,
      "learning_rate": 0.04298858069083728,
      "loss": 0.1597,
      "num_input_tokens_seen": 57955984,
      "step": 30115
    },
    {
      "epoch": 4.913614487315442,
      "grad_norm": 0.003649862250313163,
      "learning_rate": 0.04294731151320295,
      "loss": 0.2014,
      "num_input_tokens_seen": 57965856,
      "step": 30120
    },
    {
      "epoch": 4.914430214536259,
      "grad_norm": 0.0030255343299359083,
      "learning_rate": 0.04290605884443841,
      "loss": 0.1959,
      "num_input_tokens_seen": 57975328,
      "step": 30125
    },
    {
      "epoch": 4.915245941757076,
      "grad_norm": 0.001173440832644701,
      "learning_rate": 0.04286482269090545,
      "loss": 0.2066,
      "num_input_tokens_seen": 57986352,
      "step": 30130
    },
    {
      "epoch": 4.9160616689778935,
      "grad_norm": 0.0016619580565020442,
      "learning_rate": 0.04282360305896323,
      "loss": 0.216,
      "num_input_tokens_seen": 57997008,
      "step": 30135
    },
    {
      "epoch": 4.916877396198711,
      "grad_norm": 0.0032326234504580498,
      "learning_rate": 0.04278239995496822,
      "loss": 0.1539,
      "num_input_tokens_seen": 58005376,
      "step": 30140
    },
    {
      "epoch": 4.917693123419529,
      "grad_norm": 0.002259917091578245,
      "learning_rate": 0.042741213385274514,
      "loss": 0.2113,
      "num_input_tokens_seen": 58015488,
      "step": 30145
    },
    {
      "epoch": 4.918508850640346,
      "grad_norm": 0.002177976770326495,
      "learning_rate": 0.04270004335623366,
      "loss": 0.2272,
      "num_input_tokens_seen": 58025536,
      "step": 30150
    },
    {
      "epoch": 4.919324577861163,
      "grad_norm": 0.0016441888874396682,
      "learning_rate": 0.04265888987419448,
      "loss": 0.2103,
      "num_input_tokens_seen": 58035760,
      "step": 30155
    },
    {
      "epoch": 4.920140305081981,
      "grad_norm": 0.0025784096214920282,
      "learning_rate": 0.04261775294550346,
      "loss": 0.2232,
      "num_input_tokens_seen": 58044944,
      "step": 30160
    },
    {
      "epoch": 4.920956032302798,
      "grad_norm": 0.001934044761583209,
      "learning_rate": 0.042576632576504354,
      "loss": 0.1604,
      "num_input_tokens_seen": 58054672,
      "step": 30165
    },
    {
      "epoch": 4.921771759523615,
      "grad_norm": 0.0017823213711380959,
      "learning_rate": 0.0425355287735385,
      "loss": 0.1342,
      "num_input_tokens_seen": 58063248,
      "step": 30170
    },
    {
      "epoch": 4.922587486744432,
      "grad_norm": 0.0018827554304152727,
      "learning_rate": 0.0424944415429446,
      "loss": 0.1869,
      "num_input_tokens_seen": 58073168,
      "step": 30175
    },
    {
      "epoch": 4.92340321396525,
      "grad_norm": 0.0016136955237016082,
      "learning_rate": 0.04245337089105877,
      "loss": 0.2128,
      "num_input_tokens_seen": 58082112,
      "step": 30180
    },
    {
      "epoch": 4.9242189411860675,
      "grad_norm": 0.0021097941789776087,
      "learning_rate": 0.04241231682421467,
      "loss": 0.1784,
      "num_input_tokens_seen": 58091280,
      "step": 30185
    },
    {
      "epoch": 4.925034668406885,
      "grad_norm": 0.003927254118025303,
      "learning_rate": 0.04237127934874337,
      "loss": 0.132,
      "num_input_tokens_seen": 58101456,
      "step": 30190
    },
    {
      "epoch": 4.925850395627702,
      "grad_norm": 0.004219305235892534,
      "learning_rate": 0.042330258470973305,
      "loss": 0.2418,
      "num_input_tokens_seen": 58110432,
      "step": 30195
    },
    {
      "epoch": 4.92666612284852,
      "grad_norm": 0.0029445714317262173,
      "learning_rate": 0.042289254197230515,
      "loss": 0.2328,
      "num_input_tokens_seen": 58119904,
      "step": 30200
    },
    {
      "epoch": 4.92666612284852,
      "eval_loss": 0.1985388994216919,
      "eval_runtime": 67.8917,
      "eval_samples_per_second": 40.137,
      "eval_steps_per_second": 20.076,
      "num_input_tokens_seen": 58119904,
      "step": 30200
    },
    {
      "epoch": 4.927481850069337,
      "grad_norm": 0.001844537677243352,
      "learning_rate": 0.04224826653383823,
      "loss": 0.2014,
      "num_input_tokens_seen": 58129072,
      "step": 30205
    },
    {
      "epoch": 4.928297577290154,
      "grad_norm": 0.0026082457043230534,
      "learning_rate": 0.04220729548711735,
      "loss": 0.1884,
      "num_input_tokens_seen": 58138832,
      "step": 30210
    },
    {
      "epoch": 4.929113304510971,
      "grad_norm": 0.0030782094690948725,
      "learning_rate": 0.04216634106338616,
      "loss": 0.186,
      "num_input_tokens_seen": 58148592,
      "step": 30215
    },
    {
      "epoch": 4.929929031731789,
      "grad_norm": 0.0023198288399726152,
      "learning_rate": 0.04212540326896025,
      "loss": 0.1774,
      "num_input_tokens_seen": 58158576,
      "step": 30220
    },
    {
      "epoch": 4.930744758952606,
      "grad_norm": 0.0030659176409244537,
      "learning_rate": 0.0420844821101528,
      "loss": 0.2622,
      "num_input_tokens_seen": 58168384,
      "step": 30225
    },
    {
      "epoch": 4.9315604861734235,
      "grad_norm": 0.001999520231038332,
      "learning_rate": 0.04204357759327441,
      "loss": 0.1622,
      "num_input_tokens_seen": 58177856,
      "step": 30230
    },
    {
      "epoch": 4.932376213394241,
      "grad_norm": 0.0014243019977584481,
      "learning_rate": 0.042002689724632954,
      "loss": 0.1829,
      "num_input_tokens_seen": 58187584,
      "step": 30235
    },
    {
      "epoch": 4.933191940615059,
      "grad_norm": 0.002030671341344714,
      "learning_rate": 0.04196181851053398,
      "loss": 0.1493,
      "num_input_tokens_seen": 58197312,
      "step": 30240
    },
    {
      "epoch": 4.934007667835876,
      "grad_norm": 0.003345823148265481,
      "learning_rate": 0.041920963957280295,
      "loss": 0.235,
      "num_input_tokens_seen": 58205520,
      "step": 30245
    },
    {
      "epoch": 4.934823395056693,
      "grad_norm": 0.002469836501404643,
      "learning_rate": 0.04188012607117212,
      "loss": 0.2209,
      "num_input_tokens_seen": 58214000,
      "step": 30250
    },
    {
      "epoch": 4.935639122277511,
      "grad_norm": 0.00295495823957026,
      "learning_rate": 0.04183930485850725,
      "loss": 0.2739,
      "num_input_tokens_seen": 58224400,
      "step": 30255
    },
    {
      "epoch": 4.936454849498328,
      "grad_norm": 0.0015690108994022012,
      "learning_rate": 0.04179850032558078,
      "loss": 0.233,
      "num_input_tokens_seen": 58234720,
      "step": 30260
    },
    {
      "epoch": 4.937270576719145,
      "grad_norm": 0.0019523531664162874,
      "learning_rate": 0.041757712478685295,
      "loss": 0.1743,
      "num_input_tokens_seen": 58244928,
      "step": 30265
    },
    {
      "epoch": 4.938086303939962,
      "grad_norm": 0.0027681749779731035,
      "learning_rate": 0.04171694132411085,
      "loss": 0.1687,
      "num_input_tokens_seen": 58254912,
      "step": 30270
    },
    {
      "epoch": 4.938902031160779,
      "grad_norm": 0.0031966299284249544,
      "learning_rate": 0.04167618686814479,
      "loss": 0.2916,
      "num_input_tokens_seen": 58263712,
      "step": 30275
    },
    {
      "epoch": 4.939717758381597,
      "grad_norm": 0.0024754595942795277,
      "learning_rate": 0.041635449117072024,
      "loss": 0.1914,
      "num_input_tokens_seen": 58273504,
      "step": 30280
    },
    {
      "epoch": 4.940533485602415,
      "grad_norm": 0.0023399367928504944,
      "learning_rate": 0.04159472807717477,
      "loss": 0.1829,
      "num_input_tokens_seen": 58283808,
      "step": 30285
    },
    {
      "epoch": 4.941349212823232,
      "grad_norm": 0.00212470511905849,
      "learning_rate": 0.041554023754732744,
      "loss": 0.2196,
      "num_input_tokens_seen": 58294192,
      "step": 30290
    },
    {
      "epoch": 4.94216494004405,
      "grad_norm": 0.002112036105245352,
      "learning_rate": 0.04151333615602311,
      "loss": 0.2137,
      "num_input_tokens_seen": 58304512,
      "step": 30295
    },
    {
      "epoch": 4.942980667264867,
      "grad_norm": 0.0017054321942850947,
      "learning_rate": 0.04147266528732034,
      "loss": 0.2048,
      "num_input_tokens_seen": 58313760,
      "step": 30300
    },
    {
      "epoch": 4.943796394485684,
      "grad_norm": 0.002459141658619046,
      "learning_rate": 0.0414320111548964,
      "loss": 0.1995,
      "num_input_tokens_seen": 58323328,
      "step": 30305
    },
    {
      "epoch": 4.944612121706501,
      "grad_norm": 0.0020938043016940355,
      "learning_rate": 0.04139137376502076,
      "loss": 0.2242,
      "num_input_tokens_seen": 58333872,
      "step": 30310
    },
    {
      "epoch": 4.945427848927318,
      "grad_norm": 0.002377330558374524,
      "learning_rate": 0.04135075312396014,
      "loss": 0.185,
      "num_input_tokens_seen": 58342624,
      "step": 30315
    },
    {
      "epoch": 4.946243576148136,
      "grad_norm": 0.003606830956414342,
      "learning_rate": 0.04131014923797875,
      "loss": 0.2232,
      "num_input_tokens_seen": 58352272,
      "step": 30320
    },
    {
      "epoch": 4.947059303368953,
      "grad_norm": 0.0031826086342334747,
      "learning_rate": 0.04126956211333819,
      "loss": 0.2076,
      "num_input_tokens_seen": 58360704,
      "step": 30325
    },
    {
      "epoch": 4.9478750305897705,
      "grad_norm": 0.0038438960909843445,
      "learning_rate": 0.041228991756297545,
      "loss": 0.217,
      "num_input_tokens_seen": 58370464,
      "step": 30330
    },
    {
      "epoch": 4.9486907578105885,
      "grad_norm": 0.0017486705910414457,
      "learning_rate": 0.04118843817311332,
      "loss": 0.1963,
      "num_input_tokens_seen": 58380688,
      "step": 30335
    },
    {
      "epoch": 4.949506485031406,
      "grad_norm": 0.0015257814666256309,
      "learning_rate": 0.0411479013700393,
      "loss": 0.1526,
      "num_input_tokens_seen": 58389168,
      "step": 30340
    },
    {
      "epoch": 4.950322212252223,
      "grad_norm": 0.0019165880512446165,
      "learning_rate": 0.0411073813533268,
      "loss": 0.1825,
      "num_input_tokens_seen": 58398336,
      "step": 30345
    },
    {
      "epoch": 4.95113793947304,
      "grad_norm": 0.0023724085185676813,
      "learning_rate": 0.04106687812922456,
      "loss": 0.1529,
      "num_input_tokens_seen": 58407472,
      "step": 30350
    },
    {
      "epoch": 4.951953666693858,
      "grad_norm": 0.004149695858359337,
      "learning_rate": 0.041026391703978635,
      "loss": 0.2581,
      "num_input_tokens_seen": 58415936,
      "step": 30355
    },
    {
      "epoch": 4.952769393914675,
      "grad_norm": 0.0032359545584768057,
      "learning_rate": 0.04098592208383259,
      "loss": 0.2097,
      "num_input_tokens_seen": 58425920,
      "step": 30360
    },
    {
      "epoch": 4.953585121135492,
      "grad_norm": 0.00256647658534348,
      "learning_rate": 0.040945469275027256,
      "loss": 0.1667,
      "num_input_tokens_seen": 58434608,
      "step": 30365
    },
    {
      "epoch": 4.954400848356309,
      "grad_norm": 0.0016189351445063949,
      "learning_rate": 0.04090503328380104,
      "loss": 0.1971,
      "num_input_tokens_seen": 58444288,
      "step": 30370
    },
    {
      "epoch": 4.955216575577127,
      "grad_norm": 0.0016894879518076777,
      "learning_rate": 0.04086461411638971,
      "loss": 0.185,
      "num_input_tokens_seen": 58454256,
      "step": 30375
    },
    {
      "epoch": 4.9560323027979445,
      "grad_norm": 0.0027589714154601097,
      "learning_rate": 0.04082421177902631,
      "loss": 0.2147,
      "num_input_tokens_seen": 58464288,
      "step": 30380
    },
    {
      "epoch": 4.956848030018762,
      "grad_norm": 0.001961758825927973,
      "learning_rate": 0.04078382627794149,
      "loss": 0.2254,
      "num_input_tokens_seen": 58474720,
      "step": 30385
    },
    {
      "epoch": 4.957663757239579,
      "grad_norm": 0.002422418911010027,
      "learning_rate": 0.04074345761936316,
      "loss": 0.2038,
      "num_input_tokens_seen": 58485472,
      "step": 30390
    },
    {
      "epoch": 4.958479484460397,
      "grad_norm": 0.002568792551755905,
      "learning_rate": 0.04070310580951663,
      "loss": 0.2008,
      "num_input_tokens_seen": 58494768,
      "step": 30395
    },
    {
      "epoch": 4.959295211681214,
      "grad_norm": 0.002339780330657959,
      "learning_rate": 0.040662770854624726,
      "loss": 0.1302,
      "num_input_tokens_seen": 58503776,
      "step": 30400
    },
    {
      "epoch": 4.959295211681214,
      "eval_loss": 0.19306398928165436,
      "eval_runtime": 68.0092,
      "eval_samples_per_second": 40.068,
      "eval_steps_per_second": 20.041,
      "num_input_tokens_seen": 58503776,
      "step": 30400
    },
    {
      "epoch": 4.960110938902031,
      "grad_norm": 0.002249739598482847,
      "learning_rate": 0.040622452760907535,
      "loss": 0.1983,
      "num_input_tokens_seen": 58514288,
      "step": 30405
    },
    {
      "epoch": 4.960926666122848,
      "grad_norm": 0.0018091119127348065,
      "learning_rate": 0.04058215153458265,
      "loss": 0.1904,
      "num_input_tokens_seen": 58524288,
      "step": 30410
    },
    {
      "epoch": 4.961742393343666,
      "grad_norm": 0.0030627762898802757,
      "learning_rate": 0.04054186718186507,
      "loss": 0.2519,
      "num_input_tokens_seen": 58534400,
      "step": 30415
    },
    {
      "epoch": 4.962558120564483,
      "grad_norm": 0.0022035439033061266,
      "learning_rate": 0.04050159970896708,
      "loss": 0.1691,
      "num_input_tokens_seen": 58543424,
      "step": 30420
    },
    {
      "epoch": 4.9633738477853,
      "grad_norm": 0.0027370767202228308,
      "learning_rate": 0.04046134912209843,
      "loss": 0.1831,
      "num_input_tokens_seen": 58553072,
      "step": 30425
    },
    {
      "epoch": 4.964189575006118,
      "grad_norm": 0.0031547180842608213,
      "learning_rate": 0.040421115427466354,
      "loss": 0.2025,
      "num_input_tokens_seen": 58562384,
      "step": 30430
    },
    {
      "epoch": 4.965005302226936,
      "grad_norm": 0.0021651796996593475,
      "learning_rate": 0.04038089863127529,
      "loss": 0.1659,
      "num_input_tokens_seen": 58572256,
      "step": 30435
    },
    {
      "epoch": 4.965821029447753,
      "grad_norm": 0.002321979496628046,
      "learning_rate": 0.04034069873972727,
      "loss": 0.2314,
      "num_input_tokens_seen": 58582000,
      "step": 30440
    },
    {
      "epoch": 4.96663675666857,
      "grad_norm": 0.0016789333894848824,
      "learning_rate": 0.040300515759021514,
      "loss": 0.1436,
      "num_input_tokens_seen": 58590848,
      "step": 30445
    },
    {
      "epoch": 4.967452483889387,
      "grad_norm": 0.0015104326885193586,
      "learning_rate": 0.04026034969535478,
      "loss": 0.2022,
      "num_input_tokens_seen": 58602192,
      "step": 30450
    },
    {
      "epoch": 4.968268211110205,
      "grad_norm": 0.0027701430954039097,
      "learning_rate": 0.040220200554921266,
      "loss": 0.1421,
      "num_input_tokens_seen": 58611264,
      "step": 30455
    },
    {
      "epoch": 4.969083938331022,
      "grad_norm": 0.0022235452197492123,
      "learning_rate": 0.0401800683439124,
      "loss": 0.187,
      "num_input_tokens_seen": 58620720,
      "step": 30460
    },
    {
      "epoch": 4.969899665551839,
      "grad_norm": 0.0024581081233918667,
      "learning_rate": 0.04013995306851704,
      "loss": 0.2596,
      "num_input_tokens_seen": 58631104,
      "step": 30465
    },
    {
      "epoch": 4.970715392772657,
      "grad_norm": 0.0031077770981937647,
      "learning_rate": 0.040099854734921545,
      "loss": 0.1955,
      "num_input_tokens_seen": 58640672,
      "step": 30470
    },
    {
      "epoch": 4.971531119993474,
      "grad_norm": 0.00163021229673177,
      "learning_rate": 0.0400597733493095,
      "loss": 0.2393,
      "num_input_tokens_seen": 58650480,
      "step": 30475
    },
    {
      "epoch": 4.9723468472142915,
      "grad_norm": 0.0020684648770838976,
      "learning_rate": 0.04001970891786203,
      "loss": 0.1716,
      "num_input_tokens_seen": 58661280,
      "step": 30480
    },
    {
      "epoch": 4.973162574435109,
      "grad_norm": 0.0022670202888548374,
      "learning_rate": 0.03997966144675752,
      "loss": 0.1745,
      "num_input_tokens_seen": 58670832,
      "step": 30485
    },
    {
      "epoch": 4.973978301655926,
      "grad_norm": 0.0022724042646586895,
      "learning_rate": 0.039939630942171796,
      "loss": 0.2131,
      "num_input_tokens_seen": 58679776,
      "step": 30490
    },
    {
      "epoch": 4.974794028876744,
      "grad_norm": 0.0033398845698684454,
      "learning_rate": 0.03989961741027815,
      "loss": 0.1765,
      "num_input_tokens_seen": 58688880,
      "step": 30495
    },
    {
      "epoch": 4.975609756097561,
      "grad_norm": 0.0026693525724112988,
      "learning_rate": 0.03985962085724704,
      "loss": 0.1883,
      "num_input_tokens_seen": 58698480,
      "step": 30500
    },
    {
      "epoch": 4.976425483318378,
      "grad_norm": 0.0011323265498504043,
      "learning_rate": 0.03981964128924656,
      "loss": 0.2007,
      "num_input_tokens_seen": 58706944,
      "step": 30505
    },
    {
      "epoch": 4.977241210539196,
      "grad_norm": 0.0018804396968334913,
      "learning_rate": 0.03977967871244197,
      "loss": 0.158,
      "num_input_tokens_seen": 58716368,
      "step": 30510
    },
    {
      "epoch": 4.978056937760013,
      "grad_norm": 0.004491961561143398,
      "learning_rate": 0.03973973313299602,
      "loss": 0.2078,
      "num_input_tokens_seen": 58725392,
      "step": 30515
    },
    {
      "epoch": 4.97887266498083,
      "grad_norm": 0.0030296698678284883,
      "learning_rate": 0.0396998045570689,
      "loss": 0.2421,
      "num_input_tokens_seen": 58735152,
      "step": 30520
    },
    {
      "epoch": 4.9796883922016475,
      "grad_norm": 0.0021330686286091805,
      "learning_rate": 0.03965989299081798,
      "loss": 0.1785,
      "num_input_tokens_seen": 58744224,
      "step": 30525
    },
    {
      "epoch": 4.9805041194224655,
      "grad_norm": 0.0020227537024766207,
      "learning_rate": 0.039619998440398235,
      "loss": 0.1835,
      "num_input_tokens_seen": 58753552,
      "step": 30530
    },
    {
      "epoch": 4.981319846643283,
      "grad_norm": 0.001487542293034494,
      "learning_rate": 0.03958012091196184,
      "loss": 0.2201,
      "num_input_tokens_seen": 58764160,
      "step": 30535
    },
    {
      "epoch": 4.9821355738641,
      "grad_norm": 0.002102626720443368,
      "learning_rate": 0.039540260411658396,
      "loss": 0.2412,
      "num_input_tokens_seen": 58774000,
      "step": 30540
    },
    {
      "epoch": 4.982951301084917,
      "grad_norm": 0.0022565582767128944,
      "learning_rate": 0.03950041694563496,
      "loss": 0.2379,
      "num_input_tokens_seen": 58784320,
      "step": 30545
    },
    {
      "epoch": 4.983767028305735,
      "grad_norm": 0.0022846101783216,
      "learning_rate": 0.0394605905200358,
      "loss": 0.1663,
      "num_input_tokens_seen": 58794032,
      "step": 30550
    },
    {
      "epoch": 4.984582755526552,
      "grad_norm": 0.002297980012372136,
      "learning_rate": 0.03942078114100272,
      "loss": 0.1676,
      "num_input_tokens_seen": 58803808,
      "step": 30555
    },
    {
      "epoch": 4.985398482747369,
      "grad_norm": 0.002537948079407215,
      "learning_rate": 0.03938098881467485,
      "loss": 0.1791,
      "num_input_tokens_seen": 58812704,
      "step": 30560
    },
    {
      "epoch": 4.986214209968186,
      "grad_norm": 0.0024663324002176523,
      "learning_rate": 0.039341213547188586,
      "loss": 0.2188,
      "num_input_tokens_seen": 58822160,
      "step": 30565
    },
    {
      "epoch": 4.987029937189004,
      "grad_norm": 0.0019430287647992373,
      "learning_rate": 0.03930145534467782,
      "loss": 0.1526,
      "num_input_tokens_seen": 58833168,
      "step": 30570
    },
    {
      "epoch": 4.9878456644098215,
      "grad_norm": 0.0014718178426846862,
      "learning_rate": 0.0392617142132738,
      "loss": 0.1815,
      "num_input_tokens_seen": 58843072,
      "step": 30575
    },
    {
      "epoch": 4.988661391630639,
      "grad_norm": 0.0026022011879831553,
      "learning_rate": 0.03922199015910504,
      "loss": 0.2054,
      "num_input_tokens_seen": 58852832,
      "step": 30580
    },
    {
      "epoch": 4.989477118851456,
      "grad_norm": 0.004752276465296745,
      "learning_rate": 0.039182283188297556,
      "loss": 0.2134,
      "num_input_tokens_seen": 58862160,
      "step": 30585
    },
    {
      "epoch": 4.990292846072274,
      "grad_norm": 0.0015487943310290575,
      "learning_rate": 0.039142593306974595,
      "loss": 0.1512,
      "num_input_tokens_seen": 58872320,
      "step": 30590
    },
    {
      "epoch": 4.991108573293091,
      "grad_norm": 0.0025917915627360344,
      "learning_rate": 0.039102920521256856,
      "loss": 0.1734,
      "num_input_tokens_seen": 58882432,
      "step": 30595
    },
    {
      "epoch": 4.991924300513908,
      "grad_norm": 0.0017693639965727925,
      "learning_rate": 0.03906326483726243,
      "loss": 0.182,
      "num_input_tokens_seen": 58892528,
      "step": 30600
    },
    {
      "epoch": 4.991924300513908,
      "eval_loss": 0.19090847671031952,
      "eval_runtime": 68.0559,
      "eval_samples_per_second": 40.041,
      "eval_steps_per_second": 20.028,
      "num_input_tokens_seen": 58892528,
      "step": 30600
    },
    {
      "epoch": 4.992740027734725,
      "grad_norm": 0.005696638487279415,
      "learning_rate": 0.039023626261106704,
      "loss": 0.2073,
      "num_input_tokens_seen": 58901888,
      "step": 30605
    },
    {
      "epoch": 4.993555754955543,
      "grad_norm": 0.0020836731418967247,
      "learning_rate": 0.03898400479890237,
      "loss": 0.2022,
      "num_input_tokens_seen": 58911040,
      "step": 30610
    },
    {
      "epoch": 4.99437148217636,
      "grad_norm": 0.0036429923493415117,
      "learning_rate": 0.038944400456759655,
      "loss": 0.1684,
      "num_input_tokens_seen": 58920976,
      "step": 30615
    },
    {
      "epoch": 4.995187209397177,
      "grad_norm": 0.0015954882837831974,
      "learning_rate": 0.038904813240785964,
      "loss": 0.1985,
      "num_input_tokens_seen": 58930176,
      "step": 30620
    },
    {
      "epoch": 4.9960029366179945,
      "grad_norm": 0.004004445858299732,
      "learning_rate": 0.03886524315708621,
      "loss": 0.2799,
      "num_input_tokens_seen": 58938928,
      "step": 30625
    },
    {
      "epoch": 4.996818663838813,
      "grad_norm": 0.002635940443724394,
      "learning_rate": 0.03882569021176255,
      "loss": 0.2527,
      "num_input_tokens_seen": 58949632,
      "step": 30630
    },
    {
      "epoch": 4.99763439105963,
      "grad_norm": 0.0037639455404132605,
      "learning_rate": 0.038786154410914535,
      "loss": 0.2416,
      "num_input_tokens_seen": 58958944,
      "step": 30635
    },
    {
      "epoch": 4.998450118280447,
      "grad_norm": 0.0014665477210655808,
      "learning_rate": 0.03874663576063917,
      "loss": 0.1617,
      "num_input_tokens_seen": 58967296,
      "step": 30640
    },
    {
      "epoch": 4.999265845501265,
      "grad_norm": 0.0018278375500813127,
      "learning_rate": 0.038707134267030624,
      "loss": 0.18,
      "num_input_tokens_seen": 58978032,
      "step": 30645
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0037460876628756523,
      "learning_rate": 0.038667649936180555,
      "loss": 0.2009,
      "num_input_tokens_seen": 58987136,
      "step": 30650
    },
    {
      "epoch": 5.000815727220817,
      "grad_norm": 0.0027203597128391266,
      "learning_rate": 0.038628182774178,
      "loss": 0.2019,
      "num_input_tokens_seen": 58997088,
      "step": 30655
    },
    {
      "epoch": 5.001631454441635,
      "grad_norm": 0.0033448380418121815,
      "learning_rate": 0.038588732787109226,
      "loss": 0.2057,
      "num_input_tokens_seen": 59006448,
      "step": 30660
    },
    {
      "epoch": 5.002447181662452,
      "grad_norm": 0.0016575531335547566,
      "learning_rate": 0.03854929998105795,
      "loss": 0.1656,
      "num_input_tokens_seen": 59016400,
      "step": 30665
    },
    {
      "epoch": 5.003262908883269,
      "grad_norm": 0.0023792008869349957,
      "learning_rate": 0.03850988436210518,
      "loss": 0.1678,
      "num_input_tokens_seen": 59026784,
      "step": 30670
    },
    {
      "epoch": 5.0040786361040865,
      "grad_norm": 0.0024266683030873537,
      "learning_rate": 0.03847048593632933,
      "loss": 0.16,
      "num_input_tokens_seen": 59036960,
      "step": 30675
    },
    {
      "epoch": 5.004894363324905,
      "grad_norm": 0.0017631885129958391,
      "learning_rate": 0.038431104709806096,
      "loss": 0.1738,
      "num_input_tokens_seen": 59046400,
      "step": 30680
    },
    {
      "epoch": 5.005710090545722,
      "grad_norm": 0.0023316852748394012,
      "learning_rate": 0.0383917406886086,
      "loss": 0.1927,
      "num_input_tokens_seen": 59056208,
      "step": 30685
    },
    {
      "epoch": 5.006525817766539,
      "grad_norm": 0.002393920673057437,
      "learning_rate": 0.03835239387880722,
      "loss": 0.1366,
      "num_input_tokens_seen": 59065392,
      "step": 30690
    },
    {
      "epoch": 5.007341544987356,
      "grad_norm": 0.0024128658697009087,
      "learning_rate": 0.03831306428646979,
      "loss": 0.1928,
      "num_input_tokens_seen": 59076560,
      "step": 30695
    },
    {
      "epoch": 5.008157272208174,
      "grad_norm": 0.0016270180931314826,
      "learning_rate": 0.03827375191766135,
      "loss": 0.1717,
      "num_input_tokens_seen": 59085824,
      "step": 30700
    },
    {
      "epoch": 5.008972999428991,
      "grad_norm": 0.0012236819602549076,
      "learning_rate": 0.03823445677844446,
      "loss": 0.1668,
      "num_input_tokens_seen": 59095840,
      "step": 30705
    },
    {
      "epoch": 5.009788726649808,
      "grad_norm": 0.0020495287608355284,
      "learning_rate": 0.03819517887487881,
      "loss": 0.2639,
      "num_input_tokens_seen": 59104816,
      "step": 30710
    },
    {
      "epoch": 5.010604453870625,
      "grad_norm": 0.0028706274461001158,
      "learning_rate": 0.03815591821302161,
      "loss": 0.1673,
      "num_input_tokens_seen": 59113680,
      "step": 30715
    },
    {
      "epoch": 5.011420181091443,
      "grad_norm": 0.0021089082583785057,
      "learning_rate": 0.03811667479892739,
      "loss": 0.1674,
      "num_input_tokens_seen": 59122144,
      "step": 30720
    },
    {
      "epoch": 5.0122359083122605,
      "grad_norm": 0.003597815753892064,
      "learning_rate": 0.03807744863864788,
      "loss": 0.1887,
      "num_input_tokens_seen": 59132240,
      "step": 30725
    },
    {
      "epoch": 5.013051635533078,
      "grad_norm": 0.001316236099228263,
      "learning_rate": 0.03803823973823229,
      "loss": 0.164,
      "num_input_tokens_seen": 59141792,
      "step": 30730
    },
    {
      "epoch": 5.013867362753895,
      "grad_norm": 0.0020008257124572992,
      "learning_rate": 0.03799904810372719,
      "loss": 0.1824,
      "num_input_tokens_seen": 59151728,
      "step": 30735
    },
    {
      "epoch": 5.014683089974713,
      "grad_norm": 0.002534707775339484,
      "learning_rate": 0.03795987374117632,
      "loss": 0.1549,
      "num_input_tokens_seen": 59161472,
      "step": 30740
    },
    {
      "epoch": 5.01549881719553,
      "grad_norm": 0.0015503141330555081,
      "learning_rate": 0.03792071665662093,
      "loss": 0.1898,
      "num_input_tokens_seen": 59170672,
      "step": 30745
    },
    {
      "epoch": 5.016314544416347,
      "grad_norm": 0.0023627516347914934,
      "learning_rate": 0.03788157685609952,
      "loss": 0.1366,
      "num_input_tokens_seen": 59180784,
      "step": 30750
    },
    {
      "epoch": 5.017130271637164,
      "grad_norm": 0.0026046298444271088,
      "learning_rate": 0.037842454345647876,
      "loss": 0.1774,
      "num_input_tokens_seen": 59190624,
      "step": 30755
    },
    {
      "epoch": 5.017945998857982,
      "grad_norm": 0.002080318983644247,
      "learning_rate": 0.03780334913129929,
      "loss": 0.1751,
      "num_input_tokens_seen": 59200672,
      "step": 30760
    },
    {
      "epoch": 5.018761726078799,
      "grad_norm": 0.0034196183551102877,
      "learning_rate": 0.037764261219084175,
      "loss": 0.165,
      "num_input_tokens_seen": 59211680,
      "step": 30765
    },
    {
      "epoch": 5.0195774532996165,
      "grad_norm": 0.003331991611048579,
      "learning_rate": 0.037725190615030414,
      "loss": 0.1464,
      "num_input_tokens_seen": 59220528,
      "step": 30770
    },
    {
      "epoch": 5.020393180520434,
      "grad_norm": 0.0014670818345621228,
      "learning_rate": 0.037686137325163224,
      "loss": 0.16,
      "num_input_tokens_seen": 59230688,
      "step": 30775
    },
    {
      "epoch": 5.021208907741252,
      "grad_norm": 0.001705160248093307,
      "learning_rate": 0.037647101355505065,
      "loss": 0.1596,
      "num_input_tokens_seen": 59239568,
      "step": 30780
    },
    {
      "epoch": 5.022024634962069,
      "grad_norm": 0.003869543317705393,
      "learning_rate": 0.03760808271207581,
      "loss": 0.284,
      "num_input_tokens_seen": 59249376,
      "step": 30785
    },
    {
      "epoch": 5.022840362182886,
      "grad_norm": 0.0034225049894303083,
      "learning_rate": 0.03756908140089258,
      "loss": 0.259,
      "num_input_tokens_seen": 59258864,
      "step": 30790
    },
    {
      "epoch": 5.023656089403703,
      "grad_norm": 0.003495404263958335,
      "learning_rate": 0.03753009742796989,
      "loss": 0.1743,
      "num_input_tokens_seen": 59270192,
      "step": 30795
    },
    {
      "epoch": 5.024471816624521,
      "grad_norm": 0.005673824343830347,
      "learning_rate": 0.037491130799319615,
      "loss": 0.1802,
      "num_input_tokens_seen": 59278112,
      "step": 30800
    },
    {
      "epoch": 5.024471816624521,
      "eval_loss": 0.19284726679325104,
      "eval_runtime": 67.9811,
      "eval_samples_per_second": 40.085,
      "eval_steps_per_second": 20.05,
      "num_input_tokens_seen": 59278112,
      "step": 30800
    },
    {
      "epoch": 5.025287543845338,
      "grad_norm": 0.0032897002529352903,
      "learning_rate": 0.03745218152095079,
      "loss": 0.19,
      "num_input_tokens_seen": 59288064,
      "step": 30805
    },
    {
      "epoch": 5.026103271066155,
      "grad_norm": 0.003061596304178238,
      "learning_rate": 0.037413249598869935,
      "loss": 0.1882,
      "num_input_tokens_seen": 59296688,
      "step": 30810
    },
    {
      "epoch": 5.026918998286972,
      "grad_norm": 0.0017808409174904227,
      "learning_rate": 0.037374335039080886,
      "loss": 0.2199,
      "num_input_tokens_seen": 59306880,
      "step": 30815
    },
    {
      "epoch": 5.02773472550779,
      "grad_norm": 0.002727626357227564,
      "learning_rate": 0.037335437847584724,
      "loss": 0.2103,
      "num_input_tokens_seen": 59315184,
      "step": 30820
    },
    {
      "epoch": 5.028550452728608,
      "grad_norm": 0.0027123193722218275,
      "learning_rate": 0.03729655803037983,
      "loss": 0.163,
      "num_input_tokens_seen": 59324208,
      "step": 30825
    },
    {
      "epoch": 5.029366179949425,
      "grad_norm": 0.0024080886505544186,
      "learning_rate": 0.03725769559346207,
      "loss": 0.1736,
      "num_input_tokens_seen": 59333712,
      "step": 30830
    },
    {
      "epoch": 5.030181907170242,
      "grad_norm": 0.005086757708340883,
      "learning_rate": 0.03721885054282439,
      "loss": 0.2349,
      "num_input_tokens_seen": 59341360,
      "step": 30835
    },
    {
      "epoch": 5.03099763439106,
      "grad_norm": 0.0026404662057757378,
      "learning_rate": 0.03718002288445731,
      "loss": 0.2018,
      "num_input_tokens_seen": 59351184,
      "step": 30840
    },
    {
      "epoch": 5.031813361611877,
      "grad_norm": 0.0018783269915729761,
      "learning_rate": 0.03714121262434844,
      "loss": 0.1234,
      "num_input_tokens_seen": 59361520,
      "step": 30845
    },
    {
      "epoch": 5.032629088832694,
      "grad_norm": 0.0033574721310287714,
      "learning_rate": 0.037102419768482844,
      "loss": 0.1909,
      "num_input_tokens_seen": 59370736,
      "step": 30850
    },
    {
      "epoch": 5.033444816053512,
      "grad_norm": 0.004718451295047998,
      "learning_rate": 0.03706364432284293,
      "loss": 0.2156,
      "num_input_tokens_seen": 59379920,
      "step": 30855
    },
    {
      "epoch": 5.034260543274329,
      "grad_norm": 0.0020490321330726147,
      "learning_rate": 0.03702488629340828,
      "loss": 0.2164,
      "num_input_tokens_seen": 59389552,
      "step": 30860
    },
    {
      "epoch": 5.035076270495146,
      "grad_norm": 0.0050478726625442505,
      "learning_rate": 0.036986145686155915,
      "loss": 0.2027,
      "num_input_tokens_seen": 59399792,
      "step": 30865
    },
    {
      "epoch": 5.0358919977159635,
      "grad_norm": 0.002254816237837076,
      "learning_rate": 0.036947422507060075,
      "loss": 0.1575,
      "num_input_tokens_seen": 59409216,
      "step": 30870
    },
    {
      "epoch": 5.0367077249367815,
      "grad_norm": 0.0027019272092729807,
      "learning_rate": 0.0369087167620924,
      "loss": 0.1773,
      "num_input_tokens_seen": 59419760,
      "step": 30875
    },
    {
      "epoch": 5.037523452157599,
      "grad_norm": 0.002214191248640418,
      "learning_rate": 0.03687002845722183,
      "loss": 0.1649,
      "num_input_tokens_seen": 59428928,
      "step": 30880
    },
    {
      "epoch": 5.038339179378416,
      "grad_norm": 0.0017638558056205511,
      "learning_rate": 0.03683135759841451,
      "loss": 0.1463,
      "num_input_tokens_seen": 59439440,
      "step": 30885
    },
    {
      "epoch": 5.039154906599233,
      "grad_norm": 0.0035466335248202085,
      "learning_rate": 0.03679270419163406,
      "loss": 0.1956,
      "num_input_tokens_seen": 59448544,
      "step": 30890
    },
    {
      "epoch": 5.039970633820051,
      "grad_norm": 0.0018672611331567168,
      "learning_rate": 0.03675406824284127,
      "loss": 0.1927,
      "num_input_tokens_seen": 59457904,
      "step": 30895
    },
    {
      "epoch": 5.040786361040868,
      "grad_norm": 0.0022759835701435804,
      "learning_rate": 0.03671544975799425,
      "loss": 0.16,
      "num_input_tokens_seen": 59466768,
      "step": 30900
    },
    {
      "epoch": 5.041602088261685,
      "grad_norm": 0.003076797118410468,
      "learning_rate": 0.03667684874304854,
      "loss": 0.1232,
      "num_input_tokens_seen": 59476176,
      "step": 30905
    },
    {
      "epoch": 5.042417815482502,
      "grad_norm": 0.0030526157934218645,
      "learning_rate": 0.03663826520395683,
      "loss": 0.1685,
      "num_input_tokens_seen": 59485728,
      "step": 30910
    },
    {
      "epoch": 5.04323354270332,
      "grad_norm": 0.0027834437787532806,
      "learning_rate": 0.03659969914666922,
      "loss": 0.2462,
      "num_input_tokens_seen": 59495120,
      "step": 30915
    },
    {
      "epoch": 5.0440492699241375,
      "grad_norm": 0.0026760597247630358,
      "learning_rate": 0.036561150577133106,
      "loss": 0.2221,
      "num_input_tokens_seen": 59505328,
      "step": 30920
    },
    {
      "epoch": 5.044864997144955,
      "grad_norm": 0.002310805954039097,
      "learning_rate": 0.036522619501293103,
      "loss": 0.1689,
      "num_input_tokens_seen": 59515232,
      "step": 30925
    },
    {
      "epoch": 5.045680724365772,
      "grad_norm": 0.0036774754989892244,
      "learning_rate": 0.03648410592509122,
      "loss": 0.1704,
      "num_input_tokens_seen": 59524640,
      "step": 30930
    },
    {
      "epoch": 5.04649645158659,
      "grad_norm": 0.0016047765966504812,
      "learning_rate": 0.03644560985446676,
      "loss": 0.1861,
      "num_input_tokens_seen": 59534624,
      "step": 30935
    },
    {
      "epoch": 5.047312178807407,
      "grad_norm": 0.0034589292481541634,
      "learning_rate": 0.036407131295356256,
      "loss": 0.167,
      "num_input_tokens_seen": 59544128,
      "step": 30940
    },
    {
      "epoch": 5.048127906028224,
      "grad_norm": 0.0012988474918529391,
      "learning_rate": 0.03636867025369362,
      "loss": 0.1423,
      "num_input_tokens_seen": 59555344,
      "step": 30945
    },
    {
      "epoch": 5.048943633249041,
      "grad_norm": 0.0035438560880720615,
      "learning_rate": 0.03633022673540999,
      "loss": 0.2086,
      "num_input_tokens_seen": 59564832,
      "step": 30950
    },
    {
      "epoch": 5.049759360469859,
      "grad_norm": 0.0023133570794016123,
      "learning_rate": 0.03629180074643385,
      "loss": 0.1341,
      "num_input_tokens_seen": 59574176,
      "step": 30955
    },
    {
      "epoch": 5.050575087690676,
      "grad_norm": 0.0026394387241452932,
      "learning_rate": 0.03625339229269102,
      "loss": 0.2199,
      "num_input_tokens_seen": 59583504,
      "step": 30960
    },
    {
      "epoch": 5.051390814911493,
      "grad_norm": 0.002248150296509266,
      "learning_rate": 0.036215001380104535,
      "loss": 0.1332,
      "num_input_tokens_seen": 59594816,
      "step": 30965
    },
    {
      "epoch": 5.052206542132311,
      "grad_norm": 0.0016193726332858205,
      "learning_rate": 0.03617662801459471,
      "loss": 0.1949,
      "num_input_tokens_seen": 59604288,
      "step": 30970
    },
    {
      "epoch": 5.053022269353129,
      "grad_norm": 0.0024936345871537924,
      "learning_rate": 0.036138272202079276,
      "loss": 0.1434,
      "num_input_tokens_seen": 59614960,
      "step": 30975
    },
    {
      "epoch": 5.053837996573946,
      "grad_norm": 0.0032029002904891968,
      "learning_rate": 0.036099933948473106,
      "loss": 0.2178,
      "num_input_tokens_seen": 59624720,
      "step": 30980
    },
    {
      "epoch": 5.054653723794763,
      "grad_norm": 0.0015684240497648716,
      "learning_rate": 0.03606161325968851,
      "loss": 0.1407,
      "num_input_tokens_seen": 59634224,
      "step": 30985
    },
    {
      "epoch": 5.05546945101558,
      "grad_norm": 0.001954103820025921,
      "learning_rate": 0.03602331014163496,
      "loss": 0.1464,
      "num_input_tokens_seen": 59643664,
      "step": 30990
    },
    {
      "epoch": 5.056285178236398,
      "grad_norm": 0.0019793554674834013,
      "learning_rate": 0.035985024600219295,
      "loss": 0.2218,
      "num_input_tokens_seen": 59654096,
      "step": 30995
    },
    {
      "epoch": 5.057100905457215,
      "grad_norm": 0.0026544707361608744,
      "learning_rate": 0.03594675664134569,
      "loss": 0.1849,
      "num_input_tokens_seen": 59663264,
      "step": 31000
    },
    {
      "epoch": 5.057100905457215,
      "eval_loss": 0.19319574534893036,
      "eval_runtime": 68.0575,
      "eval_samples_per_second": 40.04,
      "eval_steps_per_second": 20.027,
      "num_input_tokens_seen": 59663264,
      "step": 31000
    },
    {
      "epoch": 5.057916632678032,
      "grad_norm": 0.00553317554295063,
      "learning_rate": 0.03590850627091545,
      "loss": 0.213,
      "num_input_tokens_seen": 59672192,
      "step": 31005
    },
    {
      "epoch": 5.058732359898849,
      "grad_norm": 0.0025002507027238607,
      "learning_rate": 0.03587027349482731,
      "loss": 0.1639,
      "num_input_tokens_seen": 59682496,
      "step": 31010
    },
    {
      "epoch": 5.059548087119667,
      "grad_norm": 0.0036477362737059593,
      "learning_rate": 0.035832058318977275,
      "loss": 0.1957,
      "num_input_tokens_seen": 59692608,
      "step": 31015
    },
    {
      "epoch": 5.0603638143404845,
      "grad_norm": 0.0048903850838541985,
      "learning_rate": 0.03579386074925853,
      "loss": 0.1537,
      "num_input_tokens_seen": 59702304,
      "step": 31020
    },
    {
      "epoch": 5.061179541561302,
      "grad_norm": 0.003233400173485279,
      "learning_rate": 0.035755680791561696,
      "loss": 0.1614,
      "num_input_tokens_seen": 59710688,
      "step": 31025
    },
    {
      "epoch": 5.06199526878212,
      "grad_norm": 0.003571507753804326,
      "learning_rate": 0.03571751845177454,
      "loss": 0.2636,
      "num_input_tokens_seen": 59719696,
      "step": 31030
    },
    {
      "epoch": 5.062810996002937,
      "grad_norm": 0.0018374884966760874,
      "learning_rate": 0.03567937373578225,
      "loss": 0.1548,
      "num_input_tokens_seen": 59730336,
      "step": 31035
    },
    {
      "epoch": 5.063626723223754,
      "grad_norm": 0.002927886787801981,
      "learning_rate": 0.03564124664946711,
      "loss": 0.2011,
      "num_input_tokens_seen": 59738304,
      "step": 31040
    },
    {
      "epoch": 5.064442450444571,
      "grad_norm": 0.0042747436091303825,
      "learning_rate": 0.035603137198708924,
      "loss": 0.1837,
      "num_input_tokens_seen": 59748032,
      "step": 31045
    },
    {
      "epoch": 5.065258177665389,
      "grad_norm": 0.0024903651792556047,
      "learning_rate": 0.035565045389384514,
      "loss": 0.1798,
      "num_input_tokens_seen": 59757712,
      "step": 31050
    },
    {
      "epoch": 5.066073904886206,
      "grad_norm": 0.001963977934792638,
      "learning_rate": 0.03552697122736823,
      "loss": 0.0761,
      "num_input_tokens_seen": 59767632,
      "step": 31055
    },
    {
      "epoch": 5.066889632107023,
      "grad_norm": 0.0018976108403876424,
      "learning_rate": 0.03548891471853153,
      "loss": 0.1336,
      "num_input_tokens_seen": 59777104,
      "step": 31060
    },
    {
      "epoch": 5.0677053593278405,
      "grad_norm": 0.0016581180971115828,
      "learning_rate": 0.03545087586874322,
      "loss": 0.1724,
      "num_input_tokens_seen": 59787040,
      "step": 31065
    },
    {
      "epoch": 5.0685210865486585,
      "grad_norm": 0.0010565082775428891,
      "learning_rate": 0.03541285468386935,
      "loss": 0.1196,
      "num_input_tokens_seen": 59796688,
      "step": 31070
    },
    {
      "epoch": 5.069336813769476,
      "grad_norm": 0.0023737773299217224,
      "learning_rate": 0.03537485116977327,
      "loss": 0.1627,
      "num_input_tokens_seen": 59806720,
      "step": 31075
    },
    {
      "epoch": 5.070152540990293,
      "grad_norm": 0.00215906766243279,
      "learning_rate": 0.03533686533231565,
      "loss": 0.1311,
      "num_input_tokens_seen": 59816432,
      "step": 31080
    },
    {
      "epoch": 5.07096826821111,
      "grad_norm": 0.0018752007745206356,
      "learning_rate": 0.0352988971773543,
      "loss": 0.1533,
      "num_input_tokens_seen": 59825552,
      "step": 31085
    },
    {
      "epoch": 5.071783995431928,
      "grad_norm": 0.002353373449295759,
      "learning_rate": 0.03526094671074443,
      "loss": 0.2173,
      "num_input_tokens_seen": 59835040,
      "step": 31090
    },
    {
      "epoch": 5.072599722652745,
      "grad_norm": 0.002284771064296365,
      "learning_rate": 0.03522301393833852,
      "loss": 0.1543,
      "num_input_tokens_seen": 59844928,
      "step": 31095
    },
    {
      "epoch": 5.073415449873562,
      "grad_norm": 0.007364385295659304,
      "learning_rate": 0.035185098865986204,
      "loss": 0.1486,
      "num_input_tokens_seen": 59855408,
      "step": 31100
    },
    {
      "epoch": 5.074231177094379,
      "grad_norm": 0.004401363432407379,
      "learning_rate": 0.03514720149953453,
      "loss": 0.214,
      "num_input_tokens_seen": 59865776,
      "step": 31105
    },
    {
      "epoch": 5.075046904315197,
      "grad_norm": 0.0033271959982812405,
      "learning_rate": 0.03510932184482773,
      "loss": 0.1733,
      "num_input_tokens_seen": 59875584,
      "step": 31110
    },
    {
      "epoch": 5.0758626315360145,
      "grad_norm": 0.003592481603845954,
      "learning_rate": 0.03507145990770724,
      "loss": 0.1996,
      "num_input_tokens_seen": 59884928,
      "step": 31115
    },
    {
      "epoch": 5.076678358756832,
      "grad_norm": 0.00483968760818243,
      "learning_rate": 0.035033615694011984,
      "loss": 0.1716,
      "num_input_tokens_seen": 59893936,
      "step": 31120
    },
    {
      "epoch": 5.077494085977649,
      "grad_norm": 0.004118538461625576,
      "learning_rate": 0.03499578920957788,
      "loss": 0.215,
      "num_input_tokens_seen": 59903120,
      "step": 31125
    },
    {
      "epoch": 5.078309813198467,
      "grad_norm": 0.004757577553391457,
      "learning_rate": 0.034957980460238375,
      "loss": 0.1612,
      "num_input_tokens_seen": 59911824,
      "step": 31130
    },
    {
      "epoch": 5.079125540419284,
      "grad_norm": 0.0033168538939207792,
      "learning_rate": 0.03492018945182393,
      "loss": 0.2612,
      "num_input_tokens_seen": 59921168,
      "step": 31135
    },
    {
      "epoch": 5.079941267640101,
      "grad_norm": 0.0013816022546961904,
      "learning_rate": 0.03488241619016247,
      "loss": 0.167,
      "num_input_tokens_seen": 59930144,
      "step": 31140
    },
    {
      "epoch": 5.080756994860918,
      "grad_norm": 0.0037006502971053123,
      "learning_rate": 0.03484466068107913,
      "loss": 0.217,
      "num_input_tokens_seen": 59940800,
      "step": 31145
    },
    {
      "epoch": 5.081572722081736,
      "grad_norm": 0.003579862415790558,
      "learning_rate": 0.034806922930396195,
      "loss": 0.2014,
      "num_input_tokens_seen": 59950048,
      "step": 31150
    },
    {
      "epoch": 5.082388449302553,
      "grad_norm": 0.004236187785863876,
      "learning_rate": 0.03476920294393337,
      "loss": 0.158,
      "num_input_tokens_seen": 59959584,
      "step": 31155
    },
    {
      "epoch": 5.08320417652337,
      "grad_norm": 0.002820665016770363,
      "learning_rate": 0.03473150072750755,
      "loss": 0.1835,
      "num_input_tokens_seen": 59969216,
      "step": 31160
    },
    {
      "epoch": 5.0840199037441876,
      "grad_norm": 0.0022462436463683844,
      "learning_rate": 0.03469381628693284,
      "loss": 0.2399,
      "num_input_tokens_seen": 59977616,
      "step": 31165
    },
    {
      "epoch": 5.084835630965006,
      "grad_norm": 0.003817073768004775,
      "learning_rate": 0.03465614962802072,
      "loss": 0.2282,
      "num_input_tokens_seen": 59986768,
      "step": 31170
    },
    {
      "epoch": 5.085651358185823,
      "grad_norm": 0.0033759810030460358,
      "learning_rate": 0.0346185007565798,
      "loss": 0.1848,
      "num_input_tokens_seen": 59997792,
      "step": 31175
    },
    {
      "epoch": 5.08646708540664,
      "grad_norm": 0.005181970540434122,
      "learning_rate": 0.03458086967841609,
      "loss": 0.2338,
      "num_input_tokens_seen": 60007920,
      "step": 31180
    },
    {
      "epoch": 5.087282812627457,
      "grad_norm": 0.0025694568175822496,
      "learning_rate": 0.03454325639933266,
      "loss": 0.1919,
      "num_input_tokens_seen": 60016752,
      "step": 31185
    },
    {
      "epoch": 5.088098539848275,
      "grad_norm": 0.0054982127621769905,
      "learning_rate": 0.03450566092513007,
      "loss": 0.1818,
      "num_input_tokens_seen": 60026752,
      "step": 31190
    },
    {
      "epoch": 5.088914267069092,
      "grad_norm": 0.004003638867288828,
      "learning_rate": 0.034468083261605914,
      "loss": 0.2227,
      "num_input_tokens_seen": 60036560,
      "step": 31195
    },
    {
      "epoch": 5.089729994289909,
      "grad_norm": 0.006565483752638102,
      "learning_rate": 0.03443052341455522,
      "loss": 0.196,
      "num_input_tokens_seen": 60047056,
      "step": 31200
    },
    {
      "epoch": 5.089729994289909,
      "eval_loss": 0.195194274187088,
      "eval_runtime": 68.0393,
      "eval_samples_per_second": 40.05,
      "eval_steps_per_second": 20.033,
      "num_input_tokens_seen": 60047056,
      "step": 31200
    },
    {
      "epoch": 5.090545721510727,
      "grad_norm": 0.0021097289863973856,
      "learning_rate": 0.0343929813897701,
      "loss": 0.1791,
      "num_input_tokens_seen": 60057440,
      "step": 31205
    },
    {
      "epoch": 5.091361448731544,
      "grad_norm": 0.0016160798259079456,
      "learning_rate": 0.034355457193040125,
      "loss": 0.1806,
      "num_input_tokens_seen": 60066672,
      "step": 31210
    },
    {
      "epoch": 5.0921771759523615,
      "grad_norm": 0.003251161426305771,
      "learning_rate": 0.03431795083015186,
      "loss": 0.1881,
      "num_input_tokens_seen": 60077088,
      "step": 31215
    },
    {
      "epoch": 5.092992903173179,
      "grad_norm": 0.0035842973738908768,
      "learning_rate": 0.03428046230688936,
      "loss": 0.2375,
      "num_input_tokens_seen": 60087248,
      "step": 31220
    },
    {
      "epoch": 5.093808630393997,
      "grad_norm": 0.003037200076505542,
      "learning_rate": 0.034242991629033805,
      "loss": 0.1541,
      "num_input_tokens_seen": 60097344,
      "step": 31225
    },
    {
      "epoch": 5.094624357614814,
      "grad_norm": 0.004190393723547459,
      "learning_rate": 0.03420553880236362,
      "loss": 0.1725,
      "num_input_tokens_seen": 60107488,
      "step": 31230
    },
    {
      "epoch": 5.095440084835631,
      "grad_norm": 0.005642409902065992,
      "learning_rate": 0.03416810383265449,
      "loss": 0.2064,
      "num_input_tokens_seen": 60116960,
      "step": 31235
    },
    {
      "epoch": 5.096255812056448,
      "grad_norm": 0.0033600665628910065,
      "learning_rate": 0.03413068672567944,
      "loss": 0.206,
      "num_input_tokens_seen": 60126160,
      "step": 31240
    },
    {
      "epoch": 5.097071539277266,
      "grad_norm": 0.0018978111911565065,
      "learning_rate": 0.034093287487208565,
      "loss": 0.1685,
      "num_input_tokens_seen": 60135872,
      "step": 31245
    },
    {
      "epoch": 5.097887266498083,
      "grad_norm": 0.002065549837425351,
      "learning_rate": 0.03405590612300937,
      "loss": 0.1346,
      "num_input_tokens_seen": 60145184,
      "step": 31250
    },
    {
      "epoch": 5.0987029937189,
      "grad_norm": 0.0017373096197843552,
      "learning_rate": 0.03401854263884646,
      "loss": 0.1661,
      "num_input_tokens_seen": 60155664,
      "step": 31255
    },
    {
      "epoch": 5.0995187209397175,
      "grad_norm": 0.002436535432934761,
      "learning_rate": 0.033981197040481824,
      "loss": 0.1583,
      "num_input_tokens_seen": 60165680,
      "step": 31260
    },
    {
      "epoch": 5.1003344481605355,
      "grad_norm": 0.0020086532458662987,
      "learning_rate": 0.03394386933367459,
      "loss": 0.1313,
      "num_input_tokens_seen": 60175008,
      "step": 31265
    },
    {
      "epoch": 5.101150175381353,
      "grad_norm": 0.0033370398450642824,
      "learning_rate": 0.033906559524181104,
      "loss": 0.1786,
      "num_input_tokens_seen": 60183776,
      "step": 31270
    },
    {
      "epoch": 5.10196590260217,
      "grad_norm": 0.0033627597149461508,
      "learning_rate": 0.033869267617755085,
      "loss": 0.1957,
      "num_input_tokens_seen": 60192928,
      "step": 31275
    },
    {
      "epoch": 5.102781629822987,
      "grad_norm": 0.001629007514566183,
      "learning_rate": 0.0338319936201474,
      "loss": 0.1649,
      "num_input_tokens_seen": 60202288,
      "step": 31280
    },
    {
      "epoch": 5.103597357043805,
      "grad_norm": 0.0022456867154687643,
      "learning_rate": 0.033794737537106136,
      "loss": 0.208,
      "num_input_tokens_seen": 60211872,
      "step": 31285
    },
    {
      "epoch": 5.104413084264622,
      "grad_norm": 0.004898780956864357,
      "learning_rate": 0.03375749937437671,
      "loss": 0.2251,
      "num_input_tokens_seen": 60220112,
      "step": 31290
    },
    {
      "epoch": 5.105228811485439,
      "grad_norm": 0.0036355110350996256,
      "learning_rate": 0.033720279137701634,
      "loss": 0.171,
      "num_input_tokens_seen": 60229856,
      "step": 31295
    },
    {
      "epoch": 5.106044538706256,
      "grad_norm": 0.001740667619742453,
      "learning_rate": 0.03368307683282078,
      "loss": 0.1602,
      "num_input_tokens_seen": 60239248,
      "step": 31300
    },
    {
      "epoch": 5.106860265927074,
      "grad_norm": 0.0018640790367498994,
      "learning_rate": 0.033645892465471235,
      "loss": 0.1848,
      "num_input_tokens_seen": 60249168,
      "step": 31305
    },
    {
      "epoch": 5.1076759931478914,
      "grad_norm": 0.0017366851679980755,
      "learning_rate": 0.03360872604138724,
      "loss": 0.1495,
      "num_input_tokens_seen": 60258896,
      "step": 31310
    },
    {
      "epoch": 5.108491720368709,
      "grad_norm": 0.0027098043356090784,
      "learning_rate": 0.03357157756630034,
      "loss": 0.1622,
      "num_input_tokens_seen": 60268400,
      "step": 31315
    },
    {
      "epoch": 5.109307447589526,
      "grad_norm": 0.003928487189114094,
      "learning_rate": 0.033534447045939365,
      "loss": 0.1993,
      "num_input_tokens_seen": 60278784,
      "step": 31320
    },
    {
      "epoch": 5.110123174810344,
      "grad_norm": 0.0013201053952798247,
      "learning_rate": 0.03349733448603026,
      "loss": 0.1611,
      "num_input_tokens_seen": 60289472,
      "step": 31325
    },
    {
      "epoch": 5.110938902031161,
      "grad_norm": 0.004409788176417351,
      "learning_rate": 0.03346023989229619,
      "loss": 0.1946,
      "num_input_tokens_seen": 60298432,
      "step": 31330
    },
    {
      "epoch": 5.111754629251978,
      "grad_norm": 0.004117804113775492,
      "learning_rate": 0.03342316327045769,
      "loss": 0.2421,
      "num_input_tokens_seen": 60308352,
      "step": 31335
    },
    {
      "epoch": 5.112570356472795,
      "grad_norm": 0.0018746560672298074,
      "learning_rate": 0.033386104626232385,
      "loss": 0.1548,
      "num_input_tokens_seen": 60317600,
      "step": 31340
    },
    {
      "epoch": 5.113386083693613,
      "grad_norm": 0.005978402215987444,
      "learning_rate": 0.03334906396533525,
      "loss": 0.2044,
      "num_input_tokens_seen": 60327232,
      "step": 31345
    },
    {
      "epoch": 5.11420181091443,
      "grad_norm": 0.00636946689337492,
      "learning_rate": 0.033312041293478326,
      "loss": 0.2418,
      "num_input_tokens_seen": 60337664,
      "step": 31350
    },
    {
      "epoch": 5.115017538135247,
      "grad_norm": 0.0022813095711171627,
      "learning_rate": 0.03327503661637103,
      "loss": 0.1863,
      "num_input_tokens_seen": 60347760,
      "step": 31355
    },
    {
      "epoch": 5.1158332653560645,
      "grad_norm": 0.003893311368301511,
      "learning_rate": 0.03323804993971998,
      "loss": 0.2323,
      "num_input_tokens_seen": 60358272,
      "step": 31360
    },
    {
      "epoch": 5.1166489925768825,
      "grad_norm": 0.003913565073162317,
      "learning_rate": 0.033201081269228924,
      "loss": 0.1745,
      "num_input_tokens_seen": 60368208,
      "step": 31365
    },
    {
      "epoch": 5.1174647197977,
      "grad_norm": 0.0039878604002296925,
      "learning_rate": 0.03316413061059895,
      "loss": 0.2017,
      "num_input_tokens_seen": 60377216,
      "step": 31370
    },
    {
      "epoch": 5.118280447018517,
      "grad_norm": 0.002471928484737873,
      "learning_rate": 0.03312719796952827,
      "loss": 0.1358,
      "num_input_tokens_seen": 60385888,
      "step": 31375
    },
    {
      "epoch": 5.119096174239334,
      "grad_norm": 0.0012562695192173123,
      "learning_rate": 0.03309028335171236,
      "loss": 0.1274,
      "num_input_tokens_seen": 60395456,
      "step": 31380
    },
    {
      "epoch": 5.119911901460152,
      "grad_norm": 0.0020023793913424015,
      "learning_rate": 0.03305338676284398,
      "loss": 0.1598,
      "num_input_tokens_seen": 60404944,
      "step": 31385
    },
    {
      "epoch": 5.120727628680969,
      "grad_norm": 0.0025756782852113247,
      "learning_rate": 0.03301650820861296,
      "loss": 0.1936,
      "num_input_tokens_seen": 60414816,
      "step": 31390
    },
    {
      "epoch": 5.121543355901786,
      "grad_norm": 0.002114309696480632,
      "learning_rate": 0.03297964769470652,
      "loss": 0.1511,
      "num_input_tokens_seen": 60424448,
      "step": 31395
    },
    {
      "epoch": 5.122359083122603,
      "grad_norm": 0.0036199779715389013,
      "learning_rate": 0.032942805226808945,
      "loss": 0.194,
      "num_input_tokens_seen": 60433680,
      "step": 31400
    },
    {
      "epoch": 5.122359083122603,
      "eval_loss": 0.1892799437046051,
      "eval_runtime": 67.9909,
      "eval_samples_per_second": 40.079,
      "eval_steps_per_second": 20.047,
      "num_input_tokens_seen": 60433680,
      "step": 31400
    },
    {
      "epoch": 5.123174810343421,
      "grad_norm": 0.004321994259953499,
      "learning_rate": 0.03290598081060187,
      "loss": 0.1897,
      "num_input_tokens_seen": 60442608,
      "step": 31405
    },
    {
      "epoch": 5.1239905375642385,
      "grad_norm": 0.0018722537206485868,
      "learning_rate": 0.03286917445176407,
      "loss": 0.1799,
      "num_input_tokens_seen": 60452880,
      "step": 31410
    },
    {
      "epoch": 5.124806264785056,
      "grad_norm": 0.0031823862809687853,
      "learning_rate": 0.032832386155971456,
      "loss": 0.2116,
      "num_input_tokens_seen": 60463152,
      "step": 31415
    },
    {
      "epoch": 5.125621992005874,
      "grad_norm": 0.0037841987796127796,
      "learning_rate": 0.032795615928897334,
      "loss": 0.1995,
      "num_input_tokens_seen": 60473280,
      "step": 31420
    },
    {
      "epoch": 5.126437719226691,
      "grad_norm": 0.002269917167723179,
      "learning_rate": 0.03275886377621215,
      "loss": 0.1791,
      "num_input_tokens_seen": 60482816,
      "step": 31425
    },
    {
      "epoch": 5.127253446447508,
      "grad_norm": 0.003570597618818283,
      "learning_rate": 0.03272212970358348,
      "loss": 0.2018,
      "num_input_tokens_seen": 60490704,
      "step": 31430
    },
    {
      "epoch": 5.128069173668325,
      "grad_norm": 0.0023575471714138985,
      "learning_rate": 0.032685413716676215,
      "loss": 0.1493,
      "num_input_tokens_seen": 60500032,
      "step": 31435
    },
    {
      "epoch": 5.128884900889143,
      "grad_norm": 0.0020514230709522963,
      "learning_rate": 0.032648715821152474,
      "loss": 0.1623,
      "num_input_tokens_seen": 60508752,
      "step": 31440
    },
    {
      "epoch": 5.12970062810996,
      "grad_norm": 0.0026697840075939894,
      "learning_rate": 0.03261203602267143,
      "loss": 0.1774,
      "num_input_tokens_seen": 60518352,
      "step": 31445
    },
    {
      "epoch": 5.130516355330777,
      "grad_norm": 0.007571319583803415,
      "learning_rate": 0.03257537432688966,
      "loss": 0.1761,
      "num_input_tokens_seen": 60527248,
      "step": 31450
    },
    {
      "epoch": 5.1313320825515945,
      "grad_norm": 0.004300578497350216,
      "learning_rate": 0.03253873073946077,
      "loss": 0.1722,
      "num_input_tokens_seen": 60535472,
      "step": 31455
    },
    {
      "epoch": 5.1321478097724125,
      "grad_norm": 0.0025432619731873274,
      "learning_rate": 0.03250210526603572,
      "loss": 0.202,
      "num_input_tokens_seen": 60542848,
      "step": 31460
    },
    {
      "epoch": 5.13296353699323,
      "grad_norm": 0.002264895476400852,
      "learning_rate": 0.03246549791226266,
      "loss": 0.1978,
      "num_input_tokens_seen": 60552976,
      "step": 31465
    },
    {
      "epoch": 5.133779264214047,
      "grad_norm": 0.004203835967928171,
      "learning_rate": 0.03242890868378679,
      "loss": 0.1865,
      "num_input_tokens_seen": 60562144,
      "step": 31470
    },
    {
      "epoch": 5.134594991434864,
      "grad_norm": 0.003255401039496064,
      "learning_rate": 0.03239233758625074,
      "loss": 0.2146,
      "num_input_tokens_seen": 60572768,
      "step": 31475
    },
    {
      "epoch": 5.135410718655682,
      "grad_norm": 0.0017340445192530751,
      "learning_rate": 0.032355784625294204,
      "loss": 0.1943,
      "num_input_tokens_seen": 60583472,
      "step": 31480
    },
    {
      "epoch": 5.136226445876499,
      "grad_norm": 0.0013000768376514316,
      "learning_rate": 0.03231924980655402,
      "loss": 0.1506,
      "num_input_tokens_seen": 60592832,
      "step": 31485
    },
    {
      "epoch": 5.137042173097316,
      "grad_norm": 0.0027092215605080128,
      "learning_rate": 0.032282733135664446,
      "loss": 0.1717,
      "num_input_tokens_seen": 60601184,
      "step": 31490
    },
    {
      "epoch": 5.137857900318133,
      "grad_norm": 0.0034382326994091272,
      "learning_rate": 0.03224623461825669,
      "loss": 0.2118,
      "num_input_tokens_seen": 60610512,
      "step": 31495
    },
    {
      "epoch": 5.138673627538951,
      "grad_norm": 0.0019245322328060865,
      "learning_rate": 0.03220975425995937,
      "loss": 0.1681,
      "num_input_tokens_seen": 60620048,
      "step": 31500
    },
    {
      "epoch": 5.139489354759768,
      "grad_norm": 0.00244214478880167,
      "learning_rate": 0.032173292066398206,
      "loss": 0.1479,
      "num_input_tokens_seen": 60628768,
      "step": 31505
    },
    {
      "epoch": 5.1403050819805856,
      "grad_norm": 0.003103426657617092,
      "learning_rate": 0.03213684804319606,
      "loss": 0.1924,
      "num_input_tokens_seen": 60638416,
      "step": 31510
    },
    {
      "epoch": 5.141120809201403,
      "grad_norm": 0.003271537832915783,
      "learning_rate": 0.03210042219597312,
      "loss": 0.161,
      "num_input_tokens_seen": 60648224,
      "step": 31515
    },
    {
      "epoch": 5.141936536422221,
      "grad_norm": 0.0027156639844179153,
      "learning_rate": 0.03206401453034675,
      "loss": 0.1874,
      "num_input_tokens_seen": 60658624,
      "step": 31520
    },
    {
      "epoch": 5.142752263643038,
      "grad_norm": 0.002387943910434842,
      "learning_rate": 0.03202762505193136,
      "loss": 0.1804,
      "num_input_tokens_seen": 60667984,
      "step": 31525
    },
    {
      "epoch": 5.143567990863855,
      "grad_norm": 0.0032584997825324535,
      "learning_rate": 0.031991253766338754,
      "loss": 0.1607,
      "num_input_tokens_seen": 60676800,
      "step": 31530
    },
    {
      "epoch": 5.144383718084672,
      "grad_norm": 0.001871671644039452,
      "learning_rate": 0.03195490067917778,
      "loss": 0.1715,
      "num_input_tokens_seen": 60686928,
      "step": 31535
    },
    {
      "epoch": 5.14519944530549,
      "grad_norm": 0.00244257808662951,
      "learning_rate": 0.03191856579605461,
      "loss": 0.1645,
      "num_input_tokens_seen": 60696048,
      "step": 31540
    },
    {
      "epoch": 5.146015172526307,
      "grad_norm": 0.0018024895107373595,
      "learning_rate": 0.031882249122572454,
      "loss": 0.1673,
      "num_input_tokens_seen": 60706400,
      "step": 31545
    },
    {
      "epoch": 5.146830899747124,
      "grad_norm": 0.0034112560097128153,
      "learning_rate": 0.03184595066433188,
      "loss": 0.1632,
      "num_input_tokens_seen": 60716288,
      "step": 31550
    },
    {
      "epoch": 5.1476466269679415,
      "grad_norm": 0.002756773494184017,
      "learning_rate": 0.03180967042693049,
      "loss": 0.1939,
      "num_input_tokens_seen": 60726256,
      "step": 31555
    },
    {
      "epoch": 5.1484623541887595,
      "grad_norm": 0.0020512347109615803,
      "learning_rate": 0.03177340841596323,
      "loss": 0.1341,
      "num_input_tokens_seen": 60735360,
      "step": 31560
    },
    {
      "epoch": 5.149278081409577,
      "grad_norm": 0.0020181762520223856,
      "learning_rate": 0.03173716463702209,
      "loss": 0.1744,
      "num_input_tokens_seen": 60744432,
      "step": 31565
    },
    {
      "epoch": 5.150093808630394,
      "grad_norm": 0.0015634980518370867,
      "learning_rate": 0.03170093909569638,
      "loss": 0.1734,
      "num_input_tokens_seen": 60753536,
      "step": 31570
    },
    {
      "epoch": 5.150909535851211,
      "grad_norm": 0.002657000906765461,
      "learning_rate": 0.03166473179757246,
      "loss": 0.1866,
      "num_input_tokens_seen": 60764112,
      "step": 31575
    },
    {
      "epoch": 5.151725263072029,
      "grad_norm": 0.003478836501017213,
      "learning_rate": 0.031628542748234005,
      "loss": 0.1796,
      "num_input_tokens_seen": 60772080,
      "step": 31580
    },
    {
      "epoch": 5.152540990292846,
      "grad_norm": 0.0025467940140515566,
      "learning_rate": 0.03159237195326184,
      "loss": 0.1232,
      "num_input_tokens_seen": 60781712,
      "step": 31585
    },
    {
      "epoch": 5.153356717513663,
      "grad_norm": 0.002344178268685937,
      "learning_rate": 0.031556219418233875,
      "loss": 0.1981,
      "num_input_tokens_seen": 60791104,
      "step": 31590
    },
    {
      "epoch": 5.154172444734481,
      "grad_norm": 0.004874152131378651,
      "learning_rate": 0.03152008514872533,
      "loss": 0.1988,
      "num_input_tokens_seen": 60800720,
      "step": 31595
    },
    {
      "epoch": 5.154988171955298,
      "grad_norm": 0.00271607912145555,
      "learning_rate": 0.03148396915030862,
      "loss": 0.1638,
      "num_input_tokens_seen": 60809376,
      "step": 31600
    },
    {
      "epoch": 5.154988171955298,
      "eval_loss": 0.19146056473255157,
      "eval_runtime": 68.089,
      "eval_samples_per_second": 40.021,
      "eval_steps_per_second": 20.018,
      "num_input_tokens_seen": 60809376,
      "step": 31600
    },
    {
      "epoch": 5.1558038991761155,
      "grad_norm": 0.003949540201574564,
      "learning_rate": 0.03144787142855318,
      "loss": 0.1736,
      "num_input_tokens_seen": 60819392,
      "step": 31605
    },
    {
      "epoch": 5.156619626396933,
      "grad_norm": 0.003632277250289917,
      "learning_rate": 0.031411791989025835,
      "loss": 0.1502,
      "num_input_tokens_seen": 60827888,
      "step": 31610
    },
    {
      "epoch": 5.157435353617751,
      "grad_norm": 0.0031025283969938755,
      "learning_rate": 0.031375730837290394,
      "loss": 0.2081,
      "num_input_tokens_seen": 60836384,
      "step": 31615
    },
    {
      "epoch": 5.158251080838568,
      "grad_norm": 0.00389301311224699,
      "learning_rate": 0.031339687978908015,
      "loss": 0.1713,
      "num_input_tokens_seen": 60845296,
      "step": 31620
    },
    {
      "epoch": 5.159066808059385,
      "grad_norm": 0.0033850171603262424,
      "learning_rate": 0.03130366341943694,
      "loss": 0.2143,
      "num_input_tokens_seen": 60854768,
      "step": 31625
    },
    {
      "epoch": 5.159882535280202,
      "grad_norm": 0.002019742736592889,
      "learning_rate": 0.031267657164432555,
      "loss": 0.1779,
      "num_input_tokens_seen": 60864208,
      "step": 31630
    },
    {
      "epoch": 5.16069826250102,
      "grad_norm": 0.0037279368843883276,
      "learning_rate": 0.03123166921944752,
      "loss": 0.2483,
      "num_input_tokens_seen": 60874400,
      "step": 31635
    },
    {
      "epoch": 5.161513989721837,
      "grad_norm": 0.00417878944426775,
      "learning_rate": 0.031195699590031666,
      "loss": 0.1677,
      "num_input_tokens_seen": 60884176,
      "step": 31640
    },
    {
      "epoch": 5.162329716942654,
      "grad_norm": 0.0019403138430789113,
      "learning_rate": 0.031159748281731885,
      "loss": 0.2067,
      "num_input_tokens_seen": 60894448,
      "step": 31645
    },
    {
      "epoch": 5.163145444163471,
      "grad_norm": 0.004151047673076391,
      "learning_rate": 0.031123815300092394,
      "loss": 0.1743,
      "num_input_tokens_seen": 60904816,
      "step": 31650
    },
    {
      "epoch": 5.1639611713842895,
      "grad_norm": 0.0037562635261565447,
      "learning_rate": 0.031087900650654424,
      "loss": 0.2134,
      "num_input_tokens_seen": 60914416,
      "step": 31655
    },
    {
      "epoch": 5.164776898605107,
      "grad_norm": 0.0025621443055570126,
      "learning_rate": 0.031052004338956534,
      "loss": 0.1377,
      "num_input_tokens_seen": 60922800,
      "step": 31660
    },
    {
      "epoch": 5.165592625825924,
      "grad_norm": 0.0044174352660775185,
      "learning_rate": 0.031016126370534407,
      "loss": 0.2319,
      "num_input_tokens_seen": 60932432,
      "step": 31665
    },
    {
      "epoch": 5.166408353046741,
      "grad_norm": 0.0016621018294245005,
      "learning_rate": 0.030980266750920804,
      "loss": 0.1794,
      "num_input_tokens_seen": 60942816,
      "step": 31670
    },
    {
      "epoch": 5.167224080267559,
      "grad_norm": 0.002501888433471322,
      "learning_rate": 0.030944425485645747,
      "loss": 0.2021,
      "num_input_tokens_seen": 60951264,
      "step": 31675
    },
    {
      "epoch": 5.168039807488376,
      "grad_norm": 0.004333432298153639,
      "learning_rate": 0.03090860258023647,
      "loss": 0.2407,
      "num_input_tokens_seen": 60959440,
      "step": 31680
    },
    {
      "epoch": 5.168855534709193,
      "grad_norm": 0.005217937286943197,
      "learning_rate": 0.030872798040217236,
      "loss": 0.1678,
      "num_input_tokens_seen": 60969552,
      "step": 31685
    },
    {
      "epoch": 5.16967126193001,
      "grad_norm": 0.0031162756495177746,
      "learning_rate": 0.03083701187110964,
      "loss": 0.1642,
      "num_input_tokens_seen": 60980432,
      "step": 31690
    },
    {
      "epoch": 5.170486989150828,
      "grad_norm": 0.0033705614041537046,
      "learning_rate": 0.030801244078432294,
      "loss": 0.161,
      "num_input_tokens_seen": 60988704,
      "step": 31695
    },
    {
      "epoch": 5.171302716371645,
      "grad_norm": 0.0027948126662522554,
      "learning_rate": 0.030765494667701024,
      "loss": 0.1587,
      "num_input_tokens_seen": 60997536,
      "step": 31700
    },
    {
      "epoch": 5.1721184435924625,
      "grad_norm": 0.0029083590488880873,
      "learning_rate": 0.030729763644428913,
      "loss": 0.2058,
      "num_input_tokens_seen": 61007856,
      "step": 31705
    },
    {
      "epoch": 5.17293417081328,
      "grad_norm": 0.0032397215254604816,
      "learning_rate": 0.030694051014126048,
      "loss": 0.1561,
      "num_input_tokens_seen": 61017600,
      "step": 31710
    },
    {
      "epoch": 5.173749898034098,
      "grad_norm": 0.0034669949673116207,
      "learning_rate": 0.030658356782299792,
      "loss": 0.217,
      "num_input_tokens_seen": 61027568,
      "step": 31715
    },
    {
      "epoch": 5.174565625254915,
      "grad_norm": 0.005678155459463596,
      "learning_rate": 0.030622680954454726,
      "loss": 0.2484,
      "num_input_tokens_seen": 61036672,
      "step": 31720
    },
    {
      "epoch": 5.175381352475732,
      "grad_norm": 0.003891571192070842,
      "learning_rate": 0.030587023536092398,
      "loss": 0.2027,
      "num_input_tokens_seen": 61045232,
      "step": 31725
    },
    {
      "epoch": 5.176197079696549,
      "grad_norm": 0.002179044531658292,
      "learning_rate": 0.03055138453271171,
      "loss": 0.138,
      "num_input_tokens_seen": 61055024,
      "step": 31730
    },
    {
      "epoch": 5.177012806917367,
      "grad_norm": 0.004467246122658253,
      "learning_rate": 0.03051576394980858,
      "loss": 0.2201,
      "num_input_tokens_seen": 61064624,
      "step": 31735
    },
    {
      "epoch": 5.177828534138184,
      "grad_norm": 0.001569413929246366,
      "learning_rate": 0.030480161792876187,
      "loss": 0.1705,
      "num_input_tokens_seen": 61074624,
      "step": 31740
    },
    {
      "epoch": 5.178644261359001,
      "grad_norm": 0.0030016470700502396,
      "learning_rate": 0.030444578067404846,
      "loss": 0.264,
      "num_input_tokens_seen": 61082928,
      "step": 31745
    },
    {
      "epoch": 5.1794599885798185,
      "grad_norm": 0.0023360855411738157,
      "learning_rate": 0.030409012778881975,
      "loss": 0.2208,
      "num_input_tokens_seen": 61091632,
      "step": 31750
    },
    {
      "epoch": 5.1802757158006365,
      "grad_norm": 0.0027255790773779154,
      "learning_rate": 0.030373465932792235,
      "loss": 0.1535,
      "num_input_tokens_seen": 61099520,
      "step": 31755
    },
    {
      "epoch": 5.181091443021454,
      "grad_norm": 0.0029293824918568134,
      "learning_rate": 0.030337937534617342,
      "loss": 0.2191,
      "num_input_tokens_seen": 61108784,
      "step": 31760
    },
    {
      "epoch": 5.181907170242271,
      "grad_norm": 0.0028456130530685186,
      "learning_rate": 0.030302427589836277,
      "loss": 0.1622,
      "num_input_tokens_seen": 61117712,
      "step": 31765
    },
    {
      "epoch": 5.182722897463089,
      "grad_norm": 0.0018258902709931135,
      "learning_rate": 0.030266936103925095,
      "loss": 0.2141,
      "num_input_tokens_seen": 61126384,
      "step": 31770
    },
    {
      "epoch": 5.183538624683906,
      "grad_norm": 0.002224850235506892,
      "learning_rate": 0.030231463082356982,
      "loss": 0.1559,
      "num_input_tokens_seen": 61135744,
      "step": 31775
    },
    {
      "epoch": 5.184354351904723,
      "grad_norm": 0.0020366120152175426,
      "learning_rate": 0.030196008530602367,
      "loss": 0.1657,
      "num_input_tokens_seen": 61144864,
      "step": 31780
    },
    {
      "epoch": 5.18517007912554,
      "grad_norm": 0.003660195041447878,
      "learning_rate": 0.030160572454128842,
      "loss": 0.1716,
      "num_input_tokens_seen": 61154688,
      "step": 31785
    },
    {
      "epoch": 5.185985806346358,
      "grad_norm": 0.004407871048897505,
      "learning_rate": 0.03012515485840098,
      "loss": 0.1706,
      "num_input_tokens_seen": 61165184,
      "step": 31790
    },
    {
      "epoch": 5.186801533567175,
      "grad_norm": 0.0019094664603471756,
      "learning_rate": 0.030089755748880734,
      "loss": 0.1618,
      "num_input_tokens_seen": 61176512,
      "step": 31795
    },
    {
      "epoch": 5.1876172607879925,
      "grad_norm": 0.002183450385928154,
      "learning_rate": 0.030054375131027003,
      "loss": 0.1582,
      "num_input_tokens_seen": 61186608,
      "step": 31800
    },
    {
      "epoch": 5.1876172607879925,
      "eval_loss": 0.1885540932416916,
      "eval_runtime": 68.1245,
      "eval_samples_per_second": 40.0,
      "eval_steps_per_second": 20.007,
      "num_input_tokens_seen": 61186608,
      "step": 31800
    },
    {
      "epoch": 5.18843298800881,
      "grad_norm": 0.0030873653013259172,
      "learning_rate": 0.030019013010295942,
      "loss": 0.2051,
      "num_input_tokens_seen": 61196400,
      "step": 31805
    },
    {
      "epoch": 5.189248715229628,
      "grad_norm": 0.0022992677986621857,
      "learning_rate": 0.029983669392140897,
      "loss": 0.1421,
      "num_input_tokens_seen": 61205504,
      "step": 31810
    },
    {
      "epoch": 5.190064442450445,
      "grad_norm": 0.0028770288918167353,
      "learning_rate": 0.029948344282012217,
      "loss": 0.2265,
      "num_input_tokens_seen": 61214400,
      "step": 31815
    },
    {
      "epoch": 5.190880169671262,
      "grad_norm": 0.0022579007782042027,
      "learning_rate": 0.029913037685357507,
      "loss": 0.1517,
      "num_input_tokens_seen": 61223904,
      "step": 31820
    },
    {
      "epoch": 5.191695896892079,
      "grad_norm": 0.0029233593959361315,
      "learning_rate": 0.029877749607621528,
      "loss": 0.1827,
      "num_input_tokens_seen": 61232928,
      "step": 31825
    },
    {
      "epoch": 5.192511624112897,
      "grad_norm": 0.0048292758874595165,
      "learning_rate": 0.029842480054246077,
      "loss": 0.0957,
      "num_input_tokens_seen": 61242816,
      "step": 31830
    },
    {
      "epoch": 5.193327351333714,
      "grad_norm": 0.00552166486158967,
      "learning_rate": 0.02980722903067022,
      "loss": 0.2142,
      "num_input_tokens_seen": 61252512,
      "step": 31835
    },
    {
      "epoch": 5.194143078554531,
      "grad_norm": 0.00663384236395359,
      "learning_rate": 0.029771996542330113,
      "loss": 0.2453,
      "num_input_tokens_seen": 61262016,
      "step": 31840
    },
    {
      "epoch": 5.194958805775348,
      "grad_norm": 0.0042498367838561535,
      "learning_rate": 0.029736782594658954,
      "loss": 0.1649,
      "num_input_tokens_seen": 61271904,
      "step": 31845
    },
    {
      "epoch": 5.195774532996166,
      "grad_norm": 0.0020649349316954613,
      "learning_rate": 0.029701587193087284,
      "loss": 0.206,
      "num_input_tokens_seen": 61279952,
      "step": 31850
    },
    {
      "epoch": 5.196590260216984,
      "grad_norm": 0.002897506346926093,
      "learning_rate": 0.0296664103430426,
      "loss": 0.1405,
      "num_input_tokens_seen": 61290272,
      "step": 31855
    },
    {
      "epoch": 5.197405987437801,
      "grad_norm": 0.003279141616076231,
      "learning_rate": 0.029631252049949652,
      "loss": 0.2429,
      "num_input_tokens_seen": 61299264,
      "step": 31860
    },
    {
      "epoch": 5.198221714658618,
      "grad_norm": 0.00447164848446846,
      "learning_rate": 0.02959611231923031,
      "loss": 0.2097,
      "num_input_tokens_seen": 61308032,
      "step": 31865
    },
    {
      "epoch": 5.199037441879436,
      "grad_norm": 0.0027378024533391,
      "learning_rate": 0.029560991156303507,
      "loss": 0.204,
      "num_input_tokens_seen": 61318000,
      "step": 31870
    },
    {
      "epoch": 5.199853169100253,
      "grad_norm": 0.0032827265094965696,
      "learning_rate": 0.02952588856658544,
      "loss": 0.2311,
      "num_input_tokens_seen": 61327072,
      "step": 31875
    },
    {
      "epoch": 5.20066889632107,
      "grad_norm": 0.002890777075663209,
      "learning_rate": 0.029490804555489296,
      "loss": 0.1829,
      "num_input_tokens_seen": 61337456,
      "step": 31880
    },
    {
      "epoch": 5.201484623541887,
      "grad_norm": 0.0014737073797732592,
      "learning_rate": 0.029455739128425484,
      "loss": 0.1533,
      "num_input_tokens_seen": 61345952,
      "step": 31885
    },
    {
      "epoch": 5.202300350762705,
      "grad_norm": 0.002475172281265259,
      "learning_rate": 0.029420692290801607,
      "loss": 0.2074,
      "num_input_tokens_seen": 61355584,
      "step": 31890
    },
    {
      "epoch": 5.203116077983522,
      "grad_norm": 0.0023692341055721045,
      "learning_rate": 0.02938566404802223,
      "loss": 0.1483,
      "num_input_tokens_seen": 61364512,
      "step": 31895
    },
    {
      "epoch": 5.2039318052043395,
      "grad_norm": 0.006825537420809269,
      "learning_rate": 0.029350654405489195,
      "loss": 0.2464,
      "num_input_tokens_seen": 61373072,
      "step": 31900
    },
    {
      "epoch": 5.204747532425157,
      "grad_norm": 0.003331677056849003,
      "learning_rate": 0.02931566336860145,
      "loss": 0.1622,
      "num_input_tokens_seen": 61382768,
      "step": 31905
    },
    {
      "epoch": 5.205563259645975,
      "grad_norm": 0.003117034677416086,
      "learning_rate": 0.02928069094275505,
      "loss": 0.2107,
      "num_input_tokens_seen": 61393824,
      "step": 31910
    },
    {
      "epoch": 5.206378986866792,
      "grad_norm": 0.0027659276966005564,
      "learning_rate": 0.02924573713334314,
      "loss": 0.2038,
      "num_input_tokens_seen": 61403952,
      "step": 31915
    },
    {
      "epoch": 5.207194714087609,
      "grad_norm": 0.0022500373888760805,
      "learning_rate": 0.02921080194575603,
      "loss": 0.2104,
      "num_input_tokens_seen": 61412896,
      "step": 31920
    },
    {
      "epoch": 5.208010441308426,
      "grad_norm": 0.005972761195152998,
      "learning_rate": 0.029175885385381177,
      "loss": 0.231,
      "num_input_tokens_seen": 61422160,
      "step": 31925
    },
    {
      "epoch": 5.208826168529244,
      "grad_norm": 0.003115651197731495,
      "learning_rate": 0.029140987457603223,
      "loss": 0.1699,
      "num_input_tokens_seen": 61433728,
      "step": 31930
    },
    {
      "epoch": 5.209641895750061,
      "grad_norm": 0.003151980461552739,
      "learning_rate": 0.029106108167803763,
      "loss": 0.1844,
      "num_input_tokens_seen": 61443136,
      "step": 31935
    },
    {
      "epoch": 5.210457622970878,
      "grad_norm": 0.0028092646971344948,
      "learning_rate": 0.029071247521361674,
      "loss": 0.1896,
      "num_input_tokens_seen": 61453088,
      "step": 31940
    },
    {
      "epoch": 5.211273350191696,
      "grad_norm": 0.0025266380980610847,
      "learning_rate": 0.029036405523652945,
      "loss": 0.1791,
      "num_input_tokens_seen": 61462992,
      "step": 31945
    },
    {
      "epoch": 5.2120890774125135,
      "grad_norm": 0.00810024794191122,
      "learning_rate": 0.029001582180050577,
      "loss": 0.2115,
      "num_input_tokens_seen": 61472608,
      "step": 31950
    },
    {
      "epoch": 5.212904804633331,
      "grad_norm": 0.0036303692031651735,
      "learning_rate": 0.02896677749592482,
      "loss": 0.1743,
      "num_input_tokens_seen": 61482896,
      "step": 31955
    },
    {
      "epoch": 5.213720531854148,
      "grad_norm": 0.0029942465480417013,
      "learning_rate": 0.028931991476642938,
      "loss": 0.1745,
      "num_input_tokens_seen": 61491088,
      "step": 31960
    },
    {
      "epoch": 5.214536259074965,
      "grad_norm": 0.0024012920912355185,
      "learning_rate": 0.028897224127569412,
      "loss": 0.1979,
      "num_input_tokens_seen": 61500880,
      "step": 31965
    },
    {
      "epoch": 5.215351986295783,
      "grad_norm": 0.00196919497102499,
      "learning_rate": 0.028862475454065832,
      "loss": 0.1461,
      "num_input_tokens_seen": 61510848,
      "step": 31970
    },
    {
      "epoch": 5.2161677135166,
      "grad_norm": 0.0025292409118264914,
      "learning_rate": 0.028827745461490806,
      "loss": 0.2512,
      "num_input_tokens_seen": 61520192,
      "step": 31975
    },
    {
      "epoch": 5.216983440737417,
      "grad_norm": 0.002989513333886862,
      "learning_rate": 0.028793034155200212,
      "loss": 0.1952,
      "num_input_tokens_seen": 61530480,
      "step": 31980
    },
    {
      "epoch": 5.217799167958235,
      "grad_norm": 0.0025710253976285458,
      "learning_rate": 0.028758341540546944,
      "loss": 0.1607,
      "num_input_tokens_seen": 61538704,
      "step": 31985
    },
    {
      "epoch": 5.218614895179052,
      "grad_norm": 0.0019144422840327024,
      "learning_rate": 0.02872366762288098,
      "loss": 0.2116,
      "num_input_tokens_seen": 61547792,
      "step": 31990
    },
    {
      "epoch": 5.219430622399869,
      "grad_norm": 0.00174840132240206,
      "learning_rate": 0.028689012407549567,
      "loss": 0.1541,
      "num_input_tokens_seen": 61556976,
      "step": 31995
    },
    {
      "epoch": 5.220246349620687,
      "grad_norm": 0.0032675534021109343,
      "learning_rate": 0.028654375899896892,
      "loss": 0.1677,
      "num_input_tokens_seen": 61567504,
      "step": 32000
    },
    {
      "epoch": 5.220246349620687,
      "eval_loss": 0.19102944433689117,
      "eval_runtime": 67.9695,
      "eval_samples_per_second": 40.091,
      "eval_steps_per_second": 20.053,
      "num_input_tokens_seen": 61567504,
      "step": 32000
    },
    {
      "epoch": 5.221062076841505,
      "grad_norm": 0.004371013026684523,
      "learning_rate": 0.02861975810526437,
      "loss": 0.1592,
      "num_input_tokens_seen": 61576880,
      "step": 32005
    },
    {
      "epoch": 5.221877804062322,
      "grad_norm": 0.0038793745916336775,
      "learning_rate": 0.02858515902899056,
      "loss": 0.2049,
      "num_input_tokens_seen": 61587680,
      "step": 32010
    },
    {
      "epoch": 5.222693531283139,
      "grad_norm": 0.002956029726192355,
      "learning_rate": 0.028550578676410976,
      "loss": 0.1844,
      "num_input_tokens_seen": 61597088,
      "step": 32015
    },
    {
      "epoch": 5.223509258503956,
      "grad_norm": 0.0072611551731824875,
      "learning_rate": 0.02851601705285837,
      "loss": 0.2096,
      "num_input_tokens_seen": 61607760,
      "step": 32020
    },
    {
      "epoch": 5.224324985724774,
      "grad_norm": 0.002528066746890545,
      "learning_rate": 0.028481474163662666,
      "loss": 0.236,
      "num_input_tokens_seen": 61617840,
      "step": 32025
    },
    {
      "epoch": 5.225140712945591,
      "grad_norm": 0.002638375386595726,
      "learning_rate": 0.028446950014150683,
      "loss": 0.2036,
      "num_input_tokens_seen": 61627232,
      "step": 32030
    },
    {
      "epoch": 5.225956440166408,
      "grad_norm": 0.0026335574220865965,
      "learning_rate": 0.028412444609646596,
      "loss": 0.1453,
      "num_input_tokens_seen": 61636512,
      "step": 32035
    },
    {
      "epoch": 5.226772167387225,
      "grad_norm": 0.002676488133147359,
      "learning_rate": 0.028377957955471465,
      "loss": 0.2198,
      "num_input_tokens_seen": 61647504,
      "step": 32040
    },
    {
      "epoch": 5.227587894608043,
      "grad_norm": 0.0034553343430161476,
      "learning_rate": 0.0283434900569436,
      "loss": 0.2187,
      "num_input_tokens_seen": 61656528,
      "step": 32045
    },
    {
      "epoch": 5.2284036218288605,
      "grad_norm": 0.0017185886390507221,
      "learning_rate": 0.028309040919378456,
      "loss": 0.1547,
      "num_input_tokens_seen": 61666336,
      "step": 32050
    },
    {
      "epoch": 5.229219349049678,
      "grad_norm": 0.003896046429872513,
      "learning_rate": 0.02827461054808848,
      "loss": 0.1987,
      "num_input_tokens_seen": 61675888,
      "step": 32055
    },
    {
      "epoch": 5.230035076270495,
      "grad_norm": 0.0026318093296140432,
      "learning_rate": 0.028240198948383186,
      "loss": 0.1314,
      "num_input_tokens_seen": 61685920,
      "step": 32060
    },
    {
      "epoch": 5.230850803491313,
      "grad_norm": 0.003271885449066758,
      "learning_rate": 0.028205806125569402,
      "loss": 0.1624,
      "num_input_tokens_seen": 61696720,
      "step": 32065
    },
    {
      "epoch": 5.23166653071213,
      "grad_norm": 0.0015276093035936356,
      "learning_rate": 0.028171432084950834,
      "loss": 0.1647,
      "num_input_tokens_seen": 61707632,
      "step": 32070
    },
    {
      "epoch": 5.232482257932947,
      "grad_norm": 0.006124839652329683,
      "learning_rate": 0.028137076831828478,
      "loss": 0.1428,
      "num_input_tokens_seen": 61717408,
      "step": 32075
    },
    {
      "epoch": 5.233297985153764,
      "grad_norm": 0.0035109822638332844,
      "learning_rate": 0.028102740371500238,
      "loss": 0.194,
      "num_input_tokens_seen": 61727232,
      "step": 32080
    },
    {
      "epoch": 5.234113712374582,
      "grad_norm": 0.003662301693111658,
      "learning_rate": 0.0280684227092613,
      "loss": 0.2693,
      "num_input_tokens_seen": 61737648,
      "step": 32085
    },
    {
      "epoch": 5.234929439595399,
      "grad_norm": 0.005806422792375088,
      "learning_rate": 0.02803412385040392,
      "loss": 0.2035,
      "num_input_tokens_seen": 61746640,
      "step": 32090
    },
    {
      "epoch": 5.2357451668162165,
      "grad_norm": 0.003867978462949395,
      "learning_rate": 0.027999843800217306,
      "loss": 0.2305,
      "num_input_tokens_seen": 61755536,
      "step": 32095
    },
    {
      "epoch": 5.236560894037034,
      "grad_norm": 0.0028302380815148354,
      "learning_rate": 0.027965582563987932,
      "loss": 0.2285,
      "num_input_tokens_seen": 61765712,
      "step": 32100
    },
    {
      "epoch": 5.237376621257852,
      "grad_norm": 0.0037267645820975304,
      "learning_rate": 0.027931340146999346,
      "loss": 0.1536,
      "num_input_tokens_seen": 61776784,
      "step": 32105
    },
    {
      "epoch": 5.238192348478669,
      "grad_norm": 0.0020260969176888466,
      "learning_rate": 0.02789711655453208,
      "loss": 0.1695,
      "num_input_tokens_seen": 61786336,
      "step": 32110
    },
    {
      "epoch": 5.239008075699486,
      "grad_norm": 0.0016337062697857618,
      "learning_rate": 0.02786291179186392,
      "loss": 0.1875,
      "num_input_tokens_seen": 61795328,
      "step": 32115
    },
    {
      "epoch": 5.239823802920303,
      "grad_norm": 0.0016405434580519795,
      "learning_rate": 0.02782872586426961,
      "loss": 0.1626,
      "num_input_tokens_seen": 61805600,
      "step": 32120
    },
    {
      "epoch": 5.240639530141121,
      "grad_norm": 0.002925229026004672,
      "learning_rate": 0.027794558777021083,
      "loss": 0.1967,
      "num_input_tokens_seen": 61815056,
      "step": 32125
    },
    {
      "epoch": 5.241455257361938,
      "grad_norm": 0.002907243324443698,
      "learning_rate": 0.02776041053538734,
      "loss": 0.1867,
      "num_input_tokens_seen": 61825024,
      "step": 32130
    },
    {
      "epoch": 5.242270984582755,
      "grad_norm": 0.003524775616824627,
      "learning_rate": 0.027726281144634407,
      "loss": 0.1998,
      "num_input_tokens_seen": 61834544,
      "step": 32135
    },
    {
      "epoch": 5.243086711803572,
      "grad_norm": 0.00300098885782063,
      "learning_rate": 0.02769217061002552,
      "loss": 0.1992,
      "num_input_tokens_seen": 61844816,
      "step": 32140
    },
    {
      "epoch": 5.2439024390243905,
      "grad_norm": 0.0026491160970181227,
      "learning_rate": 0.027658078936820967,
      "loss": 0.1708,
      "num_input_tokens_seen": 61854400,
      "step": 32145
    },
    {
      "epoch": 5.244718166245208,
      "grad_norm": 0.0016209479654207826,
      "learning_rate": 0.02762400613027805,
      "loss": 0.1857,
      "num_input_tokens_seen": 61863824,
      "step": 32150
    },
    {
      "epoch": 5.245533893466025,
      "grad_norm": 0.002623357344418764,
      "learning_rate": 0.027589952195651295,
      "loss": 0.2149,
      "num_input_tokens_seen": 61873152,
      "step": 32155
    },
    {
      "epoch": 5.246349620686843,
      "grad_norm": 0.004407132975757122,
      "learning_rate": 0.027555917138192186,
      "loss": 0.167,
      "num_input_tokens_seen": 61882480,
      "step": 32160
    },
    {
      "epoch": 5.24716534790766,
      "grad_norm": 0.0037562178913503885,
      "learning_rate": 0.027521900963149375,
      "loss": 0.1285,
      "num_input_tokens_seen": 61893024,
      "step": 32165
    },
    {
      "epoch": 5.247981075128477,
      "grad_norm": 0.0015302824322134256,
      "learning_rate": 0.027487903675768633,
      "loss": 0.1865,
      "num_input_tokens_seen": 61902208,
      "step": 32170
    },
    {
      "epoch": 5.248796802349294,
      "grad_norm": 0.0065763830207288265,
      "learning_rate": 0.027453925281292677,
      "loss": 0.1874,
      "num_input_tokens_seen": 61910800,
      "step": 32175
    },
    {
      "epoch": 5.249612529570112,
      "grad_norm": 0.004285306669771671,
      "learning_rate": 0.027419965784961475,
      "loss": 0.1934,
      "num_input_tokens_seen": 61919776,
      "step": 32180
    },
    {
      "epoch": 5.250428256790929,
      "grad_norm": 0.004082659725099802,
      "learning_rate": 0.027386025192012015,
      "loss": 0.2428,
      "num_input_tokens_seen": 61930528,
      "step": 32185
    },
    {
      "epoch": 5.251243984011746,
      "grad_norm": 0.005158967338502407,
      "learning_rate": 0.027352103507678277,
      "loss": 0.204,
      "num_input_tokens_seen": 61940416,
      "step": 32190
    },
    {
      "epoch": 5.2520597112325635,
      "grad_norm": 0.0028112996369600296,
      "learning_rate": 0.027318200737191527,
      "loss": 0.1918,
      "num_input_tokens_seen": 61949904,
      "step": 32195
    },
    {
      "epoch": 5.252875438453382,
      "grad_norm": 0.002726153936237097,
      "learning_rate": 0.027284316885779935,
      "loss": 0.1638,
      "num_input_tokens_seen": 61958976,
      "step": 32200
    },
    {
      "epoch": 5.252875438453382,
      "eval_loss": 0.19081522524356842,
      "eval_runtime": 67.959,
      "eval_samples_per_second": 40.098,
      "eval_steps_per_second": 20.056,
      "num_input_tokens_seen": 61958976,
      "step": 32200
    },
    {
      "epoch": 5.253691165674199,
      "grad_norm": 0.002692854031920433,
      "learning_rate": 0.027250451958668785,
      "loss": 0.1877,
      "num_input_tokens_seen": 61967648,
      "step": 32205
    },
    {
      "epoch": 5.254506892895016,
      "grad_norm": 0.0025756321847438812,
      "learning_rate": 0.027216605961080536,
      "loss": 0.1985,
      "num_input_tokens_seen": 61975904,
      "step": 32210
    },
    {
      "epoch": 5.255322620115833,
      "grad_norm": 0.003353280480951071,
      "learning_rate": 0.02718277889823461,
      "loss": 0.2175,
      "num_input_tokens_seen": 61985200,
      "step": 32215
    },
    {
      "epoch": 5.256138347336651,
      "grad_norm": 0.004737220238894224,
      "learning_rate": 0.027148970775347604,
      "loss": 0.1921,
      "num_input_tokens_seen": 61995184,
      "step": 32220
    },
    {
      "epoch": 5.256954074557468,
      "grad_norm": 0.0024227933026850224,
      "learning_rate": 0.027115181597633174,
      "loss": 0.175,
      "num_input_tokens_seen": 62005168,
      "step": 32225
    },
    {
      "epoch": 5.257769801778285,
      "grad_norm": 0.002520535374060273,
      "learning_rate": 0.027081411370301976,
      "loss": 0.1763,
      "num_input_tokens_seen": 62015536,
      "step": 32230
    },
    {
      "epoch": 5.258585528999102,
      "grad_norm": 0.004048903938382864,
      "learning_rate": 0.027047660098561875,
      "loss": 0.1696,
      "num_input_tokens_seen": 62025520,
      "step": 32235
    },
    {
      "epoch": 5.25940125621992,
      "grad_norm": 0.0038864007219672203,
      "learning_rate": 0.02701392778761766,
      "loss": 0.1997,
      "num_input_tokens_seen": 62035488,
      "step": 32240
    },
    {
      "epoch": 5.2602169834407375,
      "grad_norm": 0.0027139047160744667,
      "learning_rate": 0.02698021444267133,
      "loss": 0.1614,
      "num_input_tokens_seen": 62046032,
      "step": 32245
    },
    {
      "epoch": 5.261032710661555,
      "grad_norm": 0.001327815349213779,
      "learning_rate": 0.026946520068921915,
      "loss": 0.157,
      "num_input_tokens_seen": 62054960,
      "step": 32250
    },
    {
      "epoch": 5.261848437882372,
      "grad_norm": 0.002525953808799386,
      "learning_rate": 0.02691284467156547,
      "loss": 0.1649,
      "num_input_tokens_seen": 62063744,
      "step": 32255
    },
    {
      "epoch": 5.26266416510319,
      "grad_norm": 0.006173146888613701,
      "learning_rate": 0.026879188255795182,
      "loss": 0.2868,
      "num_input_tokens_seen": 62075056,
      "step": 32260
    },
    {
      "epoch": 5.263479892324007,
      "grad_norm": 0.0033807281870394945,
      "learning_rate": 0.026845550826801328,
      "loss": 0.1898,
      "num_input_tokens_seen": 62084224,
      "step": 32265
    },
    {
      "epoch": 5.264295619544824,
      "grad_norm": 0.0030671306885778904,
      "learning_rate": 0.02681193238977121,
      "loss": 0.1634,
      "num_input_tokens_seen": 62094032,
      "step": 32270
    },
    {
      "epoch": 5.265111346765641,
      "grad_norm": 0.0027695680037140846,
      "learning_rate": 0.026778332949889145,
      "loss": 0.1666,
      "num_input_tokens_seen": 62103312,
      "step": 32275
    },
    {
      "epoch": 5.265927073986459,
      "grad_norm": 0.002103614853695035,
      "learning_rate": 0.026744752512336673,
      "loss": 0.1834,
      "num_input_tokens_seen": 62113152,
      "step": 32280
    },
    {
      "epoch": 5.266742801207276,
      "grad_norm": 0.0018301613163203,
      "learning_rate": 0.02671119108229225,
      "loss": 0.1409,
      "num_input_tokens_seen": 62122496,
      "step": 32285
    },
    {
      "epoch": 5.2675585284280935,
      "grad_norm": 0.004134222399443388,
      "learning_rate": 0.026677648664931556,
      "loss": 0.1921,
      "num_input_tokens_seen": 62132880,
      "step": 32290
    },
    {
      "epoch": 5.268374255648911,
      "grad_norm": 0.0034890836104750633,
      "learning_rate": 0.026644125265427154,
      "loss": 0.2076,
      "num_input_tokens_seen": 62143488,
      "step": 32295
    },
    {
      "epoch": 5.269189982869729,
      "grad_norm": 0.0034381679724901915,
      "learning_rate": 0.026610620888948822,
      "loss": 0.2055,
      "num_input_tokens_seen": 62151056,
      "step": 32300
    },
    {
      "epoch": 5.270005710090546,
      "grad_norm": 0.0028287472669035196,
      "learning_rate": 0.026577135540663408,
      "loss": 0.1569,
      "num_input_tokens_seen": 62161104,
      "step": 32305
    },
    {
      "epoch": 5.270821437311363,
      "grad_norm": 0.0039465404115617275,
      "learning_rate": 0.026543669225734673,
      "loss": 0.1638,
      "num_input_tokens_seen": 62171408,
      "step": 32310
    },
    {
      "epoch": 5.27163716453218,
      "grad_norm": 0.0039285412058234215,
      "learning_rate": 0.02651022194932363,
      "loss": 0.1926,
      "num_input_tokens_seen": 62180880,
      "step": 32315
    },
    {
      "epoch": 5.272452891752998,
      "grad_norm": 0.001905878190882504,
      "learning_rate": 0.026476793716588194,
      "loss": 0.1653,
      "num_input_tokens_seen": 62190240,
      "step": 32320
    },
    {
      "epoch": 5.273268618973815,
      "grad_norm": 0.0033746894914656878,
      "learning_rate": 0.026443384532683467,
      "loss": 0.175,
      "num_input_tokens_seen": 62200704,
      "step": 32325
    },
    {
      "epoch": 5.274084346194632,
      "grad_norm": 0.004796288441866636,
      "learning_rate": 0.026409994402761584,
      "loss": 0.1688,
      "num_input_tokens_seen": 62210320,
      "step": 32330
    },
    {
      "epoch": 5.27490007341545,
      "grad_norm": 0.001975403167307377,
      "learning_rate": 0.026376623331971653,
      "loss": 0.1912,
      "num_input_tokens_seen": 62219936,
      "step": 32335
    },
    {
      "epoch": 5.275715800636267,
      "grad_norm": 0.0030263918451964855,
      "learning_rate": 0.026343271325459997,
      "loss": 0.2329,
      "num_input_tokens_seen": 62230432,
      "step": 32340
    },
    {
      "epoch": 5.276531527857085,
      "grad_norm": 0.0021971508394926786,
      "learning_rate": 0.02630993838836987,
      "loss": 0.2307,
      "num_input_tokens_seen": 62239760,
      "step": 32345
    },
    {
      "epoch": 5.277347255077902,
      "grad_norm": 0.003006287384778261,
      "learning_rate": 0.026276624525841584,
      "loss": 0.1826,
      "num_input_tokens_seen": 62248064,
      "step": 32350
    },
    {
      "epoch": 5.27816298229872,
      "grad_norm": 0.0019975050818175077,
      "learning_rate": 0.026243329743012637,
      "loss": 0.1612,
      "num_input_tokens_seen": 62258896,
      "step": 32355
    },
    {
      "epoch": 5.278978709519537,
      "grad_norm": 0.0034892114344984293,
      "learning_rate": 0.026210054045017438,
      "loss": 0.1429,
      "num_input_tokens_seen": 62269856,
      "step": 32360
    },
    {
      "epoch": 5.279794436740354,
      "grad_norm": 0.005603166297078133,
      "learning_rate": 0.02617679743698755,
      "loss": 0.192,
      "num_input_tokens_seen": 62279824,
      "step": 32365
    },
    {
      "epoch": 5.280610163961171,
      "grad_norm": 0.0027568836230784655,
      "learning_rate": 0.02614355992405158,
      "loss": 0.1864,
      "num_input_tokens_seen": 62288896,
      "step": 32370
    },
    {
      "epoch": 5.281425891181989,
      "grad_norm": 0.0018631363054737449,
      "learning_rate": 0.026110341511335115,
      "loss": 0.1683,
      "num_input_tokens_seen": 62297312,
      "step": 32375
    },
    {
      "epoch": 5.282241618402806,
      "grad_norm": 0.0022636358626186848,
      "learning_rate": 0.02607714220396093,
      "loss": 0.2407,
      "num_input_tokens_seen": 62307136,
      "step": 32380
    },
    {
      "epoch": 5.283057345623623,
      "grad_norm": 0.0031148679554462433,
      "learning_rate": 0.02604396200704869,
      "loss": 0.1526,
      "num_input_tokens_seen": 62316320,
      "step": 32385
    },
    {
      "epoch": 5.2838730728444405,
      "grad_norm": 0.003759776707738638,
      "learning_rate": 0.02601080092571523,
      "loss": 0.2162,
      "num_input_tokens_seen": 62326288,
      "step": 32390
    },
    {
      "epoch": 5.2846888000652585,
      "grad_norm": 0.0017863363027572632,
      "learning_rate": 0.025977658965074455,
      "loss": 0.1536,
      "num_input_tokens_seen": 62335936,
      "step": 32395
    },
    {
      "epoch": 5.285504527286076,
      "grad_norm": 0.0019069620175287127,
      "learning_rate": 0.02594453613023719,
      "loss": 0.1804,
      "num_input_tokens_seen": 62346176,
      "step": 32400
    },
    {
      "epoch": 5.285504527286076,
      "eval_loss": 0.18935717642307281,
      "eval_runtime": 68.085,
      "eval_samples_per_second": 40.024,
      "eval_steps_per_second": 20.019,
      "num_input_tokens_seen": 62346176,
      "step": 32400
    },
    {
      "epoch": 5.286320254506893,
      "grad_norm": 0.0033762820530682802,
      "learning_rate": 0.025911432426311443,
      "loss": 0.183,
      "num_input_tokens_seen": 62356384,
      "step": 32405
    },
    {
      "epoch": 5.28713598172771,
      "grad_norm": 0.004138768184930086,
      "learning_rate": 0.025878347858402234,
      "loss": 0.2395,
      "num_input_tokens_seen": 62366400,
      "step": 32410
    },
    {
      "epoch": 5.287951708948528,
      "grad_norm": 0.004576593171805143,
      "learning_rate": 0.025845282431611598,
      "loss": 0.1891,
      "num_input_tokens_seen": 62375216,
      "step": 32415
    },
    {
      "epoch": 5.288767436169345,
      "grad_norm": 0.002692391164600849,
      "learning_rate": 0.025812236151038608,
      "loss": 0.1906,
      "num_input_tokens_seen": 62384240,
      "step": 32420
    },
    {
      "epoch": 5.289583163390162,
      "grad_norm": 0.0023873988538980484,
      "learning_rate": 0.025779209021779468,
      "loss": 0.1833,
      "num_input_tokens_seen": 62393520,
      "step": 32425
    },
    {
      "epoch": 5.290398890610979,
      "grad_norm": 0.0034172353334724903,
      "learning_rate": 0.025746201048927324,
      "loss": 0.1857,
      "num_input_tokens_seen": 62403296,
      "step": 32430
    },
    {
      "epoch": 5.291214617831797,
      "grad_norm": 0.0017080488614737988,
      "learning_rate": 0.025713212237572485,
      "loss": 0.1499,
      "num_input_tokens_seen": 62413040,
      "step": 32435
    },
    {
      "epoch": 5.2920303450526145,
      "grad_norm": 0.0028481886256486177,
      "learning_rate": 0.025680242592802164,
      "loss": 0.1743,
      "num_input_tokens_seen": 62423056,
      "step": 32440
    },
    {
      "epoch": 5.292846072273432,
      "grad_norm": 0.0015285526169463992,
      "learning_rate": 0.02564729211970073,
      "loss": 0.1722,
      "num_input_tokens_seen": 62433344,
      "step": 32445
    },
    {
      "epoch": 5.293661799494249,
      "grad_norm": 0.0032806452363729477,
      "learning_rate": 0.025614360823349617,
      "loss": 0.1515,
      "num_input_tokens_seen": 62442224,
      "step": 32450
    },
    {
      "epoch": 5.294477526715067,
      "grad_norm": 0.0026376848109066486,
      "learning_rate": 0.025581448708827146,
      "loss": 0.1542,
      "num_input_tokens_seen": 62452880,
      "step": 32455
    },
    {
      "epoch": 5.295293253935884,
      "grad_norm": 0.003547085216268897,
      "learning_rate": 0.025548555781208876,
      "loss": 0.1868,
      "num_input_tokens_seen": 62461824,
      "step": 32460
    },
    {
      "epoch": 5.296108981156701,
      "grad_norm": 0.0038243234157562256,
      "learning_rate": 0.02551568204556721,
      "loss": 0.1972,
      "num_input_tokens_seen": 62471136,
      "step": 32465
    },
    {
      "epoch": 5.296924708377518,
      "grad_norm": 0.0017619819846004248,
      "learning_rate": 0.02548282750697173,
      "loss": 0.1344,
      "num_input_tokens_seen": 62480480,
      "step": 32470
    },
    {
      "epoch": 5.297740435598336,
      "grad_norm": 0.00193989765830338,
      "learning_rate": 0.02544999217048909,
      "loss": 0.1806,
      "num_input_tokens_seen": 62489504,
      "step": 32475
    },
    {
      "epoch": 5.298556162819153,
      "grad_norm": 0.0018886609468609095,
      "learning_rate": 0.025417176041182793,
      "loss": 0.1666,
      "num_input_tokens_seen": 62498032,
      "step": 32480
    },
    {
      "epoch": 5.2993718900399704,
      "grad_norm": 0.00463362829759717,
      "learning_rate": 0.025384379124113596,
      "loss": 0.1909,
      "num_input_tokens_seen": 62506448,
      "step": 32485
    },
    {
      "epoch": 5.300187617260788,
      "grad_norm": 0.001035036169923842,
      "learning_rate": 0.025351601424339124,
      "loss": 0.1481,
      "num_input_tokens_seen": 62516704,
      "step": 32490
    },
    {
      "epoch": 5.301003344481606,
      "grad_norm": 0.0015740799717605114,
      "learning_rate": 0.025318842946914184,
      "loss": 0.1294,
      "num_input_tokens_seen": 62526896,
      "step": 32495
    },
    {
      "epoch": 5.301819071702423,
      "grad_norm": 0.003563928883522749,
      "learning_rate": 0.025286103696890494,
      "loss": 0.211,
      "num_input_tokens_seen": 62537984,
      "step": 32500
    },
    {
      "epoch": 5.30263479892324,
      "grad_norm": 0.003360944567248225,
      "learning_rate": 0.025253383679316836,
      "loss": 0.1549,
      "num_input_tokens_seen": 62548144,
      "step": 32505
    },
    {
      "epoch": 5.303450526144058,
      "grad_norm": 0.002556940307840705,
      "learning_rate": 0.025220682899239077,
      "loss": 0.2439,
      "num_input_tokens_seen": 62557216,
      "step": 32510
    },
    {
      "epoch": 5.304266253364875,
      "grad_norm": 0.0023190900683403015,
      "learning_rate": 0.02518800136170013,
      "loss": 0.1496,
      "num_input_tokens_seen": 62568016,
      "step": 32515
    },
    {
      "epoch": 5.305081980585692,
      "grad_norm": 0.005452712997794151,
      "learning_rate": 0.02515533907173981,
      "loss": 0.1772,
      "num_input_tokens_seen": 62576336,
      "step": 32520
    },
    {
      "epoch": 5.305897707806509,
      "grad_norm": 0.004547107499092817,
      "learning_rate": 0.025122696034395115,
      "loss": 0.1857,
      "num_input_tokens_seen": 62586768,
      "step": 32525
    },
    {
      "epoch": 5.306713435027326,
      "grad_norm": 0.0029369972180575132,
      "learning_rate": 0.025090072254700023,
      "loss": 0.2065,
      "num_input_tokens_seen": 62596448,
      "step": 32530
    },
    {
      "epoch": 5.307529162248144,
      "grad_norm": 0.003958845045417547,
      "learning_rate": 0.025057467737685468,
      "loss": 0.1573,
      "num_input_tokens_seen": 62606160,
      "step": 32535
    },
    {
      "epoch": 5.3083448894689615,
      "grad_norm": 0.004521418362855911,
      "learning_rate": 0.025024882488379557,
      "loss": 0.1717,
      "num_input_tokens_seen": 62615600,
      "step": 32540
    },
    {
      "epoch": 5.309160616689779,
      "grad_norm": 0.0026967038866132498,
      "learning_rate": 0.02499231651180727,
      "loss": 0.1523,
      "num_input_tokens_seen": 62624912,
      "step": 32545
    },
    {
      "epoch": 5.309976343910597,
      "grad_norm": 0.005338952410966158,
      "learning_rate": 0.024959769812990713,
      "loss": 0.2703,
      "num_input_tokens_seen": 62634400,
      "step": 32550
    },
    {
      "epoch": 5.310792071131414,
      "grad_norm": 0.0027153801638633013,
      "learning_rate": 0.024927242396949045,
      "loss": 0.1894,
      "num_input_tokens_seen": 62643904,
      "step": 32555
    },
    {
      "epoch": 5.311607798352231,
      "grad_norm": 0.002482909942045808,
      "learning_rate": 0.02489473426869836,
      "loss": 0.1226,
      "num_input_tokens_seen": 62653504,
      "step": 32560
    },
    {
      "epoch": 5.312423525573048,
      "grad_norm": 0.0034938212484121323,
      "learning_rate": 0.024862245433251776,
      "loss": 0.2399,
      "num_input_tokens_seen": 62663168,
      "step": 32565
    },
    {
      "epoch": 5.313239252793866,
      "grad_norm": 0.0037384608294814825,
      "learning_rate": 0.024829775895619577,
      "loss": 0.154,
      "num_input_tokens_seen": 62673104,
      "step": 32570
    },
    {
      "epoch": 5.314054980014683,
      "grad_norm": 0.0031447235960513353,
      "learning_rate": 0.024797325660808882,
      "loss": 0.1521,
      "num_input_tokens_seen": 62683728,
      "step": 32575
    },
    {
      "epoch": 5.3148707072355,
      "grad_norm": 0.0018160607432946563,
      "learning_rate": 0.02476489473382401,
      "loss": 0.1925,
      "num_input_tokens_seen": 62693648,
      "step": 32580
    },
    {
      "epoch": 5.3156864344563175,
      "grad_norm": 0.004437434021383524,
      "learning_rate": 0.024732483119666127,
      "loss": 0.1983,
      "num_input_tokens_seen": 62703536,
      "step": 32585
    },
    {
      "epoch": 5.3165021616771355,
      "grad_norm": 0.005430464632809162,
      "learning_rate": 0.024700090823333548,
      "loss": 0.2119,
      "num_input_tokens_seen": 62714048,
      "step": 32590
    },
    {
      "epoch": 5.317317888897953,
      "grad_norm": 0.0024631968699395657,
      "learning_rate": 0.02466771784982163,
      "loss": 0.169,
      "num_input_tokens_seen": 62723792,
      "step": 32595
    },
    {
      "epoch": 5.31813361611877,
      "grad_norm": 0.004829746671020985,
      "learning_rate": 0.024635364204122594,
      "loss": 0.2607,
      "num_input_tokens_seen": 62734064,
      "step": 32600
    },
    {
      "epoch": 5.31813361611877,
      "eval_loss": 0.19094933569431305,
      "eval_runtime": 67.9956,
      "eval_samples_per_second": 40.076,
      "eval_steps_per_second": 20.045,
      "num_input_tokens_seen": 62734064,
      "step": 32600
    },
    {
      "epoch": 5.318949343339587,
      "grad_norm": 0.0032814701553434134,
      "learning_rate": 0.024603029891225852,
      "loss": 0.2015,
      "num_input_tokens_seen": 62744320,
      "step": 32605
    },
    {
      "epoch": 5.319765070560405,
      "grad_norm": 0.0025640144012868404,
      "learning_rate": 0.024570714916117748,
      "loss": 0.1635,
      "num_input_tokens_seen": 62755072,
      "step": 32610
    },
    {
      "epoch": 5.320580797781222,
      "grad_norm": 0.003794938325881958,
      "learning_rate": 0.024538419283781625,
      "loss": 0.2175,
      "num_input_tokens_seen": 62765504,
      "step": 32615
    },
    {
      "epoch": 5.321396525002039,
      "grad_norm": 0.0024537427816540003,
      "learning_rate": 0.024506142999197938,
      "loss": 0.1798,
      "num_input_tokens_seen": 62775504,
      "step": 32620
    },
    {
      "epoch": 5.322212252222856,
      "grad_norm": 0.002155778231099248,
      "learning_rate": 0.024473886067344002,
      "loss": 0.2063,
      "num_input_tokens_seen": 62785984,
      "step": 32625
    },
    {
      "epoch": 5.323027979443674,
      "grad_norm": 0.002605858724564314,
      "learning_rate": 0.02444164849319434,
      "loss": 0.1178,
      "num_input_tokens_seen": 62794304,
      "step": 32630
    },
    {
      "epoch": 5.3238437066644915,
      "grad_norm": 0.0027902929577976465,
      "learning_rate": 0.024409430281720306,
      "loss": 0.19,
      "num_input_tokens_seen": 62804944,
      "step": 32635
    },
    {
      "epoch": 5.324659433885309,
      "grad_norm": 0.004191721323877573,
      "learning_rate": 0.024377231437890428,
      "loss": 0.1614,
      "num_input_tokens_seen": 62815312,
      "step": 32640
    },
    {
      "epoch": 5.325475161106126,
      "grad_norm": 0.004191541578620672,
      "learning_rate": 0.024345051966670115,
      "loss": 0.1591,
      "num_input_tokens_seen": 62825168,
      "step": 32645
    },
    {
      "epoch": 5.326290888326944,
      "grad_norm": 0.0027105561457574368,
      "learning_rate": 0.024312891873021884,
      "loss": 0.1567,
      "num_input_tokens_seen": 62834976,
      "step": 32650
    },
    {
      "epoch": 5.327106615547761,
      "grad_norm": 0.0037749859038740396,
      "learning_rate": 0.024280751161905183,
      "loss": 0.2113,
      "num_input_tokens_seen": 62844256,
      "step": 32655
    },
    {
      "epoch": 5.327922342768578,
      "grad_norm": 0.0021335631608963013,
      "learning_rate": 0.02424862983827658,
      "loss": 0.1404,
      "num_input_tokens_seen": 62853152,
      "step": 32660
    },
    {
      "epoch": 5.328738069989395,
      "grad_norm": 0.00321765523403883,
      "learning_rate": 0.024216527907089495,
      "loss": 0.1887,
      "num_input_tokens_seen": 62863248,
      "step": 32665
    },
    {
      "epoch": 5.329553797210213,
      "grad_norm": 0.0018387320451438427,
      "learning_rate": 0.024184445373294505,
      "loss": 0.1515,
      "num_input_tokens_seen": 62872144,
      "step": 32670
    },
    {
      "epoch": 5.33036952443103,
      "grad_norm": 0.005462000146508217,
      "learning_rate": 0.02415238224183918,
      "loss": 0.1766,
      "num_input_tokens_seen": 62881248,
      "step": 32675
    },
    {
      "epoch": 5.331185251651847,
      "grad_norm": 0.00417410209774971,
      "learning_rate": 0.024120338517667973,
      "loss": 0.1947,
      "num_input_tokens_seen": 62891616,
      "step": 32680
    },
    {
      "epoch": 5.332000978872665,
      "grad_norm": 0.005171515978872776,
      "learning_rate": 0.02408831420572247,
      "loss": 0.1478,
      "num_input_tokens_seen": 62901184,
      "step": 32685
    },
    {
      "epoch": 5.332816706093483,
      "grad_norm": 0.00353564927354455,
      "learning_rate": 0.024056309310941264,
      "loss": 0.2703,
      "num_input_tokens_seen": 62910080,
      "step": 32690
    },
    {
      "epoch": 5.3336324333143,
      "grad_norm": 0.003173370845615864,
      "learning_rate": 0.02402432383825982,
      "loss": 0.1742,
      "num_input_tokens_seen": 62921520,
      "step": 32695
    },
    {
      "epoch": 5.334448160535117,
      "grad_norm": 0.004927566275000572,
      "learning_rate": 0.023992357792610792,
      "loss": 0.2009,
      "num_input_tokens_seen": 62932480,
      "step": 32700
    },
    {
      "epoch": 5.335263887755934,
      "grad_norm": 0.0023665735498070717,
      "learning_rate": 0.0239604111789237,
      "loss": 0.1786,
      "num_input_tokens_seen": 62943168,
      "step": 32705
    },
    {
      "epoch": 5.336079614976752,
      "grad_norm": 0.002787963952869177,
      "learning_rate": 0.023928484002125095,
      "loss": 0.1703,
      "num_input_tokens_seen": 62952848,
      "step": 32710
    },
    {
      "epoch": 5.336895342197569,
      "grad_norm": 0.003102795220911503,
      "learning_rate": 0.023896576267138595,
      "loss": 0.2194,
      "num_input_tokens_seen": 62962656,
      "step": 32715
    },
    {
      "epoch": 5.337711069418386,
      "grad_norm": 0.004342509899288416,
      "learning_rate": 0.02386468797888471,
      "loss": 0.2138,
      "num_input_tokens_seen": 62971840,
      "step": 32720
    },
    {
      "epoch": 5.338526796639204,
      "grad_norm": 0.003243209095671773,
      "learning_rate": 0.023832819142281057,
      "loss": 0.1527,
      "num_input_tokens_seen": 62980816,
      "step": 32725
    },
    {
      "epoch": 5.339342523860021,
      "grad_norm": 0.002436543582007289,
      "learning_rate": 0.02380096976224225,
      "loss": 0.1842,
      "num_input_tokens_seen": 62990720,
      "step": 32730
    },
    {
      "epoch": 5.3401582510808385,
      "grad_norm": 0.0028786801267415285,
      "learning_rate": 0.023769139843679777,
      "loss": 0.1499,
      "num_input_tokens_seen": 62999152,
      "step": 32735
    },
    {
      "epoch": 5.340973978301656,
      "grad_norm": 0.004515082575380802,
      "learning_rate": 0.023737329391502287,
      "loss": 0.198,
      "num_input_tokens_seen": 63009040,
      "step": 32740
    },
    {
      "epoch": 5.341789705522474,
      "grad_norm": 0.002774840220808983,
      "learning_rate": 0.023705538410615293,
      "loss": 0.1502,
      "num_input_tokens_seen": 63018048,
      "step": 32745
    },
    {
      "epoch": 5.342605432743291,
      "grad_norm": 0.004736543633043766,
      "learning_rate": 0.023673766905921396,
      "loss": 0.2498,
      "num_input_tokens_seen": 63027216,
      "step": 32750
    },
    {
      "epoch": 5.343421159964108,
      "grad_norm": 0.0009387094178237021,
      "learning_rate": 0.0236420148823202,
      "loss": 0.1113,
      "num_input_tokens_seen": 63036912,
      "step": 32755
    },
    {
      "epoch": 5.344236887184925,
      "grad_norm": 0.0031470567919313908,
      "learning_rate": 0.02361028234470816,
      "loss": 0.2015,
      "num_input_tokens_seen": 63046928,
      "step": 32760
    },
    {
      "epoch": 5.345052614405743,
      "grad_norm": 0.0012906900374218822,
      "learning_rate": 0.023578569297978913,
      "loss": 0.1366,
      "num_input_tokens_seen": 63056144,
      "step": 32765
    },
    {
      "epoch": 5.34586834162656,
      "grad_norm": 0.002762372372671962,
      "learning_rate": 0.023546875747023025,
      "loss": 0.1492,
      "num_input_tokens_seen": 63066208,
      "step": 32770
    },
    {
      "epoch": 5.346684068847377,
      "grad_norm": 0.0029424624517560005,
      "learning_rate": 0.02351520169672801,
      "loss": 0.1946,
      "num_input_tokens_seen": 63076080,
      "step": 32775
    },
    {
      "epoch": 5.3474997960681945,
      "grad_norm": 0.0022024367935955524,
      "learning_rate": 0.023483547151978357,
      "loss": 0.2017,
      "num_input_tokens_seen": 63087168,
      "step": 32780
    },
    {
      "epoch": 5.3483155232890125,
      "grad_norm": 0.002654964802786708,
      "learning_rate": 0.023451912117655675,
      "loss": 0.2282,
      "num_input_tokens_seen": 63097792,
      "step": 32785
    },
    {
      "epoch": 5.34913125050983,
      "grad_norm": 0.0023353819269686937,
      "learning_rate": 0.023420296598638417,
      "loss": 0.1437,
      "num_input_tokens_seen": 63106496,
      "step": 32790
    },
    {
      "epoch": 5.349946977730647,
      "grad_norm": 0.003161331173032522,
      "learning_rate": 0.023388700599802165,
      "loss": 0.1749,
      "num_input_tokens_seen": 63116016,
      "step": 32795
    },
    {
      "epoch": 5.350762704951464,
      "grad_norm": 0.004641054198145866,
      "learning_rate": 0.023357124126019334,
      "loss": 0.2548,
      "num_input_tokens_seen": 63124752,
      "step": 32800
    },
    {
      "epoch": 5.350762704951464,
      "eval_loss": 0.18904437124729156,
      "eval_runtime": 68.0794,
      "eval_samples_per_second": 40.027,
      "eval_steps_per_second": 20.021,
      "num_input_tokens_seen": 63124752,
      "step": 32800
    },
    {
      "epoch": 5.351578432172282,
      "grad_norm": 0.003027907572686672,
      "learning_rate": 0.02332556718215945,
      "loss": 0.234,
      "num_input_tokens_seen": 63134736,
      "step": 32805
    },
    {
      "epoch": 5.352394159393099,
      "grad_norm": 0.0021009936463087797,
      "learning_rate": 0.023294029773089035,
      "loss": 0.1808,
      "num_input_tokens_seen": 63143984,
      "step": 32810
    },
    {
      "epoch": 5.353209886613916,
      "grad_norm": 0.002427026629447937,
      "learning_rate": 0.023262511903671484,
      "loss": 0.1467,
      "num_input_tokens_seen": 63154128,
      "step": 32815
    },
    {
      "epoch": 5.354025613834733,
      "grad_norm": 0.003780130762606859,
      "learning_rate": 0.023231013578767324,
      "loss": 0.1944,
      "num_input_tokens_seen": 63163712,
      "step": 32820
    },
    {
      "epoch": 5.354841341055551,
      "grad_norm": 0.001857760944403708,
      "learning_rate": 0.0231995348032339,
      "loss": 0.2216,
      "num_input_tokens_seen": 63173120,
      "step": 32825
    },
    {
      "epoch": 5.3556570682763684,
      "grad_norm": 0.004431178327649832,
      "learning_rate": 0.023168075581925685,
      "loss": 0.246,
      "num_input_tokens_seen": 63183232,
      "step": 32830
    },
    {
      "epoch": 5.356472795497186,
      "grad_norm": 0.0029668754432350397,
      "learning_rate": 0.023136635919694126,
      "loss": 0.1341,
      "num_input_tokens_seen": 63192304,
      "step": 32835
    },
    {
      "epoch": 5.357288522718003,
      "grad_norm": 0.005469413008540869,
      "learning_rate": 0.02310521582138753,
      "loss": 0.2623,
      "num_input_tokens_seen": 63202560,
      "step": 32840
    },
    {
      "epoch": 5.358104249938821,
      "grad_norm": 0.003249703673645854,
      "learning_rate": 0.023073815291851357,
      "loss": 0.1931,
      "num_input_tokens_seen": 63211856,
      "step": 32845
    },
    {
      "epoch": 5.358919977159638,
      "grad_norm": 0.004454080946743488,
      "learning_rate": 0.02304243433592788,
      "loss": 0.2018,
      "num_input_tokens_seen": 63221344,
      "step": 32850
    },
    {
      "epoch": 5.359735704380455,
      "grad_norm": 0.0030432892963290215,
      "learning_rate": 0.023011072958456513,
      "loss": 0.2073,
      "num_input_tokens_seen": 63231056,
      "step": 32855
    },
    {
      "epoch": 5.360551431601272,
      "grad_norm": 0.002576084341853857,
      "learning_rate": 0.022979731164273536,
      "loss": 0.1744,
      "num_input_tokens_seen": 63241472,
      "step": 32860
    },
    {
      "epoch": 5.36136715882209,
      "grad_norm": 0.002133265370503068,
      "learning_rate": 0.022948408958212218,
      "loss": 0.2184,
      "num_input_tokens_seen": 63251120,
      "step": 32865
    },
    {
      "epoch": 5.362182886042907,
      "grad_norm": 0.0017853000899776816,
      "learning_rate": 0.022917106345102876,
      "loss": 0.1509,
      "num_input_tokens_seen": 63261152,
      "step": 32870
    },
    {
      "epoch": 5.362998613263724,
      "grad_norm": 0.0015633065486326814,
      "learning_rate": 0.022885823329772785,
      "loss": 0.1987,
      "num_input_tokens_seen": 63272096,
      "step": 32875
    },
    {
      "epoch": 5.3638143404845415,
      "grad_norm": 0.0014219204895198345,
      "learning_rate": 0.02285455991704612,
      "loss": 0.1706,
      "num_input_tokens_seen": 63280832,
      "step": 32880
    },
    {
      "epoch": 5.3646300677053596,
      "grad_norm": 0.001843627542257309,
      "learning_rate": 0.022823316111744117,
      "loss": 0.1416,
      "num_input_tokens_seen": 63290992,
      "step": 32885
    },
    {
      "epoch": 5.365445794926177,
      "grad_norm": 0.0015530670061707497,
      "learning_rate": 0.022792091918685014,
      "loss": 0.1624,
      "num_input_tokens_seen": 63300656,
      "step": 32890
    },
    {
      "epoch": 5.366261522146994,
      "grad_norm": 0.0027279802598059177,
      "learning_rate": 0.022760887342683906,
      "loss": 0.1767,
      "num_input_tokens_seen": 63309712,
      "step": 32895
    },
    {
      "epoch": 5.367077249367812,
      "grad_norm": 0.0034794325474649668,
      "learning_rate": 0.022729702388552975,
      "loss": 0.1877,
      "num_input_tokens_seen": 63319664,
      "step": 32900
    },
    {
      "epoch": 5.367892976588629,
      "grad_norm": 0.0037690370809286833,
      "learning_rate": 0.022698537061101292,
      "loss": 0.1938,
      "num_input_tokens_seen": 63330624,
      "step": 32905
    },
    {
      "epoch": 5.368708703809446,
      "grad_norm": 0.004551614169031382,
      "learning_rate": 0.022667391365134962,
      "loss": 0.186,
      "num_input_tokens_seen": 63340384,
      "step": 32910
    },
    {
      "epoch": 5.369524431030263,
      "grad_norm": 0.004567864816635847,
      "learning_rate": 0.022636265305457065,
      "loss": 0.1907,
      "num_input_tokens_seen": 63349504,
      "step": 32915
    },
    {
      "epoch": 5.370340158251081,
      "grad_norm": 0.002759229624643922,
      "learning_rate": 0.02260515888686764,
      "loss": 0.2095,
      "num_input_tokens_seen": 63358640,
      "step": 32920
    },
    {
      "epoch": 5.371155885471898,
      "grad_norm": 0.004660177510231733,
      "learning_rate": 0.022574072114163596,
      "loss": 0.2148,
      "num_input_tokens_seen": 63368448,
      "step": 32925
    },
    {
      "epoch": 5.3719716126927155,
      "grad_norm": 0.0026335790753364563,
      "learning_rate": 0.022543004992139005,
      "loss": 0.2187,
      "num_input_tokens_seen": 63377376,
      "step": 32930
    },
    {
      "epoch": 5.372787339913533,
      "grad_norm": 0.002397785196080804,
      "learning_rate": 0.022511957525584745,
      "loss": 0.2193,
      "num_input_tokens_seen": 63386784,
      "step": 32935
    },
    {
      "epoch": 5.373603067134351,
      "grad_norm": 0.00306420074775815,
      "learning_rate": 0.022480929719288778,
      "loss": 0.1614,
      "num_input_tokens_seen": 63396720,
      "step": 32940
    },
    {
      "epoch": 5.374418794355168,
      "grad_norm": 0.0020799292251467705,
      "learning_rate": 0.02244992157803592,
      "loss": 0.1509,
      "num_input_tokens_seen": 63405152,
      "step": 32945
    },
    {
      "epoch": 5.375234521575985,
      "grad_norm": 0.0027243110816925764,
      "learning_rate": 0.022418933106608047,
      "loss": 0.179,
      "num_input_tokens_seen": 63415280,
      "step": 32950
    },
    {
      "epoch": 5.376050248796802,
      "grad_norm": 0.004170479252934456,
      "learning_rate": 0.022387964309784018,
      "loss": 0.1897,
      "num_input_tokens_seen": 63424864,
      "step": 32955
    },
    {
      "epoch": 5.37686597601762,
      "grad_norm": 0.005571013782173395,
      "learning_rate": 0.022357015192339517,
      "loss": 0.2062,
      "num_input_tokens_seen": 63436464,
      "step": 32960
    },
    {
      "epoch": 5.377681703238437,
      "grad_norm": 0.004153477028012276,
      "learning_rate": 0.02232608575904734,
      "loss": 0.1968,
      "num_input_tokens_seen": 63445488,
      "step": 32965
    },
    {
      "epoch": 5.378497430459254,
      "grad_norm": 0.003917936701327562,
      "learning_rate": 0.022295176014677225,
      "loss": 0.1927,
      "num_input_tokens_seen": 63455472,
      "step": 32970
    },
    {
      "epoch": 5.3793131576800715,
      "grad_norm": 0.00371361942961812,
      "learning_rate": 0.02226428596399577,
      "loss": 0.1951,
      "num_input_tokens_seen": 63466464,
      "step": 32975
    },
    {
      "epoch": 5.3801288849008895,
      "grad_norm": 0.003452941309660673,
      "learning_rate": 0.02223341561176669,
      "loss": 0.1953,
      "num_input_tokens_seen": 63477856,
      "step": 32980
    },
    {
      "epoch": 5.380944612121707,
      "grad_norm": 0.0038588542956858873,
      "learning_rate": 0.0222025649627505,
      "loss": 0.1749,
      "num_input_tokens_seen": 63487984,
      "step": 32985
    },
    {
      "epoch": 5.381760339342524,
      "grad_norm": 0.0022552539594471455,
      "learning_rate": 0.022171734021704814,
      "loss": 0.1285,
      "num_input_tokens_seen": 63497888,
      "step": 32990
    },
    {
      "epoch": 5.382576066563341,
      "grad_norm": 0.0047652507200837135,
      "learning_rate": 0.022140922793384116,
      "loss": 0.2322,
      "num_input_tokens_seen": 63507920,
      "step": 32995
    },
    {
      "epoch": 5.383391793784159,
      "grad_norm": 0.0029305906500667334,
      "learning_rate": 0.022110131282539934,
      "loss": 0.1706,
      "num_input_tokens_seen": 63517792,
      "step": 33000
    },
    {
      "epoch": 5.383391793784159,
      "eval_loss": 0.18919792771339417,
      "eval_runtime": 68.0356,
      "eval_samples_per_second": 40.053,
      "eval_steps_per_second": 20.034,
      "num_input_tokens_seen": 63517792,
      "step": 33000
    },
    {
      "epoch": 5.384207521004976,
      "grad_norm": 0.0031509501859545708,
      "learning_rate": 0.022079359493920675,
      "loss": 0.1966,
      "num_input_tokens_seen": 63527312,
      "step": 33005
    },
    {
      "epoch": 5.385023248225793,
      "grad_norm": 0.0016094824532046914,
      "learning_rate": 0.02204860743227169,
      "loss": 0.1464,
      "num_input_tokens_seen": 63536768,
      "step": 33010
    },
    {
      "epoch": 5.38583897544661,
      "grad_norm": 0.0026404354721307755,
      "learning_rate": 0.022017875102335365,
      "loss": 0.1865,
      "num_input_tokens_seen": 63545520,
      "step": 33015
    },
    {
      "epoch": 5.386654702667428,
      "grad_norm": 0.003907662350684404,
      "learning_rate": 0.02198716250885108,
      "loss": 0.1628,
      "num_input_tokens_seen": 63555120,
      "step": 33020
    },
    {
      "epoch": 5.387470429888245,
      "grad_norm": 0.0013909615809097886,
      "learning_rate": 0.021956469656555,
      "loss": 0.1647,
      "num_input_tokens_seen": 63564496,
      "step": 33025
    },
    {
      "epoch": 5.388286157109063,
      "grad_norm": 0.0023363903164863586,
      "learning_rate": 0.0219257965501804,
      "loss": 0.2031,
      "num_input_tokens_seen": 63574736,
      "step": 33030
    },
    {
      "epoch": 5.38910188432988,
      "grad_norm": 0.003506901441141963,
      "learning_rate": 0.021895143194457494,
      "loss": 0.1737,
      "num_input_tokens_seen": 63584976,
      "step": 33035
    },
    {
      "epoch": 5.389917611550698,
      "grad_norm": 0.001738803694024682,
      "learning_rate": 0.021864509594113322,
      "loss": 0.1664,
      "num_input_tokens_seen": 63594528,
      "step": 33040
    },
    {
      "epoch": 5.390733338771515,
      "grad_norm": 0.002181906485930085,
      "learning_rate": 0.02183389575387207,
      "loss": 0.214,
      "num_input_tokens_seen": 63604992,
      "step": 33045
    },
    {
      "epoch": 5.391549065992332,
      "grad_norm": 0.0029434566386044025,
      "learning_rate": 0.021803301678454682,
      "loss": 0.1443,
      "num_input_tokens_seen": 63614048,
      "step": 33050
    },
    {
      "epoch": 5.392364793213149,
      "grad_norm": 0.00206805020570755,
      "learning_rate": 0.021772727372579213,
      "loss": 0.1824,
      "num_input_tokens_seen": 63624112,
      "step": 33055
    },
    {
      "epoch": 5.393180520433967,
      "grad_norm": 0.002038097009062767,
      "learning_rate": 0.02174217284096061,
      "loss": 0.1464,
      "num_input_tokens_seen": 63633504,
      "step": 33060
    },
    {
      "epoch": 5.393996247654784,
      "grad_norm": 0.0025156948249787092,
      "learning_rate": 0.0217116380883107,
      "loss": 0.1975,
      "num_input_tokens_seen": 63641520,
      "step": 33065
    },
    {
      "epoch": 5.394811974875601,
      "grad_norm": 0.002469914499670267,
      "learning_rate": 0.021681123119338425,
      "loss": 0.1622,
      "num_input_tokens_seen": 63651280,
      "step": 33070
    },
    {
      "epoch": 5.395627702096419,
      "grad_norm": 0.0028856315184384584,
      "learning_rate": 0.02165062793874951,
      "loss": 0.1842,
      "num_input_tokens_seen": 63661536,
      "step": 33075
    },
    {
      "epoch": 5.3964434293172365,
      "grad_norm": 0.001933643245138228,
      "learning_rate": 0.021620152551246666,
      "loss": 0.1448,
      "num_input_tokens_seen": 63670064,
      "step": 33080
    },
    {
      "epoch": 5.397259156538054,
      "grad_norm": 0.002241901820525527,
      "learning_rate": 0.02158969696152967,
      "loss": 0.127,
      "num_input_tokens_seen": 63680096,
      "step": 33085
    },
    {
      "epoch": 5.398074883758871,
      "grad_norm": 0.002790014958009124,
      "learning_rate": 0.021559261174295057,
      "loss": 0.1307,
      "num_input_tokens_seen": 63690208,
      "step": 33090
    },
    {
      "epoch": 5.398890610979688,
      "grad_norm": 0.004021522589027882,
      "learning_rate": 0.02152884519423646,
      "loss": 0.1382,
      "num_input_tokens_seen": 63700576,
      "step": 33095
    },
    {
      "epoch": 5.399706338200506,
      "grad_norm": 0.006631338968873024,
      "learning_rate": 0.021498449026044447,
      "loss": 0.2018,
      "num_input_tokens_seen": 63709216,
      "step": 33100
    },
    {
      "epoch": 5.400522065421323,
      "grad_norm": 0.003526502288877964,
      "learning_rate": 0.021468072674406414,
      "loss": 0.1935,
      "num_input_tokens_seen": 63718320,
      "step": 33105
    },
    {
      "epoch": 5.40133779264214,
      "grad_norm": 0.0020999503321945667,
      "learning_rate": 0.021437716144006795,
      "loss": 0.2149,
      "num_input_tokens_seen": 63728128,
      "step": 33110
    },
    {
      "epoch": 5.402153519862958,
      "grad_norm": 0.0034435808192938566,
      "learning_rate": 0.021407379439527002,
      "loss": 0.164,
      "num_input_tokens_seen": 63737472,
      "step": 33115
    },
    {
      "epoch": 5.402969247083775,
      "grad_norm": 0.00264205620624125,
      "learning_rate": 0.021377062565645255,
      "loss": 0.1797,
      "num_input_tokens_seen": 63745968,
      "step": 33120
    },
    {
      "epoch": 5.4037849743045925,
      "grad_norm": 0.003817421616986394,
      "learning_rate": 0.02134676552703688,
      "loss": 0.2059,
      "num_input_tokens_seen": 63755056,
      "step": 33125
    },
    {
      "epoch": 5.40460070152541,
      "grad_norm": 0.0029762363992631435,
      "learning_rate": 0.02131648832837398,
      "loss": 0.1822,
      "num_input_tokens_seen": 63764832,
      "step": 33130
    },
    {
      "epoch": 5.405416428746228,
      "grad_norm": 0.0019339809659868479,
      "learning_rate": 0.02128623097432574,
      "loss": 0.1765,
      "num_input_tokens_seen": 63774064,
      "step": 33135
    },
    {
      "epoch": 5.406232155967045,
      "grad_norm": 0.00219321739859879,
      "learning_rate": 0.021255993469558192,
      "loss": 0.1911,
      "num_input_tokens_seen": 63783680,
      "step": 33140
    },
    {
      "epoch": 5.407047883187862,
      "grad_norm": 0.00377703458070755,
      "learning_rate": 0.021225775818734364,
      "loss": 0.197,
      "num_input_tokens_seen": 63793744,
      "step": 33145
    },
    {
      "epoch": 5.407863610408679,
      "grad_norm": 0.002259765984490514,
      "learning_rate": 0.021195578026514166,
      "loss": 0.2019,
      "num_input_tokens_seen": 63801264,
      "step": 33150
    },
    {
      "epoch": 5.408679337629497,
      "grad_norm": 0.0026997660752385855,
      "learning_rate": 0.02116540009755452,
      "loss": 0.1351,
      "num_input_tokens_seen": 63810608,
      "step": 33155
    },
    {
      "epoch": 5.409495064850314,
      "grad_norm": 0.010272460989654064,
      "learning_rate": 0.021135242036509173,
      "loss": 0.1859,
      "num_input_tokens_seen": 63819760,
      "step": 33160
    },
    {
      "epoch": 5.410310792071131,
      "grad_norm": 0.005847060587257147,
      "learning_rate": 0.021105103848028967,
      "loss": 0.2396,
      "num_input_tokens_seen": 63829040,
      "step": 33165
    },
    {
      "epoch": 5.411126519291948,
      "grad_norm": 0.004171398002654314,
      "learning_rate": 0.021074985536761504,
      "loss": 0.2455,
      "num_input_tokens_seen": 63838608,
      "step": 33170
    },
    {
      "epoch": 5.4119422465127665,
      "grad_norm": 0.0025924737565219402,
      "learning_rate": 0.021044887107351435,
      "loss": 0.168,
      "num_input_tokens_seen": 63847296,
      "step": 33175
    },
    {
      "epoch": 5.412757973733584,
      "grad_norm": 0.0023489727173000574,
      "learning_rate": 0.021014808564440362,
      "loss": 0.1682,
      "num_input_tokens_seen": 63857456,
      "step": 33180
    },
    {
      "epoch": 5.413573700954401,
      "grad_norm": 0.0032608078327029943,
      "learning_rate": 0.02098474991266671,
      "loss": 0.1697,
      "num_input_tokens_seen": 63866848,
      "step": 33185
    },
    {
      "epoch": 5.414389428175218,
      "grad_norm": 0.005523335188627243,
      "learning_rate": 0.02095471115666592,
      "loss": 0.2122,
      "num_input_tokens_seen": 63877488,
      "step": 33190
    },
    {
      "epoch": 5.415205155396036,
      "grad_norm": 0.003330590669065714,
      "learning_rate": 0.020924692301070406,
      "loss": 0.1466,
      "num_input_tokens_seen": 63885824,
      "step": 33195
    },
    {
      "epoch": 5.416020882616853,
      "grad_norm": 0.0018546748906373978,
      "learning_rate": 0.020894693350509346,
      "loss": 0.2016,
      "num_input_tokens_seen": 63894896,
      "step": 33200
    },
    {
      "epoch": 5.416020882616853,
      "eval_loss": 0.18836617469787598,
      "eval_runtime": 68.0127,
      "eval_samples_per_second": 40.066,
      "eval_steps_per_second": 20.04,
      "num_input_tokens_seen": 63894896,
      "step": 33200
    },
    {
      "epoch": 5.41683660983767,
      "grad_norm": 0.007220160681754351,
      "learning_rate": 0.020864714309609057,
      "loss": 0.2497,
      "num_input_tokens_seen": 63904192,
      "step": 33205
    },
    {
      "epoch": 5.417652337058487,
      "grad_norm": 0.003736252663657069,
      "learning_rate": 0.020834755182992604,
      "loss": 0.1599,
      "num_input_tokens_seen": 63913296,
      "step": 33210
    },
    {
      "epoch": 5.418468064279305,
      "grad_norm": 0.002485757926478982,
      "learning_rate": 0.02080481597528011,
      "loss": 0.148,
      "num_input_tokens_seen": 63922624,
      "step": 33215
    },
    {
      "epoch": 5.419283791500122,
      "grad_norm": 0.003969453275203705,
      "learning_rate": 0.020774896691088583,
      "loss": 0.2189,
      "num_input_tokens_seen": 63930640,
      "step": 33220
    },
    {
      "epoch": 5.4200995187209395,
      "grad_norm": 0.002632062416523695,
      "learning_rate": 0.020744997335031882,
      "loss": 0.2311,
      "num_input_tokens_seen": 63940528,
      "step": 33225
    },
    {
      "epoch": 5.420915245941757,
      "grad_norm": 0.00369091285392642,
      "learning_rate": 0.02071511791172092,
      "loss": 0.1982,
      "num_input_tokens_seen": 63951168,
      "step": 33230
    },
    {
      "epoch": 5.421730973162575,
      "grad_norm": 0.0028968406841158867,
      "learning_rate": 0.02068525842576351,
      "loss": 0.1884,
      "num_input_tokens_seen": 63960320,
      "step": 33235
    },
    {
      "epoch": 5.422546700383392,
      "grad_norm": 0.003713390091434121,
      "learning_rate": 0.020655418881764264,
      "loss": 0.1792,
      "num_input_tokens_seen": 63969888,
      "step": 33240
    },
    {
      "epoch": 5.423362427604209,
      "grad_norm": 0.0029092729091644287,
      "learning_rate": 0.020625599284324923,
      "loss": 0.1618,
      "num_input_tokens_seen": 63979104,
      "step": 33245
    },
    {
      "epoch": 5.424178154825027,
      "grad_norm": 0.0031975044403225183,
      "learning_rate": 0.02059579963804396,
      "loss": 0.1987,
      "num_input_tokens_seen": 63988864,
      "step": 33250
    },
    {
      "epoch": 5.424993882045844,
      "grad_norm": 0.003944295458495617,
      "learning_rate": 0.02056601994751688,
      "loss": 0.1894,
      "num_input_tokens_seen": 63999488,
      "step": 33255
    },
    {
      "epoch": 5.425809609266661,
      "grad_norm": 0.003380594775080681,
      "learning_rate": 0.02053626021733614,
      "loss": 0.1886,
      "num_input_tokens_seen": 64009568,
      "step": 33260
    },
    {
      "epoch": 5.426625336487478,
      "grad_norm": 0.0025787362828850746,
      "learning_rate": 0.02050652045209097,
      "loss": 0.2092,
      "num_input_tokens_seen": 64018704,
      "step": 33265
    },
    {
      "epoch": 5.4274410637082955,
      "grad_norm": 0.005892825312912464,
      "learning_rate": 0.020476800656367672,
      "loss": 0.2092,
      "num_input_tokens_seen": 64028560,
      "step": 33270
    },
    {
      "epoch": 5.4282567909291135,
      "grad_norm": 0.003684799186885357,
      "learning_rate": 0.020447100834749425,
      "loss": 0.1767,
      "num_input_tokens_seen": 64038416,
      "step": 33275
    },
    {
      "epoch": 5.429072518149931,
      "grad_norm": 0.0034524076618254185,
      "learning_rate": 0.02041742099181627,
      "loss": 0.168,
      "num_input_tokens_seen": 64047712,
      "step": 33280
    },
    {
      "epoch": 5.429888245370748,
      "grad_norm": 0.002980235032737255,
      "learning_rate": 0.02038776113214526,
      "loss": 0.2061,
      "num_input_tokens_seen": 64057024,
      "step": 33285
    },
    {
      "epoch": 5.430703972591566,
      "grad_norm": 0.002034648321568966,
      "learning_rate": 0.0203581212603103,
      "loss": 0.1185,
      "num_input_tokens_seen": 64065648,
      "step": 33290
    },
    {
      "epoch": 5.431519699812383,
      "grad_norm": 0.0017309694085270166,
      "learning_rate": 0.02032850138088219,
      "loss": 0.0989,
      "num_input_tokens_seen": 64075248,
      "step": 33295
    },
    {
      "epoch": 5.4323354270332,
      "grad_norm": 0.0021291752345860004,
      "learning_rate": 0.020298901498428754,
      "loss": 0.1717,
      "num_input_tokens_seen": 64084368,
      "step": 33300
    },
    {
      "epoch": 5.433151154254017,
      "grad_norm": 0.0019710585474967957,
      "learning_rate": 0.020269321617514595,
      "loss": 0.1777,
      "num_input_tokens_seen": 64093424,
      "step": 33305
    },
    {
      "epoch": 5.433966881474835,
      "grad_norm": 0.0039474451914429665,
      "learning_rate": 0.020239761742701343,
      "loss": 0.2211,
      "num_input_tokens_seen": 64104384,
      "step": 33310
    },
    {
      "epoch": 5.434782608695652,
      "grad_norm": 0.0021390183828771114,
      "learning_rate": 0.02021022187854754,
      "loss": 0.1569,
      "num_input_tokens_seen": 64113664,
      "step": 33315
    },
    {
      "epoch": 5.4355983359164695,
      "grad_norm": 0.0025833987165242434,
      "learning_rate": 0.020180702029608522,
      "loss": 0.1758,
      "num_input_tokens_seen": 64123392,
      "step": 33320
    },
    {
      "epoch": 5.436414063137287,
      "grad_norm": 0.002406199462711811,
      "learning_rate": 0.020151202200436695,
      "loss": 0.1576,
      "num_input_tokens_seen": 64132352,
      "step": 33325
    },
    {
      "epoch": 5.437229790358105,
      "grad_norm": 0.0030552647076547146,
      "learning_rate": 0.020121722395581226,
      "loss": 0.1294,
      "num_input_tokens_seen": 64142416,
      "step": 33330
    },
    {
      "epoch": 5.438045517578922,
      "grad_norm": 0.004883520305156708,
      "learning_rate": 0.020092262619588342,
      "loss": 0.1738,
      "num_input_tokens_seen": 64152256,
      "step": 33335
    },
    {
      "epoch": 5.438861244799739,
      "grad_norm": 0.00309509364888072,
      "learning_rate": 0.02006282287700109,
      "loss": 0.1903,
      "num_input_tokens_seen": 64161856,
      "step": 33340
    },
    {
      "epoch": 5.439676972020556,
      "grad_norm": 0.0021456729155033827,
      "learning_rate": 0.020033403172359427,
      "loss": 0.2101,
      "num_input_tokens_seen": 64171952,
      "step": 33345
    },
    {
      "epoch": 5.440492699241374,
      "grad_norm": 0.0038908012211322784,
      "learning_rate": 0.020004003510200284,
      "loss": 0.189,
      "num_input_tokens_seen": 64182176,
      "step": 33350
    },
    {
      "epoch": 5.441308426462191,
      "grad_norm": 0.0028816231060773134,
      "learning_rate": 0.019974623895057407,
      "loss": 0.1891,
      "num_input_tokens_seen": 64192016,
      "step": 33355
    },
    {
      "epoch": 5.442124153683008,
      "grad_norm": 0.0036078677512705326,
      "learning_rate": 0.019945264331461553,
      "loss": 0.1868,
      "num_input_tokens_seen": 64201392,
      "step": 33360
    },
    {
      "epoch": 5.442939880903825,
      "grad_norm": 0.0036085161846131086,
      "learning_rate": 0.019915924823940317,
      "loss": 0.1341,
      "num_input_tokens_seen": 64211360,
      "step": 33365
    },
    {
      "epoch": 5.443755608124643,
      "grad_norm": 0.0034312037751078606,
      "learning_rate": 0.01988660537701816,
      "loss": 0.1534,
      "num_input_tokens_seen": 64221264,
      "step": 33370
    },
    {
      "epoch": 5.444571335345461,
      "grad_norm": 0.0035553360357880592,
      "learning_rate": 0.01985730599521659,
      "loss": 0.1806,
      "num_input_tokens_seen": 64231440,
      "step": 33375
    },
    {
      "epoch": 5.445387062566278,
      "grad_norm": 0.0030216400045901537,
      "learning_rate": 0.019828026683053918,
      "loss": 0.1618,
      "num_input_tokens_seen": 64239520,
      "step": 33380
    },
    {
      "epoch": 5.446202789787095,
      "grad_norm": 0.003126283409073949,
      "learning_rate": 0.01979876744504535,
      "loss": 0.2816,
      "num_input_tokens_seen": 64250096,
      "step": 33385
    },
    {
      "epoch": 5.447018517007913,
      "grad_norm": 0.0023792546708136797,
      "learning_rate": 0.019769528285703046,
      "loss": 0.2044,
      "num_input_tokens_seen": 64259184,
      "step": 33390
    },
    {
      "epoch": 5.44783424422873,
      "grad_norm": 0.0023868074640631676,
      "learning_rate": 0.019740309209536098,
      "loss": 0.1937,
      "num_input_tokens_seen": 64268912,
      "step": 33395
    },
    {
      "epoch": 5.448649971449547,
      "grad_norm": 0.0016027356032282114,
      "learning_rate": 0.019711110221050387,
      "loss": 0.1539,
      "num_input_tokens_seen": 64277584,
      "step": 33400
    },
    {
      "epoch": 5.448649971449547,
      "eval_loss": 0.18957702815532684,
      "eval_runtime": 68.048,
      "eval_samples_per_second": 40.045,
      "eval_steps_per_second": 20.03,
      "num_input_tokens_seen": 64277584,
      "step": 33400
    },
    {
      "epoch": 5.449465698670364,
      "grad_norm": 0.00557049410417676,
      "learning_rate": 0.019681931324748825,
      "loss": 0.2066,
      "num_input_tokens_seen": 64286688,
      "step": 33405
    },
    {
      "epoch": 5.450281425891182,
      "grad_norm": 0.0033919159322977066,
      "learning_rate": 0.019652772525131094,
      "loss": 0.1931,
      "num_input_tokens_seen": 64296240,
      "step": 33410
    },
    {
      "epoch": 5.451097153111999,
      "grad_norm": 0.0026682810857892036,
      "learning_rate": 0.019623633826693885,
      "loss": 0.1501,
      "num_input_tokens_seen": 64305248,
      "step": 33415
    },
    {
      "epoch": 5.4519128803328165,
      "grad_norm": 0.003492664312943816,
      "learning_rate": 0.019594515233930788,
      "loss": 0.1923,
      "num_input_tokens_seen": 64315008,
      "step": 33420
    },
    {
      "epoch": 5.4527286075536345,
      "grad_norm": 0.0022562211379408836,
      "learning_rate": 0.019565416751332186,
      "loss": 0.1537,
      "num_input_tokens_seen": 64324480,
      "step": 33425
    },
    {
      "epoch": 5.453544334774452,
      "grad_norm": 0.00401936424896121,
      "learning_rate": 0.019536338383385497,
      "loss": 0.2495,
      "num_input_tokens_seen": 64334096,
      "step": 33430
    },
    {
      "epoch": 5.454360061995269,
      "grad_norm": 0.00373516115359962,
      "learning_rate": 0.019507280134574933,
      "loss": 0.1707,
      "num_input_tokens_seen": 64344528,
      "step": 33435
    },
    {
      "epoch": 5.455175789216086,
      "grad_norm": 0.0033658291213214397,
      "learning_rate": 0.019478242009381624,
      "loss": 0.1728,
      "num_input_tokens_seen": 64354160,
      "step": 33440
    },
    {
      "epoch": 5.455991516436903,
      "grad_norm": 0.0045992848463356495,
      "learning_rate": 0.01944922401228367,
      "loss": 0.1964,
      "num_input_tokens_seen": 64363360,
      "step": 33445
    },
    {
      "epoch": 5.456807243657721,
      "grad_norm": 0.0035474395845085382,
      "learning_rate": 0.01942022614775593,
      "loss": 0.2094,
      "num_input_tokens_seen": 64372336,
      "step": 33450
    },
    {
      "epoch": 5.457622970878538,
      "grad_norm": 0.002471728017553687,
      "learning_rate": 0.01939124842027029,
      "loss": 0.2078,
      "num_input_tokens_seen": 64381488,
      "step": 33455
    },
    {
      "epoch": 5.458438698099355,
      "grad_norm": 0.002636134624481201,
      "learning_rate": 0.01936229083429551,
      "loss": 0.1694,
      "num_input_tokens_seen": 64390272,
      "step": 33460
    },
    {
      "epoch": 5.459254425320173,
      "grad_norm": 0.002330525778234005,
      "learning_rate": 0.019333353394297148,
      "loss": 0.1244,
      "num_input_tokens_seen": 64399344,
      "step": 33465
    },
    {
      "epoch": 5.4600701525409905,
      "grad_norm": 0.0043294127099215984,
      "learning_rate": 0.019304436104737754,
      "loss": 0.1648,
      "num_input_tokens_seen": 64409488,
      "step": 33470
    },
    {
      "epoch": 5.460885879761808,
      "grad_norm": 0.0016782035818323493,
      "learning_rate": 0.019275538970076778,
      "loss": 0.131,
      "num_input_tokens_seen": 64419456,
      "step": 33475
    },
    {
      "epoch": 5.461701606982625,
      "grad_norm": 0.004791988525539637,
      "learning_rate": 0.019246661994770434,
      "loss": 0.2155,
      "num_input_tokens_seen": 64429296,
      "step": 33480
    },
    {
      "epoch": 5.462517334203443,
      "grad_norm": 0.004099271725863218,
      "learning_rate": 0.019217805183271985,
      "loss": 0.2102,
      "num_input_tokens_seen": 64439744,
      "step": 33485
    },
    {
      "epoch": 5.46333306142426,
      "grad_norm": 0.002927232999354601,
      "learning_rate": 0.019188968540031465,
      "loss": 0.1946,
      "num_input_tokens_seen": 64450048,
      "step": 33490
    },
    {
      "epoch": 5.464148788645077,
      "grad_norm": 0.0030280884820967913,
      "learning_rate": 0.019160152069495867,
      "loss": 0.1781,
      "num_input_tokens_seen": 64458128,
      "step": 33495
    },
    {
      "epoch": 5.464964515865894,
      "grad_norm": 0.0018773681949824095,
      "learning_rate": 0.019131355776109103,
      "loss": 0.1368,
      "num_input_tokens_seen": 64468112,
      "step": 33500
    },
    {
      "epoch": 5.465780243086712,
      "grad_norm": 0.001720358501188457,
      "learning_rate": 0.019102579664311857,
      "loss": 0.2091,
      "num_input_tokens_seen": 64477232,
      "step": 33505
    },
    {
      "epoch": 5.466595970307529,
      "grad_norm": 0.004284876398742199,
      "learning_rate": 0.019073823738541763,
      "loss": 0.1895,
      "num_input_tokens_seen": 64486512,
      "step": 33510
    },
    {
      "epoch": 5.467411697528346,
      "grad_norm": 0.0016174643533304334,
      "learning_rate": 0.0190450880032334,
      "loss": 0.2127,
      "num_input_tokens_seen": 64495696,
      "step": 33515
    },
    {
      "epoch": 5.468227424749164,
      "grad_norm": 0.00339776324108243,
      "learning_rate": 0.019016372462818114,
      "loss": 0.1537,
      "num_input_tokens_seen": 64505440,
      "step": 33520
    },
    {
      "epoch": 5.469043151969982,
      "grad_norm": 0.0021132819820195436,
      "learning_rate": 0.018987677121724278,
      "loss": 0.1947,
      "num_input_tokens_seen": 64515728,
      "step": 33525
    },
    {
      "epoch": 5.469858879190799,
      "grad_norm": 0.004395420663058758,
      "learning_rate": 0.018959001984377,
      "loss": 0.1728,
      "num_input_tokens_seen": 64525616,
      "step": 33530
    },
    {
      "epoch": 5.470674606411616,
      "grad_norm": 0.003715881146490574,
      "learning_rate": 0.018930347055198377,
      "loss": 0.1898,
      "num_input_tokens_seen": 64535504,
      "step": 33535
    },
    {
      "epoch": 5.471490333632433,
      "grad_norm": 0.006622908636927605,
      "learning_rate": 0.01890171233860739,
      "loss": 0.1609,
      "num_input_tokens_seen": 64544864,
      "step": 33540
    },
    {
      "epoch": 5.472306060853251,
      "grad_norm": 0.004491704516112804,
      "learning_rate": 0.018873097839019807,
      "loss": 0.1609,
      "num_input_tokens_seen": 64553504,
      "step": 33545
    },
    {
      "epoch": 5.473121788074068,
      "grad_norm": 0.005502310581505299,
      "learning_rate": 0.0188445035608484,
      "loss": 0.1501,
      "num_input_tokens_seen": 64562112,
      "step": 33550
    },
    {
      "epoch": 5.473937515294885,
      "grad_norm": 0.005505615379661322,
      "learning_rate": 0.018815929508502777,
      "loss": 0.1946,
      "num_input_tokens_seen": 64572576,
      "step": 33555
    },
    {
      "epoch": 5.474753242515702,
      "grad_norm": 0.007492359261959791,
      "learning_rate": 0.01878737568638934,
      "loss": 0.2334,
      "num_input_tokens_seen": 64583456,
      "step": 33560
    },
    {
      "epoch": 5.47556896973652,
      "grad_norm": 0.003806512104347348,
      "learning_rate": 0.01875884209891152,
      "loss": 0.1479,
      "num_input_tokens_seen": 64593824,
      "step": 33565
    },
    {
      "epoch": 5.4763846969573375,
      "grad_norm": 0.0042943707667291164,
      "learning_rate": 0.018730328750469514,
      "loss": 0.1583,
      "num_input_tokens_seen": 64603968,
      "step": 33570
    },
    {
      "epoch": 5.477200424178155,
      "grad_norm": 0.0033889010082930326,
      "learning_rate": 0.018701835645460473,
      "loss": 0.1809,
      "num_input_tokens_seen": 64613808,
      "step": 33575
    },
    {
      "epoch": 5.478016151398972,
      "grad_norm": 0.00447272602468729,
      "learning_rate": 0.01867336278827838,
      "loss": 0.2248,
      "num_input_tokens_seen": 64623952,
      "step": 33580
    },
    {
      "epoch": 5.47883187861979,
      "grad_norm": 0.007538728881627321,
      "learning_rate": 0.018644910183314056,
      "loss": 0.2376,
      "num_input_tokens_seen": 64633872,
      "step": 33585
    },
    {
      "epoch": 5.479647605840607,
      "grad_norm": 0.0034967989195138216,
      "learning_rate": 0.01861647783495531,
      "loss": 0.253,
      "num_input_tokens_seen": 64642800,
      "step": 33590
    },
    {
      "epoch": 5.480463333061424,
      "grad_norm": 0.002444677986204624,
      "learning_rate": 0.01858806574758676,
      "loss": 0.1729,
      "num_input_tokens_seen": 64651808,
      "step": 33595
    },
    {
      "epoch": 5.481279060282241,
      "grad_norm": 0.002817892236635089,
      "learning_rate": 0.01855967392558988,
      "loss": 0.263,
      "num_input_tokens_seen": 64661856,
      "step": 33600
    },
    {
      "epoch": 5.481279060282241,
      "eval_loss": 0.19102953374385834,
      "eval_runtime": 68.0958,
      "eval_samples_per_second": 40.017,
      "eval_steps_per_second": 20.016,
      "num_input_tokens_seen": 64661856,
      "step": 33600
    },
    {
      "epoch": 5.482094787503059,
      "grad_norm": 0.0023357209283858538,
      "learning_rate": 0.018531302373343096,
      "loss": 0.1596,
      "num_input_tokens_seen": 64671856,
      "step": 33605
    },
    {
      "epoch": 5.482910514723876,
      "grad_norm": 0.003324100747704506,
      "learning_rate": 0.018502951095221588,
      "loss": 0.1825,
      "num_input_tokens_seen": 64681632,
      "step": 33610
    },
    {
      "epoch": 5.4837262419446935,
      "grad_norm": 0.0036079918500036,
      "learning_rate": 0.01847462009559751,
      "loss": 0.1621,
      "num_input_tokens_seen": 64691728,
      "step": 33615
    },
    {
      "epoch": 5.484541969165511,
      "grad_norm": 0.005355818197131157,
      "learning_rate": 0.01844630937883992,
      "loss": 0.175,
      "num_input_tokens_seen": 64702208,
      "step": 33620
    },
    {
      "epoch": 5.485357696386329,
      "grad_norm": 0.001386558637022972,
      "learning_rate": 0.018418018949314573,
      "loss": 0.1218,
      "num_input_tokens_seen": 64712208,
      "step": 33625
    },
    {
      "epoch": 5.486173423607146,
      "grad_norm": 0.0023900859523564577,
      "learning_rate": 0.018389748811384315,
      "loss": 0.1557,
      "num_input_tokens_seen": 64722080,
      "step": 33630
    },
    {
      "epoch": 5.486989150827963,
      "grad_norm": 0.0030763756949454546,
      "learning_rate": 0.018361498969408658,
      "loss": 0.1772,
      "num_input_tokens_seen": 64731984,
      "step": 33635
    },
    {
      "epoch": 5.487804878048781,
      "grad_norm": 0.002017237478867173,
      "learning_rate": 0.01833326942774415,
      "loss": 0.1786,
      "num_input_tokens_seen": 64741936,
      "step": 33640
    },
    {
      "epoch": 5.488620605269598,
      "grad_norm": 0.0034673172049224377,
      "learning_rate": 0.018305060190744155,
      "loss": 0.1761,
      "num_input_tokens_seen": 64752048,
      "step": 33645
    },
    {
      "epoch": 5.489436332490415,
      "grad_norm": 0.0035127580631524324,
      "learning_rate": 0.018276871262758846,
      "loss": 0.1687,
      "num_input_tokens_seen": 64761600,
      "step": 33650
    },
    {
      "epoch": 5.490252059711232,
      "grad_norm": 0.0037704543210566044,
      "learning_rate": 0.0182487026481353,
      "loss": 0.2467,
      "num_input_tokens_seen": 64770816,
      "step": 33655
    },
    {
      "epoch": 5.49106778693205,
      "grad_norm": 0.003769446397200227,
      "learning_rate": 0.018220554351217538,
      "loss": 0.2304,
      "num_input_tokens_seen": 64779536,
      "step": 33660
    },
    {
      "epoch": 5.4918835141528675,
      "grad_norm": 0.0029788664542138577,
      "learning_rate": 0.01819242637634629,
      "loss": 0.1617,
      "num_input_tokens_seen": 64789664,
      "step": 33665
    },
    {
      "epoch": 5.492699241373685,
      "grad_norm": 0.0038100816309452057,
      "learning_rate": 0.01816431872785933,
      "loss": 0.1938,
      "num_input_tokens_seen": 64799760,
      "step": 33670
    },
    {
      "epoch": 5.493514968594502,
      "grad_norm": 0.001868689083494246,
      "learning_rate": 0.018136231410091148,
      "loss": 0.1865,
      "num_input_tokens_seen": 64809264,
      "step": 33675
    },
    {
      "epoch": 5.49433069581532,
      "grad_norm": 0.005756228230893612,
      "learning_rate": 0.018108164427373175,
      "loss": 0.2196,
      "num_input_tokens_seen": 64819152,
      "step": 33680
    },
    {
      "epoch": 5.495146423036137,
      "grad_norm": 0.003831419860944152,
      "learning_rate": 0.01808011778403375,
      "loss": 0.242,
      "num_input_tokens_seen": 64827968,
      "step": 33685
    },
    {
      "epoch": 5.495962150256954,
      "grad_norm": 0.003896398702636361,
      "learning_rate": 0.01805209148439793,
      "loss": 0.1267,
      "num_input_tokens_seen": 64837088,
      "step": 33690
    },
    {
      "epoch": 5.496777877477771,
      "grad_norm": 0.004156498238444328,
      "learning_rate": 0.018024085532787757,
      "loss": 0.172,
      "num_input_tokens_seen": 64846336,
      "step": 33695
    },
    {
      "epoch": 5.497593604698589,
      "grad_norm": 0.003457619808614254,
      "learning_rate": 0.017996099933522164,
      "loss": 0.2526,
      "num_input_tokens_seen": 64856224,
      "step": 33700
    },
    {
      "epoch": 5.498409331919406,
      "grad_norm": 0.0032376116141676903,
      "learning_rate": 0.017968134690916775,
      "loss": 0.1812,
      "num_input_tokens_seen": 64865392,
      "step": 33705
    },
    {
      "epoch": 5.499225059140223,
      "grad_norm": 0.003082434181123972,
      "learning_rate": 0.017940189809284263,
      "loss": 0.1898,
      "num_input_tokens_seen": 64874848,
      "step": 33710
    },
    {
      "epoch": 5.5000407863610405,
      "grad_norm": 0.0019533170852810144,
      "learning_rate": 0.017912265292934024,
      "loss": 0.1277,
      "num_input_tokens_seen": 64884496,
      "step": 33715
    },
    {
      "epoch": 5.500856513581859,
      "grad_norm": 0.0026874272152781487,
      "learning_rate": 0.017884361146172423,
      "loss": 0.1561,
      "num_input_tokens_seen": 64894464,
      "step": 33720
    },
    {
      "epoch": 5.501672240802676,
      "grad_norm": 0.005375124979764223,
      "learning_rate": 0.01785647737330261,
      "loss": 0.1779,
      "num_input_tokens_seen": 64902192,
      "step": 33725
    },
    {
      "epoch": 5.502487968023493,
      "grad_norm": 0.0032493346370756626,
      "learning_rate": 0.017828613978624563,
      "loss": 0.1397,
      "num_input_tokens_seen": 64910208,
      "step": 33730
    },
    {
      "epoch": 5.50330369524431,
      "grad_norm": 0.0014473265036940575,
      "learning_rate": 0.01780077096643523,
      "loss": 0.1585,
      "num_input_tokens_seen": 64918928,
      "step": 33735
    },
    {
      "epoch": 5.504119422465128,
      "grad_norm": 0.003916085232049227,
      "learning_rate": 0.017772948341028345,
      "loss": 0.198,
      "num_input_tokens_seen": 64928320,
      "step": 33740
    },
    {
      "epoch": 5.504935149685945,
      "grad_norm": 0.0048216828145086765,
      "learning_rate": 0.01774514610669447,
      "loss": 0.1934,
      "num_input_tokens_seen": 64938304,
      "step": 33745
    },
    {
      "epoch": 5.505750876906762,
      "grad_norm": 0.004191816784441471,
      "learning_rate": 0.017717364267721112,
      "loss": 0.1702,
      "num_input_tokens_seen": 64948464,
      "step": 33750
    },
    {
      "epoch": 5.506566604127579,
      "grad_norm": 0.0024726015981286764,
      "learning_rate": 0.017689602828392513,
      "loss": 0.212,
      "num_input_tokens_seen": 64957856,
      "step": 33755
    },
    {
      "epoch": 5.507382331348397,
      "grad_norm": 0.003225757507607341,
      "learning_rate": 0.017661861792989897,
      "loss": 0.1612,
      "num_input_tokens_seen": 64966464,
      "step": 33760
    },
    {
      "epoch": 5.5081980585692145,
      "grad_norm": 0.0034238197840750217,
      "learning_rate": 0.017634141165791272,
      "loss": 0.1915,
      "num_input_tokens_seen": 64976512,
      "step": 33765
    },
    {
      "epoch": 5.509013785790032,
      "grad_norm": 0.0016666088486090302,
      "learning_rate": 0.017606440951071455,
      "loss": 0.1599,
      "num_input_tokens_seen": 64986880,
      "step": 33770
    },
    {
      "epoch": 5.50982951301085,
      "grad_norm": 0.0048047578893601894,
      "learning_rate": 0.017578761153102213,
      "loss": 0.1659,
      "num_input_tokens_seen": 64995648,
      "step": 33775
    },
    {
      "epoch": 5.510645240231667,
      "grad_norm": 0.002484109252691269,
      "learning_rate": 0.017551101776152146,
      "loss": 0.1883,
      "num_input_tokens_seen": 65005568,
      "step": 33780
    },
    {
      "epoch": 5.511460967452484,
      "grad_norm": 0.0020574554800987244,
      "learning_rate": 0.017523462824486608,
      "loss": 0.1877,
      "num_input_tokens_seen": 65014432,
      "step": 33785
    },
    {
      "epoch": 5.512276694673301,
      "grad_norm": 0.0009069332154467702,
      "learning_rate": 0.01749584430236794,
      "loss": 0.1648,
      "num_input_tokens_seen": 65023872,
      "step": 33790
    },
    {
      "epoch": 5.513092421894118,
      "grad_norm": 0.0033930817153304815,
      "learning_rate": 0.01746824621405524,
      "loss": 0.1648,
      "num_input_tokens_seen": 65033280,
      "step": 33795
    },
    {
      "epoch": 5.513908149114936,
      "grad_norm": 0.004821301903575659,
      "learning_rate": 0.017440668563804412,
      "loss": 0.174,
      "num_input_tokens_seen": 65043136,
      "step": 33800
    },
    {
      "epoch": 5.513908149114936,
      "eval_loss": 0.18977025151252747,
      "eval_runtime": 67.9804,
      "eval_samples_per_second": 40.085,
      "eval_steps_per_second": 20.05,
      "num_input_tokens_seen": 65043136,
      "step": 33800
    },
    {
      "epoch": 5.514723876335753,
      "grad_norm": 0.002506476826965809,
      "learning_rate": 0.017413111355868392,
      "loss": 0.1258,
      "num_input_tokens_seen": 65053504,
      "step": 33805
    },
    {
      "epoch": 5.5155396035565705,
      "grad_norm": 0.0027487692423164845,
      "learning_rate": 0.017385574594496748,
      "loss": 0.2025,
      "num_input_tokens_seen": 65063120,
      "step": 33810
    },
    {
      "epoch": 5.5163553307773885,
      "grad_norm": 0.0032394875306636095,
      "learning_rate": 0.01735805828393605,
      "loss": 0.1908,
      "num_input_tokens_seen": 65073824,
      "step": 33815
    },
    {
      "epoch": 5.517171057998206,
      "grad_norm": 0.00336196506395936,
      "learning_rate": 0.017330562428429667,
      "loss": 0.1135,
      "num_input_tokens_seen": 65083952,
      "step": 33820
    },
    {
      "epoch": 5.517986785219023,
      "grad_norm": 0.002319222316145897,
      "learning_rate": 0.01730308703221776,
      "loss": 0.2299,
      "num_input_tokens_seen": 65094704,
      "step": 33825
    },
    {
      "epoch": 5.51880251243984,
      "grad_norm": 0.0029620639979839325,
      "learning_rate": 0.01727563209953744,
      "loss": 0.1795,
      "num_input_tokens_seen": 65105392,
      "step": 33830
    },
    {
      "epoch": 5.519618239660657,
      "grad_norm": 0.002876711543649435,
      "learning_rate": 0.017248197634622535,
      "loss": 0.1473,
      "num_input_tokens_seen": 65114976,
      "step": 33835
    },
    {
      "epoch": 5.520433966881475,
      "grad_norm": 0.0021359059028327465,
      "learning_rate": 0.01722078364170383,
      "loss": 0.2112,
      "num_input_tokens_seen": 65125664,
      "step": 33840
    },
    {
      "epoch": 5.521249694102292,
      "grad_norm": 0.007416997104883194,
      "learning_rate": 0.017193390125008905,
      "loss": 0.1507,
      "num_input_tokens_seen": 65135888,
      "step": 33845
    },
    {
      "epoch": 5.522065421323109,
      "grad_norm": 0.0016778185963630676,
      "learning_rate": 0.017166017088762153,
      "loss": 0.149,
      "num_input_tokens_seen": 65146208,
      "step": 33850
    },
    {
      "epoch": 5.522881148543927,
      "grad_norm": 0.010840395465493202,
      "learning_rate": 0.017138664537184878,
      "loss": 0.2634,
      "num_input_tokens_seen": 65156512,
      "step": 33855
    },
    {
      "epoch": 5.523696875764744,
      "grad_norm": 0.00436149537563324,
      "learning_rate": 0.017111332474495172,
      "loss": 0.1619,
      "num_input_tokens_seen": 65166656,
      "step": 33860
    },
    {
      "epoch": 5.524512602985562,
      "grad_norm": 0.003390007419511676,
      "learning_rate": 0.017084020904907998,
      "loss": 0.3039,
      "num_input_tokens_seen": 65175664,
      "step": 33865
    },
    {
      "epoch": 5.525328330206379,
      "grad_norm": 0.0047015040181577206,
      "learning_rate": 0.017056729832635103,
      "loss": 0.223,
      "num_input_tokens_seen": 65186400,
      "step": 33870
    },
    {
      "epoch": 5.526144057427197,
      "grad_norm": 0.0026123900897800922,
      "learning_rate": 0.017029459261885153,
      "loss": 0.18,
      "num_input_tokens_seen": 65195664,
      "step": 33875
    },
    {
      "epoch": 5.526959784648014,
      "grad_norm": 0.0030644021462649107,
      "learning_rate": 0.01700220919686359,
      "loss": 0.181,
      "num_input_tokens_seen": 65205408,
      "step": 33880
    },
    {
      "epoch": 5.527775511868831,
      "grad_norm": 0.0026062799151986837,
      "learning_rate": 0.016974979641772723,
      "loss": 0.1766,
      "num_input_tokens_seen": 65216448,
      "step": 33885
    },
    {
      "epoch": 5.528591239089648,
      "grad_norm": 0.002292163437232375,
      "learning_rate": 0.01694777060081169,
      "loss": 0.1517,
      "num_input_tokens_seen": 65226480,
      "step": 33890
    },
    {
      "epoch": 5.529406966310466,
      "grad_norm": 0.0029395120218396187,
      "learning_rate": 0.016920582078176444,
      "loss": 0.2423,
      "num_input_tokens_seen": 65235328,
      "step": 33895
    },
    {
      "epoch": 5.530222693531283,
      "grad_norm": 0.0039231739938259125,
      "learning_rate": 0.016893414078059863,
      "loss": 0.1909,
      "num_input_tokens_seen": 65243792,
      "step": 33900
    },
    {
      "epoch": 5.5310384207521,
      "grad_norm": 0.0041934289038181305,
      "learning_rate": 0.016866266604651535,
      "loss": 0.2656,
      "num_input_tokens_seen": 65254624,
      "step": 33905
    },
    {
      "epoch": 5.5318541479729175,
      "grad_norm": 0.0025432901456952095,
      "learning_rate": 0.016839139662137976,
      "loss": 0.1588,
      "num_input_tokens_seen": 65263872,
      "step": 33910
    },
    {
      "epoch": 5.5326698751937355,
      "grad_norm": 0.001654882333241403,
      "learning_rate": 0.01681203325470245,
      "loss": 0.134,
      "num_input_tokens_seen": 65272480,
      "step": 33915
    },
    {
      "epoch": 5.533485602414553,
      "grad_norm": 0.003817860269919038,
      "learning_rate": 0.016784947386525157,
      "loss": 0.1356,
      "num_input_tokens_seen": 65281328,
      "step": 33920
    },
    {
      "epoch": 5.53430132963537,
      "grad_norm": 0.002588095609098673,
      "learning_rate": 0.01675788206178308,
      "loss": 0.1623,
      "num_input_tokens_seen": 65291344,
      "step": 33925
    },
    {
      "epoch": 5.535117056856187,
      "grad_norm": 0.0027352911420166492,
      "learning_rate": 0.016730837284649986,
      "loss": 0.1638,
      "num_input_tokens_seen": 65300304,
      "step": 33930
    },
    {
      "epoch": 5.535932784077005,
      "grad_norm": 0.0019326445180922747,
      "learning_rate": 0.016703813059296583,
      "loss": 0.1708,
      "num_input_tokens_seen": 65310272,
      "step": 33935
    },
    {
      "epoch": 5.536748511297822,
      "grad_norm": 0.0044778515584766865,
      "learning_rate": 0.016676809389890294,
      "loss": 0.1629,
      "num_input_tokens_seen": 65320928,
      "step": 33940
    },
    {
      "epoch": 5.537564238518639,
      "grad_norm": 0.0037234011106193066,
      "learning_rate": 0.016649826280595435,
      "loss": 0.261,
      "num_input_tokens_seen": 65330592,
      "step": 33945
    },
    {
      "epoch": 5.538379965739456,
      "grad_norm": 0.003410469042137265,
      "learning_rate": 0.016622863735573163,
      "loss": 0.1913,
      "num_input_tokens_seen": 65341744,
      "step": 33950
    },
    {
      "epoch": 5.539195692960274,
      "grad_norm": 0.001964805880561471,
      "learning_rate": 0.016595921758981395,
      "loss": 0.162,
      "num_input_tokens_seen": 65349856,
      "step": 33955
    },
    {
      "epoch": 5.5400114201810915,
      "grad_norm": 0.0059081618674099445,
      "learning_rate": 0.01656900035497495,
      "loss": 0.2038,
      "num_input_tokens_seen": 65357968,
      "step": 33960
    },
    {
      "epoch": 5.540827147401909,
      "grad_norm": 0.0038956268690526485,
      "learning_rate": 0.016542099527705485,
      "loss": 0.16,
      "num_input_tokens_seen": 65368368,
      "step": 33965
    },
    {
      "epoch": 5.541642874622726,
      "grad_norm": 0.0029001410584896803,
      "learning_rate": 0.01651521928132138,
      "loss": 0.1846,
      "num_input_tokens_seen": 65378064,
      "step": 33970
    },
    {
      "epoch": 5.542458601843544,
      "grad_norm": 0.002155616879463196,
      "learning_rate": 0.01648835961996794,
      "loss": 0.197,
      "num_input_tokens_seen": 65388016,
      "step": 33975
    },
    {
      "epoch": 5.543274329064361,
      "grad_norm": 0.0016005157958716154,
      "learning_rate": 0.016461520547787285,
      "loss": 0.1235,
      "num_input_tokens_seen": 65399056,
      "step": 33980
    },
    {
      "epoch": 5.544090056285178,
      "grad_norm": 0.0035117578227072954,
      "learning_rate": 0.016434702068918266,
      "loss": 0.1819,
      "num_input_tokens_seen": 65408576,
      "step": 33985
    },
    {
      "epoch": 5.544905783505996,
      "grad_norm": 0.003650801256299019,
      "learning_rate": 0.01640790418749673,
      "loss": 0.1717,
      "num_input_tokens_seen": 65419584,
      "step": 33990
    },
    {
      "epoch": 5.545721510726813,
      "grad_norm": 0.003040257142856717,
      "learning_rate": 0.016381126907655134,
      "loss": 0.1914,
      "num_input_tokens_seen": 65429952,
      "step": 33995
    },
    {
      "epoch": 5.54653723794763,
      "grad_norm": 0.0022585673723369837,
      "learning_rate": 0.016354370233522948,
      "loss": 0.1649,
      "num_input_tokens_seen": 65439360,
      "step": 34000
    },
    {
      "epoch": 5.54653723794763,
      "eval_loss": 0.18872585892677307,
      "eval_runtime": 68.0781,
      "eval_samples_per_second": 40.028,
      "eval_steps_per_second": 20.021,
      "num_input_tokens_seen": 65439360,
      "step": 34000
    },
    {
      "epoch": 5.5473529651684474,
      "grad_norm": 0.0031643102411180735,
      "learning_rate": 0.016327634169226394,
      "loss": 0.212,
      "num_input_tokens_seen": 65449056,
      "step": 34005
    },
    {
      "epoch": 5.548168692389265,
      "grad_norm": 0.008999790996313095,
      "learning_rate": 0.016300918718888485,
      "loss": 0.1887,
      "num_input_tokens_seen": 65459152,
      "step": 34010
    },
    {
      "epoch": 5.548984419610083,
      "grad_norm": 0.0016845609061419964,
      "learning_rate": 0.016274223886629052,
      "loss": 0.1965,
      "num_input_tokens_seen": 65468800,
      "step": 34015
    },
    {
      "epoch": 5.5498001468309,
      "grad_norm": 0.003212540876120329,
      "learning_rate": 0.01624754967656482,
      "loss": 0.1599,
      "num_input_tokens_seen": 65477872,
      "step": 34020
    },
    {
      "epoch": 5.550615874051717,
      "grad_norm": 0.004940592683851719,
      "learning_rate": 0.016220896092809235,
      "loss": 0.232,
      "num_input_tokens_seen": 65486512,
      "step": 34025
    },
    {
      "epoch": 5.551431601272535,
      "grad_norm": 0.0022265007719397545,
      "learning_rate": 0.01619426313947267,
      "loss": 0.1925,
      "num_input_tokens_seen": 65494816,
      "step": 34030
    },
    {
      "epoch": 5.552247328493352,
      "grad_norm": 0.003178234677761793,
      "learning_rate": 0.016167650820662228,
      "loss": 0.1788,
      "num_input_tokens_seen": 65504528,
      "step": 34035
    },
    {
      "epoch": 5.553063055714169,
      "grad_norm": 0.004552691709250212,
      "learning_rate": 0.016141059140481855,
      "loss": 0.1504,
      "num_input_tokens_seen": 65514032,
      "step": 34040
    },
    {
      "epoch": 5.553878782934986,
      "grad_norm": 0.002242104848846793,
      "learning_rate": 0.016114488103032374,
      "loss": 0.1549,
      "num_input_tokens_seen": 65523632,
      "step": 34045
    },
    {
      "epoch": 5.554694510155803,
      "grad_norm": 0.0015473061939701438,
      "learning_rate": 0.016087937712411293,
      "loss": 0.0883,
      "num_input_tokens_seen": 65532800,
      "step": 34050
    },
    {
      "epoch": 5.555510237376621,
      "grad_norm": 0.0013678541872650385,
      "learning_rate": 0.01606140797271308,
      "loss": 0.1828,
      "num_input_tokens_seen": 65541712,
      "step": 34055
    },
    {
      "epoch": 5.5563259645974385,
      "grad_norm": 0.003984248265624046,
      "learning_rate": 0.01603489888802897,
      "loss": 0.206,
      "num_input_tokens_seen": 65551968,
      "step": 34060
    },
    {
      "epoch": 5.557141691818256,
      "grad_norm": 0.0019332917872816324,
      "learning_rate": 0.016008410462446918,
      "loss": 0.1678,
      "num_input_tokens_seen": 65562880,
      "step": 34065
    },
    {
      "epoch": 5.557957419039074,
      "grad_norm": 0.0028456589207053185,
      "learning_rate": 0.01598194270005185,
      "loss": 0.1727,
      "num_input_tokens_seen": 65573552,
      "step": 34070
    },
    {
      "epoch": 5.558773146259891,
      "grad_norm": 0.005325858015567064,
      "learning_rate": 0.015955495604925356,
      "loss": 0.1711,
      "num_input_tokens_seen": 65582688,
      "step": 34075
    },
    {
      "epoch": 5.559588873480708,
      "grad_norm": 0.00564125832170248,
      "learning_rate": 0.01592906918114598,
      "loss": 0.2759,
      "num_input_tokens_seen": 65592320,
      "step": 34080
    },
    {
      "epoch": 5.560404600701525,
      "grad_norm": 0.005730916280299425,
      "learning_rate": 0.015902663432788965,
      "loss": 0.1812,
      "num_input_tokens_seen": 65602112,
      "step": 34085
    },
    {
      "epoch": 5.561220327922343,
      "grad_norm": 0.002508897799998522,
      "learning_rate": 0.01587627836392643,
      "loss": 0.2085,
      "num_input_tokens_seen": 65611328,
      "step": 34090
    },
    {
      "epoch": 5.56203605514316,
      "grad_norm": 0.0019283012952655554,
      "learning_rate": 0.01584991397862726,
      "loss": 0.1335,
      "num_input_tokens_seen": 65620992,
      "step": 34095
    },
    {
      "epoch": 5.562851782363977,
      "grad_norm": 0.0028164463583379984,
      "learning_rate": 0.015823570280957214,
      "loss": 0.1734,
      "num_input_tokens_seen": 65631200,
      "step": 34100
    },
    {
      "epoch": 5.5636675095847945,
      "grad_norm": 0.002851475728675723,
      "learning_rate": 0.015797247274978766,
      "loss": 0.1302,
      "num_input_tokens_seen": 65639168,
      "step": 34105
    },
    {
      "epoch": 5.5644832368056125,
      "grad_norm": 0.0030725672841072083,
      "learning_rate": 0.015770944964751326,
      "loss": 0.1641,
      "num_input_tokens_seen": 65647872,
      "step": 34110
    },
    {
      "epoch": 5.56529896402643,
      "grad_norm": 0.002682802267372608,
      "learning_rate": 0.015744663354330956,
      "loss": 0.1389,
      "num_input_tokens_seen": 65656096,
      "step": 34115
    },
    {
      "epoch": 5.566114691247247,
      "grad_norm": 0.0037190027069300413,
      "learning_rate": 0.015718402447770664,
      "loss": 0.2068,
      "num_input_tokens_seen": 65666528,
      "step": 34120
    },
    {
      "epoch": 5.566930418468064,
      "grad_norm": 0.0027030876372009516,
      "learning_rate": 0.015692162249120224,
      "loss": 0.1636,
      "num_input_tokens_seen": 65676240,
      "step": 34125
    },
    {
      "epoch": 5.567746145688882,
      "grad_norm": 0.005384210031479597,
      "learning_rate": 0.01566594276242615,
      "loss": 0.1908,
      "num_input_tokens_seen": 65685424,
      "step": 34130
    },
    {
      "epoch": 5.568561872909699,
      "grad_norm": 0.001662351656705141,
      "learning_rate": 0.015639743991731857,
      "loss": 0.1691,
      "num_input_tokens_seen": 65694768,
      "step": 34135
    },
    {
      "epoch": 5.569377600130516,
      "grad_norm": 0.003502328647300601,
      "learning_rate": 0.01561356594107755,
      "loss": 0.1592,
      "num_input_tokens_seen": 65704560,
      "step": 34140
    },
    {
      "epoch": 5.570193327351333,
      "grad_norm": 0.0016671806806698442,
      "learning_rate": 0.015587408614500147,
      "loss": 0.1403,
      "num_input_tokens_seen": 65714336,
      "step": 34145
    },
    {
      "epoch": 5.571009054572151,
      "grad_norm": 0.003510760609060526,
      "learning_rate": 0.015561272016033505,
      "loss": 0.1808,
      "num_input_tokens_seen": 65724176,
      "step": 34150
    },
    {
      "epoch": 5.5718247817929685,
      "grad_norm": 0.003983987495303154,
      "learning_rate": 0.015535156149708167,
      "loss": 0.2066,
      "num_input_tokens_seen": 65734048,
      "step": 34155
    },
    {
      "epoch": 5.572640509013786,
      "grad_norm": 0.002968488959595561,
      "learning_rate": 0.015509061019551528,
      "loss": 0.1686,
      "num_input_tokens_seen": 65743312,
      "step": 34160
    },
    {
      "epoch": 5.573456236234604,
      "grad_norm": 0.004216658417135477,
      "learning_rate": 0.015482986629587818,
      "loss": 0.1461,
      "num_input_tokens_seen": 65752368,
      "step": 34165
    },
    {
      "epoch": 5.574271963455421,
      "grad_norm": 0.004326097201555967,
      "learning_rate": 0.01545693298383799,
      "loss": 0.1831,
      "num_input_tokens_seen": 65761696,
      "step": 34170
    },
    {
      "epoch": 5.575087690676238,
      "grad_norm": 0.003315136069431901,
      "learning_rate": 0.015430900086319858,
      "loss": 0.1776,
      "num_input_tokens_seen": 65771904,
      "step": 34175
    },
    {
      "epoch": 5.575903417897055,
      "grad_norm": 0.005800656974315643,
      "learning_rate": 0.015404887941048084,
      "loss": 0.2018,
      "num_input_tokens_seen": 65780624,
      "step": 34180
    },
    {
      "epoch": 5.576719145117872,
      "grad_norm": 0.003022665623575449,
      "learning_rate": 0.01537889655203397,
      "loss": 0.1772,
      "num_input_tokens_seen": 65790640,
      "step": 34185
    },
    {
      "epoch": 5.57753487233869,
      "grad_norm": 0.0029796112794429064,
      "learning_rate": 0.015352925923285798,
      "loss": 0.1701,
      "num_input_tokens_seen": 65800976,
      "step": 34190
    },
    {
      "epoch": 5.578350599559507,
      "grad_norm": 0.006038009654730558,
      "learning_rate": 0.015326976058808511,
      "loss": 0.1428,
      "num_input_tokens_seen": 65810448,
      "step": 34195
    },
    {
      "epoch": 5.579166326780324,
      "grad_norm": 0.00174055527895689,
      "learning_rate": 0.015301046962603908,
      "loss": 0.1103,
      "num_input_tokens_seen": 65819600,
      "step": 34200
    },
    {
      "epoch": 5.579166326780324,
      "eval_loss": 0.1893654465675354,
      "eval_runtime": 68.0216,
      "eval_samples_per_second": 40.061,
      "eval_steps_per_second": 20.038,
      "num_input_tokens_seen": 65819600,
      "step": 34200
    },
    {
      "epoch": 5.5799820540011424,
      "grad_norm": 0.004720044322311878,
      "learning_rate": 0.015275138638670626,
      "loss": 0.1536,
      "num_input_tokens_seen": 65829056,
      "step": 34205
    },
    {
      "epoch": 5.58079778122196,
      "grad_norm": 0.003349606180563569,
      "learning_rate": 0.015249251091004001,
      "loss": 0.2211,
      "num_input_tokens_seen": 65839104,
      "step": 34210
    },
    {
      "epoch": 5.581613508442777,
      "grad_norm": 0.0022992901504039764,
      "learning_rate": 0.01522338432359624,
      "loss": 0.1545,
      "num_input_tokens_seen": 65848048,
      "step": 34215
    },
    {
      "epoch": 5.582429235663594,
      "grad_norm": 0.002166653983294964,
      "learning_rate": 0.01519753834043635,
      "loss": 0.1626,
      "num_input_tokens_seen": 65857312,
      "step": 34220
    },
    {
      "epoch": 5.583244962884411,
      "grad_norm": 0.003924970049411058,
      "learning_rate": 0.015171713145510095,
      "loss": 0.1472,
      "num_input_tokens_seen": 65867664,
      "step": 34225
    },
    {
      "epoch": 5.584060690105229,
      "grad_norm": 0.004499494563788176,
      "learning_rate": 0.01514590874279999,
      "loss": 0.1635,
      "num_input_tokens_seen": 65877312,
      "step": 34230
    },
    {
      "epoch": 5.584876417326046,
      "grad_norm": 0.002246303018182516,
      "learning_rate": 0.015120125136285467,
      "loss": 0.1654,
      "num_input_tokens_seen": 65886320,
      "step": 34235
    },
    {
      "epoch": 5.585692144546863,
      "grad_norm": 0.007255949079990387,
      "learning_rate": 0.015094362329942629,
      "loss": 0.2157,
      "num_input_tokens_seen": 65897552,
      "step": 34240
    },
    {
      "epoch": 5.586507871767681,
      "grad_norm": 0.0028727767057716846,
      "learning_rate": 0.01506862032774448,
      "loss": 0.1615,
      "num_input_tokens_seen": 65907616,
      "step": 34245
    },
    {
      "epoch": 5.587323598988498,
      "grad_norm": 0.0020282957702875137,
      "learning_rate": 0.015042899133660697,
      "loss": 0.1968,
      "num_input_tokens_seen": 65916528,
      "step": 34250
    },
    {
      "epoch": 5.5881393262093155,
      "grad_norm": 0.004785628989338875,
      "learning_rate": 0.01501719875165789,
      "loss": 0.15,
      "num_input_tokens_seen": 65925696,
      "step": 34255
    },
    {
      "epoch": 5.588955053430133,
      "grad_norm": 0.004425426479429007,
      "learning_rate": 0.014991519185699286,
      "loss": 0.1465,
      "num_input_tokens_seen": 65935888,
      "step": 34260
    },
    {
      "epoch": 5.589770780650951,
      "grad_norm": 0.0019919872283935547,
      "learning_rate": 0.014965860439745054,
      "loss": 0.2194,
      "num_input_tokens_seen": 65943296,
      "step": 34265
    },
    {
      "epoch": 5.590586507871768,
      "grad_norm": 0.00470799021422863,
      "learning_rate": 0.01494022251775211,
      "loss": 0.146,
      "num_input_tokens_seen": 65953344,
      "step": 34270
    },
    {
      "epoch": 5.591402235092585,
      "grad_norm": 0.002671673893928528,
      "learning_rate": 0.014914605423674109,
      "loss": 0.1431,
      "num_input_tokens_seen": 65962736,
      "step": 34275
    },
    {
      "epoch": 5.592217962313402,
      "grad_norm": 0.0030314517207443714,
      "learning_rate": 0.014889009161461525,
      "loss": 0.1772,
      "num_input_tokens_seen": 65971808,
      "step": 34280
    },
    {
      "epoch": 5.59303368953422,
      "grad_norm": 0.003410706529393792,
      "learning_rate": 0.014863433735061665,
      "loss": 0.1578,
      "num_input_tokens_seen": 65980784,
      "step": 34285
    },
    {
      "epoch": 5.593849416755037,
      "grad_norm": 0.0022576414048671722,
      "learning_rate": 0.014837879148418541,
      "loss": 0.18,
      "num_input_tokens_seen": 65991104,
      "step": 34290
    },
    {
      "epoch": 5.594665143975854,
      "grad_norm": 0.005815650336444378,
      "learning_rate": 0.01481234540547302,
      "loss": 0.1507,
      "num_input_tokens_seen": 66000288,
      "step": 34295
    },
    {
      "epoch": 5.5954808711966715,
      "grad_norm": 0.005142908543348312,
      "learning_rate": 0.014786832510162717,
      "loss": 0.151,
      "num_input_tokens_seen": 66009760,
      "step": 34300
    },
    {
      "epoch": 5.5962965984174895,
      "grad_norm": 0.0034648000728338957,
      "learning_rate": 0.014761340466422017,
      "loss": 0.1397,
      "num_input_tokens_seen": 66019488,
      "step": 34305
    },
    {
      "epoch": 5.597112325638307,
      "grad_norm": 0.006020873319357634,
      "learning_rate": 0.014735869278182144,
      "loss": 0.1836,
      "num_input_tokens_seen": 66028656,
      "step": 34310
    },
    {
      "epoch": 5.597928052859124,
      "grad_norm": 0.0022838388103991747,
      "learning_rate": 0.014710418949371057,
      "loss": 0.1926,
      "num_input_tokens_seen": 66037616,
      "step": 34315
    },
    {
      "epoch": 5.598743780079941,
      "grad_norm": 0.002853894140571356,
      "learning_rate": 0.014684989483913495,
      "loss": 0.1704,
      "num_input_tokens_seen": 66048224,
      "step": 34320
    },
    {
      "epoch": 5.599559507300759,
      "grad_norm": 0.0024459518026560545,
      "learning_rate": 0.014659580885731077,
      "loss": 0.1881,
      "num_input_tokens_seen": 66056656,
      "step": 34325
    },
    {
      "epoch": 5.600375234521576,
      "grad_norm": 0.006344488821923733,
      "learning_rate": 0.014634193158742047,
      "loss": 0.2143,
      "num_input_tokens_seen": 66067040,
      "step": 34330
    },
    {
      "epoch": 5.601190961742393,
      "grad_norm": 0.00424759229645133,
      "learning_rate": 0.014608826306861576,
      "loss": 0.2499,
      "num_input_tokens_seen": 66076816,
      "step": 34335
    },
    {
      "epoch": 5.602006688963211,
      "grad_norm": 0.0018562766490504146,
      "learning_rate": 0.014583480334001486,
      "loss": 0.2042,
      "num_input_tokens_seen": 66085328,
      "step": 34340
    },
    {
      "epoch": 5.602822416184028,
      "grad_norm": 0.004272765479981899,
      "learning_rate": 0.014558155244070496,
      "loss": 0.2034,
      "num_input_tokens_seen": 66095408,
      "step": 34345
    },
    {
      "epoch": 5.6036381434048455,
      "grad_norm": 0.0021157660521566868,
      "learning_rate": 0.014532851040974036,
      "loss": 0.1369,
      "num_input_tokens_seen": 66104544,
      "step": 34350
    },
    {
      "epoch": 5.604453870625663,
      "grad_norm": 0.0035447762347757816,
      "learning_rate": 0.014507567728614335,
      "loss": 0.1583,
      "num_input_tokens_seen": 66113792,
      "step": 34355
    },
    {
      "epoch": 5.60526959784648,
      "grad_norm": 0.003825337626039982,
      "learning_rate": 0.01448230531089037,
      "loss": 0.1397,
      "num_input_tokens_seen": 66122992,
      "step": 34360
    },
    {
      "epoch": 5.606085325067298,
      "grad_norm": 0.0062892744317650795,
      "learning_rate": 0.014457063791697993,
      "loss": 0.2592,
      "num_input_tokens_seen": 66132992,
      "step": 34365
    },
    {
      "epoch": 5.606901052288115,
      "grad_norm": 0.00382571411319077,
      "learning_rate": 0.01443184317492971,
      "loss": 0.174,
      "num_input_tokens_seen": 66142992,
      "step": 34370
    },
    {
      "epoch": 5.607716779508932,
      "grad_norm": 0.004044561181217432,
      "learning_rate": 0.014406643464474822,
      "loss": 0.1944,
      "num_input_tokens_seen": 66151680,
      "step": 34375
    },
    {
      "epoch": 5.60853250672975,
      "grad_norm": 0.003098484128713608,
      "learning_rate": 0.014381464664219539,
      "loss": 0.1921,
      "num_input_tokens_seen": 66161232,
      "step": 34380
    },
    {
      "epoch": 5.609348233950567,
      "grad_norm": 0.0033849102910608053,
      "learning_rate": 0.014356306778046656,
      "loss": 0.1379,
      "num_input_tokens_seen": 66171040,
      "step": 34385
    },
    {
      "epoch": 5.610163961171384,
      "grad_norm": 0.0018566444050520658,
      "learning_rate": 0.014331169809835885,
      "loss": 0.1449,
      "num_input_tokens_seen": 66181152,
      "step": 34390
    },
    {
      "epoch": 5.610979688392201,
      "grad_norm": 0.0038907427806407213,
      "learning_rate": 0.014306053763463644,
      "loss": 0.2,
      "num_input_tokens_seen": 66190528,
      "step": 34395
    },
    {
      "epoch": 5.6117954156130185,
      "grad_norm": 0.0027450541965663433,
      "learning_rate": 0.014280958642803147,
      "loss": 0.1197,
      "num_input_tokens_seen": 66199376,
      "step": 34400
    },
    {
      "epoch": 5.6117954156130185,
      "eval_loss": 0.18970376253128052,
      "eval_runtime": 68.0289,
      "eval_samples_per_second": 40.056,
      "eval_steps_per_second": 20.036,
      "num_input_tokens_seen": 66199376,
      "step": 34400
    },
    {
      "epoch": 5.6126111428338366,
      "grad_norm": 0.003530115121975541,
      "learning_rate": 0.014255884451724404,
      "loss": 0.2127,
      "num_input_tokens_seen": 66208752,
      "step": 34405
    },
    {
      "epoch": 5.613426870054654,
      "grad_norm": 0.0039468789473176,
      "learning_rate": 0.014230831194094101,
      "loss": 0.1172,
      "num_input_tokens_seen": 66217152,
      "step": 34410
    },
    {
      "epoch": 5.614242597275471,
      "grad_norm": 0.0027480325661599636,
      "learning_rate": 0.014205798873775865,
      "loss": 0.2146,
      "num_input_tokens_seen": 66227520,
      "step": 34415
    },
    {
      "epoch": 5.615058324496289,
      "grad_norm": 0.002390847774222493,
      "learning_rate": 0.014180787494629893,
      "loss": 0.1513,
      "num_input_tokens_seen": 66236816,
      "step": 34420
    },
    {
      "epoch": 5.615874051717106,
      "grad_norm": 0.0012988448143005371,
      "learning_rate": 0.014155797060513314,
      "loss": 0.1531,
      "num_input_tokens_seen": 66246416,
      "step": 34425
    },
    {
      "epoch": 5.616689778937923,
      "grad_norm": 0.002856618259102106,
      "learning_rate": 0.014130827575279963,
      "loss": 0.2345,
      "num_input_tokens_seen": 66255920,
      "step": 34430
    },
    {
      "epoch": 5.61750550615874,
      "grad_norm": 0.002402500482276082,
      "learning_rate": 0.014105879042780427,
      "loss": 0.2127,
      "num_input_tokens_seen": 66265856,
      "step": 34435
    },
    {
      "epoch": 5.618321233379558,
      "grad_norm": 0.0019351673545315862,
      "learning_rate": 0.014080951466862113,
      "loss": 0.1713,
      "num_input_tokens_seen": 66274656,
      "step": 34440
    },
    {
      "epoch": 5.619136960600375,
      "grad_norm": 0.001994193997234106,
      "learning_rate": 0.014056044851369126,
      "loss": 0.2423,
      "num_input_tokens_seen": 66285424,
      "step": 34445
    },
    {
      "epoch": 5.6199526878211925,
      "grad_norm": 0.0038000415079295635,
      "learning_rate": 0.014031159200142428,
      "loss": 0.1844,
      "num_input_tokens_seen": 66295456,
      "step": 34450
    },
    {
      "epoch": 5.62076841504201,
      "grad_norm": 0.00475184665992856,
      "learning_rate": 0.014006294517019667,
      "loss": 0.164,
      "num_input_tokens_seen": 66304928,
      "step": 34455
    },
    {
      "epoch": 5.621584142262828,
      "grad_norm": 0.002273501129820943,
      "learning_rate": 0.013981450805835276,
      "loss": 0.1543,
      "num_input_tokens_seen": 66314000,
      "step": 34460
    },
    {
      "epoch": 5.622399869483645,
      "grad_norm": 0.003197896294295788,
      "learning_rate": 0.01395662807042049,
      "loss": 0.1835,
      "num_input_tokens_seen": 66324448,
      "step": 34465
    },
    {
      "epoch": 5.623215596704462,
      "grad_norm": 0.002484917873516679,
      "learning_rate": 0.013931826314603296,
      "loss": 0.1866,
      "num_input_tokens_seen": 66334256,
      "step": 34470
    },
    {
      "epoch": 5.624031323925279,
      "grad_norm": 0.004829803481698036,
      "learning_rate": 0.013907045542208401,
      "loss": 0.1914,
      "num_input_tokens_seen": 66343728,
      "step": 34475
    },
    {
      "epoch": 5.624847051146097,
      "grad_norm": 0.002726196777075529,
      "learning_rate": 0.013882285757057333,
      "loss": 0.1644,
      "num_input_tokens_seen": 66354800,
      "step": 34480
    },
    {
      "epoch": 5.625662778366914,
      "grad_norm": 0.0028979156631976366,
      "learning_rate": 0.013857546962968403,
      "loss": 0.1289,
      "num_input_tokens_seen": 66364256,
      "step": 34485
    },
    {
      "epoch": 5.626478505587731,
      "grad_norm": 0.004740218166261911,
      "learning_rate": 0.013832829163756577,
      "loss": 0.1523,
      "num_input_tokens_seen": 66371792,
      "step": 34490
    },
    {
      "epoch": 5.6272942328085485,
      "grad_norm": 0.00173526955768466,
      "learning_rate": 0.013808132363233689,
      "loss": 0.1715,
      "num_input_tokens_seen": 66381760,
      "step": 34495
    },
    {
      "epoch": 5.6281099600293665,
      "grad_norm": 0.007723728194832802,
      "learning_rate": 0.013783456565208256,
      "loss": 0.2168,
      "num_input_tokens_seen": 66391184,
      "step": 34500
    },
    {
      "epoch": 5.628925687250184,
      "grad_norm": 0.006200533825904131,
      "learning_rate": 0.01375880177348564,
      "loss": 0.1944,
      "num_input_tokens_seen": 66400816,
      "step": 34505
    },
    {
      "epoch": 5.629741414471001,
      "grad_norm": 0.00169724365696311,
      "learning_rate": 0.013734167991867928,
      "loss": 0.1412,
      "num_input_tokens_seen": 66408848,
      "step": 34510
    },
    {
      "epoch": 5.630557141691818,
      "grad_norm": 0.004409611225128174,
      "learning_rate": 0.013709555224153935,
      "loss": 0.2293,
      "num_input_tokens_seen": 66417792,
      "step": 34515
    },
    {
      "epoch": 5.631372868912636,
      "grad_norm": 0.002676913980394602,
      "learning_rate": 0.013684963474139222,
      "loss": 0.1468,
      "num_input_tokens_seen": 66427408,
      "step": 34520
    },
    {
      "epoch": 5.632188596133453,
      "grad_norm": 0.005794290453195572,
      "learning_rate": 0.013660392745616224,
      "loss": 0.16,
      "num_input_tokens_seen": 66437824,
      "step": 34525
    },
    {
      "epoch": 5.63300432335427,
      "grad_norm": 0.0037266716826707125,
      "learning_rate": 0.013635843042373974,
      "loss": 0.1852,
      "num_input_tokens_seen": 66446448,
      "step": 34530
    },
    {
      "epoch": 5.633820050575087,
      "grad_norm": 0.0024873423390090466,
      "learning_rate": 0.01361131436819843,
      "loss": 0.2375,
      "num_input_tokens_seen": 66457488,
      "step": 34535
    },
    {
      "epoch": 5.634635777795905,
      "grad_norm": 0.0029273591935634613,
      "learning_rate": 0.013586806726872147,
      "loss": 0.1534,
      "num_input_tokens_seen": 66467744,
      "step": 34540
    },
    {
      "epoch": 5.635451505016722,
      "grad_norm": 0.0026592833455652,
      "learning_rate": 0.013562320122174537,
      "loss": 0.1634,
      "num_input_tokens_seen": 66477776,
      "step": 34545
    },
    {
      "epoch": 5.63626723223754,
      "grad_norm": 0.0024218223989009857,
      "learning_rate": 0.013537854557881762,
      "loss": 0.1654,
      "num_input_tokens_seen": 66488608,
      "step": 34550
    },
    {
      "epoch": 5.637082959458358,
      "grad_norm": 0.0018762145191431046,
      "learning_rate": 0.013513410037766687,
      "loss": 0.169,
      "num_input_tokens_seen": 66498544,
      "step": 34555
    },
    {
      "epoch": 5.637898686679175,
      "grad_norm": 0.00313694030046463,
      "learning_rate": 0.013488986565598998,
      "loss": 0.15,
      "num_input_tokens_seen": 66507776,
      "step": 34560
    },
    {
      "epoch": 5.638714413899992,
      "grad_norm": 0.0036284204106777906,
      "learning_rate": 0.013464584145145097,
      "loss": 0.1848,
      "num_input_tokens_seen": 66517792,
      "step": 34565
    },
    {
      "epoch": 5.639530141120809,
      "grad_norm": 0.0026585476007312536,
      "learning_rate": 0.013440202780168109,
      "loss": 0.1674,
      "num_input_tokens_seen": 66528128,
      "step": 34570
    },
    {
      "epoch": 5.640345868341626,
      "grad_norm": 0.0034350233618170023,
      "learning_rate": 0.01341584247442799,
      "loss": 0.1982,
      "num_input_tokens_seen": 66537984,
      "step": 34575
    },
    {
      "epoch": 5.641161595562444,
      "grad_norm": 0.003310931846499443,
      "learning_rate": 0.013391503231681355,
      "loss": 0.1115,
      "num_input_tokens_seen": 66547904,
      "step": 34580
    },
    {
      "epoch": 5.641977322783261,
      "grad_norm": 0.0021771711762994528,
      "learning_rate": 0.013367185055681685,
      "loss": 0.1909,
      "num_input_tokens_seen": 66556720,
      "step": 34585
    },
    {
      "epoch": 5.642793050004078,
      "grad_norm": 0.003439304418861866,
      "learning_rate": 0.013342887950179095,
      "loss": 0.1818,
      "num_input_tokens_seen": 66565488,
      "step": 34590
    },
    {
      "epoch": 5.643608777224896,
      "grad_norm": 0.00097986392211169,
      "learning_rate": 0.013318611918920554,
      "loss": 0.1459,
      "num_input_tokens_seen": 66574256,
      "step": 34595
    },
    {
      "epoch": 5.6444245044457135,
      "grad_norm": 0.0028859456069767475,
      "learning_rate": 0.01329435696564965,
      "loss": 0.1848,
      "num_input_tokens_seen": 66583936,
      "step": 34600
    },
    {
      "epoch": 5.6444245044457135,
      "eval_loss": 0.1898750364780426,
      "eval_runtime": 68.1353,
      "eval_samples_per_second": 39.994,
      "eval_steps_per_second": 20.004,
      "num_input_tokens_seen": 66583936,
      "step": 34600
    },
    {
      "epoch": 5.645240231666531,
      "grad_norm": 0.0027296985499560833,
      "learning_rate": 0.013270123094106894,
      "loss": 0.1819,
      "num_input_tokens_seen": 66594032,
      "step": 34605
    },
    {
      "epoch": 5.646055958887348,
      "grad_norm": 0.0019486439414322376,
      "learning_rate": 0.013245910308029395,
      "loss": 0.1974,
      "num_input_tokens_seen": 66603968,
      "step": 34610
    },
    {
      "epoch": 5.646871686108166,
      "grad_norm": 0.005293811671435833,
      "learning_rate": 0.0132217186111511,
      "loss": 0.3134,
      "num_input_tokens_seen": 66613360,
      "step": 34615
    },
    {
      "epoch": 5.647687413328983,
      "grad_norm": 0.00456391554325819,
      "learning_rate": 0.013197548007202626,
      "loss": 0.2237,
      "num_input_tokens_seen": 66622496,
      "step": 34620
    },
    {
      "epoch": 5.6485031405498,
      "grad_norm": 0.004113181959837675,
      "learning_rate": 0.01317339849991142,
      "loss": 0.2236,
      "num_input_tokens_seen": 66632544,
      "step": 34625
    },
    {
      "epoch": 5.649318867770617,
      "grad_norm": 0.003144951304420829,
      "learning_rate": 0.013149270093001675,
      "loss": 0.19,
      "num_input_tokens_seen": 66642688,
      "step": 34630
    },
    {
      "epoch": 5.650134594991435,
      "grad_norm": 0.0029435104224830866,
      "learning_rate": 0.013125162790194227,
      "loss": 0.1391,
      "num_input_tokens_seen": 66653168,
      "step": 34635
    },
    {
      "epoch": 5.650950322212252,
      "grad_norm": 0.002833125414326787,
      "learning_rate": 0.01310107659520674,
      "loss": 0.1412,
      "num_input_tokens_seen": 66662848,
      "step": 34640
    },
    {
      "epoch": 5.6517660494330695,
      "grad_norm": 0.0029297128785401583,
      "learning_rate": 0.013077011511753655,
      "loss": 0.1787,
      "num_input_tokens_seen": 66672528,
      "step": 34645
    },
    {
      "epoch": 5.652581776653887,
      "grad_norm": 0.0027698318008333445,
      "learning_rate": 0.013052967543546056,
      "loss": 0.1688,
      "num_input_tokens_seen": 66682320,
      "step": 34650
    },
    {
      "epoch": 5.653397503874705,
      "grad_norm": 0.002668774453923106,
      "learning_rate": 0.01302894469429186,
      "loss": 0.1992,
      "num_input_tokens_seen": 66691520,
      "step": 34655
    },
    {
      "epoch": 5.654213231095522,
      "grad_norm": 0.006014504469931126,
      "learning_rate": 0.013004942967695653,
      "loss": 0.1811,
      "num_input_tokens_seen": 66700416,
      "step": 34660
    },
    {
      "epoch": 5.655028958316339,
      "grad_norm": 0.003410431556403637,
      "learning_rate": 0.012980962367458859,
      "loss": 0.1864,
      "num_input_tokens_seen": 66710224,
      "step": 34665
    },
    {
      "epoch": 5.655844685537156,
      "grad_norm": 0.00255625881254673,
      "learning_rate": 0.012957002897279567,
      "loss": 0.1711,
      "num_input_tokens_seen": 66719072,
      "step": 34670
    },
    {
      "epoch": 5.656660412757974,
      "grad_norm": 0.003972623497247696,
      "learning_rate": 0.012933064560852576,
      "loss": 0.2168,
      "num_input_tokens_seen": 66728400,
      "step": 34675
    },
    {
      "epoch": 5.657476139978791,
      "grad_norm": 0.003850200679153204,
      "learning_rate": 0.012909147361869527,
      "loss": 0.195,
      "num_input_tokens_seen": 66737824,
      "step": 34680
    },
    {
      "epoch": 5.658291867199608,
      "grad_norm": 0.003145200200378895,
      "learning_rate": 0.012885251304018774,
      "loss": 0.184,
      "num_input_tokens_seen": 66746160,
      "step": 34685
    },
    {
      "epoch": 5.659107594420425,
      "grad_norm": 0.002585712820291519,
      "learning_rate": 0.012861376390985335,
      "loss": 0.1567,
      "num_input_tokens_seen": 66756640,
      "step": 34690
    },
    {
      "epoch": 5.6599233216412435,
      "grad_norm": 0.002671591006219387,
      "learning_rate": 0.012837522626451063,
      "loss": 0.2031,
      "num_input_tokens_seen": 66766496,
      "step": 34695
    },
    {
      "epoch": 5.660739048862061,
      "grad_norm": 0.0014762167120352387,
      "learning_rate": 0.01281369001409447,
      "loss": 0.1932,
      "num_input_tokens_seen": 66775520,
      "step": 34700
    },
    {
      "epoch": 5.661554776082878,
      "grad_norm": 0.005053316708654165,
      "learning_rate": 0.012789878557590877,
      "loss": 0.1965,
      "num_input_tokens_seen": 66786624,
      "step": 34705
    },
    {
      "epoch": 5.662370503303695,
      "grad_norm": 0.00161517015658319,
      "learning_rate": 0.012766088260612334,
      "loss": 0.1738,
      "num_input_tokens_seen": 66796528,
      "step": 34710
    },
    {
      "epoch": 5.663186230524513,
      "grad_norm": 0.0035551697947084904,
      "learning_rate": 0.012742319126827523,
      "loss": 0.1685,
      "num_input_tokens_seen": 66806336,
      "step": 34715
    },
    {
      "epoch": 5.66400195774533,
      "grad_norm": 0.003992021549493074,
      "learning_rate": 0.012718571159902008,
      "loss": 0.204,
      "num_input_tokens_seen": 66815728,
      "step": 34720
    },
    {
      "epoch": 5.664817684966147,
      "grad_norm": 0.003584351157769561,
      "learning_rate": 0.01269484436349803,
      "loss": 0.1541,
      "num_input_tokens_seen": 66825392,
      "step": 34725
    },
    {
      "epoch": 5.665633412186965,
      "grad_norm": 0.0032921612728387117,
      "learning_rate": 0.012671138741274528,
      "loss": 0.1457,
      "num_input_tokens_seen": 66834624,
      "step": 34730
    },
    {
      "epoch": 5.666449139407782,
      "grad_norm": 0.00413886783644557,
      "learning_rate": 0.012647454296887194,
      "loss": 0.1421,
      "num_input_tokens_seen": 66843856,
      "step": 34735
    },
    {
      "epoch": 5.667264866628599,
      "grad_norm": 0.0030138532165437937,
      "learning_rate": 0.012623791033988507,
      "loss": 0.1774,
      "num_input_tokens_seen": 66853040,
      "step": 34740
    },
    {
      "epoch": 5.6680805938494165,
      "grad_norm": 0.004286070354282856,
      "learning_rate": 0.012600148956227597,
      "loss": 0.1721,
      "num_input_tokens_seen": 66862240,
      "step": 34745
    },
    {
      "epoch": 5.668896321070234,
      "grad_norm": 0.002043703570961952,
      "learning_rate": 0.012576528067250414,
      "loss": 0.1643,
      "num_input_tokens_seen": 66872128,
      "step": 34750
    },
    {
      "epoch": 5.669712048291052,
      "grad_norm": 0.005475752521306276,
      "learning_rate": 0.012552928370699561,
      "loss": 0.1204,
      "num_input_tokens_seen": 66881344,
      "step": 34755
    },
    {
      "epoch": 5.670527775511869,
      "grad_norm": 0.0066550555638968945,
      "learning_rate": 0.012529349870214411,
      "loss": 0.2367,
      "num_input_tokens_seen": 66890656,
      "step": 34760
    },
    {
      "epoch": 5.671343502732686,
      "grad_norm": 0.0020465946290642023,
      "learning_rate": 0.012505792569431106,
      "loss": 0.1655,
      "num_input_tokens_seen": 66900720,
      "step": 34765
    },
    {
      "epoch": 5.672159229953504,
      "grad_norm": 0.002337690442800522,
      "learning_rate": 0.012482256471982422,
      "loss": 0.1441,
      "num_input_tokens_seen": 66910624,
      "step": 34770
    },
    {
      "epoch": 5.672974957174321,
      "grad_norm": 0.0027522435411810875,
      "learning_rate": 0.012458741581497956,
      "loss": 0.1557,
      "num_input_tokens_seen": 66921440,
      "step": 34775
    },
    {
      "epoch": 5.673790684395138,
      "grad_norm": 0.002267695963382721,
      "learning_rate": 0.012435247901603974,
      "loss": 0.1746,
      "num_input_tokens_seen": 66931808,
      "step": 34780
    },
    {
      "epoch": 5.674606411615955,
      "grad_norm": 0.003678312059491873,
      "learning_rate": 0.012411775435923528,
      "loss": 0.1824,
      "num_input_tokens_seen": 66940576,
      "step": 34785
    },
    {
      "epoch": 5.6754221388367725,
      "grad_norm": 0.0021532592363655567,
      "learning_rate": 0.012388324188076354,
      "loss": 0.1951,
      "num_input_tokens_seen": 66950304,
      "step": 34790
    },
    {
      "epoch": 5.6762378660575905,
      "grad_norm": 0.0016671584453433752,
      "learning_rate": 0.012364894161678913,
      "loss": 0.1227,
      "num_input_tokens_seen": 66959792,
      "step": 34795
    },
    {
      "epoch": 5.677053593278408,
      "grad_norm": 0.004241649992763996,
      "learning_rate": 0.012341485360344445,
      "loss": 0.2089,
      "num_input_tokens_seen": 66968960,
      "step": 34800
    },
    {
      "epoch": 5.677053593278408,
      "eval_loss": 0.19042296707630157,
      "eval_runtime": 68.0071,
      "eval_samples_per_second": 40.069,
      "eval_steps_per_second": 20.042,
      "num_input_tokens_seen": 66968960,
      "step": 34800
    },
    {
      "epoch": 5.677869320499225,
      "grad_norm": 0.0032861612271517515,
      "learning_rate": 0.01231809778768283,
      "loss": 0.2194,
      "num_input_tokens_seen": 66978448,
      "step": 34805
    },
    {
      "epoch": 5.678685047720043,
      "grad_norm": 0.0029668286442756653,
      "learning_rate": 0.012294731447300799,
      "loss": 0.1958,
      "num_input_tokens_seen": 66990528,
      "step": 34810
    },
    {
      "epoch": 5.67950077494086,
      "grad_norm": 0.0027236808091402054,
      "learning_rate": 0.012271386342801671,
      "loss": 0.1867,
      "num_input_tokens_seen": 67000368,
      "step": 34815
    },
    {
      "epoch": 5.680316502161677,
      "grad_norm": 0.004411738831549883,
      "learning_rate": 0.012248062477785565,
      "loss": 0.1734,
      "num_input_tokens_seen": 67009280,
      "step": 34820
    },
    {
      "epoch": 5.681132229382494,
      "grad_norm": 0.004615744575858116,
      "learning_rate": 0.012224759855849305,
      "loss": 0.1515,
      "num_input_tokens_seen": 67018960,
      "step": 34825
    },
    {
      "epoch": 5.681947956603312,
      "grad_norm": 0.005510990973562002,
      "learning_rate": 0.012201478480586513,
      "loss": 0.2472,
      "num_input_tokens_seen": 67027024,
      "step": 34830
    },
    {
      "epoch": 5.682763683824129,
      "grad_norm": 0.006973403971642256,
      "learning_rate": 0.012178218355587389,
      "loss": 0.2595,
      "num_input_tokens_seen": 67037456,
      "step": 34835
    },
    {
      "epoch": 5.6835794110449465,
      "grad_norm": 0.0015509295044466853,
      "learning_rate": 0.01215497948443896,
      "loss": 0.1912,
      "num_input_tokens_seen": 67047088,
      "step": 34840
    },
    {
      "epoch": 5.684395138265764,
      "grad_norm": 0.005013908259570599,
      "learning_rate": 0.012131761870724993,
      "loss": 0.2218,
      "num_input_tokens_seen": 67056208,
      "step": 34845
    },
    {
      "epoch": 5.685210865486582,
      "grad_norm": 0.0024201537016779184,
      "learning_rate": 0.012108565518025893,
      "loss": 0.2053,
      "num_input_tokens_seen": 67065344,
      "step": 34850
    },
    {
      "epoch": 5.686026592707399,
      "grad_norm": 0.0026446902193129063,
      "learning_rate": 0.012085390429918862,
      "loss": 0.1564,
      "num_input_tokens_seen": 67075344,
      "step": 34855
    },
    {
      "epoch": 5.686842319928216,
      "grad_norm": 0.002981368685141206,
      "learning_rate": 0.012062236609977744,
      "loss": 0.1416,
      "num_input_tokens_seen": 67085776,
      "step": 34860
    },
    {
      "epoch": 5.687658047149033,
      "grad_norm": 0.0019518714398145676,
      "learning_rate": 0.01203910406177318,
      "loss": 0.1526,
      "num_input_tokens_seen": 67094448,
      "step": 34865
    },
    {
      "epoch": 5.688473774369851,
      "grad_norm": 0.003658150089904666,
      "learning_rate": 0.01201599278887252,
      "loss": 0.2173,
      "num_input_tokens_seen": 67105088,
      "step": 34870
    },
    {
      "epoch": 5.689289501590668,
      "grad_norm": 0.004656062927097082,
      "learning_rate": 0.011992902794839744,
      "loss": 0.2439,
      "num_input_tokens_seen": 67115760,
      "step": 34875
    },
    {
      "epoch": 5.690105228811485,
      "grad_norm": 0.003251962596550584,
      "learning_rate": 0.011969834083235703,
      "loss": 0.2378,
      "num_input_tokens_seen": 67125280,
      "step": 34880
    },
    {
      "epoch": 5.690920956032302,
      "grad_norm": 0.0023435186594724655,
      "learning_rate": 0.011946786657617836,
      "loss": 0.1399,
      "num_input_tokens_seen": 67136096,
      "step": 34885
    },
    {
      "epoch": 5.69173668325312,
      "grad_norm": 0.00270193163305521,
      "learning_rate": 0.011923760521540332,
      "loss": 0.1455,
      "num_input_tokens_seen": 67145056,
      "step": 34890
    },
    {
      "epoch": 5.692552410473938,
      "grad_norm": 0.0036956104449927807,
      "learning_rate": 0.011900755678554153,
      "loss": 0.2143,
      "num_input_tokens_seen": 67154656,
      "step": 34895
    },
    {
      "epoch": 5.693368137694755,
      "grad_norm": 0.001448574592359364,
      "learning_rate": 0.011877772132206893,
      "loss": 0.1453,
      "num_input_tokens_seen": 67164656,
      "step": 34900
    },
    {
      "epoch": 5.694183864915573,
      "grad_norm": 0.002784585114568472,
      "learning_rate": 0.011854809886042915,
      "loss": 0.1398,
      "num_input_tokens_seen": 67176400,
      "step": 34905
    },
    {
      "epoch": 5.69499959213639,
      "grad_norm": 0.004913302604109049,
      "learning_rate": 0.011831868943603325,
      "loss": 0.1774,
      "num_input_tokens_seen": 67187376,
      "step": 34910
    },
    {
      "epoch": 5.695815319357207,
      "grad_norm": 0.003646242432296276,
      "learning_rate": 0.011808949308425836,
      "loss": 0.2035,
      "num_input_tokens_seen": 67197072,
      "step": 34915
    },
    {
      "epoch": 5.696631046578024,
      "grad_norm": 0.002890842268243432,
      "learning_rate": 0.01178605098404501,
      "loss": 0.1363,
      "num_input_tokens_seen": 67207328,
      "step": 34920
    },
    {
      "epoch": 5.697446773798841,
      "grad_norm": 0.0016545342514291406,
      "learning_rate": 0.011763173973992002,
      "loss": 0.1386,
      "num_input_tokens_seen": 67216768,
      "step": 34925
    },
    {
      "epoch": 5.698262501019659,
      "grad_norm": 0.004761371295899153,
      "learning_rate": 0.011740318281794776,
      "loss": 0.2305,
      "num_input_tokens_seen": 67226704,
      "step": 34930
    },
    {
      "epoch": 5.699078228240476,
      "grad_norm": 0.00330321304500103,
      "learning_rate": 0.01171748391097796,
      "loss": 0.1331,
      "num_input_tokens_seen": 67237584,
      "step": 34935
    },
    {
      "epoch": 5.6998939554612935,
      "grad_norm": 0.0019716699607670307,
      "learning_rate": 0.011694670865062873,
      "loss": 0.1612,
      "num_input_tokens_seen": 67247264,
      "step": 34940
    },
    {
      "epoch": 5.7007096826821115,
      "grad_norm": 0.0051652793772518635,
      "learning_rate": 0.011671879147567616,
      "loss": 0.1528,
      "num_input_tokens_seen": 67257040,
      "step": 34945
    },
    {
      "epoch": 5.701525409902929,
      "grad_norm": 0.0018156729638576508,
      "learning_rate": 0.011649108762006893,
      "loss": 0.1721,
      "num_input_tokens_seen": 67267888,
      "step": 34950
    },
    {
      "epoch": 5.702341137123746,
      "grad_norm": 0.0032173828221857548,
      "learning_rate": 0.011626359711892265,
      "loss": 0.141,
      "num_input_tokens_seen": 67277328,
      "step": 34955
    },
    {
      "epoch": 5.703156864344563,
      "grad_norm": 0.0023503247648477554,
      "learning_rate": 0.01160363200073189,
      "loss": 0.1486,
      "num_input_tokens_seen": 67287712,
      "step": 34960
    },
    {
      "epoch": 5.70397259156538,
      "grad_norm": 0.00234935968182981,
      "learning_rate": 0.011580925632030614,
      "loss": 0.1513,
      "num_input_tokens_seen": 67296864,
      "step": 34965
    },
    {
      "epoch": 5.704788318786198,
      "grad_norm": 0.0036103606689721346,
      "learning_rate": 0.011558240609290104,
      "loss": 0.1917,
      "num_input_tokens_seen": 67305008,
      "step": 34970
    },
    {
      "epoch": 5.705604046007015,
      "grad_norm": 0.0029369464609771967,
      "learning_rate": 0.011535576936008679,
      "loss": 0.1254,
      "num_input_tokens_seen": 67313264,
      "step": 34975
    },
    {
      "epoch": 5.706419773227832,
      "grad_norm": 0.003297429531812668,
      "learning_rate": 0.011512934615681309,
      "loss": 0.2265,
      "num_input_tokens_seen": 67324128,
      "step": 34980
    },
    {
      "epoch": 5.70723550044865,
      "grad_norm": 0.006170414853841066,
      "learning_rate": 0.011490313651799765,
      "loss": 0.1996,
      "num_input_tokens_seen": 67333056,
      "step": 34985
    },
    {
      "epoch": 5.7080512276694675,
      "grad_norm": 0.0032152326311916113,
      "learning_rate": 0.011467714047852512,
      "loss": 0.1799,
      "num_input_tokens_seen": 67343120,
      "step": 34990
    },
    {
      "epoch": 5.708866954890285,
      "grad_norm": 0.004332281183451414,
      "learning_rate": 0.011445135807324624,
      "loss": 0.2201,
      "num_input_tokens_seen": 67352720,
      "step": 34995
    },
    {
      "epoch": 5.709682682111102,
      "grad_norm": 0.002759196562692523,
      "learning_rate": 0.011422578933698002,
      "loss": 0.2098,
      "num_input_tokens_seen": 67361344,
      "step": 35000
    },
    {
      "epoch": 5.709682682111102,
      "eval_loss": 0.189099982380867,
      "eval_runtime": 67.9516,
      "eval_samples_per_second": 40.102,
      "eval_steps_per_second": 20.058,
      "num_input_tokens_seen": 67361344,
      "step": 35000
    },
    {
      "epoch": 5.71049840933192,
      "grad_norm": 0.000909821770619601,
      "learning_rate": 0.011400043430451161,
      "loss": 0.2203,
      "num_input_tokens_seen": 67370688,
      "step": 35005
    },
    {
      "epoch": 5.711314136552737,
      "grad_norm": 0.0033336160704493523,
      "learning_rate": 0.011377529301059392,
      "loss": 0.1395,
      "num_input_tokens_seen": 67378512,
      "step": 35010
    },
    {
      "epoch": 5.712129863773554,
      "grad_norm": 0.0026632219087332487,
      "learning_rate": 0.011355036548994646,
      "loss": 0.2213,
      "num_input_tokens_seen": 67387536,
      "step": 35015
    },
    {
      "epoch": 5.712945590994371,
      "grad_norm": 0.002263227943331003,
      "learning_rate": 0.011332565177725584,
      "loss": 0.1288,
      "num_input_tokens_seen": 67396752,
      "step": 35020
    },
    {
      "epoch": 5.713761318215189,
      "grad_norm": 0.0024493499659001827,
      "learning_rate": 0.011310115190717585,
      "loss": 0.1614,
      "num_input_tokens_seen": 67405632,
      "step": 35025
    },
    {
      "epoch": 5.714577045436006,
      "grad_norm": 0.002163737779483199,
      "learning_rate": 0.01128768659143271,
      "loss": 0.1619,
      "num_input_tokens_seen": 67416496,
      "step": 35030
    },
    {
      "epoch": 5.715392772656823,
      "grad_norm": 0.0026964370626956224,
      "learning_rate": 0.011265279383329713,
      "loss": 0.1315,
      "num_input_tokens_seen": 67425696,
      "step": 35035
    },
    {
      "epoch": 5.716208499877641,
      "grad_norm": 0.004089251160621643,
      "learning_rate": 0.01124289356986411,
      "loss": 0.2085,
      "num_input_tokens_seen": 67434704,
      "step": 35040
    },
    {
      "epoch": 5.717024227098459,
      "grad_norm": 0.004717097617685795,
      "learning_rate": 0.011220529154488023,
      "loss": 0.2114,
      "num_input_tokens_seen": 67443088,
      "step": 35045
    },
    {
      "epoch": 5.717839954319276,
      "grad_norm": 0.002211546991020441,
      "learning_rate": 0.011198186140650346,
      "loss": 0.2309,
      "num_input_tokens_seen": 67453984,
      "step": 35050
    },
    {
      "epoch": 5.718655681540093,
      "grad_norm": 0.002326708287000656,
      "learning_rate": 0.011175864531796685,
      "loss": 0.2017,
      "num_input_tokens_seen": 67463584,
      "step": 35055
    },
    {
      "epoch": 5.71947140876091,
      "grad_norm": 0.002644312335178256,
      "learning_rate": 0.011153564331369258,
      "loss": 0.2146,
      "num_input_tokens_seen": 67474416,
      "step": 35060
    },
    {
      "epoch": 5.720287135981728,
      "grad_norm": 0.0026551424525678158,
      "learning_rate": 0.011131285542807078,
      "loss": 0.1569,
      "num_input_tokens_seen": 67483920,
      "step": 35065
    },
    {
      "epoch": 5.721102863202545,
      "grad_norm": 0.003084180410951376,
      "learning_rate": 0.011109028169545815,
      "loss": 0.1314,
      "num_input_tokens_seen": 67493824,
      "step": 35070
    },
    {
      "epoch": 5.721918590423362,
      "grad_norm": 0.004162446595728397,
      "learning_rate": 0.011086792215017804,
      "loss": 0.1836,
      "num_input_tokens_seen": 67504336,
      "step": 35075
    },
    {
      "epoch": 5.72273431764418,
      "grad_norm": 0.004018228035420179,
      "learning_rate": 0.011064577682652137,
      "loss": 0.1726,
      "num_input_tokens_seen": 67513920,
      "step": 35080
    },
    {
      "epoch": 5.723550044864997,
      "grad_norm": 0.007017063908278942,
      "learning_rate": 0.011042384575874559,
      "loss": 0.1921,
      "num_input_tokens_seen": 67522608,
      "step": 35085
    },
    {
      "epoch": 5.7243657720858145,
      "grad_norm": 0.001973614329472184,
      "learning_rate": 0.011020212898107512,
      "loss": 0.2097,
      "num_input_tokens_seen": 67533040,
      "step": 35090
    },
    {
      "epoch": 5.725181499306632,
      "grad_norm": 0.002471742918714881,
      "learning_rate": 0.010998062652770197,
      "loss": 0.1686,
      "num_input_tokens_seen": 67542624,
      "step": 35095
    },
    {
      "epoch": 5.725997226527449,
      "grad_norm": 0.0035437438637018204,
      "learning_rate": 0.010975933843278428,
      "loss": 0.1905,
      "num_input_tokens_seen": 67552720,
      "step": 35100
    },
    {
      "epoch": 5.726812953748267,
      "grad_norm": 0.004040172323584557,
      "learning_rate": 0.010953826473044714,
      "loss": 0.1871,
      "num_input_tokens_seen": 67562576,
      "step": 35105
    },
    {
      "epoch": 5.727628680969084,
      "grad_norm": 0.0026894311886280775,
      "learning_rate": 0.010931740545478357,
      "loss": 0.1856,
      "num_input_tokens_seen": 67572512,
      "step": 35110
    },
    {
      "epoch": 5.728444408189901,
      "grad_norm": 0.0037463014014065266,
      "learning_rate": 0.010909676063985218,
      "loss": 0.1591,
      "num_input_tokens_seen": 67582240,
      "step": 35115
    },
    {
      "epoch": 5.729260135410719,
      "grad_norm": 0.0019177222857251763,
      "learning_rate": 0.010887633031967974,
      "loss": 0.2001,
      "num_input_tokens_seen": 67592096,
      "step": 35120
    },
    {
      "epoch": 5.730075862631536,
      "grad_norm": 0.0031765082385390997,
      "learning_rate": 0.01086561145282589,
      "loss": 0.2126,
      "num_input_tokens_seen": 67600672,
      "step": 35125
    },
    {
      "epoch": 5.730891589852353,
      "grad_norm": 0.004797907546162605,
      "learning_rate": 0.010843611329954983,
      "loss": 0.1791,
      "num_input_tokens_seen": 67610816,
      "step": 35130
    },
    {
      "epoch": 5.7317073170731705,
      "grad_norm": 0.00372356828302145,
      "learning_rate": 0.010821632666747988,
      "loss": 0.1392,
      "num_input_tokens_seen": 67619984,
      "step": 35135
    },
    {
      "epoch": 5.732523044293988,
      "grad_norm": 0.0021735227201133966,
      "learning_rate": 0.010799675466594244,
      "loss": 0.1582,
      "num_input_tokens_seen": 67631120,
      "step": 35140
    },
    {
      "epoch": 5.733338771514806,
      "grad_norm": 0.004910400602966547,
      "learning_rate": 0.010777739732879826,
      "loss": 0.203,
      "num_input_tokens_seen": 67641232,
      "step": 35145
    },
    {
      "epoch": 5.734154498735623,
      "grad_norm": 0.0027684608940035105,
      "learning_rate": 0.010755825468987562,
      "loss": 0.1587,
      "num_input_tokens_seen": 67651568,
      "step": 35150
    },
    {
      "epoch": 5.73497022595644,
      "grad_norm": 0.0016990646254271269,
      "learning_rate": 0.010733932678296814,
      "loss": 0.1546,
      "num_input_tokens_seen": 67662560,
      "step": 35155
    },
    {
      "epoch": 5.735785953177258,
      "grad_norm": 0.0035178312100470066,
      "learning_rate": 0.010712061364183817,
      "loss": 0.1277,
      "num_input_tokens_seen": 67672400,
      "step": 35160
    },
    {
      "epoch": 5.736601680398075,
      "grad_norm": 0.0023520151153206825,
      "learning_rate": 0.010690211530021337,
      "loss": 0.1404,
      "num_input_tokens_seen": 67682272,
      "step": 35165
    },
    {
      "epoch": 5.737417407618892,
      "grad_norm": 0.0029756261501461267,
      "learning_rate": 0.01066838317917893,
      "loss": 0.2154,
      "num_input_tokens_seen": 67692128,
      "step": 35170
    },
    {
      "epoch": 5.738233134839709,
      "grad_norm": 0.0020997142419219017,
      "learning_rate": 0.010646576315022787,
      "loss": 0.1614,
      "num_input_tokens_seen": 67700880,
      "step": 35175
    },
    {
      "epoch": 5.739048862060527,
      "grad_norm": 0.0039469706825912,
      "learning_rate": 0.010624790940915785,
      "loss": 0.1836,
      "num_input_tokens_seen": 67710432,
      "step": 35180
    },
    {
      "epoch": 5.7398645892813445,
      "grad_norm": 0.003708992851898074,
      "learning_rate": 0.0106030270602175,
      "loss": 0.2306,
      "num_input_tokens_seen": 67719680,
      "step": 35185
    },
    {
      "epoch": 5.740680316502162,
      "grad_norm": 0.0048142531886696815,
      "learning_rate": 0.010581284676284252,
      "loss": 0.23,
      "num_input_tokens_seen": 67728400,
      "step": 35190
    },
    {
      "epoch": 5.741496043722979,
      "grad_norm": 0.0018197562312707305,
      "learning_rate": 0.010559563792468923,
      "loss": 0.1401,
      "num_input_tokens_seen": 67736960,
      "step": 35195
    },
    {
      "epoch": 5.742311770943797,
      "grad_norm": 0.0033649220131337643,
      "learning_rate": 0.010537864412121217,
      "loss": 0.22,
      "num_input_tokens_seen": 67746288,
      "step": 35200
    },
    {
      "epoch": 5.742311770943797,
      "eval_loss": 0.19014392793178558,
      "eval_runtime": 68.0647,
      "eval_samples_per_second": 40.035,
      "eval_steps_per_second": 20.025,
      "num_input_tokens_seen": 67746288,
      "step": 35200
    },
    {
      "epoch": 5.743127498164614,
      "grad_norm": 0.0030702315270900726,
      "learning_rate": 0.010516186538587357,
      "loss": 0.1553,
      "num_input_tokens_seen": 67756384,
      "step": 35205
    },
    {
      "epoch": 5.743943225385431,
      "grad_norm": 0.003148446325212717,
      "learning_rate": 0.01049453017521042,
      "loss": 0.2324,
      "num_input_tokens_seen": 67767360,
      "step": 35210
    },
    {
      "epoch": 5.744758952606248,
      "grad_norm": 0.0058602807112038136,
      "learning_rate": 0.010472895325330083,
      "loss": 0.2542,
      "num_input_tokens_seen": 67777456,
      "step": 35215
    },
    {
      "epoch": 5.745574679827066,
      "grad_norm": 0.002420458709821105,
      "learning_rate": 0.010451281992282662,
      "loss": 0.1404,
      "num_input_tokens_seen": 67787472,
      "step": 35220
    },
    {
      "epoch": 5.746390407047883,
      "grad_norm": 0.002293404657393694,
      "learning_rate": 0.01042969017940124,
      "loss": 0.181,
      "num_input_tokens_seen": 67797360,
      "step": 35225
    },
    {
      "epoch": 5.7472061342687,
      "grad_norm": 0.003090775106102228,
      "learning_rate": 0.01040811989001557,
      "loss": 0.1757,
      "num_input_tokens_seen": 67807184,
      "step": 35230
    },
    {
      "epoch": 5.7480218614895175,
      "grad_norm": 0.002633779775351286,
      "learning_rate": 0.010386571127451992,
      "loss": 0.1587,
      "num_input_tokens_seen": 67817840,
      "step": 35235
    },
    {
      "epoch": 5.748837588710336,
      "grad_norm": 0.005137908272445202,
      "learning_rate": 0.010365043895033682,
      "loss": 0.294,
      "num_input_tokens_seen": 67827136,
      "step": 35240
    },
    {
      "epoch": 5.749653315931153,
      "grad_norm": 0.0022559408098459244,
      "learning_rate": 0.010343538196080365,
      "loss": 0.2185,
      "num_input_tokens_seen": 67837280,
      "step": 35245
    },
    {
      "epoch": 5.75046904315197,
      "grad_norm": 0.002949128858745098,
      "learning_rate": 0.010322054033908457,
      "loss": 0.1391,
      "num_input_tokens_seen": 67845888,
      "step": 35250
    },
    {
      "epoch": 5.751284770372787,
      "grad_norm": 0.0022050898987799883,
      "learning_rate": 0.010300591411831156,
      "loss": 0.1698,
      "num_input_tokens_seen": 67856640,
      "step": 35255
    },
    {
      "epoch": 5.752100497593605,
      "grad_norm": 0.004035883583128452,
      "learning_rate": 0.010279150333158198,
      "loss": 0.1676,
      "num_input_tokens_seen": 67865872,
      "step": 35260
    },
    {
      "epoch": 5.752916224814422,
      "grad_norm": 0.0037801344878971577,
      "learning_rate": 0.010257730801196107,
      "loss": 0.2087,
      "num_input_tokens_seen": 67875104,
      "step": 35265
    },
    {
      "epoch": 5.753731952035239,
      "grad_norm": 0.0030405938159674406,
      "learning_rate": 0.010236332819248056,
      "loss": 0.111,
      "num_input_tokens_seen": 67883520,
      "step": 35270
    },
    {
      "epoch": 5.754547679256056,
      "grad_norm": 0.005639856215566397,
      "learning_rate": 0.010214956390613854,
      "loss": 0.187,
      "num_input_tokens_seen": 67893632,
      "step": 35275
    },
    {
      "epoch": 5.755363406476874,
      "grad_norm": 0.001983066787943244,
      "learning_rate": 0.010193601518590034,
      "loss": 0.1642,
      "num_input_tokens_seen": 67902912,
      "step": 35280
    },
    {
      "epoch": 5.7561791336976915,
      "grad_norm": 0.0015229969285428524,
      "learning_rate": 0.010172268206469758,
      "loss": 0.1562,
      "num_input_tokens_seen": 67912960,
      "step": 35285
    },
    {
      "epoch": 5.756994860918509,
      "grad_norm": 0.004797075409442186,
      "learning_rate": 0.010150956457542897,
      "loss": 0.1242,
      "num_input_tokens_seen": 67923456,
      "step": 35290
    },
    {
      "epoch": 5.757810588139327,
      "grad_norm": 0.003198873484507203,
      "learning_rate": 0.010129666275096054,
      "loss": 0.1251,
      "num_input_tokens_seen": 67932544,
      "step": 35295
    },
    {
      "epoch": 5.758626315360144,
      "grad_norm": 0.004802321083843708,
      "learning_rate": 0.010108397662412338,
      "loss": 0.2441,
      "num_input_tokens_seen": 67941888,
      "step": 35300
    },
    {
      "epoch": 5.759442042580961,
      "grad_norm": 0.0035983272828161716,
      "learning_rate": 0.010087150622771707,
      "loss": 0.1917,
      "num_input_tokens_seen": 67950944,
      "step": 35305
    },
    {
      "epoch": 5.760257769801778,
      "grad_norm": 0.0031282759737223387,
      "learning_rate": 0.010065925159450739,
      "loss": 0.1372,
      "num_input_tokens_seen": 67959216,
      "step": 35310
    },
    {
      "epoch": 5.761073497022595,
      "grad_norm": 0.004331711679697037,
      "learning_rate": 0.010044721275722618,
      "loss": 0.2471,
      "num_input_tokens_seen": 67968912,
      "step": 35315
    },
    {
      "epoch": 5.761889224243413,
      "grad_norm": 0.0038657591212540865,
      "learning_rate": 0.01002353897485726,
      "loss": 0.1469,
      "num_input_tokens_seen": 67978224,
      "step": 35320
    },
    {
      "epoch": 5.76270495146423,
      "grad_norm": 0.0026199370622634888,
      "learning_rate": 0.010002378260121236,
      "loss": 0.2222,
      "num_input_tokens_seen": 67988352,
      "step": 35325
    },
    {
      "epoch": 5.7635206786850475,
      "grad_norm": 0.003180623520165682,
      "learning_rate": 0.009981239134777786,
      "loss": 0.1578,
      "num_input_tokens_seen": 67996544,
      "step": 35330
    },
    {
      "epoch": 5.7643364059058655,
      "grad_norm": 0.0030041418503969908,
      "learning_rate": 0.009960121602086884,
      "loss": 0.1268,
      "num_input_tokens_seen": 68006208,
      "step": 35335
    },
    {
      "epoch": 5.765152133126683,
      "grad_norm": 0.003527133958414197,
      "learning_rate": 0.009939025665305062,
      "loss": 0.2521,
      "num_input_tokens_seen": 68015392,
      "step": 35340
    },
    {
      "epoch": 5.7659678603475,
      "grad_norm": 0.005348314996808767,
      "learning_rate": 0.009917951327685597,
      "loss": 0.1977,
      "num_input_tokens_seen": 68025632,
      "step": 35345
    },
    {
      "epoch": 5.766783587568317,
      "grad_norm": 0.003907869569957256,
      "learning_rate": 0.009896898592478425,
      "loss": 0.2579,
      "num_input_tokens_seen": 68035104,
      "step": 35350
    },
    {
      "epoch": 5.767599314789135,
      "grad_norm": 0.002749884268268943,
      "learning_rate": 0.009875867462930132,
      "loss": 0.1755,
      "num_input_tokens_seen": 68043504,
      "step": 35355
    },
    {
      "epoch": 5.768415042009952,
      "grad_norm": 0.003185710869729519,
      "learning_rate": 0.009854857942284006,
      "loss": 0.1991,
      "num_input_tokens_seen": 68052288,
      "step": 35360
    },
    {
      "epoch": 5.769230769230769,
      "grad_norm": 0.002168115694075823,
      "learning_rate": 0.009833870033779923,
      "loss": 0.1639,
      "num_input_tokens_seen": 68062592,
      "step": 35365
    },
    {
      "epoch": 5.770046496451586,
      "grad_norm": 0.0034844642505049706,
      "learning_rate": 0.009812903740654527,
      "loss": 0.1997,
      "num_input_tokens_seen": 68072432,
      "step": 35370
    },
    {
      "epoch": 5.770862223672404,
      "grad_norm": 0.0031113876029849052,
      "learning_rate": 0.009791959066141097,
      "loss": 0.2216,
      "num_input_tokens_seen": 68082816,
      "step": 35375
    },
    {
      "epoch": 5.771677950893221,
      "grad_norm": 0.00594245595857501,
      "learning_rate": 0.009771036013469537,
      "loss": 0.1825,
      "num_input_tokens_seen": 68092512,
      "step": 35380
    },
    {
      "epoch": 5.772493678114039,
      "grad_norm": 0.004528709687292576,
      "learning_rate": 0.00975013458586646,
      "loss": 0.1657,
      "num_input_tokens_seen": 68103472,
      "step": 35385
    },
    {
      "epoch": 5.773309405334856,
      "grad_norm": 0.0018761847168207169,
      "learning_rate": 0.009729254786555107,
      "loss": 0.1351,
      "num_input_tokens_seen": 68111536,
      "step": 35390
    },
    {
      "epoch": 5.774125132555674,
      "grad_norm": 0.0038544542621821165,
      "learning_rate": 0.009708396618755421,
      "loss": 0.19,
      "num_input_tokens_seen": 68121568,
      "step": 35395
    },
    {
      "epoch": 5.774940859776491,
      "grad_norm": 0.003004673169925809,
      "learning_rate": 0.009687560085683994,
      "loss": 0.1752,
      "num_input_tokens_seen": 68131952,
      "step": 35400
    },
    {
      "epoch": 5.774940859776491,
      "eval_loss": 0.19195522367954254,
      "eval_runtime": 67.9622,
      "eval_samples_per_second": 40.096,
      "eval_steps_per_second": 20.055,
      "num_input_tokens_seen": 68131952,
      "step": 35400
    },
    {
      "epoch": 5.775756586997308,
      "grad_norm": 0.002352564362809062,
      "learning_rate": 0.009666745190554054,
      "loss": 0.1654,
      "num_input_tokens_seen": 68141456,
      "step": 35405
    },
    {
      "epoch": 5.776572314218125,
      "grad_norm": 0.003332925960421562,
      "learning_rate": 0.009645951936575553,
      "loss": 0.1689,
      "num_input_tokens_seen": 68151312,
      "step": 35410
    },
    {
      "epoch": 5.777388041438943,
      "grad_norm": 0.002298889681696892,
      "learning_rate": 0.00962518032695509,
      "loss": 0.1391,
      "num_input_tokens_seen": 68160224,
      "step": 35415
    },
    {
      "epoch": 5.77820376865976,
      "grad_norm": 0.0021364926360547543,
      "learning_rate": 0.009604430364895855,
      "loss": 0.1607,
      "num_input_tokens_seen": 68170112,
      "step": 35420
    },
    {
      "epoch": 5.779019495880577,
      "grad_norm": 0.004636113997548819,
      "learning_rate": 0.00958370205359777,
      "loss": 0.1935,
      "num_input_tokens_seen": 68178400,
      "step": 35425
    },
    {
      "epoch": 5.7798352231013945,
      "grad_norm": 0.0025831658858805895,
      "learning_rate": 0.009562995396257445,
      "loss": 0.1688,
      "num_input_tokens_seen": 68188208,
      "step": 35430
    },
    {
      "epoch": 5.7806509503222125,
      "grad_norm": 0.001882842625491321,
      "learning_rate": 0.009542310396068026,
      "loss": 0.189,
      "num_input_tokens_seen": 68197728,
      "step": 35435
    },
    {
      "epoch": 5.78146667754303,
      "grad_norm": 0.0033926237374544144,
      "learning_rate": 0.009521647056219495,
      "loss": 0.2326,
      "num_input_tokens_seen": 68206768,
      "step": 35440
    },
    {
      "epoch": 5.782282404763847,
      "grad_norm": 0.0035611288622021675,
      "learning_rate": 0.00950100537989832,
      "loss": 0.2078,
      "num_input_tokens_seen": 68216432,
      "step": 35445
    },
    {
      "epoch": 5.783098131984664,
      "grad_norm": 0.004329445771872997,
      "learning_rate": 0.00948038537028772,
      "loss": 0.1682,
      "num_input_tokens_seen": 68226512,
      "step": 35450
    },
    {
      "epoch": 5.783913859205482,
      "grad_norm": 0.003604666795581579,
      "learning_rate": 0.009459787030567617,
      "loss": 0.2198,
      "num_input_tokens_seen": 68236448,
      "step": 35455
    },
    {
      "epoch": 5.784729586426299,
      "grad_norm": 0.0030623357743024826,
      "learning_rate": 0.00943921036391449,
      "loss": 0.1915,
      "num_input_tokens_seen": 68246624,
      "step": 35460
    },
    {
      "epoch": 5.785545313647116,
      "grad_norm": 0.0031582198571413755,
      "learning_rate": 0.009418655373501483,
      "loss": 0.1999,
      "num_input_tokens_seen": 68256688,
      "step": 35465
    },
    {
      "epoch": 5.786361040867934,
      "grad_norm": 0.004383660387247801,
      "learning_rate": 0.00939812206249851,
      "loss": 0.1995,
      "num_input_tokens_seen": 68266160,
      "step": 35470
    },
    {
      "epoch": 5.787176768088751,
      "grad_norm": 0.0016278071561828256,
      "learning_rate": 0.009377610434072004,
      "loss": 0.1544,
      "num_input_tokens_seen": 68275552,
      "step": 35475
    },
    {
      "epoch": 5.7879924953095685,
      "grad_norm": 0.0010724578751251101,
      "learning_rate": 0.009357120491385167,
      "loss": 0.1836,
      "num_input_tokens_seen": 68284576,
      "step": 35480
    },
    {
      "epoch": 5.788808222530386,
      "grad_norm": 0.0016752015799283981,
      "learning_rate": 0.009336652237597743,
      "loss": 0.1955,
      "num_input_tokens_seen": 68293680,
      "step": 35485
    },
    {
      "epoch": 5.789623949751203,
      "grad_norm": 0.0027594652492552996,
      "learning_rate": 0.009316205675866251,
      "loss": 0.1321,
      "num_input_tokens_seen": 68303056,
      "step": 35490
    },
    {
      "epoch": 5.790439676972021,
      "grad_norm": 0.001716655446216464,
      "learning_rate": 0.00929578080934379,
      "loss": 0.1685,
      "num_input_tokens_seen": 68314208,
      "step": 35495
    },
    {
      "epoch": 5.791255404192838,
      "grad_norm": 0.003569555701687932,
      "learning_rate": 0.00927537764118012,
      "loss": 0.1408,
      "num_input_tokens_seen": 68325040,
      "step": 35500
    },
    {
      "epoch": 5.792071131413655,
      "grad_norm": 0.0035352613776922226,
      "learning_rate": 0.009254996174521678,
      "loss": 0.1395,
      "num_input_tokens_seen": 68334320,
      "step": 35505
    },
    {
      "epoch": 5.792886858634473,
      "grad_norm": 0.003655106294900179,
      "learning_rate": 0.009234636412511531,
      "loss": 0.2344,
      "num_input_tokens_seen": 68344112,
      "step": 35510
    },
    {
      "epoch": 5.79370258585529,
      "grad_norm": 0.004039014223963022,
      "learning_rate": 0.009214298358289418,
      "loss": 0.1618,
      "num_input_tokens_seen": 68354176,
      "step": 35515
    },
    {
      "epoch": 5.794518313076107,
      "grad_norm": 0.007336507085710764,
      "learning_rate": 0.00919398201499173,
      "loss": 0.194,
      "num_input_tokens_seen": 68364016,
      "step": 35520
    },
    {
      "epoch": 5.7953340402969244,
      "grad_norm": 0.006269774865359068,
      "learning_rate": 0.009173687385751495,
      "loss": 0.2388,
      "num_input_tokens_seen": 68372544,
      "step": 35525
    },
    {
      "epoch": 5.796149767517742,
      "grad_norm": 0.0024950469378381968,
      "learning_rate": 0.009153414473698407,
      "loss": 0.1318,
      "num_input_tokens_seen": 68381936,
      "step": 35530
    },
    {
      "epoch": 5.79696549473856,
      "grad_norm": 0.004748499486595392,
      "learning_rate": 0.009133163281958784,
      "loss": 0.1768,
      "num_input_tokens_seen": 68392272,
      "step": 35535
    },
    {
      "epoch": 5.797781221959377,
      "grad_norm": 0.0030880088452249765,
      "learning_rate": 0.009112933813655627,
      "loss": 0.2002,
      "num_input_tokens_seen": 68402000,
      "step": 35540
    },
    {
      "epoch": 5.798596949180194,
      "grad_norm": 0.005127036478370428,
      "learning_rate": 0.009092726071908573,
      "loss": 0.1162,
      "num_input_tokens_seen": 68410896,
      "step": 35545
    },
    {
      "epoch": 5.799412676401012,
      "grad_norm": 0.0020566913299262524,
      "learning_rate": 0.0090725400598339,
      "loss": 0.1702,
      "num_input_tokens_seen": 68420944,
      "step": 35550
    },
    {
      "epoch": 5.800228403621829,
      "grad_norm": 0.0037987385876476765,
      "learning_rate": 0.009052375780544563,
      "loss": 0.2177,
      "num_input_tokens_seen": 68429664,
      "step": 35555
    },
    {
      "epoch": 5.801044130842646,
      "grad_norm": 0.006278222892433405,
      "learning_rate": 0.009032233237150144,
      "loss": 0.1523,
      "num_input_tokens_seen": 68439200,
      "step": 35560
    },
    {
      "epoch": 5.801859858063463,
      "grad_norm": 0.0031605514232069254,
      "learning_rate": 0.009012112432756875,
      "loss": 0.1723,
      "num_input_tokens_seen": 68447856,
      "step": 35565
    },
    {
      "epoch": 5.802675585284281,
      "grad_norm": 0.0034207485150545835,
      "learning_rate": 0.008992013370467605,
      "loss": 0.137,
      "num_input_tokens_seen": 68456720,
      "step": 35570
    },
    {
      "epoch": 5.803491312505098,
      "grad_norm": 0.004644653294235468,
      "learning_rate": 0.008971936053381924,
      "loss": 0.2041,
      "num_input_tokens_seen": 68466800,
      "step": 35575
    },
    {
      "epoch": 5.8043070397259156,
      "grad_norm": 0.003164102789014578,
      "learning_rate": 0.008951880484595953,
      "loss": 0.2022,
      "num_input_tokens_seen": 68476496,
      "step": 35580
    },
    {
      "epoch": 5.805122766946733,
      "grad_norm": 0.0020813862793147564,
      "learning_rate": 0.008931846667202552,
      "loss": 0.1052,
      "num_input_tokens_seen": 68486432,
      "step": 35585
    },
    {
      "epoch": 5.805938494167551,
      "grad_norm": 0.002885307651013136,
      "learning_rate": 0.008911834604291152,
      "loss": 0.2006,
      "num_input_tokens_seen": 68495568,
      "step": 35590
    },
    {
      "epoch": 5.806754221388368,
      "grad_norm": 0.00283107184804976,
      "learning_rate": 0.008891844298947882,
      "loss": 0.1683,
      "num_input_tokens_seen": 68504480,
      "step": 35595
    },
    {
      "epoch": 5.807569948609185,
      "grad_norm": 0.0032880983781069517,
      "learning_rate": 0.008871875754255508,
      "loss": 0.1639,
      "num_input_tokens_seen": 68514656,
      "step": 35600
    },
    {
      "epoch": 5.807569948609185,
      "eval_loss": 0.19133777916431427,
      "eval_runtime": 68.0818,
      "eval_samples_per_second": 40.025,
      "eval_steps_per_second": 20.02,
      "num_input_tokens_seen": 68514656,
      "step": 35600
    },
    {
      "epoch": 5.808385675830002,
      "grad_norm": 0.0021244126837700605,
      "learning_rate": 0.008851928973293422,
      "loss": 0.1425,
      "num_input_tokens_seen": 68524416,
      "step": 35605
    },
    {
      "epoch": 5.80920140305082,
      "grad_norm": 0.002606586553156376,
      "learning_rate": 0.00883200395913764,
      "loss": 0.1989,
      "num_input_tokens_seen": 68534848,
      "step": 35610
    },
    {
      "epoch": 5.810017130271637,
      "grad_norm": 0.002751175547018647,
      "learning_rate": 0.00881210071486091,
      "loss": 0.2125,
      "num_input_tokens_seen": 68545344,
      "step": 35615
    },
    {
      "epoch": 5.810832857492454,
      "grad_norm": 0.0022953441366553307,
      "learning_rate": 0.008792219243532505,
      "loss": 0.1161,
      "num_input_tokens_seen": 68553952,
      "step": 35620
    },
    {
      "epoch": 5.8116485847132715,
      "grad_norm": 0.004505938850343227,
      "learning_rate": 0.008772359548218428,
      "loss": 0.1296,
      "num_input_tokens_seen": 68564416,
      "step": 35625
    },
    {
      "epoch": 5.8124643119340895,
      "grad_norm": 0.004073989111930132,
      "learning_rate": 0.008752521631981274,
      "loss": 0.1835,
      "num_input_tokens_seen": 68573360,
      "step": 35630
    },
    {
      "epoch": 5.813280039154907,
      "grad_norm": 0.004463943187147379,
      "learning_rate": 0.008732705497880315,
      "loss": 0.1876,
      "num_input_tokens_seen": 68581856,
      "step": 35635
    },
    {
      "epoch": 5.814095766375724,
      "grad_norm": 0.004252287559211254,
      "learning_rate": 0.008712911148971459,
      "loss": 0.2046,
      "num_input_tokens_seen": 68591456,
      "step": 35640
    },
    {
      "epoch": 5.814911493596542,
      "grad_norm": 0.004119445104151964,
      "learning_rate": 0.008693138588307208,
      "loss": 0.1881,
      "num_input_tokens_seen": 68600400,
      "step": 35645
    },
    {
      "epoch": 5.815727220817359,
      "grad_norm": 0.004673547111451626,
      "learning_rate": 0.008673387818936762,
      "loss": 0.228,
      "num_input_tokens_seen": 68609504,
      "step": 35650
    },
    {
      "epoch": 5.816542948038176,
      "grad_norm": 0.004188741557300091,
      "learning_rate": 0.008653658843905948,
      "loss": 0.2839,
      "num_input_tokens_seen": 68618256,
      "step": 35655
    },
    {
      "epoch": 5.817358675258993,
      "grad_norm": 0.004513585474342108,
      "learning_rate": 0.0086339516662572,
      "loss": 0.1826,
      "num_input_tokens_seen": 68626992,
      "step": 35660
    },
    {
      "epoch": 5.81817440247981,
      "grad_norm": 0.0037405784241855145,
      "learning_rate": 0.008614266289029638,
      "loss": 0.1649,
      "num_input_tokens_seen": 68638032,
      "step": 35665
    },
    {
      "epoch": 5.818990129700628,
      "grad_norm": 0.0041863517835736275,
      "learning_rate": 0.008594602715258965,
      "loss": 0.1712,
      "num_input_tokens_seen": 68646784,
      "step": 35670
    },
    {
      "epoch": 5.8198058569214455,
      "grad_norm": 0.0041234721429646015,
      "learning_rate": 0.008574960947977573,
      "loss": 0.1685,
      "num_input_tokens_seen": 68656560,
      "step": 35675
    },
    {
      "epoch": 5.820621584142263,
      "grad_norm": 0.006746771279722452,
      "learning_rate": 0.008555340990214438,
      "loss": 0.1483,
      "num_input_tokens_seen": 68666272,
      "step": 35680
    },
    {
      "epoch": 5.821437311363081,
      "grad_norm": 0.0018471042858436704,
      "learning_rate": 0.008535742844995258,
      "loss": 0.1504,
      "num_input_tokens_seen": 68677104,
      "step": 35685
    },
    {
      "epoch": 5.822253038583898,
      "grad_norm": 0.0033608085941523314,
      "learning_rate": 0.008516166515342266,
      "loss": 0.1326,
      "num_input_tokens_seen": 68686176,
      "step": 35690
    },
    {
      "epoch": 5.823068765804715,
      "grad_norm": 0.004382480401545763,
      "learning_rate": 0.008496612004274411,
      "loss": 0.1423,
      "num_input_tokens_seen": 68696352,
      "step": 35695
    },
    {
      "epoch": 5.823884493025532,
      "grad_norm": 0.004703611601144075,
      "learning_rate": 0.008477079314807201,
      "loss": 0.1245,
      "num_input_tokens_seen": 68703376,
      "step": 35700
    },
    {
      "epoch": 5.824700220246349,
      "grad_norm": 0.002779384609311819,
      "learning_rate": 0.008457568449952874,
      "loss": 0.1373,
      "num_input_tokens_seen": 68713296,
      "step": 35705
    },
    {
      "epoch": 5.825515947467167,
      "grad_norm": 0.0015154063003137708,
      "learning_rate": 0.008438079412720189,
      "loss": 0.1817,
      "num_input_tokens_seen": 68724608,
      "step": 35710
    },
    {
      "epoch": 5.826331674687984,
      "grad_norm": 0.0021068770438432693,
      "learning_rate": 0.00841861220611466,
      "loss": 0.179,
      "num_input_tokens_seen": 68734736,
      "step": 35715
    },
    {
      "epoch": 5.827147401908801,
      "grad_norm": 0.0027073936071246862,
      "learning_rate": 0.008399166833138355,
      "loss": 0.1125,
      "num_input_tokens_seen": 68745744,
      "step": 35720
    },
    {
      "epoch": 5.8279631291296194,
      "grad_norm": 0.004371614195406437,
      "learning_rate": 0.008379743296789987,
      "loss": 0.1797,
      "num_input_tokens_seen": 68756720,
      "step": 35725
    },
    {
      "epoch": 5.828778856350437,
      "grad_norm": 0.002657422097399831,
      "learning_rate": 0.008360341600064896,
      "loss": 0.1952,
      "num_input_tokens_seen": 68765376,
      "step": 35730
    },
    {
      "epoch": 5.829594583571254,
      "grad_norm": 0.004368233494460583,
      "learning_rate": 0.008340961745955121,
      "loss": 0.2075,
      "num_input_tokens_seen": 68776352,
      "step": 35735
    },
    {
      "epoch": 5.830410310792071,
      "grad_norm": 0.007203834131360054,
      "learning_rate": 0.008321603737449224,
      "loss": 0.2068,
      "num_input_tokens_seen": 68786368,
      "step": 35740
    },
    {
      "epoch": 5.831226038012889,
      "grad_norm": 0.0034685153514146805,
      "learning_rate": 0.008302267577532479,
      "loss": 0.1538,
      "num_input_tokens_seen": 68796800,
      "step": 35745
    },
    {
      "epoch": 5.832041765233706,
      "grad_norm": 0.0040082684718072414,
      "learning_rate": 0.008282953269186771,
      "loss": 0.2068,
      "num_input_tokens_seen": 68806240,
      "step": 35750
    },
    {
      "epoch": 5.832857492454523,
      "grad_norm": 0.0049104574136435986,
      "learning_rate": 0.008263660815390567,
      "loss": 0.2373,
      "num_input_tokens_seen": 68815184,
      "step": 35755
    },
    {
      "epoch": 5.83367321967534,
      "grad_norm": 0.004433753434568644,
      "learning_rate": 0.008244390219119069,
      "loss": 0.2011,
      "num_input_tokens_seen": 68825488,
      "step": 35760
    },
    {
      "epoch": 5.834488946896158,
      "grad_norm": 0.002394289243966341,
      "learning_rate": 0.008225141483343967,
      "loss": 0.1546,
      "num_input_tokens_seen": 68834560,
      "step": 35765
    },
    {
      "epoch": 5.835304674116975,
      "grad_norm": 0.0037549789994955063,
      "learning_rate": 0.00820591461103372,
      "loss": 0.197,
      "num_input_tokens_seen": 68845072,
      "step": 35770
    },
    {
      "epoch": 5.8361204013377925,
      "grad_norm": 0.00229848874732852,
      "learning_rate": 0.008186709605153358,
      "loss": 0.169,
      "num_input_tokens_seen": 68855392,
      "step": 35775
    },
    {
      "epoch": 5.83693612855861,
      "grad_norm": 0.0019361685262992978,
      "learning_rate": 0.008167526468664492,
      "loss": 0.1079,
      "num_input_tokens_seen": 68866128,
      "step": 35780
    },
    {
      "epoch": 5.837751855779428,
      "grad_norm": 0.0035621088463813066,
      "learning_rate": 0.008148365204525443,
      "loss": 0.1703,
      "num_input_tokens_seen": 68875280,
      "step": 35785
    },
    {
      "epoch": 5.838567583000245,
      "grad_norm": 0.008231868967413902,
      "learning_rate": 0.00812922581569106,
      "loss": 0.2188,
      "num_input_tokens_seen": 68883088,
      "step": 35790
    },
    {
      "epoch": 5.839383310221062,
      "grad_norm": 0.0024076439440250397,
      "learning_rate": 0.008110108305112934,
      "loss": 0.1797,
      "num_input_tokens_seen": 68894112,
      "step": 35795
    },
    {
      "epoch": 5.840199037441879,
      "grad_norm": 0.0032723622862249613,
      "learning_rate": 0.008091012675739223,
      "loss": 0.1417,
      "num_input_tokens_seen": 68904544,
      "step": 35800
    },
    {
      "epoch": 5.840199037441879,
      "eval_loss": 0.19214113056659698,
      "eval_runtime": 67.9771,
      "eval_samples_per_second": 40.087,
      "eval_steps_per_second": 20.051,
      "num_input_tokens_seen": 68904544,
      "step": 35800
    },
    {
      "epoch": 5.841014764662697,
      "grad_norm": 0.003989984281361103,
      "learning_rate": 0.008071938930514671,
      "loss": 0.2217,
      "num_input_tokens_seen": 68914592,
      "step": 35805
    },
    {
      "epoch": 5.841830491883514,
      "grad_norm": 0.005490522366017103,
      "learning_rate": 0.008052887072380726,
      "loss": 0.1957,
      "num_input_tokens_seen": 68924256,
      "step": 35810
    },
    {
      "epoch": 5.842646219104331,
      "grad_norm": 0.0036638055462390184,
      "learning_rate": 0.008033857104275437,
      "loss": 0.1778,
      "num_input_tokens_seen": 68934032,
      "step": 35815
    },
    {
      "epoch": 5.8434619463251485,
      "grad_norm": 0.0052547939121723175,
      "learning_rate": 0.008014849029133424,
      "loss": 0.2132,
      "num_input_tokens_seen": 68943408,
      "step": 35820
    },
    {
      "epoch": 5.8442776735459665,
      "grad_norm": 0.003190056188032031,
      "learning_rate": 0.007995862849885975,
      "loss": 0.1901,
      "num_input_tokens_seen": 68952304,
      "step": 35825
    },
    {
      "epoch": 5.845093400766784,
      "grad_norm": 0.0032055105548352003,
      "learning_rate": 0.007976898569461032,
      "loss": 0.2043,
      "num_input_tokens_seen": 68961488,
      "step": 35830
    },
    {
      "epoch": 5.845909127987601,
      "grad_norm": 0.004324749577790499,
      "learning_rate": 0.007957956190783088,
      "loss": 0.1779,
      "num_input_tokens_seen": 68971008,
      "step": 35835
    },
    {
      "epoch": 5.846724855208418,
      "grad_norm": 0.0015539913438260555,
      "learning_rate": 0.007939035716773324,
      "loss": 0.171,
      "num_input_tokens_seen": 68979376,
      "step": 35840
    },
    {
      "epoch": 5.847540582429236,
      "grad_norm": 0.002024256158620119,
      "learning_rate": 0.007920137150349487,
      "loss": 0.1541,
      "num_input_tokens_seen": 68988640,
      "step": 35845
    },
    {
      "epoch": 5.848356309650053,
      "grad_norm": 0.00264594703912735,
      "learning_rate": 0.007901260494425981,
      "loss": 0.132,
      "num_input_tokens_seen": 68997440,
      "step": 35850
    },
    {
      "epoch": 5.84917203687087,
      "grad_norm": 0.003938694484531879,
      "learning_rate": 0.007882405751913861,
      "loss": 0.1962,
      "num_input_tokens_seen": 69006896,
      "step": 35855
    },
    {
      "epoch": 5.849987764091688,
      "grad_norm": 0.0023566358722746372,
      "learning_rate": 0.007863572925720702,
      "loss": 0.1442,
      "num_input_tokens_seen": 69016464,
      "step": 35860
    },
    {
      "epoch": 5.850803491312505,
      "grad_norm": 0.004815423395484686,
      "learning_rate": 0.007844762018750827,
      "loss": 0.1687,
      "num_input_tokens_seen": 69025424,
      "step": 35865
    },
    {
      "epoch": 5.8516192185333225,
      "grad_norm": 0.003024268662557006,
      "learning_rate": 0.007825973033905054,
      "loss": 0.2636,
      "num_input_tokens_seen": 69035440,
      "step": 35870
    },
    {
      "epoch": 5.85243494575414,
      "grad_norm": 0.003211967647075653,
      "learning_rate": 0.007807205974080927,
      "loss": 0.1238,
      "num_input_tokens_seen": 69044128,
      "step": 35875
    },
    {
      "epoch": 5.853250672974957,
      "grad_norm": 0.004643877502530813,
      "learning_rate": 0.007788460842172551,
      "loss": 0.1619,
      "num_input_tokens_seen": 69054752,
      "step": 35880
    },
    {
      "epoch": 5.854066400195775,
      "grad_norm": 0.004700698424130678,
      "learning_rate": 0.0077697376410706285,
      "loss": 0.1559,
      "num_input_tokens_seen": 69063216,
      "step": 35885
    },
    {
      "epoch": 5.854882127416592,
      "grad_norm": 0.004008902702480555,
      "learning_rate": 0.007751036373662567,
      "loss": 0.2191,
      "num_input_tokens_seen": 69072288,
      "step": 35890
    },
    {
      "epoch": 5.855697854637409,
      "grad_norm": 0.0022495496086776257,
      "learning_rate": 0.00773235704283231,
      "loss": 0.2117,
      "num_input_tokens_seen": 69082304,
      "step": 35895
    },
    {
      "epoch": 5.856513581858227,
      "grad_norm": 0.0026850770227611065,
      "learning_rate": 0.007713699651460437,
      "loss": 0.1439,
      "num_input_tokens_seen": 69092528,
      "step": 35900
    },
    {
      "epoch": 5.857329309079044,
      "grad_norm": 0.0024478421546518803,
      "learning_rate": 0.007695064202424162,
      "loss": 0.1333,
      "num_input_tokens_seen": 69100800,
      "step": 35905
    },
    {
      "epoch": 5.858145036299861,
      "grad_norm": 0.00415376340970397,
      "learning_rate": 0.007676450698597286,
      "loss": 0.2456,
      "num_input_tokens_seen": 69110192,
      "step": 35910
    },
    {
      "epoch": 5.858960763520678,
      "grad_norm": 0.0053832996636629105,
      "learning_rate": 0.007657859142850265,
      "loss": 0.1654,
      "num_input_tokens_seen": 69119808,
      "step": 35915
    },
    {
      "epoch": 5.859776490741496,
      "grad_norm": 0.004133149515837431,
      "learning_rate": 0.0076392895380501535,
      "loss": 0.1439,
      "num_input_tokens_seen": 69130016,
      "step": 35920
    },
    {
      "epoch": 5.8605922179623136,
      "grad_norm": 0.0035995952785015106,
      "learning_rate": 0.007620741887060611,
      "loss": 0.1268,
      "num_input_tokens_seen": 69138272,
      "step": 35925
    },
    {
      "epoch": 5.861407945183131,
      "grad_norm": 0.003234164323657751,
      "learning_rate": 0.007602216192741901,
      "loss": 0.192,
      "num_input_tokens_seen": 69148512,
      "step": 35930
    },
    {
      "epoch": 5.862223672403948,
      "grad_norm": 0.0014905869029462337,
      "learning_rate": 0.007583712457950969,
      "loss": 0.2071,
      "num_input_tokens_seen": 69158784,
      "step": 35935
    },
    {
      "epoch": 5.863039399624766,
      "grad_norm": 0.0026791372802108526,
      "learning_rate": 0.007565230685541269,
      "loss": 0.2008,
      "num_input_tokens_seen": 69168912,
      "step": 35940
    },
    {
      "epoch": 5.863855126845583,
      "grad_norm": 0.005521808750927448,
      "learning_rate": 0.007546770878362968,
      "loss": 0.1737,
      "num_input_tokens_seen": 69180240,
      "step": 35945
    },
    {
      "epoch": 5.8646708540664,
      "grad_norm": 0.003118957160040736,
      "learning_rate": 0.0075283330392627405,
      "loss": 0.2271,
      "num_input_tokens_seen": 69190752,
      "step": 35950
    },
    {
      "epoch": 5.865486581287217,
      "grad_norm": 0.0020506607834249735,
      "learning_rate": 0.007509917171083979,
      "loss": 0.1553,
      "num_input_tokens_seen": 69199968,
      "step": 35955
    },
    {
      "epoch": 5.866302308508035,
      "grad_norm": 0.0026875659823417664,
      "learning_rate": 0.007491523276666662,
      "loss": 0.194,
      "num_input_tokens_seen": 69209552,
      "step": 35960
    },
    {
      "epoch": 5.867118035728852,
      "grad_norm": 0.003108591539785266,
      "learning_rate": 0.007473151358847318,
      "loss": 0.2486,
      "num_input_tokens_seen": 69218976,
      "step": 35965
    },
    {
      "epoch": 5.8679337629496695,
      "grad_norm": 0.0037041797768324614,
      "learning_rate": 0.007454801420459117,
      "loss": 0.2091,
      "num_input_tokens_seen": 69228976,
      "step": 35970
    },
    {
      "epoch": 5.868749490170487,
      "grad_norm": 0.0030465442687273026,
      "learning_rate": 0.0074364734643319105,
      "loss": 0.198,
      "num_input_tokens_seen": 69238416,
      "step": 35975
    },
    {
      "epoch": 5.869565217391305,
      "grad_norm": 0.004218382760882378,
      "learning_rate": 0.007418167493292022,
      "loss": 0.1429,
      "num_input_tokens_seen": 69248032,
      "step": 35980
    },
    {
      "epoch": 5.870380944612122,
      "grad_norm": 0.0017486143624410033,
      "learning_rate": 0.0073998835101625245,
      "loss": 0.1615,
      "num_input_tokens_seen": 69257328,
      "step": 35985
    },
    {
      "epoch": 5.871196671832939,
      "grad_norm": 0.0037378091365098953,
      "learning_rate": 0.007381621517762998,
      "loss": 0.1706,
      "num_input_tokens_seen": 69267056,
      "step": 35990
    },
    {
      "epoch": 5.872012399053756,
      "grad_norm": 0.001340976683422923,
      "learning_rate": 0.007363381518909689,
      "loss": 0.1436,
      "num_input_tokens_seen": 69277760,
      "step": 35995
    },
    {
      "epoch": 5.872828126274574,
      "grad_norm": 0.005198000930249691,
      "learning_rate": 0.007345163516415448,
      "loss": 0.1775,
      "num_input_tokens_seen": 69286320,
      "step": 36000
    },
    {
      "epoch": 5.872828126274574,
      "eval_loss": 0.18946194648742676,
      "eval_runtime": 68.112,
      "eval_samples_per_second": 40.008,
      "eval_steps_per_second": 20.011,
      "num_input_tokens_seen": 69286320,
      "step": 36000
    },
    {
      "epoch": 5.873643853495391,
      "grad_norm": 0.0026150064077228308,
      "learning_rate": 0.007326967513089693,
      "loss": 0.1741,
      "num_input_tokens_seen": 69296240,
      "step": 36005
    },
    {
      "epoch": 5.874459580716208,
      "grad_norm": 0.005289120599627495,
      "learning_rate": 0.0073087935117384815,
      "loss": 0.2751,
      "num_input_tokens_seen": 69306032,
      "step": 36010
    },
    {
      "epoch": 5.8752753079370255,
      "grad_norm": 0.002471538493409753,
      "learning_rate": 0.007290641515164503,
      "loss": 0.2221,
      "num_input_tokens_seen": 69317312,
      "step": 36015
    },
    {
      "epoch": 5.8760910351578435,
      "grad_norm": 0.00393671216443181,
      "learning_rate": 0.007272511526166986,
      "loss": 0.1633,
      "num_input_tokens_seen": 69327952,
      "step": 36020
    },
    {
      "epoch": 5.876906762378661,
      "grad_norm": 0.003013392211869359,
      "learning_rate": 0.0072544035475418265,
      "loss": 0.1927,
      "num_input_tokens_seen": 69337728,
      "step": 36025
    },
    {
      "epoch": 5.877722489599478,
      "grad_norm": 0.0043539316393435,
      "learning_rate": 0.007236317582081475,
      "loss": 0.1141,
      "num_input_tokens_seen": 69346736,
      "step": 36030
    },
    {
      "epoch": 5.878538216820296,
      "grad_norm": 0.001937944209203124,
      "learning_rate": 0.007218253632575066,
      "loss": 0.1839,
      "num_input_tokens_seen": 69355456,
      "step": 36035
    },
    {
      "epoch": 5.879353944041113,
      "grad_norm": 0.00443069264292717,
      "learning_rate": 0.007200211701808223,
      "loss": 0.2018,
      "num_input_tokens_seen": 69366448,
      "step": 36040
    },
    {
      "epoch": 5.88016967126193,
      "grad_norm": 0.005945159122347832,
      "learning_rate": 0.007182191792563286,
      "loss": 0.2305,
      "num_input_tokens_seen": 69375488,
      "step": 36045
    },
    {
      "epoch": 5.880985398482747,
      "grad_norm": 0.0036820422392338514,
      "learning_rate": 0.0071641939076191145,
      "loss": 0.1663,
      "num_input_tokens_seen": 69384240,
      "step": 36050
    },
    {
      "epoch": 5.881801125703564,
      "grad_norm": 0.003124554641544819,
      "learning_rate": 0.007146218049751257,
      "loss": 0.1586,
      "num_input_tokens_seen": 69393120,
      "step": 36055
    },
    {
      "epoch": 5.882616852924382,
      "grad_norm": 0.004334988072514534,
      "learning_rate": 0.0071282642217317775,
      "loss": 0.1868,
      "num_input_tokens_seen": 69402928,
      "step": 36060
    },
    {
      "epoch": 5.883432580145199,
      "grad_norm": 0.0036374719347804785,
      "learning_rate": 0.007110332426329396,
      "loss": 0.1606,
      "num_input_tokens_seen": 69412992,
      "step": 36065
    },
    {
      "epoch": 5.884248307366017,
      "grad_norm": 0.0029962665867060423,
      "learning_rate": 0.007092422666309417,
      "loss": 0.1104,
      "num_input_tokens_seen": 69423200,
      "step": 36070
    },
    {
      "epoch": 5.885064034586835,
      "grad_norm": 0.0032021990045905113,
      "learning_rate": 0.0070745349444337295,
      "loss": 0.2009,
      "num_input_tokens_seen": 69433168,
      "step": 36075
    },
    {
      "epoch": 5.885879761807652,
      "grad_norm": 0.0058235228061676025,
      "learning_rate": 0.007056669263460913,
      "loss": 0.1602,
      "num_input_tokens_seen": 69442976,
      "step": 36080
    },
    {
      "epoch": 5.886695489028469,
      "grad_norm": 0.003468108829110861,
      "learning_rate": 0.007038825626145995,
      "loss": 0.1764,
      "num_input_tokens_seen": 69452960,
      "step": 36085
    },
    {
      "epoch": 5.887511216249286,
      "grad_norm": 0.003052909392863512,
      "learning_rate": 0.007021004035240724,
      "loss": 0.1596,
      "num_input_tokens_seen": 69462080,
      "step": 36090
    },
    {
      "epoch": 5.888326943470103,
      "grad_norm": 0.004722762852907181,
      "learning_rate": 0.007003204493493453,
      "loss": 0.1678,
      "num_input_tokens_seen": 69471616,
      "step": 36095
    },
    {
      "epoch": 5.889142670690921,
      "grad_norm": 0.0024459317792207003,
      "learning_rate": 0.006985427003649036,
      "loss": 0.1321,
      "num_input_tokens_seen": 69482048,
      "step": 36100
    },
    {
      "epoch": 5.889958397911738,
      "grad_norm": 0.003943111747503281,
      "learning_rate": 0.006967671568449013,
      "loss": 0.2194,
      "num_input_tokens_seen": 69492144,
      "step": 36105
    },
    {
      "epoch": 5.890774125132555,
      "grad_norm": 0.004775224719196558,
      "learning_rate": 0.006949938190631511,
      "loss": 0.1698,
      "num_input_tokens_seen": 69502256,
      "step": 36110
    },
    {
      "epoch": 5.891589852353373,
      "grad_norm": 0.0027379451785236597,
      "learning_rate": 0.0069322268729311905,
      "loss": 0.1958,
      "num_input_tokens_seen": 69511472,
      "step": 36115
    },
    {
      "epoch": 5.8924055795741905,
      "grad_norm": 0.0030590256210416555,
      "learning_rate": 0.006914537618079403,
      "loss": 0.2065,
      "num_input_tokens_seen": 69520768,
      "step": 36120
    },
    {
      "epoch": 5.893221306795008,
      "grad_norm": 0.0049463436007499695,
      "learning_rate": 0.006896870428804031,
      "loss": 0.1983,
      "num_input_tokens_seen": 69530304,
      "step": 36125
    },
    {
      "epoch": 5.894037034015825,
      "grad_norm": 0.0020315086003392935,
      "learning_rate": 0.006879225307829595,
      "loss": 0.1636,
      "num_input_tokens_seen": 69539872,
      "step": 36130
    },
    {
      "epoch": 5.894852761236643,
      "grad_norm": 0.004606280475854874,
      "learning_rate": 0.00686160225787717,
      "loss": 0.215,
      "num_input_tokens_seen": 69550224,
      "step": 36135
    },
    {
      "epoch": 5.89566848845746,
      "grad_norm": 0.0057294960133731365,
      "learning_rate": 0.006844001281664463,
      "loss": 0.2408,
      "num_input_tokens_seen": 69559936,
      "step": 36140
    },
    {
      "epoch": 5.896484215678277,
      "grad_norm": 0.005438629072159529,
      "learning_rate": 0.006826422381905789,
      "loss": 0.1414,
      "num_input_tokens_seen": 69568848,
      "step": 36145
    },
    {
      "epoch": 5.897299942899094,
      "grad_norm": 0.00538145424798131,
      "learning_rate": 0.006808865561311994,
      "loss": 0.1893,
      "num_input_tokens_seen": 69579824,
      "step": 36150
    },
    {
      "epoch": 5.898115670119912,
      "grad_norm": 0.001639998983591795,
      "learning_rate": 0.00679133082259058,
      "loss": 0.1657,
      "num_input_tokens_seen": 69590336,
      "step": 36155
    },
    {
      "epoch": 5.898931397340729,
      "grad_norm": 0.0015027917688712478,
      "learning_rate": 0.00677381816844565,
      "loss": 0.1629,
      "num_input_tokens_seen": 69601056,
      "step": 36160
    },
    {
      "epoch": 5.8997471245615465,
      "grad_norm": 0.0016403441550210118,
      "learning_rate": 0.0067563276015778434,
      "loss": 0.1697,
      "num_input_tokens_seen": 69609984,
      "step": 36165
    },
    {
      "epoch": 5.900562851782364,
      "grad_norm": 0.0031246719881892204,
      "learning_rate": 0.006738859124684437,
      "loss": 0.1438,
      "num_input_tokens_seen": 69619248,
      "step": 36170
    },
    {
      "epoch": 5.901378579003182,
      "grad_norm": 0.0035750609822571278,
      "learning_rate": 0.006721412740459259,
      "loss": 0.169,
      "num_input_tokens_seen": 69628304,
      "step": 36175
    },
    {
      "epoch": 5.902194306223999,
      "grad_norm": 0.004621054045855999,
      "learning_rate": 0.006703988451592824,
      "loss": 0.2586,
      "num_input_tokens_seen": 69637184,
      "step": 36180
    },
    {
      "epoch": 5.903010033444816,
      "grad_norm": 0.00410436699166894,
      "learning_rate": 0.006686586260772114,
      "loss": 0.1942,
      "num_input_tokens_seen": 69647664,
      "step": 36185
    },
    {
      "epoch": 5.903825760665633,
      "grad_norm": 0.0023279590532183647,
      "learning_rate": 0.006669206170680819,
      "loss": 0.1365,
      "num_input_tokens_seen": 69657728,
      "step": 36190
    },
    {
      "epoch": 5.904641487886451,
      "grad_norm": 0.007089971099048853,
      "learning_rate": 0.0066518481839991095,
      "loss": 0.1994,
      "num_input_tokens_seen": 69667952,
      "step": 36195
    },
    {
      "epoch": 5.905457215107268,
      "grad_norm": 0.003981493413448334,
      "learning_rate": 0.006634512303403861,
      "loss": 0.1688,
      "num_input_tokens_seen": 69676640,
      "step": 36200
    },
    {
      "epoch": 5.905457215107268,
      "eval_loss": 0.1888953000307083,
      "eval_runtime": 68.0352,
      "eval_samples_per_second": 40.053,
      "eval_steps_per_second": 20.034,
      "num_input_tokens_seen": 69676640,
      "step": 36200
    },
    {
      "epoch": 5.906272942328085,
      "grad_norm": 0.0031293381471186876,
      "learning_rate": 0.0066171985315684355,
      "loss": 0.1658,
      "num_input_tokens_seen": 69686592,
      "step": 36205
    },
    {
      "epoch": 5.907088669548903,
      "grad_norm": 0.00430701719596982,
      "learning_rate": 0.0065999068711628806,
      "loss": 0.151,
      "num_input_tokens_seen": 69695824,
      "step": 36210
    },
    {
      "epoch": 5.9079043967697205,
      "grad_norm": 0.0036480468697845936,
      "learning_rate": 0.0065826373248537295,
      "loss": 0.1543,
      "num_input_tokens_seen": 69704240,
      "step": 36215
    },
    {
      "epoch": 5.908720123990538,
      "grad_norm": 0.003323082346469164,
      "learning_rate": 0.006565389895304218,
      "loss": 0.1807,
      "num_input_tokens_seen": 69713584,
      "step": 36220
    },
    {
      "epoch": 5.909535851211355,
      "grad_norm": 0.0042383018881082535,
      "learning_rate": 0.006548164585174104,
      "loss": 0.1824,
      "num_input_tokens_seen": 69723328,
      "step": 36225
    },
    {
      "epoch": 5.910351578432172,
      "grad_norm": 0.0029988018795847893,
      "learning_rate": 0.006530961397119728,
      "loss": 0.17,
      "num_input_tokens_seen": 69732736,
      "step": 36230
    },
    {
      "epoch": 5.91116730565299,
      "grad_norm": 0.004626339767128229,
      "learning_rate": 0.00651378033379405,
      "loss": 0.2007,
      "num_input_tokens_seen": 69742528,
      "step": 36235
    },
    {
      "epoch": 5.911983032873807,
      "grad_norm": 0.003953075036406517,
      "learning_rate": 0.006496621397846619,
      "loss": 0.1434,
      "num_input_tokens_seen": 69751232,
      "step": 36240
    },
    {
      "epoch": 5.912798760094624,
      "grad_norm": 0.002808759920299053,
      "learning_rate": 0.006479484591923518,
      "loss": 0.1914,
      "num_input_tokens_seen": 69761984,
      "step": 36245
    },
    {
      "epoch": 5.913614487315442,
      "grad_norm": 0.0015285032568499446,
      "learning_rate": 0.006462369918667515,
      "loss": 0.224,
      "num_input_tokens_seen": 69771408,
      "step": 36250
    },
    {
      "epoch": 5.914430214536259,
      "grad_norm": 0.004187643527984619,
      "learning_rate": 0.006445277380717851,
      "loss": 0.1808,
      "num_input_tokens_seen": 69779920,
      "step": 36255
    },
    {
      "epoch": 5.915245941757076,
      "grad_norm": 0.004298658575862646,
      "learning_rate": 0.006428206980710466,
      "loss": 0.2398,
      "num_input_tokens_seen": 69789648,
      "step": 36260
    },
    {
      "epoch": 5.9160616689778935,
      "grad_norm": 0.0034784392919391394,
      "learning_rate": 0.006411158721277788,
      "loss": 0.1538,
      "num_input_tokens_seen": 69799936,
      "step": 36265
    },
    {
      "epoch": 5.916877396198711,
      "grad_norm": 0.004329151939600706,
      "learning_rate": 0.00639413260504888,
      "loss": 0.1938,
      "num_input_tokens_seen": 69810096,
      "step": 36270
    },
    {
      "epoch": 5.917693123419529,
      "grad_norm": 0.003297726158052683,
      "learning_rate": 0.006377128634649376,
      "loss": 0.1431,
      "num_input_tokens_seen": 69818256,
      "step": 36275
    },
    {
      "epoch": 5.918508850640346,
      "grad_norm": 0.003753983648493886,
      "learning_rate": 0.006360146812701528,
      "loss": 0.1985,
      "num_input_tokens_seen": 69827568,
      "step": 36280
    },
    {
      "epoch": 5.919324577861163,
      "grad_norm": 0.005120395682752132,
      "learning_rate": 0.006343187141824125,
      "loss": 0.1325,
      "num_input_tokens_seen": 69836944,
      "step": 36285
    },
    {
      "epoch": 5.920140305081981,
      "grad_norm": 0.003806112566962838,
      "learning_rate": 0.00632624962463259,
      "loss": 0.1497,
      "num_input_tokens_seen": 69845680,
      "step": 36290
    },
    {
      "epoch": 5.920956032302798,
      "grad_norm": 0.0011616316623985767,
      "learning_rate": 0.006309334263738853,
      "loss": 0.1866,
      "num_input_tokens_seen": 69855840,
      "step": 36295
    },
    {
      "epoch": 5.921771759523615,
      "grad_norm": 0.0018333412008360028,
      "learning_rate": 0.006292441061751508,
      "loss": 0.1468,
      "num_input_tokens_seen": 69866448,
      "step": 36300
    },
    {
      "epoch": 5.922587486744432,
      "grad_norm": 0.004174186848104,
      "learning_rate": 0.0062755700212757054,
      "loss": 0.215,
      "num_input_tokens_seen": 69875440,
      "step": 36305
    },
    {
      "epoch": 5.92340321396525,
      "grad_norm": 0.0035546983126550913,
      "learning_rate": 0.006258721144913148,
      "loss": 0.1536,
      "num_input_tokens_seen": 69885456,
      "step": 36310
    },
    {
      "epoch": 5.9242189411860675,
      "grad_norm": 0.004804834723472595,
      "learning_rate": 0.0062418944352621575,
      "loss": 0.2525,
      "num_input_tokens_seen": 69895600,
      "step": 36315
    },
    {
      "epoch": 5.925034668406885,
      "grad_norm": 0.003077478613704443,
      "learning_rate": 0.0062250898949176405,
      "loss": 0.1764,
      "num_input_tokens_seen": 69905680,
      "step": 36320
    },
    {
      "epoch": 5.925850395627702,
      "grad_norm": 0.005651208106428385,
      "learning_rate": 0.006208307526471041,
      "loss": 0.2048,
      "num_input_tokens_seen": 69915840,
      "step": 36325
    },
    {
      "epoch": 5.92666612284852,
      "grad_norm": 0.0067503065802156925,
      "learning_rate": 0.006191547332510405,
      "loss": 0.2722,
      "num_input_tokens_seen": 69924384,
      "step": 36330
    },
    {
      "epoch": 5.927481850069337,
      "grad_norm": 0.0032591905910521746,
      "learning_rate": 0.006174809315620416,
      "loss": 0.1647,
      "num_input_tokens_seen": 69934272,
      "step": 36335
    },
    {
      "epoch": 5.928297577290154,
      "grad_norm": 0.005286764353513718,
      "learning_rate": 0.00615809347838221,
      "loss": 0.195,
      "num_input_tokens_seen": 69944336,
      "step": 36340
    },
    {
      "epoch": 5.929113304510971,
      "grad_norm": 0.0037889466620981693,
      "learning_rate": 0.006141399823373655,
      "loss": 0.1891,
      "num_input_tokens_seen": 69953744,
      "step": 36345
    },
    {
      "epoch": 5.929929031731789,
      "grad_norm": 0.0034595176111906767,
      "learning_rate": 0.0061247283531690455,
      "loss": 0.1618,
      "num_input_tokens_seen": 69962224,
      "step": 36350
    },
    {
      "epoch": 5.930744758952606,
      "grad_norm": 0.0044448841363191605,
      "learning_rate": 0.0061080790703393895,
      "loss": 0.1928,
      "num_input_tokens_seen": 69972400,
      "step": 36355
    },
    {
      "epoch": 5.9315604861734235,
      "grad_norm": 0.002783673582598567,
      "learning_rate": 0.006091451977452217,
      "loss": 0.1838,
      "num_input_tokens_seen": 69982448,
      "step": 36360
    },
    {
      "epoch": 5.932376213394241,
      "grad_norm": 0.003585024969652295,
      "learning_rate": 0.00607484707707161,
      "loss": 0.1736,
      "num_input_tokens_seen": 69991968,
      "step": 36365
    },
    {
      "epoch": 5.933191940615059,
      "grad_norm": 0.0033879559487104416,
      "learning_rate": 0.006058264371758254,
      "loss": 0.2012,
      "num_input_tokens_seen": 70000880,
      "step": 36370
    },
    {
      "epoch": 5.934007667835876,
      "grad_norm": 0.006028353702276945,
      "learning_rate": 0.00604170386406942,
      "loss": 0.1594,
      "num_input_tokens_seen": 70009136,
      "step": 36375
    },
    {
      "epoch": 5.934823395056693,
      "grad_norm": 0.0037673774641007185,
      "learning_rate": 0.006025165556558931,
      "loss": 0.1963,
      "num_input_tokens_seen": 70018368,
      "step": 36380
    },
    {
      "epoch": 5.935639122277511,
      "grad_norm": 0.003868983592838049,
      "learning_rate": 0.006008649451777248,
      "loss": 0.1584,
      "num_input_tokens_seen": 70027024,
      "step": 36385
    },
    {
      "epoch": 5.936454849498328,
      "grad_norm": 0.001699886517599225,
      "learning_rate": 0.005992155552271283,
      "loss": 0.2203,
      "num_input_tokens_seen": 70037536,
      "step": 36390
    },
    {
      "epoch": 5.937270576719145,
      "grad_norm": 0.0012419366976246238,
      "learning_rate": 0.005975683860584685,
      "loss": 0.1399,
      "num_input_tokens_seen": 70047360,
      "step": 36395
    },
    {
      "epoch": 5.938086303939962,
      "grad_norm": 0.0045072236098349094,
      "learning_rate": 0.0059592343792575385,
      "loss": 0.1898,
      "num_input_tokens_seen": 70057024,
      "step": 36400
    },
    {
      "epoch": 5.938086303939962,
      "eval_loss": 0.19027212262153625,
      "eval_runtime": 67.9944,
      "eval_samples_per_second": 40.077,
      "eval_steps_per_second": 20.046,
      "num_input_tokens_seen": 70057024,
      "step": 36400
    },
    {
      "epoch": 5.938902031160779,
      "grad_norm": 0.0042731547728180885,
      "learning_rate": 0.0059428071108265975,
      "loss": 0.1563,
      "num_input_tokens_seen": 70067024,
      "step": 36405
    },
    {
      "epoch": 5.939717758381597,
      "grad_norm": 0.001753233140334487,
      "learning_rate": 0.005926402057825136,
      "loss": 0.1507,
      "num_input_tokens_seen": 70075968,
      "step": 36410
    },
    {
      "epoch": 5.940533485602415,
      "grad_norm": 0.002466679085046053,
      "learning_rate": 0.005910019222782997,
      "loss": 0.1649,
      "num_input_tokens_seen": 70085328,
      "step": 36415
    },
    {
      "epoch": 5.941349212823232,
      "grad_norm": 0.0035069608129560947,
      "learning_rate": 0.005893658608226643,
      "loss": 0.1994,
      "num_input_tokens_seen": 70094592,
      "step": 36420
    },
    {
      "epoch": 5.94216494004405,
      "grad_norm": 0.003950281068682671,
      "learning_rate": 0.0058773202166791045,
      "loss": 0.2601,
      "num_input_tokens_seen": 70104880,
      "step": 36425
    },
    {
      "epoch": 5.942980667264867,
      "grad_norm": 0.004077192861586809,
      "learning_rate": 0.005861004050659918,
      "loss": 0.1692,
      "num_input_tokens_seen": 70111904,
      "step": 36430
    },
    {
      "epoch": 5.943796394485684,
      "grad_norm": 0.0022349704522639513,
      "learning_rate": 0.005844710112685286,
      "loss": 0.1687,
      "num_input_tokens_seen": 70120224,
      "step": 36435
    },
    {
      "epoch": 5.944612121706501,
      "grad_norm": 0.001854253700003028,
      "learning_rate": 0.005828438405267933,
      "loss": 0.1074,
      "num_input_tokens_seen": 70129552,
      "step": 36440
    },
    {
      "epoch": 5.945427848927318,
      "grad_norm": 0.003744334215298295,
      "learning_rate": 0.00581218893091715,
      "loss": 0.229,
      "num_input_tokens_seen": 70140112,
      "step": 36445
    },
    {
      "epoch": 5.946243576148136,
      "grad_norm": 0.003920827526599169,
      "learning_rate": 0.005795961692138801,
      "loss": 0.1529,
      "num_input_tokens_seen": 70150624,
      "step": 36450
    },
    {
      "epoch": 5.947059303368953,
      "grad_norm": 0.00430606584995985,
      "learning_rate": 0.00577975669143535,
      "loss": 0.1706,
      "num_input_tokens_seen": 70159120,
      "step": 36455
    },
    {
      "epoch": 5.9478750305897705,
      "grad_norm": 0.0028367764316499233,
      "learning_rate": 0.005763573931305782,
      "loss": 0.1443,
      "num_input_tokens_seen": 70168752,
      "step": 36460
    },
    {
      "epoch": 5.9486907578105885,
      "grad_norm": 0.001941136084496975,
      "learning_rate": 0.005747413414245733,
      "loss": 0.1702,
      "num_input_tokens_seen": 70177824,
      "step": 36465
    },
    {
      "epoch": 5.949506485031406,
      "grad_norm": 0.0035129315219819546,
      "learning_rate": 0.005731275142747294,
      "loss": 0.1582,
      "num_input_tokens_seen": 70186688,
      "step": 36470
    },
    {
      "epoch": 5.950322212252223,
      "grad_norm": 0.002508878707885742,
      "learning_rate": 0.005715159119299256,
      "loss": 0.1421,
      "num_input_tokens_seen": 70196384,
      "step": 36475
    },
    {
      "epoch": 5.95113793947304,
      "grad_norm": 0.005016292445361614,
      "learning_rate": 0.005699065346386867,
      "loss": 0.2361,
      "num_input_tokens_seen": 70205584,
      "step": 36480
    },
    {
      "epoch": 5.951953666693858,
      "grad_norm": 0.0011769793927669525,
      "learning_rate": 0.0056829938264919885,
      "loss": 0.1511,
      "num_input_tokens_seen": 70215600,
      "step": 36485
    },
    {
      "epoch": 5.952769393914675,
      "grad_norm": 0.0029625596944242716,
      "learning_rate": 0.005666944562093074,
      "loss": 0.1461,
      "num_input_tokens_seen": 70225408,
      "step": 36490
    },
    {
      "epoch": 5.953585121135492,
      "grad_norm": 0.0034261527471244335,
      "learning_rate": 0.005650917555665108,
      "loss": 0.2124,
      "num_input_tokens_seen": 70234592,
      "step": 36495
    },
    {
      "epoch": 5.954400848356309,
      "grad_norm": 0.0020633793901652098,
      "learning_rate": 0.005634912809679632,
      "loss": 0.153,
      "num_input_tokens_seen": 70243024,
      "step": 36500
    },
    {
      "epoch": 5.955216575577127,
      "grad_norm": 0.001208813046105206,
      "learning_rate": 0.005618930326604854,
      "loss": 0.205,
      "num_input_tokens_seen": 70252672,
      "step": 36505
    },
    {
      "epoch": 5.9560323027979445,
      "grad_norm": 0.0008958920370787382,
      "learning_rate": 0.005602970108905386,
      "loss": 0.1118,
      "num_input_tokens_seen": 70262160,
      "step": 36510
    },
    {
      "epoch": 5.956848030018762,
      "grad_norm": 0.005152690224349499,
      "learning_rate": 0.005587032159042543,
      "loss": 0.1723,
      "num_input_tokens_seen": 70270304,
      "step": 36515
    },
    {
      "epoch": 5.957663757239579,
      "grad_norm": 0.0031527315732091665,
      "learning_rate": 0.005571116479474158,
      "loss": 0.1793,
      "num_input_tokens_seen": 70280592,
      "step": 36520
    },
    {
      "epoch": 5.958479484460397,
      "grad_norm": 0.0024866119492799044,
      "learning_rate": 0.005555223072654619,
      "loss": 0.1404,
      "num_input_tokens_seen": 70288800,
      "step": 36525
    },
    {
      "epoch": 5.959295211681214,
      "grad_norm": 0.0032960055395960808,
      "learning_rate": 0.005539351941034881,
      "loss": 0.1386,
      "num_input_tokens_seen": 70298608,
      "step": 36530
    },
    {
      "epoch": 5.960110938902031,
      "grad_norm": 0.0021191793493926525,
      "learning_rate": 0.0055235030870624865,
      "loss": 0.1565,
      "num_input_tokens_seen": 70309456,
      "step": 36535
    },
    {
      "epoch": 5.960926666122848,
      "grad_norm": 0.0021292115561664104,
      "learning_rate": 0.005507676513181514,
      "loss": 0.147,
      "num_input_tokens_seen": 70319168,
      "step": 36540
    },
    {
      "epoch": 5.961742393343666,
      "grad_norm": 0.002247861586511135,
      "learning_rate": 0.005491872221832628,
      "loss": 0.144,
      "num_input_tokens_seen": 70328832,
      "step": 36545
    },
    {
      "epoch": 5.962558120564483,
      "grad_norm": 0.005912154912948608,
      "learning_rate": 0.005476090215453061,
      "loss": 0.1924,
      "num_input_tokens_seen": 70337232,
      "step": 36550
    },
    {
      "epoch": 5.9633738477853,
      "grad_norm": 0.0021084833424538374,
      "learning_rate": 0.0054603304964765675,
      "loss": 0.1232,
      "num_input_tokens_seen": 70347504,
      "step": 36555
    },
    {
      "epoch": 5.964189575006118,
      "grad_norm": 0.0033528220374137163,
      "learning_rate": 0.005444593067333519,
      "loss": 0.2439,
      "num_input_tokens_seen": 70355600,
      "step": 36560
    },
    {
      "epoch": 5.965005302226936,
      "grad_norm": 0.0050276825204491615,
      "learning_rate": 0.00542887793045081,
      "loss": 0.1681,
      "num_input_tokens_seen": 70364768,
      "step": 36565
    },
    {
      "epoch": 5.965821029447753,
      "grad_norm": 0.00324783637188375,
      "learning_rate": 0.005413185088251932,
      "loss": 0.1843,
      "num_input_tokens_seen": 70374816,
      "step": 36570
    },
    {
      "epoch": 5.96663675666857,
      "grad_norm": 0.0016848599771037698,
      "learning_rate": 0.005397514543156884,
      "loss": 0.1377,
      "num_input_tokens_seen": 70385120,
      "step": 36575
    },
    {
      "epoch": 5.967452483889387,
      "grad_norm": 0.004355729557573795,
      "learning_rate": 0.0053818662975822825,
      "loss": 0.2334,
      "num_input_tokens_seen": 70394816,
      "step": 36580
    },
    {
      "epoch": 5.968268211110205,
      "grad_norm": 0.0026954535860568285,
      "learning_rate": 0.005366240353941315,
      "loss": 0.149,
      "num_input_tokens_seen": 70404000,
      "step": 36585
    },
    {
      "epoch": 5.969083938331022,
      "grad_norm": 0.004694532137364149,
      "learning_rate": 0.005350636714643636,
      "loss": 0.151,
      "num_input_tokens_seen": 70413008,
      "step": 36590
    },
    {
      "epoch": 5.969899665551839,
      "grad_norm": 0.002893917029723525,
      "learning_rate": 0.005335055382095555,
      "loss": 0.1458,
      "num_input_tokens_seen": 70422720,
      "step": 36595
    },
    {
      "epoch": 5.970715392772657,
      "grad_norm": 0.002762838965281844,
      "learning_rate": 0.005319496358699915,
      "loss": 0.1686,
      "num_input_tokens_seen": 70432848,
      "step": 36600
    },
    {
      "epoch": 5.970715392772657,
      "eval_loss": 0.18911083042621613,
      "eval_runtime": 67.9627,
      "eval_samples_per_second": 40.096,
      "eval_steps_per_second": 20.055,
      "num_input_tokens_seen": 70432848,
      "step": 36600
    },
    {
      "epoch": 5.971531119993474,
      "grad_norm": 0.005348046310245991,
      "learning_rate": 0.005303959646856099,
      "loss": 0.204,
      "num_input_tokens_seen": 70442848,
      "step": 36605
    },
    {
      "epoch": 5.9723468472142915,
      "grad_norm": 0.0033897908870130777,
      "learning_rate": 0.005288445248960089,
      "loss": 0.1874,
      "num_input_tokens_seen": 70452112,
      "step": 36610
    },
    {
      "epoch": 5.973162574435109,
      "grad_norm": 0.0030117565765976906,
      "learning_rate": 0.005272953167404354,
      "loss": 0.1227,
      "num_input_tokens_seen": 70461680,
      "step": 36615
    },
    {
      "epoch": 5.973978301655926,
      "grad_norm": 0.0031340434215962887,
      "learning_rate": 0.005257483404578017,
      "loss": 0.2283,
      "num_input_tokens_seen": 70471680,
      "step": 36620
    },
    {
      "epoch": 5.974794028876744,
      "grad_norm": 0.004771462641656399,
      "learning_rate": 0.0052420359628666865,
      "loss": 0.2152,
      "num_input_tokens_seen": 70480176,
      "step": 36625
    },
    {
      "epoch": 5.975609756097561,
      "grad_norm": 0.004143733065575361,
      "learning_rate": 0.00522661084465254,
      "loss": 0.2147,
      "num_input_tokens_seen": 70490528,
      "step": 36630
    },
    {
      "epoch": 5.976425483318378,
      "grad_norm": 0.00507740955799818,
      "learning_rate": 0.005211208052314326,
      "loss": 0.1368,
      "num_input_tokens_seen": 70500896,
      "step": 36635
    },
    {
      "epoch": 5.977241210539196,
      "grad_norm": 0.002502464223653078,
      "learning_rate": 0.005195827588227391,
      "loss": 0.1069,
      "num_input_tokens_seen": 70510656,
      "step": 36640
    },
    {
      "epoch": 5.978056937760013,
      "grad_norm": 0.005723175592720509,
      "learning_rate": 0.0051804694547635255,
      "loss": 0.2402,
      "num_input_tokens_seen": 70520176,
      "step": 36645
    },
    {
      "epoch": 5.97887266498083,
      "grad_norm": 0.0038065474946051836,
      "learning_rate": 0.005165133654291232,
      "loss": 0.1752,
      "num_input_tokens_seen": 70529808,
      "step": 36650
    },
    {
      "epoch": 5.9796883922016475,
      "grad_norm": 0.00470677250996232,
      "learning_rate": 0.005149820189175402,
      "loss": 0.2136,
      "num_input_tokens_seen": 70540064,
      "step": 36655
    },
    {
      "epoch": 5.9805041194224655,
      "grad_norm": 0.0024196552112698555,
      "learning_rate": 0.005134529061777598,
      "loss": 0.1543,
      "num_input_tokens_seen": 70550352,
      "step": 36660
    },
    {
      "epoch": 5.981319846643283,
      "grad_norm": 0.0036959382705390453,
      "learning_rate": 0.005119260274455933,
      "loss": 0.1784,
      "num_input_tokens_seen": 70559856,
      "step": 36665
    },
    {
      "epoch": 5.9821355738641,
      "grad_norm": 0.0035723161417990923,
      "learning_rate": 0.005104013829565007,
      "loss": 0.1548,
      "num_input_tokens_seen": 70568048,
      "step": 36670
    },
    {
      "epoch": 5.982951301084917,
      "grad_norm": 0.002787345787510276,
      "learning_rate": 0.005088789729456006,
      "loss": 0.2137,
      "num_input_tokens_seen": 70578640,
      "step": 36675
    },
    {
      "epoch": 5.983767028305735,
      "grad_norm": 0.003649797523394227,
      "learning_rate": 0.005073587976476735,
      "loss": 0.1858,
      "num_input_tokens_seen": 70587232,
      "step": 36680
    },
    {
      "epoch": 5.984582755526552,
      "grad_norm": 0.003000059397891164,
      "learning_rate": 0.005058408572971418,
      "loss": 0.1754,
      "num_input_tokens_seen": 70597088,
      "step": 36685
    },
    {
      "epoch": 5.985398482747369,
      "grad_norm": 0.003190078306943178,
      "learning_rate": 0.005043251521280983,
      "loss": 0.1827,
      "num_input_tokens_seen": 70606992,
      "step": 36690
    },
    {
      "epoch": 5.986214209968186,
      "grad_norm": 0.0043268827721476555,
      "learning_rate": 0.005028116823742795,
      "loss": 0.251,
      "num_input_tokens_seen": 70617440,
      "step": 36695
    },
    {
      "epoch": 5.987029937189004,
      "grad_norm": 0.0028178978245705366,
      "learning_rate": 0.005013004482690819,
      "loss": 0.1614,
      "num_input_tokens_seen": 70627088,
      "step": 36700
    },
    {
      "epoch": 5.9878456644098215,
      "grad_norm": 0.003527650609612465,
      "learning_rate": 0.0049979145004555746,
      "loss": 0.1703,
      "num_input_tokens_seen": 70637296,
      "step": 36705
    },
    {
      "epoch": 5.988661391630639,
      "grad_norm": 0.003183530643582344,
      "learning_rate": 0.004982846879364116,
      "loss": 0.2066,
      "num_input_tokens_seen": 70647104,
      "step": 36710
    },
    {
      "epoch": 5.989477118851456,
      "grad_norm": 0.0015626595122739673,
      "learning_rate": 0.0049678016217400535,
      "loss": 0.1751,
      "num_input_tokens_seen": 70656080,
      "step": 36715
    },
    {
      "epoch": 5.990292846072274,
      "grad_norm": 0.004298172891139984,
      "learning_rate": 0.004952778729903595,
      "loss": 0.2112,
      "num_input_tokens_seen": 70664832,
      "step": 36720
    },
    {
      "epoch": 5.991108573293091,
      "grad_norm": 0.0044592516496777534,
      "learning_rate": 0.004937778206171422,
      "loss": 0.1657,
      "num_input_tokens_seen": 70674400,
      "step": 36725
    },
    {
      "epoch": 5.991924300513908,
      "grad_norm": 0.006212940905243158,
      "learning_rate": 0.004922800052856835,
      "loss": 0.212,
      "num_input_tokens_seen": 70684848,
      "step": 36730
    },
    {
      "epoch": 5.992740027734725,
      "grad_norm": 0.004342318046838045,
      "learning_rate": 0.004907844272269602,
      "loss": 0.2146,
      "num_input_tokens_seen": 70694224,
      "step": 36735
    },
    {
      "epoch": 5.993555754955543,
      "grad_norm": 0.00505404407158494,
      "learning_rate": 0.004892910866716144,
      "loss": 0.2332,
      "num_input_tokens_seen": 70705168,
      "step": 36740
    },
    {
      "epoch": 5.99437148217636,
      "grad_norm": 0.002201486611738801,
      "learning_rate": 0.004877999838499369,
      "loss": 0.165,
      "num_input_tokens_seen": 70714576,
      "step": 36745
    },
    {
      "epoch": 5.995187209397177,
      "grad_norm": 0.002618944039568305,
      "learning_rate": 0.0048631111899187065,
      "loss": 0.1624,
      "num_input_tokens_seen": 70722832,
      "step": 36750
    },
    {
      "epoch": 5.9960029366179945,
      "grad_norm": 0.0035773003473877907,
      "learning_rate": 0.0048482449232702335,
      "loss": 0.2439,
      "num_input_tokens_seen": 70732944,
      "step": 36755
    },
    {
      "epoch": 5.996818663838813,
      "grad_norm": 0.0013411788968369365,
      "learning_rate": 0.004833401040846469,
      "loss": 0.154,
      "num_input_tokens_seen": 70742704,
      "step": 36760
    },
    {
      "epoch": 5.99763439105963,
      "grad_norm": 0.0036431618500500917,
      "learning_rate": 0.004818579544936546,
      "loss": 0.1572,
      "num_input_tokens_seen": 70752240,
      "step": 36765
    },
    {
      "epoch": 5.998450118280447,
      "grad_norm": 0.004000463057309389,
      "learning_rate": 0.004803780437826121,
      "loss": 0.1624,
      "num_input_tokens_seen": 70761120,
      "step": 36770
    },
    {
      "epoch": 5.999265845501265,
      "grad_norm": 0.004751286935061216,
      "learning_rate": 0.004789003721797402,
      "loss": 0.1479,
      "num_input_tokens_seen": 70769664,
      "step": 36775
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0036899035330861807,
      "learning_rate": 0.004774249399129132,
      "loss": 0.1499,
      "num_input_tokens_seen": 70778304,
      "step": 36780
    },
    {
      "epoch": 6.000815727220817,
      "grad_norm": 0.001420512329787016,
      "learning_rate": 0.004759517472096642,
      "loss": 0.1489,
      "num_input_tokens_seen": 70788096,
      "step": 36785
    },
    {
      "epoch": 6.001631454441635,
      "grad_norm": 0.007720894180238247,
      "learning_rate": 0.004744807942971746,
      "loss": 0.216,
      "num_input_tokens_seen": 70798704,
      "step": 36790
    },
    {
      "epoch": 6.002447181662452,
      "grad_norm": 0.0038862957153469324,
      "learning_rate": 0.004730120814022881,
      "loss": 0.1718,
      "num_input_tokens_seen": 70809536,
      "step": 36795
    },
    {
      "epoch": 6.003262908883269,
      "grad_norm": 0.002694937167689204,
      "learning_rate": 0.004715456087514935,
      "loss": 0.1376,
      "num_input_tokens_seen": 70819440,
      "step": 36800
    },
    {
      "epoch": 6.003262908883269,
      "eval_loss": 0.1896362006664276,
      "eval_runtime": 67.866,
      "eval_samples_per_second": 40.153,
      "eval_steps_per_second": 20.084,
      "num_input_tokens_seen": 70819440,
      "step": 36800
    },
    {
      "epoch": 6.0040786361040865,
      "grad_norm": 0.002710500964894891,
      "learning_rate": 0.004700813765709432,
      "loss": 0.1665,
      "num_input_tokens_seen": 70828736,
      "step": 36805
    },
    {
      "epoch": 6.004894363324905,
      "grad_norm": 0.004283488262444735,
      "learning_rate": 0.004686193850864401,
      "loss": 0.1416,
      "num_input_tokens_seen": 70838128,
      "step": 36810
    },
    {
      "epoch": 6.005710090545722,
      "grad_norm": 0.003533226205036044,
      "learning_rate": 0.004671596345234385,
      "loss": 0.1897,
      "num_input_tokens_seen": 70847104,
      "step": 36815
    },
    {
      "epoch": 6.006525817766539,
      "grad_norm": 0.0023738679010421038,
      "learning_rate": 0.00465702125107052,
      "loss": 0.1508,
      "num_input_tokens_seen": 70855520,
      "step": 36820
    },
    {
      "epoch": 6.007341544987356,
      "grad_norm": 0.0027017376851290464,
      "learning_rate": 0.004642468570620506,
      "loss": 0.2009,
      "num_input_tokens_seen": 70865984,
      "step": 36825
    },
    {
      "epoch": 6.008157272208174,
      "grad_norm": 0.0016013247659429908,
      "learning_rate": 0.004627938306128482,
      "loss": 0.1225,
      "num_input_tokens_seen": 70874800,
      "step": 36830
    },
    {
      "epoch": 6.008972999428991,
      "grad_norm": 0.002159580122679472,
      "learning_rate": 0.004613430459835255,
      "loss": 0.1854,
      "num_input_tokens_seen": 70882912,
      "step": 36835
    },
    {
      "epoch": 6.009788726649808,
      "grad_norm": 0.003279642900452018,
      "learning_rate": 0.004598945033978085,
      "loss": 0.1973,
      "num_input_tokens_seen": 70892960,
      "step": 36840
    },
    {
      "epoch": 6.010604453870625,
      "grad_norm": 0.003214013995602727,
      "learning_rate": 0.004584482030790804,
      "loss": 0.2579,
      "num_input_tokens_seen": 70901744,
      "step": 36845
    },
    {
      "epoch": 6.011420181091443,
      "grad_norm": 0.003612313885241747,
      "learning_rate": 0.004570041452503826,
      "loss": 0.1919,
      "num_input_tokens_seen": 70911872,
      "step": 36850
    },
    {
      "epoch": 6.0122359083122605,
      "grad_norm": 0.0036009333562105894,
      "learning_rate": 0.004555623301344003,
      "loss": 0.1715,
      "num_input_tokens_seen": 70921696,
      "step": 36855
    },
    {
      "epoch": 6.013051635533078,
      "grad_norm": 0.0057976809330284595,
      "learning_rate": 0.004541227579534857,
      "loss": 0.2033,
      "num_input_tokens_seen": 70932000,
      "step": 36860
    },
    {
      "epoch": 6.013867362753895,
      "grad_norm": 0.003441634587943554,
      "learning_rate": 0.004526854289296378,
      "loss": 0.2106,
      "num_input_tokens_seen": 70942496,
      "step": 36865
    },
    {
      "epoch": 6.014683089974713,
      "grad_norm": 0.003823932260274887,
      "learning_rate": 0.004512503432845078,
      "loss": 0.1798,
      "num_input_tokens_seen": 70953536,
      "step": 36870
    },
    {
      "epoch": 6.01549881719553,
      "grad_norm": 0.0028949284460395575,
      "learning_rate": 0.004498175012394068,
      "loss": 0.186,
      "num_input_tokens_seen": 70963952,
      "step": 36875
    },
    {
      "epoch": 6.016314544416347,
      "grad_norm": 0.0031017435248941183,
      "learning_rate": 0.004483869030152965,
      "loss": 0.1472,
      "num_input_tokens_seen": 70974400,
      "step": 36880
    },
    {
      "epoch": 6.017130271637164,
      "grad_norm": 0.002253173151984811,
      "learning_rate": 0.004469585488327904,
      "loss": 0.1421,
      "num_input_tokens_seen": 70985296,
      "step": 36885
    },
    {
      "epoch": 6.017945998857982,
      "grad_norm": 0.004351011943072081,
      "learning_rate": 0.0044553243891216395,
      "loss": 0.1953,
      "num_input_tokens_seen": 70993104,
      "step": 36890
    },
    {
      "epoch": 6.018761726078799,
      "grad_norm": 0.002449772786349058,
      "learning_rate": 0.004441085734733363,
      "loss": 0.1927,
      "num_input_tokens_seen": 71002528,
      "step": 36895
    },
    {
      "epoch": 6.0195774532996165,
      "grad_norm": 0.0022860034368932247,
      "learning_rate": 0.004426869527358884,
      "loss": 0.1405,
      "num_input_tokens_seen": 71011712,
      "step": 36900
    },
    {
      "epoch": 6.020393180520434,
      "grad_norm": 0.00310168182477355,
      "learning_rate": 0.0044126757691905156,
      "loss": 0.1502,
      "num_input_tokens_seen": 71020528,
      "step": 36905
    },
    {
      "epoch": 6.021208907741252,
      "grad_norm": 0.002070291433483362,
      "learning_rate": 0.004398504462417107,
      "loss": 0.1759,
      "num_input_tokens_seen": 71029680,
      "step": 36910
    },
    {
      "epoch": 6.022024634962069,
      "grad_norm": 0.0024722563102841377,
      "learning_rate": 0.0043843556092240605,
      "loss": 0.1512,
      "num_input_tokens_seen": 71037536,
      "step": 36915
    },
    {
      "epoch": 6.022840362182886,
      "grad_norm": 0.002266376744955778,
      "learning_rate": 0.004370229211793281,
      "loss": 0.1278,
      "num_input_tokens_seen": 71046480,
      "step": 36920
    },
    {
      "epoch": 6.023656089403703,
      "grad_norm": 0.004192430526018143,
      "learning_rate": 0.0043561252723032405,
      "loss": 0.1871,
      "num_input_tokens_seen": 71056592,
      "step": 36925
    },
    {
      "epoch": 6.024471816624521,
      "grad_norm": 0.002189414110034704,
      "learning_rate": 0.004342043792929001,
      "loss": 0.1353,
      "num_input_tokens_seen": 71065952,
      "step": 36930
    },
    {
      "epoch": 6.025287543845338,
      "grad_norm": 0.004039567895233631,
      "learning_rate": 0.004327984775842025,
      "loss": 0.1847,
      "num_input_tokens_seen": 71076672,
      "step": 36935
    },
    {
      "epoch": 6.026103271066155,
      "grad_norm": 0.0038762183394283056,
      "learning_rate": 0.004313948223210428,
      "loss": 0.1588,
      "num_input_tokens_seen": 71085760,
      "step": 36940
    },
    {
      "epoch": 6.026918998286972,
      "grad_norm": 0.0036606958601623774,
      "learning_rate": 0.004299934137198846,
      "loss": 0.1434,
      "num_input_tokens_seen": 71095920,
      "step": 36945
    },
    {
      "epoch": 6.02773472550779,
      "grad_norm": 0.001716837752610445,
      "learning_rate": 0.004285942519968383,
      "loss": 0.1919,
      "num_input_tokens_seen": 71105264,
      "step": 36950
    },
    {
      "epoch": 6.028550452728608,
      "grad_norm": 0.003358869580551982,
      "learning_rate": 0.004271973373676746,
      "loss": 0.1659,
      "num_input_tokens_seen": 71114736,
      "step": 36955
    },
    {
      "epoch": 6.029366179949425,
      "grad_norm": 0.0032239363063126802,
      "learning_rate": 0.004258026700478146,
      "loss": 0.2075,
      "num_input_tokens_seen": 71124736,
      "step": 36960
    },
    {
      "epoch": 6.030181907170242,
      "grad_norm": 0.005596606992185116,
      "learning_rate": 0.004244102502523328,
      "loss": 0.208,
      "num_input_tokens_seen": 71135136,
      "step": 36965
    },
    {
      "epoch": 6.03099763439106,
      "grad_norm": 0.001504253363236785,
      "learning_rate": 0.004230200781959592,
      "loss": 0.1349,
      "num_input_tokens_seen": 71143680,
      "step": 36970
    },
    {
      "epoch": 6.031813361611877,
      "grad_norm": 0.003716210601851344,
      "learning_rate": 0.004216321540930756,
      "loss": 0.1101,
      "num_input_tokens_seen": 71154176,
      "step": 36975
    },
    {
      "epoch": 6.032629088832694,
      "grad_norm": 0.003423453774303198,
      "learning_rate": 0.004202464781577175,
      "loss": 0.205,
      "num_input_tokens_seen": 71163104,
      "step": 36980
    },
    {
      "epoch": 6.033444816053512,
      "grad_norm": 0.004298251587897539,
      "learning_rate": 0.00418863050603574,
      "loss": 0.2168,
      "num_input_tokens_seen": 71173616,
      "step": 36985
    },
    {
      "epoch": 6.034260543274329,
      "grad_norm": 0.0031303141731768847,
      "learning_rate": 0.004174818716439843,
      "loss": 0.1631,
      "num_input_tokens_seen": 71182832,
      "step": 36990
    },
    {
      "epoch": 6.035076270495146,
      "grad_norm": 0.0018603057833388448,
      "learning_rate": 0.004161029414919464,
      "loss": 0.1582,
      "num_input_tokens_seen": 71193280,
      "step": 36995
    },
    {
      "epoch": 6.0358919977159635,
      "grad_norm": 0.005142897367477417,
      "learning_rate": 0.004147262603601071,
      "loss": 0.2545,
      "num_input_tokens_seen": 71203008,
      "step": 37000
    },
    {
      "epoch": 6.0358919977159635,
      "eval_loss": 0.18986192345619202,
      "eval_runtime": 67.9235,
      "eval_samples_per_second": 40.119,
      "eval_steps_per_second": 20.067,
      "num_input_tokens_seen": 71203008,
      "step": 37000
    },
    {
      "epoch": 6.0367077249367815,
      "grad_norm": 0.00363901280798018,
      "learning_rate": 0.004133518284607679,
      "loss": 0.243,
      "num_input_tokens_seen": 71210928,
      "step": 37005
    },
    {
      "epoch": 6.037523452157599,
      "grad_norm": 0.003350635524839163,
      "learning_rate": 0.004119796460058861,
      "loss": 0.1471,
      "num_input_tokens_seen": 71220064,
      "step": 37010
    },
    {
      "epoch": 6.038339179378416,
      "grad_norm": 0.004590899683535099,
      "learning_rate": 0.00410609713207064,
      "loss": 0.1764,
      "num_input_tokens_seen": 71230384,
      "step": 37015
    },
    {
      "epoch": 6.039154906599233,
      "grad_norm": 0.0032076933421194553,
      "learning_rate": 0.004092420302755678,
      "loss": 0.1804,
      "num_input_tokens_seen": 71240672,
      "step": 37020
    },
    {
      "epoch": 6.039970633820051,
      "grad_norm": 0.005502693355083466,
      "learning_rate": 0.004078765974223103,
      "loss": 0.202,
      "num_input_tokens_seen": 71250528,
      "step": 37025
    },
    {
      "epoch": 6.040786361040868,
      "grad_norm": 0.0064291334711015224,
      "learning_rate": 0.004065134148578564,
      "loss": 0.2356,
      "num_input_tokens_seen": 71262000,
      "step": 37030
    },
    {
      "epoch": 6.041602088261685,
      "grad_norm": 0.002739806892350316,
      "learning_rate": 0.004051524827924279,
      "loss": 0.1549,
      "num_input_tokens_seen": 71272176,
      "step": 37035
    },
    {
      "epoch": 6.042417815482502,
      "grad_norm": 0.0026458697393536568,
      "learning_rate": 0.004037938014358955,
      "loss": 0.1767,
      "num_input_tokens_seen": 71280784,
      "step": 37040
    },
    {
      "epoch": 6.04323354270332,
      "grad_norm": 0.0017240585293620825,
      "learning_rate": 0.004024373709977863,
      "loss": 0.1058,
      "num_input_tokens_seen": 71290576,
      "step": 37045
    },
    {
      "epoch": 6.0440492699241375,
      "grad_norm": 0.004433442838490009,
      "learning_rate": 0.004010831916872814,
      "loss": 0.1996,
      "num_input_tokens_seen": 71301216,
      "step": 37050
    },
    {
      "epoch": 6.044864997144955,
      "grad_norm": 0.003360553877428174,
      "learning_rate": 0.003997312637132089,
      "loss": 0.1321,
      "num_input_tokens_seen": 71310096,
      "step": 37055
    },
    {
      "epoch": 6.045680724365772,
      "grad_norm": 0.004593903664499521,
      "learning_rate": 0.003983815872840535,
      "loss": 0.1758,
      "num_input_tokens_seen": 71321200,
      "step": 37060
    },
    {
      "epoch": 6.04649645158659,
      "grad_norm": 0.0028568364214152098,
      "learning_rate": 0.003970341626079521,
      "loss": 0.1391,
      "num_input_tokens_seen": 71331552,
      "step": 37065
    },
    {
      "epoch": 6.047312178807407,
      "grad_norm": 0.0064766607247292995,
      "learning_rate": 0.003956889898926952,
      "loss": 0.1663,
      "num_input_tokens_seen": 71340672,
      "step": 37070
    },
    {
      "epoch": 6.048127906028224,
      "grad_norm": 0.0031406281050294638,
      "learning_rate": 0.0039434606934572675,
      "loss": 0.2054,
      "num_input_tokens_seen": 71350688,
      "step": 37075
    },
    {
      "epoch": 6.048943633249041,
      "grad_norm": 0.0066606830805540085,
      "learning_rate": 0.003930054011741396,
      "loss": 0.1809,
      "num_input_tokens_seen": 71360800,
      "step": 37080
    },
    {
      "epoch": 6.049759360469859,
      "grad_norm": 0.006153131369501352,
      "learning_rate": 0.0039166698558468155,
      "loss": 0.1577,
      "num_input_tokens_seen": 71371344,
      "step": 37085
    },
    {
      "epoch": 6.050575087690676,
      "grad_norm": 0.00200170767493546,
      "learning_rate": 0.0039033082278375594,
      "loss": 0.1532,
      "num_input_tokens_seen": 71380912,
      "step": 37090
    },
    {
      "epoch": 6.051390814911493,
      "grad_norm": 0.003793652169406414,
      "learning_rate": 0.003889969129774112,
      "loss": 0.1475,
      "num_input_tokens_seen": 71391504,
      "step": 37095
    },
    {
      "epoch": 6.052206542132311,
      "grad_norm": 0.0012959401356056333,
      "learning_rate": 0.0038766525637135784,
      "loss": 0.1318,
      "num_input_tokens_seen": 71400112,
      "step": 37100
    },
    {
      "epoch": 6.053022269353129,
      "grad_norm": 0.0026892770547419786,
      "learning_rate": 0.0038633585317095318,
      "loss": 0.255,
      "num_input_tokens_seen": 71408832,
      "step": 37105
    },
    {
      "epoch": 6.053837996573946,
      "grad_norm": 0.004056860227137804,
      "learning_rate": 0.00385008703581205,
      "loss": 0.2506,
      "num_input_tokens_seen": 71418096,
      "step": 37110
    },
    {
      "epoch": 6.054653723794763,
      "grad_norm": 0.0035931614693254232,
      "learning_rate": 0.0038368380780677944,
      "loss": 0.1683,
      "num_input_tokens_seen": 71427056,
      "step": 37115
    },
    {
      "epoch": 6.05546945101558,
      "grad_norm": 0.002075619762763381,
      "learning_rate": 0.003823611660519882,
      "loss": 0.1759,
      "num_input_tokens_seen": 71436272,
      "step": 37120
    },
    {
      "epoch": 6.056285178236398,
      "grad_norm": 0.001791460206732154,
      "learning_rate": 0.0038104077852080475,
      "loss": 0.1196,
      "num_input_tokens_seen": 71445808,
      "step": 37125
    },
    {
      "epoch": 6.057100905457215,
      "grad_norm": 0.0016963094240054488,
      "learning_rate": 0.003797226454168462,
      "loss": 0.1281,
      "num_input_tokens_seen": 71455408,
      "step": 37130
    },
    {
      "epoch": 6.057916632678032,
      "grad_norm": 0.004062681458890438,
      "learning_rate": 0.003784067669433849,
      "loss": 0.1389,
      "num_input_tokens_seen": 71464960,
      "step": 37135
    },
    {
      "epoch": 6.058732359898849,
      "grad_norm": 0.002620939165353775,
      "learning_rate": 0.0037709314330334528,
      "loss": 0.1855,
      "num_input_tokens_seen": 71475696,
      "step": 37140
    },
    {
      "epoch": 6.059548087119667,
      "grad_norm": 0.0022655942011624575,
      "learning_rate": 0.003757817746993086,
      "loss": 0.1547,
      "num_input_tokens_seen": 71486368,
      "step": 37145
    },
    {
      "epoch": 6.0603638143404845,
      "grad_norm": 0.002806575270369649,
      "learning_rate": 0.0037447266133349977,
      "loss": 0.1941,
      "num_input_tokens_seen": 71496368,
      "step": 37150
    },
    {
      "epoch": 6.061179541561302,
      "grad_norm": 0.00289393262937665,
      "learning_rate": 0.003731658034078039,
      "loss": 0.1895,
      "num_input_tokens_seen": 71505712,
      "step": 37155
    },
    {
      "epoch": 6.06199526878212,
      "grad_norm": 0.0017023858381435275,
      "learning_rate": 0.0037186120112375153,
      "loss": 0.1447,
      "num_input_tokens_seen": 71515968,
      "step": 37160
    },
    {
      "epoch": 6.062810996002937,
      "grad_norm": 0.0027696944307535887,
      "learning_rate": 0.003705588546825317,
      "loss": 0.1585,
      "num_input_tokens_seen": 71525824,
      "step": 37165
    },
    {
      "epoch": 6.063626723223754,
      "grad_norm": 0.004345159977674484,
      "learning_rate": 0.0036925876428498205,
      "loss": 0.1742,
      "num_input_tokens_seen": 71535712,
      "step": 37170
    },
    {
      "epoch": 6.064442450444571,
      "grad_norm": 0.0027736215852200985,
      "learning_rate": 0.0036796093013159057,
      "loss": 0.1649,
      "num_input_tokens_seen": 71544224,
      "step": 37175
    },
    {
      "epoch": 6.065258177665389,
      "grad_norm": 0.0013211274053901434,
      "learning_rate": 0.0036666535242250217,
      "loss": 0.1629,
      "num_input_tokens_seen": 71552992,
      "step": 37180
    },
    {
      "epoch": 6.066073904886206,
      "grad_norm": 0.004810591693967581,
      "learning_rate": 0.003653720313575104,
      "loss": 0.1013,
      "num_input_tokens_seen": 71562176,
      "step": 37185
    },
    {
      "epoch": 6.066889632107023,
      "grad_norm": 0.002985366154462099,
      "learning_rate": 0.003640809671360623,
      "loss": 0.1119,
      "num_input_tokens_seen": 71570256,
      "step": 37190
    },
    {
      "epoch": 6.0677053593278405,
      "grad_norm": 0.002674244809895754,
      "learning_rate": 0.003627921599572553,
      "loss": 0.1662,
      "num_input_tokens_seen": 71579392,
      "step": 37195
    },
    {
      "epoch": 6.0685210865486585,
      "grad_norm": 0.004655764903873205,
      "learning_rate": 0.003615056100198405,
      "loss": 0.2171,
      "num_input_tokens_seen": 71588672,
      "step": 37200
    },
    {
      "epoch": 6.0685210865486585,
      "eval_loss": 0.18957319855690002,
      "eval_runtime": 68.0635,
      "eval_samples_per_second": 40.036,
      "eval_steps_per_second": 20.025,
      "num_input_tokens_seen": 71588672,
      "step": 37200
    },
    {
      "epoch": 6.069336813769476,
      "grad_norm": 0.002319245832040906,
      "learning_rate": 0.003602213175222174,
      "loss": 0.1209,
      "num_input_tokens_seen": 71597856,
      "step": 37205
    },
    {
      "epoch": 6.070152540990293,
      "grad_norm": 0.0032303479965776205,
      "learning_rate": 0.0035893928266244432,
      "loss": 0.1233,
      "num_input_tokens_seen": 71607360,
      "step": 37210
    },
    {
      "epoch": 6.07096826821111,
      "grad_norm": 0.005706100724637508,
      "learning_rate": 0.003576595056382248,
      "loss": 0.1892,
      "num_input_tokens_seen": 71616368,
      "step": 37215
    },
    {
      "epoch": 6.071783995431928,
      "grad_norm": 0.002164996461942792,
      "learning_rate": 0.0035638198664691423,
      "loss": 0.1527,
      "num_input_tokens_seen": 71626864,
      "step": 37220
    },
    {
      "epoch": 6.072599722652745,
      "grad_norm": 0.006587284151464701,
      "learning_rate": 0.003551067258855267,
      "loss": 0.181,
      "num_input_tokens_seen": 71635472,
      "step": 37225
    },
    {
      "epoch": 6.073415449873562,
      "grad_norm": 0.003972281701862812,
      "learning_rate": 0.0035383372355071996,
      "loss": 0.1346,
      "num_input_tokens_seen": 71645168,
      "step": 37230
    },
    {
      "epoch": 6.074231177094379,
      "grad_norm": 0.0033235710579901934,
      "learning_rate": 0.0035256297983881023,
      "loss": 0.1728,
      "num_input_tokens_seen": 71653776,
      "step": 37235
    },
    {
      "epoch": 6.075046904315197,
      "grad_norm": 0.0017353140283375978,
      "learning_rate": 0.0035129449494575747,
      "loss": 0.1879,
      "num_input_tokens_seen": 71662368,
      "step": 37240
    },
    {
      "epoch": 6.0758626315360145,
      "grad_norm": 0.005642223637551069,
      "learning_rate": 0.0035002826906718187,
      "loss": 0.1883,
      "num_input_tokens_seen": 71671520,
      "step": 37245
    },
    {
      "epoch": 6.076678358756832,
      "grad_norm": 0.0032239099964499474,
      "learning_rate": 0.003487643023983522,
      "loss": 0.1642,
      "num_input_tokens_seen": 71681344,
      "step": 37250
    },
    {
      "epoch": 6.077494085977649,
      "grad_norm": 0.005894879810512066,
      "learning_rate": 0.003475025951341842,
      "loss": 0.2239,
      "num_input_tokens_seen": 71690624,
      "step": 37255
    },
    {
      "epoch": 6.078309813198467,
      "grad_norm": 0.0039960043504834175,
      "learning_rate": 0.00346243147469249,
      "loss": 0.1497,
      "num_input_tokens_seen": 71700592,
      "step": 37260
    },
    {
      "epoch": 6.079125540419284,
      "grad_norm": 0.0029254157561808825,
      "learning_rate": 0.0034498595959777446,
      "loss": 0.1126,
      "num_input_tokens_seen": 71710112,
      "step": 37265
    },
    {
      "epoch": 6.079941267640101,
      "grad_norm": 0.003278814721852541,
      "learning_rate": 0.003437310317136305,
      "loss": 0.1663,
      "num_input_tokens_seen": 71720464,
      "step": 37270
    },
    {
      "epoch": 6.080756994860918,
      "grad_norm": 0.003771445946767926,
      "learning_rate": 0.0034247836401034236,
      "loss": 0.1921,
      "num_input_tokens_seen": 71730240,
      "step": 37275
    },
    {
      "epoch": 6.081572722081736,
      "grad_norm": 0.004517978522926569,
      "learning_rate": 0.003412279566810905,
      "loss": 0.1488,
      "num_input_tokens_seen": 71739888,
      "step": 37280
    },
    {
      "epoch": 6.082388449302553,
      "grad_norm": 0.002942361868917942,
      "learning_rate": 0.00339979809918699,
      "loss": 0.1166,
      "num_input_tokens_seen": 71749552,
      "step": 37285
    },
    {
      "epoch": 6.08320417652337,
      "grad_norm": 0.0036615589633584023,
      "learning_rate": 0.0033873392391565228,
      "loss": 0.1894,
      "num_input_tokens_seen": 71759424,
      "step": 37290
    },
    {
      "epoch": 6.0840199037441876,
      "grad_norm": 0.004953215830028057,
      "learning_rate": 0.003374902988640782,
      "loss": 0.1779,
      "num_input_tokens_seen": 71768080,
      "step": 37295
    },
    {
      "epoch": 6.084835630965006,
      "grad_norm": 0.003184567205607891,
      "learning_rate": 0.0033624893495576014,
      "loss": 0.1742,
      "num_input_tokens_seen": 71778208,
      "step": 37300
    },
    {
      "epoch": 6.085651358185823,
      "grad_norm": 0.001999346539378166,
      "learning_rate": 0.0033500983238213323,
      "loss": 0.165,
      "num_input_tokens_seen": 71787552,
      "step": 37305
    },
    {
      "epoch": 6.08646708540664,
      "grad_norm": 0.0026480129454284906,
      "learning_rate": 0.0033377299133428126,
      "loss": 0.1577,
      "num_input_tokens_seen": 71797888,
      "step": 37310
    },
    {
      "epoch": 6.087282812627457,
      "grad_norm": 0.004819502122700214,
      "learning_rate": 0.003325384120029434,
      "loss": 0.1745,
      "num_input_tokens_seen": 71807616,
      "step": 37315
    },
    {
      "epoch": 6.088098539848275,
      "grad_norm": 0.004974628333002329,
      "learning_rate": 0.0033130609457850233,
      "loss": 0.2607,
      "num_input_tokens_seen": 71817984,
      "step": 37320
    },
    {
      "epoch": 6.088914267069092,
      "grad_norm": 0.0027007984463125467,
      "learning_rate": 0.0033007603925100104,
      "loss": 0.1749,
      "num_input_tokens_seen": 71826768,
      "step": 37325
    },
    {
      "epoch": 6.089729994289909,
      "grad_norm": 0.003740281332284212,
      "learning_rate": 0.003288482462101294,
      "loss": 0.1805,
      "num_input_tokens_seen": 71837328,
      "step": 37330
    },
    {
      "epoch": 6.090545721510727,
      "grad_norm": 0.0032924695406109095,
      "learning_rate": 0.0032762271564522605,
      "loss": 0.1888,
      "num_input_tokens_seen": 71846528,
      "step": 37335
    },
    {
      "epoch": 6.091361448731544,
      "grad_norm": 0.004580095410346985,
      "learning_rate": 0.003263994477452864,
      "loss": 0.1509,
      "num_input_tokens_seen": 71855280,
      "step": 37340
    },
    {
      "epoch": 6.0921771759523615,
      "grad_norm": 0.0026850062422454357,
      "learning_rate": 0.0032517844269895125,
      "loss": 0.1298,
      "num_input_tokens_seen": 71864656,
      "step": 37345
    },
    {
      "epoch": 6.092992903173179,
      "grad_norm": 0.003047914244234562,
      "learning_rate": 0.0032395970069451496,
      "loss": 0.1671,
      "num_input_tokens_seen": 71874400,
      "step": 37350
    },
    {
      "epoch": 6.093808630393997,
      "grad_norm": 0.003255313029512763,
      "learning_rate": 0.0032274322191992388,
      "loss": 0.182,
      "num_input_tokens_seen": 71883984,
      "step": 37355
    },
    {
      "epoch": 6.094624357614814,
      "grad_norm": 0.004535042680799961,
      "learning_rate": 0.0032152900656277294,
      "loss": 0.1714,
      "num_input_tokens_seen": 71893856,
      "step": 37360
    },
    {
      "epoch": 6.095440084835631,
      "grad_norm": 0.004134556278586388,
      "learning_rate": 0.0032031705481030902,
      "loss": 0.198,
      "num_input_tokens_seen": 71904368,
      "step": 37365
    },
    {
      "epoch": 6.096255812056448,
      "grad_norm": 0.003512757597491145,
      "learning_rate": 0.0031910736684943428,
      "loss": 0.1271,
      "num_input_tokens_seen": 71913408,
      "step": 37370
    },
    {
      "epoch": 6.097071539277266,
      "grad_norm": 0.0059973676688969135,
      "learning_rate": 0.0031789994286669453,
      "loss": 0.1764,
      "num_input_tokens_seen": 71923296,
      "step": 37375
    },
    {
      "epoch": 6.097887266498083,
      "grad_norm": 0.002000564243644476,
      "learning_rate": 0.003166947830482908,
      "loss": 0.1283,
      "num_input_tokens_seen": 71934032,
      "step": 37380
    },
    {
      "epoch": 6.0987029937189,
      "grad_norm": 0.004406568128615618,
      "learning_rate": 0.003154918875800727,
      "loss": 0.1476,
      "num_input_tokens_seen": 71943488,
      "step": 37385
    },
    {
      "epoch": 6.0995187209397175,
      "grad_norm": 0.0032352199777960777,
      "learning_rate": 0.00314291256647542,
      "loss": 0.1933,
      "num_input_tokens_seen": 71953616,
      "step": 37390
    },
    {
      "epoch": 6.1003344481605355,
      "grad_norm": 0.0029684538021683693,
      "learning_rate": 0.0031309289043585375,
      "loss": 0.1924,
      "num_input_tokens_seen": 71962928,
      "step": 37395
    },
    {
      "epoch": 6.101150175381353,
      "grad_norm": 0.004417944233864546,
      "learning_rate": 0.003118967891298069,
      "loss": 0.1954,
      "num_input_tokens_seen": 71972608,
      "step": 37400
    },
    {
      "epoch": 6.101150175381353,
      "eval_loss": 0.18973438441753387,
      "eval_runtime": 68.076,
      "eval_samples_per_second": 40.029,
      "eval_steps_per_second": 20.022,
      "num_input_tokens_seen": 71972608,
      "step": 37400
    },
    {
      "epoch": 6.10196590260217,
      "grad_norm": 0.002047650283202529,
      "learning_rate": 0.003107029529138572,
      "loss": 0.0988,
      "num_input_tokens_seen": 71981168,
      "step": 37405
    },
    {
      "epoch": 6.102781629822987,
      "grad_norm": 0.0019011578988283873,
      "learning_rate": 0.0030951138197211235,
      "loss": 0.2072,
      "num_input_tokens_seen": 71991280,
      "step": 37410
    },
    {
      "epoch": 6.103597357043805,
      "grad_norm": 0.0036860136315226555,
      "learning_rate": 0.0030832207648832377,
      "loss": 0.202,
      "num_input_tokens_seen": 72001664,
      "step": 37415
    },
    {
      "epoch": 6.104413084264622,
      "grad_norm": 0.00432739919051528,
      "learning_rate": 0.0030713503664589635,
      "loss": 0.1475,
      "num_input_tokens_seen": 72010560,
      "step": 37420
    },
    {
      "epoch": 6.105228811485439,
      "grad_norm": 0.001551561988890171,
      "learning_rate": 0.0030595026262788872,
      "loss": 0.122,
      "num_input_tokens_seen": 72019392,
      "step": 37425
    },
    {
      "epoch": 6.106044538706256,
      "grad_norm": 0.003279993776232004,
      "learning_rate": 0.00304767754617008,
      "loss": 0.2239,
      "num_input_tokens_seen": 72028224,
      "step": 37430
    },
    {
      "epoch": 6.106860265927074,
      "grad_norm": 0.0032507700379937887,
      "learning_rate": 0.003035875127956117,
      "loss": 0.1454,
      "num_input_tokens_seen": 72037968,
      "step": 37435
    },
    {
      "epoch": 6.1076759931478914,
      "grad_norm": 0.005237598903477192,
      "learning_rate": 0.0030240953734570752,
      "loss": 0.164,
      "num_input_tokens_seen": 72047888,
      "step": 37440
    },
    {
      "epoch": 6.108491720368709,
      "grad_norm": 0.0024340117815881968,
      "learning_rate": 0.003012338284489535,
      "loss": 0.1869,
      "num_input_tokens_seen": 72056208,
      "step": 37445
    },
    {
      "epoch": 6.109307447589526,
      "grad_norm": 0.004239236935973167,
      "learning_rate": 0.0030006038628665964,
      "loss": 0.1249,
      "num_input_tokens_seen": 72064944,
      "step": 37450
    },
    {
      "epoch": 6.110123174810344,
      "grad_norm": 0.004626347683370113,
      "learning_rate": 0.002988892110397845,
      "loss": 0.231,
      "num_input_tokens_seen": 72074448,
      "step": 37455
    },
    {
      "epoch": 6.110938902031161,
      "grad_norm": 0.0027593092527240515,
      "learning_rate": 0.0029772030288894025,
      "loss": 0.1682,
      "num_input_tokens_seen": 72084352,
      "step": 37460
    },
    {
      "epoch": 6.111754629251978,
      "grad_norm": 0.0027509303763508797,
      "learning_rate": 0.0029655366201438438,
      "loss": 0.11,
      "num_input_tokens_seen": 72093776,
      "step": 37465
    },
    {
      "epoch": 6.112570356472795,
      "grad_norm": 0.0031682001426815987,
      "learning_rate": 0.0029538928859602965,
      "loss": 0.1071,
      "num_input_tokens_seen": 72102208,
      "step": 37470
    },
    {
      "epoch": 6.113386083693613,
      "grad_norm": 0.0034253187477588654,
      "learning_rate": 0.002942271828134374,
      "loss": 0.1754,
      "num_input_tokens_seen": 72111488,
      "step": 37475
    },
    {
      "epoch": 6.11420181091443,
      "grad_norm": 0.0022569652646780014,
      "learning_rate": 0.00293067344845816,
      "loss": 0.0846,
      "num_input_tokens_seen": 72121968,
      "step": 37480
    },
    {
      "epoch": 6.115017538135247,
      "grad_norm": 0.001439932850189507,
      "learning_rate": 0.0029190977487202896,
      "loss": 0.2156,
      "num_input_tokens_seen": 72132832,
      "step": 37485
    },
    {
      "epoch": 6.1158332653560645,
      "grad_norm": 0.0030130003578960896,
      "learning_rate": 0.0029075447307058853,
      "loss": 0.2271,
      "num_input_tokens_seen": 72142448,
      "step": 37490
    },
    {
      "epoch": 6.1166489925768825,
      "grad_norm": 0.0061128949746489525,
      "learning_rate": 0.0028960143961965722,
      "loss": 0.1537,
      "num_input_tokens_seen": 72151504,
      "step": 37495
    },
    {
      "epoch": 6.1174647197977,
      "grad_norm": 0.004209107253700495,
      "learning_rate": 0.002884506746970461,
      "loss": 0.1448,
      "num_input_tokens_seen": 72161472,
      "step": 37500
    },
    {
      "epoch": 6.118280447018517,
      "grad_norm": 0.008421089500188828,
      "learning_rate": 0.0028730217848021654,
      "loss": 0.198,
      "num_input_tokens_seen": 72171056,
      "step": 37505
    },
    {
      "epoch": 6.119096174239334,
      "grad_norm": 0.0024166973307728767,
      "learning_rate": 0.0028615595114628188,
      "loss": 0.2326,
      "num_input_tokens_seen": 72180816,
      "step": 37510
    },
    {
      "epoch": 6.119911901460152,
      "grad_norm": 0.003982226364314556,
      "learning_rate": 0.002850119928720074,
      "loss": 0.1371,
      "num_input_tokens_seen": 72190800,
      "step": 37515
    },
    {
      "epoch": 6.120727628680969,
      "grad_norm": 0.0037298344541341066,
      "learning_rate": 0.0028387030383380195,
      "loss": 0.1492,
      "num_input_tokens_seen": 72200224,
      "step": 37520
    },
    {
      "epoch": 6.121543355901786,
      "grad_norm": 0.003947227727621794,
      "learning_rate": 0.0028273088420772974,
      "loss": 0.1905,
      "num_input_tokens_seen": 72210912,
      "step": 37525
    },
    {
      "epoch": 6.122359083122603,
      "grad_norm": 0.0020210715010762215,
      "learning_rate": 0.002815937341695068,
      "loss": 0.1234,
      "num_input_tokens_seen": 72220960,
      "step": 37530
    },
    {
      "epoch": 6.123174810343421,
      "grad_norm": 0.0033138927537947893,
      "learning_rate": 0.0028045885389448963,
      "loss": 0.1883,
      "num_input_tokens_seen": 72231696,
      "step": 37535
    },
    {
      "epoch": 6.1239905375642385,
      "grad_norm": 0.002980302320793271,
      "learning_rate": 0.002793262435576965,
      "loss": 0.1813,
      "num_input_tokens_seen": 72240736,
      "step": 37540
    },
    {
      "epoch": 6.124806264785056,
      "grad_norm": 0.001997096696868539,
      "learning_rate": 0.0027819590333378772,
      "loss": 0.1037,
      "num_input_tokens_seen": 72250736,
      "step": 37545
    },
    {
      "epoch": 6.125621992005874,
      "grad_norm": 0.002945441985502839,
      "learning_rate": 0.002770678333970755,
      "loss": 0.2297,
      "num_input_tokens_seen": 72260272,
      "step": 37550
    },
    {
      "epoch": 6.126437719226691,
      "grad_norm": 0.003061858005821705,
      "learning_rate": 0.0027594203392152573,
      "loss": 0.1525,
      "num_input_tokens_seen": 72270432,
      "step": 37555
    },
    {
      "epoch": 6.127253446447508,
      "grad_norm": 0.0012727414723485708,
      "learning_rate": 0.002748185050807478,
      "loss": 0.137,
      "num_input_tokens_seen": 72279840,
      "step": 37560
    },
    {
      "epoch": 6.128069173668325,
      "grad_norm": 0.006617271341383457,
      "learning_rate": 0.002736972470480031,
      "loss": 0.1388,
      "num_input_tokens_seen": 72290784,
      "step": 37565
    },
    {
      "epoch": 6.128884900889143,
      "grad_norm": 0.001701225759461522,
      "learning_rate": 0.002725782599962068,
      "loss": 0.1842,
      "num_input_tokens_seen": 72301952,
      "step": 37570
    },
    {
      "epoch": 6.12970062810996,
      "grad_norm": 0.0021506233606487513,
      "learning_rate": 0.0027146154409791734,
      "loss": 0.1477,
      "num_input_tokens_seen": 72310704,
      "step": 37575
    },
    {
      "epoch": 6.130516355330777,
      "grad_norm": 0.004984232131391764,
      "learning_rate": 0.002703470995253504,
      "loss": 0.2002,
      "num_input_tokens_seen": 72319808,
      "step": 37580
    },
    {
      "epoch": 6.1313320825515945,
      "grad_norm": 0.004258709028363228,
      "learning_rate": 0.0026923492645036184,
      "loss": 0.173,
      "num_input_tokens_seen": 72329776,
      "step": 37585
    },
    {
      "epoch": 6.1321478097724125,
      "grad_norm": 0.003358955029398203,
      "learning_rate": 0.0026812502504446776,
      "loss": 0.1607,
      "num_input_tokens_seen": 72339424,
      "step": 37590
    },
    {
      "epoch": 6.13296353699323,
      "grad_norm": 0.0036002476699650288,
      "learning_rate": 0.0026701739547882798,
      "loss": 0.1653,
      "num_input_tokens_seen": 72348624,
      "step": 37595
    },
    {
      "epoch": 6.133779264214047,
      "grad_norm": 0.002141600241884589,
      "learning_rate": 0.0026591203792425077,
      "loss": 0.1308,
      "num_input_tokens_seen": 72358032,
      "step": 37600
    },
    {
      "epoch": 6.133779264214047,
      "eval_loss": 0.19082003831863403,
      "eval_runtime": 68.0647,
      "eval_samples_per_second": 40.035,
      "eval_steps_per_second": 20.025,
      "num_input_tokens_seen": 72358032,
      "step": 37600
    },
    {
      "epoch": 6.134594991434864,
      "grad_norm": 0.007148091681301594,
      "learning_rate": 0.0026480895255119818,
      "loss": 0.2133,
      "num_input_tokens_seen": 72366992,
      "step": 37605
    },
    {
      "epoch": 6.135410718655682,
      "grad_norm": 0.002802611095830798,
      "learning_rate": 0.002637081395297791,
      "loss": 0.1531,
      "num_input_tokens_seen": 72376112,
      "step": 37610
    },
    {
      "epoch": 6.136226445876499,
      "grad_norm": 0.002678866498172283,
      "learning_rate": 0.0026260959902975113,
      "loss": 0.2165,
      "num_input_tokens_seen": 72386656,
      "step": 37615
    },
    {
      "epoch": 6.137042173097316,
      "grad_norm": 0.0017429935978725553,
      "learning_rate": 0.00261513331220527,
      "loss": 0.1054,
      "num_input_tokens_seen": 72395536,
      "step": 37620
    },
    {
      "epoch": 6.137857900318133,
      "grad_norm": 0.0025860369205474854,
      "learning_rate": 0.0026041933627116154,
      "loss": 0.1207,
      "num_input_tokens_seen": 72405408,
      "step": 37625
    },
    {
      "epoch": 6.138673627538951,
      "grad_norm": 0.003315342590212822,
      "learning_rate": 0.0025932761435036476,
      "loss": 0.1437,
      "num_input_tokens_seen": 72416240,
      "step": 37630
    },
    {
      "epoch": 6.139489354759768,
      "grad_norm": 0.004453039728105068,
      "learning_rate": 0.002582381656264904,
      "loss": 0.1542,
      "num_input_tokens_seen": 72426896,
      "step": 37635
    },
    {
      "epoch": 6.1403050819805856,
      "grad_norm": 0.0037301117554306984,
      "learning_rate": 0.0025715099026754895,
      "loss": 0.1357,
      "num_input_tokens_seen": 72437952,
      "step": 37640
    },
    {
      "epoch": 6.141120809201403,
      "grad_norm": 0.0023339539766311646,
      "learning_rate": 0.002560660884411947,
      "loss": 0.1104,
      "num_input_tokens_seen": 72447552,
      "step": 37645
    },
    {
      "epoch": 6.141936536422221,
      "grad_norm": 0.0040056416764855385,
      "learning_rate": 0.0025498346031473385,
      "loss": 0.1878,
      "num_input_tokens_seen": 72456800,
      "step": 37650
    },
    {
      "epoch": 6.142752263643038,
      "grad_norm": 0.004500031936913729,
      "learning_rate": 0.0025390310605511945,
      "loss": 0.1655,
      "num_input_tokens_seen": 72467520,
      "step": 37655
    },
    {
      "epoch": 6.143567990863855,
      "grad_norm": 0.003903232282027602,
      "learning_rate": 0.0025282502582895995,
      "loss": 0.1385,
      "num_input_tokens_seen": 72476816,
      "step": 37660
    },
    {
      "epoch": 6.144383718084672,
      "grad_norm": 0.0032236964907497168,
      "learning_rate": 0.002517492198025023,
      "loss": 0.2048,
      "num_input_tokens_seen": 72486208,
      "step": 37665
    },
    {
      "epoch": 6.14519944530549,
      "grad_norm": 0.003943697549402714,
      "learning_rate": 0.0025067568814165554,
      "loss": 0.2458,
      "num_input_tokens_seen": 72496896,
      "step": 37670
    },
    {
      "epoch": 6.146015172526307,
      "grad_norm": 0.004864557646214962,
      "learning_rate": 0.0024960443101196884,
      "loss": 0.1582,
      "num_input_tokens_seen": 72506816,
      "step": 37675
    },
    {
      "epoch": 6.146830899747124,
      "grad_norm": 0.004080497194081545,
      "learning_rate": 0.002485354485786434,
      "loss": 0.2103,
      "num_input_tokens_seen": 72516528,
      "step": 37680
    },
    {
      "epoch": 6.1476466269679415,
      "grad_norm": 0.002835991093888879,
      "learning_rate": 0.002474687410065307,
      "loss": 0.172,
      "num_input_tokens_seen": 72526656,
      "step": 37685
    },
    {
      "epoch": 6.1484623541887595,
      "grad_norm": 0.0029169348999857903,
      "learning_rate": 0.002464043084601308,
      "loss": 0.191,
      "num_input_tokens_seen": 72536320,
      "step": 37690
    },
    {
      "epoch": 6.149278081409577,
      "grad_norm": 0.003398802364245057,
      "learning_rate": 0.0024534215110358915,
      "loss": 0.2,
      "num_input_tokens_seen": 72545392,
      "step": 37695
    },
    {
      "epoch": 6.150093808630394,
      "grad_norm": 0.005518733989447355,
      "learning_rate": 0.002442822691007096,
      "loss": 0.1696,
      "num_input_tokens_seen": 72555840,
      "step": 37700
    },
    {
      "epoch": 6.150909535851211,
      "grad_norm": 0.0038920934312045574,
      "learning_rate": 0.002432246626149348,
      "loss": 0.1613,
      "num_input_tokens_seen": 72565232,
      "step": 37705
    },
    {
      "epoch": 6.151725263072029,
      "grad_norm": 0.004409842658787966,
      "learning_rate": 0.002421693318093626,
      "loss": 0.1653,
      "num_input_tokens_seen": 72574944,
      "step": 37710
    },
    {
      "epoch": 6.152540990292846,
      "grad_norm": 0.003225113032385707,
      "learning_rate": 0.0024111627684673784,
      "loss": 0.2103,
      "num_input_tokens_seen": 72585088,
      "step": 37715
    },
    {
      "epoch": 6.153356717513663,
      "grad_norm": 0.0032271447125822306,
      "learning_rate": 0.0024006549788945395,
      "loss": 0.1592,
      "num_input_tokens_seen": 72595920,
      "step": 37720
    },
    {
      "epoch": 6.154172444734481,
      "grad_norm": 0.003982819151133299,
      "learning_rate": 0.0023901699509955463,
      "loss": 0.2018,
      "num_input_tokens_seen": 72605872,
      "step": 37725
    },
    {
      "epoch": 6.154988171955298,
      "grad_norm": 0.002584730740636587,
      "learning_rate": 0.0023797076863873554,
      "loss": 0.1474,
      "num_input_tokens_seen": 72615328,
      "step": 37730
    },
    {
      "epoch": 6.1558038991761155,
      "grad_norm": 0.0019780436996370554,
      "learning_rate": 0.0023692681866833262,
      "loss": 0.2,
      "num_input_tokens_seen": 72625168,
      "step": 37735
    },
    {
      "epoch": 6.156619626396933,
      "grad_norm": 0.0038476912304759026,
      "learning_rate": 0.0023588514534934046,
      "loss": 0.1457,
      "num_input_tokens_seen": 72634128,
      "step": 37740
    },
    {
      "epoch": 6.157435353617751,
      "grad_norm": 0.004370734095573425,
      "learning_rate": 0.002348457488423955,
      "loss": 0.1678,
      "num_input_tokens_seen": 72643712,
      "step": 37745
    },
    {
      "epoch": 6.158251080838568,
      "grad_norm": 0.004309885203838348,
      "learning_rate": 0.0023380862930778624,
      "loss": 0.1976,
      "num_input_tokens_seen": 72653984,
      "step": 37750
    },
    {
      "epoch": 6.159066808059385,
      "grad_norm": 0.0016284673474729061,
      "learning_rate": 0.0023277378690545135,
      "loss": 0.138,
      "num_input_tokens_seen": 72664272,
      "step": 37755
    },
    {
      "epoch": 6.159882535280202,
      "grad_norm": 0.004352119751274586,
      "learning_rate": 0.0023174122179497325,
      "loss": 0.1861,
      "num_input_tokens_seen": 72673232,
      "step": 37760
    },
    {
      "epoch": 6.16069826250102,
      "grad_norm": 0.0033487938344478607,
      "learning_rate": 0.0023071093413558784,
      "loss": 0.1536,
      "num_input_tokens_seen": 72682848,
      "step": 37765
    },
    {
      "epoch": 6.161513989721837,
      "grad_norm": 0.0032277347054332495,
      "learning_rate": 0.002296829240861814,
      "loss": 0.1282,
      "num_input_tokens_seen": 72693344,
      "step": 37770
    },
    {
      "epoch": 6.162329716942654,
      "grad_norm": 0.004644942004233599,
      "learning_rate": 0.002286571918052821,
      "loss": 0.1709,
      "num_input_tokens_seen": 72702448,
      "step": 37775
    },
    {
      "epoch": 6.163145444163471,
      "grad_norm": 0.0032771555706858635,
      "learning_rate": 0.0022763373745107174,
      "loss": 0.125,
      "num_input_tokens_seen": 72712208,
      "step": 37780
    },
    {
      "epoch": 6.1639611713842895,
      "grad_norm": 0.003547802334651351,
      "learning_rate": 0.0022661256118138074,
      "loss": 0.1577,
      "num_input_tokens_seen": 72722064,
      "step": 37785
    },
    {
      "epoch": 6.164776898605107,
      "grad_norm": 0.003872285597026348,
      "learning_rate": 0.0022559366315368645,
      "loss": 0.188,
      "num_input_tokens_seen": 72730864,
      "step": 37790
    },
    {
      "epoch": 6.165592625825924,
      "grad_norm": 0.004212879575788975,
      "learning_rate": 0.002245770435251182,
      "loss": 0.144,
      "num_input_tokens_seen": 72740576,
      "step": 37795
    },
    {
      "epoch": 6.166408353046741,
      "grad_norm": 0.0036281109787523746,
      "learning_rate": 0.002235627024524456,
      "loss": 0.1851,
      "num_input_tokens_seen": 72749840,
      "step": 37800
    },
    {
      "epoch": 6.166408353046741,
      "eval_loss": 0.19016657769680023,
      "eval_runtime": 68.1879,
      "eval_samples_per_second": 39.963,
      "eval_steps_per_second": 19.989,
      "num_input_tokens_seen": 72749840,
      "step": 37800
    },
    {
      "epoch": 6.167224080267559,
      "grad_norm": 0.0030936403200030327,
      "learning_rate": 0.0022255064009209847,
      "loss": 0.105,
      "num_input_tokens_seen": 72759584,
      "step": 37805
    },
    {
      "epoch": 6.168039807488376,
      "grad_norm": 0.004596849903464317,
      "learning_rate": 0.0022154085660014864,
      "loss": 0.1864,
      "num_input_tokens_seen": 72768576,
      "step": 37810
    },
    {
      "epoch": 6.168855534709193,
      "grad_norm": 0.002748755970969796,
      "learning_rate": 0.0022053335213231494,
      "loss": 0.128,
      "num_input_tokens_seen": 72777216,
      "step": 37815
    },
    {
      "epoch": 6.16967126193001,
      "grad_norm": 0.0019122972153127193,
      "learning_rate": 0.002195281268439697,
      "loss": 0.1707,
      "num_input_tokens_seen": 72787584,
      "step": 37820
    },
    {
      "epoch": 6.170486989150828,
      "grad_norm": 0.001666420721448958,
      "learning_rate": 0.002185251808901306,
      "loss": 0.1895,
      "num_input_tokens_seen": 72797328,
      "step": 37825
    },
    {
      "epoch": 6.171302716371645,
      "grad_norm": 0.004518695641309023,
      "learning_rate": 0.0021752451442546227,
      "loss": 0.241,
      "num_input_tokens_seen": 72807056,
      "step": 37830
    },
    {
      "epoch": 6.1721184435924625,
      "grad_norm": 0.0032987084705382586,
      "learning_rate": 0.0021652612760428456,
      "loss": 0.1394,
      "num_input_tokens_seen": 72816288,
      "step": 37835
    },
    {
      "epoch": 6.17293417081328,
      "grad_norm": 0.006710343062877655,
      "learning_rate": 0.0021553002058055603,
      "loss": 0.1653,
      "num_input_tokens_seen": 72825888,
      "step": 37840
    },
    {
      "epoch": 6.173749898034098,
      "grad_norm": 0.002576206810772419,
      "learning_rate": 0.0021453619350789376,
      "loss": 0.2254,
      "num_input_tokens_seen": 72835232,
      "step": 37845
    },
    {
      "epoch": 6.174565625254915,
      "grad_norm": 0.003842925187200308,
      "learning_rate": 0.0021354464653955516,
      "loss": 0.1342,
      "num_input_tokens_seen": 72845168,
      "step": 37850
    },
    {
      "epoch": 6.175381352475732,
      "grad_norm": 0.0029517831280827522,
      "learning_rate": 0.002125553798284513,
      "loss": 0.1615,
      "num_input_tokens_seen": 72854240,
      "step": 37855
    },
    {
      "epoch": 6.176197079696549,
      "grad_norm": 0.0026658123824745417,
      "learning_rate": 0.002115683935271384,
      "loss": 0.0804,
      "num_input_tokens_seen": 72863024,
      "step": 37860
    },
    {
      "epoch": 6.177012806917367,
      "grad_norm": 0.002726521110162139,
      "learning_rate": 0.0021058368778782144,
      "loss": 0.1429,
      "num_input_tokens_seen": 72873056,
      "step": 37865
    },
    {
      "epoch": 6.177828534138184,
      "grad_norm": 0.002374907024204731,
      "learning_rate": 0.002096012627623539,
      "loss": 0.1659,
      "num_input_tokens_seen": 72882768,
      "step": 37870
    },
    {
      "epoch": 6.178644261359001,
      "grad_norm": 0.002998692449182272,
      "learning_rate": 0.00208621118602243,
      "loss": 0.1475,
      "num_input_tokens_seen": 72892832,
      "step": 37875
    },
    {
      "epoch": 6.1794599885798185,
      "grad_norm": 0.007285616360604763,
      "learning_rate": 0.002076432554586327,
      "loss": 0.1566,
      "num_input_tokens_seen": 72902480,
      "step": 37880
    },
    {
      "epoch": 6.1802757158006365,
      "grad_norm": 0.0025497903116047382,
      "learning_rate": 0.002066676734823258,
      "loss": 0.181,
      "num_input_tokens_seen": 72912208,
      "step": 37885
    },
    {
      "epoch": 6.181091443021454,
      "grad_norm": 0.0036531202495098114,
      "learning_rate": 0.0020569437282376866,
      "loss": 0.1982,
      "num_input_tokens_seen": 72920992,
      "step": 37890
    },
    {
      "epoch": 6.181907170242271,
      "grad_norm": 0.002863998059183359,
      "learning_rate": 0.002047233536330545,
      "loss": 0.1765,
      "num_input_tokens_seen": 72930880,
      "step": 37895
    },
    {
      "epoch": 6.182722897463089,
      "grad_norm": 0.0032289260998368263,
      "learning_rate": 0.0020375461605993015,
      "loss": 0.1374,
      "num_input_tokens_seen": 72940512,
      "step": 37900
    },
    {
      "epoch": 6.183538624683906,
      "grad_norm": 0.010536914691329002,
      "learning_rate": 0.002027881602537845,
      "loss": 0.1637,
      "num_input_tokens_seen": 72950144,
      "step": 37905
    },
    {
      "epoch": 6.184354351904723,
      "grad_norm": 0.0028331344947218895,
      "learning_rate": 0.002018239863636567,
      "loss": 0.1902,
      "num_input_tokens_seen": 72960832,
      "step": 37910
    },
    {
      "epoch": 6.18517007912554,
      "grad_norm": 0.0042658220045268536,
      "learning_rate": 0.002008620945382378,
      "loss": 0.1297,
      "num_input_tokens_seen": 72969392,
      "step": 37915
    },
    {
      "epoch": 6.185985806346358,
      "grad_norm": 0.0024727347772568464,
      "learning_rate": 0.001999024849258607,
      "loss": 0.1965,
      "num_input_tokens_seen": 72978512,
      "step": 37920
    },
    {
      "epoch": 6.186801533567175,
      "grad_norm": 0.004509910009801388,
      "learning_rate": 0.001989451576745105,
      "loss": 0.1428,
      "num_input_tokens_seen": 72988032,
      "step": 37925
    },
    {
      "epoch": 6.1876172607879925,
      "grad_norm": 0.0029739905148744583,
      "learning_rate": 0.00197990112931819,
      "loss": 0.1795,
      "num_input_tokens_seen": 72997360,
      "step": 37930
    },
    {
      "epoch": 6.18843298800881,
      "grad_norm": 0.004052302334457636,
      "learning_rate": 0.0019703735084506345,
      "loss": 0.1621,
      "num_input_tokens_seen": 73007088,
      "step": 37935
    },
    {
      "epoch": 6.189248715229628,
      "grad_norm": 0.0037408247590065002,
      "learning_rate": 0.001960868715611763,
      "loss": 0.2376,
      "num_input_tokens_seen": 73016064,
      "step": 37940
    },
    {
      "epoch": 6.190064442450445,
      "grad_norm": 0.004488054662942886,
      "learning_rate": 0.0019513867522673034,
      "loss": 0.1852,
      "num_input_tokens_seen": 73026656,
      "step": 37945
    },
    {
      "epoch": 6.190880169671262,
      "grad_norm": 0.0038412397261708975,
      "learning_rate": 0.001941927619879502,
      "loss": 0.1955,
      "num_input_tokens_seen": 73037088,
      "step": 37950
    },
    {
      "epoch": 6.191695896892079,
      "grad_norm": 0.0077491034753620625,
      "learning_rate": 0.0019324913199070758,
      "loss": 0.1751,
      "num_input_tokens_seen": 73046544,
      "step": 37955
    },
    {
      "epoch": 6.192511624112897,
      "grad_norm": 0.0032791313715279102,
      "learning_rate": 0.0019230778538052106,
      "loss": 0.1907,
      "num_input_tokens_seen": 73054640,
      "step": 37960
    },
    {
      "epoch": 6.193327351333714,
      "grad_norm": 0.002601781627163291,
      "learning_rate": 0.0019136872230255952,
      "loss": 0.1503,
      "num_input_tokens_seen": 73062160,
      "step": 37965
    },
    {
      "epoch": 6.194143078554531,
      "grad_norm": 0.0023440730292350054,
      "learning_rate": 0.0019043194290164045,
      "loss": 0.1394,
      "num_input_tokens_seen": 73071152,
      "step": 37970
    },
    {
      "epoch": 6.194958805775348,
      "grad_norm": 0.003796919481828809,
      "learning_rate": 0.0018949744732222162,
      "loss": 0.2211,
      "num_input_tokens_seen": 73079808,
      "step": 37975
    },
    {
      "epoch": 6.195774532996166,
      "grad_norm": 0.005654464941471815,
      "learning_rate": 0.0018856523570841776,
      "loss": 0.2319,
      "num_input_tokens_seen": 73090320,
      "step": 37980
    },
    {
      "epoch": 6.196590260216984,
      "grad_norm": 0.004629909992218018,
      "learning_rate": 0.0018763530820398555,
      "loss": 0.2178,
      "num_input_tokens_seen": 73099200,
      "step": 37985
    },
    {
      "epoch": 6.197405987437801,
      "grad_norm": 0.0036548487842082977,
      "learning_rate": 0.0018670766495233525,
      "loss": 0.1093,
      "num_input_tokens_seen": 73109456,
      "step": 37990
    },
    {
      "epoch": 6.198221714658618,
      "grad_norm": 0.003919289913028479,
      "learning_rate": 0.001857823060965158,
      "loss": 0.1971,
      "num_input_tokens_seen": 73118912,
      "step": 37995
    },
    {
      "epoch": 6.199037441879436,
      "grad_norm": 0.0036966237239539623,
      "learning_rate": 0.0018485923177923467,
      "loss": 0.1399,
      "num_input_tokens_seen": 73128448,
      "step": 38000
    },
    {
      "epoch": 6.199037441879436,
      "eval_loss": 0.19015613198280334,
      "eval_runtime": 67.9553,
      "eval_samples_per_second": 40.1,
      "eval_steps_per_second": 20.057,
      "num_input_tokens_seen": 73128448,
      "step": 38000
    },
    {
      "epoch": 6.199853169100253,
      "grad_norm": 0.0054941326379776,
      "learning_rate": 0.001839384421428364,
      "loss": 0.2136,
      "num_input_tokens_seen": 73138224,
      "step": 38005
    },
    {
      "epoch": 6.20066889632107,
      "grad_norm": 0.002911331132054329,
      "learning_rate": 0.0018301993732932065,
      "loss": 0.1437,
      "num_input_tokens_seen": 73147328,
      "step": 38010
    },
    {
      "epoch": 6.201484623541887,
      "grad_norm": 0.0029715958517044783,
      "learning_rate": 0.0018210371748033248,
      "loss": 0.1777,
      "num_input_tokens_seen": 73158528,
      "step": 38015
    },
    {
      "epoch": 6.202300350762705,
      "grad_norm": 0.002417012583464384,
      "learning_rate": 0.0018118978273716556,
      "loss": 0.1754,
      "num_input_tokens_seen": 73169184,
      "step": 38020
    },
    {
      "epoch": 6.203116077983522,
      "grad_norm": 0.003834341187030077,
      "learning_rate": 0.001802781332407588,
      "loss": 0.216,
      "num_input_tokens_seen": 73178384,
      "step": 38025
    },
    {
      "epoch": 6.2039318052043395,
      "grad_norm": 0.002378248842433095,
      "learning_rate": 0.0017936876913169806,
      "loss": 0.1184,
      "num_input_tokens_seen": 73188864,
      "step": 38030
    },
    {
      "epoch": 6.204747532425157,
      "grad_norm": 0.0039002960547804832,
      "learning_rate": 0.0017846169055022287,
      "loss": 0.1559,
      "num_input_tokens_seen": 73197872,
      "step": 38035
    },
    {
      "epoch": 6.205563259645975,
      "grad_norm": 0.0011647818610072136,
      "learning_rate": 0.0017755689763621295,
      "loss": 0.1217,
      "num_input_tokens_seen": 73206752,
      "step": 38040
    },
    {
      "epoch": 6.206378986866792,
      "grad_norm": 0.004193603526800871,
      "learning_rate": 0.0017665439052920173,
      "loss": 0.2146,
      "num_input_tokens_seen": 73216480,
      "step": 38045
    },
    {
      "epoch": 6.207194714087609,
      "grad_norm": 0.006098713260143995,
      "learning_rate": 0.0017575416936836286,
      "loss": 0.1925,
      "num_input_tokens_seen": 73225552,
      "step": 38050
    },
    {
      "epoch": 6.208010441308426,
      "grad_norm": 0.002568450989201665,
      "learning_rate": 0.0017485623429252528,
      "loss": 0.1736,
      "num_input_tokens_seen": 73235264,
      "step": 38055
    },
    {
      "epoch": 6.208826168529244,
      "grad_norm": 0.002270784694701433,
      "learning_rate": 0.0017396058544016156,
      "loss": 0.1922,
      "num_input_tokens_seen": 73245216,
      "step": 38060
    },
    {
      "epoch": 6.209641895750061,
      "grad_norm": 0.004168434999883175,
      "learning_rate": 0.0017306722294938958,
      "loss": 0.1673,
      "num_input_tokens_seen": 73254512,
      "step": 38065
    },
    {
      "epoch": 6.210457622970878,
      "grad_norm": 0.00672774575650692,
      "learning_rate": 0.0017217614695798078,
      "loss": 0.2259,
      "num_input_tokens_seen": 73263824,
      "step": 38070
    },
    {
      "epoch": 6.211273350191696,
      "grad_norm": 0.003900641342625022,
      "learning_rate": 0.001712873576033469,
      "loss": 0.2189,
      "num_input_tokens_seen": 73273232,
      "step": 38075
    },
    {
      "epoch": 6.2120890774125135,
      "grad_norm": 0.013645845465362072,
      "learning_rate": 0.0017040085502255163,
      "loss": 0.2057,
      "num_input_tokens_seen": 73282976,
      "step": 38080
    },
    {
      "epoch": 6.212904804633331,
      "grad_norm": 0.005280466750264168,
      "learning_rate": 0.0016951663935230565,
      "loss": 0.1936,
      "num_input_tokens_seen": 73293360,
      "step": 38085
    },
    {
      "epoch": 6.213720531854148,
      "grad_norm": 0.0036654509603977203,
      "learning_rate": 0.0016863471072896485,
      "loss": 0.1689,
      "num_input_tokens_seen": 73304032,
      "step": 38090
    },
    {
      "epoch": 6.214536259074965,
      "grad_norm": 0.0024850114714354277,
      "learning_rate": 0.0016775506928853377,
      "loss": 0.1599,
      "num_input_tokens_seen": 73313424,
      "step": 38095
    },
    {
      "epoch": 6.215351986295783,
      "grad_norm": 0.0027771831955760717,
      "learning_rate": 0.001668777151666656,
      "loss": 0.2287,
      "num_input_tokens_seen": 73323152,
      "step": 38100
    },
    {
      "epoch": 6.2161677135166,
      "grad_norm": 0.005635826848447323,
      "learning_rate": 0.0016600264849865709,
      "loss": 0.1625,
      "num_input_tokens_seen": 73332736,
      "step": 38105
    },
    {
      "epoch": 6.216983440737417,
      "grad_norm": 0.0018652834696695209,
      "learning_rate": 0.0016512986941945695,
      "loss": 0.1367,
      "num_input_tokens_seen": 73342512,
      "step": 38110
    },
    {
      "epoch": 6.217799167958235,
      "grad_norm": 0.004316889215260744,
      "learning_rate": 0.0016425937806365753,
      "loss": 0.1975,
      "num_input_tokens_seen": 73351424,
      "step": 38115
    },
    {
      "epoch": 6.218614895179052,
      "grad_norm": 0.0018802196718752384,
      "learning_rate": 0.0016339117456549979,
      "loss": 0.158,
      "num_input_tokens_seen": 73360976,
      "step": 38120
    },
    {
      "epoch": 6.219430622399869,
      "grad_norm": 0.004920002073049545,
      "learning_rate": 0.0016252525905886995,
      "loss": 0.1954,
      "num_input_tokens_seen": 73371184,
      "step": 38125
    },
    {
      "epoch": 6.220246349620687,
      "grad_norm": 0.0022950018756091595,
      "learning_rate": 0.0016166163167730617,
      "loss": 0.1351,
      "num_input_tokens_seen": 73380128,
      "step": 38130
    },
    {
      "epoch": 6.221062076841505,
      "grad_norm": 0.006940577179193497,
      "learning_rate": 0.0016080029255398864,
      "loss": 0.1366,
      "num_input_tokens_seen": 73390528,
      "step": 38135
    },
    {
      "epoch": 6.221877804062322,
      "grad_norm": 0.003242013743147254,
      "learning_rate": 0.0015994124182174606,
      "loss": 0.1747,
      "num_input_tokens_seen": 73399680,
      "step": 38140
    },
    {
      "epoch": 6.222693531283139,
      "grad_norm": 0.004720110446214676,
      "learning_rate": 0.001590844796130575,
      "loss": 0.1514,
      "num_input_tokens_seen": 73409408,
      "step": 38145
    },
    {
      "epoch": 6.223509258503956,
      "grad_norm": 0.003510969690978527,
      "learning_rate": 0.001582300060600439,
      "loss": 0.137,
      "num_input_tokens_seen": 73419296,
      "step": 38150
    },
    {
      "epoch": 6.224324985724774,
      "grad_norm": 0.004052681382745504,
      "learning_rate": 0.0015737782129447652,
      "loss": 0.2046,
      "num_input_tokens_seen": 73429152,
      "step": 38155
    },
    {
      "epoch": 6.225140712945591,
      "grad_norm": 0.0030273350421339273,
      "learning_rate": 0.0015652792544777361,
      "loss": 0.1403,
      "num_input_tokens_seen": 73439040,
      "step": 38160
    },
    {
      "epoch": 6.225956440166408,
      "grad_norm": 0.003830027999356389,
      "learning_rate": 0.0015568031865099863,
      "loss": 0.1674,
      "num_input_tokens_seen": 73448112,
      "step": 38165
    },
    {
      "epoch": 6.226772167387225,
      "grad_norm": 0.0038691218942403793,
      "learning_rate": 0.0015483500103486369,
      "loss": 0.1706,
      "num_input_tokens_seen": 73458480,
      "step": 38170
    },
    {
      "epoch": 6.227587894608043,
      "grad_norm": 0.002260435139760375,
      "learning_rate": 0.0015399197272972787,
      "loss": 0.1524,
      "num_input_tokens_seen": 73468800,
      "step": 38175
    },
    {
      "epoch": 6.2284036218288605,
      "grad_norm": 0.001676948624663055,
      "learning_rate": 0.0015315123386559714,
      "loss": 0.1433,
      "num_input_tokens_seen": 73479216,
      "step": 38180
    },
    {
      "epoch": 6.229219349049678,
      "grad_norm": 0.003348248777911067,
      "learning_rate": 0.0015231278457212283,
      "loss": 0.1463,
      "num_input_tokens_seen": 73488416,
      "step": 38185
    },
    {
      "epoch": 6.230035076270495,
      "grad_norm": 0.0022730291821062565,
      "learning_rate": 0.001514766249786048,
      "loss": 0.2176,
      "num_input_tokens_seen": 73497872,
      "step": 38190
    },
    {
      "epoch": 6.230850803491313,
      "grad_norm": 0.002741565927863121,
      "learning_rate": 0.0015064275521398994,
      "loss": 0.1495,
      "num_input_tokens_seen": 73508160,
      "step": 38195
    },
    {
      "epoch": 6.23166653071213,
      "grad_norm": 0.0027952950913459063,
      "learning_rate": 0.0014981117540686872,
      "loss": 0.2122,
      "num_input_tokens_seen": 73518048,
      "step": 38200
    },
    {
      "epoch": 6.23166653071213,
      "eval_loss": 0.19022782146930695,
      "eval_runtime": 67.9901,
      "eval_samples_per_second": 40.079,
      "eval_steps_per_second": 20.047,
      "num_input_tokens_seen": 73518048,
      "step": 38200
    },
    {
      "epoch": 6.232482257932947,
      "grad_norm": 0.0034486325457692146,
      "learning_rate": 0.0014898188568548687,
      "loss": 0.1889,
      "num_input_tokens_seen": 73527520,
      "step": 38205
    },
    {
      "epoch": 6.233297985153764,
      "grad_norm": 0.0035530219320207834,
      "learning_rate": 0.0014815488617772542,
      "loss": 0.1497,
      "num_input_tokens_seen": 73537616,
      "step": 38210
    },
    {
      "epoch": 6.234113712374582,
      "grad_norm": 0.003766504116356373,
      "learning_rate": 0.0014733017701112072,
      "loss": 0.1459,
      "num_input_tokens_seen": 73547584,
      "step": 38215
    },
    {
      "epoch": 6.234929439595399,
      "grad_norm": 0.00505629600957036,
      "learning_rate": 0.0014650775831285435,
      "loss": 0.1253,
      "num_input_tokens_seen": 73557472,
      "step": 38220
    },
    {
      "epoch": 6.2357451668162165,
      "grad_norm": 0.0021885340102016926,
      "learning_rate": 0.001456876302097515,
      "loss": 0.1541,
      "num_input_tokens_seen": 73567200,
      "step": 38225
    },
    {
      "epoch": 6.236560894037034,
      "grad_norm": 0.0035677102860063314,
      "learning_rate": 0.0014486979282828604,
      "loss": 0.2569,
      "num_input_tokens_seen": 73577520,
      "step": 38230
    },
    {
      "epoch": 6.237376621257852,
      "grad_norm": 0.004453727044165134,
      "learning_rate": 0.001440542462945804,
      "loss": 0.1986,
      "num_input_tokens_seen": 73587488,
      "step": 38235
    },
    {
      "epoch": 6.238192348478669,
      "grad_norm": 0.0044677662663161755,
      "learning_rate": 0.0014324099073440232,
      "loss": 0.149,
      "num_input_tokens_seen": 73596992,
      "step": 38240
    },
    {
      "epoch": 6.239008075699486,
      "grad_norm": 0.0048743044026196,
      "learning_rate": 0.0014243002627316482,
      "loss": 0.1708,
      "num_input_tokens_seen": 73606176,
      "step": 38245
    },
    {
      "epoch": 6.239823802920303,
      "grad_norm": 0.0021983960177749395,
      "learning_rate": 0.0014162135303592781,
      "loss": 0.1623,
      "num_input_tokens_seen": 73616192,
      "step": 38250
    },
    {
      "epoch": 6.240639530141121,
      "grad_norm": 0.004677280783653259,
      "learning_rate": 0.001408149711474016,
      "loss": 0.2448,
      "num_input_tokens_seen": 73625840,
      "step": 38255
    },
    {
      "epoch": 6.241455257361938,
      "grad_norm": 0.007397428620606661,
      "learning_rate": 0.0014001088073193834,
      "loss": 0.1788,
      "num_input_tokens_seen": 73636272,
      "step": 38260
    },
    {
      "epoch": 6.242270984582755,
      "grad_norm": 0.0032150621991604567,
      "learning_rate": 0.0013920908191354052,
      "loss": 0.1443,
      "num_input_tokens_seen": 73645728,
      "step": 38265
    },
    {
      "epoch": 6.243086711803572,
      "grad_norm": 0.005313083529472351,
      "learning_rate": 0.001384095748158526,
      "loss": 0.1652,
      "num_input_tokens_seen": 73654352,
      "step": 38270
    },
    {
      "epoch": 6.2439024390243905,
      "grad_norm": 0.001293817418627441,
      "learning_rate": 0.0013761235956217255,
      "loss": 0.1327,
      "num_input_tokens_seen": 73664336,
      "step": 38275
    },
    {
      "epoch": 6.244718166245208,
      "grad_norm": 0.003082070266827941,
      "learning_rate": 0.0013681743627543873,
      "loss": 0.219,
      "num_input_tokens_seen": 73673168,
      "step": 38280
    },
    {
      "epoch": 6.245533893466025,
      "grad_norm": 0.004369497764855623,
      "learning_rate": 0.001360248050782381,
      "loss": 0.239,
      "num_input_tokens_seen": 73681296,
      "step": 38285
    },
    {
      "epoch": 6.246349620686843,
      "grad_norm": 0.003973716404289007,
      "learning_rate": 0.001352344660928062,
      "loss": 0.1513,
      "num_input_tokens_seen": 73691808,
      "step": 38290
    },
    {
      "epoch": 6.24716534790766,
      "grad_norm": 0.0059048887342214584,
      "learning_rate": 0.0013444641944102052,
      "loss": 0.1755,
      "num_input_tokens_seen": 73701904,
      "step": 38295
    },
    {
      "epoch": 6.247981075128477,
      "grad_norm": 0.00301798852160573,
      "learning_rate": 0.0013366066524441056,
      "loss": 0.1677,
      "num_input_tokens_seen": 73712496,
      "step": 38300
    },
    {
      "epoch": 6.248796802349294,
      "grad_norm": 0.003135152393952012,
      "learning_rate": 0.0013287720362414768,
      "loss": 0.1487,
      "num_input_tokens_seen": 73723040,
      "step": 38305
    },
    {
      "epoch": 6.249612529570112,
      "grad_norm": 0.003057159250602126,
      "learning_rate": 0.0013209603470105025,
      "loss": 0.1866,
      "num_input_tokens_seen": 73732192,
      "step": 38310
    },
    {
      "epoch": 6.250428256790929,
      "grad_norm": 0.004707040265202522,
      "learning_rate": 0.0013131715859558857,
      "loss": 0.2036,
      "num_input_tokens_seen": 73741536,
      "step": 38315
    },
    {
      "epoch": 6.251243984011746,
      "grad_norm": 0.0024117364082485437,
      "learning_rate": 0.001305405754278699,
      "loss": 0.128,
      "num_input_tokens_seen": 73752432,
      "step": 38320
    },
    {
      "epoch": 6.2520597112325635,
      "grad_norm": 0.002519980538636446,
      "learning_rate": 0.0012976628531765843,
      "loss": 0.1414,
      "num_input_tokens_seen": 73762176,
      "step": 38325
    },
    {
      "epoch": 6.252875438453382,
      "grad_norm": 0.0034251161850988865,
      "learning_rate": 0.0012899428838435533,
      "loss": 0.151,
      "num_input_tokens_seen": 73771904,
      "step": 38330
    },
    {
      "epoch": 6.253691165674199,
      "grad_norm": 0.0026973362546414137,
      "learning_rate": 0.001282245847470137,
      "loss": 0.2242,
      "num_input_tokens_seen": 73779840,
      "step": 38335
    },
    {
      "epoch": 6.254506892895016,
      "grad_norm": 0.0032300809398293495,
      "learning_rate": 0.001274571745243319,
      "loss": 0.1182,
      "num_input_tokens_seen": 73790368,
      "step": 38340
    },
    {
      "epoch": 6.255322620115833,
      "grad_norm": 0.0048723407089710236,
      "learning_rate": 0.0012669205783465364,
      "loss": 0.2477,
      "num_input_tokens_seen": 73801872,
      "step": 38345
    },
    {
      "epoch": 6.256138347336651,
      "grad_norm": 0.00442309956997633,
      "learning_rate": 0.001259292347959695,
      "loss": 0.1727,
      "num_input_tokens_seen": 73810944,
      "step": 38350
    },
    {
      "epoch": 6.256954074557468,
      "grad_norm": 0.001831896835938096,
      "learning_rate": 0.0012516870552591707,
      "loss": 0.1498,
      "num_input_tokens_seen": 73820944,
      "step": 38355
    },
    {
      "epoch": 6.257769801778285,
      "grad_norm": 0.003964791540056467,
      "learning_rate": 0.001244104701417792,
      "loss": 0.18,
      "num_input_tokens_seen": 73831216,
      "step": 38360
    },
    {
      "epoch": 6.258585528999102,
      "grad_norm": 0.0038164108991622925,
      "learning_rate": 0.0012365452876048565,
      "loss": 0.2142,
      "num_input_tokens_seen": 73841024,
      "step": 38365
    },
    {
      "epoch": 6.25940125621992,
      "grad_norm": 0.002296424238011241,
      "learning_rate": 0.001229008814986099,
      "loss": 0.1661,
      "num_input_tokens_seen": 73850672,
      "step": 38370
    },
    {
      "epoch": 6.2602169834407375,
      "grad_norm": 0.004223742056638002,
      "learning_rate": 0.0012214952847237725,
      "loss": 0.1841,
      "num_input_tokens_seen": 73860928,
      "step": 38375
    },
    {
      "epoch": 6.261032710661555,
      "grad_norm": 0.0034560018684715033,
      "learning_rate": 0.0012140046979765339,
      "loss": 0.1312,
      "num_input_tokens_seen": 73869824,
      "step": 38380
    },
    {
      "epoch": 6.261848437882372,
      "grad_norm": 0.004742870107293129,
      "learning_rate": 0.0012065370558995258,
      "loss": 0.1637,
      "num_input_tokens_seen": 73880880,
      "step": 38385
    },
    {
      "epoch": 6.26266416510319,
      "grad_norm": 0.004956102464348078,
      "learning_rate": 0.0011990923596443602,
      "loss": 0.1592,
      "num_input_tokens_seen": 73890352,
      "step": 38390
    },
    {
      "epoch": 6.263479892324007,
      "grad_norm": 0.004152554087340832,
      "learning_rate": 0.001191670610359119,
      "loss": 0.228,
      "num_input_tokens_seen": 73900624,
      "step": 38395
    },
    {
      "epoch": 6.264295619544824,
      "grad_norm": 0.0021977312862873077,
      "learning_rate": 0.0011842718091882865,
      "loss": 0.2377,
      "num_input_tokens_seen": 73911328,
      "step": 38400
    },
    {
      "epoch": 6.264295619544824,
      "eval_loss": 0.19073425233364105,
      "eval_runtime": 68.098,
      "eval_samples_per_second": 40.016,
      "eval_steps_per_second": 20.015,
      "num_input_tokens_seen": 73911328,
      "step": 38400
    },
    {
      "epoch": 6.265111346765641,
      "grad_norm": 0.0037544507067650557,
      "learning_rate": 0.0011768959572729,
      "loss": 0.1548,
      "num_input_tokens_seen": 73919952,
      "step": 38405
    },
    {
      "epoch": 6.265927073986459,
      "grad_norm": 0.00223355065099895,
      "learning_rate": 0.001169543055750366,
      "loss": 0.1153,
      "num_input_tokens_seen": 73928752,
      "step": 38410
    },
    {
      "epoch": 6.266742801207276,
      "grad_norm": 0.005369969643652439,
      "learning_rate": 0.0011622131057546115,
      "loss": 0.2128,
      "num_input_tokens_seen": 73938816,
      "step": 38415
    },
    {
      "epoch": 6.2675585284280935,
      "grad_norm": 0.005303267389535904,
      "learning_rate": 0.0011549061084160316,
      "loss": 0.1673,
      "num_input_tokens_seen": 73948752,
      "step": 38420
    },
    {
      "epoch": 6.268374255648911,
      "grad_norm": 0.0030894591473042965,
      "learning_rate": 0.0011476220648614088,
      "loss": 0.1225,
      "num_input_tokens_seen": 73957808,
      "step": 38425
    },
    {
      "epoch": 6.269189982869729,
      "grad_norm": 0.003729452146217227,
      "learning_rate": 0.0011403609762140777,
      "loss": 0.1801,
      "num_input_tokens_seen": 73966640,
      "step": 38430
    },
    {
      "epoch": 6.270005710090546,
      "grad_norm": 0.006811490282416344,
      "learning_rate": 0.0011331228435937756,
      "loss": 0.2352,
      "num_input_tokens_seen": 73977776,
      "step": 38435
    },
    {
      "epoch": 6.270821437311363,
      "grad_norm": 0.002992267720401287,
      "learning_rate": 0.0011259076681166935,
      "loss": 0.173,
      "num_input_tokens_seen": 73986848,
      "step": 38440
    },
    {
      "epoch": 6.27163716453218,
      "grad_norm": 0.0023872293531894684,
      "learning_rate": 0.0011187154508955244,
      "loss": 0.1466,
      "num_input_tokens_seen": 73996672,
      "step": 38445
    },
    {
      "epoch": 6.272452891752998,
      "grad_norm": 0.0039046378806233406,
      "learning_rate": 0.001111546193039381,
      "loss": 0.1866,
      "num_input_tokens_seen": 74006656,
      "step": 38450
    },
    {
      "epoch": 6.273268618973815,
      "grad_norm": 0.0033427453599870205,
      "learning_rate": 0.0011043998956538792,
      "loss": 0.1821,
      "num_input_tokens_seen": 74015248,
      "step": 38455
    },
    {
      "epoch": 6.274084346194632,
      "grad_norm": 0.007619782350957394,
      "learning_rate": 0.0010972765598410538,
      "loss": 0.1768,
      "num_input_tokens_seen": 74024848,
      "step": 38460
    },
    {
      "epoch": 6.27490007341545,
      "grad_norm": 0.004763641860336065,
      "learning_rate": 0.0010901761866993931,
      "loss": 0.164,
      "num_input_tokens_seen": 74033632,
      "step": 38465
    },
    {
      "epoch": 6.275715800636267,
      "grad_norm": 0.0027294871397316456,
      "learning_rate": 0.0010830987773238876,
      "loss": 0.1557,
      "num_input_tokens_seen": 74043120,
      "step": 38470
    },
    {
      "epoch": 6.276531527857085,
      "grad_norm": 0.002775941276922822,
      "learning_rate": 0.0010760443328059644,
      "loss": 0.2006,
      "num_input_tokens_seen": 74052656,
      "step": 38475
    },
    {
      "epoch": 6.277347255077902,
      "grad_norm": 0.0035979608073830605,
      "learning_rate": 0.001069012854233503,
      "loss": 0.154,
      "num_input_tokens_seen": 74060864,
      "step": 38480
    },
    {
      "epoch": 6.27816298229872,
      "grad_norm": 0.0032312669791281223,
      "learning_rate": 0.0010620043426908365,
      "loss": 0.1308,
      "num_input_tokens_seen": 74071360,
      "step": 38485
    },
    {
      "epoch": 6.278978709519537,
      "grad_norm": 0.0062796575948596,
      "learning_rate": 0.0010550187992587833,
      "loss": 0.2086,
      "num_input_tokens_seen": 74080736,
      "step": 38490
    },
    {
      "epoch": 6.279794436740354,
      "grad_norm": 0.00394849618896842,
      "learning_rate": 0.0010480562250145653,
      "loss": 0.1584,
      "num_input_tokens_seen": 74090400,
      "step": 38495
    },
    {
      "epoch": 6.280610163961171,
      "grad_norm": 0.0027034911327064037,
      "learning_rate": 0.0010411166210319567,
      "loss": 0.204,
      "num_input_tokens_seen": 74101312,
      "step": 38500
    },
    {
      "epoch": 6.281425891181989,
      "grad_norm": 0.003312293905764818,
      "learning_rate": 0.0010341999883810848,
      "loss": 0.2539,
      "num_input_tokens_seen": 74110832,
      "step": 38505
    },
    {
      "epoch": 6.282241618402806,
      "grad_norm": 0.007209118455648422,
      "learning_rate": 0.0010273063281285965,
      "loss": 0.238,
      "num_input_tokens_seen": 74119520,
      "step": 38510
    },
    {
      "epoch": 6.283057345623623,
      "grad_norm": 0.0037192332092672586,
      "learning_rate": 0.0010204356413375747,
      "loss": 0.162,
      "num_input_tokens_seen": 74129584,
      "step": 38515
    },
    {
      "epoch": 6.2838730728444405,
      "grad_norm": 0.003422048408538103,
      "learning_rate": 0.001013587929067572,
      "loss": 0.1441,
      "num_input_tokens_seen": 74138784,
      "step": 38520
    },
    {
      "epoch": 6.2846888000652585,
      "grad_norm": 0.003521217731758952,
      "learning_rate": 0.00100676319237461,
      "loss": 0.1443,
      "num_input_tokens_seen": 74147776,
      "step": 38525
    },
    {
      "epoch": 6.285504527286076,
      "grad_norm": 0.0037178124766796827,
      "learning_rate": 0.0009999614323110972,
      "loss": 0.126,
      "num_input_tokens_seen": 74157008,
      "step": 38530
    },
    {
      "epoch": 6.286320254506893,
      "grad_norm": 0.0024189529940485954,
      "learning_rate": 0.000993182649926011,
      "loss": 0.1597,
      "num_input_tokens_seen": 74165888,
      "step": 38535
    },
    {
      "epoch": 6.28713598172771,
      "grad_norm": 0.0034904752392321825,
      "learning_rate": 0.000986426846264682,
      "loss": 0.1144,
      "num_input_tokens_seen": 74175232,
      "step": 38540
    },
    {
      "epoch": 6.287951708948528,
      "grad_norm": 0.003476334735751152,
      "learning_rate": 0.00097969402236896,
      "loss": 0.2978,
      "num_input_tokens_seen": 74185360,
      "step": 38545
    },
    {
      "epoch": 6.288767436169345,
      "grad_norm": 0.002344410168007016,
      "learning_rate": 0.0009729841792771143,
      "loss": 0.163,
      "num_input_tokens_seen": 74194368,
      "step": 38550
    },
    {
      "epoch": 6.289583163390162,
      "grad_norm": 0.0017495794454589486,
      "learning_rate": 0.0009662973180239176,
      "loss": 0.2023,
      "num_input_tokens_seen": 74204448,
      "step": 38555
    },
    {
      "epoch": 6.290398890610979,
      "grad_norm": 0.0018813546048477292,
      "learning_rate": 0.0009596334396405448,
      "loss": 0.1777,
      "num_input_tokens_seen": 74214272,
      "step": 38560
    },
    {
      "epoch": 6.291214617831797,
      "grad_norm": 0.004279403481632471,
      "learning_rate": 0.0009529925451546406,
      "loss": 0.188,
      "num_input_tokens_seen": 74224416,
      "step": 38565
    },
    {
      "epoch": 6.2920303450526145,
      "grad_norm": 0.00379734649322927,
      "learning_rate": 0.0009463746355903357,
      "loss": 0.1639,
      "num_input_tokens_seen": 74233920,
      "step": 38570
    },
    {
      "epoch": 6.292846072273432,
      "grad_norm": 0.004427447449415922,
      "learning_rate": 0.0009397797119681971,
      "loss": 0.2162,
      "num_input_tokens_seen": 74243600,
      "step": 38575
    },
    {
      "epoch": 6.293661799494249,
      "grad_norm": 0.0038895299658179283,
      "learning_rate": 0.0009332077753052281,
      "loss": 0.132,
      "num_input_tokens_seen": 74252464,
      "step": 38580
    },
    {
      "epoch": 6.294477526715067,
      "grad_norm": 0.0016950578428804874,
      "learning_rate": 0.0009266588266149011,
      "loss": 0.1664,
      "num_input_tokens_seen": 74261504,
      "step": 38585
    },
    {
      "epoch": 6.295293253935884,
      "grad_norm": 0.0035810042172670364,
      "learning_rate": 0.0009201328669071584,
      "loss": 0.1781,
      "num_input_tokens_seen": 74271296,
      "step": 38590
    },
    {
      "epoch": 6.296108981156701,
      "grad_norm": 0.004891968797892332,
      "learning_rate": 0.0009136298971883949,
      "loss": 0.2194,
      "num_input_tokens_seen": 74283088,
      "step": 38595
    },
    {
      "epoch": 6.296924708377518,
      "grad_norm": 0.005069867242127657,
      "learning_rate": 0.0009071499184614251,
      "loss": 0.213,
      "num_input_tokens_seen": 74293168,
      "step": 38600
    },
    {
      "epoch": 6.296924708377518,
      "eval_loss": 0.19038179516792297,
      "eval_runtime": 67.9975,
      "eval_samples_per_second": 40.075,
      "eval_steps_per_second": 20.045,
      "num_input_tokens_seen": 74293168,
      "step": 38600
    },
    {
      "epoch": 6.297740435598336,
      "grad_norm": 0.0052665104158222675,
      "learning_rate": 0.0009006929317255663,
      "loss": 0.15,
      "num_input_tokens_seen": 74302128,
      "step": 38605
    },
    {
      "epoch": 6.298556162819153,
      "grad_norm": 0.002108219778165221,
      "learning_rate": 0.0008942589379765387,
      "loss": 0.1846,
      "num_input_tokens_seen": 74312512,
      "step": 38610
    },
    {
      "epoch": 6.2993718900399704,
      "grad_norm": 0.002879201900213957,
      "learning_rate": 0.0008878479382065817,
      "loss": 0.17,
      "num_input_tokens_seen": 74322560,
      "step": 38615
    },
    {
      "epoch": 6.300187617260788,
      "grad_norm": 0.0031297444365918636,
      "learning_rate": 0.0008814599334043215,
      "loss": 0.1687,
      "num_input_tokens_seen": 74333008,
      "step": 38620
    },
    {
      "epoch": 6.301003344481606,
      "grad_norm": 0.003520137630403042,
      "learning_rate": 0.0008750949245548866,
      "loss": 0.161,
      "num_input_tokens_seen": 74341984,
      "step": 38625
    },
    {
      "epoch": 6.301819071702423,
      "grad_norm": 0.0026086720172315836,
      "learning_rate": 0.0008687529126398252,
      "loss": 0.1266,
      "num_input_tokens_seen": 74351888,
      "step": 38630
    },
    {
      "epoch": 6.30263479892324,
      "grad_norm": 0.0058916793204844,
      "learning_rate": 0.0008624338986371715,
      "loss": 0.2268,
      "num_input_tokens_seen": 74360832,
      "step": 38635
    },
    {
      "epoch": 6.303450526144058,
      "grad_norm": 0.005179365631192923,
      "learning_rate": 0.0008561378835213962,
      "loss": 0.2016,
      "num_input_tokens_seen": 74370576,
      "step": 38640
    },
    {
      "epoch": 6.304266253364875,
      "grad_norm": 0.0055434308014810085,
      "learning_rate": 0.0008498648682634058,
      "loss": 0.1999,
      "num_input_tokens_seen": 74379120,
      "step": 38645
    },
    {
      "epoch": 6.305081980585692,
      "grad_norm": 0.0018847957253456116,
      "learning_rate": 0.0008436148538306099,
      "loss": 0.1243,
      "num_input_tokens_seen": 74388704,
      "step": 38650
    },
    {
      "epoch": 6.305897707806509,
      "grad_norm": 0.004722807090729475,
      "learning_rate": 0.0008373878411868041,
      "loss": 0.2221,
      "num_input_tokens_seen": 74397840,
      "step": 38655
    },
    {
      "epoch": 6.306713435027326,
      "grad_norm": 0.0031186891719698906,
      "learning_rate": 0.000831183831292287,
      "loss": 0.1109,
      "num_input_tokens_seen": 74406608,
      "step": 38660
    },
    {
      "epoch": 6.307529162248144,
      "grad_norm": 0.003760380670428276,
      "learning_rate": 0.0008250028251037933,
      "loss": 0.1826,
      "num_input_tokens_seen": 74416336,
      "step": 38665
    },
    {
      "epoch": 6.3083448894689615,
      "grad_norm": 0.0031788782216608524,
      "learning_rate": 0.0008188448235745271,
      "loss": 0.142,
      "num_input_tokens_seen": 74425520,
      "step": 38670
    },
    {
      "epoch": 6.309160616689779,
      "grad_norm": 0.003535233438014984,
      "learning_rate": 0.0008127098276541122,
      "loss": 0.1346,
      "num_input_tokens_seen": 74435584,
      "step": 38675
    },
    {
      "epoch": 6.309976343910597,
      "grad_norm": 0.0017152174841612577,
      "learning_rate": 0.0008065978382886418,
      "loss": 0.1288,
      "num_input_tokens_seen": 74445120,
      "step": 38680
    },
    {
      "epoch": 6.310792071131414,
      "grad_norm": 0.004207239020615816,
      "learning_rate": 0.0008005088564206785,
      "loss": 0.2556,
      "num_input_tokens_seen": 74453584,
      "step": 38685
    },
    {
      "epoch": 6.311607798352231,
      "grad_norm": 0.004728192463517189,
      "learning_rate": 0.0007944428829891881,
      "loss": 0.1377,
      "num_input_tokens_seen": 74463088,
      "step": 38690
    },
    {
      "epoch": 6.312423525573048,
      "grad_norm": 0.006381343584507704,
      "learning_rate": 0.0007883999189296386,
      "loss": 0.2039,
      "num_input_tokens_seen": 74472640,
      "step": 38695
    },
    {
      "epoch": 6.313239252793866,
      "grad_norm": 0.002540601883083582,
      "learning_rate": 0.0007823799651739515,
      "loss": 0.1454,
      "num_input_tokens_seen": 74483376,
      "step": 38700
    },
    {
      "epoch": 6.314054980014683,
      "grad_norm": 0.004267572890967131,
      "learning_rate": 0.0007763830226504509,
      "loss": 0.1526,
      "num_input_tokens_seen": 74492144,
      "step": 38705
    },
    {
      "epoch": 6.3148707072355,
      "grad_norm": 0.004385190550237894,
      "learning_rate": 0.0007704090922839468,
      "loss": 0.1141,
      "num_input_tokens_seen": 74499600,
      "step": 38710
    },
    {
      "epoch": 6.3156864344563175,
      "grad_norm": 0.001964489696547389,
      "learning_rate": 0.0007644581749957025,
      "loss": 0.1359,
      "num_input_tokens_seen": 74509200,
      "step": 38715
    },
    {
      "epoch": 6.3165021616771355,
      "grad_norm": 0.0011504536960273981,
      "learning_rate": 0.000758530271703417,
      "loss": 0.126,
      "num_input_tokens_seen": 74517664,
      "step": 38720
    },
    {
      "epoch": 6.317317888897953,
      "grad_norm": 0.007335934322327375,
      "learning_rate": 0.0007526253833212426,
      "loss": 0.2154,
      "num_input_tokens_seen": 74527152,
      "step": 38725
    },
    {
      "epoch": 6.31813361611877,
      "grad_norm": 0.0034502288326621056,
      "learning_rate": 0.0007467435107598008,
      "loss": 0.2103,
      "num_input_tokens_seen": 74537024,
      "step": 38730
    },
    {
      "epoch": 6.318949343339587,
      "grad_norm": 0.004672536626458168,
      "learning_rate": 0.0007408846549261328,
      "loss": 0.1861,
      "num_input_tokens_seen": 74546800,
      "step": 38735
    },
    {
      "epoch": 6.319765070560405,
      "grad_norm": 0.0020524600986391306,
      "learning_rate": 0.0007350488167237656,
      "loss": 0.1536,
      "num_input_tokens_seen": 74556496,
      "step": 38740
    },
    {
      "epoch": 6.320580797781222,
      "grad_norm": 0.0030109200160950422,
      "learning_rate": 0.0007292359970526629,
      "loss": 0.1554,
      "num_input_tokens_seen": 74565808,
      "step": 38745
    },
    {
      "epoch": 6.321396525002039,
      "grad_norm": 0.0004618000821210444,
      "learning_rate": 0.0007234461968092076,
      "loss": 0.1353,
      "num_input_tokens_seen": 74574560,
      "step": 38750
    },
    {
      "epoch": 6.322212252222856,
      "grad_norm": 0.0035547513980418444,
      "learning_rate": 0.0007176794168862854,
      "loss": 0.27,
      "num_input_tokens_seen": 74584384,
      "step": 38755
    },
    {
      "epoch": 6.323027979443674,
      "grad_norm": 0.002416956704109907,
      "learning_rate": 0.000711935658173185,
      "loss": 0.1646,
      "num_input_tokens_seen": 74593808,
      "step": 38760
    },
    {
      "epoch": 6.3238437066644915,
      "grad_norm": 0.002104389714077115,
      "learning_rate": 0.0007062149215556812,
      "loss": 0.0907,
      "num_input_tokens_seen": 74602256,
      "step": 38765
    },
    {
      "epoch": 6.324659433885309,
      "grad_norm": 0.004782311152666807,
      "learning_rate": 0.0007005172079159849,
      "loss": 0.1529,
      "num_input_tokens_seen": 74611376,
      "step": 38770
    },
    {
      "epoch": 6.325475161106126,
      "grad_norm": 0.0020198358688503504,
      "learning_rate": 0.0006948425181327267,
      "loss": 0.0974,
      "num_input_tokens_seen": 74621088,
      "step": 38775
    },
    {
      "epoch": 6.326290888326944,
      "grad_norm": 0.004302664659917355,
      "learning_rate": 0.000689190853081073,
      "loss": 0.1826,
      "num_input_tokens_seen": 74631120,
      "step": 38780
    },
    {
      "epoch": 6.327106615547761,
      "grad_norm": 0.003738028695806861,
      "learning_rate": 0.000683562213632527,
      "loss": 0.1434,
      "num_input_tokens_seen": 74640768,
      "step": 38785
    },
    {
      "epoch": 6.327922342768578,
      "grad_norm": 0.003534656250849366,
      "learning_rate": 0.0006779566006551108,
      "loss": 0.1536,
      "num_input_tokens_seen": 74650480,
      "step": 38790
    },
    {
      "epoch": 6.328738069989395,
      "grad_norm": 0.0036769318394362926,
      "learning_rate": 0.0006723740150132995,
      "loss": 0.2205,
      "num_input_tokens_seen": 74660368,
      "step": 38795
    },
    {
      "epoch": 6.329553797210213,
      "grad_norm": 0.0041419691406190395,
      "learning_rate": 0.0006668144575679713,
      "loss": 0.1727,
      "num_input_tokens_seen": 74668864,
      "step": 38800
    },
    {
      "epoch": 6.329553797210213,
      "eval_loss": 0.18995895981788635,
      "eval_runtime": 67.9733,
      "eval_samples_per_second": 40.089,
      "eval_steps_per_second": 20.052,
      "num_input_tokens_seen": 74668864,
      "step": 38800
    },
    {
      "epoch": 6.33036952443103,
      "grad_norm": 0.0037689630407840014,
      "learning_rate": 0.0006612779291765069,
      "loss": 0.1694,
      "num_input_tokens_seen": 74678944,
      "step": 38805
    },
    {
      "epoch": 6.331185251651847,
      "grad_norm": 0.0039654881693422794,
      "learning_rate": 0.0006557644306926736,
      "loss": 0.2047,
      "num_input_tokens_seen": 74688112,
      "step": 38810
    },
    {
      "epoch": 6.332000978872665,
      "grad_norm": 0.0025548378471285105,
      "learning_rate": 0.0006502739629667575,
      "loss": 0.1613,
      "num_input_tokens_seen": 74697840,
      "step": 38815
    },
    {
      "epoch": 6.332816706093483,
      "grad_norm": 0.006673446856439114,
      "learning_rate": 0.0006448065268454317,
      "loss": 0.2383,
      "num_input_tokens_seen": 74707040,
      "step": 38820
    },
    {
      "epoch": 6.3336324333143,
      "grad_norm": 0.003281016368418932,
      "learning_rate": 0.0006393621231718549,
      "loss": 0.1928,
      "num_input_tokens_seen": 74717936,
      "step": 38825
    },
    {
      "epoch": 6.334448160535117,
      "grad_norm": 0.0033348584547638893,
      "learning_rate": 0.0006339407527856389,
      "loss": 0.21,
      "num_input_tokens_seen": 74727856,
      "step": 38830
    },
    {
      "epoch": 6.335263887755934,
      "grad_norm": 0.0024068430066108704,
      "learning_rate": 0.0006285424165227982,
      "loss": 0.1762,
      "num_input_tokens_seen": 74739008,
      "step": 38835
    },
    {
      "epoch": 6.336079614976752,
      "grad_norm": 0.005117053631693125,
      "learning_rate": 0.0006231671152158169,
      "loss": 0.2383,
      "num_input_tokens_seen": 74748544,
      "step": 38840
    },
    {
      "epoch": 6.336895342197569,
      "grad_norm": 0.004422876983880997,
      "learning_rate": 0.0006178148496936819,
      "loss": 0.169,
      "num_input_tokens_seen": 74759920,
      "step": 38845
    },
    {
      "epoch": 6.337711069418386,
      "grad_norm": 0.006128957495093346,
      "learning_rate": 0.000612485620781733,
      "loss": 0.1881,
      "num_input_tokens_seen": 74770832,
      "step": 38850
    },
    {
      "epoch": 6.338526796639204,
      "grad_norm": 0.0028901174664497375,
      "learning_rate": 0.0006071794293018296,
      "loss": 0.1507,
      "num_input_tokens_seen": 74781008,
      "step": 38855
    },
    {
      "epoch": 6.339342523860021,
      "grad_norm": 0.0034581744112074375,
      "learning_rate": 0.0006018962760722501,
      "loss": 0.1424,
      "num_input_tokens_seen": 74791392,
      "step": 38860
    },
    {
      "epoch": 6.3401582510808385,
      "grad_norm": 0.004816803615540266,
      "learning_rate": 0.0005966361619077098,
      "loss": 0.2088,
      "num_input_tokens_seen": 74801888,
      "step": 38865
    },
    {
      "epoch": 6.340973978301656,
      "grad_norm": 0.004845455288887024,
      "learning_rate": 0.000591399087619393,
      "loss": 0.2177,
      "num_input_tokens_seen": 74811872,
      "step": 38870
    },
    {
      "epoch": 6.341789705522474,
      "grad_norm": 0.005634006578475237,
      "learning_rate": 0.0005861850540149371,
      "loss": 0.1889,
      "num_input_tokens_seen": 74821312,
      "step": 38875
    },
    {
      "epoch": 6.342605432743291,
      "grad_norm": 0.002613007090985775,
      "learning_rate": 0.0005809940618983822,
      "loss": 0.1612,
      "num_input_tokens_seen": 74829152,
      "step": 38880
    },
    {
      "epoch": 6.343421159964108,
      "grad_norm": 0.0020787157118320465,
      "learning_rate": 0.0005758261120702712,
      "loss": 0.111,
      "num_input_tokens_seen": 74839424,
      "step": 38885
    },
    {
      "epoch": 6.344236887184925,
      "grad_norm": 0.0045687113888561726,
      "learning_rate": 0.0005706812053275501,
      "loss": 0.2047,
      "num_input_tokens_seen": 74848800,
      "step": 38890
    },
    {
      "epoch": 6.345052614405743,
      "grad_norm": 0.00381510891020298,
      "learning_rate": 0.0005655593424636173,
      "loss": 0.1572,
      "num_input_tokens_seen": 74859744,
      "step": 38895
    },
    {
      "epoch": 6.34586834162656,
      "grad_norm": 0.002408453496173024,
      "learning_rate": 0.0005604605242683746,
      "loss": 0.1291,
      "num_input_tokens_seen": 74869168,
      "step": 38900
    },
    {
      "epoch": 6.346684068847377,
      "grad_norm": 0.002232619794085622,
      "learning_rate": 0.0005553847515280596,
      "loss": 0.1582,
      "num_input_tokens_seen": 74879104,
      "step": 38905
    },
    {
      "epoch": 6.3474997960681945,
      "grad_norm": 0.003108711214736104,
      "learning_rate": 0.0005503320250254795,
      "loss": 0.1409,
      "num_input_tokens_seen": 74889440,
      "step": 38910
    },
    {
      "epoch": 6.3483155232890125,
      "grad_norm": 0.005297530442476273,
      "learning_rate": 0.0005453023455397943,
      "loss": 0.2014,
      "num_input_tokens_seen": 74898160,
      "step": 38915
    },
    {
      "epoch": 6.34913125050983,
      "grad_norm": 0.0016516668256372213,
      "learning_rate": 0.0005402957138466502,
      "loss": 0.1413,
      "num_input_tokens_seen": 74907456,
      "step": 38920
    },
    {
      "epoch": 6.349946977730647,
      "grad_norm": 0.0018899725982919335,
      "learning_rate": 0.0005353121307181463,
      "loss": 0.1736,
      "num_input_tokens_seen": 74916752,
      "step": 38925
    },
    {
      "epoch": 6.350762704951464,
      "grad_norm": 0.0026749465614557266,
      "learning_rate": 0.0005303515969227845,
      "loss": 0.1417,
      "num_input_tokens_seen": 74926864,
      "step": 38930
    },
    {
      "epoch": 6.351578432172282,
      "grad_norm": 0.004385913722217083,
      "learning_rate": 0.0005254141132255862,
      "loss": 0.2056,
      "num_input_tokens_seen": 74936976,
      "step": 38935
    },
    {
      "epoch": 6.352394159393099,
      "grad_norm": 0.0029559500981122255,
      "learning_rate": 0.0005204996803879258,
      "loss": 0.1696,
      "num_input_tokens_seen": 74947168,
      "step": 38940
    },
    {
      "epoch": 6.353209886613916,
      "grad_norm": 0.0022050575353205204,
      "learning_rate": 0.0005156082991676969,
      "loss": 0.1266,
      "num_input_tokens_seen": 74957056,
      "step": 38945
    },
    {
      "epoch": 6.354025613834733,
      "grad_norm": 0.00286508584395051,
      "learning_rate": 0.0005107399703192127,
      "loss": 0.1838,
      "num_input_tokens_seen": 74965824,
      "step": 38950
    },
    {
      "epoch": 6.354841341055551,
      "grad_norm": 0.0020242836326360703,
      "learning_rate": 0.0005058946945932063,
      "loss": 0.1665,
      "num_input_tokens_seen": 74974528,
      "step": 38955
    },
    {
      "epoch": 6.3556570682763684,
      "grad_norm": 0.005294668022543192,
      "learning_rate": 0.0005010724727369131,
      "loss": 0.131,
      "num_input_tokens_seen": 74983248,
      "step": 38960
    },
    {
      "epoch": 6.356472795497186,
      "grad_norm": 0.003526918590068817,
      "learning_rate": 0.000496273305493955,
      "loss": 0.2152,
      "num_input_tokens_seen": 74993056,
      "step": 38965
    },
    {
      "epoch": 6.357288522718003,
      "grad_norm": 0.0030375360511243343,
      "learning_rate": 0.0004914971936044399,
      "loss": 0.1643,
      "num_input_tokens_seen": 75003616,
      "step": 38970
    },
    {
      "epoch": 6.358104249938821,
      "grad_norm": 0.002246707212179899,
      "learning_rate": 0.00048674413780491196,
      "loss": 0.1807,
      "num_input_tokens_seen": 75012576,
      "step": 38975
    },
    {
      "epoch": 6.358919977159638,
      "grad_norm": 0.016358744353055954,
      "learning_rate": 0.0004820141388283183,
      "loss": 0.2761,
      "num_input_tokens_seen": 75020416,
      "step": 38980
    },
    {
      "epoch": 6.359735704380455,
      "grad_norm": 0.008266828954219818,
      "learning_rate": 0.00047730719740410874,
      "loss": 0.2227,
      "num_input_tokens_seen": 75030560,
      "step": 38985
    },
    {
      "epoch": 6.360551431601272,
      "grad_norm": 0.005487764719873667,
      "learning_rate": 0.00047262331425816927,
      "loss": 0.138,
      "num_input_tokens_seen": 75040880,
      "step": 38990
    },
    {
      "epoch": 6.36136715882209,
      "grad_norm": 0.0028414218686521053,
      "learning_rate": 0.00046796249011277213,
      "loss": 0.1682,
      "num_input_tokens_seen": 75050464,
      "step": 38995
    },
    {
      "epoch": 6.362182886042907,
      "grad_norm": 0.002262038877233863,
      "learning_rate": 0.00046332472568669236,
      "loss": 0.1088,
      "num_input_tokens_seen": 75058640,
      "step": 39000
    },
    {
      "epoch": 6.362182886042907,
      "eval_loss": 0.19013573229312897,
      "eval_runtime": 68.1197,
      "eval_samples_per_second": 40.003,
      "eval_steps_per_second": 20.009,
      "num_input_tokens_seen": 75058640,
      "step": 39000
    },
    {
      "epoch": 6.362998613263724,
      "grad_norm": 0.004407414235174656,
      "learning_rate": 0.0004587100216951578,
      "loss": 0.2104,
      "num_input_tokens_seen": 75068624,
      "step": 39005
    },
    {
      "epoch": 6.3638143404845415,
      "grad_norm": 0.0024027847684919834,
      "learning_rate": 0.00045411837884978265,
      "loss": 0.1133,
      "num_input_tokens_seen": 75078016,
      "step": 39010
    },
    {
      "epoch": 6.3646300677053596,
      "grad_norm": 0.004391211085021496,
      "learning_rate": 0.00044954979785865045,
      "loss": 0.1619,
      "num_input_tokens_seen": 75088048,
      "step": 39015
    },
    {
      "epoch": 6.365445794926177,
      "grad_norm": 0.0049465796910226345,
      "learning_rate": 0.00044500427942631426,
      "loss": 0.2044,
      "num_input_tokens_seen": 75098080,
      "step": 39020
    },
    {
      "epoch": 6.366261522146994,
      "grad_norm": 0.0025782836601138115,
      "learning_rate": 0.0004404818242537467,
      "loss": 0.1579,
      "num_input_tokens_seen": 75106912,
      "step": 39025
    },
    {
      "epoch": 6.367077249367812,
      "grad_norm": 0.0040235091000795364,
      "learning_rate": 0.00043598243303837324,
      "loss": 0.1149,
      "num_input_tokens_seen": 75116416,
      "step": 39030
    },
    {
      "epoch": 6.367892976588629,
      "grad_norm": 0.003136499784886837,
      "learning_rate": 0.00043150610647403885,
      "loss": 0.1795,
      "num_input_tokens_seen": 75126352,
      "step": 39035
    },
    {
      "epoch": 6.368708703809446,
      "grad_norm": 0.003367869881913066,
      "learning_rate": 0.00042705284525104134,
      "loss": 0.1388,
      "num_input_tokens_seen": 75134768,
      "step": 39040
    },
    {
      "epoch": 6.369524431030263,
      "grad_norm": 0.007319363299757242,
      "learning_rate": 0.0004226226500561647,
      "loss": 0.2051,
      "num_input_tokens_seen": 75144512,
      "step": 39045
    },
    {
      "epoch": 6.370340158251081,
      "grad_norm": 0.0030574530828744173,
      "learning_rate": 0.0004182155215725791,
      "loss": 0.1447,
      "num_input_tokens_seen": 75152752,
      "step": 39050
    },
    {
      "epoch": 6.371155885471898,
      "grad_norm": 0.003861707402393222,
      "learning_rate": 0.00041383146047992424,
      "loss": 0.203,
      "num_input_tokens_seen": 75162736,
      "step": 39055
    },
    {
      "epoch": 6.3719716126927155,
      "grad_norm": 0.0017316552111878991,
      "learning_rate": 0.00040947046745427597,
      "loss": 0.1972,
      "num_input_tokens_seen": 75172544,
      "step": 39060
    },
    {
      "epoch": 6.372787339913533,
      "grad_norm": 0.002395526971668005,
      "learning_rate": 0.00040513254316814625,
      "loss": 0.1456,
      "num_input_tokens_seen": 75182800,
      "step": 39065
    },
    {
      "epoch": 6.373603067134351,
      "grad_norm": 0.002805202966555953,
      "learning_rate": 0.0004008176882905168,
      "loss": 0.1451,
      "num_input_tokens_seen": 75192144,
      "step": 39070
    },
    {
      "epoch": 6.374418794355168,
      "grad_norm": 0.00403608987107873,
      "learning_rate": 0.00039652590348677184,
      "loss": 0.1681,
      "num_input_tokens_seen": 75201312,
      "step": 39075
    },
    {
      "epoch": 6.375234521575985,
      "grad_norm": 0.00333785661496222,
      "learning_rate": 0.00039225718941878206,
      "loss": 0.1787,
      "num_input_tokens_seen": 75211440,
      "step": 39080
    },
    {
      "epoch": 6.376050248796802,
      "grad_norm": 0.00304041407071054,
      "learning_rate": 0.00038801154674480417,
      "loss": 0.119,
      "num_input_tokens_seen": 75220704,
      "step": 39085
    },
    {
      "epoch": 6.37686597601762,
      "grad_norm": 0.003385892603546381,
      "learning_rate": 0.00038378897611959784,
      "loss": 0.1679,
      "num_input_tokens_seen": 75230368,
      "step": 39090
    },
    {
      "epoch": 6.377681703238437,
      "grad_norm": 0.00512880552560091,
      "learning_rate": 0.00037958947819430875,
      "loss": 0.1792,
      "num_input_tokens_seen": 75238832,
      "step": 39095
    },
    {
      "epoch": 6.378497430459254,
      "grad_norm": 0.004972186405211687,
      "learning_rate": 0.0003754130536165856,
      "loss": 0.1539,
      "num_input_tokens_seen": 75247280,
      "step": 39100
    },
    {
      "epoch": 6.3793131576800715,
      "grad_norm": 0.005114618688821793,
      "learning_rate": 0.0003712597030304632,
      "loss": 0.1636,
      "num_input_tokens_seen": 75257568,
      "step": 39105
    },
    {
      "epoch": 6.3801288849008895,
      "grad_norm": 0.004254648927599192,
      "learning_rate": 0.00036712942707646247,
      "loss": 0.123,
      "num_input_tokens_seen": 75266160,
      "step": 39110
    },
    {
      "epoch": 6.380944612121707,
      "grad_norm": 0.0047112260945141315,
      "learning_rate": 0.00036302222639149063,
      "loss": 0.1663,
      "num_input_tokens_seen": 75275376,
      "step": 39115
    },
    {
      "epoch": 6.381760339342524,
      "grad_norm": 0.003416541963815689,
      "learning_rate": 0.000358938101608941,
      "loss": 0.1339,
      "num_input_tokens_seen": 75284880,
      "step": 39120
    },
    {
      "epoch": 6.382576066563341,
      "grad_norm": 0.0014407529961317778,
      "learning_rate": 0.0003548770533586598,
      "loss": 0.1349,
      "num_input_tokens_seen": 75294544,
      "step": 39125
    },
    {
      "epoch": 6.383391793784159,
      "grad_norm": 0.0022468343377113342,
      "learning_rate": 0.0003508390822668961,
      "loss": 0.18,
      "num_input_tokens_seen": 75305056,
      "step": 39130
    },
    {
      "epoch": 6.384207521004976,
      "grad_norm": 0.002991152461618185,
      "learning_rate": 0.00034682418895633503,
      "loss": 0.1316,
      "num_input_tokens_seen": 75315792,
      "step": 39135
    },
    {
      "epoch": 6.385023248225793,
      "grad_norm": 0.004908763337880373,
      "learning_rate": 0.0003428323740461647,
      "loss": 0.1563,
      "num_input_tokens_seen": 75325824,
      "step": 39140
    },
    {
      "epoch": 6.38583897544661,
      "grad_norm": 0.0030041246209293604,
      "learning_rate": 0.00033886363815194276,
      "loss": 0.181,
      "num_input_tokens_seen": 75336112,
      "step": 39145
    },
    {
      "epoch": 6.386654702667428,
      "grad_norm": 0.004620629362761974,
      "learning_rate": 0.0003349179818857129,
      "loss": 0.1224,
      "num_input_tokens_seen": 75345536,
      "step": 39150
    },
    {
      "epoch": 6.387470429888245,
      "grad_norm": 0.0011623292230069637,
      "learning_rate": 0.0003309954058559383,
      "loss": 0.1626,
      "num_input_tokens_seen": 75355232,
      "step": 39155
    },
    {
      "epoch": 6.388286157109063,
      "grad_norm": 0.002692318521440029,
      "learning_rate": 0.0003270959106675186,
      "loss": 0.1696,
      "num_input_tokens_seen": 75364368,
      "step": 39160
    },
    {
      "epoch": 6.38910188432988,
      "grad_norm": 0.004125090781599283,
      "learning_rate": 0.0003232194969218227,
      "loss": 0.1305,
      "num_input_tokens_seen": 75374464,
      "step": 39165
    },
    {
      "epoch": 6.389917611550698,
      "grad_norm": 0.004170611966401339,
      "learning_rate": 0.00031936616521663905,
      "loss": 0.1787,
      "num_input_tokens_seen": 75384224,
      "step": 39170
    },
    {
      "epoch": 6.390733338771515,
      "grad_norm": 0.007201667875051498,
      "learning_rate": 0.00031553591614619236,
      "loss": 0.1749,
      "num_input_tokens_seen": 75394144,
      "step": 39175
    },
    {
      "epoch": 6.391549065992332,
      "grad_norm": 0.002884375397115946,
      "learning_rate": 0.00031172875030117676,
      "loss": 0.1609,
      "num_input_tokens_seen": 75403552,
      "step": 39180
    },
    {
      "epoch": 6.392364793213149,
      "grad_norm": 0.0026491458993405104,
      "learning_rate": 0.0003079446682686726,
      "loss": 0.1698,
      "num_input_tokens_seen": 75413424,
      "step": 39185
    },
    {
      "epoch": 6.393180520433967,
      "grad_norm": 0.004718845710158348,
      "learning_rate": 0.0003041836706322465,
      "loss": 0.1976,
      "num_input_tokens_seen": 75421776,
      "step": 39190
    },
    {
      "epoch": 6.393996247654784,
      "grad_norm": 0.0025597293861210346,
      "learning_rate": 0.0003004457579719011,
      "loss": 0.2004,
      "num_input_tokens_seen": 75431552,
      "step": 39195
    },
    {
      "epoch": 6.394811974875601,
      "grad_norm": 0.004850753117352724,
      "learning_rate": 0.00029673093086405867,
      "loss": 0.2513,
      "num_input_tokens_seen": 75440784,
      "step": 39200
    },
    {
      "epoch": 6.394811974875601,
      "eval_loss": 0.19021406769752502,
      "eval_runtime": 68.2201,
      "eval_samples_per_second": 39.944,
      "eval_steps_per_second": 19.979,
      "num_input_tokens_seen": 75440784,
      "step": 39200
    },
    {
      "epoch": 6.395627702096419,
      "grad_norm": 0.005779858212918043,
      "learning_rate": 0.00029303918988159426,
      "loss": 0.1131,
      "num_input_tokens_seen": 75449888,
      "step": 39205
    },
    {
      "epoch": 6.3964434293172365,
      "grad_norm": 0.003276464994996786,
      "learning_rate": 0.0002893705355938192,
      "loss": 0.1643,
      "num_input_tokens_seen": 75459856,
      "step": 39210
    },
    {
      "epoch": 6.397259156538054,
      "grad_norm": 0.0038615581579506397,
      "learning_rate": 0.0002857249685664975,
      "loss": 0.146,
      "num_input_tokens_seen": 75469872,
      "step": 39215
    },
    {
      "epoch": 6.398074883758871,
      "grad_norm": 0.002455157460644841,
      "learning_rate": 0.0002821024893618129,
      "loss": 0.1614,
      "num_input_tokens_seen": 75479680,
      "step": 39220
    },
    {
      "epoch": 6.398890610979688,
      "grad_norm": 0.0035996194928884506,
      "learning_rate": 0.0002785030985383852,
      "loss": 0.2153,
      "num_input_tokens_seen": 75488208,
      "step": 39225
    },
    {
      "epoch": 6.399706338200506,
      "grad_norm": 0.0025736785028129816,
      "learning_rate": 0.00027492679665130356,
      "loss": 0.1297,
      "num_input_tokens_seen": 75496928,
      "step": 39230
    },
    {
      "epoch": 6.400522065421323,
      "grad_norm": 0.0021218503825366497,
      "learning_rate": 0.000271373584252077,
      "loss": 0.1344,
      "num_input_tokens_seen": 75505184,
      "step": 39235
    },
    {
      "epoch": 6.40133779264214,
      "grad_norm": 0.0026669749058783054,
      "learning_rate": 0.00026784346188865046,
      "loss": 0.1675,
      "num_input_tokens_seen": 75514464,
      "step": 39240
    },
    {
      "epoch": 6.402153519862958,
      "grad_norm": 0.004410305991768837,
      "learning_rate": 0.0002643364301054218,
      "loss": 0.1964,
      "num_input_tokens_seen": 75523584,
      "step": 39245
    },
    {
      "epoch": 6.402969247083775,
      "grad_norm": 0.0033664866350591183,
      "learning_rate": 0.0002608524894431918,
      "loss": 0.1893,
      "num_input_tokens_seen": 75533280,
      "step": 39250
    },
    {
      "epoch": 6.4037849743045925,
      "grad_norm": 0.0018259932985529304,
      "learning_rate": 0.000257391640439264,
      "loss": 0.1373,
      "num_input_tokens_seen": 75543536,
      "step": 39255
    },
    {
      "epoch": 6.40460070152541,
      "grad_norm": 0.004521674942225218,
      "learning_rate": 0.00025395388362732806,
      "loss": 0.1638,
      "num_input_tokens_seen": 75552992,
      "step": 39260
    },
    {
      "epoch": 6.405416428746228,
      "grad_norm": 0.003245942061766982,
      "learning_rate": 0.00025053921953751,
      "loss": 0.2095,
      "num_input_tokens_seen": 75562112,
      "step": 39265
    },
    {
      "epoch": 6.406232155967045,
      "grad_norm": 0.005720309913158417,
      "learning_rate": 0.00024714764869643855,
      "loss": 0.1898,
      "num_input_tokens_seen": 75571840,
      "step": 39270
    },
    {
      "epoch": 6.407047883187862,
      "grad_norm": 0.004629319999366999,
      "learning_rate": 0.0002437791716270954,
      "loss": 0.1936,
      "num_input_tokens_seen": 75580048,
      "step": 39275
    },
    {
      "epoch": 6.407863610408679,
      "grad_norm": 0.0031378783751279116,
      "learning_rate": 0.00024043378884896493,
      "loss": 0.1721,
      "num_input_tokens_seen": 75589552,
      "step": 39280
    },
    {
      "epoch": 6.408679337629497,
      "grad_norm": 0.0043150135315954685,
      "learning_rate": 0.00023711150087793453,
      "loss": 0.2167,
      "num_input_tokens_seen": 75597440,
      "step": 39285
    },
    {
      "epoch": 6.409495064850314,
      "grad_norm": 0.0024909875355660915,
      "learning_rate": 0.000233812308226361,
      "loss": 0.1932,
      "num_input_tokens_seen": 75607232,
      "step": 39290
    },
    {
      "epoch": 6.410310792071131,
      "grad_norm": 0.0022239619866013527,
      "learning_rate": 0.00023053621140300406,
      "loss": 0.1938,
      "num_input_tokens_seen": 75616464,
      "step": 39295
    },
    {
      "epoch": 6.411126519291948,
      "grad_norm": 0.003520806785672903,
      "learning_rate": 0.00022728321091307623,
      "loss": 0.196,
      "num_input_tokens_seen": 75625520,
      "step": 39300
    },
    {
      "epoch": 6.4119422465127665,
      "grad_norm": 0.005590024404227734,
      "learning_rate": 0.0002240533072582429,
      "loss": 0.2019,
      "num_input_tokens_seen": 75634480,
      "step": 39305
    },
    {
      "epoch": 6.412757973733584,
      "grad_norm": 0.0035862212534993887,
      "learning_rate": 0.00022084650093658897,
      "loss": 0.1758,
      "num_input_tokens_seen": 75645504,
      "step": 39310
    },
    {
      "epoch": 6.413573700954401,
      "grad_norm": 0.0022443998605012894,
      "learning_rate": 0.0002176627924426522,
      "loss": 0.1167,
      "num_input_tokens_seen": 75654736,
      "step": 39315
    },
    {
      "epoch": 6.414389428175218,
      "grad_norm": 0.0029115602374076843,
      "learning_rate": 0.0002145021822673898,
      "loss": 0.1226,
      "num_input_tokens_seen": 75665584,
      "step": 39320
    },
    {
      "epoch": 6.415205155396036,
      "grad_norm": 0.0034359730780124664,
      "learning_rate": 0.00021136467089822862,
      "loss": 0.2184,
      "num_input_tokens_seen": 75674896,
      "step": 39325
    },
    {
      "epoch": 6.416020882616853,
      "grad_norm": 0.0030280037317425013,
      "learning_rate": 0.00020825025881898162,
      "loss": 0.1831,
      "num_input_tokens_seen": 75684992,
      "step": 39330
    },
    {
      "epoch": 6.41683660983767,
      "grad_norm": 0.0030520171858370304,
      "learning_rate": 0.0002051589465099479,
      "loss": 0.1741,
      "num_input_tokens_seen": 75694816,
      "step": 39335
    },
    {
      "epoch": 6.417652337058487,
      "grad_norm": 0.0028747518081218004,
      "learning_rate": 0.0002020907344478462,
      "loss": 0.1448,
      "num_input_tokens_seen": 75705248,
      "step": 39340
    },
    {
      "epoch": 6.418468064279305,
      "grad_norm": 0.00508781848475337,
      "learning_rate": 0.0001990456231058313,
      "loss": 0.156,
      "num_input_tokens_seen": 75714608,
      "step": 39345
    },
    {
      "epoch": 6.419283791500122,
      "grad_norm": 0.002461967756971717,
      "learning_rate": 0.00019602361295349423,
      "loss": 0.1635,
      "num_input_tokens_seen": 75725152,
      "step": 39350
    },
    {
      "epoch": 6.4200995187209395,
      "grad_norm": 0.005410638637840748,
      "learning_rate": 0.0001930247044568789,
      "loss": 0.2253,
      "num_input_tokens_seen": 75733744,
      "step": 39355
    },
    {
      "epoch": 6.420915245941757,
      "grad_norm": 0.004920526407659054,
      "learning_rate": 0.00019004889807843205,
      "loss": 0.1507,
      "num_input_tokens_seen": 75742560,
      "step": 39360
    },
    {
      "epoch": 6.421730973162575,
      "grad_norm": 0.0018603382632136345,
      "learning_rate": 0.00018709619427708656,
      "loss": 0.1647,
      "num_input_tokens_seen": 75753744,
      "step": 39365
    },
    {
      "epoch": 6.422546700383392,
      "grad_norm": 0.0029258395079523325,
      "learning_rate": 0.00018416659350817822,
      "loss": 0.2002,
      "num_input_tokens_seen": 75764064,
      "step": 39370
    },
    {
      "epoch": 6.423362427604209,
      "grad_norm": 0.0016649164026603103,
      "learning_rate": 0.00018126009622346229,
      "loss": 0.1362,
      "num_input_tokens_seen": 75774416,
      "step": 39375
    },
    {
      "epoch": 6.424178154825027,
      "grad_norm": 0.0032183327712118626,
      "learning_rate": 0.00017837670287119687,
      "loss": 0.1576,
      "num_input_tokens_seen": 75784112,
      "step": 39380
    },
    {
      "epoch": 6.424993882045844,
      "grad_norm": 0.004693891387432814,
      "learning_rate": 0.00017551641389602633,
      "loss": 0.2513,
      "num_input_tokens_seen": 75793904,
      "step": 39385
    },
    {
      "epoch": 6.425809609266661,
      "grad_norm": 0.003804979147389531,
      "learning_rate": 0.00017267922973903115,
      "loss": 0.2051,
      "num_input_tokens_seen": 75803088,
      "step": 39390
    },
    {
      "epoch": 6.426625336487478,
      "grad_norm": 0.006690874695777893,
      "learning_rate": 0.00016986515083774467,
      "loss": 0.2197,
      "num_input_tokens_seen": 75812512,
      "step": 39395
    },
    {
      "epoch": 6.4274410637082955,
      "grad_norm": 0.0037238257937133312,
      "learning_rate": 0.00016707417762611975,
      "loss": 0.1817,
      "num_input_tokens_seen": 75822528,
      "step": 39400
    },
    {
      "epoch": 6.4274410637082955,
      "eval_loss": 0.1903531849384308,
      "eval_runtime": 68.0662,
      "eval_samples_per_second": 40.035,
      "eval_steps_per_second": 20.025,
      "num_input_tokens_seen": 75822528,
      "step": 39400
    },
    {
      "epoch": 6.4282567909291135,
      "grad_norm": 0.0034939071629196405,
      "learning_rate": 0.00016430631053459543,
      "loss": 0.1749,
      "num_input_tokens_seen": 75832176,
      "step": 39405
    },
    {
      "epoch": 6.429072518149931,
      "grad_norm": 0.0019916961900889874,
      "learning_rate": 0.0001615615499899803,
      "loss": 0.1762,
      "num_input_tokens_seen": 75842256,
      "step": 39410
    },
    {
      "epoch": 6.429888245370748,
      "grad_norm": 0.0028379373252391815,
      "learning_rate": 0.00015883989641556905,
      "loss": 0.166,
      "num_input_tokens_seen": 75851920,
      "step": 39415
    },
    {
      "epoch": 6.430703972591566,
      "grad_norm": 0.0019750380888581276,
      "learning_rate": 0.00015614135023105934,
      "loss": 0.1218,
      "num_input_tokens_seen": 75860928,
      "step": 39420
    },
    {
      "epoch": 6.431519699812383,
      "grad_norm": 0.003646970260888338,
      "learning_rate": 0.00015346591185261827,
      "loss": 0.208,
      "num_input_tokens_seen": 75871200,
      "step": 39425
    },
    {
      "epoch": 6.4323354270332,
      "grad_norm": 0.0037205335684120655,
      "learning_rate": 0.00015081358169281576,
      "loss": 0.1674,
      "num_input_tokens_seen": 75879296,
      "step": 39430
    },
    {
      "epoch": 6.433151154254017,
      "grad_norm": 0.004531739745289087,
      "learning_rate": 0.00014818436016069135,
      "loss": 0.1573,
      "num_input_tokens_seen": 75887344,
      "step": 39435
    },
    {
      "epoch": 6.433966881474835,
      "grad_norm": 0.00281383260153234,
      "learning_rate": 0.00014557824766168735,
      "loss": 0.1907,
      "num_input_tokens_seen": 75895312,
      "step": 39440
    },
    {
      "epoch": 6.434782608695652,
      "grad_norm": 0.004710738081485033,
      "learning_rate": 0.00014299524459769896,
      "loss": 0.2162,
      "num_input_tokens_seen": 75904752,
      "step": 39445
    },
    {
      "epoch": 6.4355983359164695,
      "grad_norm": 0.005755741149187088,
      "learning_rate": 0.0001404353513670742,
      "loss": 0.1772,
      "num_input_tokens_seen": 75912960,
      "step": 39450
    },
    {
      "epoch": 6.436414063137287,
      "grad_norm": 0.004512040410190821,
      "learning_rate": 0.0001378985683645806,
      "loss": 0.149,
      "num_input_tokens_seen": 75923440,
      "step": 39455
    },
    {
      "epoch": 6.437229790358105,
      "grad_norm": 0.001966046169400215,
      "learning_rate": 0.0001353848959813886,
      "loss": 0.1588,
      "num_input_tokens_seen": 75934720,
      "step": 39460
    },
    {
      "epoch": 6.438045517578922,
      "grad_norm": 0.003665718249976635,
      "learning_rate": 0.00013289433460517142,
      "loss": 0.1464,
      "num_input_tokens_seen": 75943648,
      "step": 39465
    },
    {
      "epoch": 6.438861244799739,
      "grad_norm": 0.0022319729905575514,
      "learning_rate": 0.00013042688462000518,
      "loss": 0.1253,
      "num_input_tokens_seen": 75952736,
      "step": 39470
    },
    {
      "epoch": 6.439676972020556,
      "grad_norm": 0.0024913318920880556,
      "learning_rate": 0.0001279825464063855,
      "loss": 0.2122,
      "num_input_tokens_seen": 75961376,
      "step": 39475
    },
    {
      "epoch": 6.440492699241374,
      "grad_norm": 0.006035480182617903,
      "learning_rate": 0.00012556132034126087,
      "loss": 0.2209,
      "num_input_tokens_seen": 75971040,
      "step": 39480
    },
    {
      "epoch": 6.441308426462191,
      "grad_norm": 0.0030888367909938097,
      "learning_rate": 0.0001231632067980326,
      "loss": 0.1984,
      "num_input_tokens_seen": 75979136,
      "step": 39485
    },
    {
      "epoch": 6.442124153683008,
      "grad_norm": 0.003259108169004321,
      "learning_rate": 0.00012078820614650486,
      "loss": 0.1698,
      "num_input_tokens_seen": 75988176,
      "step": 39490
    },
    {
      "epoch": 6.442939880903825,
      "grad_norm": 0.003314873669296503,
      "learning_rate": 0.00011843631875291804,
      "loss": 0.1724,
      "num_input_tokens_seen": 75997360,
      "step": 39495
    },
    {
      "epoch": 6.443755608124643,
      "grad_norm": 0.0036445611622184515,
      "learning_rate": 0.00011610754497999863,
      "loss": 0.1277,
      "num_input_tokens_seen": 76007952,
      "step": 39500
    },
    {
      "epoch": 6.444571335345461,
      "grad_norm": 0.004523038398474455,
      "learning_rate": 0.0001138018851868594,
      "loss": 0.235,
      "num_input_tokens_seen": 76019344,
      "step": 39505
    },
    {
      "epoch": 6.445387062566278,
      "grad_norm": 0.003418676322326064,
      "learning_rate": 0.0001115193397290326,
      "loss": 0.1399,
      "num_input_tokens_seen": 76027248,
      "step": 39510
    },
    {
      "epoch": 6.446202789787095,
      "grad_norm": 0.001347652287222445,
      "learning_rate": 0.00010925990895856996,
      "loss": 0.1552,
      "num_input_tokens_seen": 76037808,
      "step": 39515
    },
    {
      "epoch": 6.447018517007913,
      "grad_norm": 0.003951231483370066,
      "learning_rate": 0.00010702359322385946,
      "loss": 0.1907,
      "num_input_tokens_seen": 76046928,
      "step": 39520
    },
    {
      "epoch": 6.44783424422873,
      "grad_norm": 0.005307634361088276,
      "learning_rate": 0.00010481039286977523,
      "loss": 0.1848,
      "num_input_tokens_seen": 76056928,
      "step": 39525
    },
    {
      "epoch": 6.448649971449547,
      "grad_norm": 0.003511273767799139,
      "learning_rate": 0.00010262030823764423,
      "loss": 0.1939,
      "num_input_tokens_seen": 76066656,
      "step": 39530
    },
    {
      "epoch": 6.449465698670364,
      "grad_norm": 0.0024607875384390354,
      "learning_rate": 0.00010045333966517966,
      "loss": 0.1476,
      "num_input_tokens_seen": 76076256,
      "step": 39535
    },
    {
      "epoch": 6.450281425891182,
      "grad_norm": 0.002453185385093093,
      "learning_rate": 9.83094874865642e-05,
      "loss": 0.1912,
      "num_input_tokens_seen": 76083840,
      "step": 39540
    },
    {
      "epoch": 6.451097153111999,
      "grad_norm": 0.0032928523141890764,
      "learning_rate": 9.618875203241672e-05,
      "loss": 0.1603,
      "num_input_tokens_seen": 76093440,
      "step": 39545
    },
    {
      "epoch": 6.4519128803328165,
      "grad_norm": 0.002645404078066349,
      "learning_rate": 9.409113362977561e-05,
      "loss": 0.1718,
      "num_input_tokens_seen": 76103008,
      "step": 39550
    },
    {
      "epoch": 6.4527286075536345,
      "grad_norm": 0.004383414052426815,
      "learning_rate": 9.20166326020988e-05,
      "loss": 0.205,
      "num_input_tokens_seen": 76111200,
      "step": 39555
    },
    {
      "epoch": 6.453544334774452,
      "grad_norm": 0.0041679032146930695,
      "learning_rate": 8.996524926933035e-05,
      "loss": 0.2059,
      "num_input_tokens_seen": 76120384,
      "step": 39560
    },
    {
      "epoch": 6.454360061995269,
      "grad_norm": 0.003859882242977619,
      "learning_rate": 8.793698394781723e-05,
      "loss": 0.2197,
      "num_input_tokens_seen": 76130464,
      "step": 39565
    },
    {
      "epoch": 6.455175789216086,
      "grad_norm": 0.0019468554528430104,
      "learning_rate": 8.593183695030926e-05,
      "loss": 0.1075,
      "num_input_tokens_seen": 76140464,
      "step": 39570
    },
    {
      "epoch": 6.455991516436903,
      "grad_norm": 0.002309437608346343,
      "learning_rate": 8.39498085860757e-05,
      "loss": 0.1669,
      "num_input_tokens_seen": 76150736,
      "step": 39575
    },
    {
      "epoch": 6.456807243657721,
      "grad_norm": 0.0027700390201061964,
      "learning_rate": 8.199089916072211e-05,
      "loss": 0.1207,
      "num_input_tokens_seen": 76160512,
      "step": 39580
    },
    {
      "epoch": 6.457622970878538,
      "grad_norm": 0.004957363475114107,
      "learning_rate": 8.005510897637346e-05,
      "loss": 0.1533,
      "num_input_tokens_seen": 76169312,
      "step": 39585
    },
    {
      "epoch": 6.458438698099355,
      "grad_norm": 0.00177411874756217,
      "learning_rate": 7.8142438331541e-05,
      "loss": 0.1364,
      "num_input_tokens_seen": 76179264,
      "step": 39590
    },
    {
      "epoch": 6.459254425320173,
      "grad_norm": 0.0012155435979366302,
      "learning_rate": 7.625288752117209e-05,
      "loss": 0.2042,
      "num_input_tokens_seen": 76188512,
      "step": 39595
    },
    {
      "epoch": 6.4600701525409905,
      "grad_norm": 0.011279961094260216,
      "learning_rate": 7.4386456836667e-05,
      "loss": 0.1354,
      "num_input_tokens_seen": 76198368,
      "step": 39600
    },
    {
      "epoch": 6.4600701525409905,
      "eval_loss": 0.19006377458572388,
      "eval_runtime": 68.1036,
      "eval_samples_per_second": 40.013,
      "eval_steps_per_second": 20.014,
      "num_input_tokens_seen": 76198368,
      "step": 39600
    },
    {
      "epoch": 6.460885879761808,
      "grad_norm": 0.004234002903103828,
      "learning_rate": 7.254314656586214e-05,
      "loss": 0.1371,
      "num_input_tokens_seen": 76209312,
      "step": 39605
    },
    {
      "epoch": 6.461701606982625,
      "grad_norm": 0.0026327415835112333,
      "learning_rate": 7.07229569929968e-05,
      "loss": 0.1502,
      "num_input_tokens_seen": 76218528,
      "step": 39610
    },
    {
      "epoch": 6.462517334203443,
      "grad_norm": 0.004837165120989084,
      "learning_rate": 6.892588839879643e-05,
      "loss": 0.1688,
      "num_input_tokens_seen": 76229056,
      "step": 39615
    },
    {
      "epoch": 6.46333306142426,
      "grad_norm": 0.002837577136233449,
      "learning_rate": 6.71519410603727e-05,
      "loss": 0.1266,
      "num_input_tokens_seen": 76238912,
      "step": 39620
    },
    {
      "epoch": 6.464148788645077,
      "grad_norm": 0.005297747440636158,
      "learning_rate": 6.540111525129011e-05,
      "loss": 0.1629,
      "num_input_tokens_seen": 76249632,
      "step": 39625
    },
    {
      "epoch": 6.464964515865894,
      "grad_norm": 0.0049360995180904865,
      "learning_rate": 6.367341124154934e-05,
      "loss": 0.2294,
      "num_input_tokens_seen": 76258144,
      "step": 39630
    },
    {
      "epoch": 6.465780243086712,
      "grad_norm": 0.0028645333368331194,
      "learning_rate": 6.19688292975873e-05,
      "loss": 0.1402,
      "num_input_tokens_seen": 76267360,
      "step": 39635
    },
    {
      "epoch": 6.466595970307529,
      "grad_norm": 0.002569186268374324,
      "learning_rate": 6.0287369682260336e-05,
      "loss": 0.1453,
      "num_input_tokens_seen": 76276000,
      "step": 39640
    },
    {
      "epoch": 6.467411697528346,
      "grad_norm": 0.005155154038220644,
      "learning_rate": 5.8629032654894384e-05,
      "loss": 0.1822,
      "num_input_tokens_seen": 76285744,
      "step": 39645
    },
    {
      "epoch": 6.468227424749164,
      "grad_norm": 0.0034680068492889404,
      "learning_rate": 5.699381847120155e-05,
      "loss": 0.1379,
      "num_input_tokens_seen": 76295280,
      "step": 39650
    },
    {
      "epoch": 6.469043151969982,
      "grad_norm": 0.001481320010498166,
      "learning_rate": 5.5381727383380094e-05,
      "loss": 0.1458,
      "num_input_tokens_seen": 76304304,
      "step": 39655
    },
    {
      "epoch": 6.469858879190799,
      "grad_norm": 0.0031864922493696213,
      "learning_rate": 5.379275964001451e-05,
      "loss": 0.1831,
      "num_input_tokens_seen": 76313600,
      "step": 39660
    },
    {
      "epoch": 6.470674606411616,
      "grad_norm": 0.0032889251597225666,
      "learning_rate": 5.222691548614211e-05,
      "loss": 0.1507,
      "num_input_tokens_seen": 76322496,
      "step": 39665
    },
    {
      "epoch": 6.471490333632433,
      "grad_norm": 0.004213120322674513,
      "learning_rate": 5.068419516323641e-05,
      "loss": 0.1576,
      "num_input_tokens_seen": 76331520,
      "step": 39670
    },
    {
      "epoch": 6.472306060853251,
      "grad_norm": 0.009887050837278366,
      "learning_rate": 4.91645989092071e-05,
      "loss": 0.1918,
      "num_input_tokens_seen": 76341344,
      "step": 39675
    },
    {
      "epoch": 6.473121788074068,
      "grad_norm": 0.002712773159146309,
      "learning_rate": 4.7668126958400056e-05,
      "loss": 0.1557,
      "num_input_tokens_seen": 76351664,
      "step": 39680
    },
    {
      "epoch": 6.473937515294885,
      "grad_norm": 0.004216181579977274,
      "learning_rate": 4.619477954159734e-05,
      "loss": 0.1432,
      "num_input_tokens_seen": 76360352,
      "step": 39685
    },
    {
      "epoch": 6.474753242515702,
      "grad_norm": 0.00417858874425292,
      "learning_rate": 4.4744556885983884e-05,
      "loss": 0.1859,
      "num_input_tokens_seen": 76370240,
      "step": 39690
    },
    {
      "epoch": 6.47556896973652,
      "grad_norm": 0.0016148792346939445,
      "learning_rate": 4.331745921523078e-05,
      "loss": 0.1587,
      "num_input_tokens_seen": 76379024,
      "step": 39695
    },
    {
      "epoch": 6.4763846969573375,
      "grad_norm": 0.003169122152030468,
      "learning_rate": 4.191348674937867e-05,
      "loss": 0.1813,
      "num_input_tokens_seen": 76388832,
      "step": 39700
    },
    {
      "epoch": 6.477200424178155,
      "grad_norm": 0.0033894460648298264,
      "learning_rate": 4.0532639704971006e-05,
      "loss": 0.1897,
      "num_input_tokens_seen": 76398400,
      "step": 39705
    },
    {
      "epoch": 6.478016151398972,
      "grad_norm": 0.0023546034935861826,
      "learning_rate": 3.917491829493747e-05,
      "loss": 0.111,
      "num_input_tokens_seen": 76408144,
      "step": 39710
    },
    {
      "epoch": 6.47883187861979,
      "grad_norm": 0.0035648702178150415,
      "learning_rate": 3.78403227286439e-05,
      "loss": 0.1823,
      "num_input_tokens_seen": 76416624,
      "step": 39715
    },
    {
      "epoch": 6.479647605840607,
      "grad_norm": 0.005682039074599743,
      "learning_rate": 3.652885321192567e-05,
      "loss": 0.1502,
      "num_input_tokens_seen": 76426912,
      "step": 39720
    },
    {
      "epoch": 6.480463333061424,
      "grad_norm": 0.003121935296803713,
      "learning_rate": 3.524050994702099e-05,
      "loss": 0.1607,
      "num_input_tokens_seen": 76437568,
      "step": 39725
    },
    {
      "epoch": 6.481279060282241,
      "grad_norm": 0.003606152720749378,
      "learning_rate": 3.3975293132604276e-05,
      "loss": 0.1191,
      "num_input_tokens_seen": 76447328,
      "step": 39730
    },
    {
      "epoch": 6.482094787503059,
      "grad_norm": 0.002003840170800686,
      "learning_rate": 3.2733202963786125e-05,
      "loss": 0.107,
      "num_input_tokens_seen": 76457584,
      "step": 39735
    },
    {
      "epoch": 6.482910514723876,
      "grad_norm": 0.002548077842220664,
      "learning_rate": 3.15142396321133e-05,
      "loss": 0.1515,
      "num_input_tokens_seen": 76467152,
      "step": 39740
    },
    {
      "epoch": 6.4837262419446935,
      "grad_norm": 0.0017219556029886007,
      "learning_rate": 3.0318403325552132e-05,
      "loss": 0.1528,
      "num_input_tokens_seen": 76478304,
      "step": 39745
    },
    {
      "epoch": 6.484541969165511,
      "grad_norm": 0.004168633371591568,
      "learning_rate": 2.914569422855506e-05,
      "loss": 0.1626,
      "num_input_tokens_seen": 76486464,
      "step": 39750
    },
    {
      "epoch": 6.485357696386329,
      "grad_norm": 0.0034401477314531803,
      "learning_rate": 2.7996112521927462e-05,
      "loss": 0.1843,
      "num_input_tokens_seen": 76496176,
      "step": 39755
    },
    {
      "epoch": 6.486173423607146,
      "grad_norm": 0.008364561013877392,
      "learning_rate": 2.68696583829775e-05,
      "loss": 0.2221,
      "num_input_tokens_seen": 76505120,
      "step": 39760
    },
    {
      "epoch": 6.486989150827963,
      "grad_norm": 0.0025289002805948257,
      "learning_rate": 2.576633198539957e-05,
      "loss": 0.164,
      "num_input_tokens_seen": 76513968,
      "step": 39765
    },
    {
      "epoch": 6.487804878048781,
      "grad_norm": 0.0032105478458106518,
      "learning_rate": 2.46861334993409e-05,
      "loss": 0.2566,
      "num_input_tokens_seen": 76522912,
      "step": 39770
    },
    {
      "epoch": 6.488620605269598,
      "grad_norm": 0.010897310450673103,
      "learning_rate": 2.3629063091384903e-05,
      "loss": 0.2317,
      "num_input_tokens_seen": 76533248,
      "step": 39775
    },
    {
      "epoch": 6.489436332490415,
      "grad_norm": 0.0028476344887167215,
      "learning_rate": 2.2595120924567834e-05,
      "loss": 0.1944,
      "num_input_tokens_seen": 76543376,
      "step": 39780
    },
    {
      "epoch": 6.490252059711232,
      "grad_norm": 0.003166517009958625,
      "learning_rate": 2.158430715829551e-05,
      "loss": 0.1715,
      "num_input_tokens_seen": 76552096,
      "step": 39785
    },
    {
      "epoch": 6.49106778693205,
      "grad_norm": 0.002204363467171788,
      "learning_rate": 2.059662194849321e-05,
      "loss": 0.1234,
      "num_input_tokens_seen": 76560832,
      "step": 39790
    },
    {
      "epoch": 6.4918835141528675,
      "grad_norm": 0.0032613789662718773,
      "learning_rate": 1.9632065447422463e-05,
      "loss": 0.1732,
      "num_input_tokens_seen": 76570480,
      "step": 39795
    },
    {
      "epoch": 6.492699241373685,
      "grad_norm": 0.003317461349070072,
      "learning_rate": 1.8690637803880916e-05,
      "loss": 0.1974,
      "num_input_tokens_seen": 76581104,
      "step": 39800
    },
    {
      "epoch": 6.492699241373685,
      "eval_loss": 0.1904086470603943,
      "eval_runtime": 68.1245,
      "eval_samples_per_second": 40.0,
      "eval_steps_per_second": 20.007,
      "num_input_tokens_seen": 76581104,
      "step": 39800
    },
    {
      "epoch": 6.493514968594502,
      "grad_norm": 0.004934626165777445,
      "learning_rate": 1.7772339163019123e-05,
      "loss": 0.1806,
      "num_input_tokens_seen": 76590896,
      "step": 39805
    },
    {
      "epoch": 6.49433069581532,
      "grad_norm": 0.0038629386108368635,
      "learning_rate": 1.6877169666457138e-05,
      "loss": 0.2058,
      "num_input_tokens_seen": 76600464,
      "step": 39810
    },
    {
      "epoch": 6.495146423036137,
      "grad_norm": 0.0033426315058022738,
      "learning_rate": 1.6005129452234532e-05,
      "loss": 0.1725,
      "num_input_tokens_seen": 76610864,
      "step": 39815
    },
    {
      "epoch": 6.495962150256954,
      "grad_norm": 0.0072455997578799725,
      "learning_rate": 1.5156218654843733e-05,
      "loss": 0.2651,
      "num_input_tokens_seen": 76620688,
      "step": 39820
    },
    {
      "epoch": 6.496777877477771,
      "grad_norm": 0.002377530559897423,
      "learning_rate": 1.4330437405196683e-05,
      "loss": 0.1395,
      "num_input_tokens_seen": 76630128,
      "step": 39825
    },
    {
      "epoch": 6.497593604698589,
      "grad_norm": 0.003568626008927822,
      "learning_rate": 1.352778583062486e-05,
      "loss": 0.2264,
      "num_input_tokens_seen": 76638128,
      "step": 39830
    },
    {
      "epoch": 6.498409331919406,
      "grad_norm": 0.002445274032652378,
      "learning_rate": 1.2748264054929237e-05,
      "loss": 0.1109,
      "num_input_tokens_seen": 76646032,
      "step": 39835
    },
    {
      "epoch": 6.499225059140223,
      "grad_norm": 0.0029177218675613403,
      "learning_rate": 1.1991872198297004e-05,
      "loss": 0.229,
      "num_input_tokens_seen": 76656768,
      "step": 39840
    },
    {
      "epoch": 6.5000407863610405,
      "grad_norm": 0.0022760722786188126,
      "learning_rate": 1.1258610377384847e-05,
      "loss": 0.1825,
      "num_input_tokens_seen": 76667984,
      "step": 39845
    },
    {
      "epoch": 6.500856513581859,
      "grad_norm": 0.0033981201704591513,
      "learning_rate": 1.0548478705268982e-05,
      "loss": 0.1799,
      "num_input_tokens_seen": 76677392,
      "step": 39850
    },
    {
      "epoch": 6.501672240802676,
      "grad_norm": 0.002810397185385227,
      "learning_rate": 9.86147729147846e-06,
      "loss": 0.145,
      "num_input_tokens_seen": 76687072,
      "step": 39855
    },
    {
      "epoch": 6.502487968023493,
      "grad_norm": 0.0032772142440080643,
      "learning_rate": 9.197606241928557e-06,
      "loss": 0.1557,
      "num_input_tokens_seen": 76696976,
      "step": 39860
    },
    {
      "epoch": 6.50330369524431,
      "grad_norm": 0.0024252801667898893,
      "learning_rate": 8.556865659004042e-06,
      "loss": 0.1433,
      "num_input_tokens_seen": 76706512,
      "step": 39865
    },
    {
      "epoch": 6.504119422465128,
      "grad_norm": 0.003230872331187129,
      "learning_rate": 7.939255641525867e-06,
      "loss": 0.1818,
      "num_input_tokens_seen": 76716496,
      "step": 39870
    },
    {
      "epoch": 6.504935149685945,
      "grad_norm": 0.0021485062316060066,
      "learning_rate": 7.344776284751164e-06,
      "loss": 0.1476,
      "num_input_tokens_seen": 76725744,
      "step": 39875
    },
    {
      "epoch": 6.505750876906762,
      "grad_norm": 0.0029211286455392838,
      "learning_rate": 6.773427680323296e-06,
      "loss": 0.1769,
      "num_input_tokens_seen": 76736368,
      "step": 39880
    },
    {
      "epoch": 6.506566604127579,
      "grad_norm": 0.0045824190601706505,
      "learning_rate": 6.225209916355112e-06,
      "loss": 0.2424,
      "num_input_tokens_seen": 76747024,
      "step": 39885
    },
    {
      "epoch": 6.507382331348397,
      "grad_norm": 0.0021174901630729437,
      "learning_rate": 5.7001230774123e-06,
      "loss": 0.108,
      "num_input_tokens_seen": 76754944,
      "step": 39890
    },
    {
      "epoch": 6.5081980585692145,
      "grad_norm": 0.0027258337941020727,
      "learning_rate": 5.198167244446772e-06,
      "loss": 0.2331,
      "num_input_tokens_seen": 76764368,
      "step": 39895
    },
    {
      "epoch": 6.509013785790032,
      "grad_norm": 0.003880405332893133,
      "learning_rate": 4.71934249487993e-06,
      "loss": 0.1131,
      "num_input_tokens_seen": 76774384,
      "step": 39900
    },
    {
      "epoch": 6.50982951301085,
      "grad_norm": 0.005381994880735874,
      "learning_rate": 4.2636489025527075e-06,
      "loss": 0.1656,
      "num_input_tokens_seen": 76784208,
      "step": 39905
    },
    {
      "epoch": 6.510645240231667,
      "grad_norm": 0.0033271911088377237,
      "learning_rate": 3.831086537742223e-06,
      "loss": 0.1535,
      "num_input_tokens_seen": 76793600,
      "step": 39910
    },
    {
      "epoch": 6.511460967452484,
      "grad_norm": 0.004060279577970505,
      "learning_rate": 3.4216554671451236e-06,
      "loss": 0.109,
      "num_input_tokens_seen": 76802192,
      "step": 39915
    },
    {
      "epoch": 6.512276694673301,
      "grad_norm": 0.004036723170429468,
      "learning_rate": 3.035355753894242e-06,
      "loss": 0.1754,
      "num_input_tokens_seen": 76811264,
      "step": 39920
    },
    {
      "epoch": 6.513092421894118,
      "grad_norm": 0.003489480121061206,
      "learning_rate": 2.6721874575752477e-06,
      "loss": 0.0973,
      "num_input_tokens_seen": 76820224,
      "step": 39925
    },
    {
      "epoch": 6.513908149114936,
      "grad_norm": 0.004523979499936104,
      "learning_rate": 2.3321506341933418e-06,
      "loss": 0.1774,
      "num_input_tokens_seen": 76829824,
      "step": 39930
    },
    {
      "epoch": 6.514723876335753,
      "grad_norm": 0.0053523085080087185,
      "learning_rate": 2.0152453361732546e-06,
      "loss": 0.2046,
      "num_input_tokens_seen": 76838992,
      "step": 39935
    },
    {
      "epoch": 6.5155396035565705,
      "grad_norm": 0.004667972214519978,
      "learning_rate": 1.7214716123925554e-06,
      "loss": 0.1736,
      "num_input_tokens_seen": 76848464,
      "step": 39940
    },
    {
      "epoch": 6.5163553307773885,
      "grad_norm": 0.0028745271265506744,
      "learning_rate": 1.4508295081649968e-06,
      "loss": 0.1486,
      "num_input_tokens_seen": 76858288,
      "step": 39945
    },
    {
      "epoch": 6.517171057998206,
      "grad_norm": 0.0018199475016444921,
      "learning_rate": 1.2033190652238623e-06,
      "loss": 0.1376,
      "num_input_tokens_seen": 76867280,
      "step": 39950
    },
    {
      "epoch": 6.517986785219023,
      "grad_norm": 0.0026946146972477436,
      "learning_rate": 9.78940321721966e-07,
      "loss": 0.1764,
      "num_input_tokens_seen": 76876992,
      "step": 39955
    },
    {
      "epoch": 6.51880251243984,
      "grad_norm": 0.003235730342566967,
      "learning_rate": 7.776933122816132e-07,
      "loss": 0.1644,
      "num_input_tokens_seen": 76887008,
      "step": 39960
    },
    {
      "epoch": 6.519618239660657,
      "grad_norm": 0.004453983157873154,
      "learning_rate": 5.99578067927986e-07,
      "loss": 0.174,
      "num_input_tokens_seen": 76896288,
      "step": 39965
    },
    {
      "epoch": 6.520433966881475,
      "grad_norm": 0.002131012501195073,
      "learning_rate": 4.445946161224512e-07,
      "loss": 0.1172,
      "num_input_tokens_seen": 76906288,
      "step": 39970
    },
    {
      "epoch": 6.521249694102292,
      "grad_norm": 0.0028948646504431963,
      "learning_rate": 3.127429807792126e-07,
      "loss": 0.1329,
      "num_input_tokens_seen": 76914000,
      "step": 39975
    },
    {
      "epoch": 6.522065421323109,
      "grad_norm": 0.0019652743358165026,
      "learning_rate": 2.040231822320049e-07,
      "loss": 0.1612,
      "num_input_tokens_seen": 76923104,
      "step": 39980
    },
    {
      "epoch": 6.522881148543927,
      "grad_norm": 0.0022867931984364986,
      "learning_rate": 1.1843523723409354e-07,
      "loss": 0.1581,
      "num_input_tokens_seen": 76933424,
      "step": 39985
    },
    {
      "epoch": 6.523696875764744,
      "grad_norm": 0.003440155880525708,
      "learning_rate": 5.597915897492811e-08,
      "loss": 0.2303,
      "num_input_tokens_seen": 76943616,
      "step": 39990
    },
    {
      "epoch": 6.524512602985562,
      "grad_norm": 0.002332102507352829,
      "learning_rate": 1.6654957113448885e-08,
      "loss": 0.1853,
      "num_input_tokens_seen": 76953296,
      "step": 39995
    },
    {
      "epoch": 6.525328330206379,
      "grad_norm": 0.003981535788625479,
      "learning_rate": 4.626377114735902e-10,
      "loss": 0.1542,
      "num_input_tokens_seen": 76963024,
      "step": 40000
    },
    {
      "epoch": 6.525328330206379,
      "eval_loss": 0.18989315629005432,
      "eval_runtime": 68.0487,
      "eval_samples_per_second": 40.045,
      "eval_steps_per_second": 20.03,
      "num_input_tokens_seen": 76963024,
      "step": 40000
    },
    {
      "epoch": 6.525328330206379,
      "num_input_tokens_seen": 76963024,
      "step": 40000,
      "total_flos": 3.222731491388252e+17,
      "train_loss": 0.2213717069283128,
      "train_runtime": 23397.8026,
      "train_samples_per_second": 6.838,
      "train_steps_per_second": 1.71
    }
  ],
  "logging_steps": 5,
  "max_steps": 40000,
  "num_input_tokens_seen": 76963024,
  "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": 3.222731491388252e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}