{
  "best_global_step": 18400,
  "best_metric": 4.457133769989014,
  "best_model_checkpoint": "saves/lntuning/llama-3-8b-instruct/train_record_1745950253/checkpoint-18400",
  "epoch": 1.2803277639075603,
  "eval_steps": 200,
  "global_step": 40000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00016004097048844505,
      "grad_norm": 31.0,
      "learning_rate": 4.999999876629946e-05,
      "loss": 5.3873,
      "num_input_tokens_seen": 6960,
      "step": 5
    },
    {
      "epoch": 0.0003200819409768901,
      "grad_norm": 23.625,
      "learning_rate": 4.999999375439123e-05,
      "loss": 5.2115,
      "num_input_tokens_seen": 13728,
      "step": 10
    },
    {
      "epoch": 0.0004801229114653351,
      "grad_norm": 17.625,
      "learning_rate": 4.9999984887169785e-05,
      "loss": 4.8528,
      "num_input_tokens_seen": 20688,
      "step": 15
    },
    {
      "epoch": 0.0006401638819537802,
      "grad_norm": 32.25,
      "learning_rate": 4.9999972164636506e-05,
      "loss": 4.4496,
      "num_input_tokens_seen": 27616,
      "step": 20
    },
    {
      "epoch": 0.0008002048524422252,
      "grad_norm": 54.25,
      "learning_rate": 4.999995558679334e-05,
      "loss": 5.2182,
      "num_input_tokens_seen": 34368,
      "step": 25
    },
    {
      "epoch": 0.0009602458229306702,
      "grad_norm": 23.25,
      "learning_rate": 4.999993515364287e-05,
      "loss": 5.9712,
      "num_input_tokens_seen": 41664,
      "step": 30
    },
    {
      "epoch": 0.0011202867934191152,
      "grad_norm": 26.125,
      "learning_rate": 4.999991086518822e-05,
      "loss": 4.7434,
      "num_input_tokens_seen": 48544,
      "step": 35
    },
    {
      "epoch": 0.0012803277639075604,
      "grad_norm": 33.0,
      "learning_rate": 4.999988272143315e-05,
      "loss": 5.0899,
      "num_input_tokens_seen": 55360,
      "step": 40
    },
    {
      "epoch": 0.0014403687343960054,
      "grad_norm": 33.25,
      "learning_rate": 4.999985072238199e-05,
      "loss": 5.1986,
      "num_input_tokens_seen": 62176,
      "step": 45
    },
    {
      "epoch": 0.0016004097048844504,
      "grad_norm": 20.75,
      "learning_rate": 4.999981486803969e-05,
      "loss": 5.3302,
      "num_input_tokens_seen": 68768,
      "step": 50
    },
    {
      "epoch": 0.0017604506753728954,
      "grad_norm": 34.75,
      "learning_rate": 4.999977515841176e-05,
      "loss": 3.9874,
      "num_input_tokens_seen": 75536,
      "step": 55
    },
    {
      "epoch": 0.0019204916458613404,
      "grad_norm": 42.75,
      "learning_rate": 4.9999731593504344e-05,
      "loss": 5.3052,
      "num_input_tokens_seen": 82320,
      "step": 60
    },
    {
      "epoch": 0.0020805326163497854,
      "grad_norm": 26.625,
      "learning_rate": 4.999968417332415e-05,
      "loss": 4.1852,
      "num_input_tokens_seen": 89184,
      "step": 65
    },
    {
      "epoch": 0.0022405735868382304,
      "grad_norm": 33.25,
      "learning_rate": 4.999963289787848e-05,
      "loss": 3.996,
      "num_input_tokens_seen": 96384,
      "step": 70
    },
    {
      "epoch": 0.002400614557326676,
      "grad_norm": 29.625,
      "learning_rate": 4.999957776717526e-05,
      "loss": 4.1978,
      "num_input_tokens_seen": 102960,
      "step": 75
    },
    {
      "epoch": 0.002560655527815121,
      "grad_norm": 29.0,
      "learning_rate": 4.9999518781222984e-05,
      "loss": 4.2056,
      "num_input_tokens_seen": 109664,
      "step": 80
    },
    {
      "epoch": 0.002720696498303566,
      "grad_norm": 28.125,
      "learning_rate": 4.9999455940030746e-05,
      "loss": 4.9472,
      "num_input_tokens_seen": 116896,
      "step": 85
    },
    {
      "epoch": 0.002880737468792011,
      "grad_norm": 25.25,
      "learning_rate": 4.999938924360824e-05,
      "loss": 5.4813,
      "num_input_tokens_seen": 123808,
      "step": 90
    },
    {
      "epoch": 0.003040778439280456,
      "grad_norm": 31.375,
      "learning_rate": 4.999931869196575e-05,
      "loss": 4.6269,
      "num_input_tokens_seen": 130928,
      "step": 95
    },
    {
      "epoch": 0.003200819409768901,
      "grad_norm": 23.875,
      "learning_rate": 4.999924428511416e-05,
      "loss": 5.1459,
      "num_input_tokens_seen": 137840,
      "step": 100
    },
    {
      "epoch": 0.003360860380257346,
      "grad_norm": 18.5,
      "learning_rate": 4.999916602306494e-05,
      "loss": 4.26,
      "num_input_tokens_seen": 144720,
      "step": 105
    },
    {
      "epoch": 0.003520901350745791,
      "grad_norm": 15.0625,
      "learning_rate": 4.999908390583016e-05,
      "loss": 3.2622,
      "num_input_tokens_seen": 151488,
      "step": 110
    },
    {
      "epoch": 0.003680942321234236,
      "grad_norm": 35.0,
      "learning_rate": 4.999899793342247e-05,
      "loss": 4.8306,
      "num_input_tokens_seen": 158416,
      "step": 115
    },
    {
      "epoch": 0.003840983291722681,
      "grad_norm": 22.625,
      "learning_rate": 4.999890810585516e-05,
      "loss": 5.0245,
      "num_input_tokens_seen": 165024,
      "step": 120
    },
    {
      "epoch": 0.004001024262211126,
      "grad_norm": 37.0,
      "learning_rate": 4.999881442314206e-05,
      "loss": 5.189,
      "num_input_tokens_seen": 171600,
      "step": 125
    },
    {
      "epoch": 0.004161065232699571,
      "grad_norm": 46.25,
      "learning_rate": 4.9998716885297617e-05,
      "loss": 4.8355,
      "num_input_tokens_seen": 178304,
      "step": 130
    },
    {
      "epoch": 0.004321106203188016,
      "grad_norm": 18.875,
      "learning_rate": 4.999861549233688e-05,
      "loss": 4.3904,
      "num_input_tokens_seen": 185216,
      "step": 135
    },
    {
      "epoch": 0.004481147173676461,
      "grad_norm": 34.75,
      "learning_rate": 4.999851024427548e-05,
      "loss": 5.377,
      "num_input_tokens_seen": 191680,
      "step": 140
    },
    {
      "epoch": 0.004641188144164906,
      "grad_norm": 23.75,
      "learning_rate": 4.999840114112965e-05,
      "loss": 4.3272,
      "num_input_tokens_seen": 198080,
      "step": 145
    },
    {
      "epoch": 0.004801229114653352,
      "grad_norm": 32.0,
      "learning_rate": 4.999828818291621e-05,
      "loss": 6.6272,
      "num_input_tokens_seen": 204752,
      "step": 150
    },
    {
      "epoch": 0.004961270085141797,
      "grad_norm": 40.5,
      "learning_rate": 4.999817136965259e-05,
      "loss": 4.4723,
      "num_input_tokens_seen": 211168,
      "step": 155
    },
    {
      "epoch": 0.005121311055630242,
      "grad_norm": 27.75,
      "learning_rate": 4.9998050701356794e-05,
      "loss": 5.4354,
      "num_input_tokens_seen": 217552,
      "step": 160
    },
    {
      "epoch": 0.005281352026118687,
      "grad_norm": 38.25,
      "learning_rate": 4.999792617804744e-05,
      "loss": 4.5776,
      "num_input_tokens_seen": 224464,
      "step": 165
    },
    {
      "epoch": 0.005441392996607132,
      "grad_norm": 22.25,
      "learning_rate": 4.9997797799743724e-05,
      "loss": 5.5844,
      "num_input_tokens_seen": 231232,
      "step": 170
    },
    {
      "epoch": 0.005601433967095577,
      "grad_norm": 20.5,
      "learning_rate": 4.999766556646545e-05,
      "loss": 4.4541,
      "num_input_tokens_seen": 238368,
      "step": 175
    },
    {
      "epoch": 0.005761474937584022,
      "grad_norm": 23.75,
      "learning_rate": 4.9997529478232996e-05,
      "loss": 4.856,
      "num_input_tokens_seen": 245248,
      "step": 180
    },
    {
      "epoch": 0.005921515908072467,
      "grad_norm": 17.75,
      "learning_rate": 4.9997389535067365e-05,
      "loss": 4.9755,
      "num_input_tokens_seen": 252128,
      "step": 185
    },
    {
      "epoch": 0.006081556878560912,
      "grad_norm": 38.5,
      "learning_rate": 4.999724573699012e-05,
      "loss": 4.1852,
      "num_input_tokens_seen": 259232,
      "step": 190
    },
    {
      "epoch": 0.006241597849049357,
      "grad_norm": 46.5,
      "learning_rate": 4.9997098084023457e-05,
      "loss": 5.136,
      "num_input_tokens_seen": 266192,
      "step": 195
    },
    {
      "epoch": 0.006401638819537802,
      "grad_norm": 20.5,
      "learning_rate": 4.999694657619013e-05,
      "loss": 3.4622,
      "num_input_tokens_seen": 272992,
      "step": 200
    },
    {
      "epoch": 0.006401638819537802,
      "eval_loss": 4.9109392166137695,
      "eval_runtime": 487.841,
      "eval_samples_per_second": 28.464,
      "eval_steps_per_second": 14.232,
      "num_input_tokens_seen": 272992,
      "step": 200
    },
    {
      "epoch": 0.006561679790026247,
      "grad_norm": 36.75,
      "learning_rate": 4.999679121351352e-05,
      "loss": 5.1296,
      "num_input_tokens_seen": 279584,
      "step": 205
    },
    {
      "epoch": 0.006721720760514692,
      "grad_norm": 20.0,
      "learning_rate": 4.9996631996017565e-05,
      "loss": 4.6603,
      "num_input_tokens_seen": 286704,
      "step": 210
    },
    {
      "epoch": 0.006881761731003137,
      "grad_norm": 26.375,
      "learning_rate": 4.9996468923726835e-05,
      "loss": 4.3854,
      "num_input_tokens_seen": 293520,
      "step": 215
    },
    {
      "epoch": 0.007041802701491582,
      "grad_norm": 25.625,
      "learning_rate": 4.999630199666647e-05,
      "loss": 5.2557,
      "num_input_tokens_seen": 300032,
      "step": 220
    },
    {
      "epoch": 0.007201843671980027,
      "grad_norm": 21.875,
      "learning_rate": 4.999613121486222e-05,
      "loss": 3.887,
      "num_input_tokens_seen": 306832,
      "step": 225
    },
    {
      "epoch": 0.007361884642468472,
      "grad_norm": 32.25,
      "learning_rate": 4.999595657834041e-05,
      "loss": 4.1801,
      "num_input_tokens_seen": 313664,
      "step": 230
    },
    {
      "epoch": 0.007521925612956917,
      "grad_norm": 19.25,
      "learning_rate": 4.999577808712798e-05,
      "loss": 5.0298,
      "num_input_tokens_seen": 320384,
      "step": 235
    },
    {
      "epoch": 0.007681966583445362,
      "grad_norm": 31.625,
      "learning_rate": 4.999559574125244e-05,
      "loss": 4.8527,
      "num_input_tokens_seen": 326896,
      "step": 240
    },
    {
      "epoch": 0.007842007553933807,
      "grad_norm": 34.75,
      "learning_rate": 4.9995409540741934e-05,
      "loss": 5.0211,
      "num_input_tokens_seen": 333760,
      "step": 245
    },
    {
      "epoch": 0.008002048524422252,
      "grad_norm": 19.625,
      "learning_rate": 4.999521948562516e-05,
      "loss": 4.6165,
      "num_input_tokens_seen": 340288,
      "step": 250
    },
    {
      "epoch": 0.008162089494910697,
      "grad_norm": 25.5,
      "learning_rate": 4.999502557593143e-05,
      "loss": 5.3668,
      "num_input_tokens_seen": 347136,
      "step": 255
    },
    {
      "epoch": 0.008322130465399142,
      "grad_norm": 27.25,
      "learning_rate": 4.999482781169066e-05,
      "loss": 4.8469,
      "num_input_tokens_seen": 353664,
      "step": 260
    },
    {
      "epoch": 0.008482171435887587,
      "grad_norm": 23.125,
      "learning_rate": 4.9994626192933324e-05,
      "loss": 4.5681,
      "num_input_tokens_seen": 360368,
      "step": 265
    },
    {
      "epoch": 0.008642212406376032,
      "grad_norm": 41.5,
      "learning_rate": 4.999442071969054e-05,
      "loss": 4.4311,
      "num_input_tokens_seen": 366928,
      "step": 270
    },
    {
      "epoch": 0.008802253376864477,
      "grad_norm": 71.5,
      "learning_rate": 4.999421139199397e-05,
      "loss": 3.9936,
      "num_input_tokens_seen": 373904,
      "step": 275
    },
    {
      "epoch": 0.008962294347352922,
      "grad_norm": 28.875,
      "learning_rate": 4.999399820987592e-05,
      "loss": 4.2904,
      "num_input_tokens_seen": 380432,
      "step": 280
    },
    {
      "epoch": 0.009122335317841367,
      "grad_norm": 31.25,
      "learning_rate": 4.999378117336924e-05,
      "loss": 4.7234,
      "num_input_tokens_seen": 387312,
      "step": 285
    },
    {
      "epoch": 0.009282376288329812,
      "grad_norm": 25.125,
      "learning_rate": 4.9993560282507415e-05,
      "loss": 5.1889,
      "num_input_tokens_seen": 393904,
      "step": 290
    },
    {
      "epoch": 0.009442417258818258,
      "grad_norm": 34.25,
      "learning_rate": 4.9993335537324495e-05,
      "loss": 4.6285,
      "num_input_tokens_seen": 400688,
      "step": 295
    },
    {
      "epoch": 0.009602458229306703,
      "grad_norm": 12.3125,
      "learning_rate": 4.999310693785516e-05,
      "loss": 3.6803,
      "num_input_tokens_seen": 407248,
      "step": 300
    },
    {
      "epoch": 0.009762499199795148,
      "grad_norm": 35.25,
      "learning_rate": 4.9992874484134653e-05,
      "loss": 4.0898,
      "num_input_tokens_seen": 414096,
      "step": 305
    },
    {
      "epoch": 0.009922540170283593,
      "grad_norm": 41.0,
      "learning_rate": 4.999263817619882e-05,
      "loss": 5.751,
      "num_input_tokens_seen": 420832,
      "step": 310
    },
    {
      "epoch": 0.010082581140772038,
      "grad_norm": 12.9375,
      "learning_rate": 4.9992398014084105e-05,
      "loss": 4.6977,
      "num_input_tokens_seen": 427312,
      "step": 315
    },
    {
      "epoch": 0.010242622111260483,
      "grad_norm": 46.25,
      "learning_rate": 4.999215399782754e-05,
      "loss": 4.8153,
      "num_input_tokens_seen": 433680,
      "step": 320
    },
    {
      "epoch": 0.010402663081748928,
      "grad_norm": 20.5,
      "learning_rate": 4.999190612746675e-05,
      "loss": 4.7057,
      "num_input_tokens_seen": 440368,
      "step": 325
    },
    {
      "epoch": 0.010562704052237373,
      "grad_norm": 30.25,
      "learning_rate": 4.999165440303998e-05,
      "loss": 4.1927,
      "num_input_tokens_seen": 446736,
      "step": 330
    },
    {
      "epoch": 0.010722745022725818,
      "grad_norm": 12.5,
      "learning_rate": 4.999139882458603e-05,
      "loss": 4.6765,
      "num_input_tokens_seen": 453120,
      "step": 335
    },
    {
      "epoch": 0.010882785993214263,
      "grad_norm": 41.75,
      "learning_rate": 4.9991139392144314e-05,
      "loss": 5.1898,
      "num_input_tokens_seen": 459840,
      "step": 340
    },
    {
      "epoch": 0.011042826963702708,
      "grad_norm": 18.5,
      "learning_rate": 4.999087610575485e-05,
      "loss": 4.231,
      "num_input_tokens_seen": 466384,
      "step": 345
    },
    {
      "epoch": 0.011202867934191153,
      "grad_norm": 27.625,
      "learning_rate": 4.999060896545824e-05,
      "loss": 5.2959,
      "num_input_tokens_seen": 473328,
      "step": 350
    },
    {
      "epoch": 0.011362908904679598,
      "grad_norm": 13.75,
      "learning_rate": 4.999033797129568e-05,
      "loss": 4.0298,
      "num_input_tokens_seen": 480448,
      "step": 355
    },
    {
      "epoch": 0.011522949875168043,
      "grad_norm": 25.625,
      "learning_rate": 4.999006312330894e-05,
      "loss": 5.4019,
      "num_input_tokens_seen": 487408,
      "step": 360
    },
    {
      "epoch": 0.011682990845656488,
      "grad_norm": 24.875,
      "learning_rate": 4.998978442154043e-05,
      "loss": 5.0847,
      "num_input_tokens_seen": 494160,
      "step": 365
    },
    {
      "epoch": 0.011843031816144933,
      "grad_norm": 24.125,
      "learning_rate": 4.9989501866033125e-05,
      "loss": 5.8768,
      "num_input_tokens_seen": 500480,
      "step": 370
    },
    {
      "epoch": 0.012003072786633378,
      "grad_norm": 26.375,
      "learning_rate": 4.998921545683059e-05,
      "loss": 4.2189,
      "num_input_tokens_seen": 507424,
      "step": 375
    },
    {
      "epoch": 0.012163113757121823,
      "grad_norm": 32.25,
      "learning_rate": 4.9988925193976996e-05,
      "loss": 3.638,
      "num_input_tokens_seen": 514064,
      "step": 380
    },
    {
      "epoch": 0.012323154727610268,
      "grad_norm": 17.0,
      "learning_rate": 4.998863107751711e-05,
      "loss": 3.8511,
      "num_input_tokens_seen": 520816,
      "step": 385
    },
    {
      "epoch": 0.012483195698098713,
      "grad_norm": 28.125,
      "learning_rate": 4.998833310749629e-05,
      "loss": 4.3198,
      "num_input_tokens_seen": 527328,
      "step": 390
    },
    {
      "epoch": 0.012643236668587158,
      "grad_norm": 31.125,
      "learning_rate": 4.998803128396047e-05,
      "loss": 3.6676,
      "num_input_tokens_seen": 534656,
      "step": 395
    },
    {
      "epoch": 0.012803277639075603,
      "grad_norm": 17.875,
      "learning_rate": 4.9987725606956215e-05,
      "loss": 4.4474,
      "num_input_tokens_seen": 541536,
      "step": 400
    },
    {
      "epoch": 0.012803277639075603,
      "eval_loss": 4.677643299102783,
      "eval_runtime": 488.0443,
      "eval_samples_per_second": 28.452,
      "eval_steps_per_second": 14.226,
      "num_input_tokens_seen": 541536,
      "step": 400
    },
    {
      "epoch": 0.012963318609564048,
      "grad_norm": 22.125,
      "learning_rate": 4.998741607653066e-05,
      "loss": 4.1912,
      "num_input_tokens_seen": 548160,
      "step": 405
    },
    {
      "epoch": 0.013123359580052493,
      "grad_norm": 20.25,
      "learning_rate": 4.9987102692731523e-05,
      "loss": 4.6597,
      "num_input_tokens_seen": 555024,
      "step": 410
    },
    {
      "epoch": 0.013283400550540938,
      "grad_norm": 40.75,
      "learning_rate": 4.9986785455607157e-05,
      "loss": 4.6052,
      "num_input_tokens_seen": 561792,
      "step": 415
    },
    {
      "epoch": 0.013443441521029383,
      "grad_norm": 29.375,
      "learning_rate": 4.9986464365206456e-05,
      "loss": 4.3191,
      "num_input_tokens_seen": 568464,
      "step": 420
    },
    {
      "epoch": 0.013603482491517828,
      "grad_norm": 37.75,
      "learning_rate": 4.9986139421578956e-05,
      "loss": 4.4763,
      "num_input_tokens_seen": 574928,
      "step": 425
    },
    {
      "epoch": 0.013763523462006273,
      "grad_norm": 24.875,
      "learning_rate": 4.998581062477477e-05,
      "loss": 3.8146,
      "num_input_tokens_seen": 581792,
      "step": 430
    },
    {
      "epoch": 0.013923564432494718,
      "grad_norm": 24.125,
      "learning_rate": 4.998547797484458e-05,
      "loss": 4.5159,
      "num_input_tokens_seen": 588512,
      "step": 435
    },
    {
      "epoch": 0.014083605402983163,
      "grad_norm": 31.125,
      "learning_rate": 4.9985141471839706e-05,
      "loss": 4.7663,
      "num_input_tokens_seen": 595168,
      "step": 440
    },
    {
      "epoch": 0.014243646373471608,
      "grad_norm": 36.25,
      "learning_rate": 4.998480111581203e-05,
      "loss": 4.2657,
      "num_input_tokens_seen": 601968,
      "step": 445
    },
    {
      "epoch": 0.014403687343960053,
      "grad_norm": 35.25,
      "learning_rate": 4.998445690681405e-05,
      "loss": 5.4711,
      "num_input_tokens_seen": 608432,
      "step": 450
    },
    {
      "epoch": 0.014563728314448498,
      "grad_norm": 29.0,
      "learning_rate": 4.9984108844898834e-05,
      "loss": 4.3682,
      "num_input_tokens_seen": 615248,
      "step": 455
    },
    {
      "epoch": 0.014723769284936943,
      "grad_norm": 22.75,
      "learning_rate": 4.9983756930120076e-05,
      "loss": 4.0379,
      "num_input_tokens_seen": 622432,
      "step": 460
    },
    {
      "epoch": 0.014883810255425388,
      "grad_norm": 32.25,
      "learning_rate": 4.9983401162532025e-05,
      "loss": 3.9806,
      "num_input_tokens_seen": 629600,
      "step": 465
    },
    {
      "epoch": 0.015043851225913833,
      "grad_norm": 36.5,
      "learning_rate": 4.998304154218955e-05,
      "loss": 3.9275,
      "num_input_tokens_seen": 636640,
      "step": 470
    },
    {
      "epoch": 0.015203892196402278,
      "grad_norm": 23.625,
      "learning_rate": 4.998267806914812e-05,
      "loss": 3.9174,
      "num_input_tokens_seen": 644128,
      "step": 475
    },
    {
      "epoch": 0.015363933166890723,
      "grad_norm": 28.25,
      "learning_rate": 4.998231074346378e-05,
      "loss": 4.5791,
      "num_input_tokens_seen": 650416,
      "step": 480
    },
    {
      "epoch": 0.015523974137379168,
      "grad_norm": 18.375,
      "learning_rate": 4.998193956519317e-05,
      "loss": 3.7922,
      "num_input_tokens_seen": 656896,
      "step": 485
    },
    {
      "epoch": 0.015684015107867613,
      "grad_norm": 12.4375,
      "learning_rate": 4.9981564534393545e-05,
      "loss": 4.1018,
      "num_input_tokens_seen": 663440,
      "step": 490
    },
    {
      "epoch": 0.01584405607835606,
      "grad_norm": 43.0,
      "learning_rate": 4.998118565112272e-05,
      "loss": 3.956,
      "num_input_tokens_seen": 670128,
      "step": 495
    },
    {
      "epoch": 0.016004097048844503,
      "grad_norm": 30.625,
      "learning_rate": 4.998080291543914e-05,
      "loss": 5.4005,
      "num_input_tokens_seen": 676656,
      "step": 500
    },
    {
      "epoch": 0.01616413801933295,
      "grad_norm": 16.875,
      "learning_rate": 4.9980416327401826e-05,
      "loss": 4.1971,
      "num_input_tokens_seen": 683392,
      "step": 505
    },
    {
      "epoch": 0.016324178989821393,
      "grad_norm": 38.5,
      "learning_rate": 4.998002588707038e-05,
      "loss": 3.4296,
      "num_input_tokens_seen": 690496,
      "step": 510
    },
    {
      "epoch": 0.01648421996030984,
      "grad_norm": 27.625,
      "learning_rate": 4.997963159450503e-05,
      "loss": 3.8914,
      "num_input_tokens_seen": 697632,
      "step": 515
    },
    {
      "epoch": 0.016644260930798283,
      "grad_norm": 44.25,
      "learning_rate": 4.9979233449766575e-05,
      "loss": 5.0921,
      "num_input_tokens_seen": 705136,
      "step": 520
    },
    {
      "epoch": 0.01680430190128673,
      "grad_norm": 22.875,
      "learning_rate": 4.997883145291641e-05,
      "loss": 5.3732,
      "num_input_tokens_seen": 711712,
      "step": 525
    },
    {
      "epoch": 0.016964342871775173,
      "grad_norm": 29.625,
      "learning_rate": 4.9978425604016536e-05,
      "loss": 4.3989,
      "num_input_tokens_seen": 718688,
      "step": 530
    },
    {
      "epoch": 0.01712438384226362,
      "grad_norm": 37.0,
      "learning_rate": 4.9978015903129536e-05,
      "loss": 4.4974,
      "num_input_tokens_seen": 725664,
      "step": 535
    },
    {
      "epoch": 0.017284424812752063,
      "grad_norm": 29.0,
      "learning_rate": 4.997760235031859e-05,
      "loss": 6.2601,
      "num_input_tokens_seen": 732336,
      "step": 540
    },
    {
      "epoch": 0.01744446578324051,
      "grad_norm": 32.25,
      "learning_rate": 4.9977184945647473e-05,
      "loss": 3.9743,
      "num_input_tokens_seen": 739024,
      "step": 545
    },
    {
      "epoch": 0.017604506753728953,
      "grad_norm": 27.875,
      "learning_rate": 4.997676368918055e-05,
      "loss": 4.5499,
      "num_input_tokens_seen": 745952,
      "step": 550
    },
    {
      "epoch": 0.0177645477242174,
      "grad_norm": 35.5,
      "learning_rate": 4.9976338580982794e-05,
      "loss": 5.5751,
      "num_input_tokens_seen": 752336,
      "step": 555
    },
    {
      "epoch": 0.017924588694705843,
      "grad_norm": 19.625,
      "learning_rate": 4.9975909621119755e-05,
      "loss": 4.7718,
      "num_input_tokens_seen": 759616,
      "step": 560
    },
    {
      "epoch": 0.01808462966519429,
      "grad_norm": 23.25,
      "learning_rate": 4.997547680965758e-05,
      "loss": 3.9131,
      "num_input_tokens_seen": 766768,
      "step": 565
    },
    {
      "epoch": 0.018244670635682733,
      "grad_norm": 48.0,
      "learning_rate": 4.997504014666302e-05,
      "loss": 4.9897,
      "num_input_tokens_seen": 773200,
      "step": 570
    },
    {
      "epoch": 0.01840471160617118,
      "grad_norm": 19.375,
      "learning_rate": 4.997459963220342e-05,
      "loss": 4.0444,
      "num_input_tokens_seen": 779920,
      "step": 575
    },
    {
      "epoch": 0.018564752576659623,
      "grad_norm": 13.625,
      "learning_rate": 4.997415526634671e-05,
      "loss": 3.6871,
      "num_input_tokens_seen": 786688,
      "step": 580
    },
    {
      "epoch": 0.01872479354714807,
      "grad_norm": 44.5,
      "learning_rate": 4.99737070491614e-05,
      "loss": 4.7699,
      "num_input_tokens_seen": 793840,
      "step": 585
    },
    {
      "epoch": 0.018884834517636517,
      "grad_norm": 50.5,
      "learning_rate": 4.997325498071663e-05,
      "loss": 4.2241,
      "num_input_tokens_seen": 800448,
      "step": 590
    },
    {
      "epoch": 0.01904487548812496,
      "grad_norm": 29.875,
      "learning_rate": 4.997279906108211e-05,
      "loss": 4.6113,
      "num_input_tokens_seen": 807456,
      "step": 595
    },
    {
      "epoch": 0.019204916458613407,
      "grad_norm": 26.125,
      "learning_rate": 4.9972339290328155e-05,
      "loss": 4.3464,
      "num_input_tokens_seen": 813648,
      "step": 600
    },
    {
      "epoch": 0.019204916458613407,
      "eval_loss": 4.574725151062012,
      "eval_runtime": 487.297,
      "eval_samples_per_second": 28.496,
      "eval_steps_per_second": 14.248,
      "num_input_tokens_seen": 813648,
      "step": 600
    },
    {
      "epoch": 0.01936495742910185,
      "grad_norm": 19.625,
      "learning_rate": 4.9971875668525646e-05,
      "loss": 3.7043,
      "num_input_tokens_seen": 820368,
      "step": 605
    },
    {
      "epoch": 0.019524998399590297,
      "grad_norm": 23.75,
      "learning_rate": 4.997140819574609e-05,
      "loss": 5.1095,
      "num_input_tokens_seen": 827792,
      "step": 610
    },
    {
      "epoch": 0.01968503937007874,
      "grad_norm": 36.0,
      "learning_rate": 4.997093687206159e-05,
      "loss": 4.0532,
      "num_input_tokens_seen": 834256,
      "step": 615
    },
    {
      "epoch": 0.019845080340567187,
      "grad_norm": 39.25,
      "learning_rate": 4.997046169754482e-05,
      "loss": 4.6611,
      "num_input_tokens_seen": 840960,
      "step": 620
    },
    {
      "epoch": 0.02000512131105563,
      "grad_norm": 25.125,
      "learning_rate": 4.996998267226905e-05,
      "loss": 4.4177,
      "num_input_tokens_seen": 847616,
      "step": 625
    },
    {
      "epoch": 0.020165162281544077,
      "grad_norm": 79.0,
      "learning_rate": 4.996949979630817e-05,
      "loss": 3.764,
      "num_input_tokens_seen": 854320,
      "step": 630
    },
    {
      "epoch": 0.02032520325203252,
      "grad_norm": 57.25,
      "learning_rate": 4.996901306973663e-05,
      "loss": 4.6675,
      "num_input_tokens_seen": 860752,
      "step": 635
    },
    {
      "epoch": 0.020485244222520967,
      "grad_norm": 31.125,
      "learning_rate": 4.996852249262949e-05,
      "loss": 4.4102,
      "num_input_tokens_seen": 867792,
      "step": 640
    },
    {
      "epoch": 0.02064528519300941,
      "grad_norm": 27.625,
      "learning_rate": 4.996802806506241e-05,
      "loss": 5.2286,
      "num_input_tokens_seen": 874768,
      "step": 645
    },
    {
      "epoch": 0.020805326163497857,
      "grad_norm": 28.0,
      "learning_rate": 4.996752978711164e-05,
      "loss": 4.8788,
      "num_input_tokens_seen": 881712,
      "step": 650
    },
    {
      "epoch": 0.0209653671339863,
      "grad_norm": 56.25,
      "learning_rate": 4.996702765885401e-05,
      "loss": 4.5701,
      "num_input_tokens_seen": 888848,
      "step": 655
    },
    {
      "epoch": 0.021125408104474747,
      "grad_norm": 36.5,
      "learning_rate": 4.9966521680366964e-05,
      "loss": 3.8523,
      "num_input_tokens_seen": 895168,
      "step": 660
    },
    {
      "epoch": 0.02128544907496319,
      "grad_norm": 30.25,
      "learning_rate": 4.9966011851728524e-05,
      "loss": 4.771,
      "num_input_tokens_seen": 902224,
      "step": 665
    },
    {
      "epoch": 0.021445490045451637,
      "grad_norm": 33.75,
      "learning_rate": 4.996549817301731e-05,
      "loss": 4.3128,
      "num_input_tokens_seen": 909136,
      "step": 670
    },
    {
      "epoch": 0.02160553101594008,
      "grad_norm": 41.0,
      "learning_rate": 4.9964980644312544e-05,
      "loss": 4.3373,
      "num_input_tokens_seen": 915776,
      "step": 675
    },
    {
      "epoch": 0.021765571986428527,
      "grad_norm": 42.5,
      "learning_rate": 4.996445926569403e-05,
      "loss": 4.6962,
      "num_input_tokens_seen": 922544,
      "step": 680
    },
    {
      "epoch": 0.02192561295691697,
      "grad_norm": 30.625,
      "learning_rate": 4.996393403724218e-05,
      "loss": 4.8212,
      "num_input_tokens_seen": 929120,
      "step": 685
    },
    {
      "epoch": 0.022085653927405417,
      "grad_norm": 42.5,
      "learning_rate": 4.9963404959037985e-05,
      "loss": 5.6861,
      "num_input_tokens_seen": 936528,
      "step": 690
    },
    {
      "epoch": 0.02224569489789386,
      "grad_norm": 55.0,
      "learning_rate": 4.996287203116303e-05,
      "loss": 4.8151,
      "num_input_tokens_seen": 943296,
      "step": 695
    },
    {
      "epoch": 0.022405735868382307,
      "grad_norm": 30.75,
      "learning_rate": 4.996233525369951e-05,
      "loss": 4.498,
      "num_input_tokens_seen": 949616,
      "step": 700
    },
    {
      "epoch": 0.02256577683887075,
      "grad_norm": 33.0,
      "learning_rate": 4.99617946267302e-05,
      "loss": 3.9566,
      "num_input_tokens_seen": 956048,
      "step": 705
    },
    {
      "epoch": 0.022725817809359197,
      "grad_norm": 19.25,
      "learning_rate": 4.996125015033846e-05,
      "loss": 3.6571,
      "num_input_tokens_seen": 962736,
      "step": 710
    },
    {
      "epoch": 0.02288585877984764,
      "grad_norm": 39.25,
      "learning_rate": 4.996070182460827e-05,
      "loss": 4.7216,
      "num_input_tokens_seen": 969856,
      "step": 715
    },
    {
      "epoch": 0.023045899750336087,
      "grad_norm": 37.5,
      "learning_rate": 4.996014964962418e-05,
      "loss": 3.9294,
      "num_input_tokens_seen": 976672,
      "step": 720
    },
    {
      "epoch": 0.02320594072082453,
      "grad_norm": 33.5,
      "learning_rate": 4.9959593625471344e-05,
      "loss": 4.161,
      "num_input_tokens_seen": 983696,
      "step": 725
    },
    {
      "epoch": 0.023365981691312977,
      "grad_norm": 39.75,
      "learning_rate": 4.995903375223552e-05,
      "loss": 4.5668,
      "num_input_tokens_seen": 990512,
      "step": 730
    },
    {
      "epoch": 0.02352602266180142,
      "grad_norm": 18.125,
      "learning_rate": 4.995847003000302e-05,
      "loss": 4.7046,
      "num_input_tokens_seen": 997312,
      "step": 735
    },
    {
      "epoch": 0.023686063632289867,
      "grad_norm": 24.0,
      "learning_rate": 4.9957902458860804e-05,
      "loss": 4.6541,
      "num_input_tokens_seen": 1003712,
      "step": 740
    },
    {
      "epoch": 0.02384610460277831,
      "grad_norm": 20.5,
      "learning_rate": 4.995733103889639e-05,
      "loss": 3.9137,
      "num_input_tokens_seen": 1010368,
      "step": 745
    },
    {
      "epoch": 0.024006145573266757,
      "grad_norm": 32.5,
      "learning_rate": 4.99567557701979e-05,
      "loss": 4.8488,
      "num_input_tokens_seen": 1017152,
      "step": 750
    },
    {
      "epoch": 0.0241661865437552,
      "grad_norm": 30.0,
      "learning_rate": 4.995617665285403e-05,
      "loss": 4.2658,
      "num_input_tokens_seen": 1023808,
      "step": 755
    },
    {
      "epoch": 0.024326227514243647,
      "grad_norm": 21.875,
      "learning_rate": 4.99555936869541e-05,
      "loss": 4.503,
      "num_input_tokens_seen": 1030656,
      "step": 760
    },
    {
      "epoch": 0.02448626848473209,
      "grad_norm": 20.0,
      "learning_rate": 4.995500687258803e-05,
      "loss": 4.4156,
      "num_input_tokens_seen": 1037696,
      "step": 765
    },
    {
      "epoch": 0.024646309455220537,
      "grad_norm": 44.5,
      "learning_rate": 4.995441620984628e-05,
      "loss": 5.1225,
      "num_input_tokens_seen": 1044240,
      "step": 770
    },
    {
      "epoch": 0.02480635042570898,
      "grad_norm": 32.25,
      "learning_rate": 4.995382169881996e-05,
      "loss": 5.3643,
      "num_input_tokens_seen": 1050528,
      "step": 775
    },
    {
      "epoch": 0.024966391396197427,
      "grad_norm": 50.5,
      "learning_rate": 4.9953223339600755e-05,
      "loss": 4.677,
      "num_input_tokens_seen": 1057408,
      "step": 780
    },
    {
      "epoch": 0.025126432366685873,
      "grad_norm": 15.125,
      "learning_rate": 4.995262113228091e-05,
      "loss": 3.8362,
      "num_input_tokens_seen": 1063984,
      "step": 785
    },
    {
      "epoch": 0.025286473337174317,
      "grad_norm": 33.75,
      "learning_rate": 4.995201507695332e-05,
      "loss": 4.3126,
      "num_input_tokens_seen": 1070640,
      "step": 790
    },
    {
      "epoch": 0.025446514307662763,
      "grad_norm": 29.375,
      "learning_rate": 4.995140517371144e-05,
      "loss": 4.155,
      "num_input_tokens_seen": 1077456,
      "step": 795
    },
    {
      "epoch": 0.025606555278151207,
      "grad_norm": 68.0,
      "learning_rate": 4.995079142264932e-05,
      "loss": 4.3474,
      "num_input_tokens_seen": 1084496,
      "step": 800
    },
    {
      "epoch": 0.025606555278151207,
      "eval_loss": 4.540164947509766,
      "eval_runtime": 487.8109,
      "eval_samples_per_second": 28.466,
      "eval_steps_per_second": 14.233,
      "num_input_tokens_seen": 1084496,
      "step": 800
    },
    {
      "epoch": 0.025766596248639653,
      "grad_norm": 30.375,
      "learning_rate": 4.995017382386162e-05,
      "loss": 4.9233,
      "num_input_tokens_seen": 1091504,
      "step": 805
    },
    {
      "epoch": 0.025926637219128097,
      "grad_norm": 28.625,
      "learning_rate": 4.994955237744356e-05,
      "loss": 3.8379,
      "num_input_tokens_seen": 1098496,
      "step": 810
    },
    {
      "epoch": 0.026086678189616543,
      "grad_norm": 30.0,
      "learning_rate": 4.994892708349101e-05,
      "loss": 4.0409,
      "num_input_tokens_seen": 1105024,
      "step": 815
    },
    {
      "epoch": 0.026246719160104987,
      "grad_norm": 28.0,
      "learning_rate": 4.994829794210035e-05,
      "loss": 4.3391,
      "num_input_tokens_seen": 1111408,
      "step": 820
    },
    {
      "epoch": 0.026406760130593433,
      "grad_norm": 17.5,
      "learning_rate": 4.994766495336864e-05,
      "loss": 4.1037,
      "num_input_tokens_seen": 1118496,
      "step": 825
    },
    {
      "epoch": 0.026566801101081877,
      "grad_norm": 16.25,
      "learning_rate": 4.994702811739348e-05,
      "loss": 4.3899,
      "num_input_tokens_seen": 1125504,
      "step": 830
    },
    {
      "epoch": 0.026726842071570323,
      "grad_norm": 28.5,
      "learning_rate": 4.994638743427308e-05,
      "loss": 3.9608,
      "num_input_tokens_seen": 1132544,
      "step": 835
    },
    {
      "epoch": 0.026886883042058767,
      "grad_norm": 16.5,
      "learning_rate": 4.994574290410624e-05,
      "loss": 4.3352,
      "num_input_tokens_seen": 1139296,
      "step": 840
    },
    {
      "epoch": 0.027046924012547213,
      "grad_norm": 21.625,
      "learning_rate": 4.9945094526992364e-05,
      "loss": 4.3339,
      "num_input_tokens_seen": 1145696,
      "step": 845
    },
    {
      "epoch": 0.027206964983035656,
      "grad_norm": 47.25,
      "learning_rate": 4.994444230303142e-05,
      "loss": 4.6906,
      "num_input_tokens_seen": 1152416,
      "step": 850
    },
    {
      "epoch": 0.027367005953524103,
      "grad_norm": 33.25,
      "learning_rate": 4.994378623232402e-05,
      "loss": 4.4818,
      "num_input_tokens_seen": 1159072,
      "step": 855
    },
    {
      "epoch": 0.027527046924012546,
      "grad_norm": 36.75,
      "learning_rate": 4.99431263149713e-05,
      "loss": 5.3903,
      "num_input_tokens_seen": 1165696,
      "step": 860
    },
    {
      "epoch": 0.027687087894500993,
      "grad_norm": 22.75,
      "learning_rate": 4.9942462551075056e-05,
      "loss": 5.3147,
      "num_input_tokens_seen": 1172416,
      "step": 865
    },
    {
      "epoch": 0.027847128864989436,
      "grad_norm": 30.25,
      "learning_rate": 4.994179494073764e-05,
      "loss": 4.603,
      "num_input_tokens_seen": 1179136,
      "step": 870
    },
    {
      "epoch": 0.028007169835477883,
      "grad_norm": 20.125,
      "learning_rate": 4.9941123484062e-05,
      "loss": 4.0198,
      "num_input_tokens_seen": 1186144,
      "step": 875
    },
    {
      "epoch": 0.028167210805966326,
      "grad_norm": 35.75,
      "learning_rate": 4.99404481811517e-05,
      "loss": 4.1003,
      "num_input_tokens_seen": 1192624,
      "step": 880
    },
    {
      "epoch": 0.028327251776454773,
      "grad_norm": 26.25,
      "learning_rate": 4.9939769032110864e-05,
      "loss": 4.394,
      "num_input_tokens_seen": 1199088,
      "step": 885
    },
    {
      "epoch": 0.028487292746943216,
      "grad_norm": 22.125,
      "learning_rate": 4.993908603704423e-05,
      "loss": 4.2519,
      "num_input_tokens_seen": 1205520,
      "step": 890
    },
    {
      "epoch": 0.028647333717431663,
      "grad_norm": 23.0,
      "learning_rate": 4.9938399196057126e-05,
      "loss": 4.0645,
      "num_input_tokens_seen": 1212288,
      "step": 895
    },
    {
      "epoch": 0.028807374687920106,
      "grad_norm": 18.75,
      "learning_rate": 4.993770850925547e-05,
      "loss": 3.9336,
      "num_input_tokens_seen": 1219104,
      "step": 900
    },
    {
      "epoch": 0.028967415658408553,
      "grad_norm": 25.375,
      "learning_rate": 4.993701397674577e-05,
      "loss": 5.1141,
      "num_input_tokens_seen": 1225728,
      "step": 905
    },
    {
      "epoch": 0.029127456628896996,
      "grad_norm": 20.0,
      "learning_rate": 4.993631559863515e-05,
      "loss": 3.6785,
      "num_input_tokens_seen": 1232592,
      "step": 910
    },
    {
      "epoch": 0.029287497599385443,
      "grad_norm": 14.9375,
      "learning_rate": 4.9935613375031283e-05,
      "loss": 4.2343,
      "num_input_tokens_seen": 1239056,
      "step": 915
    },
    {
      "epoch": 0.029447538569873886,
      "grad_norm": 38.5,
      "learning_rate": 4.993490730604248e-05,
      "loss": 4.9118,
      "num_input_tokens_seen": 1245776,
      "step": 920
    },
    {
      "epoch": 0.029607579540362333,
      "grad_norm": 24.875,
      "learning_rate": 4.993419739177761e-05,
      "loss": 4.1061,
      "num_input_tokens_seen": 1252896,
      "step": 925
    },
    {
      "epoch": 0.029767620510850776,
      "grad_norm": 18.25,
      "learning_rate": 4.9933483632346164e-05,
      "loss": 4.3388,
      "num_input_tokens_seen": 1259840,
      "step": 930
    },
    {
      "epoch": 0.029927661481339223,
      "grad_norm": 22.5,
      "learning_rate": 4.993276602785821e-05,
      "loss": 3.974,
      "num_input_tokens_seen": 1266336,
      "step": 935
    },
    {
      "epoch": 0.030087702451827666,
      "grad_norm": 43.25,
      "learning_rate": 4.993204457842441e-05,
      "loss": 4.3133,
      "num_input_tokens_seen": 1272720,
      "step": 940
    },
    {
      "epoch": 0.030247743422316113,
      "grad_norm": 34.5,
      "learning_rate": 4.993131928415602e-05,
      "loss": 5.4432,
      "num_input_tokens_seen": 1279072,
      "step": 945
    },
    {
      "epoch": 0.030407784392804556,
      "grad_norm": 29.125,
      "learning_rate": 4.993059014516489e-05,
      "loss": 3.759,
      "num_input_tokens_seen": 1285712,
      "step": 950
    },
    {
      "epoch": 0.030567825363293003,
      "grad_norm": 60.25,
      "learning_rate": 4.9929857161563464e-05,
      "loss": 6.0525,
      "num_input_tokens_seen": 1293488,
      "step": 955
    },
    {
      "epoch": 0.030727866333781446,
      "grad_norm": 18.375,
      "learning_rate": 4.992912033346477e-05,
      "loss": 4.7683,
      "num_input_tokens_seen": 1300432,
      "step": 960
    },
    {
      "epoch": 0.030887907304269893,
      "grad_norm": 26.75,
      "learning_rate": 4.992837966098245e-05,
      "loss": 4.0814,
      "num_input_tokens_seen": 1307568,
      "step": 965
    },
    {
      "epoch": 0.031047948274758336,
      "grad_norm": 23.75,
      "learning_rate": 4.992763514423071e-05,
      "loss": 4.5507,
      "num_input_tokens_seen": 1314464,
      "step": 970
    },
    {
      "epoch": 0.031207989245246783,
      "grad_norm": 34.75,
      "learning_rate": 4.992688678332437e-05,
      "loss": 4.8111,
      "num_input_tokens_seen": 1321120,
      "step": 975
    },
    {
      "epoch": 0.031368030215735226,
      "grad_norm": 33.0,
      "learning_rate": 4.992613457837884e-05,
      "loss": 4.8879,
      "num_input_tokens_seen": 1327680,
      "step": 980
    },
    {
      "epoch": 0.03152807118622367,
      "grad_norm": 23.125,
      "learning_rate": 4.992537852951011e-05,
      "loss": 4.5731,
      "num_input_tokens_seen": 1334576,
      "step": 985
    },
    {
      "epoch": 0.03168811215671212,
      "grad_norm": 25.375,
      "learning_rate": 4.9924618636834785e-05,
      "loss": 4.173,
      "num_input_tokens_seen": 1341824,
      "step": 990
    },
    {
      "epoch": 0.03184815312720057,
      "grad_norm": 29.875,
      "learning_rate": 4.9923854900470046e-05,
      "loss": 5.3346,
      "num_input_tokens_seen": 1348512,
      "step": 995
    },
    {
      "epoch": 0.032008194097689006,
      "grad_norm": 32.0,
      "learning_rate": 4.992308732053367e-05,
      "loss": 3.707,
      "num_input_tokens_seen": 1355472,
      "step": 1000
    },
    {
      "epoch": 0.032008194097689006,
      "eval_loss": 4.54565954208374,
      "eval_runtime": 488.029,
      "eval_samples_per_second": 28.453,
      "eval_steps_per_second": 14.227,
      "num_input_tokens_seen": 1355472,
      "step": 1000
    },
    {
      "epoch": 0.03216823506817745,
      "grad_norm": 22.75,
      "learning_rate": 4.992231589714402e-05,
      "loss": 5.0739,
      "num_input_tokens_seen": 1362192,
      "step": 1005
    },
    {
      "epoch": 0.0323282760386659,
      "grad_norm": 45.75,
      "learning_rate": 4.992154063042007e-05,
      "loss": 4.7601,
      "num_input_tokens_seen": 1369040,
      "step": 1010
    },
    {
      "epoch": 0.03248831700915435,
      "grad_norm": 20.375,
      "learning_rate": 4.992076152048136e-05,
      "loss": 4.156,
      "num_input_tokens_seen": 1375616,
      "step": 1015
    },
    {
      "epoch": 0.032648357979642786,
      "grad_norm": 32.25,
      "learning_rate": 4.991997856744807e-05,
      "loss": 5.2873,
      "num_input_tokens_seen": 1382240,
      "step": 1020
    },
    {
      "epoch": 0.03280839895013123,
      "grad_norm": 26.125,
      "learning_rate": 4.9919191771440905e-05,
      "loss": 4.4496,
      "num_input_tokens_seen": 1388752,
      "step": 1025
    },
    {
      "epoch": 0.03296843992061968,
      "grad_norm": 29.0,
      "learning_rate": 4.991840113258122e-05,
      "loss": 4.5485,
      "num_input_tokens_seen": 1395648,
      "step": 1030
    },
    {
      "epoch": 0.03312848089110813,
      "grad_norm": 28.25,
      "learning_rate": 4.9917606650990933e-05,
      "loss": 4.1785,
      "num_input_tokens_seen": 1402480,
      "step": 1035
    },
    {
      "epoch": 0.033288521861596566,
      "grad_norm": 24.75,
      "learning_rate": 4.9916808326792566e-05,
      "loss": 4.684,
      "num_input_tokens_seen": 1409552,
      "step": 1040
    },
    {
      "epoch": 0.03344856283208501,
      "grad_norm": 32.5,
      "learning_rate": 4.9916006160109235e-05,
      "loss": 5.0028,
      "num_input_tokens_seen": 1416160,
      "step": 1045
    },
    {
      "epoch": 0.03360860380257346,
      "grad_norm": 21.625,
      "learning_rate": 4.991520015106464e-05,
      "loss": 4.6189,
      "num_input_tokens_seen": 1423072,
      "step": 1050
    },
    {
      "epoch": 0.03376864477306191,
      "grad_norm": 37.25,
      "learning_rate": 4.991439029978308e-05,
      "loss": 4.699,
      "num_input_tokens_seen": 1429680,
      "step": 1055
    },
    {
      "epoch": 0.033928685743550346,
      "grad_norm": 25.375,
      "learning_rate": 4.9913576606389434e-05,
      "loss": 4.1606,
      "num_input_tokens_seen": 1436096,
      "step": 1060
    },
    {
      "epoch": 0.03408872671403879,
      "grad_norm": 40.0,
      "learning_rate": 4.991275907100919e-05,
      "loss": 4.7423,
      "num_input_tokens_seen": 1442704,
      "step": 1065
    },
    {
      "epoch": 0.03424876768452724,
      "grad_norm": 27.25,
      "learning_rate": 4.9911937693768434e-05,
      "loss": 4.3887,
      "num_input_tokens_seen": 1449472,
      "step": 1070
    },
    {
      "epoch": 0.03440880865501569,
      "grad_norm": 48.5,
      "learning_rate": 4.991111247479382e-05,
      "loss": 4.7425,
      "num_input_tokens_seen": 1456560,
      "step": 1075
    },
    {
      "epoch": 0.034568849625504126,
      "grad_norm": 27.375,
      "learning_rate": 4.9910283414212605e-05,
      "loss": 3.829,
      "num_input_tokens_seen": 1463232,
      "step": 1080
    },
    {
      "epoch": 0.03472889059599257,
      "grad_norm": 30.0,
      "learning_rate": 4.990945051215265e-05,
      "loss": 4.7076,
      "num_input_tokens_seen": 1469904,
      "step": 1085
    },
    {
      "epoch": 0.03488893156648102,
      "grad_norm": 30.75,
      "learning_rate": 4.99086137687424e-05,
      "loss": 4.2374,
      "num_input_tokens_seen": 1476464,
      "step": 1090
    },
    {
      "epoch": 0.03504897253696947,
      "grad_norm": 23.625,
      "learning_rate": 4.9907773184110874e-05,
      "loss": 4.0153,
      "num_input_tokens_seen": 1483248,
      "step": 1095
    },
    {
      "epoch": 0.035209013507457906,
      "grad_norm": 33.25,
      "learning_rate": 4.9906928758387715e-05,
      "loss": 3.6822,
      "num_input_tokens_seen": 1490464,
      "step": 1100
    },
    {
      "epoch": 0.03536905447794635,
      "grad_norm": 24.875,
      "learning_rate": 4.9906080491703146e-05,
      "loss": 4.8102,
      "num_input_tokens_seen": 1497328,
      "step": 1105
    },
    {
      "epoch": 0.0355290954484348,
      "grad_norm": 35.25,
      "learning_rate": 4.990522838418797e-05,
      "loss": 4.4738,
      "num_input_tokens_seen": 1503808,
      "step": 1110
    },
    {
      "epoch": 0.03568913641892325,
      "grad_norm": 34.75,
      "learning_rate": 4.9904372435973604e-05,
      "loss": 4.2679,
      "num_input_tokens_seen": 1510336,
      "step": 1115
    },
    {
      "epoch": 0.035849177389411686,
      "grad_norm": 30.125,
      "learning_rate": 4.990351264719203e-05,
      "loss": 4.2457,
      "num_input_tokens_seen": 1517392,
      "step": 1120
    },
    {
      "epoch": 0.03600921835990013,
      "grad_norm": 27.375,
      "learning_rate": 4.990264901797586e-05,
      "loss": 3.4367,
      "num_input_tokens_seen": 1524032,
      "step": 1125
    },
    {
      "epoch": 0.03616925933038858,
      "grad_norm": 26.75,
      "learning_rate": 4.990178154845826e-05,
      "loss": 4.3477,
      "num_input_tokens_seen": 1530944,
      "step": 1130
    },
    {
      "epoch": 0.03632930030087703,
      "grad_norm": 23.25,
      "learning_rate": 4.9900910238773014e-05,
      "loss": 4.5997,
      "num_input_tokens_seen": 1537328,
      "step": 1135
    },
    {
      "epoch": 0.036489341271365466,
      "grad_norm": 28.625,
      "learning_rate": 4.990003508905448e-05,
      "loss": 4.9302,
      "num_input_tokens_seen": 1544112,
      "step": 1140
    },
    {
      "epoch": 0.03664938224185391,
      "grad_norm": 18.25,
      "learning_rate": 4.989915609943763e-05,
      "loss": 4.1591,
      "num_input_tokens_seen": 1550640,
      "step": 1145
    },
    {
      "epoch": 0.03680942321234236,
      "grad_norm": 33.5,
      "learning_rate": 4.9898273270058e-05,
      "loss": 4.2781,
      "num_input_tokens_seen": 1557440,
      "step": 1150
    },
    {
      "epoch": 0.03696946418283081,
      "grad_norm": 28.0,
      "learning_rate": 4.989738660105174e-05,
      "loss": 4.562,
      "num_input_tokens_seen": 1563920,
      "step": 1155
    },
    {
      "epoch": 0.037129505153319246,
      "grad_norm": 21.0,
      "learning_rate": 4.989649609255559e-05,
      "loss": 4.5061,
      "num_input_tokens_seen": 1570800,
      "step": 1160
    },
    {
      "epoch": 0.03728954612380769,
      "grad_norm": 36.5,
      "learning_rate": 4.989560174470687e-05,
      "loss": 3.569,
      "num_input_tokens_seen": 1577408,
      "step": 1165
    },
    {
      "epoch": 0.03744958709429614,
      "grad_norm": 46.5,
      "learning_rate": 4.989470355764351e-05,
      "loss": 4.2957,
      "num_input_tokens_seen": 1583792,
      "step": 1170
    },
    {
      "epoch": 0.03760962806478459,
      "grad_norm": 19.125,
      "learning_rate": 4.9893801531504e-05,
      "loss": 4.3366,
      "num_input_tokens_seen": 1590432,
      "step": 1175
    },
    {
      "epoch": 0.03776966903527303,
      "grad_norm": 21.0,
      "learning_rate": 4.9892895666427475e-05,
      "loss": 4.1318,
      "num_input_tokens_seen": 1597248,
      "step": 1180
    },
    {
      "epoch": 0.03792971000576147,
      "grad_norm": 36.0,
      "learning_rate": 4.9891985962553606e-05,
      "loss": 4.6394,
      "num_input_tokens_seen": 1603952,
      "step": 1185
    },
    {
      "epoch": 0.03808975097624992,
      "grad_norm": 22.375,
      "learning_rate": 4.989107242002269e-05,
      "loss": 3.4509,
      "num_input_tokens_seen": 1610672,
      "step": 1190
    },
    {
      "epoch": 0.038249791946738367,
      "grad_norm": 37.75,
      "learning_rate": 4.989015503897561e-05,
      "loss": 4.329,
      "num_input_tokens_seen": 1617520,
      "step": 1195
    },
    {
      "epoch": 0.03840983291722681,
      "grad_norm": 20.0,
      "learning_rate": 4.988923381955383e-05,
      "loss": 4.4764,
      "num_input_tokens_seen": 1624048,
      "step": 1200
    },
    {
      "epoch": 0.03840983291722681,
      "eval_loss": 4.514975070953369,
      "eval_runtime": 487.4267,
      "eval_samples_per_second": 28.488,
      "eval_steps_per_second": 14.244,
      "num_input_tokens_seen": 1624048,
      "step": 1200
    },
    {
      "epoch": 0.03856987388771525,
      "grad_norm": 33.5,
      "learning_rate": 4.988830876189942e-05,
      "loss": 4.4731,
      "num_input_tokens_seen": 1630848,
      "step": 1205
    },
    {
      "epoch": 0.0387299148582037,
      "grad_norm": 34.0,
      "learning_rate": 4.988737986615503e-05,
      "loss": 4.8188,
      "num_input_tokens_seen": 1637504,
      "step": 1210
    },
    {
      "epoch": 0.038889955828692147,
      "grad_norm": 41.75,
      "learning_rate": 4.988644713246391e-05,
      "loss": 3.9107,
      "num_input_tokens_seen": 1644272,
      "step": 1215
    },
    {
      "epoch": 0.03904999679918059,
      "grad_norm": 17.625,
      "learning_rate": 4.988551056096991e-05,
      "loss": 4.6399,
      "num_input_tokens_seen": 1650992,
      "step": 1220
    },
    {
      "epoch": 0.03921003776966903,
      "grad_norm": 46.25,
      "learning_rate": 4.988457015181743e-05,
      "loss": 4.1269,
      "num_input_tokens_seen": 1657840,
      "step": 1225
    },
    {
      "epoch": 0.03937007874015748,
      "grad_norm": 21.75,
      "learning_rate": 4.988362590515153e-05,
      "loss": 3.9249,
      "num_input_tokens_seen": 1664608,
      "step": 1230
    },
    {
      "epoch": 0.039530119710645927,
      "grad_norm": 23.875,
      "learning_rate": 4.9882677821117805e-05,
      "loss": 4.4155,
      "num_input_tokens_seen": 1671568,
      "step": 1235
    },
    {
      "epoch": 0.03969016068113437,
      "grad_norm": 31.25,
      "learning_rate": 4.988172589986246e-05,
      "loss": 5.3577,
      "num_input_tokens_seen": 1678048,
      "step": 1240
    },
    {
      "epoch": 0.03985020165162281,
      "grad_norm": 32.5,
      "learning_rate": 4.9880770141532304e-05,
      "loss": 4.1314,
      "num_input_tokens_seen": 1684688,
      "step": 1245
    },
    {
      "epoch": 0.04001024262211126,
      "grad_norm": 29.625,
      "learning_rate": 4.987981054627472e-05,
      "loss": 5.2751,
      "num_input_tokens_seen": 1691360,
      "step": 1250
    },
    {
      "epoch": 0.040170283592599706,
      "grad_norm": 25.875,
      "learning_rate": 4.987884711423769e-05,
      "loss": 4.3146,
      "num_input_tokens_seen": 1698064,
      "step": 1255
    },
    {
      "epoch": 0.04033032456308815,
      "grad_norm": 47.5,
      "learning_rate": 4.9877879845569784e-05,
      "loss": 3.4933,
      "num_input_tokens_seen": 1705040,
      "step": 1260
    },
    {
      "epoch": 0.04049036553357659,
      "grad_norm": 32.75,
      "learning_rate": 4.9876908740420175e-05,
      "loss": 3.456,
      "num_input_tokens_seen": 1711952,
      "step": 1265
    },
    {
      "epoch": 0.04065040650406504,
      "grad_norm": 25.125,
      "learning_rate": 4.987593379893861e-05,
      "loss": 3.1882,
      "num_input_tokens_seen": 1718480,
      "step": 1270
    },
    {
      "epoch": 0.040810447474553486,
      "grad_norm": 17.875,
      "learning_rate": 4.987495502127545e-05,
      "loss": 4.5449,
      "num_input_tokens_seen": 1725264,
      "step": 1275
    },
    {
      "epoch": 0.04097048844504193,
      "grad_norm": 37.75,
      "learning_rate": 4.987397240758162e-05,
      "loss": 4.1276,
      "num_input_tokens_seen": 1731792,
      "step": 1280
    },
    {
      "epoch": 0.04113052941553037,
      "grad_norm": 36.25,
      "learning_rate": 4.9872985958008664e-05,
      "loss": 4.4167,
      "num_input_tokens_seen": 1738880,
      "step": 1285
    },
    {
      "epoch": 0.04129057038601882,
      "grad_norm": 23.75,
      "learning_rate": 4.987199567270871e-05,
      "loss": 3.9588,
      "num_input_tokens_seen": 1745712,
      "step": 1290
    },
    {
      "epoch": 0.041450611356507266,
      "grad_norm": 25.25,
      "learning_rate": 4.9871001551834444e-05,
      "loss": 3.7285,
      "num_input_tokens_seen": 1752400,
      "step": 1295
    },
    {
      "epoch": 0.04161065232699571,
      "grad_norm": 40.5,
      "learning_rate": 4.98700035955392e-05,
      "loss": 4.8406,
      "num_input_tokens_seen": 1758960,
      "step": 1300
    },
    {
      "epoch": 0.04177069329748415,
      "grad_norm": 31.5,
      "learning_rate": 4.986900180397686e-05,
      "loss": 5.44,
      "num_input_tokens_seen": 1765840,
      "step": 1305
    },
    {
      "epoch": 0.0419307342679726,
      "grad_norm": 11.125,
      "learning_rate": 4.9867996177301926e-05,
      "loss": 3.9759,
      "num_input_tokens_seen": 1772208,
      "step": 1310
    },
    {
      "epoch": 0.042090775238461046,
      "grad_norm": 30.125,
      "learning_rate": 4.9866986715669464e-05,
      "loss": 4.7627,
      "num_input_tokens_seen": 1778976,
      "step": 1315
    },
    {
      "epoch": 0.04225081620894949,
      "grad_norm": 15.0625,
      "learning_rate": 4.9865973419235155e-05,
      "loss": 3.5789,
      "num_input_tokens_seen": 1786400,
      "step": 1320
    },
    {
      "epoch": 0.04241085717943793,
      "grad_norm": 40.75,
      "learning_rate": 4.986495628815526e-05,
      "loss": 4.82,
      "num_input_tokens_seen": 1793200,
      "step": 1325
    },
    {
      "epoch": 0.04257089814992638,
      "grad_norm": 21.75,
      "learning_rate": 4.986393532258663e-05,
      "loss": 4.0956,
      "num_input_tokens_seen": 1799824,
      "step": 1330
    },
    {
      "epoch": 0.042730939120414826,
      "grad_norm": 15.0,
      "learning_rate": 4.986291052268671e-05,
      "loss": 4.1851,
      "num_input_tokens_seen": 1806880,
      "step": 1335
    },
    {
      "epoch": 0.04289098009090327,
      "grad_norm": 27.125,
      "learning_rate": 4.986188188861355e-05,
      "loss": 4.5034,
      "num_input_tokens_seen": 1813328,
      "step": 1340
    },
    {
      "epoch": 0.04305102106139171,
      "grad_norm": 22.5,
      "learning_rate": 4.9860849420525766e-05,
      "loss": 4.6454,
      "num_input_tokens_seen": 1819904,
      "step": 1345
    },
    {
      "epoch": 0.04321106203188016,
      "grad_norm": 50.5,
      "learning_rate": 4.9859813118582575e-05,
      "loss": 4.4188,
      "num_input_tokens_seen": 1826896,
      "step": 1350
    },
    {
      "epoch": 0.043371103002368606,
      "grad_norm": 30.125,
      "learning_rate": 4.98587729829438e-05,
      "loss": 3.9147,
      "num_input_tokens_seen": 1833520,
      "step": 1355
    },
    {
      "epoch": 0.04353114397285705,
      "grad_norm": 34.75,
      "learning_rate": 4.985772901376983e-05,
      "loss": 4.207,
      "num_input_tokens_seen": 1840032,
      "step": 1360
    },
    {
      "epoch": 0.0436911849433455,
      "grad_norm": 28.0,
      "learning_rate": 4.9856681211221666e-05,
      "loss": 5.6264,
      "num_input_tokens_seen": 1846464,
      "step": 1365
    },
    {
      "epoch": 0.04385122591383394,
      "grad_norm": 37.25,
      "learning_rate": 4.985562957546089e-05,
      "loss": 4.6332,
      "num_input_tokens_seen": 1853296,
      "step": 1370
    },
    {
      "epoch": 0.044011266884322386,
      "grad_norm": 40.25,
      "learning_rate": 4.9854574106649686e-05,
      "loss": 4.8578,
      "num_input_tokens_seen": 1859904,
      "step": 1375
    },
    {
      "epoch": 0.04417130785481083,
      "grad_norm": 45.75,
      "learning_rate": 4.985351480495081e-05,
      "loss": 4.2502,
      "num_input_tokens_seen": 1867024,
      "step": 1380
    },
    {
      "epoch": 0.04433134882529928,
      "grad_norm": 22.625,
      "learning_rate": 4.985245167052762e-05,
      "loss": 3.7949,
      "num_input_tokens_seen": 1873664,
      "step": 1385
    },
    {
      "epoch": 0.04449138979578772,
      "grad_norm": 36.5,
      "learning_rate": 4.9851384703544066e-05,
      "loss": 4.6745,
      "num_input_tokens_seen": 1880320,
      "step": 1390
    },
    {
      "epoch": 0.044651430766276166,
      "grad_norm": 50.0,
      "learning_rate": 4.985031390416469e-05,
      "loss": 4.374,
      "num_input_tokens_seen": 1887168,
      "step": 1395
    },
    {
      "epoch": 0.04481147173676461,
      "grad_norm": 41.25,
      "learning_rate": 4.984923927255461e-05,
      "loss": 4.6147,
      "num_input_tokens_seen": 1893968,
      "step": 1400
    },
    {
      "epoch": 0.04481147173676461,
      "eval_loss": 4.516072750091553,
      "eval_runtime": 487.4716,
      "eval_samples_per_second": 28.486,
      "eval_steps_per_second": 14.243,
      "num_input_tokens_seen": 1893968,
      "step": 1400
    },
    {
      "epoch": 0.04497151270725306,
      "grad_norm": 29.625,
      "learning_rate": 4.984816080887958e-05,
      "loss": 4.4118,
      "num_input_tokens_seen": 1900480,
      "step": 1405
    },
    {
      "epoch": 0.0451315536777415,
      "grad_norm": 21.625,
      "learning_rate": 4.9847078513305875e-05,
      "loss": 4.5218,
      "num_input_tokens_seen": 1907296,
      "step": 1410
    },
    {
      "epoch": 0.045291594648229946,
      "grad_norm": 20.625,
      "learning_rate": 4.984599238600043e-05,
      "loss": 3.7753,
      "num_input_tokens_seen": 1913744,
      "step": 1415
    },
    {
      "epoch": 0.04545163561871839,
      "grad_norm": 22.5,
      "learning_rate": 4.9844902427130716e-05,
      "loss": 3.3907,
      "num_input_tokens_seen": 1920304,
      "step": 1420
    },
    {
      "epoch": 0.04561167658920684,
      "grad_norm": 42.5,
      "learning_rate": 4.984380863686482e-05,
      "loss": 4.1633,
      "num_input_tokens_seen": 1926736,
      "step": 1425
    },
    {
      "epoch": 0.04577171755969528,
      "grad_norm": 31.5,
      "learning_rate": 4.984271101537143e-05,
      "loss": 3.8036,
      "num_input_tokens_seen": 1933792,
      "step": 1430
    },
    {
      "epoch": 0.045931758530183726,
      "grad_norm": 27.125,
      "learning_rate": 4.9841609562819816e-05,
      "loss": 4.5617,
      "num_input_tokens_seen": 1940176,
      "step": 1435
    },
    {
      "epoch": 0.04609179950067217,
      "grad_norm": 59.75,
      "learning_rate": 4.984050427937983e-05,
      "loss": 5.0865,
      "num_input_tokens_seen": 1946784,
      "step": 1440
    },
    {
      "epoch": 0.04625184047116062,
      "grad_norm": 52.0,
      "learning_rate": 4.983939516522191e-05,
      "loss": 3.5381,
      "num_input_tokens_seen": 1953792,
      "step": 1445
    },
    {
      "epoch": 0.04641188144164906,
      "grad_norm": 23.25,
      "learning_rate": 4.983828222051711e-05,
      "loss": 3.9802,
      "num_input_tokens_seen": 1960704,
      "step": 1450
    },
    {
      "epoch": 0.046571922412137506,
      "grad_norm": 27.25,
      "learning_rate": 4.983716544543705e-05,
      "loss": 3.9115,
      "num_input_tokens_seen": 1967280,
      "step": 1455
    },
    {
      "epoch": 0.04673196338262595,
      "grad_norm": 40.0,
      "learning_rate": 4.983604484015395e-05,
      "loss": 4.1863,
      "num_input_tokens_seen": 1973936,
      "step": 1460
    },
    {
      "epoch": 0.0468920043531144,
      "grad_norm": 16.375,
      "learning_rate": 4.983492040484064e-05,
      "loss": 4.5885,
      "num_input_tokens_seen": 1980624,
      "step": 1465
    },
    {
      "epoch": 0.04705204532360284,
      "grad_norm": 19.125,
      "learning_rate": 4.98337921396705e-05,
      "loss": 3.633,
      "num_input_tokens_seen": 1987056,
      "step": 1470
    },
    {
      "epoch": 0.047212086294091286,
      "grad_norm": 22.0,
      "learning_rate": 4.983266004481753e-05,
      "loss": 4.7012,
      "num_input_tokens_seen": 1993888,
      "step": 1475
    },
    {
      "epoch": 0.04737212726457973,
      "grad_norm": 43.0,
      "learning_rate": 4.9831524120456316e-05,
      "loss": 4.6905,
      "num_input_tokens_seen": 2001008,
      "step": 1480
    },
    {
      "epoch": 0.04753216823506818,
      "grad_norm": 33.5,
      "learning_rate": 4.9830384366762026e-05,
      "loss": 4.5459,
      "num_input_tokens_seen": 2007312,
      "step": 1485
    },
    {
      "epoch": 0.04769220920555662,
      "grad_norm": 27.625,
      "learning_rate": 4.9829240783910436e-05,
      "loss": 4.7483,
      "num_input_tokens_seen": 2013904,
      "step": 1490
    },
    {
      "epoch": 0.047852250176045066,
      "grad_norm": 48.5,
      "learning_rate": 4.982809337207789e-05,
      "loss": 5.1669,
      "num_input_tokens_seen": 2020384,
      "step": 1495
    },
    {
      "epoch": 0.04801229114653351,
      "grad_norm": 26.0,
      "learning_rate": 4.9826942131441337e-05,
      "loss": 4.5644,
      "num_input_tokens_seen": 2027104,
      "step": 1500
    },
    {
      "epoch": 0.04817233211702196,
      "grad_norm": 27.5,
      "learning_rate": 4.9825787062178315e-05,
      "loss": 3.9383,
      "num_input_tokens_seen": 2034240,
      "step": 1505
    },
    {
      "epoch": 0.0483323730875104,
      "grad_norm": 49.0,
      "learning_rate": 4.9824628164466945e-05,
      "loss": 4.6437,
      "num_input_tokens_seen": 2040672,
      "step": 1510
    },
    {
      "epoch": 0.048492414057998846,
      "grad_norm": 60.0,
      "learning_rate": 4.982346543848595e-05,
      "loss": 4.6784,
      "num_input_tokens_seen": 2047136,
      "step": 1515
    },
    {
      "epoch": 0.04865245502848729,
      "grad_norm": 38.0,
      "learning_rate": 4.9822298884414626e-05,
      "loss": 4.1881,
      "num_input_tokens_seen": 2053520,
      "step": 1520
    },
    {
      "epoch": 0.04881249599897574,
      "grad_norm": 24.625,
      "learning_rate": 4.982112850243288e-05,
      "loss": 4.1183,
      "num_input_tokens_seen": 2060896,
      "step": 1525
    },
    {
      "epoch": 0.04897253696946418,
      "grad_norm": 38.0,
      "learning_rate": 4.98199542927212e-05,
      "loss": 5.8589,
      "num_input_tokens_seen": 2067136,
      "step": 1530
    },
    {
      "epoch": 0.049132577939952626,
      "grad_norm": 20.25,
      "learning_rate": 4.981877625546066e-05,
      "loss": 4.4336,
      "num_input_tokens_seen": 2074048,
      "step": 1535
    },
    {
      "epoch": 0.04929261891044107,
      "grad_norm": 30.125,
      "learning_rate": 4.981759439083293e-05,
      "loss": 5.7292,
      "num_input_tokens_seen": 2080416,
      "step": 1540
    },
    {
      "epoch": 0.04945265988092952,
      "grad_norm": 34.25,
      "learning_rate": 4.981640869902027e-05,
      "loss": 4.4475,
      "num_input_tokens_seen": 2087072,
      "step": 1545
    },
    {
      "epoch": 0.04961270085141796,
      "grad_norm": 24.0,
      "learning_rate": 4.9815219180205517e-05,
      "loss": 3.4111,
      "num_input_tokens_seen": 2093872,
      "step": 1550
    },
    {
      "epoch": 0.049772741821906406,
      "grad_norm": 15.6875,
      "learning_rate": 4.9814025834572126e-05,
      "loss": 3.8696,
      "num_input_tokens_seen": 2100784,
      "step": 1555
    },
    {
      "epoch": 0.04993278279239485,
      "grad_norm": 17.25,
      "learning_rate": 4.981282866230411e-05,
      "loss": 3.8305,
      "num_input_tokens_seen": 2107488,
      "step": 1560
    },
    {
      "epoch": 0.0500928237628833,
      "grad_norm": 30.75,
      "learning_rate": 4.981162766358611e-05,
      "loss": 4.4382,
      "num_input_tokens_seen": 2114224,
      "step": 1565
    },
    {
      "epoch": 0.05025286473337175,
      "grad_norm": 45.75,
      "learning_rate": 4.9810422838603316e-05,
      "loss": 4.3934,
      "num_input_tokens_seen": 2120880,
      "step": 1570
    },
    {
      "epoch": 0.050412905703860186,
      "grad_norm": 28.625,
      "learning_rate": 4.9809214187541533e-05,
      "loss": 5.1501,
      "num_input_tokens_seen": 2128112,
      "step": 1575
    },
    {
      "epoch": 0.05057294667434863,
      "grad_norm": 35.75,
      "learning_rate": 4.980800171058715e-05,
      "loss": 4.2702,
      "num_input_tokens_seen": 2134928,
      "step": 1580
    },
    {
      "epoch": 0.05073298764483708,
      "grad_norm": 19.125,
      "learning_rate": 4.980678540792715e-05,
      "loss": 4.8613,
      "num_input_tokens_seen": 2141808,
      "step": 1585
    },
    {
      "epoch": 0.050893028615325527,
      "grad_norm": 14.6875,
      "learning_rate": 4.980556527974909e-05,
      "loss": 4.4869,
      "num_input_tokens_seen": 2149376,
      "step": 1590
    },
    {
      "epoch": 0.051053069585813966,
      "grad_norm": 26.0,
      "learning_rate": 4.980434132624114e-05,
      "loss": 5.0712,
      "num_input_tokens_seen": 2156352,
      "step": 1595
    },
    {
      "epoch": 0.05121311055630241,
      "grad_norm": 13.0,
      "learning_rate": 4.980311354759205e-05,
      "loss": 3.7767,
      "num_input_tokens_seen": 2163024,
      "step": 1600
    },
    {
      "epoch": 0.05121311055630241,
      "eval_loss": 4.505253791809082,
      "eval_runtime": 487.7227,
      "eval_samples_per_second": 28.471,
      "eval_steps_per_second": 14.236,
      "num_input_tokens_seen": 2163024,
      "step": 1600
    },
    {
      "epoch": 0.05137315152679086,
      "grad_norm": 29.25,
      "learning_rate": 4.980188194399116e-05,
      "loss": 4.3623,
      "num_input_tokens_seen": 2169744,
      "step": 1605
    },
    {
      "epoch": 0.051533192497279307,
      "grad_norm": 20.625,
      "learning_rate": 4.9800646515628384e-05,
      "loss": 4.6435,
      "num_input_tokens_seen": 2176896,
      "step": 1610
    },
    {
      "epoch": 0.051693233467767746,
      "grad_norm": 20.75,
      "learning_rate": 4.979940726269426e-05,
      "loss": 4.3078,
      "num_input_tokens_seen": 2183696,
      "step": 1615
    },
    {
      "epoch": 0.05185327443825619,
      "grad_norm": 28.875,
      "learning_rate": 4.979816418537988e-05,
      "loss": 4.0691,
      "num_input_tokens_seen": 2190400,
      "step": 1620
    },
    {
      "epoch": 0.05201331540874464,
      "grad_norm": 25.375,
      "learning_rate": 4.979691728387696e-05,
      "loss": 3.8661,
      "num_input_tokens_seen": 2197088,
      "step": 1625
    },
    {
      "epoch": 0.052173356379233086,
      "grad_norm": 21.5,
      "learning_rate": 4.979566655837776e-05,
      "loss": 3.8088,
      "num_input_tokens_seen": 2204064,
      "step": 1630
    },
    {
      "epoch": 0.052333397349721526,
      "grad_norm": 31.125,
      "learning_rate": 4.9794412009075184e-05,
      "loss": 4.5676,
      "num_input_tokens_seen": 2210704,
      "step": 1635
    },
    {
      "epoch": 0.05249343832020997,
      "grad_norm": 20.375,
      "learning_rate": 4.979315363616269e-05,
      "loss": 4.4096,
      "num_input_tokens_seen": 2217536,
      "step": 1640
    },
    {
      "epoch": 0.05265347929069842,
      "grad_norm": 23.125,
      "learning_rate": 4.979189143983434e-05,
      "loss": 4.1958,
      "num_input_tokens_seen": 2224080,
      "step": 1645
    },
    {
      "epoch": 0.052813520261186866,
      "grad_norm": 30.375,
      "learning_rate": 4.979062542028478e-05,
      "loss": 3.9808,
      "num_input_tokens_seen": 2230848,
      "step": 1650
    },
    {
      "epoch": 0.052973561231675306,
      "grad_norm": 21.875,
      "learning_rate": 4.978935557770923e-05,
      "loss": 4.629,
      "num_input_tokens_seen": 2237888,
      "step": 1655
    },
    {
      "epoch": 0.05313360220216375,
      "grad_norm": 15.5,
      "learning_rate": 4.978808191230353e-05,
      "loss": 3.9863,
      "num_input_tokens_seen": 2244640,
      "step": 1660
    },
    {
      "epoch": 0.0532936431726522,
      "grad_norm": 20.625,
      "learning_rate": 4.9786804424264085e-05,
      "loss": 4.1341,
      "num_input_tokens_seen": 2251760,
      "step": 1665
    },
    {
      "epoch": 0.053453684143140646,
      "grad_norm": 25.125,
      "learning_rate": 4.978552311378792e-05,
      "loss": 3.9331,
      "num_input_tokens_seen": 2258208,
      "step": 1670
    },
    {
      "epoch": 0.053613725113629086,
      "grad_norm": 32.75,
      "learning_rate": 4.978423798107261e-05,
      "loss": 3.8847,
      "num_input_tokens_seen": 2265088,
      "step": 1675
    },
    {
      "epoch": 0.05377376608411753,
      "grad_norm": 33.75,
      "learning_rate": 4.978294902631635e-05,
      "loss": 4.0751,
      "num_input_tokens_seen": 2271888,
      "step": 1680
    },
    {
      "epoch": 0.05393380705460598,
      "grad_norm": 38.0,
      "learning_rate": 4.9781656249717914e-05,
      "loss": 4.9875,
      "num_input_tokens_seen": 2278976,
      "step": 1685
    },
    {
      "epoch": 0.054093848025094426,
      "grad_norm": 20.875,
      "learning_rate": 4.9780359651476645e-05,
      "loss": 4.0462,
      "num_input_tokens_seen": 2285520,
      "step": 1690
    },
    {
      "epoch": 0.054253888995582866,
      "grad_norm": 18.75,
      "learning_rate": 4.977905923179251e-05,
      "loss": 3.809,
      "num_input_tokens_seen": 2292352,
      "step": 1695
    },
    {
      "epoch": 0.05441392996607131,
      "grad_norm": 21.25,
      "learning_rate": 4.977775499086606e-05,
      "loss": 4.5572,
      "num_input_tokens_seen": 2299696,
      "step": 1700
    },
    {
      "epoch": 0.05457397093655976,
      "grad_norm": 11.0,
      "learning_rate": 4.97764469288984e-05,
      "loss": 3.9047,
      "num_input_tokens_seen": 2306496,
      "step": 1705
    },
    {
      "epoch": 0.054734011907048206,
      "grad_norm": 17.375,
      "learning_rate": 4.977513504609127e-05,
      "loss": 3.6829,
      "num_input_tokens_seen": 2313600,
      "step": 1710
    },
    {
      "epoch": 0.054894052877536646,
      "grad_norm": 28.25,
      "learning_rate": 4.9773819342646965e-05,
      "loss": 4.5149,
      "num_input_tokens_seen": 2320480,
      "step": 1715
    },
    {
      "epoch": 0.05505409384802509,
      "grad_norm": 38.75,
      "learning_rate": 4.97724998187684e-05,
      "loss": 4.425,
      "num_input_tokens_seen": 2327904,
      "step": 1720
    },
    {
      "epoch": 0.05521413481851354,
      "grad_norm": 23.875,
      "learning_rate": 4.9771176474659045e-05,
      "loss": 4.9255,
      "num_input_tokens_seen": 2334400,
      "step": 1725
    },
    {
      "epoch": 0.055374175789001986,
      "grad_norm": 20.75,
      "learning_rate": 4.976984931052299e-05,
      "loss": 4.0485,
      "num_input_tokens_seen": 2341264,
      "step": 1730
    },
    {
      "epoch": 0.055534216759490426,
      "grad_norm": 21.5,
      "learning_rate": 4.976851832656489e-05,
      "loss": 3.7852,
      "num_input_tokens_seen": 2348080,
      "step": 1735
    },
    {
      "epoch": 0.05569425772997887,
      "grad_norm": 23.25,
      "learning_rate": 4.9767183522990004e-05,
      "loss": 4.1353,
      "num_input_tokens_seen": 2354800,
      "step": 1740
    },
    {
      "epoch": 0.05585429870046732,
      "grad_norm": 27.25,
      "learning_rate": 4.9765844900004176e-05,
      "loss": 4.6811,
      "num_input_tokens_seen": 2361968,
      "step": 1745
    },
    {
      "epoch": 0.056014339670955766,
      "grad_norm": 43.5,
      "learning_rate": 4.9764502457813834e-05,
      "loss": 5.01,
      "num_input_tokens_seen": 2368448,
      "step": 1750
    },
    {
      "epoch": 0.05617438064144421,
      "grad_norm": 31.0,
      "learning_rate": 4.9763156196626005e-05,
      "loss": 4.5149,
      "num_input_tokens_seen": 2375248,
      "step": 1755
    },
    {
      "epoch": 0.05633442161193265,
      "grad_norm": 23.25,
      "learning_rate": 4.97618061166483e-05,
      "loss": 4.5423,
      "num_input_tokens_seen": 2381920,
      "step": 1760
    },
    {
      "epoch": 0.0564944625824211,
      "grad_norm": 40.75,
      "learning_rate": 4.9760452218088915e-05,
      "loss": 4.3787,
      "num_input_tokens_seen": 2388320,
      "step": 1765
    },
    {
      "epoch": 0.056654503552909546,
      "grad_norm": 44.25,
      "learning_rate": 4.975909450115663e-05,
      "loss": 4.4625,
      "num_input_tokens_seen": 2395200,
      "step": 1770
    },
    {
      "epoch": 0.05681454452339799,
      "grad_norm": 30.125,
      "learning_rate": 4.975773296606084e-05,
      "loss": 5.0833,
      "num_input_tokens_seen": 2402032,
      "step": 1775
    },
    {
      "epoch": 0.05697458549388643,
      "grad_norm": 23.625,
      "learning_rate": 4.97563676130115e-05,
      "loss": 4.4891,
      "num_input_tokens_seen": 2409296,
      "step": 1780
    },
    {
      "epoch": 0.05713462646437488,
      "grad_norm": 20.5,
      "learning_rate": 4.9754998442219166e-05,
      "loss": 4.8565,
      "num_input_tokens_seen": 2416096,
      "step": 1785
    },
    {
      "epoch": 0.057294667434863326,
      "grad_norm": 39.75,
      "learning_rate": 4.9753625453894984e-05,
      "loss": 4.8936,
      "num_input_tokens_seen": 2422992,
      "step": 1790
    },
    {
      "epoch": 0.05745470840535177,
      "grad_norm": 35.5,
      "learning_rate": 4.975224864825068e-05,
      "loss": 5.0471,
      "num_input_tokens_seen": 2429280,
      "step": 1795
    },
    {
      "epoch": 0.05761474937584021,
      "grad_norm": 41.0,
      "learning_rate": 4.9750868025498576e-05,
      "loss": 4.0815,
      "num_input_tokens_seen": 2436032,
      "step": 1800
    },
    {
      "epoch": 0.05761474937584021,
      "eval_loss": 4.5198869705200195,
      "eval_runtime": 487.3976,
      "eval_samples_per_second": 28.49,
      "eval_steps_per_second": 14.245,
      "num_input_tokens_seen": 2436032,
      "step": 1800
    },
    {
      "epoch": 0.05777479034632866,
      "grad_norm": 11.75,
      "learning_rate": 4.974948358585158e-05,
      "loss": 4.4113,
      "num_input_tokens_seen": 2442544,
      "step": 1805
    },
    {
      "epoch": 0.057934831316817106,
      "grad_norm": 29.125,
      "learning_rate": 4.9748095329523205e-05,
      "loss": 4.6435,
      "num_input_tokens_seen": 2449312,
      "step": 1810
    },
    {
      "epoch": 0.05809487228730555,
      "grad_norm": 22.625,
      "learning_rate": 4.974670325672752e-05,
      "loss": 3.5787,
      "num_input_tokens_seen": 2456208,
      "step": 1815
    },
    {
      "epoch": 0.05825491325779399,
      "grad_norm": 31.25,
      "learning_rate": 4.974530736767921e-05,
      "loss": 3.6336,
      "num_input_tokens_seen": 2462896,
      "step": 1820
    },
    {
      "epoch": 0.05841495422828244,
      "grad_norm": 19.875,
      "learning_rate": 4.9743907662593524e-05,
      "loss": 4.5508,
      "num_input_tokens_seen": 2469296,
      "step": 1825
    },
    {
      "epoch": 0.058574995198770886,
      "grad_norm": 35.25,
      "learning_rate": 4.974250414168633e-05,
      "loss": 4.0816,
      "num_input_tokens_seen": 2476064,
      "step": 1830
    },
    {
      "epoch": 0.05873503616925933,
      "grad_norm": 22.125,
      "learning_rate": 4.974109680517407e-05,
      "loss": 4.7094,
      "num_input_tokens_seen": 2482832,
      "step": 1835
    },
    {
      "epoch": 0.05889507713974777,
      "grad_norm": 36.5,
      "learning_rate": 4.973968565327376e-05,
      "loss": 4.8071,
      "num_input_tokens_seen": 2489472,
      "step": 1840
    },
    {
      "epoch": 0.05905511811023622,
      "grad_norm": 57.0,
      "learning_rate": 4.973827068620303e-05,
      "loss": 4.2339,
      "num_input_tokens_seen": 2496528,
      "step": 1845
    },
    {
      "epoch": 0.059215159080724666,
      "grad_norm": 27.5,
      "learning_rate": 4.973685190418008e-05,
      "loss": 4.2914,
      "num_input_tokens_seen": 2503184,
      "step": 1850
    },
    {
      "epoch": 0.05937520005121311,
      "grad_norm": 42.25,
      "learning_rate": 4.97354293074237e-05,
      "loss": 4.6522,
      "num_input_tokens_seen": 2509504,
      "step": 1855
    },
    {
      "epoch": 0.05953524102170155,
      "grad_norm": 25.25,
      "learning_rate": 4.9734002896153276e-05,
      "loss": 3.8361,
      "num_input_tokens_seen": 2516304,
      "step": 1860
    },
    {
      "epoch": 0.05969528199219,
      "grad_norm": 28.625,
      "learning_rate": 4.973257267058877e-05,
      "loss": 4.0424,
      "num_input_tokens_seen": 2522736,
      "step": 1865
    },
    {
      "epoch": 0.059855322962678446,
      "grad_norm": 25.875,
      "learning_rate": 4.973113863095076e-05,
      "loss": 3.9261,
      "num_input_tokens_seen": 2530128,
      "step": 1870
    },
    {
      "epoch": 0.06001536393316689,
      "grad_norm": 25.5,
      "learning_rate": 4.9729700777460384e-05,
      "loss": 4.7405,
      "num_input_tokens_seen": 2536752,
      "step": 1875
    },
    {
      "epoch": 0.06017540490365533,
      "grad_norm": 37.0,
      "learning_rate": 4.972825911033937e-05,
      "loss": 3.4801,
      "num_input_tokens_seen": 2543920,
      "step": 1880
    },
    {
      "epoch": 0.06033544587414378,
      "grad_norm": 17.5,
      "learning_rate": 4.9726813629810056e-05,
      "loss": 4.1063,
      "num_input_tokens_seen": 2550528,
      "step": 1885
    },
    {
      "epoch": 0.060495486844632226,
      "grad_norm": 28.375,
      "learning_rate": 4.9725364336095326e-05,
      "loss": 5.1018,
      "num_input_tokens_seen": 2557184,
      "step": 1890
    },
    {
      "epoch": 0.06065552781512067,
      "grad_norm": 42.25,
      "learning_rate": 4.972391122941871e-05,
      "loss": 3.2609,
      "num_input_tokens_seen": 2563536,
      "step": 1895
    },
    {
      "epoch": 0.06081556878560911,
      "grad_norm": 44.5,
      "learning_rate": 4.972245431000428e-05,
      "loss": 4.8014,
      "num_input_tokens_seen": 2570656,
      "step": 1900
    },
    {
      "epoch": 0.06097560975609756,
      "grad_norm": 33.5,
      "learning_rate": 4.972099357807671e-05,
      "loss": 4.9721,
      "num_input_tokens_seen": 2577248,
      "step": 1905
    },
    {
      "epoch": 0.061135650726586006,
      "grad_norm": 30.625,
      "learning_rate": 4.971952903386127e-05,
      "loss": 4.2986,
      "num_input_tokens_seen": 2583904,
      "step": 1910
    },
    {
      "epoch": 0.06129569169707445,
      "grad_norm": 25.5,
      "learning_rate": 4.971806067758381e-05,
      "loss": 4.3017,
      "num_input_tokens_seen": 2590848,
      "step": 1915
    },
    {
      "epoch": 0.06145573266756289,
      "grad_norm": 21.625,
      "learning_rate": 4.971658850947076e-05,
      "loss": 3.7463,
      "num_input_tokens_seen": 2597392,
      "step": 1920
    },
    {
      "epoch": 0.06161577363805134,
      "grad_norm": 45.0,
      "learning_rate": 4.9715112529749165e-05,
      "loss": 4.8169,
      "num_input_tokens_seen": 2603712,
      "step": 1925
    },
    {
      "epoch": 0.061775814608539786,
      "grad_norm": 29.0,
      "learning_rate": 4.9713632738646624e-05,
      "loss": 5.2083,
      "num_input_tokens_seen": 2610528,
      "step": 1930
    },
    {
      "epoch": 0.06193585557902823,
      "grad_norm": 25.125,
      "learning_rate": 4.971214913639134e-05,
      "loss": 4.1879,
      "num_input_tokens_seen": 2617456,
      "step": 1935
    },
    {
      "epoch": 0.06209589654951667,
      "grad_norm": 23.75,
      "learning_rate": 4.9710661723212104e-05,
      "loss": 3.6422,
      "num_input_tokens_seen": 2624384,
      "step": 1940
    },
    {
      "epoch": 0.06225593752000512,
      "grad_norm": 30.125,
      "learning_rate": 4.9709170499338295e-05,
      "loss": 4.4535,
      "num_input_tokens_seen": 2630992,
      "step": 1945
    },
    {
      "epoch": 0.062415978490493566,
      "grad_norm": 19.5,
      "learning_rate": 4.9707675464999895e-05,
      "loss": 3.6784,
      "num_input_tokens_seen": 2637808,
      "step": 1950
    },
    {
      "epoch": 0.06257601946098201,
      "grad_norm": 35.25,
      "learning_rate": 4.970617662042743e-05,
      "loss": 3.8492,
      "num_input_tokens_seen": 2644688,
      "step": 1955
    },
    {
      "epoch": 0.06273606043147045,
      "grad_norm": 54.25,
      "learning_rate": 4.970467396585206e-05,
      "loss": 4.7991,
      "num_input_tokens_seen": 2651328,
      "step": 1960
    },
    {
      "epoch": 0.0628961014019589,
      "grad_norm": 24.75,
      "learning_rate": 4.97031675015055e-05,
      "loss": 4.2392,
      "num_input_tokens_seen": 2657760,
      "step": 1965
    },
    {
      "epoch": 0.06305614237244735,
      "grad_norm": 24.5,
      "learning_rate": 4.9701657227620075e-05,
      "loss": 3.9346,
      "num_input_tokens_seen": 2666016,
      "step": 1970
    },
    {
      "epoch": 0.06321618334293579,
      "grad_norm": 16.0,
      "learning_rate": 4.9700143144428685e-05,
      "loss": 5.2777,
      "num_input_tokens_seen": 2673008,
      "step": 1975
    },
    {
      "epoch": 0.06337622431342424,
      "grad_norm": 27.25,
      "learning_rate": 4.969862525216482e-05,
      "loss": 5.5554,
      "num_input_tokens_seen": 2679808,
      "step": 1980
    },
    {
      "epoch": 0.06353626528391268,
      "grad_norm": 134.0,
      "learning_rate": 4.9697103551062556e-05,
      "loss": 4.1596,
      "num_input_tokens_seen": 2686640,
      "step": 1985
    },
    {
      "epoch": 0.06369630625440113,
      "grad_norm": 38.5,
      "learning_rate": 4.9695578041356565e-05,
      "loss": 4.3765,
      "num_input_tokens_seen": 2693696,
      "step": 1990
    },
    {
      "epoch": 0.06385634722488957,
      "grad_norm": 39.75,
      "learning_rate": 4.969404872328209e-05,
      "loss": 4.6469,
      "num_input_tokens_seen": 2700240,
      "step": 1995
    },
    {
      "epoch": 0.06401638819537801,
      "grad_norm": 18.625,
      "learning_rate": 4.969251559707498e-05,
      "loss": 4.5046,
      "num_input_tokens_seen": 2706960,
      "step": 2000
    },
    {
      "epoch": 0.06401638819537801,
      "eval_loss": 4.511370658874512,
      "eval_runtime": 487.87,
      "eval_samples_per_second": 28.463,
      "eval_steps_per_second": 14.231,
      "num_input_tokens_seen": 2706960,
      "step": 2000
    },
    {
      "epoch": 0.06417642916586647,
      "grad_norm": 37.0,
      "learning_rate": 4.9690978662971674e-05,
      "loss": 4.3309,
      "num_input_tokens_seen": 2714000,
      "step": 2005
    },
    {
      "epoch": 0.0643364701363549,
      "grad_norm": 23.875,
      "learning_rate": 4.968943792120916e-05,
      "loss": 4.8545,
      "num_input_tokens_seen": 2720784,
      "step": 2010
    },
    {
      "epoch": 0.06449651110684335,
      "grad_norm": 23.5,
      "learning_rate": 4.9687893372025046e-05,
      "loss": 4.4763,
      "num_input_tokens_seen": 2727136,
      "step": 2015
    },
    {
      "epoch": 0.0646565520773318,
      "grad_norm": 22.875,
      "learning_rate": 4.9686345015657535e-05,
      "loss": 4.3016,
      "num_input_tokens_seen": 2733776,
      "step": 2020
    },
    {
      "epoch": 0.06481659304782024,
      "grad_norm": 26.0,
      "learning_rate": 4.968479285234538e-05,
      "loss": 3.7708,
      "num_input_tokens_seen": 2740512,
      "step": 2025
    },
    {
      "epoch": 0.0649766340183087,
      "grad_norm": 20.125,
      "learning_rate": 4.9683236882327974e-05,
      "loss": 3.8676,
      "num_input_tokens_seen": 2747200,
      "step": 2030
    },
    {
      "epoch": 0.06513667498879713,
      "grad_norm": 23.625,
      "learning_rate": 4.968167710584526e-05,
      "loss": 4.1902,
      "num_input_tokens_seen": 2753824,
      "step": 2035
    },
    {
      "epoch": 0.06529671595928557,
      "grad_norm": 29.875,
      "learning_rate": 4.968011352313775e-05,
      "loss": 5.7319,
      "num_input_tokens_seen": 2760384,
      "step": 2040
    },
    {
      "epoch": 0.06545675692977403,
      "grad_norm": 23.875,
      "learning_rate": 4.967854613444659e-05,
      "loss": 3.5772,
      "num_input_tokens_seen": 2767312,
      "step": 2045
    },
    {
      "epoch": 0.06561679790026247,
      "grad_norm": 28.375,
      "learning_rate": 4.967697494001349e-05,
      "loss": 4.151,
      "num_input_tokens_seen": 2774544,
      "step": 2050
    },
    {
      "epoch": 0.0657768388707509,
      "grad_norm": 40.5,
      "learning_rate": 4.9675399940080736e-05,
      "loss": 4.5804,
      "num_input_tokens_seen": 2781280,
      "step": 2055
    },
    {
      "epoch": 0.06593687984123936,
      "grad_norm": 35.75,
      "learning_rate": 4.9673821134891226e-05,
      "loss": 4.747,
      "num_input_tokens_seen": 2788416,
      "step": 2060
    },
    {
      "epoch": 0.0660969208117278,
      "grad_norm": 17.125,
      "learning_rate": 4.967223852468842e-05,
      "loss": 5.6032,
      "num_input_tokens_seen": 2794752,
      "step": 2065
    },
    {
      "epoch": 0.06625696178221625,
      "grad_norm": 24.75,
      "learning_rate": 4.967065210971639e-05,
      "loss": 4.1692,
      "num_input_tokens_seen": 2801328,
      "step": 2070
    },
    {
      "epoch": 0.06641700275270469,
      "grad_norm": 27.75,
      "learning_rate": 4.966906189021977e-05,
      "loss": 3.6154,
      "num_input_tokens_seen": 2808224,
      "step": 2075
    },
    {
      "epoch": 0.06657704372319313,
      "grad_norm": 21.5,
      "learning_rate": 4.966746786644379e-05,
      "loss": 4.8695,
      "num_input_tokens_seen": 2815184,
      "step": 2080
    },
    {
      "epoch": 0.06673708469368159,
      "grad_norm": 35.0,
      "learning_rate": 4.966587003863429e-05,
      "loss": 3.8771,
      "num_input_tokens_seen": 2822032,
      "step": 2085
    },
    {
      "epoch": 0.06689712566417003,
      "grad_norm": 27.875,
      "learning_rate": 4.966426840703765e-05,
      "loss": 4.2668,
      "num_input_tokens_seen": 2829248,
      "step": 2090
    },
    {
      "epoch": 0.06705716663465847,
      "grad_norm": 43.25,
      "learning_rate": 4.9662662971900875e-05,
      "loss": 5.2681,
      "num_input_tokens_seen": 2835568,
      "step": 2095
    },
    {
      "epoch": 0.06721720760514692,
      "grad_norm": 28.375,
      "learning_rate": 4.9661053733471534e-05,
      "loss": 4.4478,
      "num_input_tokens_seen": 2842224,
      "step": 2100
    },
    {
      "epoch": 0.06737724857563536,
      "grad_norm": 25.875,
      "learning_rate": 4.965944069199781e-05,
      "loss": 4.6504,
      "num_input_tokens_seen": 2848880,
      "step": 2105
    },
    {
      "epoch": 0.06753728954612381,
      "grad_norm": 32.5,
      "learning_rate": 4.965782384772842e-05,
      "loss": 3.8665,
      "num_input_tokens_seen": 2855664,
      "step": 2110
    },
    {
      "epoch": 0.06769733051661225,
      "grad_norm": 62.0,
      "learning_rate": 4.9656203200912734e-05,
      "loss": 5.094,
      "num_input_tokens_seen": 2862720,
      "step": 2115
    },
    {
      "epoch": 0.06785737148710069,
      "grad_norm": 32.0,
      "learning_rate": 4.965457875180067e-05,
      "loss": 4.3075,
      "num_input_tokens_seen": 2869376,
      "step": 2120
    },
    {
      "epoch": 0.06801741245758915,
      "grad_norm": 30.5,
      "learning_rate": 4.9652950500642724e-05,
      "loss": 4.4661,
      "num_input_tokens_seen": 2876272,
      "step": 2125
    },
    {
      "epoch": 0.06817745342807759,
      "grad_norm": 28.5,
      "learning_rate": 4.965131844769001e-05,
      "loss": 4.3992,
      "num_input_tokens_seen": 2882848,
      "step": 2130
    },
    {
      "epoch": 0.06833749439856603,
      "grad_norm": 23.125,
      "learning_rate": 4.96496825931942e-05,
      "loss": 4.2714,
      "num_input_tokens_seen": 2889856,
      "step": 2135
    },
    {
      "epoch": 0.06849753536905448,
      "grad_norm": 27.375,
      "learning_rate": 4.9648042937407566e-05,
      "loss": 3.1333,
      "num_input_tokens_seen": 2896768,
      "step": 2140
    },
    {
      "epoch": 0.06865757633954292,
      "grad_norm": 20.875,
      "learning_rate": 4.964639948058297e-05,
      "loss": 4.6239,
      "num_input_tokens_seen": 2903344,
      "step": 2145
    },
    {
      "epoch": 0.06881761731003137,
      "grad_norm": 22.25,
      "learning_rate": 4.9644752222973846e-05,
      "loss": 4.3524,
      "num_input_tokens_seen": 2910080,
      "step": 2150
    },
    {
      "epoch": 0.06897765828051981,
      "grad_norm": 29.5,
      "learning_rate": 4.964310116483422e-05,
      "loss": 4.8904,
      "num_input_tokens_seen": 2916688,
      "step": 2155
    },
    {
      "epoch": 0.06913769925100825,
      "grad_norm": 33.5,
      "learning_rate": 4.964144630641872e-05,
      "loss": 3.975,
      "num_input_tokens_seen": 2923168,
      "step": 2160
    },
    {
      "epoch": 0.0692977402214967,
      "grad_norm": 12.0,
      "learning_rate": 4.9639787647982525e-05,
      "loss": 4.2116,
      "num_input_tokens_seen": 2930032,
      "step": 2165
    },
    {
      "epoch": 0.06945778119198515,
      "grad_norm": 24.375,
      "learning_rate": 4.963812518978143e-05,
      "loss": 4.9,
      "num_input_tokens_seen": 2936640,
      "step": 2170
    },
    {
      "epoch": 0.0696178221624736,
      "grad_norm": 27.25,
      "learning_rate": 4.963645893207182e-05,
      "loss": 4.3489,
      "num_input_tokens_seen": 2943536,
      "step": 2175
    },
    {
      "epoch": 0.06977786313296204,
      "grad_norm": 28.125,
      "learning_rate": 4.963478887511063e-05,
      "loss": 4.9938,
      "num_input_tokens_seen": 2950016,
      "step": 2180
    },
    {
      "epoch": 0.06993790410345048,
      "grad_norm": 25.25,
      "learning_rate": 4.963311501915542e-05,
      "loss": 3.4801,
      "num_input_tokens_seen": 2956528,
      "step": 2185
    },
    {
      "epoch": 0.07009794507393893,
      "grad_norm": 38.75,
      "learning_rate": 4.963143736446432e-05,
      "loss": 5.1497,
      "num_input_tokens_seen": 2962848,
      "step": 2190
    },
    {
      "epoch": 0.07025798604442737,
      "grad_norm": 23.625,
      "learning_rate": 4.962975591129603e-05,
      "loss": 3.6706,
      "num_input_tokens_seen": 2969248,
      "step": 2195
    },
    {
      "epoch": 0.07041802701491581,
      "grad_norm": 37.75,
      "learning_rate": 4.962807065990986e-05,
      "loss": 6.1891,
      "num_input_tokens_seen": 2976144,
      "step": 2200
    },
    {
      "epoch": 0.07041802701491581,
      "eval_loss": 4.510693073272705,
      "eval_runtime": 487.7166,
      "eval_samples_per_second": 28.471,
      "eval_steps_per_second": 14.236,
      "num_input_tokens_seen": 2976144,
      "step": 2200
    },
    {
      "epoch": 0.07057806798540427,
      "grad_norm": 34.75,
      "learning_rate": 4.9626381610565714e-05,
      "loss": 3.6036,
      "num_input_tokens_seen": 2982992,
      "step": 2205
    },
    {
      "epoch": 0.0707381089558927,
      "grad_norm": 35.5,
      "learning_rate": 4.9624688763524043e-05,
      "loss": 4.0685,
      "num_input_tokens_seen": 2989568,
      "step": 2210
    },
    {
      "epoch": 0.07089814992638116,
      "grad_norm": 33.25,
      "learning_rate": 4.962299211904591e-05,
      "loss": 4.4228,
      "num_input_tokens_seen": 2996336,
      "step": 2215
    },
    {
      "epoch": 0.0710581908968696,
      "grad_norm": 31.875,
      "learning_rate": 4.962129167739296e-05,
      "loss": 4.3961,
      "num_input_tokens_seen": 3003232,
      "step": 2220
    },
    {
      "epoch": 0.07121823186735804,
      "grad_norm": 25.75,
      "learning_rate": 4.961958743882742e-05,
      "loss": 3.4919,
      "num_input_tokens_seen": 3010144,
      "step": 2225
    },
    {
      "epoch": 0.0713782728378465,
      "grad_norm": 31.375,
      "learning_rate": 4.961787940361211e-05,
      "loss": 3.5666,
      "num_input_tokens_seen": 3017264,
      "step": 2230
    },
    {
      "epoch": 0.07153831380833493,
      "grad_norm": 29.5,
      "learning_rate": 4.961616757201043e-05,
      "loss": 5.595,
      "num_input_tokens_seen": 3023968,
      "step": 2235
    },
    {
      "epoch": 0.07169835477882337,
      "grad_norm": 27.75,
      "learning_rate": 4.961445194428637e-05,
      "loss": 3.7699,
      "num_input_tokens_seen": 3030480,
      "step": 2240
    },
    {
      "epoch": 0.07185839574931183,
      "grad_norm": 37.25,
      "learning_rate": 4.9612732520704486e-05,
      "loss": 3.3139,
      "num_input_tokens_seen": 3037520,
      "step": 2245
    },
    {
      "epoch": 0.07201843671980027,
      "grad_norm": 20.5,
      "learning_rate": 4.961100930152994e-05,
      "loss": 5.2164,
      "num_input_tokens_seen": 3043808,
      "step": 2250
    },
    {
      "epoch": 0.07217847769028872,
      "grad_norm": 44.5,
      "learning_rate": 4.960928228702849e-05,
      "loss": 4.7677,
      "num_input_tokens_seen": 3050544,
      "step": 2255
    },
    {
      "epoch": 0.07233851866077716,
      "grad_norm": 34.0,
      "learning_rate": 4.960755147746645e-05,
      "loss": 3.7818,
      "num_input_tokens_seen": 3057408,
      "step": 2260
    },
    {
      "epoch": 0.0724985596312656,
      "grad_norm": 14.75,
      "learning_rate": 4.9605816873110736e-05,
      "loss": 3.5586,
      "num_input_tokens_seen": 3063792,
      "step": 2265
    },
    {
      "epoch": 0.07265860060175405,
      "grad_norm": 33.25,
      "learning_rate": 4.960407847422883e-05,
      "loss": 3.8746,
      "num_input_tokens_seen": 3070672,
      "step": 2270
    },
    {
      "epoch": 0.07281864157224249,
      "grad_norm": 19.5,
      "learning_rate": 4.960233628108885e-05,
      "loss": 4.9879,
      "num_input_tokens_seen": 3077200,
      "step": 2275
    },
    {
      "epoch": 0.07297868254273093,
      "grad_norm": 22.625,
      "learning_rate": 4.960059029395942e-05,
      "loss": 4.4438,
      "num_input_tokens_seen": 3084784,
      "step": 2280
    },
    {
      "epoch": 0.07313872351321939,
      "grad_norm": 31.75,
      "learning_rate": 4.959884051310983e-05,
      "loss": 3.9192,
      "num_input_tokens_seen": 3091712,
      "step": 2285
    },
    {
      "epoch": 0.07329876448370783,
      "grad_norm": 36.0,
      "learning_rate": 4.959708693880991e-05,
      "loss": 5.2383,
      "num_input_tokens_seen": 3098512,
      "step": 2290
    },
    {
      "epoch": 0.07345880545419628,
      "grad_norm": 36.0,
      "learning_rate": 4.9595329571330074e-05,
      "loss": 4.2746,
      "num_input_tokens_seen": 3105648,
      "step": 2295
    },
    {
      "epoch": 0.07361884642468472,
      "grad_norm": 35.25,
      "learning_rate": 4.9593568410941326e-05,
      "loss": 4.6477,
      "num_input_tokens_seen": 3112608,
      "step": 2300
    },
    {
      "epoch": 0.07377888739517316,
      "grad_norm": 29.125,
      "learning_rate": 4.959180345791528e-05,
      "loss": 4.609,
      "num_input_tokens_seen": 3119088,
      "step": 2305
    },
    {
      "epoch": 0.07393892836566161,
      "grad_norm": 48.25,
      "learning_rate": 4.9590034712524086e-05,
      "loss": 5.0851,
      "num_input_tokens_seen": 3126336,
      "step": 2310
    },
    {
      "epoch": 0.07409896933615005,
      "grad_norm": 28.0,
      "learning_rate": 4.958826217504053e-05,
      "loss": 3.5807,
      "num_input_tokens_seen": 3133200,
      "step": 2315
    },
    {
      "epoch": 0.07425901030663849,
      "grad_norm": 29.125,
      "learning_rate": 4.958648584573795e-05,
      "loss": 3.5284,
      "num_input_tokens_seen": 3140096,
      "step": 2320
    },
    {
      "epoch": 0.07441905127712695,
      "grad_norm": 32.0,
      "learning_rate": 4.958470572489028e-05,
      "loss": 4.5168,
      "num_input_tokens_seen": 3146880,
      "step": 2325
    },
    {
      "epoch": 0.07457909224761539,
      "grad_norm": 18.0,
      "learning_rate": 4.958292181277203e-05,
      "loss": 4.2675,
      "num_input_tokens_seen": 3153808,
      "step": 2330
    },
    {
      "epoch": 0.07473913321810384,
      "grad_norm": 17.25,
      "learning_rate": 4.958113410965832e-05,
      "loss": 3.8033,
      "num_input_tokens_seen": 3160288,
      "step": 2335
    },
    {
      "epoch": 0.07489917418859228,
      "grad_norm": 41.0,
      "learning_rate": 4.957934261582481e-05,
      "loss": 4.9943,
      "num_input_tokens_seen": 3166624,
      "step": 2340
    },
    {
      "epoch": 0.07505921515908072,
      "grad_norm": 22.25,
      "learning_rate": 4.95775473315478e-05,
      "loss": 4.4381,
      "num_input_tokens_seen": 3173232,
      "step": 2345
    },
    {
      "epoch": 0.07521925612956917,
      "grad_norm": 27.0,
      "learning_rate": 4.9575748257104124e-05,
      "loss": 4.216,
      "num_input_tokens_seen": 3180000,
      "step": 2350
    },
    {
      "epoch": 0.07537929710005761,
      "grad_norm": 28.375,
      "learning_rate": 4.9573945392771224e-05,
      "loss": 4.574,
      "num_input_tokens_seen": 3186544,
      "step": 2355
    },
    {
      "epoch": 0.07553933807054607,
      "grad_norm": 28.625,
      "learning_rate": 4.9572138738827134e-05,
      "loss": 3.5742,
      "num_input_tokens_seen": 3193376,
      "step": 2360
    },
    {
      "epoch": 0.0756993790410345,
      "grad_norm": 42.5,
      "learning_rate": 4.957032829555046e-05,
      "loss": 4.3855,
      "num_input_tokens_seen": 3200000,
      "step": 2365
    },
    {
      "epoch": 0.07585942001152295,
      "grad_norm": 18.375,
      "learning_rate": 4.956851406322039e-05,
      "loss": 4.8576,
      "num_input_tokens_seen": 3206720,
      "step": 2370
    },
    {
      "epoch": 0.0760194609820114,
      "grad_norm": 15.6875,
      "learning_rate": 4.9566696042116704e-05,
      "loss": 4.5718,
      "num_input_tokens_seen": 3213248,
      "step": 2375
    },
    {
      "epoch": 0.07617950195249984,
      "grad_norm": 40.5,
      "learning_rate": 4.9564874232519766e-05,
      "loss": 4.1799,
      "num_input_tokens_seen": 3220704,
      "step": 2380
    },
    {
      "epoch": 0.07633954292298828,
      "grad_norm": 20.75,
      "learning_rate": 4.9563048634710516e-05,
      "loss": 3.6001,
      "num_input_tokens_seen": 3228048,
      "step": 2385
    },
    {
      "epoch": 0.07649958389347673,
      "grad_norm": 31.75,
      "learning_rate": 4.956121924897049e-05,
      "loss": 4.8026,
      "num_input_tokens_seen": 3234832,
      "step": 2390
    },
    {
      "epoch": 0.07665962486396517,
      "grad_norm": 27.375,
      "learning_rate": 4.955938607558181e-05,
      "loss": 3.89,
      "num_input_tokens_seen": 3241952,
      "step": 2395
    },
    {
      "epoch": 0.07681966583445363,
      "grad_norm": 35.75,
      "learning_rate": 4.955754911482715e-05,
      "loss": 3.7462,
      "num_input_tokens_seen": 3248384,
      "step": 2400
    },
    {
      "epoch": 0.07681966583445363,
      "eval_loss": 4.510420322418213,
      "eval_runtime": 487.425,
      "eval_samples_per_second": 28.488,
      "eval_steps_per_second": 14.244,
      "num_input_tokens_seen": 3248384,
      "step": 2400
    },
    {
      "epoch": 0.07697970680494207,
      "grad_norm": 42.0,
      "learning_rate": 4.9555708366989804e-05,
      "loss": 4.2226,
      "num_input_tokens_seen": 3254784,
      "step": 2405
    },
    {
      "epoch": 0.0771397477754305,
      "grad_norm": 29.5,
      "learning_rate": 4.9553863832353655e-05,
      "loss": 4.8066,
      "num_input_tokens_seen": 3261504,
      "step": 2410
    },
    {
      "epoch": 0.07729978874591896,
      "grad_norm": 20.625,
      "learning_rate": 4.955201551120313e-05,
      "loss": 4.093,
      "num_input_tokens_seen": 3268160,
      "step": 2415
    },
    {
      "epoch": 0.0774598297164074,
      "grad_norm": 22.625,
      "learning_rate": 4.955016340382328e-05,
      "loss": 4.057,
      "num_input_tokens_seen": 3275184,
      "step": 2420
    },
    {
      "epoch": 0.07761987068689584,
      "grad_norm": 24.0,
      "learning_rate": 4.954830751049972e-05,
      "loss": 3.7462,
      "num_input_tokens_seen": 3282064,
      "step": 2425
    },
    {
      "epoch": 0.07777991165738429,
      "grad_norm": 44.0,
      "learning_rate": 4.954644783151864e-05,
      "loss": 5.0383,
      "num_input_tokens_seen": 3288704,
      "step": 2430
    },
    {
      "epoch": 0.07793995262787273,
      "grad_norm": 27.875,
      "learning_rate": 4.954458436716684e-05,
      "loss": 3.8268,
      "num_input_tokens_seen": 3295984,
      "step": 2435
    },
    {
      "epoch": 0.07809999359836119,
      "grad_norm": 35.75,
      "learning_rate": 4.954271711773168e-05,
      "loss": 4.0962,
      "num_input_tokens_seen": 3302592,
      "step": 2440
    },
    {
      "epoch": 0.07826003456884963,
      "grad_norm": 44.75,
      "learning_rate": 4.9540846083501115e-05,
      "loss": 4.5044,
      "num_input_tokens_seen": 3309424,
      "step": 2445
    },
    {
      "epoch": 0.07842007553933807,
      "grad_norm": 24.75,
      "learning_rate": 4.953897126476369e-05,
      "loss": 4.6272,
      "num_input_tokens_seen": 3315728,
      "step": 2450
    },
    {
      "epoch": 0.07858011650982652,
      "grad_norm": 29.5,
      "learning_rate": 4.9537092661808514e-05,
      "loss": 3.8657,
      "num_input_tokens_seen": 3322448,
      "step": 2455
    },
    {
      "epoch": 0.07874015748031496,
      "grad_norm": 21.625,
      "learning_rate": 4.9535210274925306e-05,
      "loss": 5.0862,
      "num_input_tokens_seen": 3329456,
      "step": 2460
    },
    {
      "epoch": 0.0789001984508034,
      "grad_norm": 24.625,
      "learning_rate": 4.953332410440435e-05,
      "loss": 4.232,
      "num_input_tokens_seen": 3336368,
      "step": 2465
    },
    {
      "epoch": 0.07906023942129185,
      "grad_norm": 39.0,
      "learning_rate": 4.9531434150536496e-05,
      "loss": 5.42,
      "num_input_tokens_seen": 3343280,
      "step": 2470
    },
    {
      "epoch": 0.07922028039178029,
      "grad_norm": 24.625,
      "learning_rate": 4.952954041361322e-05,
      "loss": 3.8515,
      "num_input_tokens_seen": 3350112,
      "step": 2475
    },
    {
      "epoch": 0.07938032136226875,
      "grad_norm": 23.375,
      "learning_rate": 4.952764289392655e-05,
      "loss": 3.435,
      "num_input_tokens_seen": 3356528,
      "step": 2480
    },
    {
      "epoch": 0.07954036233275719,
      "grad_norm": 50.5,
      "learning_rate": 4.952574159176912e-05,
      "loss": 4.545,
      "num_input_tokens_seen": 3363024,
      "step": 2485
    },
    {
      "epoch": 0.07970040330324563,
      "grad_norm": 32.25,
      "learning_rate": 4.952383650743413e-05,
      "loss": 4.2361,
      "num_input_tokens_seen": 3370016,
      "step": 2490
    },
    {
      "epoch": 0.07986044427373408,
      "grad_norm": 45.0,
      "learning_rate": 4.952192764121536e-05,
      "loss": 3.9761,
      "num_input_tokens_seen": 3377120,
      "step": 2495
    },
    {
      "epoch": 0.08002048524422252,
      "grad_norm": 39.5,
      "learning_rate": 4.9520014993407185e-05,
      "loss": 5.3796,
      "num_input_tokens_seen": 3383760,
      "step": 2500
    },
    {
      "epoch": 0.08018052621471096,
      "grad_norm": 17.625,
      "learning_rate": 4.951809856430456e-05,
      "loss": 3.7864,
      "num_input_tokens_seen": 3390400,
      "step": 2505
    },
    {
      "epoch": 0.08034056718519941,
      "grad_norm": 29.375,
      "learning_rate": 4.951617835420303e-05,
      "loss": 4.1848,
      "num_input_tokens_seen": 3397168,
      "step": 2510
    },
    {
      "epoch": 0.08050060815568785,
      "grad_norm": 26.5,
      "learning_rate": 4.951425436339869e-05,
      "loss": 3.8872,
      "num_input_tokens_seen": 3403824,
      "step": 2515
    },
    {
      "epoch": 0.0806606491261763,
      "grad_norm": 35.75,
      "learning_rate": 4.9512326592188274e-05,
      "loss": 4.3462,
      "num_input_tokens_seen": 3410288,
      "step": 2520
    },
    {
      "epoch": 0.08082069009666475,
      "grad_norm": 32.5,
      "learning_rate": 4.9510395040869054e-05,
      "loss": 4.9761,
      "num_input_tokens_seen": 3416864,
      "step": 2525
    },
    {
      "epoch": 0.08098073106715319,
      "grad_norm": 23.0,
      "learning_rate": 4.9508459709738905e-05,
      "loss": 4.3965,
      "num_input_tokens_seen": 3423568,
      "step": 2530
    },
    {
      "epoch": 0.08114077203764164,
      "grad_norm": 34.5,
      "learning_rate": 4.950652059909627e-05,
      "loss": 4.9572,
      "num_input_tokens_seen": 3431104,
      "step": 2535
    },
    {
      "epoch": 0.08130081300813008,
      "grad_norm": 19.625,
      "learning_rate": 4.95045777092402e-05,
      "loss": 3.2393,
      "num_input_tokens_seen": 3437680,
      "step": 2540
    },
    {
      "epoch": 0.08146085397861853,
      "grad_norm": 37.75,
      "learning_rate": 4.950263104047031e-05,
      "loss": 4.5807,
      "num_input_tokens_seen": 3444208,
      "step": 2545
    },
    {
      "epoch": 0.08162089494910697,
      "grad_norm": 32.5,
      "learning_rate": 4.9500680593086775e-05,
      "loss": 4.6393,
      "num_input_tokens_seen": 3450864,
      "step": 2550
    },
    {
      "epoch": 0.08178093591959541,
      "grad_norm": 20.5,
      "learning_rate": 4.94987263673904e-05,
      "loss": 4.8016,
      "num_input_tokens_seen": 3457312,
      "step": 2555
    },
    {
      "epoch": 0.08194097689008387,
      "grad_norm": 16.625,
      "learning_rate": 4.949676836368256e-05,
      "loss": 4.5002,
      "num_input_tokens_seen": 3464272,
      "step": 2560
    },
    {
      "epoch": 0.0821010178605723,
      "grad_norm": 24.75,
      "learning_rate": 4.949480658226518e-05,
      "loss": 4.533,
      "num_input_tokens_seen": 3470720,
      "step": 2565
    },
    {
      "epoch": 0.08226105883106075,
      "grad_norm": 25.875,
      "learning_rate": 4.949284102344082e-05,
      "loss": 4.8493,
      "num_input_tokens_seen": 3477744,
      "step": 2570
    },
    {
      "epoch": 0.0824210998015492,
      "grad_norm": 21.375,
      "learning_rate": 4.9490871687512565e-05,
      "loss": 4.6284,
      "num_input_tokens_seen": 3484688,
      "step": 2575
    },
    {
      "epoch": 0.08258114077203764,
      "grad_norm": 21.125,
      "learning_rate": 4.948889857478413e-05,
      "loss": 4.8743,
      "num_input_tokens_seen": 3492096,
      "step": 2580
    },
    {
      "epoch": 0.0827411817425261,
      "grad_norm": 22.625,
      "learning_rate": 4.948692168555978e-05,
      "loss": 3.8457,
      "num_input_tokens_seen": 3498784,
      "step": 2585
    },
    {
      "epoch": 0.08290122271301453,
      "grad_norm": 22.25,
      "learning_rate": 4.94849410201444e-05,
      "loss": 4.2596,
      "num_input_tokens_seen": 3505648,
      "step": 2590
    },
    {
      "epoch": 0.08306126368350297,
      "grad_norm": 44.5,
      "learning_rate": 4.948295657884341e-05,
      "loss": 5.5255,
      "num_input_tokens_seen": 3512272,
      "step": 2595
    },
    {
      "epoch": 0.08322130465399143,
      "grad_norm": 35.5,
      "learning_rate": 4.9480968361962835e-05,
      "loss": 4.6499,
      "num_input_tokens_seen": 3519088,
      "step": 2600
    },
    {
      "epoch": 0.08322130465399143,
      "eval_loss": 4.503515720367432,
      "eval_runtime": 487.3837,
      "eval_samples_per_second": 28.491,
      "eval_steps_per_second": 14.245,
      "num_input_tokens_seen": 3519088,
      "step": 2600
    },
    {
      "epoch": 0.08338134562447987,
      "grad_norm": 27.375,
      "learning_rate": 4.9478976369809305e-05,
      "loss": 4.4503,
      "num_input_tokens_seen": 3525696,
      "step": 2605
    },
    {
      "epoch": 0.0835413865949683,
      "grad_norm": 22.75,
      "learning_rate": 4.947698060268999e-05,
      "loss": 4.7051,
      "num_input_tokens_seen": 3532640,
      "step": 2610
    },
    {
      "epoch": 0.08370142756545676,
      "grad_norm": 23.25,
      "learning_rate": 4.9474981060912665e-05,
      "loss": 4.1823,
      "num_input_tokens_seen": 3539840,
      "step": 2615
    },
    {
      "epoch": 0.0838614685359452,
      "grad_norm": 49.5,
      "learning_rate": 4.94729777447857e-05,
      "loss": 5.1655,
      "num_input_tokens_seen": 3546544,
      "step": 2620
    },
    {
      "epoch": 0.08402150950643365,
      "grad_norm": 16.5,
      "learning_rate": 4.947097065461801e-05,
      "loss": 3.9573,
      "num_input_tokens_seen": 3553584,
      "step": 2625
    },
    {
      "epoch": 0.08418155047692209,
      "grad_norm": 24.125,
      "learning_rate": 4.9468959790719125e-05,
      "loss": 3.9223,
      "num_input_tokens_seen": 3560368,
      "step": 2630
    },
    {
      "epoch": 0.08434159144741053,
      "grad_norm": 48.75,
      "learning_rate": 4.9466945153399146e-05,
      "loss": 5.5976,
      "num_input_tokens_seen": 3566768,
      "step": 2635
    },
    {
      "epoch": 0.08450163241789899,
      "grad_norm": 34.5,
      "learning_rate": 4.9464926742968755e-05,
      "loss": 4.9323,
      "num_input_tokens_seen": 3574112,
      "step": 2640
    },
    {
      "epoch": 0.08466167338838743,
      "grad_norm": 43.5,
      "learning_rate": 4.946290455973921e-05,
      "loss": 4.052,
      "num_input_tokens_seen": 3581024,
      "step": 2645
    },
    {
      "epoch": 0.08482171435887587,
      "grad_norm": 30.625,
      "learning_rate": 4.9460878604022365e-05,
      "loss": 4.674,
      "num_input_tokens_seen": 3588016,
      "step": 2650
    },
    {
      "epoch": 0.08498175532936432,
      "grad_norm": 27.625,
      "learning_rate": 4.945884887613065e-05,
      "loss": 4.9814,
      "num_input_tokens_seen": 3594416,
      "step": 2655
    },
    {
      "epoch": 0.08514179629985276,
      "grad_norm": 25.125,
      "learning_rate": 4.9456815376377055e-05,
      "loss": 4.2029,
      "num_input_tokens_seen": 3601072,
      "step": 2660
    },
    {
      "epoch": 0.08530183727034121,
      "grad_norm": 26.0,
      "learning_rate": 4.9454778105075195e-05,
      "loss": 4.419,
      "num_input_tokens_seen": 3607456,
      "step": 2665
    },
    {
      "epoch": 0.08546187824082965,
      "grad_norm": 36.25,
      "learning_rate": 4.945273706253924e-05,
      "loss": 4.7082,
      "num_input_tokens_seen": 3614352,
      "step": 2670
    },
    {
      "epoch": 0.08562191921131809,
      "grad_norm": 26.375,
      "learning_rate": 4.9450692249083925e-05,
      "loss": 3.5042,
      "num_input_tokens_seen": 3621376,
      "step": 2675
    },
    {
      "epoch": 0.08578196018180655,
      "grad_norm": 26.375,
      "learning_rate": 4.9448643665024605e-05,
      "loss": 4.7783,
      "num_input_tokens_seen": 3628176,
      "step": 2680
    },
    {
      "epoch": 0.08594200115229499,
      "grad_norm": 53.5,
      "learning_rate": 4.944659131067719e-05,
      "loss": 3.605,
      "num_input_tokens_seen": 3634832,
      "step": 2685
    },
    {
      "epoch": 0.08610204212278343,
      "grad_norm": 27.0,
      "learning_rate": 4.944453518635818e-05,
      "loss": 3.5888,
      "num_input_tokens_seen": 3641360,
      "step": 2690
    },
    {
      "epoch": 0.08626208309327188,
      "grad_norm": 25.875,
      "learning_rate": 4.944247529238465e-05,
      "loss": 4.1159,
      "num_input_tokens_seen": 3648560,
      "step": 2695
    },
    {
      "epoch": 0.08642212406376032,
      "grad_norm": 41.25,
      "learning_rate": 4.944041162907427e-05,
      "loss": 4.1814,
      "num_input_tokens_seen": 3655088,
      "step": 2700
    },
    {
      "epoch": 0.08658216503424877,
      "grad_norm": 44.0,
      "learning_rate": 4.943834419674529e-05,
      "loss": 4.9668,
      "num_input_tokens_seen": 3661744,
      "step": 2705
    },
    {
      "epoch": 0.08674220600473721,
      "grad_norm": 17.875,
      "learning_rate": 4.9436272995716506e-05,
      "loss": 3.7282,
      "num_input_tokens_seen": 3668448,
      "step": 2710
    },
    {
      "epoch": 0.08690224697522565,
      "grad_norm": 26.375,
      "learning_rate": 4.943419802630735e-05,
      "loss": 3.5763,
      "num_input_tokens_seen": 3674944,
      "step": 2715
    },
    {
      "epoch": 0.0870622879457141,
      "grad_norm": 29.0,
      "learning_rate": 4.94321192888378e-05,
      "loss": 4.3304,
      "num_input_tokens_seen": 3682128,
      "step": 2720
    },
    {
      "epoch": 0.08722232891620255,
      "grad_norm": 39.75,
      "learning_rate": 4.943003678362842e-05,
      "loss": 4.6834,
      "num_input_tokens_seen": 3689152,
      "step": 2725
    },
    {
      "epoch": 0.087382369886691,
      "grad_norm": 33.75,
      "learning_rate": 4.942795051100036e-05,
      "loss": 4.4113,
      "num_input_tokens_seen": 3696144,
      "step": 2730
    },
    {
      "epoch": 0.08754241085717944,
      "grad_norm": 40.75,
      "learning_rate": 4.942586047127536e-05,
      "loss": 5.1449,
      "num_input_tokens_seen": 3702512,
      "step": 2735
    },
    {
      "epoch": 0.08770245182766788,
      "grad_norm": 44.75,
      "learning_rate": 4.942376666477571e-05,
      "loss": 3.8144,
      "num_input_tokens_seen": 3709744,
      "step": 2740
    },
    {
      "epoch": 0.08786249279815633,
      "grad_norm": 31.25,
      "learning_rate": 4.9421669091824304e-05,
      "loss": 3.8711,
      "num_input_tokens_seen": 3716176,
      "step": 2745
    },
    {
      "epoch": 0.08802253376864477,
      "grad_norm": 14.5,
      "learning_rate": 4.9419567752744634e-05,
      "loss": 3.6784,
      "num_input_tokens_seen": 3723168,
      "step": 2750
    },
    {
      "epoch": 0.08818257473913321,
      "grad_norm": 20.875,
      "learning_rate": 4.941746264786074e-05,
      "loss": 4.4276,
      "num_input_tokens_seen": 3729584,
      "step": 2755
    },
    {
      "epoch": 0.08834261570962167,
      "grad_norm": 16.75,
      "learning_rate": 4.9415353777497254e-05,
      "loss": 3.5831,
      "num_input_tokens_seen": 3736304,
      "step": 2760
    },
    {
      "epoch": 0.0885026566801101,
      "grad_norm": 38.5,
      "learning_rate": 4.9413241141979394e-05,
      "loss": 4.2757,
      "num_input_tokens_seen": 3743088,
      "step": 2765
    },
    {
      "epoch": 0.08866269765059856,
      "grad_norm": 30.0,
      "learning_rate": 4.9411124741632956e-05,
      "loss": 4.6545,
      "num_input_tokens_seen": 3749776,
      "step": 2770
    },
    {
      "epoch": 0.088822738621087,
      "grad_norm": 24.25,
      "learning_rate": 4.940900457678431e-05,
      "loss": 4.1232,
      "num_input_tokens_seen": 3756464,
      "step": 2775
    },
    {
      "epoch": 0.08898277959157544,
      "grad_norm": 37.5,
      "learning_rate": 4.9406880647760425e-05,
      "loss": 3.823,
      "num_input_tokens_seen": 3763136,
      "step": 2780
    },
    {
      "epoch": 0.08914282056206389,
      "grad_norm": 45.25,
      "learning_rate": 4.9404752954888824e-05,
      "loss": 5.0209,
      "num_input_tokens_seen": 3769680,
      "step": 2785
    },
    {
      "epoch": 0.08930286153255233,
      "grad_norm": 21.5,
      "learning_rate": 4.940262149849762e-05,
      "loss": 4.0674,
      "num_input_tokens_seen": 3776512,
      "step": 2790
    },
    {
      "epoch": 0.08946290250304077,
      "grad_norm": 45.75,
      "learning_rate": 4.9400486278915526e-05,
      "loss": 4.7249,
      "num_input_tokens_seen": 3783424,
      "step": 2795
    },
    {
      "epoch": 0.08962294347352923,
      "grad_norm": 21.0,
      "learning_rate": 4.939834729647181e-05,
      "loss": 4.3216,
      "num_input_tokens_seen": 3790208,
      "step": 2800
    },
    {
      "epoch": 0.08962294347352923,
      "eval_loss": 4.518871307373047,
      "eval_runtime": 487.6066,
      "eval_samples_per_second": 28.478,
      "eval_steps_per_second": 14.239,
      "num_input_tokens_seen": 3790208,
      "step": 2800
    },
    {
      "epoch": 0.08978298444401767,
      "grad_norm": 34.75,
      "learning_rate": 4.9396204551496326e-05,
      "loss": 3.8919,
      "num_input_tokens_seen": 3796864,
      "step": 2805
    },
    {
      "epoch": 0.08994302541450612,
      "grad_norm": 20.125,
      "learning_rate": 4.939405804431952e-05,
      "loss": 4.2581,
      "num_input_tokens_seen": 3803632,
      "step": 2810
    },
    {
      "epoch": 0.09010306638499456,
      "grad_norm": 30.125,
      "learning_rate": 4.9391907775272414e-05,
      "loss": 3.7003,
      "num_input_tokens_seen": 3810224,
      "step": 2815
    },
    {
      "epoch": 0.090263107355483,
      "grad_norm": 27.875,
      "learning_rate": 4.9389753744686604e-05,
      "loss": 3.7033,
      "num_input_tokens_seen": 3817040,
      "step": 2820
    },
    {
      "epoch": 0.09042314832597145,
      "grad_norm": 15.5,
      "learning_rate": 4.938759595289426e-05,
      "loss": 3.2828,
      "num_input_tokens_seen": 3823760,
      "step": 2825
    },
    {
      "epoch": 0.09058318929645989,
      "grad_norm": 38.0,
      "learning_rate": 4.938543440022815e-05,
      "loss": 5.1159,
      "num_input_tokens_seen": 3830864,
      "step": 2830
    },
    {
      "epoch": 0.09074323026694833,
      "grad_norm": 39.25,
      "learning_rate": 4.938326908702161e-05,
      "loss": 4.8743,
      "num_input_tokens_seen": 3837616,
      "step": 2835
    },
    {
      "epoch": 0.09090327123743679,
      "grad_norm": 31.875,
      "learning_rate": 4.9381100013608554e-05,
      "loss": 4.5856,
      "num_input_tokens_seen": 3844080,
      "step": 2840
    },
    {
      "epoch": 0.09106331220792523,
      "grad_norm": 39.75,
      "learning_rate": 4.9378927180323485e-05,
      "loss": 3.6099,
      "num_input_tokens_seen": 3850848,
      "step": 2845
    },
    {
      "epoch": 0.09122335317841368,
      "grad_norm": 35.25,
      "learning_rate": 4.937675058750148e-05,
      "loss": 4.2813,
      "num_input_tokens_seen": 3857344,
      "step": 2850
    },
    {
      "epoch": 0.09138339414890212,
      "grad_norm": 28.75,
      "learning_rate": 4.937457023547819e-05,
      "loss": 4.3579,
      "num_input_tokens_seen": 3864144,
      "step": 2855
    },
    {
      "epoch": 0.09154343511939056,
      "grad_norm": 19.25,
      "learning_rate": 4.9372386124589876e-05,
      "loss": 4.6254,
      "num_input_tokens_seen": 3870672,
      "step": 2860
    },
    {
      "epoch": 0.09170347608987901,
      "grad_norm": 43.25,
      "learning_rate": 4.937019825517333e-05,
      "loss": 4.9959,
      "num_input_tokens_seen": 3877664,
      "step": 2865
    },
    {
      "epoch": 0.09186351706036745,
      "grad_norm": 34.75,
      "learning_rate": 4.9368006627565954e-05,
      "loss": 4.4652,
      "num_input_tokens_seen": 3884336,
      "step": 2870
    },
    {
      "epoch": 0.09202355803085589,
      "grad_norm": 35.0,
      "learning_rate": 4.936581124210573e-05,
      "loss": 3.9248,
      "num_input_tokens_seen": 3891280,
      "step": 2875
    },
    {
      "epoch": 0.09218359900134435,
      "grad_norm": 93.0,
      "learning_rate": 4.9363612099131216e-05,
      "loss": 3.6839,
      "num_input_tokens_seen": 3897536,
      "step": 2880
    },
    {
      "epoch": 0.09234363997183279,
      "grad_norm": 28.625,
      "learning_rate": 4.936140919898155e-05,
      "loss": 4.4714,
      "num_input_tokens_seen": 3904624,
      "step": 2885
    },
    {
      "epoch": 0.09250368094232124,
      "grad_norm": 12.125,
      "learning_rate": 4.9359202541996426e-05,
      "loss": 4.1772,
      "num_input_tokens_seen": 3911040,
      "step": 2890
    },
    {
      "epoch": 0.09266372191280968,
      "grad_norm": 24.625,
      "learning_rate": 4.935699212851616e-05,
      "loss": 4.3633,
      "num_input_tokens_seen": 3917680,
      "step": 2895
    },
    {
      "epoch": 0.09282376288329812,
      "grad_norm": 16.625,
      "learning_rate": 4.935477795888162e-05,
      "loss": 4.5867,
      "num_input_tokens_seen": 3924112,
      "step": 2900
    },
    {
      "epoch": 0.09298380385378657,
      "grad_norm": 32.0,
      "learning_rate": 4.935256003343426e-05,
      "loss": 5.3185,
      "num_input_tokens_seen": 3930816,
      "step": 2905
    },
    {
      "epoch": 0.09314384482427501,
      "grad_norm": 26.0,
      "learning_rate": 4.93503383525161e-05,
      "loss": 4.112,
      "num_input_tokens_seen": 3937744,
      "step": 2910
    },
    {
      "epoch": 0.09330388579476345,
      "grad_norm": 25.0,
      "learning_rate": 4.934811291646977e-05,
      "loss": 4.3673,
      "num_input_tokens_seen": 3944272,
      "step": 2915
    },
    {
      "epoch": 0.0934639267652519,
      "grad_norm": 34.75,
      "learning_rate": 4.934588372563845e-05,
      "loss": 4.7257,
      "num_input_tokens_seen": 3951024,
      "step": 2920
    },
    {
      "epoch": 0.09362396773574035,
      "grad_norm": 47.5,
      "learning_rate": 4.93436507803659e-05,
      "loss": 4.691,
      "num_input_tokens_seen": 3957456,
      "step": 2925
    },
    {
      "epoch": 0.0937840087062288,
      "grad_norm": 40.25,
      "learning_rate": 4.934141408099649e-05,
      "loss": 4.3695,
      "num_input_tokens_seen": 3963856,
      "step": 2930
    },
    {
      "epoch": 0.09394404967671724,
      "grad_norm": 28.625,
      "learning_rate": 4.9339173627875135e-05,
      "loss": 4.0251,
      "num_input_tokens_seen": 3970528,
      "step": 2935
    },
    {
      "epoch": 0.09410409064720568,
      "grad_norm": 33.0,
      "learning_rate": 4.9336929421347335e-05,
      "loss": 4.1521,
      "num_input_tokens_seen": 3977056,
      "step": 2940
    },
    {
      "epoch": 0.09426413161769413,
      "grad_norm": 28.25,
      "learning_rate": 4.933468146175918e-05,
      "loss": 4.1054,
      "num_input_tokens_seen": 3983952,
      "step": 2945
    },
    {
      "epoch": 0.09442417258818257,
      "grad_norm": 24.25,
      "learning_rate": 4.933242974945734e-05,
      "loss": 4.0759,
      "num_input_tokens_seen": 3990752,
      "step": 2950
    },
    {
      "epoch": 0.09458421355867103,
      "grad_norm": 21.75,
      "learning_rate": 4.933017428478906e-05,
      "loss": 4.6373,
      "num_input_tokens_seen": 3997344,
      "step": 2955
    },
    {
      "epoch": 0.09474425452915947,
      "grad_norm": 21.625,
      "learning_rate": 4.932791506810214e-05,
      "loss": 4.4958,
      "num_input_tokens_seen": 4004016,
      "step": 2960
    },
    {
      "epoch": 0.0949042954996479,
      "grad_norm": 25.375,
      "learning_rate": 4.932565209974499e-05,
      "loss": 4.2779,
      "num_input_tokens_seen": 4010912,
      "step": 2965
    },
    {
      "epoch": 0.09506433647013636,
      "grad_norm": 18.75,
      "learning_rate": 4.93233853800666e-05,
      "loss": 4.2983,
      "num_input_tokens_seen": 4017536,
      "step": 2970
    },
    {
      "epoch": 0.0952243774406248,
      "grad_norm": 37.25,
      "learning_rate": 4.932111490941651e-05,
      "loss": 3.9336,
      "num_input_tokens_seen": 4024624,
      "step": 2975
    },
    {
      "epoch": 0.09538441841111324,
      "grad_norm": 22.75,
      "learning_rate": 4.9318840688144876e-05,
      "loss": 4.5564,
      "num_input_tokens_seen": 4031776,
      "step": 2980
    },
    {
      "epoch": 0.09554445938160169,
      "grad_norm": 45.0,
      "learning_rate": 4.9316562716602387e-05,
      "loss": 4.655,
      "num_input_tokens_seen": 4039104,
      "step": 2985
    },
    {
      "epoch": 0.09570450035209013,
      "grad_norm": 28.375,
      "learning_rate": 4.9314280995140346e-05,
      "loss": 4.1266,
      "num_input_tokens_seen": 4046048,
      "step": 2990
    },
    {
      "epoch": 0.09586454132257859,
      "grad_norm": 26.5,
      "learning_rate": 4.931199552411063e-05,
      "loss": 3.969,
      "num_input_tokens_seen": 4052512,
      "step": 2995
    },
    {
      "epoch": 0.09602458229306703,
      "grad_norm": 36.5,
      "learning_rate": 4.930970630386568e-05,
      "loss": 4.4919,
      "num_input_tokens_seen": 4059472,
      "step": 3000
    },
    {
      "epoch": 0.09602458229306703,
      "eval_loss": 4.507416248321533,
      "eval_runtime": 487.4016,
      "eval_samples_per_second": 28.49,
      "eval_steps_per_second": 14.245,
      "num_input_tokens_seen": 4059472,
      "step": 3000
    },
    {
      "epoch": 0.09618462326355547,
      "grad_norm": 48.75,
      "learning_rate": 4.9307413334758524e-05,
      "loss": 5.336,
      "num_input_tokens_seen": 4066288,
      "step": 3005
    },
    {
      "epoch": 0.09634466423404392,
      "grad_norm": 14.1875,
      "learning_rate": 4.930511661714276e-05,
      "loss": 3.65,
      "num_input_tokens_seen": 4072912,
      "step": 3010
    },
    {
      "epoch": 0.09650470520453236,
      "grad_norm": 24.0,
      "learning_rate": 4.9302816151372576e-05,
      "loss": 3.667,
      "num_input_tokens_seen": 4080160,
      "step": 3015
    },
    {
      "epoch": 0.0966647461750208,
      "grad_norm": 21.75,
      "learning_rate": 4.930051193780274e-05,
      "loss": 4.1351,
      "num_input_tokens_seen": 4086848,
      "step": 3020
    },
    {
      "epoch": 0.09682478714550925,
      "grad_norm": 28.625,
      "learning_rate": 4.929820397678858e-05,
      "loss": 4.0308,
      "num_input_tokens_seen": 4093392,
      "step": 3025
    },
    {
      "epoch": 0.09698482811599769,
      "grad_norm": 23.375,
      "learning_rate": 4.9295892268686015e-05,
      "loss": 4.4142,
      "num_input_tokens_seen": 4100160,
      "step": 3030
    },
    {
      "epoch": 0.09714486908648615,
      "grad_norm": 37.25,
      "learning_rate": 4.9293576813851536e-05,
      "loss": 3.5231,
      "num_input_tokens_seen": 4107232,
      "step": 3035
    },
    {
      "epoch": 0.09730491005697459,
      "grad_norm": 39.5,
      "learning_rate": 4.929125761264223e-05,
      "loss": 4.2197,
      "num_input_tokens_seen": 4113856,
      "step": 3040
    },
    {
      "epoch": 0.09746495102746303,
      "grad_norm": 33.0,
      "learning_rate": 4.928893466541573e-05,
      "loss": 3.7872,
      "num_input_tokens_seen": 4120528,
      "step": 3045
    },
    {
      "epoch": 0.09762499199795148,
      "grad_norm": 31.75,
      "learning_rate": 4.928660797253027e-05,
      "loss": 5.7949,
      "num_input_tokens_seen": 4127088,
      "step": 3050
    },
    {
      "epoch": 0.09778503296843992,
      "grad_norm": 52.5,
      "learning_rate": 4.928427753434467e-05,
      "loss": 4.5217,
      "num_input_tokens_seen": 4133392,
      "step": 3055
    },
    {
      "epoch": 0.09794507393892836,
      "grad_norm": 26.125,
      "learning_rate": 4.9281943351218286e-05,
      "loss": 4.6442,
      "num_input_tokens_seen": 4140320,
      "step": 3060
    },
    {
      "epoch": 0.09810511490941681,
      "grad_norm": 29.125,
      "learning_rate": 4.9279605423511095e-05,
      "loss": 4.5513,
      "num_input_tokens_seen": 4147104,
      "step": 3065
    },
    {
      "epoch": 0.09826515587990525,
      "grad_norm": 47.25,
      "learning_rate": 4.927726375158363e-05,
      "loss": 3.8334,
      "num_input_tokens_seen": 4153360,
      "step": 3070
    },
    {
      "epoch": 0.0984251968503937,
      "grad_norm": 21.125,
      "learning_rate": 4.9274918335797004e-05,
      "loss": 3.8935,
      "num_input_tokens_seen": 4160000,
      "step": 3075
    },
    {
      "epoch": 0.09858523782088215,
      "grad_norm": 29.375,
      "learning_rate": 4.927256917651292e-05,
      "loss": 4.0506,
      "num_input_tokens_seen": 4166880,
      "step": 3080
    },
    {
      "epoch": 0.09874527879137059,
      "grad_norm": 25.5,
      "learning_rate": 4.927021627409364e-05,
      "loss": 3.8714,
      "num_input_tokens_seen": 4173568,
      "step": 3085
    },
    {
      "epoch": 0.09890531976185904,
      "grad_norm": 23.5,
      "learning_rate": 4.9267859628902005e-05,
      "loss": 4.8269,
      "num_input_tokens_seen": 4180000,
      "step": 3090
    },
    {
      "epoch": 0.09906536073234748,
      "grad_norm": 27.625,
      "learning_rate": 4.9265499241301454e-05,
      "loss": 4.8675,
      "num_input_tokens_seen": 4186736,
      "step": 3095
    },
    {
      "epoch": 0.09922540170283592,
      "grad_norm": 25.0,
      "learning_rate": 4.926313511165598e-05,
      "loss": 4.6747,
      "num_input_tokens_seen": 4193344,
      "step": 3100
    },
    {
      "epoch": 0.09938544267332437,
      "grad_norm": 53.0,
      "learning_rate": 4.926076724033016e-05,
      "loss": 4.8949,
      "num_input_tokens_seen": 4200512,
      "step": 3105
    },
    {
      "epoch": 0.09954548364381281,
      "grad_norm": 25.5,
      "learning_rate": 4.9258395627689146e-05,
      "loss": 4.0335,
      "num_input_tokens_seen": 4207104,
      "step": 3110
    },
    {
      "epoch": 0.09970552461430127,
      "grad_norm": 26.25,
      "learning_rate": 4.925602027409868e-05,
      "loss": 5.2034,
      "num_input_tokens_seen": 4214240,
      "step": 3115
    },
    {
      "epoch": 0.0998655655847897,
      "grad_norm": 44.25,
      "learning_rate": 4.925364117992507e-05,
      "loss": 4.9292,
      "num_input_tokens_seen": 4220784,
      "step": 3120
    },
    {
      "epoch": 0.10002560655527815,
      "grad_norm": 26.0,
      "learning_rate": 4.92512583455352e-05,
      "loss": 4.6564,
      "num_input_tokens_seen": 4227728,
      "step": 3125
    },
    {
      "epoch": 0.1001856475257666,
      "grad_norm": 37.75,
      "learning_rate": 4.9248871771296536e-05,
      "loss": 4.5735,
      "num_input_tokens_seen": 4234544,
      "step": 3130
    },
    {
      "epoch": 0.10034568849625504,
      "grad_norm": 39.0,
      "learning_rate": 4.924648145757711e-05,
      "loss": 4.1186,
      "num_input_tokens_seen": 4241024,
      "step": 3135
    },
    {
      "epoch": 0.1005057294667435,
      "grad_norm": 37.25,
      "learning_rate": 4.924408740474554e-05,
      "loss": 3.9531,
      "num_input_tokens_seen": 4248288,
      "step": 3140
    },
    {
      "epoch": 0.10066577043723193,
      "grad_norm": 23.375,
      "learning_rate": 4.924168961317103e-05,
      "loss": 4.1656,
      "num_input_tokens_seen": 4255104,
      "step": 3145
    },
    {
      "epoch": 0.10082581140772037,
      "grad_norm": 19.25,
      "learning_rate": 4.9239288083223334e-05,
      "loss": 3.7703,
      "num_input_tokens_seen": 4262400,
      "step": 3150
    },
    {
      "epoch": 0.10098585237820883,
      "grad_norm": 30.25,
      "learning_rate": 4.9236882815272803e-05,
      "loss": 3.9127,
      "num_input_tokens_seen": 4269504,
      "step": 3155
    },
    {
      "epoch": 0.10114589334869727,
      "grad_norm": 24.625,
      "learning_rate": 4.9234473809690365e-05,
      "loss": 4.4563,
      "num_input_tokens_seen": 4276320,
      "step": 3160
    },
    {
      "epoch": 0.1013059343191857,
      "grad_norm": 31.375,
      "learning_rate": 4.923206106684752e-05,
      "loss": 4.1223,
      "num_input_tokens_seen": 4283392,
      "step": 3165
    },
    {
      "epoch": 0.10146597528967416,
      "grad_norm": 22.625,
      "learning_rate": 4.922964458711634e-05,
      "loss": 4.0271,
      "num_input_tokens_seen": 4290080,
      "step": 3170
    },
    {
      "epoch": 0.1016260162601626,
      "grad_norm": 25.75,
      "learning_rate": 4.9227224370869474e-05,
      "loss": 4.763,
      "num_input_tokens_seen": 4296688,
      "step": 3175
    },
    {
      "epoch": 0.10178605723065105,
      "grad_norm": 25.0,
      "learning_rate": 4.9224800418480155e-05,
      "loss": 5.1876,
      "num_input_tokens_seen": 4303792,
      "step": 3180
    },
    {
      "epoch": 0.10194609820113949,
      "grad_norm": 28.75,
      "learning_rate": 4.9222372730322176e-05,
      "loss": 4.4969,
      "num_input_tokens_seen": 4310784,
      "step": 3185
    },
    {
      "epoch": 0.10210613917162793,
      "grad_norm": 30.125,
      "learning_rate": 4.921994130676993e-05,
      "loss": 4.4218,
      "num_input_tokens_seen": 4317680,
      "step": 3190
    },
    {
      "epoch": 0.10226618014211639,
      "grad_norm": 35.75,
      "learning_rate": 4.9217506148198366e-05,
      "loss": 3.7545,
      "num_input_tokens_seen": 4324368,
      "step": 3195
    },
    {
      "epoch": 0.10242622111260483,
      "grad_norm": 17.125,
      "learning_rate": 4.921506725498302e-05,
      "loss": 3.3845,
      "num_input_tokens_seen": 4331088,
      "step": 3200
    },
    {
      "epoch": 0.10242622111260483,
      "eval_loss": 4.505590915679932,
      "eval_runtime": 487.6984,
      "eval_samples_per_second": 28.473,
      "eval_steps_per_second": 14.236,
      "num_input_tokens_seen": 4331088,
      "step": 3200
    },
    {
      "epoch": 0.10258626208309327,
      "grad_norm": 26.875,
      "learning_rate": 4.9212624627499994e-05,
      "loss": 3.4527,
      "num_input_tokens_seen": 4337776,
      "step": 3205
    },
    {
      "epoch": 0.10274630305358172,
      "grad_norm": 32.25,
      "learning_rate": 4.921017826612597e-05,
      "loss": 4.2486,
      "num_input_tokens_seen": 4344624,
      "step": 3210
    },
    {
      "epoch": 0.10290634402407016,
      "grad_norm": 36.75,
      "learning_rate": 4.9207728171238223e-05,
      "loss": 4.1449,
      "num_input_tokens_seen": 4351664,
      "step": 3215
    },
    {
      "epoch": 0.10306638499455861,
      "grad_norm": 29.5,
      "learning_rate": 4.920527434321458e-05,
      "loss": 3.5826,
      "num_input_tokens_seen": 4358576,
      "step": 3220
    },
    {
      "epoch": 0.10322642596504705,
      "grad_norm": 19.375,
      "learning_rate": 4.920281678243345e-05,
      "loss": 3.765,
      "num_input_tokens_seen": 4365216,
      "step": 3225
    },
    {
      "epoch": 0.10338646693553549,
      "grad_norm": 24.0,
      "learning_rate": 4.920035548927381e-05,
      "loss": 4.4335,
      "num_input_tokens_seen": 4371936,
      "step": 3230
    },
    {
      "epoch": 0.10354650790602395,
      "grad_norm": 22.5,
      "learning_rate": 4.919789046411525e-05,
      "loss": 4.8827,
      "num_input_tokens_seen": 4378640,
      "step": 3235
    },
    {
      "epoch": 0.10370654887651239,
      "grad_norm": 45.5,
      "learning_rate": 4.919542170733787e-05,
      "loss": 4.4314,
      "num_input_tokens_seen": 4385248,
      "step": 3240
    },
    {
      "epoch": 0.10386658984700083,
      "grad_norm": 24.0,
      "learning_rate": 4.919294921932242e-05,
      "loss": 4.9181,
      "num_input_tokens_seen": 4392080,
      "step": 3245
    },
    {
      "epoch": 0.10402663081748928,
      "grad_norm": 33.25,
      "learning_rate": 4.919047300045016e-05,
      "loss": 4.6706,
      "num_input_tokens_seen": 4398784,
      "step": 3250
    },
    {
      "epoch": 0.10418667178797772,
      "grad_norm": 19.625,
      "learning_rate": 4.918799305110299e-05,
      "loss": 4.7292,
      "num_input_tokens_seen": 4405376,
      "step": 3255
    },
    {
      "epoch": 0.10434671275846617,
      "grad_norm": 63.25,
      "learning_rate": 4.918550937166331e-05,
      "loss": 5.2017,
      "num_input_tokens_seen": 4412112,
      "step": 3260
    },
    {
      "epoch": 0.10450675372895461,
      "grad_norm": 24.0,
      "learning_rate": 4.918302196251415e-05,
      "loss": 3.9382,
      "num_input_tokens_seen": 4418816,
      "step": 3265
    },
    {
      "epoch": 0.10466679469944305,
      "grad_norm": 26.875,
      "learning_rate": 4.91805308240391e-05,
      "loss": 4.5427,
      "num_input_tokens_seen": 4425472,
      "step": 3270
    },
    {
      "epoch": 0.1048268356699315,
      "grad_norm": 19.875,
      "learning_rate": 4.9178035956622326e-05,
      "loss": 4.0524,
      "num_input_tokens_seen": 4432176,
      "step": 3275
    },
    {
      "epoch": 0.10498687664041995,
      "grad_norm": 12.5,
      "learning_rate": 4.917553736064857e-05,
      "loss": 4.5076,
      "num_input_tokens_seen": 4438928,
      "step": 3280
    },
    {
      "epoch": 0.10514691761090839,
      "grad_norm": 34.25,
      "learning_rate": 4.917303503650314e-05,
      "loss": 4.1736,
      "num_input_tokens_seen": 4445216,
      "step": 3285
    },
    {
      "epoch": 0.10530695858139684,
      "grad_norm": 24.25,
      "learning_rate": 4.917052898457194e-05,
      "loss": 4.3112,
      "num_input_tokens_seen": 4451760,
      "step": 3290
    },
    {
      "epoch": 0.10546699955188528,
      "grad_norm": 19.625,
      "learning_rate": 4.916801920524141e-05,
      "loss": 4.1713,
      "num_input_tokens_seen": 4458752,
      "step": 3295
    },
    {
      "epoch": 0.10562704052237373,
      "grad_norm": 22.375,
      "learning_rate": 4.916550569889862e-05,
      "loss": 3.0676,
      "num_input_tokens_seen": 4465632,
      "step": 3300
    },
    {
      "epoch": 0.10578708149286217,
      "grad_norm": 20.25,
      "learning_rate": 4.916298846593116e-05,
      "loss": 4.0078,
      "num_input_tokens_seen": 4472224,
      "step": 3305
    },
    {
      "epoch": 0.10594712246335061,
      "grad_norm": 35.75,
      "learning_rate": 4.916046750672722e-05,
      "loss": 4.1726,
      "num_input_tokens_seen": 4478624,
      "step": 3310
    },
    {
      "epoch": 0.10610716343383907,
      "grad_norm": 42.5,
      "learning_rate": 4.915794282167559e-05,
      "loss": 5.1088,
      "num_input_tokens_seen": 4485488,
      "step": 3315
    },
    {
      "epoch": 0.1062672044043275,
      "grad_norm": 29.375,
      "learning_rate": 4.915541441116558e-05,
      "loss": 3.6216,
      "num_input_tokens_seen": 4492736,
      "step": 3320
    },
    {
      "epoch": 0.10642724537481596,
      "grad_norm": 30.375,
      "learning_rate": 4.915288227558711e-05,
      "loss": 4.448,
      "num_input_tokens_seen": 4499504,
      "step": 3325
    },
    {
      "epoch": 0.1065872863453044,
      "grad_norm": 11.75,
      "learning_rate": 4.915034641533066e-05,
      "loss": 4.5616,
      "num_input_tokens_seen": 4506192,
      "step": 3330
    },
    {
      "epoch": 0.10674732731579284,
      "grad_norm": 21.5,
      "learning_rate": 4.914780683078731e-05,
      "loss": 3.7923,
      "num_input_tokens_seen": 4512880,
      "step": 3335
    },
    {
      "epoch": 0.10690736828628129,
      "grad_norm": 26.5,
      "learning_rate": 4.9145263522348695e-05,
      "loss": 5.3423,
      "num_input_tokens_seen": 4519824,
      "step": 3340
    },
    {
      "epoch": 0.10706740925676973,
      "grad_norm": 18.375,
      "learning_rate": 4.9142716490407e-05,
      "loss": 4.0557,
      "num_input_tokens_seen": 4526880,
      "step": 3345
    },
    {
      "epoch": 0.10722745022725817,
      "grad_norm": 22.625,
      "learning_rate": 4.914016573535504e-05,
      "loss": 4.6242,
      "num_input_tokens_seen": 4534304,
      "step": 3350
    },
    {
      "epoch": 0.10738749119774663,
      "grad_norm": 35.75,
      "learning_rate": 4.9137611257586154e-05,
      "loss": 5.1233,
      "num_input_tokens_seen": 4540880,
      "step": 3355
    },
    {
      "epoch": 0.10754753216823507,
      "grad_norm": 35.25,
      "learning_rate": 4.9135053057494274e-05,
      "loss": 4.4458,
      "num_input_tokens_seen": 4547520,
      "step": 3360
    },
    {
      "epoch": 0.10770757313872352,
      "grad_norm": 29.75,
      "learning_rate": 4.913249113547392e-05,
      "loss": 4.0475,
      "num_input_tokens_seen": 4554240,
      "step": 3365
    },
    {
      "epoch": 0.10786761410921196,
      "grad_norm": 19.25,
      "learning_rate": 4.912992549192016e-05,
      "loss": 3.896,
      "num_input_tokens_seen": 4561040,
      "step": 3370
    },
    {
      "epoch": 0.1080276550797004,
      "grad_norm": 29.0,
      "learning_rate": 4.9127356127228665e-05,
      "loss": 4.7645,
      "num_input_tokens_seen": 4567920,
      "step": 3375
    },
    {
      "epoch": 0.10818769605018885,
      "grad_norm": 37.0,
      "learning_rate": 4.912478304179564e-05,
      "loss": 4.2522,
      "num_input_tokens_seen": 4574592,
      "step": 3380
    },
    {
      "epoch": 0.10834773702067729,
      "grad_norm": 21.5,
      "learning_rate": 4.9122206236017896e-05,
      "loss": 3.7656,
      "num_input_tokens_seen": 4581184,
      "step": 3385
    },
    {
      "epoch": 0.10850777799116573,
      "grad_norm": 33.0,
      "learning_rate": 4.911962571029282e-05,
      "loss": 4.0757,
      "num_input_tokens_seen": 4587968,
      "step": 3390
    },
    {
      "epoch": 0.10866781896165419,
      "grad_norm": 18.75,
      "learning_rate": 4.9117041465018353e-05,
      "loss": 4.9454,
      "num_input_tokens_seen": 4594368,
      "step": 3395
    },
    {
      "epoch": 0.10882785993214263,
      "grad_norm": 20.5,
      "learning_rate": 4.911445350059302e-05,
      "loss": 4.6709,
      "num_input_tokens_seen": 4601728,
      "step": 3400
    },
    {
      "epoch": 0.10882785993214263,
      "eval_loss": 4.507922649383545,
      "eval_runtime": 487.645,
      "eval_samples_per_second": 28.476,
      "eval_steps_per_second": 14.238,
      "num_input_tokens_seen": 4601728,
      "step": 3400
    },
    {
      "epoch": 0.10898790090263108,
      "grad_norm": 50.25,
      "learning_rate": 4.9111861817415905e-05,
      "loss": 4.3739,
      "num_input_tokens_seen": 4608336,
      "step": 3405
    },
    {
      "epoch": 0.10914794187311952,
      "grad_norm": 34.75,
      "learning_rate": 4.91092664158867e-05,
      "loss": 4.9336,
      "num_input_tokens_seen": 4614848,
      "step": 3410
    },
    {
      "epoch": 0.10930798284360796,
      "grad_norm": 23.375,
      "learning_rate": 4.910666729640563e-05,
      "loss": 4.7634,
      "num_input_tokens_seen": 4622000,
      "step": 3415
    },
    {
      "epoch": 0.10946802381409641,
      "grad_norm": 36.75,
      "learning_rate": 4.910406445937353e-05,
      "loss": 4.3127,
      "num_input_tokens_seen": 4628848,
      "step": 3420
    },
    {
      "epoch": 0.10962806478458485,
      "grad_norm": 34.25,
      "learning_rate": 4.9101457905191774e-05,
      "loss": 4.7651,
      "num_input_tokens_seen": 4635952,
      "step": 3425
    },
    {
      "epoch": 0.10978810575507329,
      "grad_norm": 17.625,
      "learning_rate": 4.909884763426233e-05,
      "loss": 3.8417,
      "num_input_tokens_seen": 4642480,
      "step": 3430
    },
    {
      "epoch": 0.10994814672556175,
      "grad_norm": 26.375,
      "learning_rate": 4.9096233646987736e-05,
      "loss": 4.5304,
      "num_input_tokens_seen": 4650032,
      "step": 3435
    },
    {
      "epoch": 0.11010818769605019,
      "grad_norm": 22.625,
      "learning_rate": 4.9093615943771104e-05,
      "loss": 5.1215,
      "num_input_tokens_seen": 4657488,
      "step": 3440
    },
    {
      "epoch": 0.11026822866653864,
      "grad_norm": 35.5,
      "learning_rate": 4.909099452501611e-05,
      "loss": 5.3899,
      "num_input_tokens_seen": 4663776,
      "step": 3445
    },
    {
      "epoch": 0.11042826963702708,
      "grad_norm": 28.875,
      "learning_rate": 4.908836939112702e-05,
      "loss": 4.2126,
      "num_input_tokens_seen": 4670608,
      "step": 3450
    },
    {
      "epoch": 0.11058831060751552,
      "grad_norm": 16.375,
      "learning_rate": 4.908574054250865e-05,
      "loss": 3.5569,
      "num_input_tokens_seen": 4677520,
      "step": 3455
    },
    {
      "epoch": 0.11074835157800397,
      "grad_norm": 23.25,
      "learning_rate": 4.9083107979566414e-05,
      "loss": 4.1134,
      "num_input_tokens_seen": 4684320,
      "step": 3460
    },
    {
      "epoch": 0.11090839254849241,
      "grad_norm": 23.375,
      "learning_rate": 4.908047170270628e-05,
      "loss": 4.337,
      "num_input_tokens_seen": 4691280,
      "step": 3465
    },
    {
      "epoch": 0.11106843351898085,
      "grad_norm": 30.75,
      "learning_rate": 4.9077831712334784e-05,
      "loss": 4.2479,
      "num_input_tokens_seen": 4697648,
      "step": 3470
    },
    {
      "epoch": 0.1112284744894693,
      "grad_norm": 28.0,
      "learning_rate": 4.907518800885907e-05,
      "loss": 5.0494,
      "num_input_tokens_seen": 4704224,
      "step": 3475
    },
    {
      "epoch": 0.11138851545995775,
      "grad_norm": 15.8125,
      "learning_rate": 4.907254059268681e-05,
      "loss": 4.4881,
      "num_input_tokens_seen": 4711216,
      "step": 3480
    },
    {
      "epoch": 0.1115485564304462,
      "grad_norm": 21.125,
      "learning_rate": 4.906988946422628e-05,
      "loss": 4.4392,
      "num_input_tokens_seen": 4717984,
      "step": 3485
    },
    {
      "epoch": 0.11170859740093464,
      "grad_norm": 25.125,
      "learning_rate": 4.9067234623886315e-05,
      "loss": 3.8601,
      "num_input_tokens_seen": 4724816,
      "step": 3490
    },
    {
      "epoch": 0.11186863837142308,
      "grad_norm": 20.125,
      "learning_rate": 4.9064576072076316e-05,
      "loss": 3.5321,
      "num_input_tokens_seen": 4731584,
      "step": 3495
    },
    {
      "epoch": 0.11202867934191153,
      "grad_norm": 14.875,
      "learning_rate": 4.906191380920628e-05,
      "loss": 3.4207,
      "num_input_tokens_seen": 4738640,
      "step": 3500
    },
    {
      "epoch": 0.11218872031239997,
      "grad_norm": 20.25,
      "learning_rate": 4.905924783568675e-05,
      "loss": 4.426,
      "num_input_tokens_seen": 4745872,
      "step": 3505
    },
    {
      "epoch": 0.11234876128288843,
      "grad_norm": 14.0,
      "learning_rate": 4.905657815192886e-05,
      "loss": 4.3493,
      "num_input_tokens_seen": 4752672,
      "step": 3510
    },
    {
      "epoch": 0.11250880225337687,
      "grad_norm": 36.5,
      "learning_rate": 4.90539047583443e-05,
      "loss": 3.7217,
      "num_input_tokens_seen": 4759488,
      "step": 3515
    },
    {
      "epoch": 0.1126688432238653,
      "grad_norm": 46.75,
      "learning_rate": 4.905122765534534e-05,
      "loss": 4.437,
      "num_input_tokens_seen": 4765904,
      "step": 3520
    },
    {
      "epoch": 0.11282888419435376,
      "grad_norm": 19.75,
      "learning_rate": 4.9048546843344846e-05,
      "loss": 4.7195,
      "num_input_tokens_seen": 4773488,
      "step": 3525
    },
    {
      "epoch": 0.1129889251648422,
      "grad_norm": 37.75,
      "learning_rate": 4.9045862322756206e-05,
      "loss": 4.4697,
      "num_input_tokens_seen": 4780400,
      "step": 3530
    },
    {
      "epoch": 0.11314896613533064,
      "grad_norm": 41.0,
      "learning_rate": 4.904317409399342e-05,
      "loss": 3.7747,
      "num_input_tokens_seen": 4787152,
      "step": 3535
    },
    {
      "epoch": 0.11330900710581909,
      "grad_norm": 39.5,
      "learning_rate": 4.904048215747104e-05,
      "loss": 4.177,
      "num_input_tokens_seen": 4793664,
      "step": 3540
    },
    {
      "epoch": 0.11346904807630753,
      "grad_norm": 19.125,
      "learning_rate": 4.90377865136042e-05,
      "loss": 3.8187,
      "num_input_tokens_seen": 4803888,
      "step": 3545
    },
    {
      "epoch": 0.11362908904679599,
      "grad_norm": 49.75,
      "learning_rate": 4.90350871628086e-05,
      "loss": 4.6425,
      "num_input_tokens_seen": 4810480,
      "step": 3550
    },
    {
      "epoch": 0.11378913001728443,
      "grad_norm": 39.5,
      "learning_rate": 4.903238410550052e-05,
      "loss": 4.2115,
      "num_input_tokens_seen": 4817424,
      "step": 3555
    },
    {
      "epoch": 0.11394917098777287,
      "grad_norm": 16.625,
      "learning_rate": 4.90296773420968e-05,
      "loss": 4.6538,
      "num_input_tokens_seen": 4824032,
      "step": 3560
    },
    {
      "epoch": 0.11410921195826132,
      "grad_norm": 29.625,
      "learning_rate": 4.902696687301486e-05,
      "loss": 4.4094,
      "num_input_tokens_seen": 4830848,
      "step": 3565
    },
    {
      "epoch": 0.11426925292874976,
      "grad_norm": 31.625,
      "learning_rate": 4.902425269867268e-05,
      "loss": 5.5465,
      "num_input_tokens_seen": 4837424,
      "step": 3570
    },
    {
      "epoch": 0.1144292938992382,
      "grad_norm": 21.75,
      "learning_rate": 4.902153481948883e-05,
      "loss": 3.6289,
      "num_input_tokens_seen": 4843952,
      "step": 3575
    },
    {
      "epoch": 0.11458933486972665,
      "grad_norm": 40.5,
      "learning_rate": 4.901881323588244e-05,
      "loss": 4.3222,
      "num_input_tokens_seen": 4850512,
      "step": 3580
    },
    {
      "epoch": 0.11474937584021509,
      "grad_norm": 31.5,
      "learning_rate": 4.90160879482732e-05,
      "loss": 3.9644,
      "num_input_tokens_seen": 4857264,
      "step": 3585
    },
    {
      "epoch": 0.11490941681070355,
      "grad_norm": 20.5,
      "learning_rate": 4.9013358957081405e-05,
      "loss": 3.9531,
      "num_input_tokens_seen": 4864064,
      "step": 3590
    },
    {
      "epoch": 0.11506945778119199,
      "grad_norm": 16.375,
      "learning_rate": 4.901062626272789e-05,
      "loss": 3.9761,
      "num_input_tokens_seen": 4870512,
      "step": 3595
    },
    {
      "epoch": 0.11522949875168043,
      "grad_norm": 29.125,
      "learning_rate": 4.900788986563406e-05,
      "loss": 3.8271,
      "num_input_tokens_seen": 4877104,
      "step": 3600
    },
    {
      "epoch": 0.11522949875168043,
      "eval_loss": 4.491082668304443,
      "eval_runtime": 487.3663,
      "eval_samples_per_second": 28.492,
      "eval_steps_per_second": 14.246,
      "num_input_tokens_seen": 4877104,
      "step": 3600
    },
    {
      "epoch": 0.11538953972216888,
      "grad_norm": 26.375,
      "learning_rate": 4.9005149766221915e-05,
      "loss": 4.2601,
      "num_input_tokens_seen": 4884000,
      "step": 3605
    },
    {
      "epoch": 0.11554958069265732,
      "grad_norm": 30.125,
      "learning_rate": 4.9002405964914e-05,
      "loss": 4.0343,
      "num_input_tokens_seen": 4891184,
      "step": 3610
    },
    {
      "epoch": 0.11570962166314576,
      "grad_norm": 37.25,
      "learning_rate": 4.899965846213346e-05,
      "loss": 3.2192,
      "num_input_tokens_seen": 4897792,
      "step": 3615
    },
    {
      "epoch": 0.11586966263363421,
      "grad_norm": 31.625,
      "learning_rate": 4.899690725830399e-05,
      "loss": 5.2291,
      "num_input_tokens_seen": 4904864,
      "step": 3620
    },
    {
      "epoch": 0.11602970360412265,
      "grad_norm": 24.625,
      "learning_rate": 4.899415235384985e-05,
      "loss": 4.6103,
      "num_input_tokens_seen": 4911936,
      "step": 3625
    },
    {
      "epoch": 0.1161897445746111,
      "grad_norm": 44.75,
      "learning_rate": 4.899139374919589e-05,
      "loss": 4.3111,
      "num_input_tokens_seen": 4918368,
      "step": 3630
    },
    {
      "epoch": 0.11634978554509955,
      "grad_norm": 20.125,
      "learning_rate": 4.898863144476752e-05,
      "loss": 4.7455,
      "num_input_tokens_seen": 4925616,
      "step": 3635
    },
    {
      "epoch": 0.11650982651558799,
      "grad_norm": 29.625,
      "learning_rate": 4.898586544099072e-05,
      "loss": 4.2184,
      "num_input_tokens_seen": 4932416,
      "step": 3640
    },
    {
      "epoch": 0.11666986748607644,
      "grad_norm": 22.375,
      "learning_rate": 4.898309573829204e-05,
      "loss": 4.5121,
      "num_input_tokens_seen": 4938976,
      "step": 3645
    },
    {
      "epoch": 0.11682990845656488,
      "grad_norm": 28.0,
      "learning_rate": 4.898032233709862e-05,
      "loss": 4.0777,
      "num_input_tokens_seen": 4946112,
      "step": 3650
    },
    {
      "epoch": 0.11698994942705332,
      "grad_norm": 38.5,
      "learning_rate": 4.8977545237838123e-05,
      "loss": 4.6732,
      "num_input_tokens_seen": 4952512,
      "step": 3655
    },
    {
      "epoch": 0.11714999039754177,
      "grad_norm": 21.75,
      "learning_rate": 4.8974764440938836e-05,
      "loss": 4.0297,
      "num_input_tokens_seen": 4959008,
      "step": 3660
    },
    {
      "epoch": 0.11731003136803021,
      "grad_norm": 38.0,
      "learning_rate": 4.897197994682959e-05,
      "loss": 3.4812,
      "num_input_tokens_seen": 4965872,
      "step": 3665
    },
    {
      "epoch": 0.11747007233851867,
      "grad_norm": 30.0,
      "learning_rate": 4.8969191755939786e-05,
      "loss": 3.9493,
      "num_input_tokens_seen": 4972976,
      "step": 3670
    },
    {
      "epoch": 0.1176301133090071,
      "grad_norm": 29.25,
      "learning_rate": 4.8966399868699396e-05,
      "loss": 4.6778,
      "num_input_tokens_seen": 4979984,
      "step": 3675
    },
    {
      "epoch": 0.11779015427949555,
      "grad_norm": 23.875,
      "learning_rate": 4.8963604285538965e-05,
      "loss": 3.9687,
      "num_input_tokens_seen": 4987152,
      "step": 3680
    },
    {
      "epoch": 0.117950195249984,
      "grad_norm": 25.25,
      "learning_rate": 4.8960805006889604e-05,
      "loss": 3.5245,
      "num_input_tokens_seen": 4994912,
      "step": 3685
    },
    {
      "epoch": 0.11811023622047244,
      "grad_norm": 27.0,
      "learning_rate": 4.8958002033183004e-05,
      "loss": 3.6727,
      "num_input_tokens_seen": 5001824,
      "step": 3690
    },
    {
      "epoch": 0.11827027719096088,
      "grad_norm": 25.0,
      "learning_rate": 4.8955195364851414e-05,
      "loss": 5.1071,
      "num_input_tokens_seen": 5008992,
      "step": 3695
    },
    {
      "epoch": 0.11843031816144933,
      "grad_norm": 24.875,
      "learning_rate": 4.895238500232766e-05,
      "loss": 4.7543,
      "num_input_tokens_seen": 5015616,
      "step": 3700
    },
    {
      "epoch": 0.11859035913193777,
      "grad_norm": 49.0,
      "learning_rate": 4.8949570946045143e-05,
      "loss": 4.9461,
      "num_input_tokens_seen": 5022208,
      "step": 3705
    },
    {
      "epoch": 0.11875040010242623,
      "grad_norm": 31.875,
      "learning_rate": 4.89467531964378e-05,
      "loss": 4.1585,
      "num_input_tokens_seen": 5028480,
      "step": 3710
    },
    {
      "epoch": 0.11891044107291467,
      "grad_norm": 30.125,
      "learning_rate": 4.894393175394019e-05,
      "loss": 4.8595,
      "num_input_tokens_seen": 5035360,
      "step": 3715
    },
    {
      "epoch": 0.1190704820434031,
      "grad_norm": 33.25,
      "learning_rate": 4.8941106618987406e-05,
      "loss": 5.2811,
      "num_input_tokens_seen": 5042112,
      "step": 3720
    },
    {
      "epoch": 0.11923052301389156,
      "grad_norm": 29.375,
      "learning_rate": 4.893827779201512e-05,
      "loss": 4.0346,
      "num_input_tokens_seen": 5049072,
      "step": 3725
    },
    {
      "epoch": 0.11939056398438,
      "grad_norm": 28.125,
      "learning_rate": 4.893544527345957e-05,
      "loss": 4.3606,
      "num_input_tokens_seen": 5055440,
      "step": 3730
    },
    {
      "epoch": 0.11955060495486845,
      "grad_norm": 34.25,
      "learning_rate": 4.8932609063757563e-05,
      "loss": 3.534,
      "num_input_tokens_seen": 5062432,
      "step": 3735
    },
    {
      "epoch": 0.11971064592535689,
      "grad_norm": 35.25,
      "learning_rate": 4.8929769163346484e-05,
      "loss": 4.2832,
      "num_input_tokens_seen": 5069280,
      "step": 3740
    },
    {
      "epoch": 0.11987068689584533,
      "grad_norm": 34.5,
      "learning_rate": 4.892692557266429e-05,
      "loss": 3.9878,
      "num_input_tokens_seen": 5076016,
      "step": 3745
    },
    {
      "epoch": 0.12003072786633379,
      "grad_norm": 25.625,
      "learning_rate": 4.8924078292149464e-05,
      "loss": 4.1706,
      "num_input_tokens_seen": 5083024,
      "step": 3750
    },
    {
      "epoch": 0.12019076883682223,
      "grad_norm": 20.625,
      "learning_rate": 4.892122732224114e-05,
      "loss": 4.7869,
      "num_input_tokens_seen": 5089840,
      "step": 3755
    },
    {
      "epoch": 0.12035080980731067,
      "grad_norm": 27.875,
      "learning_rate": 4.8918372663378944e-05,
      "loss": 4.7784,
      "num_input_tokens_seen": 5096768,
      "step": 3760
    },
    {
      "epoch": 0.12051085077779912,
      "grad_norm": 24.75,
      "learning_rate": 4.89155143160031e-05,
      "loss": 4.991,
      "num_input_tokens_seen": 5104032,
      "step": 3765
    },
    {
      "epoch": 0.12067089174828756,
      "grad_norm": 44.75,
      "learning_rate": 4.891265228055441e-05,
      "loss": 4.2415,
      "num_input_tokens_seen": 5110544,
      "step": 3770
    },
    {
      "epoch": 0.12083093271877601,
      "grad_norm": 34.5,
      "learning_rate": 4.890978655747424e-05,
      "loss": 4.3551,
      "num_input_tokens_seen": 5117344,
      "step": 3775
    },
    {
      "epoch": 0.12099097368926445,
      "grad_norm": 28.25,
      "learning_rate": 4.89069171472045e-05,
      "loss": 4.0447,
      "num_input_tokens_seen": 5124064,
      "step": 3780
    },
    {
      "epoch": 0.12115101465975289,
      "grad_norm": 34.5,
      "learning_rate": 4.890404405018772e-05,
      "loss": 4.7234,
      "num_input_tokens_seen": 5130560,
      "step": 3785
    },
    {
      "epoch": 0.12131105563024135,
      "grad_norm": 34.0,
      "learning_rate": 4.8901167266866934e-05,
      "loss": 5.0481,
      "num_input_tokens_seen": 5137088,
      "step": 3790
    },
    {
      "epoch": 0.12147109660072979,
      "grad_norm": 21.375,
      "learning_rate": 4.88982867976858e-05,
      "loss": 4.1655,
      "num_input_tokens_seen": 5143792,
      "step": 3795
    },
    {
      "epoch": 0.12163113757121823,
      "grad_norm": 39.5,
      "learning_rate": 4.889540264308852e-05,
      "loss": 4.429,
      "num_input_tokens_seen": 5150656,
      "step": 3800
    },
    {
      "epoch": 0.12163113757121823,
      "eval_loss": 4.488132953643799,
      "eval_runtime": 487.6461,
      "eval_samples_per_second": 28.476,
      "eval_steps_per_second": 14.238,
      "num_input_tokens_seen": 5150656,
      "step": 3800
    },
    {
      "epoch": 0.12179117854170668,
      "grad_norm": 37.25,
      "learning_rate": 4.889251480351986e-05,
      "loss": 5.1296,
      "num_input_tokens_seen": 5157456,
      "step": 3805
    },
    {
      "epoch": 0.12195121951219512,
      "grad_norm": 35.75,
      "learning_rate": 4.888962327942517e-05,
      "loss": 3.951,
      "num_input_tokens_seen": 5164240,
      "step": 3810
    },
    {
      "epoch": 0.12211126048268357,
      "grad_norm": 31.5,
      "learning_rate": 4.8886728071250356e-05,
      "loss": 3.9952,
      "num_input_tokens_seen": 5171040,
      "step": 3815
    },
    {
      "epoch": 0.12227130145317201,
      "grad_norm": 49.75,
      "learning_rate": 4.8883829179441884e-05,
      "loss": 4.116,
      "num_input_tokens_seen": 5178016,
      "step": 3820
    },
    {
      "epoch": 0.12243134242366045,
      "grad_norm": 16.0,
      "learning_rate": 4.888092660444682e-05,
      "loss": 4.6684,
      "num_input_tokens_seen": 5184528,
      "step": 3825
    },
    {
      "epoch": 0.1225913833941489,
      "grad_norm": 45.5,
      "learning_rate": 4.887802034671276e-05,
      "loss": 3.6181,
      "num_input_tokens_seen": 5191328,
      "step": 3830
    },
    {
      "epoch": 0.12275142436463735,
      "grad_norm": 42.0,
      "learning_rate": 4.88751104066879e-05,
      "loss": 4.4272,
      "num_input_tokens_seen": 5198960,
      "step": 3835
    },
    {
      "epoch": 0.12291146533512579,
      "grad_norm": 35.5,
      "learning_rate": 4.887219678482098e-05,
      "loss": 4.4863,
      "num_input_tokens_seen": 5205536,
      "step": 3840
    },
    {
      "epoch": 0.12307150630561424,
      "grad_norm": 28.125,
      "learning_rate": 4.8869279481561316e-05,
      "loss": 4.4667,
      "num_input_tokens_seen": 5212368,
      "step": 3845
    },
    {
      "epoch": 0.12323154727610268,
      "grad_norm": 19.5,
      "learning_rate": 4.88663584973588e-05,
      "loss": 3.398,
      "num_input_tokens_seen": 5218912,
      "step": 3850
    },
    {
      "epoch": 0.12339158824659113,
      "grad_norm": 24.5,
      "learning_rate": 4.8863433832663874e-05,
      "loss": 4.4181,
      "num_input_tokens_seen": 5225408,
      "step": 3855
    },
    {
      "epoch": 0.12355162921707957,
      "grad_norm": 34.75,
      "learning_rate": 4.886050548792757e-05,
      "loss": 4.0348,
      "num_input_tokens_seen": 5232016,
      "step": 3860
    },
    {
      "epoch": 0.12371167018756801,
      "grad_norm": 28.0,
      "learning_rate": 4.8857573463601465e-05,
      "loss": 4.3582,
      "num_input_tokens_seen": 5238560,
      "step": 3865
    },
    {
      "epoch": 0.12387171115805647,
      "grad_norm": 29.5,
      "learning_rate": 4.885463776013772e-05,
      "loss": 3.9096,
      "num_input_tokens_seen": 5245664,
      "step": 3870
    },
    {
      "epoch": 0.1240317521285449,
      "grad_norm": 39.0,
      "learning_rate": 4.8851698377989056e-05,
      "loss": 5.2664,
      "num_input_tokens_seen": 5252288,
      "step": 3875
    },
    {
      "epoch": 0.12419179309903335,
      "grad_norm": 37.25,
      "learning_rate": 4.884875531760876e-05,
      "loss": 4.7875,
      "num_input_tokens_seen": 5258528,
      "step": 3880
    },
    {
      "epoch": 0.1243518340695218,
      "grad_norm": 29.0,
      "learning_rate": 4.88458085794507e-05,
      "loss": 4.4204,
      "num_input_tokens_seen": 5265264,
      "step": 3885
    },
    {
      "epoch": 0.12451187504001024,
      "grad_norm": 27.625,
      "learning_rate": 4.884285816396929e-05,
      "loss": 4.6685,
      "num_input_tokens_seen": 5272032,
      "step": 3890
    },
    {
      "epoch": 0.12467191601049869,
      "grad_norm": 20.75,
      "learning_rate": 4.8839904071619526e-05,
      "loss": 3.8431,
      "num_input_tokens_seen": 5278896,
      "step": 3895
    },
    {
      "epoch": 0.12483195698098713,
      "grad_norm": 27.75,
      "learning_rate": 4.8836946302856955e-05,
      "loss": 5.0415,
      "num_input_tokens_seen": 5285808,
      "step": 3900
    },
    {
      "epoch": 0.12499199795147557,
      "grad_norm": 16.625,
      "learning_rate": 4.8833984858137715e-05,
      "loss": 5.5456,
      "num_input_tokens_seen": 5292624,
      "step": 3905
    },
    {
      "epoch": 0.12515203892196403,
      "grad_norm": 29.125,
      "learning_rate": 4.8831019737918494e-05,
      "loss": 4.0128,
      "num_input_tokens_seen": 5300064,
      "step": 3910
    },
    {
      "epoch": 0.12531207989245247,
      "grad_norm": 41.0,
      "learning_rate": 4.882805094265655e-05,
      "loss": 4.1239,
      "num_input_tokens_seen": 5307056,
      "step": 3915
    },
    {
      "epoch": 0.1254721208629409,
      "grad_norm": 16.75,
      "learning_rate": 4.8825078472809706e-05,
      "loss": 4.6648,
      "num_input_tokens_seen": 5313680,
      "step": 3920
    },
    {
      "epoch": 0.12563216183342935,
      "grad_norm": 44.0,
      "learning_rate": 4.882210232883635e-05,
      "loss": 5.4454,
      "num_input_tokens_seen": 5320368,
      "step": 3925
    },
    {
      "epoch": 0.1257922028039178,
      "grad_norm": 39.0,
      "learning_rate": 4.881912251119546e-05,
      "loss": 3.742,
      "num_input_tokens_seen": 5326912,
      "step": 3930
    },
    {
      "epoch": 0.12595224377440625,
      "grad_norm": 18.25,
      "learning_rate": 4.881613902034654e-05,
      "loss": 4.2434,
      "num_input_tokens_seen": 5333856,
      "step": 3935
    },
    {
      "epoch": 0.1261122847448947,
      "grad_norm": 25.75,
      "learning_rate": 4.88131518567497e-05,
      "loss": 4.8378,
      "num_input_tokens_seen": 5340880,
      "step": 3940
    },
    {
      "epoch": 0.12627232571538313,
      "grad_norm": 20.875,
      "learning_rate": 4.881016102086558e-05,
      "loss": 4.3066,
      "num_input_tokens_seen": 5347424,
      "step": 3945
    },
    {
      "epoch": 0.12643236668587157,
      "grad_norm": 18.25,
      "learning_rate": 4.8807166513155425e-05,
      "loss": 3.5431,
      "num_input_tokens_seen": 5354736,
      "step": 3950
    },
    {
      "epoch": 0.12659240765636004,
      "grad_norm": 22.125,
      "learning_rate": 4.8804168334081004e-05,
      "loss": 3.8725,
      "num_input_tokens_seen": 5361584,
      "step": 3955
    },
    {
      "epoch": 0.12675244862684848,
      "grad_norm": 34.75,
      "learning_rate": 4.880116648410468e-05,
      "loss": 4.572,
      "num_input_tokens_seen": 5368368,
      "step": 3960
    },
    {
      "epoch": 0.12691248959733692,
      "grad_norm": 19.25,
      "learning_rate": 4.879816096368939e-05,
      "loss": 4.2702,
      "num_input_tokens_seen": 5375136,
      "step": 3965
    },
    {
      "epoch": 0.12707253056782536,
      "grad_norm": 21.375,
      "learning_rate": 4.879515177329861e-05,
      "loss": 4.5557,
      "num_input_tokens_seen": 5382048,
      "step": 3970
    },
    {
      "epoch": 0.1272325715383138,
      "grad_norm": 32.0,
      "learning_rate": 4.8792138913396394e-05,
      "loss": 5.4958,
      "num_input_tokens_seen": 5389056,
      "step": 3975
    },
    {
      "epoch": 0.12739261250880227,
      "grad_norm": 25.25,
      "learning_rate": 4.8789122384447374e-05,
      "loss": 4.136,
      "num_input_tokens_seen": 5395904,
      "step": 3980
    },
    {
      "epoch": 0.1275526534792907,
      "grad_norm": 27.25,
      "learning_rate": 4.878610218691673e-05,
      "loss": 4.1919,
      "num_input_tokens_seen": 5402688,
      "step": 3985
    },
    {
      "epoch": 0.12771269444977915,
      "grad_norm": 24.75,
      "learning_rate": 4.87830783212702e-05,
      "loss": 4.4171,
      "num_input_tokens_seen": 5409600,
      "step": 3990
    },
    {
      "epoch": 0.12787273542026759,
      "grad_norm": 25.75,
      "learning_rate": 4.878005078797413e-05,
      "loss": 4.7384,
      "num_input_tokens_seen": 5416384,
      "step": 3995
    },
    {
      "epoch": 0.12803277639075603,
      "grad_norm": 28.0,
      "learning_rate": 4.877701958749539e-05,
      "loss": 3.6511,
      "num_input_tokens_seen": 5422944,
      "step": 4000
    },
    {
      "epoch": 0.12803277639075603,
      "eval_loss": 4.495148658752441,
      "eval_runtime": 487.4503,
      "eval_samples_per_second": 28.487,
      "eval_steps_per_second": 14.244,
      "num_input_tokens_seen": 5422944,
      "step": 4000
    },
    {
      "epoch": 0.12819281736124447,
      "grad_norm": 26.375,
      "learning_rate": 4.877398472030142e-05,
      "loss": 4.0132,
      "num_input_tokens_seen": 5429664,
      "step": 4005
    },
    {
      "epoch": 0.12835285833173293,
      "grad_norm": 33.75,
      "learning_rate": 4.877094618686024e-05,
      "loss": 4.9132,
      "num_input_tokens_seen": 5436208,
      "step": 4010
    },
    {
      "epoch": 0.12851289930222137,
      "grad_norm": 49.5,
      "learning_rate": 4.876790398764045e-05,
      "loss": 4.9844,
      "num_input_tokens_seen": 5442704,
      "step": 4015
    },
    {
      "epoch": 0.1286729402727098,
      "grad_norm": 53.5,
      "learning_rate": 4.8764858123111167e-05,
      "loss": 4.3852,
      "num_input_tokens_seen": 5449808,
      "step": 4020
    },
    {
      "epoch": 0.12883298124319825,
      "grad_norm": 24.0,
      "learning_rate": 4.876180859374212e-05,
      "loss": 4.172,
      "num_input_tokens_seen": 5456288,
      "step": 4025
    },
    {
      "epoch": 0.1289930222136867,
      "grad_norm": 27.5,
      "learning_rate": 4.875875540000357e-05,
      "loss": 4.5449,
      "num_input_tokens_seen": 5463120,
      "step": 4030
    },
    {
      "epoch": 0.12915306318417516,
      "grad_norm": 31.25,
      "learning_rate": 4.8755698542366376e-05,
      "loss": 4.017,
      "num_input_tokens_seen": 5469552,
      "step": 4035
    },
    {
      "epoch": 0.1293131041546636,
      "grad_norm": 26.625,
      "learning_rate": 4.875263802130193e-05,
      "loss": 5.0089,
      "num_input_tokens_seen": 5475840,
      "step": 4040
    },
    {
      "epoch": 0.12947314512515204,
      "grad_norm": 45.25,
      "learning_rate": 4.8749573837282207e-05,
      "loss": 3.6751,
      "num_input_tokens_seen": 5482384,
      "step": 4045
    },
    {
      "epoch": 0.12963318609564048,
      "grad_norm": 26.375,
      "learning_rate": 4.874650599077974e-05,
      "loss": 4.6691,
      "num_input_tokens_seen": 5488544,
      "step": 4050
    },
    {
      "epoch": 0.12979322706612892,
      "grad_norm": 14.6875,
      "learning_rate": 4.874343448226764e-05,
      "loss": 3.9732,
      "num_input_tokens_seen": 5495344,
      "step": 4055
    },
    {
      "epoch": 0.1299532680366174,
      "grad_norm": 22.125,
      "learning_rate": 4.874035931221955e-05,
      "loss": 4.6924,
      "num_input_tokens_seen": 5501824,
      "step": 4060
    },
    {
      "epoch": 0.13011330900710583,
      "grad_norm": 19.75,
      "learning_rate": 4.8737280481109724e-05,
      "loss": 3.6389,
      "num_input_tokens_seen": 5508928,
      "step": 4065
    },
    {
      "epoch": 0.13027334997759427,
      "grad_norm": 24.5,
      "learning_rate": 4.873419798941294e-05,
      "loss": 4.6614,
      "num_input_tokens_seen": 5515440,
      "step": 4070
    },
    {
      "epoch": 0.1304333909480827,
      "grad_norm": 19.375,
      "learning_rate": 4.873111183760458e-05,
      "loss": 4.3339,
      "num_input_tokens_seen": 5522448,
      "step": 4075
    },
    {
      "epoch": 0.13059343191857115,
      "grad_norm": 26.625,
      "learning_rate": 4.8728022026160537e-05,
      "loss": 4.8237,
      "num_input_tokens_seen": 5528992,
      "step": 4080
    },
    {
      "epoch": 0.13075347288905959,
      "grad_norm": 46.0,
      "learning_rate": 4.872492855555732e-05,
      "loss": 3.4552,
      "num_input_tokens_seen": 5535520,
      "step": 4085
    },
    {
      "epoch": 0.13091351385954805,
      "grad_norm": 36.0,
      "learning_rate": 4.8721831426271956e-05,
      "loss": 4.035,
      "num_input_tokens_seen": 5542224,
      "step": 4090
    },
    {
      "epoch": 0.1310735548300365,
      "grad_norm": 19.5,
      "learning_rate": 4.87187306387821e-05,
      "loss": 3.525,
      "num_input_tokens_seen": 5549024,
      "step": 4095
    },
    {
      "epoch": 0.13123359580052493,
      "grad_norm": 18.25,
      "learning_rate": 4.87156261935659e-05,
      "loss": 3.9317,
      "num_input_tokens_seen": 5556032,
      "step": 4100
    },
    {
      "epoch": 0.13139363677101337,
      "grad_norm": 32.0,
      "learning_rate": 4.871251809110211e-05,
      "loss": 4.0562,
      "num_input_tokens_seen": 5562416,
      "step": 4105
    },
    {
      "epoch": 0.1315536777415018,
      "grad_norm": 22.875,
      "learning_rate": 4.8709406331870044e-05,
      "loss": 3.8557,
      "num_input_tokens_seen": 5569488,
      "step": 4110
    },
    {
      "epoch": 0.13171371871199028,
      "grad_norm": 25.875,
      "learning_rate": 4.8706290916349574e-05,
      "loss": 5.0253,
      "num_input_tokens_seen": 5575936,
      "step": 4115
    },
    {
      "epoch": 0.13187375968247872,
      "grad_norm": 27.125,
      "learning_rate": 4.8703171845021134e-05,
      "loss": 4.0505,
      "num_input_tokens_seen": 5582496,
      "step": 4120
    },
    {
      "epoch": 0.13203380065296716,
      "grad_norm": 37.75,
      "learning_rate": 4.870004911836572e-05,
      "loss": 5.773,
      "num_input_tokens_seen": 5589536,
      "step": 4125
    },
    {
      "epoch": 0.1321938416234556,
      "grad_norm": 14.875,
      "learning_rate": 4.869692273686489e-05,
      "loss": 4.6191,
      "num_input_tokens_seen": 5596304,
      "step": 4130
    },
    {
      "epoch": 0.13235388259394404,
      "grad_norm": 28.125,
      "learning_rate": 4.869379270100079e-05,
      "loss": 4.0143,
      "num_input_tokens_seen": 5603488,
      "step": 4135
    },
    {
      "epoch": 0.1325139235644325,
      "grad_norm": 29.125,
      "learning_rate": 4.86906590112561e-05,
      "loss": 4.2029,
      "num_input_tokens_seen": 5610304,
      "step": 4140
    },
    {
      "epoch": 0.13267396453492095,
      "grad_norm": 18.875,
      "learning_rate": 4.8687521668114064e-05,
      "loss": 4.5218,
      "num_input_tokens_seen": 5617136,
      "step": 4145
    },
    {
      "epoch": 0.13283400550540939,
      "grad_norm": 33.25,
      "learning_rate": 4.868438067205853e-05,
      "loss": 4.7192,
      "num_input_tokens_seen": 5623808,
      "step": 4150
    },
    {
      "epoch": 0.13299404647589783,
      "grad_norm": 40.0,
      "learning_rate": 4.8681236023573844e-05,
      "loss": 4.2523,
      "num_input_tokens_seen": 5630672,
      "step": 4155
    },
    {
      "epoch": 0.13315408744638627,
      "grad_norm": 19.25,
      "learning_rate": 4.867808772314497e-05,
      "loss": 4.6517,
      "num_input_tokens_seen": 5637152,
      "step": 4160
    },
    {
      "epoch": 0.13331412841687473,
      "grad_norm": 27.25,
      "learning_rate": 4.867493577125741e-05,
      "loss": 4.4181,
      "num_input_tokens_seen": 5643552,
      "step": 4165
    },
    {
      "epoch": 0.13347416938736317,
      "grad_norm": 23.5,
      "learning_rate": 4.867178016839725e-05,
      "loss": 3.11,
      "num_input_tokens_seen": 5650512,
      "step": 4170
    },
    {
      "epoch": 0.1336342103578516,
      "grad_norm": 30.75,
      "learning_rate": 4.8668620915051094e-05,
      "loss": 3.1405,
      "num_input_tokens_seen": 5657040,
      "step": 4175
    },
    {
      "epoch": 0.13379425132834005,
      "grad_norm": 21.0,
      "learning_rate": 4.866545801170616e-05,
      "loss": 4.0191,
      "num_input_tokens_seen": 5663744,
      "step": 4180
    },
    {
      "epoch": 0.1339542922988285,
      "grad_norm": 23.5,
      "learning_rate": 4.86622914588502e-05,
      "loss": 4.291,
      "num_input_tokens_seen": 5671200,
      "step": 4185
    },
    {
      "epoch": 0.13411433326931693,
      "grad_norm": 37.25,
      "learning_rate": 4.865912125697154e-05,
      "loss": 4.1158,
      "num_input_tokens_seen": 5678112,
      "step": 4190
    },
    {
      "epoch": 0.1342743742398054,
      "grad_norm": 21.875,
      "learning_rate": 4.865594740655907e-05,
      "loss": 3.7361,
      "num_input_tokens_seen": 5685136,
      "step": 4195
    },
    {
      "epoch": 0.13443441521029384,
      "grad_norm": 42.25,
      "learning_rate": 4.865276990810222e-05,
      "loss": 5.3519,
      "num_input_tokens_seen": 5692368,
      "step": 4200
    },
    {
      "epoch": 0.13443441521029384,
      "eval_loss": 4.485454559326172,
      "eval_runtime": 487.2794,
      "eval_samples_per_second": 28.497,
      "eval_steps_per_second": 14.248,
      "num_input_tokens_seen": 5692368,
      "step": 4200
    },
    {
      "epoch": 0.13459445618078228,
      "grad_norm": 40.5,
      "learning_rate": 4.8649588762091016e-05,
      "loss": 4.5315,
      "num_input_tokens_seen": 5698944,
      "step": 4205
    },
    {
      "epoch": 0.13475449715127072,
      "grad_norm": 18.75,
      "learning_rate": 4.8646403969016016e-05,
      "loss": 3.937,
      "num_input_tokens_seen": 5705952,
      "step": 4210
    },
    {
      "epoch": 0.13491453812175916,
      "grad_norm": 31.75,
      "learning_rate": 4.864321552936838e-05,
      "loss": 4.6013,
      "num_input_tokens_seen": 5712464,
      "step": 4215
    },
    {
      "epoch": 0.13507457909224763,
      "grad_norm": 32.75,
      "learning_rate": 4.864002344363978e-05,
      "loss": 4.7426,
      "num_input_tokens_seen": 5719392,
      "step": 4220
    },
    {
      "epoch": 0.13523462006273607,
      "grad_norm": 27.875,
      "learning_rate": 4.863682771232248e-05,
      "loss": 3.9824,
      "num_input_tokens_seen": 5726304,
      "step": 4225
    },
    {
      "epoch": 0.1353946610332245,
      "grad_norm": 31.875,
      "learning_rate": 4.8633628335909324e-05,
      "loss": 3.5724,
      "num_input_tokens_seen": 5733360,
      "step": 4230
    },
    {
      "epoch": 0.13555470200371295,
      "grad_norm": 20.625,
      "learning_rate": 4.8630425314893676e-05,
      "loss": 4.51,
      "num_input_tokens_seen": 5740432,
      "step": 4235
    },
    {
      "epoch": 0.13571474297420139,
      "grad_norm": 19.25,
      "learning_rate": 4.862721864976948e-05,
      "loss": 4.6871,
      "num_input_tokens_seen": 5747280,
      "step": 4240
    },
    {
      "epoch": 0.13587478394468985,
      "grad_norm": 41.0,
      "learning_rate": 4.862400834103125e-05,
      "loss": 3.9747,
      "num_input_tokens_seen": 5754176,
      "step": 4245
    },
    {
      "epoch": 0.1360348249151783,
      "grad_norm": 29.375,
      "learning_rate": 4.862079438917406e-05,
      "loss": 4.4841,
      "num_input_tokens_seen": 5761008,
      "step": 4250
    },
    {
      "epoch": 0.13619486588566673,
      "grad_norm": 22.5,
      "learning_rate": 4.8617576794693536e-05,
      "loss": 5.2104,
      "num_input_tokens_seen": 5767584,
      "step": 4255
    },
    {
      "epoch": 0.13635490685615517,
      "grad_norm": 24.25,
      "learning_rate": 4.8614355558085875e-05,
      "loss": 3.5065,
      "num_input_tokens_seen": 5775072,
      "step": 4260
    },
    {
      "epoch": 0.1365149478266436,
      "grad_norm": 27.625,
      "learning_rate": 4.861113067984783e-05,
      "loss": 3.6157,
      "num_input_tokens_seen": 5781792,
      "step": 4265
    },
    {
      "epoch": 0.13667498879713205,
      "grad_norm": 21.0,
      "learning_rate": 4.860790216047671e-05,
      "loss": 3.752,
      "num_input_tokens_seen": 5788176,
      "step": 4270
    },
    {
      "epoch": 0.13683502976762052,
      "grad_norm": 20.875,
      "learning_rate": 4.860467000047041e-05,
      "loss": 4.5783,
      "num_input_tokens_seen": 5794608,
      "step": 4275
    },
    {
      "epoch": 0.13699507073810896,
      "grad_norm": 28.125,
      "learning_rate": 4.860143420032737e-05,
      "loss": 2.8798,
      "num_input_tokens_seen": 5801584,
      "step": 4280
    },
    {
      "epoch": 0.1371551117085974,
      "grad_norm": 32.25,
      "learning_rate": 4.859819476054657e-05,
      "loss": 5.3983,
      "num_input_tokens_seen": 5808144,
      "step": 4285
    },
    {
      "epoch": 0.13731515267908584,
      "grad_norm": 34.75,
      "learning_rate": 4.859495168162758e-05,
      "loss": 3.7836,
      "num_input_tokens_seen": 5815072,
      "step": 4290
    },
    {
      "epoch": 0.13747519364957428,
      "grad_norm": 31.0,
      "learning_rate": 4.859170496407054e-05,
      "loss": 5.6536,
      "num_input_tokens_seen": 5822000,
      "step": 4295
    },
    {
      "epoch": 0.13763523462006275,
      "grad_norm": 37.5,
      "learning_rate": 4.8588454608376114e-05,
      "loss": 4.6008,
      "num_input_tokens_seen": 5828464,
      "step": 4300
    },
    {
      "epoch": 0.1377952755905512,
      "grad_norm": 26.25,
      "learning_rate": 4.8585200615045555e-05,
      "loss": 4.0428,
      "num_input_tokens_seen": 5835376,
      "step": 4305
    },
    {
      "epoch": 0.13795531656103963,
      "grad_norm": 35.5,
      "learning_rate": 4.8581942984580674e-05,
      "loss": 3.8567,
      "num_input_tokens_seen": 5842320,
      "step": 4310
    },
    {
      "epoch": 0.13811535753152807,
      "grad_norm": 16.375,
      "learning_rate": 4.857868171748384e-05,
      "loss": 4.5142,
      "num_input_tokens_seen": 5849168,
      "step": 4315
    },
    {
      "epoch": 0.1382753985020165,
      "grad_norm": 33.75,
      "learning_rate": 4.8575416814257976e-05,
      "loss": 4.8309,
      "num_input_tokens_seen": 5856112,
      "step": 4320
    },
    {
      "epoch": 0.13843543947250497,
      "grad_norm": 21.125,
      "learning_rate": 4.857214827540657e-05,
      "loss": 3.7462,
      "num_input_tokens_seen": 5863424,
      "step": 4325
    },
    {
      "epoch": 0.1385954804429934,
      "grad_norm": 38.5,
      "learning_rate": 4.856887610143367e-05,
      "loss": 4.9814,
      "num_input_tokens_seen": 5870592,
      "step": 4330
    },
    {
      "epoch": 0.13875552141348185,
      "grad_norm": 26.125,
      "learning_rate": 4.8565600292843896e-05,
      "loss": 4.3537,
      "num_input_tokens_seen": 5877088,
      "step": 4335
    },
    {
      "epoch": 0.1389155623839703,
      "grad_norm": 28.625,
      "learning_rate": 4.856232085014241e-05,
      "loss": 5.3693,
      "num_input_tokens_seen": 5884112,
      "step": 4340
    },
    {
      "epoch": 0.13907560335445873,
      "grad_norm": 34.5,
      "learning_rate": 4.855903777383495e-05,
      "loss": 4.9333,
      "num_input_tokens_seen": 5890640,
      "step": 4345
    },
    {
      "epoch": 0.1392356443249472,
      "grad_norm": 45.0,
      "learning_rate": 4.85557510644278e-05,
      "loss": 4.6244,
      "num_input_tokens_seen": 5897312,
      "step": 4350
    },
    {
      "epoch": 0.13939568529543564,
      "grad_norm": 32.25,
      "learning_rate": 4.855246072242782e-05,
      "loss": 5.0779,
      "num_input_tokens_seen": 5904320,
      "step": 4355
    },
    {
      "epoch": 0.13955572626592408,
      "grad_norm": 29.0,
      "learning_rate": 4.8549166748342414e-05,
      "loss": 4.5205,
      "num_input_tokens_seen": 5911072,
      "step": 4360
    },
    {
      "epoch": 0.13971576723641252,
      "grad_norm": 33.75,
      "learning_rate": 4.8545869142679556e-05,
      "loss": 4.2113,
      "num_input_tokens_seen": 5917984,
      "step": 4365
    },
    {
      "epoch": 0.13987580820690096,
      "grad_norm": 44.5,
      "learning_rate": 4.8542567905947776e-05,
      "loss": 4.0603,
      "num_input_tokens_seen": 5925872,
      "step": 4370
    },
    {
      "epoch": 0.1400358491773894,
      "grad_norm": 27.875,
      "learning_rate": 4.853926303865618e-05,
      "loss": 4.3486,
      "num_input_tokens_seen": 5932544,
      "step": 4375
    },
    {
      "epoch": 0.14019589014787787,
      "grad_norm": 31.625,
      "learning_rate": 4.853595454131441e-05,
      "loss": 4.5498,
      "num_input_tokens_seen": 5939008,
      "step": 4380
    },
    {
      "epoch": 0.1403559311183663,
      "grad_norm": 41.25,
      "learning_rate": 4.8532642414432674e-05,
      "loss": 4.5614,
      "num_input_tokens_seen": 5945824,
      "step": 4385
    },
    {
      "epoch": 0.14051597208885475,
      "grad_norm": 20.875,
      "learning_rate": 4.8529326658521754e-05,
      "loss": 4.3129,
      "num_input_tokens_seen": 5952384,
      "step": 4390
    },
    {
      "epoch": 0.14067601305934319,
      "grad_norm": 41.5,
      "learning_rate": 4.8526007274092965e-05,
      "loss": 4.0418,
      "num_input_tokens_seen": 5958864,
      "step": 4395
    },
    {
      "epoch": 0.14083605402983163,
      "grad_norm": 32.75,
      "learning_rate": 4.852268426165822e-05,
      "loss": 4.2255,
      "num_input_tokens_seen": 5965440,
      "step": 4400
    },
    {
      "epoch": 0.14083605402983163,
      "eval_loss": 4.498112201690674,
      "eval_runtime": 487.6837,
      "eval_samples_per_second": 28.473,
      "eval_steps_per_second": 14.237,
      "num_input_tokens_seen": 5965440,
      "step": 4400
    },
    {
      "epoch": 0.1409960950003201,
      "grad_norm": 24.0,
      "learning_rate": 4.851935762172995e-05,
      "loss": 3.2734,
      "num_input_tokens_seen": 5972464,
      "step": 4405
    },
    {
      "epoch": 0.14115613597080853,
      "grad_norm": 24.5,
      "learning_rate": 4.8516027354821175e-05,
      "loss": 4.23,
      "num_input_tokens_seen": 5979344,
      "step": 4410
    },
    {
      "epoch": 0.14131617694129697,
      "grad_norm": 37.25,
      "learning_rate": 4.851269346144546e-05,
      "loss": 4.1418,
      "num_input_tokens_seen": 5986384,
      "step": 4415
    },
    {
      "epoch": 0.1414762179117854,
      "grad_norm": 21.375,
      "learning_rate": 4.850935594211693e-05,
      "loss": 3.6272,
      "num_input_tokens_seen": 5993840,
      "step": 4420
    },
    {
      "epoch": 0.14163625888227385,
      "grad_norm": 29.75,
      "learning_rate": 4.850601479735029e-05,
      "loss": 3.6661,
      "num_input_tokens_seen": 6000672,
      "step": 4425
    },
    {
      "epoch": 0.14179629985276232,
      "grad_norm": 34.5,
      "learning_rate": 4.850267002766076e-05,
      "loss": 4.6692,
      "num_input_tokens_seen": 6007408,
      "step": 4430
    },
    {
      "epoch": 0.14195634082325076,
      "grad_norm": 22.375,
      "learning_rate": 4.849932163356417e-05,
      "loss": 3.9078,
      "num_input_tokens_seen": 6014096,
      "step": 4435
    },
    {
      "epoch": 0.1421163817937392,
      "grad_norm": 37.5,
      "learning_rate": 4.8495969615576864e-05,
      "loss": 3.8658,
      "num_input_tokens_seen": 6020592,
      "step": 4440
    },
    {
      "epoch": 0.14227642276422764,
      "grad_norm": 30.375,
      "learning_rate": 4.849261397421577e-05,
      "loss": 5.0738,
      "num_input_tokens_seen": 6026976,
      "step": 4445
    },
    {
      "epoch": 0.14243646373471608,
      "grad_norm": 25.375,
      "learning_rate": 4.848925470999839e-05,
      "loss": 4.8309,
      "num_input_tokens_seen": 6033648,
      "step": 4450
    },
    {
      "epoch": 0.14259650470520452,
      "grad_norm": 35.25,
      "learning_rate": 4.848589182344273e-05,
      "loss": 4.8045,
      "num_input_tokens_seen": 6040384,
      "step": 4455
    },
    {
      "epoch": 0.142756545675693,
      "grad_norm": 34.25,
      "learning_rate": 4.848252531506742e-05,
      "loss": 5.0988,
      "num_input_tokens_seen": 6047440,
      "step": 4460
    },
    {
      "epoch": 0.14291658664618143,
      "grad_norm": 22.625,
      "learning_rate": 4.847915518539161e-05,
      "loss": 3.7277,
      "num_input_tokens_seen": 6054400,
      "step": 4465
    },
    {
      "epoch": 0.14307662761666987,
      "grad_norm": 35.75,
      "learning_rate": 4.847578143493501e-05,
      "loss": 4.2931,
      "num_input_tokens_seen": 6061264,
      "step": 4470
    },
    {
      "epoch": 0.1432366685871583,
      "grad_norm": 34.75,
      "learning_rate": 4.847240406421789e-05,
      "loss": 3.9602,
      "num_input_tokens_seen": 6067888,
      "step": 4475
    },
    {
      "epoch": 0.14339670955764675,
      "grad_norm": 34.25,
      "learning_rate": 4.84690230737611e-05,
      "loss": 4.839,
      "num_input_tokens_seen": 6074640,
      "step": 4480
    },
    {
      "epoch": 0.1435567505281352,
      "grad_norm": 18.875,
      "learning_rate": 4.846563846408602e-05,
      "loss": 4.0363,
      "num_input_tokens_seen": 6080992,
      "step": 4485
    },
    {
      "epoch": 0.14371679149862365,
      "grad_norm": 30.75,
      "learning_rate": 4.84622502357146e-05,
      "loss": 4.2636,
      "num_input_tokens_seen": 6088160,
      "step": 4490
    },
    {
      "epoch": 0.1438768324691121,
      "grad_norm": 51.25,
      "learning_rate": 4.8458858389169345e-05,
      "loss": 4.9206,
      "num_input_tokens_seen": 6095008,
      "step": 4495
    },
    {
      "epoch": 0.14403687343960053,
      "grad_norm": 27.125,
      "learning_rate": 4.8455462924973334e-05,
      "loss": 4.6437,
      "num_input_tokens_seen": 6102608,
      "step": 4500
    },
    {
      "epoch": 0.14419691441008897,
      "grad_norm": 31.0,
      "learning_rate": 4.845206384365018e-05,
      "loss": 3.7585,
      "num_input_tokens_seen": 6109216,
      "step": 4505
    },
    {
      "epoch": 0.14435695538057744,
      "grad_norm": 23.875,
      "learning_rate": 4.844866114572405e-05,
      "loss": 4.8846,
      "num_input_tokens_seen": 6115760,
      "step": 4510
    },
    {
      "epoch": 0.14451699635106588,
      "grad_norm": 30.75,
      "learning_rate": 4.8445254831719706e-05,
      "loss": 3.49,
      "num_input_tokens_seen": 6122352,
      "step": 4515
    },
    {
      "epoch": 0.14467703732155432,
      "grad_norm": 18.625,
      "learning_rate": 4.8441844902162434e-05,
      "loss": 3.5193,
      "num_input_tokens_seen": 6129024,
      "step": 4520
    },
    {
      "epoch": 0.14483707829204276,
      "grad_norm": 17.875,
      "learning_rate": 4.843843135757809e-05,
      "loss": 3.92,
      "num_input_tokens_seen": 6135952,
      "step": 4525
    },
    {
      "epoch": 0.1449971192625312,
      "grad_norm": 25.125,
      "learning_rate": 4.843501419849308e-05,
      "loss": 4.2074,
      "num_input_tokens_seen": 6142656,
      "step": 4530
    },
    {
      "epoch": 0.14515716023301967,
      "grad_norm": 22.875,
      "learning_rate": 4.8431593425434386e-05,
      "loss": 3.8866,
      "num_input_tokens_seen": 6149904,
      "step": 4535
    },
    {
      "epoch": 0.1453172012035081,
      "grad_norm": 30.375,
      "learning_rate": 4.8428169038929526e-05,
      "loss": 3.4904,
      "num_input_tokens_seen": 6156672,
      "step": 4540
    },
    {
      "epoch": 0.14547724217399655,
      "grad_norm": 46.0,
      "learning_rate": 4.8424741039506575e-05,
      "loss": 4.3155,
      "num_input_tokens_seen": 6163296,
      "step": 4545
    },
    {
      "epoch": 0.14563728314448499,
      "grad_norm": 19.375,
      "learning_rate": 4.842130942769419e-05,
      "loss": 4.1846,
      "num_input_tokens_seen": 6169952,
      "step": 4550
    },
    {
      "epoch": 0.14579732411497343,
      "grad_norm": 24.125,
      "learning_rate": 4.841787420402156e-05,
      "loss": 4.0395,
      "num_input_tokens_seen": 6176512,
      "step": 4555
    },
    {
      "epoch": 0.14595736508546187,
      "grad_norm": 31.625,
      "learning_rate": 4.841443536901844e-05,
      "loss": 4.445,
      "num_input_tokens_seen": 6183248,
      "step": 4560
    },
    {
      "epoch": 0.14611740605595033,
      "grad_norm": 28.125,
      "learning_rate": 4.841099292321514e-05,
      "loss": 3.7798,
      "num_input_tokens_seen": 6189872,
      "step": 4565
    },
    {
      "epoch": 0.14627744702643877,
      "grad_norm": 21.625,
      "learning_rate": 4.8407546867142525e-05,
      "loss": 3.9995,
      "num_input_tokens_seen": 6196512,
      "step": 4570
    },
    {
      "epoch": 0.1464374879969272,
      "grad_norm": 12.4375,
      "learning_rate": 4.840409720133203e-05,
      "loss": 3.8552,
      "num_input_tokens_seen": 6203424,
      "step": 4575
    },
    {
      "epoch": 0.14659752896741565,
      "grad_norm": 34.75,
      "learning_rate": 4.8400643926315634e-05,
      "loss": 4.0488,
      "num_input_tokens_seen": 6210320,
      "step": 4580
    },
    {
      "epoch": 0.1467575699379041,
      "grad_norm": 33.5,
      "learning_rate": 4.839718704262587e-05,
      "loss": 3.8226,
      "num_input_tokens_seen": 6217216,
      "step": 4585
    },
    {
      "epoch": 0.14691761090839256,
      "grad_norm": 20.25,
      "learning_rate": 4.839372655079585e-05,
      "loss": 4.0209,
      "num_input_tokens_seen": 6223952,
      "step": 4590
    },
    {
      "epoch": 0.147077651878881,
      "grad_norm": 21.125,
      "learning_rate": 4.83902624513592e-05,
      "loss": 4.1914,
      "num_input_tokens_seen": 6230544,
      "step": 4595
    },
    {
      "epoch": 0.14723769284936944,
      "grad_norm": 27.5,
      "learning_rate": 4.838679474485014e-05,
      "loss": 4.9348,
      "num_input_tokens_seen": 6237632,
      "step": 4600
    },
    {
      "epoch": 0.14723769284936944,
      "eval_loss": 4.49765157699585,
      "eval_runtime": 487.9194,
      "eval_samples_per_second": 28.46,
      "eval_steps_per_second": 14.23,
      "num_input_tokens_seen": 6237632,
      "step": 4600
    },
    {
      "epoch": 0.14739773381985788,
      "grad_norm": 43.25,
      "learning_rate": 4.838332343180343e-05,
      "loss": 5.1507,
      "num_input_tokens_seen": 6244192,
      "step": 4605
    },
    {
      "epoch": 0.14755777479034632,
      "grad_norm": 14.9375,
      "learning_rate": 4.83798485127544e-05,
      "loss": 3.431,
      "num_input_tokens_seen": 6251280,
      "step": 4610
    },
    {
      "epoch": 0.1477178157608348,
      "grad_norm": 35.0,
      "learning_rate": 4.837636998823892e-05,
      "loss": 4.8992,
      "num_input_tokens_seen": 6257680,
      "step": 4615
    },
    {
      "epoch": 0.14787785673132323,
      "grad_norm": 29.875,
      "learning_rate": 4.8372887858793414e-05,
      "loss": 4.7823,
      "num_input_tokens_seen": 6264480,
      "step": 4620
    },
    {
      "epoch": 0.14803789770181167,
      "grad_norm": 35.0,
      "learning_rate": 4.836940212495489e-05,
      "loss": 4.2364,
      "num_input_tokens_seen": 6271296,
      "step": 4625
    },
    {
      "epoch": 0.1481979386723001,
      "grad_norm": 23.0,
      "learning_rate": 4.836591278726087e-05,
      "loss": 4.1741,
      "num_input_tokens_seen": 6278528,
      "step": 4630
    },
    {
      "epoch": 0.14835797964278855,
      "grad_norm": 17.875,
      "learning_rate": 4.836241984624947e-05,
      "loss": 5.09,
      "num_input_tokens_seen": 6284928,
      "step": 4635
    },
    {
      "epoch": 0.14851802061327699,
      "grad_norm": 20.5,
      "learning_rate": 4.8358923302459336e-05,
      "loss": 3.345,
      "num_input_tokens_seen": 6291488,
      "step": 4640
    },
    {
      "epoch": 0.14867806158376545,
      "grad_norm": 32.75,
      "learning_rate": 4.835542315642968e-05,
      "loss": 4.4556,
      "num_input_tokens_seen": 6298368,
      "step": 4645
    },
    {
      "epoch": 0.1488381025542539,
      "grad_norm": 24.625,
      "learning_rate": 4.8351919408700274e-05,
      "loss": 4.4239,
      "num_input_tokens_seen": 6304976,
      "step": 4650
    },
    {
      "epoch": 0.14899814352474233,
      "grad_norm": 15.375,
      "learning_rate": 4.834841205981144e-05,
      "loss": 3.8284,
      "num_input_tokens_seen": 6311376,
      "step": 4655
    },
    {
      "epoch": 0.14915818449523077,
      "grad_norm": 43.25,
      "learning_rate": 4.8344901110304054e-05,
      "loss": 5.0477,
      "num_input_tokens_seen": 6318608,
      "step": 4660
    },
    {
      "epoch": 0.1493182254657192,
      "grad_norm": 26.875,
      "learning_rate": 4.8341386560719534e-05,
      "loss": 4.0157,
      "num_input_tokens_seen": 6325008,
      "step": 4665
    },
    {
      "epoch": 0.14947826643620768,
      "grad_norm": 30.5,
      "learning_rate": 4.833786841159989e-05,
      "loss": 4.6216,
      "num_input_tokens_seen": 6331568,
      "step": 4670
    },
    {
      "epoch": 0.14963830740669612,
      "grad_norm": 30.875,
      "learning_rate": 4.833434666348765e-05,
      "loss": 4.8461,
      "num_input_tokens_seen": 6338448,
      "step": 4675
    },
    {
      "epoch": 0.14979834837718456,
      "grad_norm": 25.625,
      "learning_rate": 4.833082131692592e-05,
      "loss": 3.6666,
      "num_input_tokens_seen": 6344976,
      "step": 4680
    },
    {
      "epoch": 0.149958389347673,
      "grad_norm": 26.0,
      "learning_rate": 4.832729237245835e-05,
      "loss": 4.2447,
      "num_input_tokens_seen": 6351360,
      "step": 4685
    },
    {
      "epoch": 0.15011843031816144,
      "grad_norm": 50.5,
      "learning_rate": 4.8323759830629145e-05,
      "loss": 4.6556,
      "num_input_tokens_seen": 6357952,
      "step": 4690
    },
    {
      "epoch": 0.1502784712886499,
      "grad_norm": 32.5,
      "learning_rate": 4.8320223691983066e-05,
      "loss": 4.4483,
      "num_input_tokens_seen": 6365088,
      "step": 4695
    },
    {
      "epoch": 0.15043851225913835,
      "grad_norm": 18.5,
      "learning_rate": 4.831668395706544e-05,
      "loss": 4.9799,
      "num_input_tokens_seen": 6371312,
      "step": 4700
    },
    {
      "epoch": 0.15059855322962679,
      "grad_norm": 42.75,
      "learning_rate": 4.8313140626422125e-05,
      "loss": 4.7922,
      "num_input_tokens_seen": 6377840,
      "step": 4705
    },
    {
      "epoch": 0.15075859420011523,
      "grad_norm": 25.75,
      "learning_rate": 4.830959370059956e-05,
      "loss": 4.5168,
      "num_input_tokens_seen": 6384496,
      "step": 4710
    },
    {
      "epoch": 0.15091863517060367,
      "grad_norm": 23.625,
      "learning_rate": 4.830604318014472e-05,
      "loss": 3.7427,
      "num_input_tokens_seen": 6391120,
      "step": 4715
    },
    {
      "epoch": 0.15107867614109213,
      "grad_norm": 21.5,
      "learning_rate": 4.830248906560514e-05,
      "loss": 4.0071,
      "num_input_tokens_seen": 6398256,
      "step": 4720
    },
    {
      "epoch": 0.15123871711158057,
      "grad_norm": 27.125,
      "learning_rate": 4.829893135752891e-05,
      "loss": 4.324,
      "num_input_tokens_seen": 6405552,
      "step": 4725
    },
    {
      "epoch": 0.151398758082069,
      "grad_norm": 16.125,
      "learning_rate": 4.829537005646466e-05,
      "loss": 3.9937,
      "num_input_tokens_seen": 6412336,
      "step": 4730
    },
    {
      "epoch": 0.15155879905255745,
      "grad_norm": 41.25,
      "learning_rate": 4.8291805162961615e-05,
      "loss": 4.2059,
      "num_input_tokens_seen": 6419248,
      "step": 4735
    },
    {
      "epoch": 0.1517188400230459,
      "grad_norm": 26.125,
      "learning_rate": 4.82882366775695e-05,
      "loss": 3.7523,
      "num_input_tokens_seen": 6425808,
      "step": 4740
    },
    {
      "epoch": 0.15187888099353433,
      "grad_norm": 30.125,
      "learning_rate": 4.828466460083864e-05,
      "loss": 3.9734,
      "num_input_tokens_seen": 6432432,
      "step": 4745
    },
    {
      "epoch": 0.1520389219640228,
      "grad_norm": 42.5,
      "learning_rate": 4.8281088933319877e-05,
      "loss": 5.0478,
      "num_input_tokens_seen": 6438960,
      "step": 4750
    },
    {
      "epoch": 0.15219896293451124,
      "grad_norm": 27.875,
      "learning_rate": 4.827750967556464e-05,
      "loss": 4.7265,
      "num_input_tokens_seen": 6445568,
      "step": 4755
    },
    {
      "epoch": 0.15235900390499968,
      "grad_norm": 28.875,
      "learning_rate": 4.827392682812488e-05,
      "loss": 5.2109,
      "num_input_tokens_seen": 6452096,
      "step": 4760
    },
    {
      "epoch": 0.15251904487548812,
      "grad_norm": 38.25,
      "learning_rate": 4.827034039155312e-05,
      "loss": 4.7692,
      "num_input_tokens_seen": 6458848,
      "step": 4765
    },
    {
      "epoch": 0.15267908584597656,
      "grad_norm": 47.75,
      "learning_rate": 4.8266750366402445e-05,
      "loss": 4.3876,
      "num_input_tokens_seen": 6465440,
      "step": 4770
    },
    {
      "epoch": 0.15283912681646503,
      "grad_norm": 26.875,
      "learning_rate": 4.8263156753226476e-05,
      "loss": 5.1206,
      "num_input_tokens_seen": 6471984,
      "step": 4775
    },
    {
      "epoch": 0.15299916778695347,
      "grad_norm": 28.0,
      "learning_rate": 4.8259559552579394e-05,
      "loss": 4.412,
      "num_input_tokens_seen": 6478720,
      "step": 4780
    },
    {
      "epoch": 0.1531592087574419,
      "grad_norm": 24.5,
      "learning_rate": 4.825595876501593e-05,
      "loss": 3.9805,
      "num_input_tokens_seen": 6485728,
      "step": 4785
    },
    {
      "epoch": 0.15331924972793035,
      "grad_norm": 27.875,
      "learning_rate": 4.825235439109137e-05,
      "loss": 3.9733,
      "num_input_tokens_seen": 6492528,
      "step": 4790
    },
    {
      "epoch": 0.15347929069841879,
      "grad_norm": 39.25,
      "learning_rate": 4.824874643136156e-05,
      "loss": 4.6847,
      "num_input_tokens_seen": 6499648,
      "step": 4795
    },
    {
      "epoch": 0.15363933166890725,
      "grad_norm": 31.0,
      "learning_rate": 4.824513488638288e-05,
      "loss": 3.4591,
      "num_input_tokens_seen": 6506256,
      "step": 4800
    },
    {
      "epoch": 0.15363933166890725,
      "eval_loss": 4.5017242431640625,
      "eval_runtime": 487.0535,
      "eval_samples_per_second": 28.51,
      "eval_steps_per_second": 14.255,
      "num_input_tokens_seen": 6506256,
      "step": 4800
    },
    {
      "epoch": 0.1537993726393957,
      "grad_norm": 27.25,
      "learning_rate": 4.8241519756712293e-05,
      "loss": 3.8674,
      "num_input_tokens_seen": 6513024,
      "step": 4805
    },
    {
      "epoch": 0.15395941360988413,
      "grad_norm": 28.875,
      "learning_rate": 4.8237901042907285e-05,
      "loss": 4.3207,
      "num_input_tokens_seen": 6520368,
      "step": 4810
    },
    {
      "epoch": 0.15411945458037257,
      "grad_norm": 27.875,
      "learning_rate": 4.823427874552591e-05,
      "loss": 4.29,
      "num_input_tokens_seen": 6527136,
      "step": 4815
    },
    {
      "epoch": 0.154279495550861,
      "grad_norm": 45.5,
      "learning_rate": 4.823065286512677e-05,
      "loss": 4.3932,
      "num_input_tokens_seen": 6533760,
      "step": 4820
    },
    {
      "epoch": 0.15443953652134945,
      "grad_norm": 24.625,
      "learning_rate": 4.8227023402269025e-05,
      "loss": 3.6026,
      "num_input_tokens_seen": 6540624,
      "step": 4825
    },
    {
      "epoch": 0.15459957749183792,
      "grad_norm": 23.75,
      "learning_rate": 4.822339035751239e-05,
      "loss": 3.8808,
      "num_input_tokens_seen": 6547392,
      "step": 4830
    },
    {
      "epoch": 0.15475961846232636,
      "grad_norm": 29.625,
      "learning_rate": 4.8219753731417104e-05,
      "loss": 4.6501,
      "num_input_tokens_seen": 6554192,
      "step": 4835
    },
    {
      "epoch": 0.1549196594328148,
      "grad_norm": 32.25,
      "learning_rate": 4.821611352454401e-05,
      "loss": 4.6093,
      "num_input_tokens_seen": 6561024,
      "step": 4840
    },
    {
      "epoch": 0.15507970040330324,
      "grad_norm": 28.625,
      "learning_rate": 4.8212469737454444e-05,
      "loss": 4.594,
      "num_input_tokens_seen": 6567792,
      "step": 4845
    },
    {
      "epoch": 0.15523974137379168,
      "grad_norm": 18.25,
      "learning_rate": 4.820882237071035e-05,
      "loss": 4.4441,
      "num_input_tokens_seen": 6574400,
      "step": 4850
    },
    {
      "epoch": 0.15539978234428015,
      "grad_norm": 38.25,
      "learning_rate": 4.820517142487417e-05,
      "loss": 4.2989,
      "num_input_tokens_seen": 6581408,
      "step": 4855
    },
    {
      "epoch": 0.15555982331476859,
      "grad_norm": 22.125,
      "learning_rate": 4.8201516900508956e-05,
      "loss": 4.1149,
      "num_input_tokens_seen": 6587904,
      "step": 4860
    },
    {
      "epoch": 0.15571986428525703,
      "grad_norm": 33.75,
      "learning_rate": 4.819785879817827e-05,
      "loss": 4.9501,
      "num_input_tokens_seen": 6594560,
      "step": 4865
    },
    {
      "epoch": 0.15587990525574547,
      "grad_norm": 33.25,
      "learning_rate": 4.8194197118446226e-05,
      "loss": 4.203,
      "num_input_tokens_seen": 6601312,
      "step": 4870
    },
    {
      "epoch": 0.1560399462262339,
      "grad_norm": 34.0,
      "learning_rate": 4.819053186187752e-05,
      "loss": 3.7988,
      "num_input_tokens_seen": 6607872,
      "step": 4875
    },
    {
      "epoch": 0.15619998719672237,
      "grad_norm": 24.25,
      "learning_rate": 4.818686302903736e-05,
      "loss": 5.1559,
      "num_input_tokens_seen": 6614352,
      "step": 4880
    },
    {
      "epoch": 0.1563600281672108,
      "grad_norm": 24.0,
      "learning_rate": 4.818319062049154e-05,
      "loss": 3.8431,
      "num_input_tokens_seen": 6621424,
      "step": 4885
    },
    {
      "epoch": 0.15652006913769925,
      "grad_norm": 18.5,
      "learning_rate": 4.817951463680639e-05,
      "loss": 4.1553,
      "num_input_tokens_seen": 6629040,
      "step": 4890
    },
    {
      "epoch": 0.1566801101081877,
      "grad_norm": 19.5,
      "learning_rate": 4.817583507854879e-05,
      "loss": 4.4823,
      "num_input_tokens_seen": 6635776,
      "step": 4895
    },
    {
      "epoch": 0.15684015107867613,
      "grad_norm": 15.5625,
      "learning_rate": 4.817215194628617e-05,
      "loss": 4.0685,
      "num_input_tokens_seen": 6643008,
      "step": 4900
    },
    {
      "epoch": 0.1570001920491646,
      "grad_norm": 26.375,
      "learning_rate": 4.816846524058653e-05,
      "loss": 4.6489,
      "num_input_tokens_seen": 6649872,
      "step": 4905
    },
    {
      "epoch": 0.15716023301965304,
      "grad_norm": 34.75,
      "learning_rate": 4.816477496201839e-05,
      "loss": 4.3837,
      "num_input_tokens_seen": 6656640,
      "step": 4910
    },
    {
      "epoch": 0.15732027399014148,
      "grad_norm": 26.0,
      "learning_rate": 4.8161081111150845e-05,
      "loss": 5.3152,
      "num_input_tokens_seen": 6663264,
      "step": 4915
    },
    {
      "epoch": 0.15748031496062992,
      "grad_norm": 23.5,
      "learning_rate": 4.815738368855354e-05,
      "loss": 3.7981,
      "num_input_tokens_seen": 6669888,
      "step": 4920
    },
    {
      "epoch": 0.15764035593111836,
      "grad_norm": 17.75,
      "learning_rate": 4.815368269479664e-05,
      "loss": 3.7073,
      "num_input_tokens_seen": 6676672,
      "step": 4925
    },
    {
      "epoch": 0.1578003969016068,
      "grad_norm": 37.25,
      "learning_rate": 4.814997813045092e-05,
      "loss": 4.9449,
      "num_input_tokens_seen": 6683744,
      "step": 4930
    },
    {
      "epoch": 0.15796043787209527,
      "grad_norm": 21.25,
      "learning_rate": 4.814626999608764e-05,
      "loss": 3.5363,
      "num_input_tokens_seen": 6690480,
      "step": 4935
    },
    {
      "epoch": 0.1581204788425837,
      "grad_norm": 55.5,
      "learning_rate": 4.814255829227865e-05,
      "loss": 5.5486,
      "num_input_tokens_seen": 6696880,
      "step": 4940
    },
    {
      "epoch": 0.15828051981307215,
      "grad_norm": 37.25,
      "learning_rate": 4.813884301959635e-05,
      "loss": 4.2208,
      "num_input_tokens_seen": 6704016,
      "step": 4945
    },
    {
      "epoch": 0.15844056078356059,
      "grad_norm": 19.375,
      "learning_rate": 4.813512417861368e-05,
      "loss": 4.5549,
      "num_input_tokens_seen": 6711008,
      "step": 4950
    },
    {
      "epoch": 0.15860060175404903,
      "grad_norm": 46.75,
      "learning_rate": 4.813140176990411e-05,
      "loss": 4.318,
      "num_input_tokens_seen": 6717776,
      "step": 4955
    },
    {
      "epoch": 0.1587606427245375,
      "grad_norm": 37.75,
      "learning_rate": 4.8127675794041714e-05,
      "loss": 4.4552,
      "num_input_tokens_seen": 6724304,
      "step": 4960
    },
    {
      "epoch": 0.15892068369502593,
      "grad_norm": 41.25,
      "learning_rate": 4.812394625160107e-05,
      "loss": 4.8554,
      "num_input_tokens_seen": 6731424,
      "step": 4965
    },
    {
      "epoch": 0.15908072466551437,
      "grad_norm": 33.0,
      "learning_rate": 4.812021314315732e-05,
      "loss": 4.0083,
      "num_input_tokens_seen": 6738032,
      "step": 4970
    },
    {
      "epoch": 0.1592407656360028,
      "grad_norm": 24.75,
      "learning_rate": 4.811647646928616e-05,
      "loss": 3.8059,
      "num_input_tokens_seen": 6745280,
      "step": 4975
    },
    {
      "epoch": 0.15940080660649125,
      "grad_norm": 49.5,
      "learning_rate": 4.8112736230563814e-05,
      "loss": 3.5211,
      "num_input_tokens_seen": 6752272,
      "step": 4980
    },
    {
      "epoch": 0.15956084757697972,
      "grad_norm": 18.375,
      "learning_rate": 4.81089924275671e-05,
      "loss": 4.811,
      "num_input_tokens_seen": 6758912,
      "step": 4985
    },
    {
      "epoch": 0.15972088854746816,
      "grad_norm": 38.0,
      "learning_rate": 4.810524506087335e-05,
      "loss": 4.2435,
      "num_input_tokens_seen": 6765808,
      "step": 4990
    },
    {
      "epoch": 0.1598809295179566,
      "grad_norm": 27.25,
      "learning_rate": 4.810149413106044e-05,
      "loss": 4.394,
      "num_input_tokens_seen": 6772880,
      "step": 4995
    },
    {
      "epoch": 0.16004097048844504,
      "grad_norm": 31.875,
      "learning_rate": 4.809773963870684e-05,
      "loss": 4.3076,
      "num_input_tokens_seen": 6779376,
      "step": 5000
    },
    {
      "epoch": 0.16004097048844504,
      "eval_loss": 4.497137069702148,
      "eval_runtime": 487.9795,
      "eval_samples_per_second": 28.456,
      "eval_steps_per_second": 14.228,
      "num_input_tokens_seen": 6779376,
      "step": 5000
    },
    {
      "epoch": 0.16020101145893348,
      "grad_norm": 58.25,
      "learning_rate": 4.809398158439151e-05,
      "loss": 4.4914,
      "num_input_tokens_seen": 6786432,
      "step": 5005
    },
    {
      "epoch": 0.16036105242942192,
      "grad_norm": 21.25,
      "learning_rate": 4.8090219968694005e-05,
      "loss": 4.4754,
      "num_input_tokens_seen": 6793456,
      "step": 5010
    },
    {
      "epoch": 0.1605210933999104,
      "grad_norm": 15.4375,
      "learning_rate": 4.808645479219442e-05,
      "loss": 4.0049,
      "num_input_tokens_seen": 6800064,
      "step": 5015
    },
    {
      "epoch": 0.16068113437039883,
      "grad_norm": 29.25,
      "learning_rate": 4.8082686055473375e-05,
      "loss": 4.713,
      "num_input_tokens_seen": 6807728,
      "step": 5020
    },
    {
      "epoch": 0.16084117534088727,
      "grad_norm": 28.5,
      "learning_rate": 4.8078913759112066e-05,
      "loss": 4.494,
      "num_input_tokens_seen": 6814304,
      "step": 5025
    },
    {
      "epoch": 0.1610012163113757,
      "grad_norm": 25.5,
      "learning_rate": 4.807513790369223e-05,
      "loss": 3.4489,
      "num_input_tokens_seen": 6821232,
      "step": 5030
    },
    {
      "epoch": 0.16116125728186415,
      "grad_norm": 17.75,
      "learning_rate": 4.8071358489796145e-05,
      "loss": 3.8137,
      "num_input_tokens_seen": 6828048,
      "step": 5035
    },
    {
      "epoch": 0.1613212982523526,
      "grad_norm": 43.0,
      "learning_rate": 4.806757551800665e-05,
      "loss": 5.1147,
      "num_input_tokens_seen": 6834832,
      "step": 5040
    },
    {
      "epoch": 0.16148133922284105,
      "grad_norm": 40.0,
      "learning_rate": 4.806378898890713e-05,
      "loss": 5.4316,
      "num_input_tokens_seen": 6841216,
      "step": 5045
    },
    {
      "epoch": 0.1616413801933295,
      "grad_norm": 40.5,
      "learning_rate": 4.80599989030815e-05,
      "loss": 4.783,
      "num_input_tokens_seen": 6848176,
      "step": 5050
    },
    {
      "epoch": 0.16180142116381793,
      "grad_norm": 39.0,
      "learning_rate": 4.805620526111426e-05,
      "loss": 4.3281,
      "num_input_tokens_seen": 6854944,
      "step": 5055
    },
    {
      "epoch": 0.16196146213430637,
      "grad_norm": 32.75,
      "learning_rate": 4.805240806359042e-05,
      "loss": 4.4453,
      "num_input_tokens_seen": 6861808,
      "step": 5060
    },
    {
      "epoch": 0.16212150310479484,
      "grad_norm": 19.25,
      "learning_rate": 4.804860731109557e-05,
      "loss": 4.3362,
      "num_input_tokens_seen": 6868416,
      "step": 5065
    },
    {
      "epoch": 0.16228154407528328,
      "grad_norm": 22.5,
      "learning_rate": 4.804480300421581e-05,
      "loss": 3.7124,
      "num_input_tokens_seen": 6875248,
      "step": 5070
    },
    {
      "epoch": 0.16244158504577172,
      "grad_norm": 38.5,
      "learning_rate": 4.804099514353784e-05,
      "loss": 4.8385,
      "num_input_tokens_seen": 6882176,
      "step": 5075
    },
    {
      "epoch": 0.16260162601626016,
      "grad_norm": 32.5,
      "learning_rate": 4.8037183729648867e-05,
      "loss": 3.8085,
      "num_input_tokens_seen": 6889056,
      "step": 5080
    },
    {
      "epoch": 0.1627616669867486,
      "grad_norm": 32.25,
      "learning_rate": 4.803336876313666e-05,
      "loss": 4.6843,
      "num_input_tokens_seen": 6895856,
      "step": 5085
    },
    {
      "epoch": 0.16292170795723707,
      "grad_norm": 21.875,
      "learning_rate": 4.802955024458953e-05,
      "loss": 4.0995,
      "num_input_tokens_seen": 6902752,
      "step": 5090
    },
    {
      "epoch": 0.1630817489277255,
      "grad_norm": 33.75,
      "learning_rate": 4.802572817459634e-05,
      "loss": 3.8916,
      "num_input_tokens_seen": 6909104,
      "step": 5095
    },
    {
      "epoch": 0.16324178989821395,
      "grad_norm": 25.625,
      "learning_rate": 4.802190255374651e-05,
      "loss": 4.5369,
      "num_input_tokens_seen": 6916144,
      "step": 5100
    },
    {
      "epoch": 0.16340183086870239,
      "grad_norm": 25.375,
      "learning_rate": 4.801807338263e-05,
      "loss": 4.3675,
      "num_input_tokens_seen": 6922608,
      "step": 5105
    },
    {
      "epoch": 0.16356187183919083,
      "grad_norm": 17.875,
      "learning_rate": 4.8014240661837306e-05,
      "loss": 4.0628,
      "num_input_tokens_seen": 6929680,
      "step": 5110
    },
    {
      "epoch": 0.16372191280967927,
      "grad_norm": 77.5,
      "learning_rate": 4.80104043919595e-05,
      "loss": 3.8185,
      "num_input_tokens_seen": 6936352,
      "step": 5115
    },
    {
      "epoch": 0.16388195378016773,
      "grad_norm": 27.5,
      "learning_rate": 4.800656457358815e-05,
      "loss": 3.579,
      "num_input_tokens_seen": 6943072,
      "step": 5120
    },
    {
      "epoch": 0.16404199475065617,
      "grad_norm": 23.375,
      "learning_rate": 4.800272120731544e-05,
      "loss": 4.0197,
      "num_input_tokens_seen": 6949504,
      "step": 5125
    },
    {
      "epoch": 0.1642020357211446,
      "grad_norm": 29.375,
      "learning_rate": 4.799887429373404e-05,
      "loss": 3.8411,
      "num_input_tokens_seen": 6956544,
      "step": 5130
    },
    {
      "epoch": 0.16436207669163305,
      "grad_norm": 20.5,
      "learning_rate": 4.79950238334372e-05,
      "loss": 3.2779,
      "num_input_tokens_seen": 6963632,
      "step": 5135
    },
    {
      "epoch": 0.1645221176621215,
      "grad_norm": 23.125,
      "learning_rate": 4.799116982701872e-05,
      "loss": 5.1485,
      "num_input_tokens_seen": 6970320,
      "step": 5140
    },
    {
      "epoch": 0.16468215863260996,
      "grad_norm": 40.0,
      "learning_rate": 4.7987312275072926e-05,
      "loss": 6.1168,
      "num_input_tokens_seen": 6977040,
      "step": 5145
    },
    {
      "epoch": 0.1648421996030984,
      "grad_norm": 21.625,
      "learning_rate": 4.79834511781947e-05,
      "loss": 3.9336,
      "num_input_tokens_seen": 6984096,
      "step": 5150
    },
    {
      "epoch": 0.16500224057358684,
      "grad_norm": 23.125,
      "learning_rate": 4.797958653697947e-05,
      "loss": 3.6146,
      "num_input_tokens_seen": 6991088,
      "step": 5155
    },
    {
      "epoch": 0.16516228154407528,
      "grad_norm": 15.0,
      "learning_rate": 4.7975718352023225e-05,
      "loss": 3.898,
      "num_input_tokens_seen": 6997840,
      "step": 5160
    },
    {
      "epoch": 0.16532232251456372,
      "grad_norm": 23.0,
      "learning_rate": 4.7971846623922476e-05,
      "loss": 3.8871,
      "num_input_tokens_seen": 7004880,
      "step": 5165
    },
    {
      "epoch": 0.1654823634850522,
      "grad_norm": 23.0,
      "learning_rate": 4.7967971353274294e-05,
      "loss": 3.3871,
      "num_input_tokens_seen": 7011616,
      "step": 5170
    },
    {
      "epoch": 0.16564240445554063,
      "grad_norm": 52.75,
      "learning_rate": 4.79640925406763e-05,
      "loss": 3.6004,
      "num_input_tokens_seen": 7018112,
      "step": 5175
    },
    {
      "epoch": 0.16580244542602907,
      "grad_norm": 38.5,
      "learning_rate": 4.796021018672664e-05,
      "loss": 4.7444,
      "num_input_tokens_seen": 7024528,
      "step": 5180
    },
    {
      "epoch": 0.1659624863965175,
      "grad_norm": 42.25,
      "learning_rate": 4.795632429202405e-05,
      "loss": 4.9511,
      "num_input_tokens_seen": 7031232,
      "step": 5185
    },
    {
      "epoch": 0.16612252736700595,
      "grad_norm": 40.5,
      "learning_rate": 4.795243485716775e-05,
      "loss": 4.1883,
      "num_input_tokens_seen": 7037728,
      "step": 5190
    },
    {
      "epoch": 0.16628256833749439,
      "grad_norm": 42.75,
      "learning_rate": 4.794854188275757e-05,
      "loss": 3.8984,
      "num_input_tokens_seen": 7044608,
      "step": 5195
    },
    {
      "epoch": 0.16644260930798285,
      "grad_norm": 33.5,
      "learning_rate": 4.794464536939384e-05,
      "loss": 4.1459,
      "num_input_tokens_seen": 7051504,
      "step": 5200
    },
    {
      "epoch": 0.16644260930798285,
      "eval_loss": 4.504709243774414,
      "eval_runtime": 487.7129,
      "eval_samples_per_second": 28.472,
      "eval_steps_per_second": 14.236,
      "num_input_tokens_seen": 7051504,
      "step": 5200
    },
    {
      "epoch": 0.1666026502784713,
      "grad_norm": 39.75,
      "learning_rate": 4.794074531767745e-05,
      "loss": 4.0721,
      "num_input_tokens_seen": 7058224,
      "step": 5205
    },
    {
      "epoch": 0.16676269124895973,
      "grad_norm": 31.375,
      "learning_rate": 4.7936841728209834e-05,
      "loss": 4.7401,
      "num_input_tokens_seen": 7065088,
      "step": 5210
    },
    {
      "epoch": 0.16692273221944817,
      "grad_norm": 28.75,
      "learning_rate": 4.7932934601593e-05,
      "loss": 3.9665,
      "num_input_tokens_seen": 7071632,
      "step": 5215
    },
    {
      "epoch": 0.1670827731899366,
      "grad_norm": 28.25,
      "learning_rate": 4.792902393842943e-05,
      "loss": 4.4557,
      "num_input_tokens_seen": 7078416,
      "step": 5220
    },
    {
      "epoch": 0.16724281416042508,
      "grad_norm": 30.875,
      "learning_rate": 4.792510973932225e-05,
      "loss": 4.3638,
      "num_input_tokens_seen": 7085248,
      "step": 5225
    },
    {
      "epoch": 0.16740285513091352,
      "grad_norm": 28.0,
      "learning_rate": 4.7921192004875036e-05,
      "loss": 3.6881,
      "num_input_tokens_seen": 7092464,
      "step": 5230
    },
    {
      "epoch": 0.16756289610140196,
      "grad_norm": 27.0,
      "learning_rate": 4.791727073569198e-05,
      "loss": 3.7939,
      "num_input_tokens_seen": 7098832,
      "step": 5235
    },
    {
      "epoch": 0.1677229370718904,
      "grad_norm": 28.75,
      "learning_rate": 4.7913345932377775e-05,
      "loss": 4.1741,
      "num_input_tokens_seen": 7105904,
      "step": 5240
    },
    {
      "epoch": 0.16788297804237884,
      "grad_norm": 23.875,
      "learning_rate": 4.790941759553769e-05,
      "loss": 4.1528,
      "num_input_tokens_seen": 7112496,
      "step": 5245
    },
    {
      "epoch": 0.1680430190128673,
      "grad_norm": 22.75,
      "learning_rate": 4.79054857257775e-05,
      "loss": 3.8899,
      "num_input_tokens_seen": 7118880,
      "step": 5250
    },
    {
      "epoch": 0.16820305998335575,
      "grad_norm": 45.25,
      "learning_rate": 4.790155032370357e-05,
      "loss": 4.3477,
      "num_input_tokens_seen": 7125344,
      "step": 5255
    },
    {
      "epoch": 0.16836310095384419,
      "grad_norm": 30.875,
      "learning_rate": 4.789761138992278e-05,
      "loss": 5.1062,
      "num_input_tokens_seen": 7132128,
      "step": 5260
    },
    {
      "epoch": 0.16852314192433263,
      "grad_norm": 20.75,
      "learning_rate": 4.7893668925042565e-05,
      "loss": 4.4213,
      "num_input_tokens_seen": 7138880,
      "step": 5265
    },
    {
      "epoch": 0.16868318289482107,
      "grad_norm": 22.375,
      "learning_rate": 4.78897229296709e-05,
      "loss": 5.0178,
      "num_input_tokens_seen": 7145632,
      "step": 5270
    },
    {
      "epoch": 0.16884322386530953,
      "grad_norm": 36.0,
      "learning_rate": 4.7885773404416315e-05,
      "loss": 3.5617,
      "num_input_tokens_seen": 7152896,
      "step": 5275
    },
    {
      "epoch": 0.16900326483579797,
      "grad_norm": 31.625,
      "learning_rate": 4.788182034988786e-05,
      "loss": 4.8095,
      "num_input_tokens_seen": 7159920,
      "step": 5280
    },
    {
      "epoch": 0.1691633058062864,
      "grad_norm": 23.625,
      "learning_rate": 4.787786376669516e-05,
      "loss": 4.0639,
      "num_input_tokens_seen": 7166512,
      "step": 5285
    },
    {
      "epoch": 0.16932334677677485,
      "grad_norm": 26.75,
      "learning_rate": 4.787390365544837e-05,
      "loss": 3.6846,
      "num_input_tokens_seen": 7173296,
      "step": 5290
    },
    {
      "epoch": 0.1694833877472633,
      "grad_norm": 28.625,
      "learning_rate": 4.786994001675818e-05,
      "loss": 5.1753,
      "num_input_tokens_seen": 7179824,
      "step": 5295
    },
    {
      "epoch": 0.16964342871775173,
      "grad_norm": 17.875,
      "learning_rate": 4.786597285123584e-05,
      "loss": 4.5958,
      "num_input_tokens_seen": 7186720,
      "step": 5300
    },
    {
      "epoch": 0.1698034696882402,
      "grad_norm": 23.625,
      "learning_rate": 4.7862002159493135e-05,
      "loss": 4.5668,
      "num_input_tokens_seen": 7193216,
      "step": 5305
    },
    {
      "epoch": 0.16996351065872864,
      "grad_norm": 27.125,
      "learning_rate": 4.785802794214239e-05,
      "loss": 5.0498,
      "num_input_tokens_seen": 7200240,
      "step": 5310
    },
    {
      "epoch": 0.17012355162921708,
      "grad_norm": 23.5,
      "learning_rate": 4.7854050199796495e-05,
      "loss": 4.0007,
      "num_input_tokens_seen": 7207264,
      "step": 5315
    },
    {
      "epoch": 0.17028359259970552,
      "grad_norm": 34.0,
      "learning_rate": 4.7850068933068845e-05,
      "loss": 4.3053,
      "num_input_tokens_seen": 7213984,
      "step": 5320
    },
    {
      "epoch": 0.17044363357019396,
      "grad_norm": 27.0,
      "learning_rate": 4.7846084142573425e-05,
      "loss": 5.6306,
      "num_input_tokens_seen": 7220592,
      "step": 5325
    },
    {
      "epoch": 0.17060367454068243,
      "grad_norm": 32.25,
      "learning_rate": 4.7842095828924725e-05,
      "loss": 4.9042,
      "num_input_tokens_seen": 7227232,
      "step": 5330
    },
    {
      "epoch": 0.17076371551117087,
      "grad_norm": 26.375,
      "learning_rate": 4.783810399273779e-05,
      "loss": 5.6261,
      "num_input_tokens_seen": 7234128,
      "step": 5335
    },
    {
      "epoch": 0.1709237564816593,
      "grad_norm": 36.75,
      "learning_rate": 4.7834108634628226e-05,
      "loss": 4.3912,
      "num_input_tokens_seen": 7240656,
      "step": 5340
    },
    {
      "epoch": 0.17108379745214775,
      "grad_norm": 58.0,
      "learning_rate": 4.783010975521216e-05,
      "loss": 4.7664,
      "num_input_tokens_seen": 7247440,
      "step": 5345
    },
    {
      "epoch": 0.17124383842263619,
      "grad_norm": 30.875,
      "learning_rate": 4.782610735510626e-05,
      "loss": 4.5167,
      "num_input_tokens_seen": 7254224,
      "step": 5350
    },
    {
      "epoch": 0.17140387939312465,
      "grad_norm": 59.25,
      "learning_rate": 4.782210143492776e-05,
      "loss": 4.8347,
      "num_input_tokens_seen": 7261136,
      "step": 5355
    },
    {
      "epoch": 0.1715639203636131,
      "grad_norm": 37.0,
      "learning_rate": 4.781809199529442e-05,
      "loss": 5.1358,
      "num_input_tokens_seen": 7267504,
      "step": 5360
    },
    {
      "epoch": 0.17172396133410153,
      "grad_norm": 38.5,
      "learning_rate": 4.781407903682454e-05,
      "loss": 4.2233,
      "num_input_tokens_seen": 7274352,
      "step": 5365
    },
    {
      "epoch": 0.17188400230458997,
      "grad_norm": 32.25,
      "learning_rate": 4.781006256013698e-05,
      "loss": 4.1652,
      "num_input_tokens_seen": 7281136,
      "step": 5370
    },
    {
      "epoch": 0.1720440432750784,
      "grad_norm": 40.0,
      "learning_rate": 4.7806042565851115e-05,
      "loss": 4.6687,
      "num_input_tokens_seen": 7287680,
      "step": 5375
    },
    {
      "epoch": 0.17220408424556685,
      "grad_norm": 28.625,
      "learning_rate": 4.7802019054586895e-05,
      "loss": 4.0428,
      "num_input_tokens_seen": 7294224,
      "step": 5380
    },
    {
      "epoch": 0.17236412521605532,
      "grad_norm": 43.0,
      "learning_rate": 4.779799202696479e-05,
      "loss": 4.3473,
      "num_input_tokens_seen": 7300848,
      "step": 5385
    },
    {
      "epoch": 0.17252416618654376,
      "grad_norm": 15.875,
      "learning_rate": 4.779396148360581e-05,
      "loss": 4.4186,
      "num_input_tokens_seen": 7308080,
      "step": 5390
    },
    {
      "epoch": 0.1726842071570322,
      "grad_norm": 20.875,
      "learning_rate": 4.7789927425131517e-05,
      "loss": 4.3427,
      "num_input_tokens_seen": 7315152,
      "step": 5395
    },
    {
      "epoch": 0.17284424812752064,
      "grad_norm": 32.25,
      "learning_rate": 4.778588985216403e-05,
      "loss": 3.943,
      "num_input_tokens_seen": 7321552,
      "step": 5400
    },
    {
      "epoch": 0.17284424812752064,
      "eval_loss": 4.488741397857666,
      "eval_runtime": 487.0441,
      "eval_samples_per_second": 28.511,
      "eval_steps_per_second": 14.255,
      "num_input_tokens_seen": 7321552,
      "step": 5400
    },
    {
      "epoch": 0.17300428909800908,
      "grad_norm": 39.75,
      "learning_rate": 4.778184876532598e-05,
      "loss": 4.7971,
      "num_input_tokens_seen": 7328288,
      "step": 5405
    },
    {
      "epoch": 0.17316433006849755,
      "grad_norm": 30.625,
      "learning_rate": 4.7777804165240556e-05,
      "loss": 5.3668,
      "num_input_tokens_seen": 7334784,
      "step": 5410
    },
    {
      "epoch": 0.17332437103898599,
      "grad_norm": 27.75,
      "learning_rate": 4.7773756052531485e-05,
      "loss": 4.1584,
      "num_input_tokens_seen": 7341904,
      "step": 5415
    },
    {
      "epoch": 0.17348441200947443,
      "grad_norm": 32.25,
      "learning_rate": 4.7769704427823035e-05,
      "loss": 4.4452,
      "num_input_tokens_seen": 7348368,
      "step": 5420
    },
    {
      "epoch": 0.17364445297996287,
      "grad_norm": 24.125,
      "learning_rate": 4.776564929174003e-05,
      "loss": 3.7112,
      "num_input_tokens_seen": 7355040,
      "step": 5425
    },
    {
      "epoch": 0.1738044939504513,
      "grad_norm": 40.5,
      "learning_rate": 4.7761590644907806e-05,
      "loss": 4.3411,
      "num_input_tokens_seen": 7361792,
      "step": 5430
    },
    {
      "epoch": 0.17396453492093977,
      "grad_norm": 32.0,
      "learning_rate": 4.7757528487952263e-05,
      "loss": 3.9839,
      "num_input_tokens_seen": 7368720,
      "step": 5435
    },
    {
      "epoch": 0.1741245758914282,
      "grad_norm": 39.25,
      "learning_rate": 4.7753462821499836e-05,
      "loss": 3.7032,
      "num_input_tokens_seen": 7375408,
      "step": 5440
    },
    {
      "epoch": 0.17428461686191665,
      "grad_norm": 25.0,
      "learning_rate": 4.774939364617751e-05,
      "loss": 3.8028,
      "num_input_tokens_seen": 7381952,
      "step": 5445
    },
    {
      "epoch": 0.1744446578324051,
      "grad_norm": 27.875,
      "learning_rate": 4.7745320962612795e-05,
      "loss": 5.1947,
      "num_input_tokens_seen": 7388688,
      "step": 5450
    },
    {
      "epoch": 0.17460469880289353,
      "grad_norm": 52.0,
      "learning_rate": 4.7741244771433756e-05,
      "loss": 5.1015,
      "num_input_tokens_seen": 7395392,
      "step": 5455
    },
    {
      "epoch": 0.174764739773382,
      "grad_norm": 21.125,
      "learning_rate": 4.7737165073268985e-05,
      "loss": 3.8987,
      "num_input_tokens_seen": 7402320,
      "step": 5460
    },
    {
      "epoch": 0.17492478074387044,
      "grad_norm": 21.75,
      "learning_rate": 4.7733081868747626e-05,
      "loss": 3.3869,
      "num_input_tokens_seen": 7409040,
      "step": 5465
    },
    {
      "epoch": 0.17508482171435888,
      "grad_norm": 30.25,
      "learning_rate": 4.772899515849936e-05,
      "loss": 4.3086,
      "num_input_tokens_seen": 7415904,
      "step": 5470
    },
    {
      "epoch": 0.17524486268484732,
      "grad_norm": 31.25,
      "learning_rate": 4.7724904943154414e-05,
      "loss": 5.2278,
      "num_input_tokens_seen": 7422768,
      "step": 5475
    },
    {
      "epoch": 0.17540490365533576,
      "grad_norm": 25.5,
      "learning_rate": 4.772081122334354e-05,
      "loss": 4.9198,
      "num_input_tokens_seen": 7429712,
      "step": 5480
    },
    {
      "epoch": 0.1755649446258242,
      "grad_norm": 30.875,
      "learning_rate": 4.771671399969806e-05,
      "loss": 4.868,
      "num_input_tokens_seen": 7436480,
      "step": 5485
    },
    {
      "epoch": 0.17572498559631267,
      "grad_norm": 23.25,
      "learning_rate": 4.7712613272849794e-05,
      "loss": 3.936,
      "num_input_tokens_seen": 7443648,
      "step": 5490
    },
    {
      "epoch": 0.1758850265668011,
      "grad_norm": 84.0,
      "learning_rate": 4.770850904343114e-05,
      "loss": 4.4062,
      "num_input_tokens_seen": 7450816,
      "step": 5495
    },
    {
      "epoch": 0.17604506753728955,
      "grad_norm": 16.375,
      "learning_rate": 4.770440131207502e-05,
      "loss": 3.6426,
      "num_input_tokens_seen": 7457536,
      "step": 5500
    },
    {
      "epoch": 0.17620510850777799,
      "grad_norm": 25.625,
      "learning_rate": 4.7700290079414896e-05,
      "loss": 4.8701,
      "num_input_tokens_seen": 7464336,
      "step": 5505
    },
    {
      "epoch": 0.17636514947826643,
      "grad_norm": 51.25,
      "learning_rate": 4.769617534608477e-05,
      "loss": 3.7336,
      "num_input_tokens_seen": 7471136,
      "step": 5510
    },
    {
      "epoch": 0.1765251904487549,
      "grad_norm": 25.875,
      "learning_rate": 4.7692057112719193e-05,
      "loss": 4.0497,
      "num_input_tokens_seen": 7478192,
      "step": 5515
    },
    {
      "epoch": 0.17668523141924333,
      "grad_norm": 32.5,
      "learning_rate": 4.7687935379953234e-05,
      "loss": 4.3384,
      "num_input_tokens_seen": 7484752,
      "step": 5520
    },
    {
      "epoch": 0.17684527238973177,
      "grad_norm": 27.625,
      "learning_rate": 4.7683810148422534e-05,
      "loss": 5.5303,
      "num_input_tokens_seen": 7491648,
      "step": 5525
    },
    {
      "epoch": 0.1770053133602202,
      "grad_norm": 20.25,
      "learning_rate": 4.767968141876324e-05,
      "loss": 4.7169,
      "num_input_tokens_seen": 7498496,
      "step": 5530
    },
    {
      "epoch": 0.17716535433070865,
      "grad_norm": 32.75,
      "learning_rate": 4.767554919161207e-05,
      "loss": 4.613,
      "num_input_tokens_seen": 7504992,
      "step": 5535
    },
    {
      "epoch": 0.17732539530119712,
      "grad_norm": 40.75,
      "learning_rate": 4.767141346760624e-05,
      "loss": 4.0051,
      "num_input_tokens_seen": 7511520,
      "step": 5540
    },
    {
      "epoch": 0.17748543627168556,
      "grad_norm": 25.25,
      "learning_rate": 4.766727424738356e-05,
      "loss": 3.3941,
      "num_input_tokens_seen": 7518464,
      "step": 5545
    },
    {
      "epoch": 0.177645477242174,
      "grad_norm": 32.0,
      "learning_rate": 4.7663131531582325e-05,
      "loss": 4.07,
      "num_input_tokens_seen": 7524896,
      "step": 5550
    },
    {
      "epoch": 0.17780551821266244,
      "grad_norm": 46.5,
      "learning_rate": 4.765898532084142e-05,
      "loss": 3.9924,
      "num_input_tokens_seen": 7531568,
      "step": 5555
    },
    {
      "epoch": 0.17796555918315088,
      "grad_norm": 30.5,
      "learning_rate": 4.765483561580022e-05,
      "loss": 4.5661,
      "num_input_tokens_seen": 7538352,
      "step": 5560
    },
    {
      "epoch": 0.17812560015363932,
      "grad_norm": 34.5,
      "learning_rate": 4.7650682417098666e-05,
      "loss": 4.4192,
      "num_input_tokens_seen": 7545440,
      "step": 5565
    },
    {
      "epoch": 0.17828564112412779,
      "grad_norm": 23.75,
      "learning_rate": 4.7646525725377244e-05,
      "loss": 3.949,
      "num_input_tokens_seen": 7551808,
      "step": 5570
    },
    {
      "epoch": 0.17844568209461623,
      "grad_norm": 26.125,
      "learning_rate": 4.764236554127696e-05,
      "loss": 3.9253,
      "num_input_tokens_seen": 7558848,
      "step": 5575
    },
    {
      "epoch": 0.17860572306510467,
      "grad_norm": 37.0,
      "learning_rate": 4.7638201865439356e-05,
      "loss": 4.1684,
      "num_input_tokens_seen": 7565440,
      "step": 5580
    },
    {
      "epoch": 0.1787657640355931,
      "grad_norm": 24.875,
      "learning_rate": 4.7634034698506545e-05,
      "loss": 4.4785,
      "num_input_tokens_seen": 7572320,
      "step": 5585
    },
    {
      "epoch": 0.17892580500608155,
      "grad_norm": 23.0,
      "learning_rate": 4.762986404112115e-05,
      "loss": 4.7833,
      "num_input_tokens_seen": 7578880,
      "step": 5590
    },
    {
      "epoch": 0.17908584597657,
      "grad_norm": 19.5,
      "learning_rate": 4.762568989392633e-05,
      "loss": 3.969,
      "num_input_tokens_seen": 7585728,
      "step": 5595
    },
    {
      "epoch": 0.17924588694705845,
      "grad_norm": 28.125,
      "learning_rate": 4.76215122575658e-05,
      "loss": 3.7287,
      "num_input_tokens_seen": 7592304,
      "step": 5600
    },
    {
      "epoch": 0.17924588694705845,
      "eval_loss": 4.471382141113281,
      "eval_runtime": 487.5691,
      "eval_samples_per_second": 28.48,
      "eval_steps_per_second": 14.24,
      "num_input_tokens_seen": 7592304,
      "step": 5600
    },
    {
      "epoch": 0.1794059279175469,
      "grad_norm": 26.125,
      "learning_rate": 4.7617331132683795e-05,
      "loss": 3.9313,
      "num_input_tokens_seen": 7598784,
      "step": 5605
    },
    {
      "epoch": 0.17956596888803533,
      "grad_norm": 36.25,
      "learning_rate": 4.7613146519925105e-05,
      "loss": 4.0202,
      "num_input_tokens_seen": 7605536,
      "step": 5610
    },
    {
      "epoch": 0.17972600985852377,
      "grad_norm": 24.25,
      "learning_rate": 4.7608958419935045e-05,
      "loss": 3.8896,
      "num_input_tokens_seen": 7612576,
      "step": 5615
    },
    {
      "epoch": 0.17988605082901224,
      "grad_norm": 70.5,
      "learning_rate": 4.760476683335948e-05,
      "loss": 4.8746,
      "num_input_tokens_seen": 7619104,
      "step": 5620
    },
    {
      "epoch": 0.18004609179950068,
      "grad_norm": 32.5,
      "learning_rate": 4.760057176084479e-05,
      "loss": 4.6588,
      "num_input_tokens_seen": 7625584,
      "step": 5625
    },
    {
      "epoch": 0.18020613276998912,
      "grad_norm": 26.875,
      "learning_rate": 4.759637320303793e-05,
      "loss": 4.4081,
      "num_input_tokens_seen": 7632128,
      "step": 5630
    },
    {
      "epoch": 0.18036617374047756,
      "grad_norm": 33.25,
      "learning_rate": 4.759217116058635e-05,
      "loss": 4.1384,
      "num_input_tokens_seen": 7638720,
      "step": 5635
    },
    {
      "epoch": 0.180526214710966,
      "grad_norm": 18.875,
      "learning_rate": 4.758796563413807e-05,
      "loss": 4.4549,
      "num_input_tokens_seen": 7645856,
      "step": 5640
    },
    {
      "epoch": 0.18068625568145444,
      "grad_norm": 29.875,
      "learning_rate": 4.758375662434163e-05,
      "loss": 4.1053,
      "num_input_tokens_seen": 7652544,
      "step": 5645
    },
    {
      "epoch": 0.1808462966519429,
      "grad_norm": 27.375,
      "learning_rate": 4.7579544131846114e-05,
      "loss": 4.9998,
      "num_input_tokens_seen": 7659568,
      "step": 5650
    },
    {
      "epoch": 0.18100633762243135,
      "grad_norm": 21.75,
      "learning_rate": 4.757532815730114e-05,
      "loss": 4.3697,
      "num_input_tokens_seen": 7666384,
      "step": 5655
    },
    {
      "epoch": 0.18116637859291979,
      "grad_norm": 36.75,
      "learning_rate": 4.7571108701356865e-05,
      "loss": 4.0533,
      "num_input_tokens_seen": 7672944,
      "step": 5660
    },
    {
      "epoch": 0.18132641956340823,
      "grad_norm": 19.875,
      "learning_rate": 4.756688576466398e-05,
      "loss": 4.63,
      "num_input_tokens_seen": 7679792,
      "step": 5665
    },
    {
      "epoch": 0.18148646053389667,
      "grad_norm": 28.625,
      "learning_rate": 4.756265934787372e-05,
      "loss": 3.8748,
      "num_input_tokens_seen": 7686944,
      "step": 5670
    },
    {
      "epoch": 0.18164650150438513,
      "grad_norm": 29.0,
      "learning_rate": 4.755842945163785e-05,
      "loss": 4.7054,
      "num_input_tokens_seen": 7694416,
      "step": 5675
    },
    {
      "epoch": 0.18180654247487357,
      "grad_norm": 14.375,
      "learning_rate": 4.755419607660867e-05,
      "loss": 4.2227,
      "num_input_tokens_seen": 7701504,
      "step": 5680
    },
    {
      "epoch": 0.181966583445362,
      "grad_norm": 29.375,
      "learning_rate": 4.7549959223439016e-05,
      "loss": 4.2579,
      "num_input_tokens_seen": 7708080,
      "step": 5685
    },
    {
      "epoch": 0.18212662441585045,
      "grad_norm": 34.5,
      "learning_rate": 4.754571889278228e-05,
      "loss": 5.0665,
      "num_input_tokens_seen": 7714896,
      "step": 5690
    },
    {
      "epoch": 0.1822866653863389,
      "grad_norm": 34.75,
      "learning_rate": 4.754147508529235e-05,
      "loss": 4.563,
      "num_input_tokens_seen": 7721376,
      "step": 5695
    },
    {
      "epoch": 0.18244670635682736,
      "grad_norm": 33.5,
      "learning_rate": 4.75372278016237e-05,
      "loss": 4.0098,
      "num_input_tokens_seen": 7728160,
      "step": 5700
    },
    {
      "epoch": 0.1826067473273158,
      "grad_norm": 31.0,
      "learning_rate": 4.753297704243129e-05,
      "loss": 3.756,
      "num_input_tokens_seen": 7734672,
      "step": 5705
    },
    {
      "epoch": 0.18276678829780424,
      "grad_norm": 27.75,
      "learning_rate": 4.752872280837066e-05,
      "loss": 3.9439,
      "num_input_tokens_seen": 7741760,
      "step": 5710
    },
    {
      "epoch": 0.18292682926829268,
      "grad_norm": 27.25,
      "learning_rate": 4.752446510009786e-05,
      "loss": 4.371,
      "num_input_tokens_seen": 7748992,
      "step": 5715
    },
    {
      "epoch": 0.18308687023878112,
      "grad_norm": 22.375,
      "learning_rate": 4.7520203918269476e-05,
      "loss": 3.9334,
      "num_input_tokens_seen": 7756016,
      "step": 5720
    },
    {
      "epoch": 0.1832469112092696,
      "grad_norm": 22.375,
      "learning_rate": 4.751593926354265e-05,
      "loss": 4.0611,
      "num_input_tokens_seen": 7762800,
      "step": 5725
    },
    {
      "epoch": 0.18340695217975803,
      "grad_norm": 28.5,
      "learning_rate": 4.751167113657503e-05,
      "loss": 4.6979,
      "num_input_tokens_seen": 7769616,
      "step": 5730
    },
    {
      "epoch": 0.18356699315024647,
      "grad_norm": 26.75,
      "learning_rate": 4.7507399538024834e-05,
      "loss": 3.8687,
      "num_input_tokens_seen": 7776208,
      "step": 5735
    },
    {
      "epoch": 0.1837270341207349,
      "grad_norm": 38.0,
      "learning_rate": 4.750312446855077e-05,
      "loss": 4.6713,
      "num_input_tokens_seen": 7783200,
      "step": 5740
    },
    {
      "epoch": 0.18388707509122335,
      "grad_norm": 43.5,
      "learning_rate": 4.749884592881212e-05,
      "loss": 4.5244,
      "num_input_tokens_seen": 7789776,
      "step": 5745
    },
    {
      "epoch": 0.18404711606171179,
      "grad_norm": 35.75,
      "learning_rate": 4.74945639194687e-05,
      "loss": 4.2023,
      "num_input_tokens_seen": 7796544,
      "step": 5750
    },
    {
      "epoch": 0.18420715703220025,
      "grad_norm": 21.25,
      "learning_rate": 4.749027844118083e-05,
      "loss": 4.4357,
      "num_input_tokens_seen": 7803344,
      "step": 5755
    },
    {
      "epoch": 0.1843671980026887,
      "grad_norm": 20.375,
      "learning_rate": 4.7485989494609395e-05,
      "loss": 4.7052,
      "num_input_tokens_seen": 7810272,
      "step": 5760
    },
    {
      "epoch": 0.18452723897317713,
      "grad_norm": 16.5,
      "learning_rate": 4.748169708041581e-05,
      "loss": 3.7967,
      "num_input_tokens_seen": 7817072,
      "step": 5765
    },
    {
      "epoch": 0.18468727994366557,
      "grad_norm": 28.0,
      "learning_rate": 4.7477401199262004e-05,
      "loss": 4.0093,
      "num_input_tokens_seen": 7823984,
      "step": 5770
    },
    {
      "epoch": 0.184847320914154,
      "grad_norm": 39.25,
      "learning_rate": 4.747310185181048e-05,
      "loss": 4.0201,
      "num_input_tokens_seen": 7831088,
      "step": 5775
    },
    {
      "epoch": 0.18500736188464248,
      "grad_norm": 26.75,
      "learning_rate": 4.746879903872422e-05,
      "loss": 4.4199,
      "num_input_tokens_seen": 7838000,
      "step": 5780
    },
    {
      "epoch": 0.18516740285513092,
      "grad_norm": 19.25,
      "learning_rate": 4.746449276066679e-05,
      "loss": 3.2885,
      "num_input_tokens_seen": 7844912,
      "step": 5785
    },
    {
      "epoch": 0.18532744382561936,
      "grad_norm": 31.625,
      "learning_rate": 4.746018301830227e-05,
      "loss": 3.2575,
      "num_input_tokens_seen": 7851488,
      "step": 5790
    },
    {
      "epoch": 0.1854874847961078,
      "grad_norm": 29.125,
      "learning_rate": 4.7455869812295275e-05,
      "loss": 4.3763,
      "num_input_tokens_seen": 7858736,
      "step": 5795
    },
    {
      "epoch": 0.18564752576659624,
      "grad_norm": 21.75,
      "learning_rate": 4.7451553143310964e-05,
      "loss": 4.5494,
      "num_input_tokens_seen": 7865632,
      "step": 5800
    },
    {
      "epoch": 0.18564752576659624,
      "eval_loss": 4.484038829803467,
      "eval_runtime": 487.8283,
      "eval_samples_per_second": 28.465,
      "eval_steps_per_second": 14.232,
      "num_input_tokens_seen": 7865632,
      "step": 5800
    },
    {
      "epoch": 0.1858075667370847,
      "grad_norm": 25.75,
      "learning_rate": 4.744723301201501e-05,
      "loss": 3.9798,
      "num_input_tokens_seen": 7872240,
      "step": 5805
    },
    {
      "epoch": 0.18596760770757315,
      "grad_norm": 14.0625,
      "learning_rate": 4.744290941907364e-05,
      "loss": 4.0921,
      "num_input_tokens_seen": 7879648,
      "step": 5810
    },
    {
      "epoch": 0.18612764867806159,
      "grad_norm": 33.75,
      "learning_rate": 4.7438582365153594e-05,
      "loss": 4.0556,
      "num_input_tokens_seen": 7886656,
      "step": 5815
    },
    {
      "epoch": 0.18628768964855003,
      "grad_norm": 20.875,
      "learning_rate": 4.743425185092217e-05,
      "loss": 3.5841,
      "num_input_tokens_seen": 7893280,
      "step": 5820
    },
    {
      "epoch": 0.18644773061903847,
      "grad_norm": 15.5625,
      "learning_rate": 4.742991787704719e-05,
      "loss": 4.1423,
      "num_input_tokens_seen": 7900176,
      "step": 5825
    },
    {
      "epoch": 0.1866077715895269,
      "grad_norm": 21.25,
      "learning_rate": 4.7425580444196994e-05,
      "loss": 3.3455,
      "num_input_tokens_seen": 7906784,
      "step": 5830
    },
    {
      "epoch": 0.18676781256001537,
      "grad_norm": 48.75,
      "learning_rate": 4.742123955304048e-05,
      "loss": 4.1579,
      "num_input_tokens_seen": 7914016,
      "step": 5835
    },
    {
      "epoch": 0.1869278535305038,
      "grad_norm": 32.5,
      "learning_rate": 4.741689520424706e-05,
      "loss": 4.1574,
      "num_input_tokens_seen": 7920528,
      "step": 5840
    },
    {
      "epoch": 0.18708789450099225,
      "grad_norm": 43.5,
      "learning_rate": 4.741254739848669e-05,
      "loss": 4.6524,
      "num_input_tokens_seen": 7927504,
      "step": 5845
    },
    {
      "epoch": 0.1872479354714807,
      "grad_norm": 33.75,
      "learning_rate": 4.740819613642987e-05,
      "loss": 4.4901,
      "num_input_tokens_seen": 7934688,
      "step": 5850
    },
    {
      "epoch": 0.18740797644196913,
      "grad_norm": 28.75,
      "learning_rate": 4.74038414187476e-05,
      "loss": 3.9867,
      "num_input_tokens_seen": 7941264,
      "step": 5855
    },
    {
      "epoch": 0.1875680174124576,
      "grad_norm": 20.875,
      "learning_rate": 4.739948324611144e-05,
      "loss": 4.9667,
      "num_input_tokens_seen": 7948112,
      "step": 5860
    },
    {
      "epoch": 0.18772805838294604,
      "grad_norm": 23.375,
      "learning_rate": 4.7395121619193465e-05,
      "loss": 3.4122,
      "num_input_tokens_seen": 7954672,
      "step": 5865
    },
    {
      "epoch": 0.18788809935343448,
      "grad_norm": 35.5,
      "learning_rate": 4.7390756538666313e-05,
      "loss": 4.6721,
      "num_input_tokens_seen": 7961360,
      "step": 5870
    },
    {
      "epoch": 0.18804814032392292,
      "grad_norm": 32.75,
      "learning_rate": 4.738638800520311e-05,
      "loss": 4.2334,
      "num_input_tokens_seen": 7967984,
      "step": 5875
    },
    {
      "epoch": 0.18820818129441136,
      "grad_norm": 16.375,
      "learning_rate": 4.738201601947757e-05,
      "loss": 4.0962,
      "num_input_tokens_seen": 7974736,
      "step": 5880
    },
    {
      "epoch": 0.18836822226489983,
      "grad_norm": 45.25,
      "learning_rate": 4.7377640582163876e-05,
      "loss": 4.823,
      "num_input_tokens_seen": 7981280,
      "step": 5885
    },
    {
      "epoch": 0.18852826323538827,
      "grad_norm": 26.5,
      "learning_rate": 4.7373261693936786e-05,
      "loss": 3.9782,
      "num_input_tokens_seen": 7987728,
      "step": 5890
    },
    {
      "epoch": 0.1886883042058767,
      "grad_norm": 44.75,
      "learning_rate": 4.7368879355471595e-05,
      "loss": 4.6064,
      "num_input_tokens_seen": 7994368,
      "step": 5895
    },
    {
      "epoch": 0.18884834517636515,
      "grad_norm": 28.0,
      "learning_rate": 4.736449356744409e-05,
      "loss": 3.7916,
      "num_input_tokens_seen": 8001216,
      "step": 5900
    },
    {
      "epoch": 0.18900838614685359,
      "grad_norm": 44.0,
      "learning_rate": 4.736010433053064e-05,
      "loss": 3.9717,
      "num_input_tokens_seen": 8008048,
      "step": 5905
    },
    {
      "epoch": 0.18916842711734205,
      "grad_norm": 44.5,
      "learning_rate": 4.73557116454081e-05,
      "loss": 5.1763,
      "num_input_tokens_seen": 8015040,
      "step": 5910
    },
    {
      "epoch": 0.1893284680878305,
      "grad_norm": 40.25,
      "learning_rate": 4.735131551275389e-05,
      "loss": 4.5885,
      "num_input_tokens_seen": 8021600,
      "step": 5915
    },
    {
      "epoch": 0.18948850905831893,
      "grad_norm": 47.0,
      "learning_rate": 4.734691593324594e-05,
      "loss": 4.2328,
      "num_input_tokens_seen": 8027856,
      "step": 5920
    },
    {
      "epoch": 0.18964855002880737,
      "grad_norm": 15.25,
      "learning_rate": 4.734251290756272e-05,
      "loss": 4.4037,
      "num_input_tokens_seen": 8034560,
      "step": 5925
    },
    {
      "epoch": 0.1898085909992958,
      "grad_norm": 23.625,
      "learning_rate": 4.7338106436383246e-05,
      "loss": 4.2355,
      "num_input_tokens_seen": 8041088,
      "step": 5930
    },
    {
      "epoch": 0.18996863196978425,
      "grad_norm": 48.75,
      "learning_rate": 4.733369652038703e-05,
      "loss": 5.0149,
      "num_input_tokens_seen": 8047792,
      "step": 5935
    },
    {
      "epoch": 0.19012867294027272,
      "grad_norm": 38.25,
      "learning_rate": 4.7329283160254156e-05,
      "loss": 4.5263,
      "num_input_tokens_seen": 8054736,
      "step": 5940
    },
    {
      "epoch": 0.19028871391076116,
      "grad_norm": 30.5,
      "learning_rate": 4.732486635666521e-05,
      "loss": 3.6695,
      "num_input_tokens_seen": 8061872,
      "step": 5945
    },
    {
      "epoch": 0.1904487548812496,
      "grad_norm": 32.25,
      "learning_rate": 4.732044611030132e-05,
      "loss": 4.4423,
      "num_input_tokens_seen": 8068384,
      "step": 5950
    },
    {
      "epoch": 0.19060879585173804,
      "grad_norm": 24.625,
      "learning_rate": 4.731602242184414e-05,
      "loss": 4.1324,
      "num_input_tokens_seen": 8075312,
      "step": 5955
    },
    {
      "epoch": 0.19076883682222648,
      "grad_norm": 26.5,
      "learning_rate": 4.7311595291975864e-05,
      "loss": 3.547,
      "num_input_tokens_seen": 8081808,
      "step": 5960
    },
    {
      "epoch": 0.19092887779271495,
      "grad_norm": 41.25,
      "learning_rate": 4.7307164721379216e-05,
      "loss": 4.2905,
      "num_input_tokens_seen": 8088496,
      "step": 5965
    },
    {
      "epoch": 0.19108891876320339,
      "grad_norm": 24.25,
      "learning_rate": 4.730273071073743e-05,
      "loss": 3.5896,
      "num_input_tokens_seen": 8095104,
      "step": 5970
    },
    {
      "epoch": 0.19124895973369183,
      "grad_norm": 25.5,
      "learning_rate": 4.729829326073429e-05,
      "loss": 3.1223,
      "num_input_tokens_seen": 8102144,
      "step": 5975
    },
    {
      "epoch": 0.19140900070418027,
      "grad_norm": 23.125,
      "learning_rate": 4.7293852372054126e-05,
      "loss": 3.7314,
      "num_input_tokens_seen": 8109088,
      "step": 5980
    },
    {
      "epoch": 0.1915690416746687,
      "grad_norm": 38.75,
      "learning_rate": 4.728940804538176e-05,
      "loss": 4.1127,
      "num_input_tokens_seen": 8115792,
      "step": 5985
    },
    {
      "epoch": 0.19172908264515717,
      "grad_norm": 36.5,
      "learning_rate": 4.7284960281402556e-05,
      "loss": 4.2568,
      "num_input_tokens_seen": 8122208,
      "step": 5990
    },
    {
      "epoch": 0.1918891236156456,
      "grad_norm": 27.625,
      "learning_rate": 4.728050908080244e-05,
      "loss": 4.5962,
      "num_input_tokens_seen": 8128912,
      "step": 5995
    },
    {
      "epoch": 0.19204916458613405,
      "grad_norm": 32.5,
      "learning_rate": 4.727605444426782e-05,
      "loss": 3.8875,
      "num_input_tokens_seen": 8135936,
      "step": 6000
    },
    {
      "epoch": 0.19204916458613405,
      "eval_loss": 4.490610122680664,
      "eval_runtime": 486.934,
      "eval_samples_per_second": 28.517,
      "eval_steps_per_second": 14.259,
      "num_input_tokens_seen": 8135936,
      "step": 6000
    },
    {
      "epoch": 0.1922092055566225,
      "grad_norm": 33.25,
      "learning_rate": 4.727159637248567e-05,
      "loss": 3.6267,
      "num_input_tokens_seen": 8142464,
      "step": 6005
    },
    {
      "epoch": 0.19236924652711093,
      "grad_norm": 23.875,
      "learning_rate": 4.7267134866143474e-05,
      "loss": 4.0803,
      "num_input_tokens_seen": 8149456,
      "step": 6010
    },
    {
      "epoch": 0.19252928749759937,
      "grad_norm": 27.125,
      "learning_rate": 4.726266992592926e-05,
      "loss": 3.5603,
      "num_input_tokens_seen": 8156192,
      "step": 6015
    },
    {
      "epoch": 0.19268932846808784,
      "grad_norm": 23.25,
      "learning_rate": 4.725820155253157e-05,
      "loss": 3.6621,
      "num_input_tokens_seen": 8162992,
      "step": 6020
    },
    {
      "epoch": 0.19284936943857628,
      "grad_norm": 20.375,
      "learning_rate": 4.725372974663948e-05,
      "loss": 3.3174,
      "num_input_tokens_seen": 8169984,
      "step": 6025
    },
    {
      "epoch": 0.19300941040906472,
      "grad_norm": 29.5,
      "learning_rate": 4.724925450894262e-05,
      "loss": 3.6122,
      "num_input_tokens_seen": 8177008,
      "step": 6030
    },
    {
      "epoch": 0.19316945137955316,
      "grad_norm": 16.625,
      "learning_rate": 4.72447758401311e-05,
      "loss": 3.487,
      "num_input_tokens_seen": 8183408,
      "step": 6035
    },
    {
      "epoch": 0.1933294923500416,
      "grad_norm": 31.0,
      "learning_rate": 4.7240293740895616e-05,
      "loss": 4.5461,
      "num_input_tokens_seen": 8190368,
      "step": 6040
    },
    {
      "epoch": 0.19348953332053007,
      "grad_norm": 15.9375,
      "learning_rate": 4.723580821192733e-05,
      "loss": 4.3724,
      "num_input_tokens_seen": 8197152,
      "step": 6045
    },
    {
      "epoch": 0.1936495742910185,
      "grad_norm": 21.75,
      "learning_rate": 4.7231319253917996e-05,
      "loss": 4.1212,
      "num_input_tokens_seen": 8203920,
      "step": 6050
    },
    {
      "epoch": 0.19380961526150695,
      "grad_norm": 22.125,
      "learning_rate": 4.722682686755986e-05,
      "loss": 3.8605,
      "num_input_tokens_seen": 8210656,
      "step": 6055
    },
    {
      "epoch": 0.19396965623199539,
      "grad_norm": 34.25,
      "learning_rate": 4.722233105354569e-05,
      "loss": 4.0257,
      "num_input_tokens_seen": 8217568,
      "step": 6060
    },
    {
      "epoch": 0.19412969720248383,
      "grad_norm": 20.375,
      "learning_rate": 4.7217831812568815e-05,
      "loss": 3.6936,
      "num_input_tokens_seen": 8224704,
      "step": 6065
    },
    {
      "epoch": 0.1942897381729723,
      "grad_norm": 25.75,
      "learning_rate": 4.721332914532307e-05,
      "loss": 5.1552,
      "num_input_tokens_seen": 8231600,
      "step": 6070
    },
    {
      "epoch": 0.19444977914346073,
      "grad_norm": 40.25,
      "learning_rate": 4.720882305250281e-05,
      "loss": 5.3589,
      "num_input_tokens_seen": 8238032,
      "step": 6075
    },
    {
      "epoch": 0.19460982011394917,
      "grad_norm": 18.875,
      "learning_rate": 4.720431353480295e-05,
      "loss": 4.1426,
      "num_input_tokens_seen": 8244848,
      "step": 6080
    },
    {
      "epoch": 0.1947698610844376,
      "grad_norm": 28.875,
      "learning_rate": 4.719980059291891e-05,
      "loss": 4.3477,
      "num_input_tokens_seen": 8251376,
      "step": 6085
    },
    {
      "epoch": 0.19492990205492605,
      "grad_norm": 28.75,
      "learning_rate": 4.7195284227546634e-05,
      "loss": 3.8707,
      "num_input_tokens_seen": 8258032,
      "step": 6090
    },
    {
      "epoch": 0.19508994302541452,
      "grad_norm": 17.875,
      "learning_rate": 4.7190764439382604e-05,
      "loss": 3.1462,
      "num_input_tokens_seen": 8264784,
      "step": 6095
    },
    {
      "epoch": 0.19524998399590296,
      "grad_norm": 24.375,
      "learning_rate": 4.7186241229123826e-05,
      "loss": 4.1194,
      "num_input_tokens_seen": 8271792,
      "step": 6100
    },
    {
      "epoch": 0.1954100249663914,
      "grad_norm": 28.375,
      "learning_rate": 4.718171459746785e-05,
      "loss": 4.2387,
      "num_input_tokens_seen": 8279008,
      "step": 6105
    },
    {
      "epoch": 0.19557006593687984,
      "grad_norm": 20.25,
      "learning_rate": 4.717718454511273e-05,
      "loss": 4.3928,
      "num_input_tokens_seen": 8285760,
      "step": 6110
    },
    {
      "epoch": 0.19573010690736828,
      "grad_norm": 52.75,
      "learning_rate": 4.7172651072757056e-05,
      "loss": 4.6032,
      "num_input_tokens_seen": 8292768,
      "step": 6115
    },
    {
      "epoch": 0.19589014787785672,
      "grad_norm": 32.5,
      "learning_rate": 4.7168114181099945e-05,
      "loss": 4.8025,
      "num_input_tokens_seen": 8299440,
      "step": 6120
    },
    {
      "epoch": 0.19605018884834519,
      "grad_norm": 21.75,
      "learning_rate": 4.716357387084105e-05,
      "loss": 3.2358,
      "num_input_tokens_seen": 8306288,
      "step": 6125
    },
    {
      "epoch": 0.19621022981883363,
      "grad_norm": 29.375,
      "learning_rate": 4.715903014268054e-05,
      "loss": 4.4766,
      "num_input_tokens_seen": 8313376,
      "step": 6130
    },
    {
      "epoch": 0.19637027078932207,
      "grad_norm": 34.25,
      "learning_rate": 4.715448299731911e-05,
      "loss": 4.4301,
      "num_input_tokens_seen": 8319648,
      "step": 6135
    },
    {
      "epoch": 0.1965303117598105,
      "grad_norm": 27.0,
      "learning_rate": 4.7149932435457986e-05,
      "loss": 4.918,
      "num_input_tokens_seen": 8326592,
      "step": 6140
    },
    {
      "epoch": 0.19669035273029895,
      "grad_norm": 20.25,
      "learning_rate": 4.714537845779894e-05,
      "loss": 3.7924,
      "num_input_tokens_seen": 8333920,
      "step": 6145
    },
    {
      "epoch": 0.1968503937007874,
      "grad_norm": 28.375,
      "learning_rate": 4.714082106504423e-05,
      "loss": 3.308,
      "num_input_tokens_seen": 8340896,
      "step": 6150
    },
    {
      "epoch": 0.19701043467127585,
      "grad_norm": 28.875,
      "learning_rate": 4.713626025789667e-05,
      "loss": 4.5593,
      "num_input_tokens_seen": 8347488,
      "step": 6155
    },
    {
      "epoch": 0.1971704756417643,
      "grad_norm": 36.25,
      "learning_rate": 4.7131696037059606e-05,
      "loss": 3.8587,
      "num_input_tokens_seen": 8354528,
      "step": 6160
    },
    {
      "epoch": 0.19733051661225273,
      "grad_norm": 19.625,
      "learning_rate": 4.712712840323689e-05,
      "loss": 4.2529,
      "num_input_tokens_seen": 8361376,
      "step": 6165
    },
    {
      "epoch": 0.19749055758274117,
      "grad_norm": 32.0,
      "learning_rate": 4.71225573571329e-05,
      "loss": 4.3646,
      "num_input_tokens_seen": 8368176,
      "step": 6170
    },
    {
      "epoch": 0.19765059855322964,
      "grad_norm": 30.0,
      "learning_rate": 4.711798289945256e-05,
      "loss": 4.2389,
      "num_input_tokens_seen": 8375024,
      "step": 6175
    },
    {
      "epoch": 0.19781063952371808,
      "grad_norm": 35.5,
      "learning_rate": 4.71134050309013e-05,
      "loss": 3.6716,
      "num_input_tokens_seen": 8381504,
      "step": 6180
    },
    {
      "epoch": 0.19797068049420652,
      "grad_norm": 27.875,
      "learning_rate": 4.710882375218509e-05,
      "loss": 4.6263,
      "num_input_tokens_seen": 8388160,
      "step": 6185
    },
    {
      "epoch": 0.19813072146469496,
      "grad_norm": 26.75,
      "learning_rate": 4.7104239064010424e-05,
      "loss": 3.8779,
      "num_input_tokens_seen": 8394864,
      "step": 6190
    },
    {
      "epoch": 0.1982907624351834,
      "grad_norm": 33.5,
      "learning_rate": 4.709965096708432e-05,
      "loss": 5.0922,
      "num_input_tokens_seen": 8402048,
      "step": 6195
    },
    {
      "epoch": 0.19845080340567184,
      "grad_norm": 29.375,
      "learning_rate": 4.709505946211431e-05,
      "loss": 4.7084,
      "num_input_tokens_seen": 8408624,
      "step": 6200
    },
    {
      "epoch": 0.19845080340567184,
      "eval_loss": 4.499368667602539,
      "eval_runtime": 487.8212,
      "eval_samples_per_second": 28.465,
      "eval_steps_per_second": 14.233,
      "num_input_tokens_seen": 8408624,
      "step": 6200
    },
    {
      "epoch": 0.1986108443761603,
      "grad_norm": 27.25,
      "learning_rate": 4.709046454980846e-05,
      "loss": 4.8341,
      "num_input_tokens_seen": 8415440,
      "step": 6205
    },
    {
      "epoch": 0.19877088534664875,
      "grad_norm": 23.625,
      "learning_rate": 4.708586623087538e-05,
      "loss": 5.5311,
      "num_input_tokens_seen": 8421888,
      "step": 6210
    },
    {
      "epoch": 0.19893092631713719,
      "grad_norm": 26.25,
      "learning_rate": 4.708126450602418e-05,
      "loss": 3.9823,
      "num_input_tokens_seen": 8428560,
      "step": 6215
    },
    {
      "epoch": 0.19909096728762563,
      "grad_norm": 34.25,
      "learning_rate": 4.7076659375964495e-05,
      "loss": 3.5745,
      "num_input_tokens_seen": 8435520,
      "step": 6220
    },
    {
      "epoch": 0.19925100825811407,
      "grad_norm": 17.125,
      "learning_rate": 4.707205084140651e-05,
      "loss": 4.0274,
      "num_input_tokens_seen": 8442336,
      "step": 6225
    },
    {
      "epoch": 0.19941104922860253,
      "grad_norm": 13.0625,
      "learning_rate": 4.7067438903060904e-05,
      "loss": 4.2417,
      "num_input_tokens_seen": 8449056,
      "step": 6230
    },
    {
      "epoch": 0.19957109019909097,
      "grad_norm": 51.25,
      "learning_rate": 4.70628235616389e-05,
      "loss": 6.1243,
      "num_input_tokens_seen": 8455632,
      "step": 6235
    },
    {
      "epoch": 0.1997311311695794,
      "grad_norm": 19.375,
      "learning_rate": 4.7058204817852256e-05,
      "loss": 3.6683,
      "num_input_tokens_seen": 8462864,
      "step": 6240
    },
    {
      "epoch": 0.19989117214006785,
      "grad_norm": 41.75,
      "learning_rate": 4.705358267241322e-05,
      "loss": 5.3675,
      "num_input_tokens_seen": 8469520,
      "step": 6245
    },
    {
      "epoch": 0.2000512131105563,
      "grad_norm": 33.75,
      "learning_rate": 4.704895712603459e-05,
      "loss": 4.2464,
      "num_input_tokens_seen": 8475936,
      "step": 6250
    },
    {
      "epoch": 0.20021125408104476,
      "grad_norm": 14.375,
      "learning_rate": 4.704432817942969e-05,
      "loss": 3.9577,
      "num_input_tokens_seen": 8482928,
      "step": 6255
    },
    {
      "epoch": 0.2003712950515332,
      "grad_norm": 58.25,
      "learning_rate": 4.703969583331236e-05,
      "loss": 5.1744,
      "num_input_tokens_seen": 8489536,
      "step": 6260
    },
    {
      "epoch": 0.20053133602202164,
      "grad_norm": 17.125,
      "learning_rate": 4.7035060088396965e-05,
      "loss": 4.1553,
      "num_input_tokens_seen": 8496400,
      "step": 6265
    },
    {
      "epoch": 0.20069137699251008,
      "grad_norm": 25.625,
      "learning_rate": 4.703042094539839e-05,
      "loss": 3.6884,
      "num_input_tokens_seen": 8502784,
      "step": 6270
    },
    {
      "epoch": 0.20085141796299852,
      "grad_norm": 21.125,
      "learning_rate": 4.702577840503206e-05,
      "loss": 3.6835,
      "num_input_tokens_seen": 8509232,
      "step": 6275
    },
    {
      "epoch": 0.201011458933487,
      "grad_norm": 17.625,
      "learning_rate": 4.70211324680139e-05,
      "loss": 3.5627,
      "num_input_tokens_seen": 8515904,
      "step": 6280
    },
    {
      "epoch": 0.20117149990397543,
      "grad_norm": 33.0,
      "learning_rate": 4.7016483135060386e-05,
      "loss": 3.0798,
      "num_input_tokens_seen": 8522736,
      "step": 6285
    },
    {
      "epoch": 0.20133154087446387,
      "grad_norm": 29.625,
      "learning_rate": 4.701183040688849e-05,
      "loss": 5.0861,
      "num_input_tokens_seen": 8529232,
      "step": 6290
    },
    {
      "epoch": 0.2014915818449523,
      "grad_norm": 26.5,
      "learning_rate": 4.700717428421573e-05,
      "loss": 4.1342,
      "num_input_tokens_seen": 8536080,
      "step": 6295
    },
    {
      "epoch": 0.20165162281544075,
      "grad_norm": 37.5,
      "learning_rate": 4.700251476776014e-05,
      "loss": 4.56,
      "num_input_tokens_seen": 8542464,
      "step": 6300
    },
    {
      "epoch": 0.20181166378592919,
      "grad_norm": 26.125,
      "learning_rate": 4.699785185824026e-05,
      "loss": 3.5589,
      "num_input_tokens_seen": 8549408,
      "step": 6305
    },
    {
      "epoch": 0.20197170475641765,
      "grad_norm": 37.5,
      "learning_rate": 4.699318555637519e-05,
      "loss": 5.0139,
      "num_input_tokens_seen": 8556208,
      "step": 6310
    },
    {
      "epoch": 0.2021317457269061,
      "grad_norm": 19.75,
      "learning_rate": 4.6988515862884525e-05,
      "loss": 4.5774,
      "num_input_tokens_seen": 8563040,
      "step": 6315
    },
    {
      "epoch": 0.20229178669739453,
      "grad_norm": 25.25,
      "learning_rate": 4.698384277848838e-05,
      "loss": 4.8149,
      "num_input_tokens_seen": 8569712,
      "step": 6320
    },
    {
      "epoch": 0.20245182766788297,
      "grad_norm": 26.125,
      "learning_rate": 4.6979166303907425e-05,
      "loss": 3.7535,
      "num_input_tokens_seen": 8576752,
      "step": 6325
    },
    {
      "epoch": 0.2026118686383714,
      "grad_norm": 33.25,
      "learning_rate": 4.697448643986281e-05,
      "loss": 3.446,
      "num_input_tokens_seen": 8583632,
      "step": 6330
    },
    {
      "epoch": 0.20277190960885988,
      "grad_norm": 29.625,
      "learning_rate": 4.696980318707624e-05,
      "loss": 5.5927,
      "num_input_tokens_seen": 8590320,
      "step": 6335
    },
    {
      "epoch": 0.20293195057934832,
      "grad_norm": 21.125,
      "learning_rate": 4.6965116546269924e-05,
      "loss": 4.4026,
      "num_input_tokens_seen": 8596992,
      "step": 6340
    },
    {
      "epoch": 0.20309199154983676,
      "grad_norm": 22.5,
      "learning_rate": 4.6960426518166615e-05,
      "loss": 4.6622,
      "num_input_tokens_seen": 8603760,
      "step": 6345
    },
    {
      "epoch": 0.2032520325203252,
      "grad_norm": 16.75,
      "learning_rate": 4.6955733103489556e-05,
      "loss": 4.0451,
      "num_input_tokens_seen": 8610448,
      "step": 6350
    },
    {
      "epoch": 0.20341207349081364,
      "grad_norm": 28.375,
      "learning_rate": 4.695103630296255e-05,
      "loss": 4.9537,
      "num_input_tokens_seen": 8617488,
      "step": 6355
    },
    {
      "epoch": 0.2035721144613021,
      "grad_norm": 28.375,
      "learning_rate": 4.694633611730988e-05,
      "loss": 3.9834,
      "num_input_tokens_seen": 8624096,
      "step": 6360
    },
    {
      "epoch": 0.20373215543179055,
      "grad_norm": 18.625,
      "learning_rate": 4.694163254725639e-05,
      "loss": 4.374,
      "num_input_tokens_seen": 8630688,
      "step": 6365
    },
    {
      "epoch": 0.20389219640227899,
      "grad_norm": 25.375,
      "learning_rate": 4.693692559352743e-05,
      "loss": 4.8135,
      "num_input_tokens_seen": 8637216,
      "step": 6370
    },
    {
      "epoch": 0.20405223737276743,
      "grad_norm": 41.25,
      "learning_rate": 4.693221525684886e-05,
      "loss": 4.0846,
      "num_input_tokens_seen": 8644224,
      "step": 6375
    },
    {
      "epoch": 0.20421227834325587,
      "grad_norm": 20.5,
      "learning_rate": 4.6927501537947084e-05,
      "loss": 4.2133,
      "num_input_tokens_seen": 8651088,
      "step": 6380
    },
    {
      "epoch": 0.2043723193137443,
      "grad_norm": 36.0,
      "learning_rate": 4.692278443754901e-05,
      "loss": 4.8452,
      "num_input_tokens_seen": 8657680,
      "step": 6385
    },
    {
      "epoch": 0.20453236028423277,
      "grad_norm": 27.75,
      "learning_rate": 4.691806395638208e-05,
      "loss": 4.9199,
      "num_input_tokens_seen": 8664560,
      "step": 6390
    },
    {
      "epoch": 0.2046924012547212,
      "grad_norm": 35.25,
      "learning_rate": 4.6913340095174255e-05,
      "loss": 3.6547,
      "num_input_tokens_seen": 8671136,
      "step": 6395
    },
    {
      "epoch": 0.20485244222520965,
      "grad_norm": 18.5,
      "learning_rate": 4.690861285465399e-05,
      "loss": 3.6507,
      "num_input_tokens_seen": 8677888,
      "step": 6400
    },
    {
      "epoch": 0.20485244222520965,
      "eval_loss": 4.485253810882568,
      "eval_runtime": 487.8717,
      "eval_samples_per_second": 28.462,
      "eval_steps_per_second": 14.231,
      "num_input_tokens_seen": 8677888,
      "step": 6400
    },
    {
      "epoch": 0.2050124831956981,
      "grad_norm": 33.25,
      "learning_rate": 4.690388223555031e-05,
      "loss": 3.9879,
      "num_input_tokens_seen": 8684240,
      "step": 6405
    },
    {
      "epoch": 0.20517252416618653,
      "grad_norm": 17.25,
      "learning_rate": 4.689914823859273e-05,
      "loss": 4.7873,
      "num_input_tokens_seen": 8691184,
      "step": 6410
    },
    {
      "epoch": 0.205332565136675,
      "grad_norm": 39.75,
      "learning_rate": 4.689441086451129e-05,
      "loss": 3.4342,
      "num_input_tokens_seen": 8698096,
      "step": 6415
    },
    {
      "epoch": 0.20549260610716344,
      "grad_norm": 27.75,
      "learning_rate": 4.688967011403655e-05,
      "loss": 2.8253,
      "num_input_tokens_seen": 8705024,
      "step": 6420
    },
    {
      "epoch": 0.20565264707765188,
      "grad_norm": 24.25,
      "learning_rate": 4.68849259878996e-05,
      "loss": 4.1888,
      "num_input_tokens_seen": 8711520,
      "step": 6425
    },
    {
      "epoch": 0.20581268804814032,
      "grad_norm": 19.25,
      "learning_rate": 4.6880178486832036e-05,
      "loss": 3.6314,
      "num_input_tokens_seen": 8718208,
      "step": 6430
    },
    {
      "epoch": 0.20597272901862876,
      "grad_norm": 22.125,
      "learning_rate": 4.687542761156598e-05,
      "loss": 3.9445,
      "num_input_tokens_seen": 8724592,
      "step": 6435
    },
    {
      "epoch": 0.20613276998911723,
      "grad_norm": 43.0,
      "learning_rate": 4.6870673362834096e-05,
      "loss": 4.6634,
      "num_input_tokens_seen": 8731760,
      "step": 6440
    },
    {
      "epoch": 0.20629281095960567,
      "grad_norm": 21.0,
      "learning_rate": 4.6865915741369526e-05,
      "loss": 3.5823,
      "num_input_tokens_seen": 8738352,
      "step": 6445
    },
    {
      "epoch": 0.2064528519300941,
      "grad_norm": 31.375,
      "learning_rate": 4.686115474790597e-05,
      "loss": 4.3423,
      "num_input_tokens_seen": 8745376,
      "step": 6450
    },
    {
      "epoch": 0.20661289290058255,
      "grad_norm": 14.375,
      "learning_rate": 4.685639038317762e-05,
      "loss": 4.5133,
      "num_input_tokens_seen": 8752304,
      "step": 6455
    },
    {
      "epoch": 0.20677293387107099,
      "grad_norm": 41.0,
      "learning_rate": 4.685162264791921e-05,
      "loss": 3.2839,
      "num_input_tokens_seen": 8758784,
      "step": 6460
    },
    {
      "epoch": 0.20693297484155945,
      "grad_norm": 34.5,
      "learning_rate": 4.684685154286599e-05,
      "loss": 4.1022,
      "num_input_tokens_seen": 8765728,
      "step": 6465
    },
    {
      "epoch": 0.2070930158120479,
      "grad_norm": 36.5,
      "learning_rate": 4.684207706875371e-05,
      "loss": 5.0715,
      "num_input_tokens_seen": 8772224,
      "step": 6470
    },
    {
      "epoch": 0.20725305678253633,
      "grad_norm": 32.0,
      "learning_rate": 4.683729922631866e-05,
      "loss": 5.7111,
      "num_input_tokens_seen": 8778688,
      "step": 6475
    },
    {
      "epoch": 0.20741309775302477,
      "grad_norm": 35.0,
      "learning_rate": 4.683251801629765e-05,
      "loss": 4.8612,
      "num_input_tokens_seen": 8785312,
      "step": 6480
    },
    {
      "epoch": 0.2075731387235132,
      "grad_norm": 21.75,
      "learning_rate": 4.6827733439428e-05,
      "loss": 4.6457,
      "num_input_tokens_seen": 8791968,
      "step": 6485
    },
    {
      "epoch": 0.20773317969400165,
      "grad_norm": 23.75,
      "learning_rate": 4.682294549644754e-05,
      "loss": 5.0327,
      "num_input_tokens_seen": 8798784,
      "step": 6490
    },
    {
      "epoch": 0.20789322066449012,
      "grad_norm": 23.75,
      "learning_rate": 4.681815418809464e-05,
      "loss": 4.3472,
      "num_input_tokens_seen": 8805728,
      "step": 6495
    },
    {
      "epoch": 0.20805326163497856,
      "grad_norm": 19.375,
      "learning_rate": 4.681335951510819e-05,
      "loss": 3.6427,
      "num_input_tokens_seen": 8812416,
      "step": 6500
    },
    {
      "epoch": 0.208213302605467,
      "grad_norm": 49.25,
      "learning_rate": 4.6808561478227576e-05,
      "loss": 4.4811,
      "num_input_tokens_seen": 8819312,
      "step": 6505
    },
    {
      "epoch": 0.20837334357595544,
      "grad_norm": 37.5,
      "learning_rate": 4.680376007819271e-05,
      "loss": 5.057,
      "num_input_tokens_seen": 8825776,
      "step": 6510
    },
    {
      "epoch": 0.20853338454644388,
      "grad_norm": 19.875,
      "learning_rate": 4.679895531574405e-05,
      "loss": 4.4693,
      "num_input_tokens_seen": 8832384,
      "step": 6515
    },
    {
      "epoch": 0.20869342551693235,
      "grad_norm": 32.75,
      "learning_rate": 4.679414719162253e-05,
      "loss": 4.0698,
      "num_input_tokens_seen": 8839040,
      "step": 6520
    },
    {
      "epoch": 0.20885346648742079,
      "grad_norm": 36.0,
      "learning_rate": 4.6789335706569635e-05,
      "loss": 3.8198,
      "num_input_tokens_seen": 8845696,
      "step": 6525
    },
    {
      "epoch": 0.20901350745790923,
      "grad_norm": 39.25,
      "learning_rate": 4.678452086132734e-05,
      "loss": 4.4376,
      "num_input_tokens_seen": 8852352,
      "step": 6530
    },
    {
      "epoch": 0.20917354842839767,
      "grad_norm": 24.75,
      "learning_rate": 4.677970265663818e-05,
      "loss": 4.2597,
      "num_input_tokens_seen": 8858944,
      "step": 6535
    },
    {
      "epoch": 0.2093335893988861,
      "grad_norm": 33.25,
      "learning_rate": 4.677488109324517e-05,
      "loss": 3.7035,
      "num_input_tokens_seen": 8865808,
      "step": 6540
    },
    {
      "epoch": 0.20949363036937457,
      "grad_norm": 23.25,
      "learning_rate": 4.6770056171891846e-05,
      "loss": 4.4214,
      "num_input_tokens_seen": 8873168,
      "step": 6545
    },
    {
      "epoch": 0.209653671339863,
      "grad_norm": 38.0,
      "learning_rate": 4.6765227893322286e-05,
      "loss": 3.7824,
      "num_input_tokens_seen": 8879984,
      "step": 6550
    },
    {
      "epoch": 0.20981371231035145,
      "grad_norm": 29.5,
      "learning_rate": 4.676039625828107e-05,
      "loss": 4.7801,
      "num_input_tokens_seen": 8886768,
      "step": 6555
    },
    {
      "epoch": 0.2099737532808399,
      "grad_norm": 15.75,
      "learning_rate": 4.675556126751328e-05,
      "loss": 4.7943,
      "num_input_tokens_seen": 8893664,
      "step": 6560
    },
    {
      "epoch": 0.21013379425132833,
      "grad_norm": 18.5,
      "learning_rate": 4.6750722921764556e-05,
      "loss": 4.0718,
      "num_input_tokens_seen": 8900592,
      "step": 6565
    },
    {
      "epoch": 0.21029383522181677,
      "grad_norm": 18.625,
      "learning_rate": 4.674588122178102e-05,
      "loss": 4.4983,
      "num_input_tokens_seen": 8907104,
      "step": 6570
    },
    {
      "epoch": 0.21045387619230524,
      "grad_norm": 55.5,
      "learning_rate": 4.674103616830931e-05,
      "loss": 3.9732,
      "num_input_tokens_seen": 8913856,
      "step": 6575
    },
    {
      "epoch": 0.21061391716279368,
      "grad_norm": 14.75,
      "learning_rate": 4.673618776209663e-05,
      "loss": 3.8891,
      "num_input_tokens_seen": 8920384,
      "step": 6580
    },
    {
      "epoch": 0.21077395813328212,
      "grad_norm": 18.875,
      "learning_rate": 4.673133600389063e-05,
      "loss": 4.5162,
      "num_input_tokens_seen": 8926816,
      "step": 6585
    },
    {
      "epoch": 0.21093399910377056,
      "grad_norm": 31.0,
      "learning_rate": 4.672648089443953e-05,
      "loss": 4.9048,
      "num_input_tokens_seen": 8933424,
      "step": 6590
    },
    {
      "epoch": 0.211094040074259,
      "grad_norm": 25.75,
      "learning_rate": 4.672162243449204e-05,
      "loss": 4.1864,
      "num_input_tokens_seen": 8940096,
      "step": 6595
    },
    {
      "epoch": 0.21125408104474747,
      "grad_norm": 26.75,
      "learning_rate": 4.67167606247974e-05,
      "loss": 4.6074,
      "num_input_tokens_seen": 8947120,
      "step": 6600
    },
    {
      "epoch": 0.21125408104474747,
      "eval_loss": 4.488213539123535,
      "eval_runtime": 486.9394,
      "eval_samples_per_second": 28.517,
      "eval_steps_per_second": 14.258,
      "num_input_tokens_seen": 8947120,
      "step": 6600
    },
    {
      "epoch": 0.2114141220152359,
      "grad_norm": 32.75,
      "learning_rate": 4.671189546610536e-05,
      "loss": 4.0762,
      "num_input_tokens_seen": 8954176,
      "step": 6605
    },
    {
      "epoch": 0.21157416298572435,
      "grad_norm": 35.5,
      "learning_rate": 4.67070269591662e-05,
      "loss": 4.6447,
      "num_input_tokens_seen": 8960528,
      "step": 6610
    },
    {
      "epoch": 0.21173420395621279,
      "grad_norm": 28.375,
      "learning_rate": 4.670215510473068e-05,
      "loss": 5.0551,
      "num_input_tokens_seen": 8967152,
      "step": 6615
    },
    {
      "epoch": 0.21189424492670123,
      "grad_norm": 34.75,
      "learning_rate": 4.669727990355013e-05,
      "loss": 5.331,
      "num_input_tokens_seen": 8973808,
      "step": 6620
    },
    {
      "epoch": 0.2120542858971897,
      "grad_norm": 20.5,
      "learning_rate": 4.669240135637635e-05,
      "loss": 4.2938,
      "num_input_tokens_seen": 8980352,
      "step": 6625
    },
    {
      "epoch": 0.21221432686767813,
      "grad_norm": 45.75,
      "learning_rate": 4.6687519463961675e-05,
      "loss": 4.8072,
      "num_input_tokens_seen": 8986928,
      "step": 6630
    },
    {
      "epoch": 0.21237436783816657,
      "grad_norm": 32.0,
      "learning_rate": 4.668263422705896e-05,
      "loss": 3.71,
      "num_input_tokens_seen": 8993472,
      "step": 6635
    },
    {
      "epoch": 0.212534408808655,
      "grad_norm": 26.375,
      "learning_rate": 4.667774564642156e-05,
      "loss": 4.0462,
      "num_input_tokens_seen": 9000320,
      "step": 6640
    },
    {
      "epoch": 0.21269444977914345,
      "grad_norm": 52.25,
      "learning_rate": 4.6672853722803365e-05,
      "loss": 6.2584,
      "num_input_tokens_seen": 9007040,
      "step": 6645
    },
    {
      "epoch": 0.21285449074963192,
      "grad_norm": 22.875,
      "learning_rate": 4.666795845695877e-05,
      "loss": 4.3074,
      "num_input_tokens_seen": 9013664,
      "step": 6650
    },
    {
      "epoch": 0.21301453172012036,
      "grad_norm": 17.75,
      "learning_rate": 4.666305984964269e-05,
      "loss": 3.8565,
      "num_input_tokens_seen": 9020624,
      "step": 6655
    },
    {
      "epoch": 0.2131745726906088,
      "grad_norm": 43.0,
      "learning_rate": 4.6658157901610535e-05,
      "loss": 4.7825,
      "num_input_tokens_seen": 9027376,
      "step": 6660
    },
    {
      "epoch": 0.21333461366109724,
      "grad_norm": 18.625,
      "learning_rate": 4.665325261361826e-05,
      "loss": 4.4871,
      "num_input_tokens_seen": 9033696,
      "step": 6665
    },
    {
      "epoch": 0.21349465463158568,
      "grad_norm": 17.5,
      "learning_rate": 4.664834398642232e-05,
      "loss": 4.427,
      "num_input_tokens_seen": 9039936,
      "step": 6670
    },
    {
      "epoch": 0.21365469560207412,
      "grad_norm": 15.25,
      "learning_rate": 4.6643432020779686e-05,
      "loss": 5.4005,
      "num_input_tokens_seen": 9047056,
      "step": 6675
    },
    {
      "epoch": 0.21381473657256259,
      "grad_norm": 16.75,
      "learning_rate": 4.663851671744786e-05,
      "loss": 3.4596,
      "num_input_tokens_seen": 9054464,
      "step": 6680
    },
    {
      "epoch": 0.21397477754305103,
      "grad_norm": 33.5,
      "learning_rate": 4.6633598077184815e-05,
      "loss": 4.8485,
      "num_input_tokens_seen": 9061040,
      "step": 6685
    },
    {
      "epoch": 0.21413481851353947,
      "grad_norm": 41.75,
      "learning_rate": 4.662867610074908e-05,
      "loss": 4.2204,
      "num_input_tokens_seen": 9067776,
      "step": 6690
    },
    {
      "epoch": 0.2142948594840279,
      "grad_norm": 26.625,
      "learning_rate": 4.6623750788899696e-05,
      "loss": 4.463,
      "num_input_tokens_seen": 9074576,
      "step": 6695
    },
    {
      "epoch": 0.21445490045451635,
      "grad_norm": 15.75,
      "learning_rate": 4.6618822142396195e-05,
      "loss": 4.6316,
      "num_input_tokens_seen": 9081312,
      "step": 6700
    },
    {
      "epoch": 0.2146149414250048,
      "grad_norm": 23.125,
      "learning_rate": 4.661389016199864e-05,
      "loss": 4.2496,
      "num_input_tokens_seen": 9088144,
      "step": 6705
    },
    {
      "epoch": 0.21477498239549325,
      "grad_norm": 37.0,
      "learning_rate": 4.660895484846761e-05,
      "loss": 4.4822,
      "num_input_tokens_seen": 9095104,
      "step": 6710
    },
    {
      "epoch": 0.2149350233659817,
      "grad_norm": 21.875,
      "learning_rate": 4.660401620256418e-05,
      "loss": 3.8268,
      "num_input_tokens_seen": 9101808,
      "step": 6715
    },
    {
      "epoch": 0.21509506433647013,
      "grad_norm": 18.75,
      "learning_rate": 4.659907422504997e-05,
      "loss": 4.3527,
      "num_input_tokens_seen": 9109088,
      "step": 6720
    },
    {
      "epoch": 0.21525510530695857,
      "grad_norm": 28.875,
      "learning_rate": 4.6594128916687074e-05,
      "loss": 3.7215,
      "num_input_tokens_seen": 9115504,
      "step": 6725
    },
    {
      "epoch": 0.21541514627744704,
      "grad_norm": 45.0,
      "learning_rate": 4.658918027823813e-05,
      "loss": 3.1181,
      "num_input_tokens_seen": 9122144,
      "step": 6730
    },
    {
      "epoch": 0.21557518724793548,
      "grad_norm": 27.5,
      "learning_rate": 4.658422831046628e-05,
      "loss": 5.2128,
      "num_input_tokens_seen": 9129008,
      "step": 6735
    },
    {
      "epoch": 0.21573522821842392,
      "grad_norm": 35.25,
      "learning_rate": 4.657927301413518e-05,
      "loss": 4.9361,
      "num_input_tokens_seen": 9135520,
      "step": 6740
    },
    {
      "epoch": 0.21589526918891236,
      "grad_norm": 37.5,
      "learning_rate": 4.657431439000901e-05,
      "loss": 3.7383,
      "num_input_tokens_seen": 9142336,
      "step": 6745
    },
    {
      "epoch": 0.2160553101594008,
      "grad_norm": 38.25,
      "learning_rate": 4.656935243885243e-05,
      "loss": 4.3987,
      "num_input_tokens_seen": 9148576,
      "step": 6750
    },
    {
      "epoch": 0.21621535112988924,
      "grad_norm": 26.125,
      "learning_rate": 4.656438716143066e-05,
      "loss": 4.4299,
      "num_input_tokens_seen": 9155488,
      "step": 6755
    },
    {
      "epoch": 0.2163753921003777,
      "grad_norm": 15.9375,
      "learning_rate": 4.6559418558509384e-05,
      "loss": 5.2126,
      "num_input_tokens_seen": 9162208,
      "step": 6760
    },
    {
      "epoch": 0.21653543307086615,
      "grad_norm": 35.25,
      "learning_rate": 4.6554446630854833e-05,
      "loss": 3.4551,
      "num_input_tokens_seen": 9168912,
      "step": 6765
    },
    {
      "epoch": 0.21669547404135459,
      "grad_norm": 42.25,
      "learning_rate": 4.654947137923374e-05,
      "loss": 4.6757,
      "num_input_tokens_seen": 9175328,
      "step": 6770
    },
    {
      "epoch": 0.21685551501184303,
      "grad_norm": 23.0,
      "learning_rate": 4.654449280441335e-05,
      "loss": 3.6151,
      "num_input_tokens_seen": 9182464,
      "step": 6775
    },
    {
      "epoch": 0.21701555598233147,
      "grad_norm": 29.0,
      "learning_rate": 4.653951090716143e-05,
      "loss": 3.3092,
      "num_input_tokens_seen": 9189456,
      "step": 6780
    },
    {
      "epoch": 0.21717559695281993,
      "grad_norm": 29.875,
      "learning_rate": 4.653452568824625e-05,
      "loss": 3.3658,
      "num_input_tokens_seen": 9196080,
      "step": 6785
    },
    {
      "epoch": 0.21733563792330837,
      "grad_norm": 23.0,
      "learning_rate": 4.6529537148436585e-05,
      "loss": 4.3841,
      "num_input_tokens_seen": 9202736,
      "step": 6790
    },
    {
      "epoch": 0.2174956788937968,
      "grad_norm": 32.75,
      "learning_rate": 4.6524545288501734e-05,
      "loss": 3.9999,
      "num_input_tokens_seen": 9209472,
      "step": 6795
    },
    {
      "epoch": 0.21765571986428525,
      "grad_norm": 47.0,
      "learning_rate": 4.6519550109211506e-05,
      "loss": 3.9965,
      "num_input_tokens_seen": 9216336,
      "step": 6800
    },
    {
      "epoch": 0.21765571986428525,
      "eval_loss": 4.497682094573975,
      "eval_runtime": 487.6978,
      "eval_samples_per_second": 28.473,
      "eval_steps_per_second": 14.236,
      "num_input_tokens_seen": 9216336,
      "step": 6800
    },
    {
      "epoch": 0.2178157608347737,
      "grad_norm": 25.625,
      "learning_rate": 4.651455161133622e-05,
      "loss": 4.4166,
      "num_input_tokens_seen": 9222688,
      "step": 6805
    },
    {
      "epoch": 0.21797580180526216,
      "grad_norm": 36.75,
      "learning_rate": 4.6509549795646704e-05,
      "loss": 3.5613,
      "num_input_tokens_seen": 9229472,
      "step": 6810
    },
    {
      "epoch": 0.2181358427757506,
      "grad_norm": 44.25,
      "learning_rate": 4.6504544662914306e-05,
      "loss": 4.6715,
      "num_input_tokens_seen": 9236128,
      "step": 6815
    },
    {
      "epoch": 0.21829588374623904,
      "grad_norm": 32.75,
      "learning_rate": 4.6499536213910876e-05,
      "loss": 3.9094,
      "num_input_tokens_seen": 9242832,
      "step": 6820
    },
    {
      "epoch": 0.21845592471672748,
      "grad_norm": 27.0,
      "learning_rate": 4.6494524449408786e-05,
      "loss": 4.0543,
      "num_input_tokens_seen": 9249408,
      "step": 6825
    },
    {
      "epoch": 0.21861596568721592,
      "grad_norm": 33.5,
      "learning_rate": 4.6489509370180903e-05,
      "loss": 4.3521,
      "num_input_tokens_seen": 9255968,
      "step": 6830
    },
    {
      "epoch": 0.21877600665770439,
      "grad_norm": 30.5,
      "learning_rate": 4.648449097700063e-05,
      "loss": 4.8862,
      "num_input_tokens_seen": 9263136,
      "step": 6835
    },
    {
      "epoch": 0.21893604762819283,
      "grad_norm": 29.875,
      "learning_rate": 4.647946927064185e-05,
      "loss": 4.7988,
      "num_input_tokens_seen": 9269744,
      "step": 6840
    },
    {
      "epoch": 0.21909608859868127,
      "grad_norm": 48.75,
      "learning_rate": 4.647444425187898e-05,
      "loss": 4.4673,
      "num_input_tokens_seen": 9276560,
      "step": 6845
    },
    {
      "epoch": 0.2192561295691697,
      "grad_norm": 39.5,
      "learning_rate": 4.646941592148695e-05,
      "loss": 4.5849,
      "num_input_tokens_seen": 9283408,
      "step": 6850
    },
    {
      "epoch": 0.21941617053965815,
      "grad_norm": 33.5,
      "learning_rate": 4.646438428024117e-05,
      "loss": 4.1307,
      "num_input_tokens_seen": 9290560,
      "step": 6855
    },
    {
      "epoch": 0.21957621151014659,
      "grad_norm": 25.5,
      "learning_rate": 4.64593493289176e-05,
      "loss": 5.0019,
      "num_input_tokens_seen": 9297168,
      "step": 6860
    },
    {
      "epoch": 0.21973625248063505,
      "grad_norm": 27.5,
      "learning_rate": 4.64543110682927e-05,
      "loss": 3.9346,
      "num_input_tokens_seen": 9303504,
      "step": 6865
    },
    {
      "epoch": 0.2198962934511235,
      "grad_norm": 44.25,
      "learning_rate": 4.644926949914341e-05,
      "loss": 4.7093,
      "num_input_tokens_seen": 9310400,
      "step": 6870
    },
    {
      "epoch": 0.22005633442161193,
      "grad_norm": 26.0,
      "learning_rate": 4.644422462224722e-05,
      "loss": 4.5093,
      "num_input_tokens_seen": 9317168,
      "step": 6875
    },
    {
      "epoch": 0.22021637539210037,
      "grad_norm": 32.25,
      "learning_rate": 4.643917643838211e-05,
      "loss": 4.6978,
      "num_input_tokens_seen": 9324176,
      "step": 6880
    },
    {
      "epoch": 0.2203764163625888,
      "grad_norm": 28.875,
      "learning_rate": 4.6434124948326564e-05,
      "loss": 4.4034,
      "num_input_tokens_seen": 9331040,
      "step": 6885
    },
    {
      "epoch": 0.22053645733307728,
      "grad_norm": 31.375,
      "learning_rate": 4.6429070152859594e-05,
      "loss": 4.2831,
      "num_input_tokens_seen": 9337712,
      "step": 6890
    },
    {
      "epoch": 0.22069649830356572,
      "grad_norm": 36.5,
      "learning_rate": 4.6424012052760714e-05,
      "loss": 4.3808,
      "num_input_tokens_seen": 9344624,
      "step": 6895
    },
    {
      "epoch": 0.22085653927405416,
      "grad_norm": 26.0,
      "learning_rate": 4.6418950648809945e-05,
      "loss": 4.2145,
      "num_input_tokens_seen": 9351104,
      "step": 6900
    },
    {
      "epoch": 0.2210165802445426,
      "grad_norm": 40.0,
      "learning_rate": 4.641388594178782e-05,
      "loss": 4.4788,
      "num_input_tokens_seen": 9358160,
      "step": 6905
    },
    {
      "epoch": 0.22117662121503104,
      "grad_norm": 20.25,
      "learning_rate": 4.640881793247538e-05,
      "loss": 4.717,
      "num_input_tokens_seen": 9364592,
      "step": 6910
    },
    {
      "epoch": 0.2213366621855195,
      "grad_norm": 23.625,
      "learning_rate": 4.6403746621654173e-05,
      "loss": 4.7155,
      "num_input_tokens_seen": 9371312,
      "step": 6915
    },
    {
      "epoch": 0.22149670315600795,
      "grad_norm": 29.375,
      "learning_rate": 4.639867201010626e-05,
      "loss": 3.296,
      "num_input_tokens_seen": 9377792,
      "step": 6920
    },
    {
      "epoch": 0.22165674412649639,
      "grad_norm": 31.875,
      "learning_rate": 4.6393594098614204e-05,
      "loss": 4.6354,
      "num_input_tokens_seen": 9384592,
      "step": 6925
    },
    {
      "epoch": 0.22181678509698483,
      "grad_norm": 30.75,
      "learning_rate": 4.63885128879611e-05,
      "loss": 4.0053,
      "num_input_tokens_seen": 9391424,
      "step": 6930
    },
    {
      "epoch": 0.22197682606747327,
      "grad_norm": 26.25,
      "learning_rate": 4.638342837893052e-05,
      "loss": 4.8063,
      "num_input_tokens_seen": 9398464,
      "step": 6935
    },
    {
      "epoch": 0.2221368670379617,
      "grad_norm": 40.25,
      "learning_rate": 4.6378340572306565e-05,
      "loss": 3.9236,
      "num_input_tokens_seen": 9405488,
      "step": 6940
    },
    {
      "epoch": 0.22229690800845017,
      "grad_norm": 36.5,
      "learning_rate": 4.6373249468873833e-05,
      "loss": 4.3132,
      "num_input_tokens_seen": 9412304,
      "step": 6945
    },
    {
      "epoch": 0.2224569489789386,
      "grad_norm": 22.0,
      "learning_rate": 4.636815506941744e-05,
      "loss": 4.5488,
      "num_input_tokens_seen": 9418976,
      "step": 6950
    },
    {
      "epoch": 0.22261698994942705,
      "grad_norm": 34.25,
      "learning_rate": 4.6363057374723004e-05,
      "loss": 3.8418,
      "num_input_tokens_seen": 9425472,
      "step": 6955
    },
    {
      "epoch": 0.2227770309199155,
      "grad_norm": 25.25,
      "learning_rate": 4.635795638557666e-05,
      "loss": 3.4371,
      "num_input_tokens_seen": 9431920,
      "step": 6960
    },
    {
      "epoch": 0.22293707189040393,
      "grad_norm": 17.0,
      "learning_rate": 4.635285210276504e-05,
      "loss": 3.9929,
      "num_input_tokens_seen": 9438928,
      "step": 6965
    },
    {
      "epoch": 0.2230971128608924,
      "grad_norm": 32.75,
      "learning_rate": 4.6347744527075295e-05,
      "loss": 3.6772,
      "num_input_tokens_seen": 9445680,
      "step": 6970
    },
    {
      "epoch": 0.22325715383138084,
      "grad_norm": 25.875,
      "learning_rate": 4.634263365929506e-05,
      "loss": 4.1547,
      "num_input_tokens_seen": 9452272,
      "step": 6975
    },
    {
      "epoch": 0.22341719480186928,
      "grad_norm": 29.625,
      "learning_rate": 4.6337519500212515e-05,
      "loss": 3.3522,
      "num_input_tokens_seen": 9458672,
      "step": 6980
    },
    {
      "epoch": 0.22357723577235772,
      "grad_norm": 42.75,
      "learning_rate": 4.633240205061632e-05,
      "loss": 4.5186,
      "num_input_tokens_seen": 9465168,
      "step": 6985
    },
    {
      "epoch": 0.22373727674284616,
      "grad_norm": 35.75,
      "learning_rate": 4.632728131129565e-05,
      "loss": 3.8705,
      "num_input_tokens_seen": 9471872,
      "step": 6990
    },
    {
      "epoch": 0.22389731771333463,
      "grad_norm": 37.25,
      "learning_rate": 4.632215728304018e-05,
      "loss": 5.4072,
      "num_input_tokens_seen": 9478736,
      "step": 6995
    },
    {
      "epoch": 0.22405735868382307,
      "grad_norm": 21.125,
      "learning_rate": 4.63170299666401e-05,
      "loss": 3.3807,
      "num_input_tokens_seen": 9485568,
      "step": 7000
    },
    {
      "epoch": 0.22405735868382307,
      "eval_loss": 4.491085052490234,
      "eval_runtime": 487.5483,
      "eval_samples_per_second": 28.481,
      "eval_steps_per_second": 14.241,
      "num_input_tokens_seen": 9485568,
      "step": 7000
    },
    {
      "epoch": 0.2242173996543115,
      "grad_norm": 26.625,
      "learning_rate": 4.631189936288612e-05,
      "loss": 4.6085,
      "num_input_tokens_seen": 9492112,
      "step": 7005
    },
    {
      "epoch": 0.22437744062479995,
      "grad_norm": 32.75,
      "learning_rate": 4.630676547256944e-05,
      "loss": 4.1677,
      "num_input_tokens_seen": 9498816,
      "step": 7010
    },
    {
      "epoch": 0.22453748159528839,
      "grad_norm": 26.75,
      "learning_rate": 4.630162829648176e-05,
      "loss": 4.5055,
      "num_input_tokens_seen": 9506112,
      "step": 7015
    },
    {
      "epoch": 0.22469752256577685,
      "grad_norm": 23.625,
      "learning_rate": 4.629648783541531e-05,
      "loss": 4.7532,
      "num_input_tokens_seen": 9512864,
      "step": 7020
    },
    {
      "epoch": 0.2248575635362653,
      "grad_norm": 37.25,
      "learning_rate": 4.6291344090162804e-05,
      "loss": 4.4224,
      "num_input_tokens_seen": 9520224,
      "step": 7025
    },
    {
      "epoch": 0.22501760450675373,
      "grad_norm": 26.125,
      "learning_rate": 4.628619706151748e-05,
      "loss": 5.1177,
      "num_input_tokens_seen": 9526848,
      "step": 7030
    },
    {
      "epoch": 0.22517764547724217,
      "grad_norm": 24.5,
      "learning_rate": 4.628104675027306e-05,
      "loss": 3.5097,
      "num_input_tokens_seen": 9533520,
      "step": 7035
    },
    {
      "epoch": 0.2253376864477306,
      "grad_norm": 23.125,
      "learning_rate": 4.6275893157223805e-05,
      "loss": 3.5847,
      "num_input_tokens_seen": 9540624,
      "step": 7040
    },
    {
      "epoch": 0.22549772741821905,
      "grad_norm": 28.5,
      "learning_rate": 4.627073628316445e-05,
      "loss": 4.8027,
      "num_input_tokens_seen": 9547712,
      "step": 7045
    },
    {
      "epoch": 0.22565776838870752,
      "grad_norm": 26.0,
      "learning_rate": 4.626557612889026e-05,
      "loss": 3.6811,
      "num_input_tokens_seen": 9554160,
      "step": 7050
    },
    {
      "epoch": 0.22581780935919596,
      "grad_norm": 20.375,
      "learning_rate": 4.626041269519699e-05,
      "loss": 4.063,
      "num_input_tokens_seen": 9560848,
      "step": 7055
    },
    {
      "epoch": 0.2259778503296844,
      "grad_norm": 37.5,
      "learning_rate": 4.6255245982880905e-05,
      "loss": 4.4261,
      "num_input_tokens_seen": 9567808,
      "step": 7060
    },
    {
      "epoch": 0.22613789130017284,
      "grad_norm": 20.75,
      "learning_rate": 4.625007599273879e-05,
      "loss": 3.5727,
      "num_input_tokens_seen": 9574864,
      "step": 7065
    },
    {
      "epoch": 0.22629793227066128,
      "grad_norm": 26.5,
      "learning_rate": 4.6244902725567895e-05,
      "loss": 4.3643,
      "num_input_tokens_seen": 9581776,
      "step": 7070
    },
    {
      "epoch": 0.22645797324114975,
      "grad_norm": 24.5,
      "learning_rate": 4.6239726182166024e-05,
      "loss": 4.7221,
      "num_input_tokens_seen": 9588208,
      "step": 7075
    },
    {
      "epoch": 0.22661801421163819,
      "grad_norm": 31.25,
      "learning_rate": 4.623454636333147e-05,
      "loss": 3.7702,
      "num_input_tokens_seen": 9595024,
      "step": 7080
    },
    {
      "epoch": 0.22677805518212663,
      "grad_norm": 38.25,
      "learning_rate": 4.622936326986301e-05,
      "loss": 4.2371,
      "num_input_tokens_seen": 9601680,
      "step": 7085
    },
    {
      "epoch": 0.22693809615261507,
      "grad_norm": 18.375,
      "learning_rate": 4.6224176902559946e-05,
      "loss": 3.0676,
      "num_input_tokens_seen": 9608560,
      "step": 7090
    },
    {
      "epoch": 0.2270981371231035,
      "grad_norm": 71.0,
      "learning_rate": 4.621898726222209e-05,
      "loss": 3.9551,
      "num_input_tokens_seen": 9615136,
      "step": 7095
    },
    {
      "epoch": 0.22725817809359197,
      "grad_norm": 29.75,
      "learning_rate": 4.6213794349649744e-05,
      "loss": 4.7992,
      "num_input_tokens_seen": 9622112,
      "step": 7100
    },
    {
      "epoch": 0.2274182190640804,
      "grad_norm": 40.5,
      "learning_rate": 4.6208598165643715e-05,
      "loss": 3.6625,
      "num_input_tokens_seen": 9628576,
      "step": 7105
    },
    {
      "epoch": 0.22757826003456885,
      "grad_norm": 27.875,
      "learning_rate": 4.620339871100533e-05,
      "loss": 4.6431,
      "num_input_tokens_seen": 9635392,
      "step": 7110
    },
    {
      "epoch": 0.2277383010050573,
      "grad_norm": 42.25,
      "learning_rate": 4.6198195986536394e-05,
      "loss": 4.6298,
      "num_input_tokens_seen": 9641936,
      "step": 7115
    },
    {
      "epoch": 0.22789834197554573,
      "grad_norm": 33.5,
      "learning_rate": 4.619298999303926e-05,
      "loss": 4.0716,
      "num_input_tokens_seen": 9648672,
      "step": 7120
    },
    {
      "epoch": 0.22805838294603417,
      "grad_norm": 30.5,
      "learning_rate": 4.618778073131673e-05,
      "loss": 3.6259,
      "num_input_tokens_seen": 9655568,
      "step": 7125
    },
    {
      "epoch": 0.22821842391652264,
      "grad_norm": 29.875,
      "learning_rate": 4.618256820217215e-05,
      "loss": 4.3848,
      "num_input_tokens_seen": 9662016,
      "step": 7130
    },
    {
      "epoch": 0.22837846488701108,
      "grad_norm": 49.5,
      "learning_rate": 4.617735240640936e-05,
      "loss": 4.8789,
      "num_input_tokens_seen": 9668592,
      "step": 7135
    },
    {
      "epoch": 0.22853850585749952,
      "grad_norm": 40.25,
      "learning_rate": 4.6172133344832705e-05,
      "loss": 4.9896,
      "num_input_tokens_seen": 9675504,
      "step": 7140
    },
    {
      "epoch": 0.22869854682798796,
      "grad_norm": 31.875,
      "learning_rate": 4.6166911018247004e-05,
      "loss": 5.5022,
      "num_input_tokens_seen": 9682224,
      "step": 7145
    },
    {
      "epoch": 0.2288585877984764,
      "grad_norm": 31.125,
      "learning_rate": 4.616168542745764e-05,
      "loss": 4.8816,
      "num_input_tokens_seen": 9688864,
      "step": 7150
    },
    {
      "epoch": 0.22901862876896487,
      "grad_norm": 29.375,
      "learning_rate": 4.6156456573270446e-05,
      "loss": 3.7868,
      "num_input_tokens_seen": 9696368,
      "step": 7155
    },
    {
      "epoch": 0.2291786697394533,
      "grad_norm": 36.25,
      "learning_rate": 4.615122445649177e-05,
      "loss": 4.8694,
      "num_input_tokens_seen": 9703136,
      "step": 7160
    },
    {
      "epoch": 0.22933871070994175,
      "grad_norm": 20.25,
      "learning_rate": 4.6145989077928486e-05,
      "loss": 4.067,
      "num_input_tokens_seen": 9709600,
      "step": 7165
    },
    {
      "epoch": 0.22949875168043019,
      "grad_norm": 50.5,
      "learning_rate": 4.6140750438387953e-05,
      "loss": 4.0522,
      "num_input_tokens_seen": 9716336,
      "step": 7170
    },
    {
      "epoch": 0.22965879265091863,
      "grad_norm": 32.25,
      "learning_rate": 4.613550853867803e-05,
      "loss": 3.8112,
      "num_input_tokens_seen": 9723232,
      "step": 7175
    },
    {
      "epoch": 0.2298188336214071,
      "grad_norm": 31.625,
      "learning_rate": 4.613026337960708e-05,
      "loss": 4.397,
      "num_input_tokens_seen": 9730128,
      "step": 7180
    },
    {
      "epoch": 0.22997887459189553,
      "grad_norm": 25.5,
      "learning_rate": 4.612501496198398e-05,
      "loss": 4.3655,
      "num_input_tokens_seen": 9737104,
      "step": 7185
    },
    {
      "epoch": 0.23013891556238397,
      "grad_norm": 29.625,
      "learning_rate": 4.61197632866181e-05,
      "loss": 3.5478,
      "num_input_tokens_seen": 9743872,
      "step": 7190
    },
    {
      "epoch": 0.2302989565328724,
      "grad_norm": 33.0,
      "learning_rate": 4.611450835431931e-05,
      "loss": 4.2123,
      "num_input_tokens_seen": 9750784,
      "step": 7195
    },
    {
      "epoch": 0.23045899750336085,
      "grad_norm": 20.375,
      "learning_rate": 4.6109250165898e-05,
      "loss": 4.1261,
      "num_input_tokens_seen": 9758160,
      "step": 7200
    },
    {
      "epoch": 0.23045899750336085,
      "eval_loss": 4.491462230682373,
      "eval_runtime": 487.0163,
      "eval_samples_per_second": 28.512,
      "eval_steps_per_second": 14.256,
      "num_input_tokens_seen": 9758160,
      "step": 7200
    },
    {
      "epoch": 0.2306190384738493,
      "grad_norm": 24.0,
      "learning_rate": 4.610398872216503e-05,
      "loss": 3.9774,
      "num_input_tokens_seen": 9765360,
      "step": 7205
    },
    {
      "epoch": 0.23077907944433776,
      "grad_norm": 22.125,
      "learning_rate": 4.6098724023931796e-05,
      "loss": 2.8219,
      "num_input_tokens_seen": 9772176,
      "step": 7210
    },
    {
      "epoch": 0.2309391204148262,
      "grad_norm": 35.0,
      "learning_rate": 4.609345607201017e-05,
      "loss": 4.8123,
      "num_input_tokens_seen": 9778928,
      "step": 7215
    },
    {
      "epoch": 0.23109916138531464,
      "grad_norm": 24.25,
      "learning_rate": 4.608818486721254e-05,
      "loss": 3.8565,
      "num_input_tokens_seen": 9785424,
      "step": 7220
    },
    {
      "epoch": 0.23125920235580308,
      "grad_norm": 42.0,
      "learning_rate": 4.608291041035179e-05,
      "loss": 4.1801,
      "num_input_tokens_seen": 9792096,
      "step": 7225
    },
    {
      "epoch": 0.23141924332629152,
      "grad_norm": 37.0,
      "learning_rate": 4.607763270224132e-05,
      "loss": 4.2772,
      "num_input_tokens_seen": 9798944,
      "step": 7230
    },
    {
      "epoch": 0.23157928429677999,
      "grad_norm": 34.25,
      "learning_rate": 4.6072351743695e-05,
      "loss": 3.8587,
      "num_input_tokens_seen": 9805552,
      "step": 7235
    },
    {
      "epoch": 0.23173932526726843,
      "grad_norm": 24.875,
      "learning_rate": 4.606706753552723e-05,
      "loss": 4.0619,
      "num_input_tokens_seen": 9812592,
      "step": 7240
    },
    {
      "epoch": 0.23189936623775687,
      "grad_norm": 23.5,
      "learning_rate": 4.6061780078552906e-05,
      "loss": 3.9772,
      "num_input_tokens_seen": 9819184,
      "step": 7245
    },
    {
      "epoch": 0.2320594072082453,
      "grad_norm": 24.125,
      "learning_rate": 4.605648937358742e-05,
      "loss": 4.7286,
      "num_input_tokens_seen": 9825984,
      "step": 7250
    },
    {
      "epoch": 0.23221944817873375,
      "grad_norm": 23.25,
      "learning_rate": 4.605119542144665e-05,
      "loss": 4.0619,
      "num_input_tokens_seen": 9832736,
      "step": 7255
    },
    {
      "epoch": 0.2323794891492222,
      "grad_norm": 34.75,
      "learning_rate": 4.604589822294701e-05,
      "loss": 4.0224,
      "num_input_tokens_seen": 9839392,
      "step": 7260
    },
    {
      "epoch": 0.23253953011971065,
      "grad_norm": 41.25,
      "learning_rate": 4.604059777890537e-05,
      "loss": 4.8065,
      "num_input_tokens_seen": 9846512,
      "step": 7265
    },
    {
      "epoch": 0.2326995710901991,
      "grad_norm": 33.75,
      "learning_rate": 4.6035294090139145e-05,
      "loss": 3.689,
      "num_input_tokens_seen": 9853680,
      "step": 7270
    },
    {
      "epoch": 0.23285961206068753,
      "grad_norm": 19.625,
      "learning_rate": 4.6029987157466226e-05,
      "loss": 3.5293,
      "num_input_tokens_seen": 9860448,
      "step": 7275
    },
    {
      "epoch": 0.23301965303117597,
      "grad_norm": 33.75,
      "learning_rate": 4.602467698170502e-05,
      "loss": 4.8206,
      "num_input_tokens_seen": 9867712,
      "step": 7280
    },
    {
      "epoch": 0.23317969400166444,
      "grad_norm": 38.25,
      "learning_rate": 4.601936356367439e-05,
      "loss": 4.0398,
      "num_input_tokens_seen": 9874528,
      "step": 7285
    },
    {
      "epoch": 0.23333973497215288,
      "grad_norm": 22.25,
      "learning_rate": 4.601404690419377e-05,
      "loss": 4.7844,
      "num_input_tokens_seen": 9881392,
      "step": 7290
    },
    {
      "epoch": 0.23349977594264132,
      "grad_norm": 23.25,
      "learning_rate": 4.600872700408303e-05,
      "loss": 5.11,
      "num_input_tokens_seen": 9888144,
      "step": 7295
    },
    {
      "epoch": 0.23365981691312976,
      "grad_norm": 28.625,
      "learning_rate": 4.600340386416258e-05,
      "loss": 3.7796,
      "num_input_tokens_seen": 9894576,
      "step": 7300
    },
    {
      "epoch": 0.2338198578836182,
      "grad_norm": 32.25,
      "learning_rate": 4.5998077485253296e-05,
      "loss": 4.2178,
      "num_input_tokens_seen": 9901408,
      "step": 7305
    },
    {
      "epoch": 0.23397989885410664,
      "grad_norm": 26.125,
      "learning_rate": 4.59927478681766e-05,
      "loss": 5.1253,
      "num_input_tokens_seen": 9908096,
      "step": 7310
    },
    {
      "epoch": 0.2341399398245951,
      "grad_norm": 19.25,
      "learning_rate": 4.5987415013754366e-05,
      "loss": 4.8376,
      "num_input_tokens_seen": 9915376,
      "step": 7315
    },
    {
      "epoch": 0.23429998079508355,
      "grad_norm": 21.125,
      "learning_rate": 4.598207892280899e-05,
      "loss": 3.8192,
      "num_input_tokens_seen": 9922000,
      "step": 7320
    },
    {
      "epoch": 0.23446002176557199,
      "grad_norm": 34.75,
      "learning_rate": 4.597673959616337e-05,
      "loss": 4.7887,
      "num_input_tokens_seen": 9928528,
      "step": 7325
    },
    {
      "epoch": 0.23462006273606043,
      "grad_norm": 39.75,
      "learning_rate": 4.597139703464089e-05,
      "loss": 4.8974,
      "num_input_tokens_seen": 9935360,
      "step": 7330
    },
    {
      "epoch": 0.23478010370654886,
      "grad_norm": 31.5,
      "learning_rate": 4.596605123906545e-05,
      "loss": 4.2881,
      "num_input_tokens_seen": 9941936,
      "step": 7335
    },
    {
      "epoch": 0.23494014467703733,
      "grad_norm": 30.625,
      "learning_rate": 4.596070221026143e-05,
      "loss": 3.7792,
      "num_input_tokens_seen": 9948288,
      "step": 7340
    },
    {
      "epoch": 0.23510018564752577,
      "grad_norm": 32.25,
      "learning_rate": 4.595534994905372e-05,
      "loss": 4.2239,
      "num_input_tokens_seen": 9954992,
      "step": 7345
    },
    {
      "epoch": 0.2352602266180142,
      "grad_norm": 42.75,
      "learning_rate": 4.594999445626771e-05,
      "loss": 4.2673,
      "num_input_tokens_seen": 9961616,
      "step": 7350
    },
    {
      "epoch": 0.23542026758850265,
      "grad_norm": 32.25,
      "learning_rate": 4.5944635732729276e-05,
      "loss": 4.2305,
      "num_input_tokens_seen": 9968064,
      "step": 7355
    },
    {
      "epoch": 0.2355803085589911,
      "grad_norm": 27.5,
      "learning_rate": 4.5939273779264804e-05,
      "loss": 4.2624,
      "num_input_tokens_seen": 9974832,
      "step": 7360
    },
    {
      "epoch": 0.23574034952947956,
      "grad_norm": 38.75,
      "learning_rate": 4.593390859670118e-05,
      "loss": 6.4372,
      "num_input_tokens_seen": 9981536,
      "step": 7365
    },
    {
      "epoch": 0.235900390499968,
      "grad_norm": 18.5,
      "learning_rate": 4.5928540185865776e-05,
      "loss": 4.2349,
      "num_input_tokens_seen": 9988128,
      "step": 7370
    },
    {
      "epoch": 0.23606043147045644,
      "grad_norm": 29.875,
      "learning_rate": 4.592316854758648e-05,
      "loss": 4.9978,
      "num_input_tokens_seen": 9994704,
      "step": 7375
    },
    {
      "epoch": 0.23622047244094488,
      "grad_norm": 41.75,
      "learning_rate": 4.5917793682691646e-05,
      "loss": 4.0774,
      "num_input_tokens_seen": 10001280,
      "step": 7380
    },
    {
      "epoch": 0.23638051341143332,
      "grad_norm": 39.75,
      "learning_rate": 4.5912415592010164e-05,
      "loss": 4.4464,
      "num_input_tokens_seen": 10007760,
      "step": 7385
    },
    {
      "epoch": 0.23654055438192176,
      "grad_norm": 42.25,
      "learning_rate": 4.5907034276371386e-05,
      "loss": 4.583,
      "num_input_tokens_seen": 10014032,
      "step": 7390
    },
    {
      "epoch": 0.23670059535241023,
      "grad_norm": 25.625,
      "learning_rate": 4.5901649736605196e-05,
      "loss": 3.8831,
      "num_input_tokens_seen": 10021232,
      "step": 7395
    },
    {
      "epoch": 0.23686063632289867,
      "grad_norm": 21.5,
      "learning_rate": 4.589626197354195e-05,
      "loss": 4.3934,
      "num_input_tokens_seen": 10028256,
      "step": 7400
    },
    {
      "epoch": 0.23686063632289867,
      "eval_loss": 4.4810943603515625,
      "eval_runtime": 487.9238,
      "eval_samples_per_second": 28.459,
      "eval_steps_per_second": 14.23,
      "num_input_tokens_seen": 10028256,
      "step": 7400
    },
    {
      "epoch": 0.2370206772933871,
      "grad_norm": 25.75,
      "learning_rate": 4.5890870988012504e-05,
      "loss": 3.6175,
      "num_input_tokens_seen": 10034672,
      "step": 7405
    },
    {
      "epoch": 0.23718071826387555,
      "grad_norm": 47.25,
      "learning_rate": 4.5885476780848226e-05,
      "loss": 4.7821,
      "num_input_tokens_seen": 10041936,
      "step": 7410
    },
    {
      "epoch": 0.23734075923436398,
      "grad_norm": 32.75,
      "learning_rate": 4.5880079352880964e-05,
      "loss": 3.9653,
      "num_input_tokens_seen": 10048704,
      "step": 7415
    },
    {
      "epoch": 0.23750080020485245,
      "grad_norm": 31.0,
      "learning_rate": 4.5874678704943065e-05,
      "loss": 3.8104,
      "num_input_tokens_seen": 10055696,
      "step": 7420
    },
    {
      "epoch": 0.2376608411753409,
      "grad_norm": 44.25,
      "learning_rate": 4.5869274837867394e-05,
      "loss": 4.0289,
      "num_input_tokens_seen": 10062272,
      "step": 7425
    },
    {
      "epoch": 0.23782088214582933,
      "grad_norm": 36.75,
      "learning_rate": 4.5863867752487275e-05,
      "loss": 4.1296,
      "num_input_tokens_seen": 10068816,
      "step": 7430
    },
    {
      "epoch": 0.23798092311631777,
      "grad_norm": 23.375,
      "learning_rate": 4.5858457449636554e-05,
      "loss": 3.7634,
      "num_input_tokens_seen": 10075936,
      "step": 7435
    },
    {
      "epoch": 0.2381409640868062,
      "grad_norm": 29.375,
      "learning_rate": 4.5853043930149574e-05,
      "loss": 4.526,
      "num_input_tokens_seen": 10082720,
      "step": 7440
    },
    {
      "epoch": 0.23830100505729468,
      "grad_norm": 29.25,
      "learning_rate": 4.584762719486117e-05,
      "loss": 4.3541,
      "num_input_tokens_seen": 10089264,
      "step": 7445
    },
    {
      "epoch": 0.23846104602778312,
      "grad_norm": 44.75,
      "learning_rate": 4.584220724460665e-05,
      "loss": 4.7323,
      "num_input_tokens_seen": 10095744,
      "step": 7450
    },
    {
      "epoch": 0.23862108699827156,
      "grad_norm": 13.9375,
      "learning_rate": 4.5836784080221865e-05,
      "loss": 4.3047,
      "num_input_tokens_seen": 10102432,
      "step": 7455
    },
    {
      "epoch": 0.23878112796876,
      "grad_norm": 24.0,
      "learning_rate": 4.583135770254312e-05,
      "loss": 4.6457,
      "num_input_tokens_seen": 10109648,
      "step": 7460
    },
    {
      "epoch": 0.23894116893924844,
      "grad_norm": 21.0,
      "learning_rate": 4.5825928112407236e-05,
      "loss": 4.9857,
      "num_input_tokens_seen": 10116384,
      "step": 7465
    },
    {
      "epoch": 0.2391012099097369,
      "grad_norm": 20.5,
      "learning_rate": 4.582049531065152e-05,
      "loss": 3.552,
      "num_input_tokens_seen": 10123424,
      "step": 7470
    },
    {
      "epoch": 0.23926125088022535,
      "grad_norm": 65.0,
      "learning_rate": 4.5815059298113783e-05,
      "loss": 4.213,
      "num_input_tokens_seen": 10129968,
      "step": 7475
    },
    {
      "epoch": 0.23942129185071379,
      "grad_norm": 22.375,
      "learning_rate": 4.580962007563232e-05,
      "loss": 4.2836,
      "num_input_tokens_seen": 10136784,
      "step": 7480
    },
    {
      "epoch": 0.23958133282120223,
      "grad_norm": 18.125,
      "learning_rate": 4.5804177644045935e-05,
      "loss": 4.3396,
      "num_input_tokens_seen": 10143600,
      "step": 7485
    },
    {
      "epoch": 0.23974137379169067,
      "grad_norm": 13.8125,
      "learning_rate": 4.579873200419391e-05,
      "loss": 3.5552,
      "num_input_tokens_seen": 10150448,
      "step": 7490
    },
    {
      "epoch": 0.2399014147621791,
      "grad_norm": 31.875,
      "learning_rate": 4.5793283156916046e-05,
      "loss": 4.8694,
      "num_input_tokens_seen": 10157024,
      "step": 7495
    },
    {
      "epoch": 0.24006145573266757,
      "grad_norm": 20.125,
      "learning_rate": 4.578783110305261e-05,
      "loss": 4.3151,
      "num_input_tokens_seen": 10164208,
      "step": 7500
    },
    {
      "epoch": 0.240221496703156,
      "grad_norm": 24.25,
      "learning_rate": 4.578237584344438e-05,
      "loss": 3.9774,
      "num_input_tokens_seen": 10170976,
      "step": 7505
    },
    {
      "epoch": 0.24038153767364445,
      "grad_norm": 14.9375,
      "learning_rate": 4.577691737893263e-05,
      "loss": 3.8977,
      "num_input_tokens_seen": 10177840,
      "step": 7510
    },
    {
      "epoch": 0.2405415786441329,
      "grad_norm": 20.75,
      "learning_rate": 4.577145571035912e-05,
      "loss": 3.8836,
      "num_input_tokens_seen": 10184560,
      "step": 7515
    },
    {
      "epoch": 0.24070161961462133,
      "grad_norm": 28.0,
      "learning_rate": 4.576599083856611e-05,
      "loss": 4.3243,
      "num_input_tokens_seen": 10191456,
      "step": 7520
    },
    {
      "epoch": 0.2408616605851098,
      "grad_norm": 18.5,
      "learning_rate": 4.576052276439635e-05,
      "loss": 3.97,
      "num_input_tokens_seen": 10198000,
      "step": 7525
    },
    {
      "epoch": 0.24102170155559824,
      "grad_norm": 44.0,
      "learning_rate": 4.575505148869308e-05,
      "loss": 4.6802,
      "num_input_tokens_seen": 10204800,
      "step": 7530
    },
    {
      "epoch": 0.24118174252608668,
      "grad_norm": 16.25,
      "learning_rate": 4.574957701230006e-05,
      "loss": 3.5175,
      "num_input_tokens_seen": 10211808,
      "step": 7535
    },
    {
      "epoch": 0.24134178349657512,
      "grad_norm": 26.25,
      "learning_rate": 4.57440993360615e-05,
      "loss": 5.5474,
      "num_input_tokens_seen": 10219008,
      "step": 7540
    },
    {
      "epoch": 0.24150182446706356,
      "grad_norm": 21.5,
      "learning_rate": 4.5738618460822134e-05,
      "loss": 4.2653,
      "num_input_tokens_seen": 10225872,
      "step": 7545
    },
    {
      "epoch": 0.24166186543755203,
      "grad_norm": 15.1875,
      "learning_rate": 4.573313438742719e-05,
      "loss": 4.2805,
      "num_input_tokens_seen": 10232704,
      "step": 7550
    },
    {
      "epoch": 0.24182190640804047,
      "grad_norm": 60.75,
      "learning_rate": 4.5727647116722374e-05,
      "loss": 3.6647,
      "num_input_tokens_seen": 10239744,
      "step": 7555
    },
    {
      "epoch": 0.2419819473785289,
      "grad_norm": 22.0,
      "learning_rate": 4.5722156649553884e-05,
      "loss": 4.0837,
      "num_input_tokens_seen": 10246704,
      "step": 7560
    },
    {
      "epoch": 0.24214198834901735,
      "grad_norm": 35.0,
      "learning_rate": 4.571666298676843e-05,
      "loss": 4.5049,
      "num_input_tokens_seen": 10253264,
      "step": 7565
    },
    {
      "epoch": 0.24230202931950579,
      "grad_norm": 45.0,
      "learning_rate": 4.571116612921321e-05,
      "loss": 3.8061,
      "num_input_tokens_seen": 10259712,
      "step": 7570
    },
    {
      "epoch": 0.24246207028999422,
      "grad_norm": 24.25,
      "learning_rate": 4.57056660777359e-05,
      "loss": 3.8726,
      "num_input_tokens_seen": 10266400,
      "step": 7575
    },
    {
      "epoch": 0.2426221112604827,
      "grad_norm": 16.5,
      "learning_rate": 4.5700162833184666e-05,
      "loss": 3.5059,
      "num_input_tokens_seen": 10273568,
      "step": 7580
    },
    {
      "epoch": 0.24278215223097113,
      "grad_norm": 66.0,
      "learning_rate": 4.5694656396408195e-05,
      "loss": 3.7319,
      "num_input_tokens_seen": 10280432,
      "step": 7585
    },
    {
      "epoch": 0.24294219320145957,
      "grad_norm": 24.125,
      "learning_rate": 4.5689146768255646e-05,
      "loss": 4.5582,
      "num_input_tokens_seen": 10286880,
      "step": 7590
    },
    {
      "epoch": 0.243102234171948,
      "grad_norm": 34.75,
      "learning_rate": 4.568363394957667e-05,
      "loss": 5.5914,
      "num_input_tokens_seen": 10293984,
      "step": 7595
    },
    {
      "epoch": 0.24326227514243645,
      "grad_norm": 30.625,
      "learning_rate": 4.567811794122141e-05,
      "loss": 5.2692,
      "num_input_tokens_seen": 10300544,
      "step": 7600
    },
    {
      "epoch": 0.24326227514243645,
      "eval_loss": 4.493503570556641,
      "eval_runtime": 487.7458,
      "eval_samples_per_second": 28.47,
      "eval_steps_per_second": 14.235,
      "num_input_tokens_seen": 10300544,
      "step": 7600
    },
    {
      "epoch": 0.24342231611292492,
      "grad_norm": 30.625,
      "learning_rate": 4.56725987440405e-05,
      "loss": 4.4173,
      "num_input_tokens_seen": 10307504,
      "step": 7605
    },
    {
      "epoch": 0.24358235708341336,
      "grad_norm": 20.375,
      "learning_rate": 4.566707635888508e-05,
      "loss": 4.6899,
      "num_input_tokens_seen": 10314432,
      "step": 7610
    },
    {
      "epoch": 0.2437423980539018,
      "grad_norm": 36.0,
      "learning_rate": 4.566155078660677e-05,
      "loss": 4.5143,
      "num_input_tokens_seen": 10321472,
      "step": 7615
    },
    {
      "epoch": 0.24390243902439024,
      "grad_norm": 26.125,
      "learning_rate": 4.565602202805768e-05,
      "loss": 4.554,
      "num_input_tokens_seen": 10328160,
      "step": 7620
    },
    {
      "epoch": 0.24406247999487868,
      "grad_norm": 27.75,
      "learning_rate": 4.56504900840904e-05,
      "loss": 4.441,
      "num_input_tokens_seen": 10335360,
      "step": 7625
    },
    {
      "epoch": 0.24422252096536715,
      "grad_norm": 20.125,
      "learning_rate": 4.564495495555805e-05,
      "loss": 3.8692,
      "num_input_tokens_seen": 10341888,
      "step": 7630
    },
    {
      "epoch": 0.24438256193585559,
      "grad_norm": 31.625,
      "learning_rate": 4.5639416643314204e-05,
      "loss": 4.1565,
      "num_input_tokens_seen": 10348528,
      "step": 7635
    },
    {
      "epoch": 0.24454260290634403,
      "grad_norm": 25.125,
      "learning_rate": 4.5633875148212946e-05,
      "loss": 4.6097,
      "num_input_tokens_seen": 10355440,
      "step": 7640
    },
    {
      "epoch": 0.24470264387683247,
      "grad_norm": 25.75,
      "learning_rate": 4.562833047110883e-05,
      "loss": 4.455,
      "num_input_tokens_seen": 10362592,
      "step": 7645
    },
    {
      "epoch": 0.2448626848473209,
      "grad_norm": 44.5,
      "learning_rate": 4.5622782612856923e-05,
      "loss": 4.2408,
      "num_input_tokens_seen": 10369392,
      "step": 7650
    },
    {
      "epoch": 0.24502272581780937,
      "grad_norm": 24.5,
      "learning_rate": 4.561723157431278e-05,
      "loss": 3.7785,
      "num_input_tokens_seen": 10376144,
      "step": 7655
    },
    {
      "epoch": 0.2451827667882978,
      "grad_norm": 18.75,
      "learning_rate": 4.5611677356332435e-05,
      "loss": 3.729,
      "num_input_tokens_seen": 10383136,
      "step": 7660
    },
    {
      "epoch": 0.24534280775878625,
      "grad_norm": 31.375,
      "learning_rate": 4.560611995977242e-05,
      "loss": 3.2471,
      "num_input_tokens_seen": 10389968,
      "step": 7665
    },
    {
      "epoch": 0.2455028487292747,
      "grad_norm": 30.0,
      "learning_rate": 4.560055938548975e-05,
      "loss": 4.1778,
      "num_input_tokens_seen": 10396528,
      "step": 7670
    },
    {
      "epoch": 0.24566288969976313,
      "grad_norm": 36.25,
      "learning_rate": 4.5594995634341944e-05,
      "loss": 3.6353,
      "num_input_tokens_seen": 10403664,
      "step": 7675
    },
    {
      "epoch": 0.24582293067025157,
      "grad_norm": 30.75,
      "learning_rate": 4.5589428707187e-05,
      "loss": 4.0625,
      "num_input_tokens_seen": 10410256,
      "step": 7680
    },
    {
      "epoch": 0.24598297164074004,
      "grad_norm": 22.125,
      "learning_rate": 4.55838586048834e-05,
      "loss": 4.2944,
      "num_input_tokens_seen": 10416560,
      "step": 7685
    },
    {
      "epoch": 0.24614301261122848,
      "grad_norm": 45.0,
      "learning_rate": 4.557828532829013e-05,
      "loss": 4.6006,
      "num_input_tokens_seen": 10423296,
      "step": 7690
    },
    {
      "epoch": 0.24630305358171692,
      "grad_norm": 30.75,
      "learning_rate": 4.557270887826667e-05,
      "loss": 4.3384,
      "num_input_tokens_seen": 10430192,
      "step": 7695
    },
    {
      "epoch": 0.24646309455220536,
      "grad_norm": 9.8125,
      "learning_rate": 4.556712925567296e-05,
      "loss": 4.4368,
      "num_input_tokens_seen": 10437136,
      "step": 7700
    },
    {
      "epoch": 0.2466231355226938,
      "grad_norm": 21.0,
      "learning_rate": 4.5561546461369454e-05,
      "loss": 4.3265,
      "num_input_tokens_seen": 10444176,
      "step": 7705
    },
    {
      "epoch": 0.24678317649318227,
      "grad_norm": 41.75,
      "learning_rate": 4.55559604962171e-05,
      "loss": 3.9428,
      "num_input_tokens_seen": 10451504,
      "step": 7710
    },
    {
      "epoch": 0.2469432174636707,
      "grad_norm": 14.5,
      "learning_rate": 4.55503713610773e-05,
      "loss": 3.3788,
      "num_input_tokens_seen": 10458288,
      "step": 7715
    },
    {
      "epoch": 0.24710325843415915,
      "grad_norm": 40.25,
      "learning_rate": 4.5544779056812e-05,
      "loss": 4.4809,
      "num_input_tokens_seen": 10464992,
      "step": 7720
    },
    {
      "epoch": 0.24726329940464759,
      "grad_norm": 27.125,
      "learning_rate": 4.553918358428358e-05,
      "loss": 4.382,
      "num_input_tokens_seen": 10471520,
      "step": 7725
    },
    {
      "epoch": 0.24742334037513602,
      "grad_norm": 28.0,
      "learning_rate": 4.553358494435494e-05,
      "loss": 4.7734,
      "num_input_tokens_seen": 10478032,
      "step": 7730
    },
    {
      "epoch": 0.2475833813456245,
      "grad_norm": 28.75,
      "learning_rate": 4.5527983137889464e-05,
      "loss": 4.3933,
      "num_input_tokens_seen": 10484864,
      "step": 7735
    },
    {
      "epoch": 0.24774342231611293,
      "grad_norm": 25.625,
      "learning_rate": 4.5522378165751015e-05,
      "loss": 4.0699,
      "num_input_tokens_seen": 10492272,
      "step": 7740
    },
    {
      "epoch": 0.24790346328660137,
      "grad_norm": 25.125,
      "learning_rate": 4.5516770028803954e-05,
      "loss": 4.0569,
      "num_input_tokens_seen": 10499104,
      "step": 7745
    },
    {
      "epoch": 0.2480635042570898,
      "grad_norm": 24.0,
      "learning_rate": 4.5511158727913116e-05,
      "loss": 5.6388,
      "num_input_tokens_seen": 10505808,
      "step": 7750
    },
    {
      "epoch": 0.24822354522757825,
      "grad_norm": 24.625,
      "learning_rate": 4.5505544263943856e-05,
      "loss": 4.2801,
      "num_input_tokens_seen": 10513200,
      "step": 7755
    },
    {
      "epoch": 0.2483835861980667,
      "grad_norm": 23.25,
      "learning_rate": 4.549992663776197e-05,
      "loss": 3.9208,
      "num_input_tokens_seen": 10520000,
      "step": 7760
    },
    {
      "epoch": 0.24854362716855516,
      "grad_norm": 24.125,
      "learning_rate": 4.5494305850233786e-05,
      "loss": 5.0065,
      "num_input_tokens_seen": 10526320,
      "step": 7765
    },
    {
      "epoch": 0.2487036681390436,
      "grad_norm": 25.375,
      "learning_rate": 4.5488681902226094e-05,
      "loss": 4.2047,
      "num_input_tokens_seen": 10533280,
      "step": 7770
    },
    {
      "epoch": 0.24886370910953204,
      "grad_norm": 17.875,
      "learning_rate": 4.5483054794606174e-05,
      "loss": 4.2566,
      "num_input_tokens_seen": 10540736,
      "step": 7775
    },
    {
      "epoch": 0.24902375008002048,
      "grad_norm": 39.5,
      "learning_rate": 4.547742452824179e-05,
      "loss": 4.2789,
      "num_input_tokens_seen": 10547264,
      "step": 7780
    },
    {
      "epoch": 0.24918379105050892,
      "grad_norm": 26.625,
      "learning_rate": 4.5471791104001215e-05,
      "loss": 4.2495,
      "num_input_tokens_seen": 10554048,
      "step": 7785
    },
    {
      "epoch": 0.24934383202099739,
      "grad_norm": 40.25,
      "learning_rate": 4.546615452275319e-05,
      "loss": 4.871,
      "num_input_tokens_seen": 10560928,
      "step": 7790
    },
    {
      "epoch": 0.24950387299148583,
      "grad_norm": 35.0,
      "learning_rate": 4.5460514785366944e-05,
      "loss": 4.7073,
      "num_input_tokens_seen": 10567408,
      "step": 7795
    },
    {
      "epoch": 0.24966391396197427,
      "grad_norm": 31.625,
      "learning_rate": 4.545487189271219e-05,
      "loss": 4.0686,
      "num_input_tokens_seen": 10574192,
      "step": 7800
    },
    {
      "epoch": 0.24966391396197427,
      "eval_loss": 4.481276035308838,
      "eval_runtime": 487.1331,
      "eval_samples_per_second": 28.506,
      "eval_steps_per_second": 14.253,
      "num_input_tokens_seen": 10574192,
      "step": 7800
    },
    {
      "epoch": 0.2498239549324627,
      "grad_norm": 28.375,
      "learning_rate": 4.544922584565914e-05,
      "loss": 4.8595,
      "num_input_tokens_seen": 10580624,
      "step": 7805
    },
    {
      "epoch": 0.24998399590295114,
      "grad_norm": 25.875,
      "learning_rate": 4.544357664507848e-05,
      "loss": 4.3558,
      "num_input_tokens_seen": 10587536,
      "step": 7810
    },
    {
      "epoch": 0.2501440368734396,
      "grad_norm": 18.875,
      "learning_rate": 4.54379242918414e-05,
      "loss": 3.4562,
      "num_input_tokens_seen": 10594592,
      "step": 7815
    },
    {
      "epoch": 0.25030407784392805,
      "grad_norm": 21.125,
      "learning_rate": 4.543226878681955e-05,
      "loss": 4.1951,
      "num_input_tokens_seen": 10601232,
      "step": 7820
    },
    {
      "epoch": 0.2504641188144165,
      "grad_norm": 19.75,
      "learning_rate": 4.5426610130885087e-05,
      "loss": 4.6025,
      "num_input_tokens_seen": 10607616,
      "step": 7825
    },
    {
      "epoch": 0.25062415978490493,
      "grad_norm": 16.625,
      "learning_rate": 4.542094832491064e-05,
      "loss": 4.9582,
      "num_input_tokens_seen": 10614400,
      "step": 7830
    },
    {
      "epoch": 0.25078420075539337,
      "grad_norm": 24.375,
      "learning_rate": 4.541528336976934e-05,
      "loss": 4.5293,
      "num_input_tokens_seen": 10620992,
      "step": 7835
    },
    {
      "epoch": 0.2509442417258818,
      "grad_norm": 19.25,
      "learning_rate": 4.540961526633479e-05,
      "loss": 3.4573,
      "num_input_tokens_seen": 10627792,
      "step": 7840
    },
    {
      "epoch": 0.25110428269637025,
      "grad_norm": 29.625,
      "learning_rate": 4.540394401548108e-05,
      "loss": 4.3842,
      "num_input_tokens_seen": 10634656,
      "step": 7845
    },
    {
      "epoch": 0.2512643236668587,
      "grad_norm": 27.5,
      "learning_rate": 4.539826961808279e-05,
      "loss": 5.271,
      "num_input_tokens_seen": 10641616,
      "step": 7850
    },
    {
      "epoch": 0.2514243646373472,
      "grad_norm": 28.875,
      "learning_rate": 4.5392592075014994e-05,
      "loss": 4.7363,
      "num_input_tokens_seen": 10648832,
      "step": 7855
    },
    {
      "epoch": 0.2515844056078356,
      "grad_norm": 34.25,
      "learning_rate": 4.538691138715322e-05,
      "loss": 5.1903,
      "num_input_tokens_seen": 10655616,
      "step": 7860
    },
    {
      "epoch": 0.25174444657832407,
      "grad_norm": 27.625,
      "learning_rate": 4.5381227555373516e-05,
      "loss": 3.8515,
      "num_input_tokens_seen": 10662256,
      "step": 7865
    },
    {
      "epoch": 0.2519044875488125,
      "grad_norm": 27.375,
      "learning_rate": 4.537554058055239e-05,
      "loss": 3.7612,
      "num_input_tokens_seen": 10669136,
      "step": 7870
    },
    {
      "epoch": 0.25206452851930095,
      "grad_norm": 33.25,
      "learning_rate": 4.5369850463566865e-05,
      "loss": 4.6569,
      "num_input_tokens_seen": 10675744,
      "step": 7875
    },
    {
      "epoch": 0.2522245694897894,
      "grad_norm": 30.25,
      "learning_rate": 4.5364157205294404e-05,
      "loss": 4.0992,
      "num_input_tokens_seen": 10682464,
      "step": 7880
    },
    {
      "epoch": 0.2523846104602778,
      "grad_norm": 55.5,
      "learning_rate": 4.5358460806612996e-05,
      "loss": 4.7124,
      "num_input_tokens_seen": 10689696,
      "step": 7885
    },
    {
      "epoch": 0.25254465143076626,
      "grad_norm": 35.0,
      "learning_rate": 4.535276126840109e-05,
      "loss": 4.2993,
      "num_input_tokens_seen": 10696608,
      "step": 7890
    },
    {
      "epoch": 0.2527046924012547,
      "grad_norm": 15.625,
      "learning_rate": 4.5347058591537626e-05,
      "loss": 4.6014,
      "num_input_tokens_seen": 10703424,
      "step": 7895
    },
    {
      "epoch": 0.25286473337174314,
      "grad_norm": 35.5,
      "learning_rate": 4.534135277690203e-05,
      "loss": 5.1637,
      "num_input_tokens_seen": 10710800,
      "step": 7900
    },
    {
      "epoch": 0.2530247743422316,
      "grad_norm": 18.125,
      "learning_rate": 4.533564382537421e-05,
      "loss": 4.52,
      "num_input_tokens_seen": 10717120,
      "step": 7905
    },
    {
      "epoch": 0.2531848153127201,
      "grad_norm": 22.25,
      "learning_rate": 4.532993173783456e-05,
      "loss": 4.2125,
      "num_input_tokens_seen": 10723776,
      "step": 7910
    },
    {
      "epoch": 0.2533448562832085,
      "grad_norm": 47.5,
      "learning_rate": 4.5324216515163954e-05,
      "loss": 5.0143,
      "num_input_tokens_seen": 10730512,
      "step": 7915
    },
    {
      "epoch": 0.25350489725369696,
      "grad_norm": 41.25,
      "learning_rate": 4.531849815824375e-05,
      "loss": 4.3006,
      "num_input_tokens_seen": 10737472,
      "step": 7920
    },
    {
      "epoch": 0.2536649382241854,
      "grad_norm": 54.5,
      "learning_rate": 4.5312776667955795e-05,
      "loss": 4.076,
      "num_input_tokens_seen": 10744304,
      "step": 7925
    },
    {
      "epoch": 0.25382497919467384,
      "grad_norm": 69.5,
      "learning_rate": 4.5307052045182405e-05,
      "loss": 6.1498,
      "num_input_tokens_seen": 10750848,
      "step": 7930
    },
    {
      "epoch": 0.2539850201651623,
      "grad_norm": 34.5,
      "learning_rate": 4.53013242908064e-05,
      "loss": 3.7437,
      "num_input_tokens_seen": 10757456,
      "step": 7935
    },
    {
      "epoch": 0.2541450611356507,
      "grad_norm": 38.25,
      "learning_rate": 4.529559340571107e-05,
      "loss": 4.0618,
      "num_input_tokens_seen": 10763808,
      "step": 7940
    },
    {
      "epoch": 0.25430510210613916,
      "grad_norm": 33.25,
      "learning_rate": 4.528985939078018e-05,
      "loss": 4.13,
      "num_input_tokens_seen": 10770112,
      "step": 7945
    },
    {
      "epoch": 0.2544651430766276,
      "grad_norm": 17.5,
      "learning_rate": 4.5284122246898e-05,
      "loss": 4.1755,
      "num_input_tokens_seen": 10777408,
      "step": 7950
    },
    {
      "epoch": 0.25462518404711604,
      "grad_norm": 26.0,
      "learning_rate": 4.527838197494926e-05,
      "loss": 4.3099,
      "num_input_tokens_seen": 10784272,
      "step": 7955
    },
    {
      "epoch": 0.25478522501760453,
      "grad_norm": 23.0,
      "learning_rate": 4.527263857581918e-05,
      "loss": 4.2976,
      "num_input_tokens_seen": 10791056,
      "step": 7960
    },
    {
      "epoch": 0.254945265988093,
      "grad_norm": 51.0,
      "learning_rate": 4.526689205039347e-05,
      "loss": 4.7298,
      "num_input_tokens_seen": 10797664,
      "step": 7965
    },
    {
      "epoch": 0.2551053069585814,
      "grad_norm": 21.125,
      "learning_rate": 4.5261142399558324e-05,
      "loss": 2.9771,
      "num_input_tokens_seen": 10804272,
      "step": 7970
    },
    {
      "epoch": 0.25526534792906985,
      "grad_norm": 25.125,
      "learning_rate": 4.525538962420041e-05,
      "loss": 3.718,
      "num_input_tokens_seen": 10811168,
      "step": 7975
    },
    {
      "epoch": 0.2554253888995583,
      "grad_norm": 42.75,
      "learning_rate": 4.524963372520685e-05,
      "loss": 3.742,
      "num_input_tokens_seen": 10818016,
      "step": 7980
    },
    {
      "epoch": 0.25558542987004673,
      "grad_norm": 14.0,
      "learning_rate": 4.524387470346531e-05,
      "loss": 4.0074,
      "num_input_tokens_seen": 10824704,
      "step": 7985
    },
    {
      "epoch": 0.25574547084053517,
      "grad_norm": 24.0,
      "learning_rate": 4.5238112559863885e-05,
      "loss": 5.0085,
      "num_input_tokens_seen": 10831552,
      "step": 7990
    },
    {
      "epoch": 0.2559055118110236,
      "grad_norm": 21.0,
      "learning_rate": 4.5232347295291175e-05,
      "loss": 3.6891,
      "num_input_tokens_seen": 10838192,
      "step": 7995
    },
    {
      "epoch": 0.25606555278151205,
      "grad_norm": 34.25,
      "learning_rate": 4.522657891063626e-05,
      "loss": 4.6755,
      "num_input_tokens_seen": 10844928,
      "step": 8000
    },
    {
      "epoch": 0.25606555278151205,
      "eval_loss": 4.492510795593262,
      "eval_runtime": 487.303,
      "eval_samples_per_second": 28.496,
      "eval_steps_per_second": 14.248,
      "num_input_tokens_seen": 10844928,
      "step": 8000
    },
    {
      "epoch": 0.2562255937520005,
      "grad_norm": 43.5,
      "learning_rate": 4.52208074067887e-05,
      "loss": 3.7405,
      "num_input_tokens_seen": 10851856,
      "step": 8005
    },
    {
      "epoch": 0.25638563472248893,
      "grad_norm": 42.25,
      "learning_rate": 4.5215032784638516e-05,
      "loss": 3.6878,
      "num_input_tokens_seen": 10858496,
      "step": 8010
    },
    {
      "epoch": 0.2565456756929774,
      "grad_norm": 44.0,
      "learning_rate": 4.5209255045076245e-05,
      "loss": 4.2909,
      "num_input_tokens_seen": 10865152,
      "step": 8015
    },
    {
      "epoch": 0.25670571666346587,
      "grad_norm": 18.0,
      "learning_rate": 4.5203474188992875e-05,
      "loss": 4.8338,
      "num_input_tokens_seen": 10872016,
      "step": 8020
    },
    {
      "epoch": 0.2568657576339543,
      "grad_norm": 31.625,
      "learning_rate": 4.51976902172799e-05,
      "loss": 3.5832,
      "num_input_tokens_seen": 10878624,
      "step": 8025
    },
    {
      "epoch": 0.25702579860444275,
      "grad_norm": 38.0,
      "learning_rate": 4.519190313082927e-05,
      "loss": 5.8551,
      "num_input_tokens_seen": 10885312,
      "step": 8030
    },
    {
      "epoch": 0.2571858395749312,
      "grad_norm": 36.75,
      "learning_rate": 4.518611293053343e-05,
      "loss": 3.7876,
      "num_input_tokens_seen": 10892416,
      "step": 8035
    },
    {
      "epoch": 0.2573458805454196,
      "grad_norm": 26.125,
      "learning_rate": 4.51803196172853e-05,
      "loss": 4.1427,
      "num_input_tokens_seen": 10899136,
      "step": 8040
    },
    {
      "epoch": 0.25750592151590807,
      "grad_norm": 42.0,
      "learning_rate": 4.517452319197828e-05,
      "loss": 4.3843,
      "num_input_tokens_seen": 10906048,
      "step": 8045
    },
    {
      "epoch": 0.2576659624863965,
      "grad_norm": 27.25,
      "learning_rate": 4.5168723655506265e-05,
      "loss": 4.6699,
      "num_input_tokens_seen": 10912800,
      "step": 8050
    },
    {
      "epoch": 0.25782600345688494,
      "grad_norm": 35.5,
      "learning_rate": 4.51629210087636e-05,
      "loss": 4.1075,
      "num_input_tokens_seen": 10919584,
      "step": 8055
    },
    {
      "epoch": 0.2579860444273734,
      "grad_norm": 29.875,
      "learning_rate": 4.515711525264513e-05,
      "loss": 3.8531,
      "num_input_tokens_seen": 10926576,
      "step": 8060
    },
    {
      "epoch": 0.2581460853978619,
      "grad_norm": 34.25,
      "learning_rate": 4.5151306388046175e-05,
      "loss": 4.945,
      "num_input_tokens_seen": 10933376,
      "step": 8065
    },
    {
      "epoch": 0.2583061263683503,
      "grad_norm": 18.125,
      "learning_rate": 4.514549441586255e-05,
      "loss": 3.4658,
      "num_input_tokens_seen": 10940448,
      "step": 8070
    },
    {
      "epoch": 0.25846616733883876,
      "grad_norm": 27.125,
      "learning_rate": 4.513967933699051e-05,
      "loss": 4.8936,
      "num_input_tokens_seen": 10947216,
      "step": 8075
    },
    {
      "epoch": 0.2586262083093272,
      "grad_norm": 23.125,
      "learning_rate": 4.513386115232684e-05,
      "loss": 3.9693,
      "num_input_tokens_seen": 10954080,
      "step": 8080
    },
    {
      "epoch": 0.25878624927981564,
      "grad_norm": 36.0,
      "learning_rate": 4.5128039862768745e-05,
      "loss": 5.2932,
      "num_input_tokens_seen": 10960624,
      "step": 8085
    },
    {
      "epoch": 0.2589462902503041,
      "grad_norm": 20.0,
      "learning_rate": 4.512221546921397e-05,
      "loss": 4.1937,
      "num_input_tokens_seen": 10967504,
      "step": 8090
    },
    {
      "epoch": 0.2591063312207925,
      "grad_norm": 31.625,
      "learning_rate": 4.5116387972560694e-05,
      "loss": 4.4937,
      "num_input_tokens_seen": 10974144,
      "step": 8095
    },
    {
      "epoch": 0.25926637219128096,
      "grad_norm": 53.0,
      "learning_rate": 4.511055737370759e-05,
      "loss": 4.2388,
      "num_input_tokens_seen": 10980592,
      "step": 8100
    },
    {
      "epoch": 0.2594264131617694,
      "grad_norm": 25.25,
      "learning_rate": 4.510472367355383e-05,
      "loss": 4.5681,
      "num_input_tokens_seen": 10987344,
      "step": 8105
    },
    {
      "epoch": 0.25958645413225784,
      "grad_norm": 29.0,
      "learning_rate": 4.509888687299901e-05,
      "loss": 4.2597,
      "num_input_tokens_seen": 10993872,
      "step": 8110
    },
    {
      "epoch": 0.2597464951027463,
      "grad_norm": 33.0,
      "learning_rate": 4.5093046972943266e-05,
      "loss": 4.1446,
      "num_input_tokens_seen": 11001152,
      "step": 8115
    },
    {
      "epoch": 0.2599065360732348,
      "grad_norm": 19.25,
      "learning_rate": 4.508720397428717e-05,
      "loss": 3.8108,
      "num_input_tokens_seen": 11007904,
      "step": 8120
    },
    {
      "epoch": 0.2600665770437232,
      "grad_norm": 27.0,
      "learning_rate": 4.508135787793178e-05,
      "loss": 4.3892,
      "num_input_tokens_seen": 11014624,
      "step": 8125
    },
    {
      "epoch": 0.26022661801421165,
      "grad_norm": 19.25,
      "learning_rate": 4.5075508684778664e-05,
      "loss": 4.6209,
      "num_input_tokens_seen": 11021632,
      "step": 8130
    },
    {
      "epoch": 0.2603866589847001,
      "grad_norm": 35.0,
      "learning_rate": 4.506965639572982e-05,
      "loss": 5.3289,
      "num_input_tokens_seen": 11028688,
      "step": 8135
    },
    {
      "epoch": 0.26054669995518853,
      "grad_norm": 21.875,
      "learning_rate": 4.506380101168774e-05,
      "loss": 4.4377,
      "num_input_tokens_seen": 11035120,
      "step": 8140
    },
    {
      "epoch": 0.26070674092567697,
      "grad_norm": 30.5,
      "learning_rate": 4.505794253355542e-05,
      "loss": 3.2056,
      "num_input_tokens_seen": 11042208,
      "step": 8145
    },
    {
      "epoch": 0.2608667818961654,
      "grad_norm": 24.625,
      "learning_rate": 4.5052080962236286e-05,
      "loss": 4.5235,
      "num_input_tokens_seen": 11048976,
      "step": 8150
    },
    {
      "epoch": 0.26102682286665385,
      "grad_norm": 39.0,
      "learning_rate": 4.504621629863428e-05,
      "loss": 5.0216,
      "num_input_tokens_seen": 11055376,
      "step": 8155
    },
    {
      "epoch": 0.2611868638371423,
      "grad_norm": 35.0,
      "learning_rate": 4.504034854365381e-05,
      "loss": 4.864,
      "num_input_tokens_seen": 11061952,
      "step": 8160
    },
    {
      "epoch": 0.26134690480763073,
      "grad_norm": 24.25,
      "learning_rate": 4.503447769819974e-05,
      "loss": 4.0143,
      "num_input_tokens_seen": 11068304,
      "step": 8165
    },
    {
      "epoch": 0.26150694577811917,
      "grad_norm": 27.5,
      "learning_rate": 4.502860376317745e-05,
      "loss": 4.4841,
      "num_input_tokens_seen": 11075168,
      "step": 8170
    },
    {
      "epoch": 0.26166698674860767,
      "grad_norm": 33.75,
      "learning_rate": 4.502272673949276e-05,
      "loss": 3.5393,
      "num_input_tokens_seen": 11081728,
      "step": 8175
    },
    {
      "epoch": 0.2618270277190961,
      "grad_norm": 27.375,
      "learning_rate": 4.501684662805199e-05,
      "loss": 4.3644,
      "num_input_tokens_seen": 11088016,
      "step": 8180
    },
    {
      "epoch": 0.26198706868958455,
      "grad_norm": 23.75,
      "learning_rate": 4.5010963429761924e-05,
      "loss": 4.9627,
      "num_input_tokens_seen": 11094816,
      "step": 8185
    },
    {
      "epoch": 0.262147109660073,
      "grad_norm": 31.375,
      "learning_rate": 4.500507714552982e-05,
      "loss": 3.9927,
      "num_input_tokens_seen": 11101824,
      "step": 8190
    },
    {
      "epoch": 0.2623071506305614,
      "grad_norm": 29.125,
      "learning_rate": 4.499918777626342e-05,
      "loss": 4.4934,
      "num_input_tokens_seen": 11108192,
      "step": 8195
    },
    {
      "epoch": 0.26246719160104987,
      "grad_norm": 25.875,
      "learning_rate": 4.499329532287093e-05,
      "loss": 3.8404,
      "num_input_tokens_seen": 11114800,
      "step": 8200
    },
    {
      "epoch": 0.26246719160104987,
      "eval_loss": 4.48245096206665,
      "eval_runtime": 487.6894,
      "eval_samples_per_second": 28.473,
      "eval_steps_per_second": 14.237,
      "num_input_tokens_seen": 11114800,
      "step": 8200
    },
    {
      "epoch": 0.2626272325715383,
      "grad_norm": 27.375,
      "learning_rate": 4.4987399786261064e-05,
      "loss": 4.1044,
      "num_input_tokens_seen": 11121440,
      "step": 8205
    },
    {
      "epoch": 0.26278727354202674,
      "grad_norm": 27.875,
      "learning_rate": 4.498150116734297e-05,
      "loss": 4.3994,
      "num_input_tokens_seen": 11127584,
      "step": 8210
    },
    {
      "epoch": 0.2629473145125152,
      "grad_norm": 34.0,
      "learning_rate": 4.4975599467026294e-05,
      "loss": 5.5417,
      "num_input_tokens_seen": 11134496,
      "step": 8215
    },
    {
      "epoch": 0.2631073554830036,
      "grad_norm": 31.5,
      "learning_rate": 4.496969468622114e-05,
      "loss": 4.6976,
      "num_input_tokens_seen": 11140928,
      "step": 8220
    },
    {
      "epoch": 0.2632673964534921,
      "grad_norm": 26.625,
      "learning_rate": 4.496378682583813e-05,
      "loss": 3.9957,
      "num_input_tokens_seen": 11147568,
      "step": 8225
    },
    {
      "epoch": 0.26342743742398056,
      "grad_norm": 14.4375,
      "learning_rate": 4.495787588678829e-05,
      "loss": 4.7343,
      "num_input_tokens_seen": 11154400,
      "step": 8230
    },
    {
      "epoch": 0.263587478394469,
      "grad_norm": 35.0,
      "learning_rate": 4.4951961869983196e-05,
      "loss": 4.9514,
      "num_input_tokens_seen": 11161024,
      "step": 8235
    },
    {
      "epoch": 0.26374751936495744,
      "grad_norm": 26.375,
      "learning_rate": 4.494604477633485e-05,
      "loss": 4.6574,
      "num_input_tokens_seen": 11167536,
      "step": 8240
    },
    {
      "epoch": 0.2639075603354459,
      "grad_norm": 43.25,
      "learning_rate": 4.4940124606755734e-05,
      "loss": 4.7451,
      "num_input_tokens_seen": 11174608,
      "step": 8245
    },
    {
      "epoch": 0.2640676013059343,
      "grad_norm": 24.125,
      "learning_rate": 4.493420136215882e-05,
      "loss": 4.5522,
      "num_input_tokens_seen": 11181360,
      "step": 8250
    },
    {
      "epoch": 0.26422764227642276,
      "grad_norm": 17.5,
      "learning_rate": 4.492827504345756e-05,
      "loss": 4.283,
      "num_input_tokens_seen": 11188576,
      "step": 8255
    },
    {
      "epoch": 0.2643876832469112,
      "grad_norm": 19.625,
      "learning_rate": 4.492234565156584e-05,
      "loss": 3.9288,
      "num_input_tokens_seen": 11194976,
      "step": 8260
    },
    {
      "epoch": 0.26454772421739964,
      "grad_norm": 16.125,
      "learning_rate": 4.491641318739807e-05,
      "loss": 4.8168,
      "num_input_tokens_seen": 11201712,
      "step": 8265
    },
    {
      "epoch": 0.2647077651878881,
      "grad_norm": 31.375,
      "learning_rate": 4.4910477651869096e-05,
      "loss": 4.9047,
      "num_input_tokens_seen": 11208096,
      "step": 8270
    },
    {
      "epoch": 0.2648678061583765,
      "grad_norm": 26.0,
      "learning_rate": 4.4904539045894254e-05,
      "loss": 3.6468,
      "num_input_tokens_seen": 11214752,
      "step": 8275
    },
    {
      "epoch": 0.265027847128865,
      "grad_norm": 25.5,
      "learning_rate": 4.4898597370389364e-05,
      "loss": 3.0806,
      "num_input_tokens_seen": 11221840,
      "step": 8280
    },
    {
      "epoch": 0.26518788809935345,
      "grad_norm": 25.125,
      "learning_rate": 4.489265262627069e-05,
      "loss": 3.8427,
      "num_input_tokens_seen": 11228256,
      "step": 8285
    },
    {
      "epoch": 0.2653479290698419,
      "grad_norm": 26.0,
      "learning_rate": 4.488670481445499e-05,
      "loss": 3.9925,
      "num_input_tokens_seen": 11234896,
      "step": 8290
    },
    {
      "epoch": 0.26550797004033033,
      "grad_norm": 17.75,
      "learning_rate": 4.488075393585951e-05,
      "loss": 3.4629,
      "num_input_tokens_seen": 11241648,
      "step": 8295
    },
    {
      "epoch": 0.26566801101081877,
      "grad_norm": 21.375,
      "learning_rate": 4.487479999140193e-05,
      "loss": 3.592,
      "num_input_tokens_seen": 11248800,
      "step": 8300
    },
    {
      "epoch": 0.2658280519813072,
      "grad_norm": 15.6875,
      "learning_rate": 4.4868842982000425e-05,
      "loss": 3.588,
      "num_input_tokens_seen": 11255120,
      "step": 8305
    },
    {
      "epoch": 0.26598809295179565,
      "grad_norm": 41.25,
      "learning_rate": 4.486288290857365e-05,
      "loss": 5.0872,
      "num_input_tokens_seen": 11261760,
      "step": 8310
    },
    {
      "epoch": 0.2661481339222841,
      "grad_norm": 34.5,
      "learning_rate": 4.4856919772040715e-05,
      "loss": 3.9405,
      "num_input_tokens_seen": 11268384,
      "step": 8315
    },
    {
      "epoch": 0.26630817489277253,
      "grad_norm": 25.5,
      "learning_rate": 4.485095357332122e-05,
      "loss": 4.9755,
      "num_input_tokens_seen": 11274880,
      "step": 8320
    },
    {
      "epoch": 0.26646821586326097,
      "grad_norm": 29.75,
      "learning_rate": 4.484498431333521e-05,
      "loss": 4.0566,
      "num_input_tokens_seen": 11281696,
      "step": 8325
    },
    {
      "epoch": 0.26662825683374947,
      "grad_norm": 29.0,
      "learning_rate": 4.4839011993003245e-05,
      "loss": 4.0929,
      "num_input_tokens_seen": 11288640,
      "step": 8330
    },
    {
      "epoch": 0.2667882978042379,
      "grad_norm": 21.75,
      "learning_rate": 4.4833036613246305e-05,
      "loss": 4.578,
      "num_input_tokens_seen": 11295648,
      "step": 8335
    },
    {
      "epoch": 0.26694833877472635,
      "grad_norm": 30.375,
      "learning_rate": 4.482705817498589e-05,
      "loss": 3.9722,
      "num_input_tokens_seen": 11302176,
      "step": 8340
    },
    {
      "epoch": 0.2671083797452148,
      "grad_norm": 17.875,
      "learning_rate": 4.4821076679143934e-05,
      "loss": 3.4029,
      "num_input_tokens_seen": 11308816,
      "step": 8345
    },
    {
      "epoch": 0.2672684207157032,
      "grad_norm": 23.75,
      "learning_rate": 4.481509212664288e-05,
      "loss": 3.9069,
      "num_input_tokens_seen": 11316736,
      "step": 8350
    },
    {
      "epoch": 0.26742846168619167,
      "grad_norm": 28.0,
      "learning_rate": 4.480910451840559e-05,
      "loss": 6.0627,
      "num_input_tokens_seen": 11323408,
      "step": 8355
    },
    {
      "epoch": 0.2675885026566801,
      "grad_norm": 29.375,
      "learning_rate": 4.480311385535546e-05,
      "loss": 4.7256,
      "num_input_tokens_seen": 11330256,
      "step": 8360
    },
    {
      "epoch": 0.26774854362716854,
      "grad_norm": 34.75,
      "learning_rate": 4.47971201384163e-05,
      "loss": 4.3341,
      "num_input_tokens_seen": 11336768,
      "step": 8365
    },
    {
      "epoch": 0.267908584597657,
      "grad_norm": 40.25,
      "learning_rate": 4.4791123368512446e-05,
      "loss": 4.4465,
      "num_input_tokens_seen": 11343296,
      "step": 8370
    },
    {
      "epoch": 0.2680686255681454,
      "grad_norm": 16.25,
      "learning_rate": 4.478512354656864e-05,
      "loss": 4.4561,
      "num_input_tokens_seen": 11349728,
      "step": 8375
    },
    {
      "epoch": 0.26822866653863386,
      "grad_norm": 19.75,
      "learning_rate": 4.477912067351016e-05,
      "loss": 4.4129,
      "num_input_tokens_seen": 11356624,
      "step": 8380
    },
    {
      "epoch": 0.26838870750912236,
      "grad_norm": 36.25,
      "learning_rate": 4.477311475026271e-05,
      "loss": 4.4363,
      "num_input_tokens_seen": 11363136,
      "step": 8385
    },
    {
      "epoch": 0.2685487484796108,
      "grad_norm": 12.5625,
      "learning_rate": 4.476710577775248e-05,
      "loss": 3.298,
      "num_input_tokens_seen": 11370112,
      "step": 8390
    },
    {
      "epoch": 0.26870878945009924,
      "grad_norm": 31.5,
      "learning_rate": 4.476109375690612e-05,
      "loss": 5.1154,
      "num_input_tokens_seen": 11376688,
      "step": 8395
    },
    {
      "epoch": 0.2688688304205877,
      "grad_norm": 21.0,
      "learning_rate": 4.4755078688650784e-05,
      "loss": 4.5697,
      "num_input_tokens_seen": 11383280,
      "step": 8400
    },
    {
      "epoch": 0.2688688304205877,
      "eval_loss": 4.4838948249816895,
      "eval_runtime": 487.0382,
      "eval_samples_per_second": 28.511,
      "eval_steps_per_second": 14.256,
      "num_input_tokens_seen": 11383280,
      "step": 8400
    },
    {
      "epoch": 0.2690288713910761,
      "grad_norm": 27.375,
      "learning_rate": 4.474906057391406e-05,
      "loss": 3.5569,
      "num_input_tokens_seen": 11390208,
      "step": 8405
    },
    {
      "epoch": 0.26918891236156456,
      "grad_norm": 30.875,
      "learning_rate": 4.4743039413624e-05,
      "loss": 4.0056,
      "num_input_tokens_seen": 11397072,
      "step": 8410
    },
    {
      "epoch": 0.269348953332053,
      "grad_norm": 49.5,
      "learning_rate": 4.473701520870916e-05,
      "loss": 4.3525,
      "num_input_tokens_seen": 11403584,
      "step": 8415
    },
    {
      "epoch": 0.26950899430254144,
      "grad_norm": 44.75,
      "learning_rate": 4.4730987960098544e-05,
      "loss": 3.6571,
      "num_input_tokens_seen": 11409904,
      "step": 8420
    },
    {
      "epoch": 0.2696690352730299,
      "grad_norm": 56.75,
      "learning_rate": 4.4724957668721635e-05,
      "loss": 4.6051,
      "num_input_tokens_seen": 11416336,
      "step": 8425
    },
    {
      "epoch": 0.2698290762435183,
      "grad_norm": 26.875,
      "learning_rate": 4.471892433550836e-05,
      "loss": 4.442,
      "num_input_tokens_seen": 11423184,
      "step": 8430
    },
    {
      "epoch": 0.2699891172140068,
      "grad_norm": 38.25,
      "learning_rate": 4.471288796138916e-05,
      "loss": 4.4392,
      "num_input_tokens_seen": 11429968,
      "step": 8435
    },
    {
      "epoch": 0.27014915818449525,
      "grad_norm": 18.0,
      "learning_rate": 4.470684854729491e-05,
      "loss": 4.4202,
      "num_input_tokens_seen": 11436816,
      "step": 8440
    },
    {
      "epoch": 0.2703091991549837,
      "grad_norm": 19.625,
      "learning_rate": 4.4700806094156955e-05,
      "loss": 4.0926,
      "num_input_tokens_seen": 11444192,
      "step": 8445
    },
    {
      "epoch": 0.27046924012547213,
      "grad_norm": 18.0,
      "learning_rate": 4.469476060290713e-05,
      "loss": 3.9648,
      "num_input_tokens_seen": 11450976,
      "step": 8450
    },
    {
      "epoch": 0.27062928109596057,
      "grad_norm": 24.5,
      "learning_rate": 4.468871207447772e-05,
      "loss": 3.2567,
      "num_input_tokens_seen": 11457616,
      "step": 8455
    },
    {
      "epoch": 0.270789322066449,
      "grad_norm": 26.125,
      "learning_rate": 4.4682660509801486e-05,
      "loss": 4.7008,
      "num_input_tokens_seen": 11464096,
      "step": 8460
    },
    {
      "epoch": 0.27094936303693745,
      "grad_norm": 23.125,
      "learning_rate": 4.467660590981165e-05,
      "loss": 3.969,
      "num_input_tokens_seen": 11470800,
      "step": 8465
    },
    {
      "epoch": 0.2711094040074259,
      "grad_norm": 29.5,
      "learning_rate": 4.467054827544191e-05,
      "loss": 4.6958,
      "num_input_tokens_seen": 11477184,
      "step": 8470
    },
    {
      "epoch": 0.27126944497791433,
      "grad_norm": 19.5,
      "learning_rate": 4.4664487607626434e-05,
      "loss": 3.5614,
      "num_input_tokens_seen": 11484096,
      "step": 8475
    },
    {
      "epoch": 0.27142948594840277,
      "grad_norm": 24.75,
      "learning_rate": 4.4658423907299845e-05,
      "loss": 4.9294,
      "num_input_tokens_seen": 11490864,
      "step": 8480
    },
    {
      "epoch": 0.2715895269188912,
      "grad_norm": 21.75,
      "learning_rate": 4.465235717539725e-05,
      "loss": 3.4586,
      "num_input_tokens_seen": 11497456,
      "step": 8485
    },
    {
      "epoch": 0.2717495678893797,
      "grad_norm": 27.25,
      "learning_rate": 4.464628741285421e-05,
      "loss": 3.6255,
      "num_input_tokens_seen": 11503968,
      "step": 8490
    },
    {
      "epoch": 0.27190960885986815,
      "grad_norm": 49.5,
      "learning_rate": 4.4640214620606754e-05,
      "loss": 4.2184,
      "num_input_tokens_seen": 11510336,
      "step": 8495
    },
    {
      "epoch": 0.2720696498303566,
      "grad_norm": 24.0,
      "learning_rate": 4.46341387995914e-05,
      "loss": 3.4543,
      "num_input_tokens_seen": 11517136,
      "step": 8500
    },
    {
      "epoch": 0.272229690800845,
      "grad_norm": 28.625,
      "learning_rate": 4.4628059950745106e-05,
      "loss": 4.9673,
      "num_input_tokens_seen": 11524240,
      "step": 8505
    },
    {
      "epoch": 0.27238973177133347,
      "grad_norm": 30.5,
      "learning_rate": 4.4621978075005297e-05,
      "loss": 4.2333,
      "num_input_tokens_seen": 11531456,
      "step": 8510
    },
    {
      "epoch": 0.2725497727418219,
      "grad_norm": 40.5,
      "learning_rate": 4.461589317330989e-05,
      "loss": 4.6975,
      "num_input_tokens_seen": 11538816,
      "step": 8515
    },
    {
      "epoch": 0.27270981371231034,
      "grad_norm": 24.25,
      "learning_rate": 4.460980524659724e-05,
      "loss": 3.9576,
      "num_input_tokens_seen": 11545520,
      "step": 8520
    },
    {
      "epoch": 0.2728698546827988,
      "grad_norm": 43.75,
      "learning_rate": 4.46037142958062e-05,
      "loss": 4.3148,
      "num_input_tokens_seen": 11552256,
      "step": 8525
    },
    {
      "epoch": 0.2730298956532872,
      "grad_norm": 32.0,
      "learning_rate": 4.4597620321876046e-05,
      "loss": 4.4898,
      "num_input_tokens_seen": 11559040,
      "step": 8530
    },
    {
      "epoch": 0.27318993662377566,
      "grad_norm": 17.125,
      "learning_rate": 4.459152332574656e-05,
      "loss": 4.1011,
      "num_input_tokens_seen": 11565616,
      "step": 8535
    },
    {
      "epoch": 0.2733499775942641,
      "grad_norm": 24.125,
      "learning_rate": 4.4585423308357985e-05,
      "loss": 3.6085,
      "num_input_tokens_seen": 11572256,
      "step": 8540
    },
    {
      "epoch": 0.2735100185647526,
      "grad_norm": 29.75,
      "learning_rate": 4.457932027065102e-05,
      "loss": 4.412,
      "num_input_tokens_seen": 11579136,
      "step": 8545
    },
    {
      "epoch": 0.27367005953524104,
      "grad_norm": 21.0,
      "learning_rate": 4.45732142135668e-05,
      "loss": 3.6916,
      "num_input_tokens_seen": 11586080,
      "step": 8550
    },
    {
      "epoch": 0.2738301005057295,
      "grad_norm": 28.125,
      "learning_rate": 4.4567105138046986e-05,
      "loss": 3.9703,
      "num_input_tokens_seen": 11592800,
      "step": 8555
    },
    {
      "epoch": 0.2739901414762179,
      "grad_norm": 27.125,
      "learning_rate": 4.456099304503365e-05,
      "loss": 3.7982,
      "num_input_tokens_seen": 11599680,
      "step": 8560
    },
    {
      "epoch": 0.27415018244670636,
      "grad_norm": 31.75,
      "learning_rate": 4.455487793546939e-05,
      "loss": 4.0288,
      "num_input_tokens_seen": 11606464,
      "step": 8565
    },
    {
      "epoch": 0.2743102234171948,
      "grad_norm": 32.5,
      "learning_rate": 4.454875981029719e-05,
      "loss": 3.9312,
      "num_input_tokens_seen": 11613216,
      "step": 8570
    },
    {
      "epoch": 0.27447026438768324,
      "grad_norm": 30.0,
      "learning_rate": 4.454263867046057e-05,
      "loss": 4.1271,
      "num_input_tokens_seen": 11619632,
      "step": 8575
    },
    {
      "epoch": 0.2746303053581717,
      "grad_norm": 29.75,
      "learning_rate": 4.4536514516903484e-05,
      "loss": 5.147,
      "num_input_tokens_seen": 11625952,
      "step": 8580
    },
    {
      "epoch": 0.2747903463286601,
      "grad_norm": 29.625,
      "learning_rate": 4.453038735057034e-05,
      "loss": 4.6404,
      "num_input_tokens_seen": 11632528,
      "step": 8585
    },
    {
      "epoch": 0.27495038729914856,
      "grad_norm": 37.75,
      "learning_rate": 4.4524257172406034e-05,
      "loss": 3.5549,
      "num_input_tokens_seen": 11639200,
      "step": 8590
    },
    {
      "epoch": 0.27511042826963705,
      "grad_norm": 13.875,
      "learning_rate": 4.451812398335592e-05,
      "loss": 4.565,
      "num_input_tokens_seen": 11645696,
      "step": 8595
    },
    {
      "epoch": 0.2752704692401255,
      "grad_norm": 39.5,
      "learning_rate": 4.4511987784365805e-05,
      "loss": 5.3488,
      "num_input_tokens_seen": 11652336,
      "step": 8600
    },
    {
      "epoch": 0.2752704692401255,
      "eval_loss": 4.4746503829956055,
      "eval_runtime": 486.9899,
      "eval_samples_per_second": 28.514,
      "eval_steps_per_second": 14.257,
      "num_input_tokens_seen": 11652336,
      "step": 8600
    },
    {
      "epoch": 0.27543051021061393,
      "grad_norm": 35.75,
      "learning_rate": 4.450584857638197e-05,
      "loss": 4.3436,
      "num_input_tokens_seen": 11658688,
      "step": 8605
    },
    {
      "epoch": 0.2755905511811024,
      "grad_norm": 37.25,
      "learning_rate": 4.449970636035116e-05,
      "loss": 4.0956,
      "num_input_tokens_seen": 11665488,
      "step": 8610
    },
    {
      "epoch": 0.2757505921515908,
      "grad_norm": 39.75,
      "learning_rate": 4.4493561137220574e-05,
      "loss": 3.971,
      "num_input_tokens_seen": 11672528,
      "step": 8615
    },
    {
      "epoch": 0.27591063312207925,
      "grad_norm": 38.5,
      "learning_rate": 4.44874129079379e-05,
      "loss": 3.6719,
      "num_input_tokens_seen": 11679248,
      "step": 8620
    },
    {
      "epoch": 0.2760706740925677,
      "grad_norm": 31.5,
      "learning_rate": 4.4481261673451255e-05,
      "loss": 4.1315,
      "num_input_tokens_seen": 11685968,
      "step": 8625
    },
    {
      "epoch": 0.27623071506305613,
      "grad_norm": 30.125,
      "learning_rate": 4.4475107434709245e-05,
      "loss": 5.2953,
      "num_input_tokens_seen": 11692688,
      "step": 8630
    },
    {
      "epoch": 0.27639075603354457,
      "grad_norm": 27.375,
      "learning_rate": 4.446895019266093e-05,
      "loss": 5.074,
      "num_input_tokens_seen": 11699376,
      "step": 8635
    },
    {
      "epoch": 0.276550797004033,
      "grad_norm": 25.875,
      "learning_rate": 4.446278994825583e-05,
      "loss": 4.4267,
      "num_input_tokens_seen": 11705776,
      "step": 8640
    },
    {
      "epoch": 0.27671083797452145,
      "grad_norm": 22.875,
      "learning_rate": 4.445662670244394e-05,
      "loss": 4.2185,
      "num_input_tokens_seen": 11712480,
      "step": 8645
    },
    {
      "epoch": 0.27687087894500995,
      "grad_norm": 20.0,
      "learning_rate": 4.44504604561757e-05,
      "loss": 3.3415,
      "num_input_tokens_seen": 11719376,
      "step": 8650
    },
    {
      "epoch": 0.2770309199154984,
      "grad_norm": 17.625,
      "learning_rate": 4.4444291210402035e-05,
      "loss": 4.7519,
      "num_input_tokens_seen": 11725776,
      "step": 8655
    },
    {
      "epoch": 0.2771909608859868,
      "grad_norm": 23.125,
      "learning_rate": 4.443811896607431e-05,
      "loss": 3.8262,
      "num_input_tokens_seen": 11732880,
      "step": 8660
    },
    {
      "epoch": 0.27735100185647527,
      "grad_norm": 49.5,
      "learning_rate": 4.443194372414436e-05,
      "loss": 4.4489,
      "num_input_tokens_seen": 11739536,
      "step": 8665
    },
    {
      "epoch": 0.2775110428269637,
      "grad_norm": 41.25,
      "learning_rate": 4.442576548556449e-05,
      "loss": 4.6553,
      "num_input_tokens_seen": 11746544,
      "step": 8670
    },
    {
      "epoch": 0.27767108379745215,
      "grad_norm": 23.125,
      "learning_rate": 4.441958425128747e-05,
      "loss": 4.8804,
      "num_input_tokens_seen": 11753472,
      "step": 8675
    },
    {
      "epoch": 0.2778311247679406,
      "grad_norm": 24.75,
      "learning_rate": 4.4413400022266515e-05,
      "loss": 4.8071,
      "num_input_tokens_seen": 11760400,
      "step": 8680
    },
    {
      "epoch": 0.277991165738429,
      "grad_norm": 41.0,
      "learning_rate": 4.4407212799455313e-05,
      "loss": 5.1601,
      "num_input_tokens_seen": 11767440,
      "step": 8685
    },
    {
      "epoch": 0.27815120670891746,
      "grad_norm": 51.5,
      "learning_rate": 4.4401022583808003e-05,
      "loss": 4.0669,
      "num_input_tokens_seen": 11774192,
      "step": 8690
    },
    {
      "epoch": 0.2783112476794059,
      "grad_norm": 23.125,
      "learning_rate": 4.439482937627921e-05,
      "loss": 3.7152,
      "num_input_tokens_seen": 11780960,
      "step": 8695
    },
    {
      "epoch": 0.2784712886498944,
      "grad_norm": 30.75,
      "learning_rate": 4.4388633177824004e-05,
      "loss": 4.1284,
      "num_input_tokens_seen": 11787312,
      "step": 8700
    },
    {
      "epoch": 0.27863132962038284,
      "grad_norm": 36.0,
      "learning_rate": 4.4382433989397895e-05,
      "loss": 4.379,
      "num_input_tokens_seen": 11793920,
      "step": 8705
    },
    {
      "epoch": 0.2787913705908713,
      "grad_norm": 22.875,
      "learning_rate": 4.4376231811956895e-05,
      "loss": 4.2268,
      "num_input_tokens_seen": 11800848,
      "step": 8710
    },
    {
      "epoch": 0.2789514115613597,
      "grad_norm": 28.125,
      "learning_rate": 4.437002664645745e-05,
      "loss": 4.3836,
      "num_input_tokens_seen": 11807888,
      "step": 8715
    },
    {
      "epoch": 0.27911145253184816,
      "grad_norm": 41.75,
      "learning_rate": 4.436381849385649e-05,
      "loss": 5.1916,
      "num_input_tokens_seen": 11814688,
      "step": 8720
    },
    {
      "epoch": 0.2792714935023366,
      "grad_norm": 23.625,
      "learning_rate": 4.435760735511136e-05,
      "loss": 3.8916,
      "num_input_tokens_seen": 11821696,
      "step": 8725
    },
    {
      "epoch": 0.27943153447282504,
      "grad_norm": 34.75,
      "learning_rate": 4.435139323117992e-05,
      "loss": 5.3303,
      "num_input_tokens_seen": 11828368,
      "step": 8730
    },
    {
      "epoch": 0.2795915754433135,
      "grad_norm": 50.25,
      "learning_rate": 4.434517612302046e-05,
      "loss": 4.4308,
      "num_input_tokens_seen": 11834896,
      "step": 8735
    },
    {
      "epoch": 0.2797516164138019,
      "grad_norm": 17.375,
      "learning_rate": 4.433895603159174e-05,
      "loss": 4.0983,
      "num_input_tokens_seen": 11841584,
      "step": 8740
    },
    {
      "epoch": 0.27991165738429036,
      "grad_norm": 34.0,
      "learning_rate": 4.433273295785296e-05,
      "loss": 4.0606,
      "num_input_tokens_seen": 11848304,
      "step": 8745
    },
    {
      "epoch": 0.2800716983547788,
      "grad_norm": 29.25,
      "learning_rate": 4.432650690276382e-05,
      "loss": 4.0285,
      "num_input_tokens_seen": 11855200,
      "step": 8750
    },
    {
      "epoch": 0.2802317393252673,
      "grad_norm": 27.75,
      "learning_rate": 4.4320277867284435e-05,
      "loss": 3.9377,
      "num_input_tokens_seen": 11862400,
      "step": 8755
    },
    {
      "epoch": 0.28039178029575573,
      "grad_norm": 30.5,
      "learning_rate": 4.431404585237541e-05,
      "loss": 4.7813,
      "num_input_tokens_seen": 11869232,
      "step": 8760
    },
    {
      "epoch": 0.2805518212662442,
      "grad_norm": 25.0,
      "learning_rate": 4.43078108589978e-05,
      "loss": 4.0827,
      "num_input_tokens_seen": 11875760,
      "step": 8765
    },
    {
      "epoch": 0.2807118622367326,
      "grad_norm": 22.25,
      "learning_rate": 4.4301572888113116e-05,
      "loss": 4.4674,
      "num_input_tokens_seen": 11882480,
      "step": 8770
    },
    {
      "epoch": 0.28087190320722105,
      "grad_norm": 22.5,
      "learning_rate": 4.4295331940683337e-05,
      "loss": 3.5522,
      "num_input_tokens_seen": 11889472,
      "step": 8775
    },
    {
      "epoch": 0.2810319441777095,
      "grad_norm": 29.125,
      "learning_rate": 4.428908801767089e-05,
      "loss": 4.2354,
      "num_input_tokens_seen": 11896464,
      "step": 8780
    },
    {
      "epoch": 0.28119198514819793,
      "grad_norm": 34.25,
      "learning_rate": 4.428284112003868e-05,
      "loss": 5.0189,
      "num_input_tokens_seen": 11902912,
      "step": 8785
    },
    {
      "epoch": 0.28135202611868637,
      "grad_norm": 38.5,
      "learning_rate": 4.4276591248750033e-05,
      "loss": 5.6576,
      "num_input_tokens_seen": 11910192,
      "step": 8790
    },
    {
      "epoch": 0.2815120670891748,
      "grad_norm": 48.25,
      "learning_rate": 4.4270338404768774e-05,
      "loss": 4.7188,
      "num_input_tokens_seen": 11917424,
      "step": 8795
    },
    {
      "epoch": 0.28167210805966325,
      "grad_norm": 24.375,
      "learning_rate": 4.426408258905917e-05,
      "loss": 4.5469,
      "num_input_tokens_seen": 11924224,
      "step": 8800
    },
    {
      "epoch": 0.28167210805966325,
      "eval_loss": 4.492716312408447,
      "eval_runtime": 487.2717,
      "eval_samples_per_second": 28.497,
      "eval_steps_per_second": 14.249,
      "num_input_tokens_seen": 11924224,
      "step": 8800
    },
    {
      "epoch": 0.28183214903015175,
      "grad_norm": 44.75,
      "learning_rate": 4.425782380258594e-05,
      "loss": 4.4708,
      "num_input_tokens_seen": 11930928,
      "step": 8805
    },
    {
      "epoch": 0.2819921900006402,
      "grad_norm": 31.75,
      "learning_rate": 4.425156204631427e-05,
      "loss": 4.9354,
      "num_input_tokens_seen": 11937632,
      "step": 8810
    },
    {
      "epoch": 0.2821522309711286,
      "grad_norm": 35.5,
      "learning_rate": 4.424529732120981e-05,
      "loss": 4.2925,
      "num_input_tokens_seen": 11944208,
      "step": 8815
    },
    {
      "epoch": 0.28231227194161707,
      "grad_norm": 21.0,
      "learning_rate": 4.423902962823864e-05,
      "loss": 5.4318,
      "num_input_tokens_seen": 11951056,
      "step": 8820
    },
    {
      "epoch": 0.2824723129121055,
      "grad_norm": 21.375,
      "learning_rate": 4.423275896836733e-05,
      "loss": 4.004,
      "num_input_tokens_seen": 11958192,
      "step": 8825
    },
    {
      "epoch": 0.28263235388259395,
      "grad_norm": 39.75,
      "learning_rate": 4.42264853425629e-05,
      "loss": 3.9658,
      "num_input_tokens_seen": 11965024,
      "step": 8830
    },
    {
      "epoch": 0.2827923948530824,
      "grad_norm": 56.25,
      "learning_rate": 4.4220208751792816e-05,
      "loss": 4.2482,
      "num_input_tokens_seen": 11971968,
      "step": 8835
    },
    {
      "epoch": 0.2829524358235708,
      "grad_norm": 35.25,
      "learning_rate": 4.421392919702499e-05,
      "loss": 4.4549,
      "num_input_tokens_seen": 11978560,
      "step": 8840
    },
    {
      "epoch": 0.28311247679405926,
      "grad_norm": 20.625,
      "learning_rate": 4.4207646679227846e-05,
      "loss": 4.4658,
      "num_input_tokens_seen": 11985296,
      "step": 8845
    },
    {
      "epoch": 0.2832725177645477,
      "grad_norm": 58.5,
      "learning_rate": 4.42013611993702e-05,
      "loss": 4.2692,
      "num_input_tokens_seen": 11991888,
      "step": 8850
    },
    {
      "epoch": 0.28343255873503614,
      "grad_norm": 19.125,
      "learning_rate": 4.419507275842135e-05,
      "loss": 3.9911,
      "num_input_tokens_seen": 11998512,
      "step": 8855
    },
    {
      "epoch": 0.28359259970552464,
      "grad_norm": 39.0,
      "learning_rate": 4.418878135735106e-05,
      "loss": 5.0946,
      "num_input_tokens_seen": 12005696,
      "step": 8860
    },
    {
      "epoch": 0.2837526406760131,
      "grad_norm": 33.75,
      "learning_rate": 4.418248699712955e-05,
      "loss": 5.1307,
      "num_input_tokens_seen": 12012128,
      "step": 8865
    },
    {
      "epoch": 0.2839126816465015,
      "grad_norm": 32.25,
      "learning_rate": 4.417618967872748e-05,
      "loss": 5.1847,
      "num_input_tokens_seen": 12018704,
      "step": 8870
    },
    {
      "epoch": 0.28407272261698996,
      "grad_norm": 40.75,
      "learning_rate": 4.4169889403115985e-05,
      "loss": 5.1113,
      "num_input_tokens_seen": 12025616,
      "step": 8875
    },
    {
      "epoch": 0.2842327635874784,
      "grad_norm": 29.375,
      "learning_rate": 4.4163586171266627e-05,
      "loss": 4.7912,
      "num_input_tokens_seen": 12032304,
      "step": 8880
    },
    {
      "epoch": 0.28439280455796684,
      "grad_norm": 12.9375,
      "learning_rate": 4.415727998415147e-05,
      "loss": 4.3261,
      "num_input_tokens_seen": 12039344,
      "step": 8885
    },
    {
      "epoch": 0.2845528455284553,
      "grad_norm": 27.25,
      "learning_rate": 4.4150970842742985e-05,
      "loss": 4.3232,
      "num_input_tokens_seen": 12046064,
      "step": 8890
    },
    {
      "epoch": 0.2847128864989437,
      "grad_norm": 22.625,
      "learning_rate": 4.4144658748014134e-05,
      "loss": 4.9497,
      "num_input_tokens_seen": 12052832,
      "step": 8895
    },
    {
      "epoch": 0.28487292746943216,
      "grad_norm": 37.0,
      "learning_rate": 4.413834370093831e-05,
      "loss": 4.1006,
      "num_input_tokens_seen": 12059632,
      "step": 8900
    },
    {
      "epoch": 0.2850329684399206,
      "grad_norm": 32.0,
      "learning_rate": 4.413202570248939e-05,
      "loss": 4.7615,
      "num_input_tokens_seen": 12066880,
      "step": 8905
    },
    {
      "epoch": 0.28519300941040904,
      "grad_norm": 36.0,
      "learning_rate": 4.412570475364167e-05,
      "loss": 4.7059,
      "num_input_tokens_seen": 12073600,
      "step": 8910
    },
    {
      "epoch": 0.28535305038089753,
      "grad_norm": 24.5,
      "learning_rate": 4.411938085536994e-05,
      "loss": 4.6845,
      "num_input_tokens_seen": 12080368,
      "step": 8915
    },
    {
      "epoch": 0.285513091351386,
      "grad_norm": 19.5,
      "learning_rate": 4.41130540086494e-05,
      "loss": 4.0489,
      "num_input_tokens_seen": 12087120,
      "step": 8920
    },
    {
      "epoch": 0.2856731323218744,
      "grad_norm": 40.25,
      "learning_rate": 4.4106724214455754e-05,
      "loss": 4.3872,
      "num_input_tokens_seen": 12093776,
      "step": 8925
    },
    {
      "epoch": 0.28583317329236285,
      "grad_norm": 29.125,
      "learning_rate": 4.4100391473765115e-05,
      "loss": 4.1187,
      "num_input_tokens_seen": 12100208,
      "step": 8930
    },
    {
      "epoch": 0.2859932142628513,
      "grad_norm": 43.75,
      "learning_rate": 4.409405578755408e-05,
      "loss": 4.0915,
      "num_input_tokens_seen": 12107008,
      "step": 8935
    },
    {
      "epoch": 0.28615325523333973,
      "grad_norm": 17.875,
      "learning_rate": 4.4087717156799705e-05,
      "loss": 4.1618,
      "num_input_tokens_seen": 12113520,
      "step": 8940
    },
    {
      "epoch": 0.28631329620382817,
      "grad_norm": 18.875,
      "learning_rate": 4.408137558247946e-05,
      "loss": 4.6166,
      "num_input_tokens_seen": 12120192,
      "step": 8945
    },
    {
      "epoch": 0.2864733371743166,
      "grad_norm": 60.25,
      "learning_rate": 4.4075031065571306e-05,
      "loss": 3.8721,
      "num_input_tokens_seen": 12127328,
      "step": 8950
    },
    {
      "epoch": 0.28663337814480505,
      "grad_norm": 14.25,
      "learning_rate": 4.406868360705366e-05,
      "loss": 4.6686,
      "num_input_tokens_seen": 12134000,
      "step": 8955
    },
    {
      "epoch": 0.2867934191152935,
      "grad_norm": 44.75,
      "learning_rate": 4.406233320790536e-05,
      "loss": 3.2749,
      "num_input_tokens_seen": 12140736,
      "step": 8960
    },
    {
      "epoch": 0.286953460085782,
      "grad_norm": 37.25,
      "learning_rate": 4.4055979869105734e-05,
      "loss": 4.8961,
      "num_input_tokens_seen": 12147408,
      "step": 8965
    },
    {
      "epoch": 0.2871135010562704,
      "grad_norm": 32.75,
      "learning_rate": 4.404962359163454e-05,
      "loss": 4.2887,
      "num_input_tokens_seen": 12154464,
      "step": 8970
    },
    {
      "epoch": 0.28727354202675887,
      "grad_norm": 22.375,
      "learning_rate": 4.404326437647199e-05,
      "loss": 4.131,
      "num_input_tokens_seen": 12161136,
      "step": 8975
    },
    {
      "epoch": 0.2874335829972473,
      "grad_norm": 24.75,
      "learning_rate": 4.403690222459877e-05,
      "loss": 4.1572,
      "num_input_tokens_seen": 12168000,
      "step": 8980
    },
    {
      "epoch": 0.28759362396773575,
      "grad_norm": 28.25,
      "learning_rate": 4.4030537136995984e-05,
      "loss": 4.0448,
      "num_input_tokens_seen": 12174480,
      "step": 8985
    },
    {
      "epoch": 0.2877536649382242,
      "grad_norm": 35.75,
      "learning_rate": 4.402416911464523e-05,
      "loss": 4.3819,
      "num_input_tokens_seen": 12181296,
      "step": 8990
    },
    {
      "epoch": 0.2879137059087126,
      "grad_norm": 30.875,
      "learning_rate": 4.4017798158528516e-05,
      "loss": 4.3392,
      "num_input_tokens_seen": 12188192,
      "step": 8995
    },
    {
      "epoch": 0.28807374687920106,
      "grad_norm": 34.25,
      "learning_rate": 4.401142426962834e-05,
      "loss": 3.6975,
      "num_input_tokens_seen": 12194800,
      "step": 9000
    },
    {
      "epoch": 0.28807374687920106,
      "eval_loss": 4.499340534210205,
      "eval_runtime": 487.8898,
      "eval_samples_per_second": 28.461,
      "eval_steps_per_second": 14.231,
      "num_input_tokens_seen": 12194800,
      "step": 9000
    },
    {
      "epoch": 0.2882337878496895,
      "grad_norm": 32.0,
      "learning_rate": 4.400504744892763e-05,
      "loss": 4.8765,
      "num_input_tokens_seen": 12201840,
      "step": 9005
    },
    {
      "epoch": 0.28839382882017794,
      "grad_norm": 23.5,
      "learning_rate": 4.399866769740975e-05,
      "loss": 3.998,
      "num_input_tokens_seen": 12208672,
      "step": 9010
    },
    {
      "epoch": 0.2885538697906664,
      "grad_norm": 37.0,
      "learning_rate": 4.399228501605859e-05,
      "loss": 4.6068,
      "num_input_tokens_seen": 12215936,
      "step": 9015
    },
    {
      "epoch": 0.2887139107611549,
      "grad_norm": 26.625,
      "learning_rate": 4.398589940585839e-05,
      "loss": 4.3099,
      "num_input_tokens_seen": 12222640,
      "step": 9020
    },
    {
      "epoch": 0.2888739517316433,
      "grad_norm": 30.5,
      "learning_rate": 4.3979510867793917e-05,
      "loss": 3.8735,
      "num_input_tokens_seen": 12229872,
      "step": 9025
    },
    {
      "epoch": 0.28903399270213176,
      "grad_norm": 42.75,
      "learning_rate": 4.3973119402850346e-05,
      "loss": 5.3768,
      "num_input_tokens_seen": 12236480,
      "step": 9030
    },
    {
      "epoch": 0.2891940336726202,
      "grad_norm": 28.625,
      "learning_rate": 4.396672501201334e-05,
      "loss": 3.0792,
      "num_input_tokens_seen": 12243328,
      "step": 9035
    },
    {
      "epoch": 0.28935407464310864,
      "grad_norm": 19.375,
      "learning_rate": 4.396032769626899e-05,
      "loss": 4.0655,
      "num_input_tokens_seen": 12250128,
      "step": 9040
    },
    {
      "epoch": 0.2895141156135971,
      "grad_norm": 26.875,
      "learning_rate": 4.395392745660384e-05,
      "loss": 4.4526,
      "num_input_tokens_seen": 12256864,
      "step": 9045
    },
    {
      "epoch": 0.2896741565840855,
      "grad_norm": 34.75,
      "learning_rate": 4.394752429400488e-05,
      "loss": 3.6947,
      "num_input_tokens_seen": 12263568,
      "step": 9050
    },
    {
      "epoch": 0.28983419755457396,
      "grad_norm": 20.375,
      "learning_rate": 4.394111820945957e-05,
      "loss": 3.8584,
      "num_input_tokens_seen": 12270624,
      "step": 9055
    },
    {
      "epoch": 0.2899942385250624,
      "grad_norm": 25.25,
      "learning_rate": 4.393470920395579e-05,
      "loss": 4.2094,
      "num_input_tokens_seen": 12277088,
      "step": 9060
    },
    {
      "epoch": 0.29015427949555084,
      "grad_norm": 28.25,
      "learning_rate": 4.392829727848192e-05,
      "loss": 3.7748,
      "num_input_tokens_seen": 12283792,
      "step": 9065
    },
    {
      "epoch": 0.29031432046603933,
      "grad_norm": 25.25,
      "learning_rate": 4.392188243402673e-05,
      "loss": 4.8178,
      "num_input_tokens_seen": 12290768,
      "step": 9070
    },
    {
      "epoch": 0.2904743614365278,
      "grad_norm": 14.25,
      "learning_rate": 4.391546467157949e-05,
      "loss": 3.9302,
      "num_input_tokens_seen": 12297600,
      "step": 9075
    },
    {
      "epoch": 0.2906344024070162,
      "grad_norm": 39.0,
      "learning_rate": 4.390904399212988e-05,
      "loss": 4.7117,
      "num_input_tokens_seen": 12304112,
      "step": 9080
    },
    {
      "epoch": 0.29079444337750465,
      "grad_norm": 28.625,
      "learning_rate": 4.390262039666807e-05,
      "loss": 4.9481,
      "num_input_tokens_seen": 12310960,
      "step": 9085
    },
    {
      "epoch": 0.2909544843479931,
      "grad_norm": 26.125,
      "learning_rate": 4.389619388618464e-05,
      "loss": 4.3926,
      "num_input_tokens_seen": 12317776,
      "step": 9090
    },
    {
      "epoch": 0.29111452531848153,
      "grad_norm": 23.125,
      "learning_rate": 4.3889764461670655e-05,
      "loss": 4.3785,
      "num_input_tokens_seen": 12324288,
      "step": 9095
    },
    {
      "epoch": 0.29127456628896997,
      "grad_norm": 19.625,
      "learning_rate": 4.38833321241176e-05,
      "loss": 5.305,
      "num_input_tokens_seen": 12330848,
      "step": 9100
    },
    {
      "epoch": 0.2914346072594584,
      "grad_norm": 28.5,
      "learning_rate": 4.3876896874517434e-05,
      "loss": 3.7234,
      "num_input_tokens_seen": 12337696,
      "step": 9105
    },
    {
      "epoch": 0.29159464822994685,
      "grad_norm": 22.75,
      "learning_rate": 4.3870458713862554e-05,
      "loss": 3.7743,
      "num_input_tokens_seen": 12344688,
      "step": 9110
    },
    {
      "epoch": 0.2917546892004353,
      "grad_norm": 29.0,
      "learning_rate": 4.386401764314579e-05,
      "loss": 4.8694,
      "num_input_tokens_seen": 12351776,
      "step": 9115
    },
    {
      "epoch": 0.29191473017092373,
      "grad_norm": 38.25,
      "learning_rate": 4.385757366336045e-05,
      "loss": 3.5431,
      "num_input_tokens_seen": 12358384,
      "step": 9120
    },
    {
      "epoch": 0.2920747711414122,
      "grad_norm": 40.75,
      "learning_rate": 4.385112677550027e-05,
      "loss": 4.5977,
      "num_input_tokens_seen": 12365056,
      "step": 9125
    },
    {
      "epoch": 0.29223481211190067,
      "grad_norm": 25.75,
      "learning_rate": 4.384467698055945e-05,
      "loss": 3.9754,
      "num_input_tokens_seen": 12371904,
      "step": 9130
    },
    {
      "epoch": 0.2923948530823891,
      "grad_norm": 29.875,
      "learning_rate": 4.383822427953261e-05,
      "loss": 4.9813,
      "num_input_tokens_seen": 12378752,
      "step": 9135
    },
    {
      "epoch": 0.29255489405287755,
      "grad_norm": 29.75,
      "learning_rate": 4.3831768673414864e-05,
      "loss": 4.4887,
      "num_input_tokens_seen": 12385680,
      "step": 9140
    },
    {
      "epoch": 0.292714935023366,
      "grad_norm": 28.75,
      "learning_rate": 4.382531016320173e-05,
      "loss": 5.1855,
      "num_input_tokens_seen": 12392224,
      "step": 9145
    },
    {
      "epoch": 0.2928749759938544,
      "grad_norm": 29.875,
      "learning_rate": 4.3818848749889184e-05,
      "loss": 5.0176,
      "num_input_tokens_seen": 12399008,
      "step": 9150
    },
    {
      "epoch": 0.29303501696434286,
      "grad_norm": 29.25,
      "learning_rate": 4.381238443447368e-05,
      "loss": 4.5109,
      "num_input_tokens_seen": 12405920,
      "step": 9155
    },
    {
      "epoch": 0.2931950579348313,
      "grad_norm": 31.875,
      "learning_rate": 4.380591721795208e-05,
      "loss": 5.0025,
      "num_input_tokens_seen": 12412800,
      "step": 9160
    },
    {
      "epoch": 0.29335509890531974,
      "grad_norm": 35.25,
      "learning_rate": 4.3799447101321723e-05,
      "loss": 4.0146,
      "num_input_tokens_seen": 12419552,
      "step": 9165
    },
    {
      "epoch": 0.2935151398758082,
      "grad_norm": 17.625,
      "learning_rate": 4.379297408558036e-05,
      "loss": 4.0949,
      "num_input_tokens_seen": 12426304,
      "step": 9170
    },
    {
      "epoch": 0.2936751808462967,
      "grad_norm": 37.75,
      "learning_rate": 4.378649817172624e-05,
      "loss": 4.1193,
      "num_input_tokens_seen": 12432832,
      "step": 9175
    },
    {
      "epoch": 0.2938352218167851,
      "grad_norm": 46.0,
      "learning_rate": 4.378001936075801e-05,
      "loss": 3.8781,
      "num_input_tokens_seen": 12439616,
      "step": 9180
    },
    {
      "epoch": 0.29399526278727356,
      "grad_norm": 30.0,
      "learning_rate": 4.377353765367479e-05,
      "loss": 4.4287,
      "num_input_tokens_seen": 12446048,
      "step": 9185
    },
    {
      "epoch": 0.294155303757762,
      "grad_norm": 21.625,
      "learning_rate": 4.376705305147614e-05,
      "loss": 4.1676,
      "num_input_tokens_seen": 12452496,
      "step": 9190
    },
    {
      "epoch": 0.29431534472825044,
      "grad_norm": 30.75,
      "learning_rate": 4.376056555516206e-05,
      "loss": 3.773,
      "num_input_tokens_seen": 12459296,
      "step": 9195
    },
    {
      "epoch": 0.2944753856987389,
      "grad_norm": 32.25,
      "learning_rate": 4.375407516573302e-05,
      "loss": 3.8447,
      "num_input_tokens_seen": 12466288,
      "step": 9200
    },
    {
      "epoch": 0.2944753856987389,
      "eval_loss": 4.479452610015869,
      "eval_runtime": 487.3604,
      "eval_samples_per_second": 28.492,
      "eval_steps_per_second": 14.246,
      "num_input_tokens_seen": 12466288,
      "step": 9200
    },
    {
      "epoch": 0.2946354266692273,
      "grad_norm": 18.75,
      "learning_rate": 4.3747581884189913e-05,
      "loss": 4.584,
      "num_input_tokens_seen": 12472896,
      "step": 9205
    },
    {
      "epoch": 0.29479546763971576,
      "grad_norm": 46.25,
      "learning_rate": 4.374108571153408e-05,
      "loss": 3.897,
      "num_input_tokens_seen": 12479584,
      "step": 9210
    },
    {
      "epoch": 0.2949555086102042,
      "grad_norm": 21.625,
      "learning_rate": 4.3734586648767316e-05,
      "loss": 4.8855,
      "num_input_tokens_seen": 12486304,
      "step": 9215
    },
    {
      "epoch": 0.29511554958069264,
      "grad_norm": 59.25,
      "learning_rate": 4.372808469689186e-05,
      "loss": 5.3024,
      "num_input_tokens_seen": 12493136,
      "step": 9220
    },
    {
      "epoch": 0.2952755905511811,
      "grad_norm": 31.375,
      "learning_rate": 4.372157985691039e-05,
      "loss": 4.3215,
      "num_input_tokens_seen": 12499808,
      "step": 9225
    },
    {
      "epoch": 0.2954356315216696,
      "grad_norm": 55.75,
      "learning_rate": 4.371507212982603e-05,
      "loss": 4.7319,
      "num_input_tokens_seen": 12506432,
      "step": 9230
    },
    {
      "epoch": 0.295595672492158,
      "grad_norm": 37.0,
      "learning_rate": 4.370856151664236e-05,
      "loss": 3.836,
      "num_input_tokens_seen": 12512848,
      "step": 9235
    },
    {
      "epoch": 0.29575571346264645,
      "grad_norm": 25.0,
      "learning_rate": 4.3702048018363404e-05,
      "loss": 4.3063,
      "num_input_tokens_seen": 12519856,
      "step": 9240
    },
    {
      "epoch": 0.2959157544331349,
      "grad_norm": 22.75,
      "learning_rate": 4.369553163599362e-05,
      "loss": 4.9518,
      "num_input_tokens_seen": 12526688,
      "step": 9245
    },
    {
      "epoch": 0.29607579540362333,
      "grad_norm": 39.0,
      "learning_rate": 4.3689012370537904e-05,
      "loss": 4.3652,
      "num_input_tokens_seen": 12533696,
      "step": 9250
    },
    {
      "epoch": 0.29623583637411177,
      "grad_norm": 30.0,
      "learning_rate": 4.368249022300164e-05,
      "loss": 4.3569,
      "num_input_tokens_seen": 12539952,
      "step": 9255
    },
    {
      "epoch": 0.2963958773446002,
      "grad_norm": 25.25,
      "learning_rate": 4.367596519439059e-05,
      "loss": 4.3954,
      "num_input_tokens_seen": 12546976,
      "step": 9260
    },
    {
      "epoch": 0.29655591831508865,
      "grad_norm": 15.8125,
      "learning_rate": 4.366943728571101e-05,
      "loss": 3.9751,
      "num_input_tokens_seen": 12553264,
      "step": 9265
    },
    {
      "epoch": 0.2967159592855771,
      "grad_norm": 23.0,
      "learning_rate": 4.366290649796959e-05,
      "loss": 4.0959,
      "num_input_tokens_seen": 12560208,
      "step": 9270
    },
    {
      "epoch": 0.29687600025606553,
      "grad_norm": 16.75,
      "learning_rate": 4.3656372832173456e-05,
      "loss": 4.1917,
      "num_input_tokens_seen": 12566832,
      "step": 9275
    },
    {
      "epoch": 0.29703604122655397,
      "grad_norm": 25.625,
      "learning_rate": 4.364983628933017e-05,
      "loss": 4.6967,
      "num_input_tokens_seen": 12573264,
      "step": 9280
    },
    {
      "epoch": 0.29719608219704247,
      "grad_norm": 70.0,
      "learning_rate": 4.364329687044777e-05,
      "loss": 4.7274,
      "num_input_tokens_seen": 12580224,
      "step": 9285
    },
    {
      "epoch": 0.2973561231675309,
      "grad_norm": 20.375,
      "learning_rate": 4.36367545765347e-05,
      "loss": 4.5886,
      "num_input_tokens_seen": 12586560,
      "step": 9290
    },
    {
      "epoch": 0.29751616413801935,
      "grad_norm": 25.5,
      "learning_rate": 4.363020940859988e-05,
      "loss": 4.7777,
      "num_input_tokens_seen": 12592944,
      "step": 9295
    },
    {
      "epoch": 0.2976762051085078,
      "grad_norm": 45.0,
      "learning_rate": 4.362366136765263e-05,
      "loss": 4.8877,
      "num_input_tokens_seen": 12599776,
      "step": 9300
    },
    {
      "epoch": 0.2978362460789962,
      "grad_norm": 13.5,
      "learning_rate": 4.361711045470278e-05,
      "loss": 4.3631,
      "num_input_tokens_seen": 12606640,
      "step": 9305
    },
    {
      "epoch": 0.29799628704948466,
      "grad_norm": 19.0,
      "learning_rate": 4.3610556670760524e-05,
      "loss": 3.4033,
      "num_input_tokens_seen": 12613136,
      "step": 9310
    },
    {
      "epoch": 0.2981563280199731,
      "grad_norm": 22.0,
      "learning_rate": 4.360400001683657e-05,
      "loss": 4.3421,
      "num_input_tokens_seen": 12620240,
      "step": 9315
    },
    {
      "epoch": 0.29831636899046154,
      "grad_norm": 26.875,
      "learning_rate": 4.3597440493942e-05,
      "loss": 4.6424,
      "num_input_tokens_seen": 12626768,
      "step": 9320
    },
    {
      "epoch": 0.29847640996095,
      "grad_norm": 20.625,
      "learning_rate": 4.3590878103088405e-05,
      "loss": 4.2299,
      "num_input_tokens_seen": 12633360,
      "step": 9325
    },
    {
      "epoch": 0.2986364509314384,
      "grad_norm": 32.25,
      "learning_rate": 4.358431284528779e-05,
      "loss": 4.195,
      "num_input_tokens_seen": 12640288,
      "step": 9330
    },
    {
      "epoch": 0.2987964919019269,
      "grad_norm": 26.125,
      "learning_rate": 4.357774472155257e-05,
      "loss": 3.9228,
      "num_input_tokens_seen": 12647232,
      "step": 9335
    },
    {
      "epoch": 0.29895653287241536,
      "grad_norm": 26.625,
      "learning_rate": 4.3571173732895664e-05,
      "loss": 4.5167,
      "num_input_tokens_seen": 12653760,
      "step": 9340
    },
    {
      "epoch": 0.2991165738429038,
      "grad_norm": 23.875,
      "learning_rate": 4.356459988033039e-05,
      "loss": 4.9829,
      "num_input_tokens_seen": 12660592,
      "step": 9345
    },
    {
      "epoch": 0.29927661481339224,
      "grad_norm": 31.75,
      "learning_rate": 4.355802316487051e-05,
      "loss": 4.0862,
      "num_input_tokens_seen": 12667024,
      "step": 9350
    },
    {
      "epoch": 0.2994366557838807,
      "grad_norm": 36.25,
      "learning_rate": 4.355144358753025e-05,
      "loss": 4.127,
      "num_input_tokens_seen": 12673584,
      "step": 9355
    },
    {
      "epoch": 0.2995966967543691,
      "grad_norm": 42.75,
      "learning_rate": 4.354486114932425e-05,
      "loss": 4.002,
      "num_input_tokens_seen": 12680096,
      "step": 9360
    },
    {
      "epoch": 0.29975673772485756,
      "grad_norm": 24.0,
      "learning_rate": 4.353827585126762e-05,
      "loss": 4.3846,
      "num_input_tokens_seen": 12686720,
      "step": 9365
    },
    {
      "epoch": 0.299916778695346,
      "grad_norm": 19.5,
      "learning_rate": 4.353168769437588e-05,
      "loss": 4.0999,
      "num_input_tokens_seen": 12693840,
      "step": 9370
    },
    {
      "epoch": 0.30007681966583444,
      "grad_norm": 37.25,
      "learning_rate": 4.3525096679665014e-05,
      "loss": 4.6476,
      "num_input_tokens_seen": 12700640,
      "step": 9375
    },
    {
      "epoch": 0.3002368606363229,
      "grad_norm": 32.25,
      "learning_rate": 4.351850280815144e-05,
      "loss": 4.4209,
      "num_input_tokens_seen": 12707888,
      "step": 9380
    },
    {
      "epoch": 0.3003969016068113,
      "grad_norm": 38.75,
      "learning_rate": 4.3511906080852014e-05,
      "loss": 4.0941,
      "num_input_tokens_seen": 12714992,
      "step": 9385
    },
    {
      "epoch": 0.3005569425772998,
      "grad_norm": 23.75,
      "learning_rate": 4.350530649878404e-05,
      "loss": 3.8614,
      "num_input_tokens_seen": 12721776,
      "step": 9390
    },
    {
      "epoch": 0.30071698354778825,
      "grad_norm": 12.8125,
      "learning_rate": 4.3498704062965246e-05,
      "loss": 3.3043,
      "num_input_tokens_seen": 12728608,
      "step": 9395
    },
    {
      "epoch": 0.3008770245182767,
      "grad_norm": 26.5,
      "learning_rate": 4.3492098774413815e-05,
      "loss": 4.2442,
      "num_input_tokens_seen": 12735104,
      "step": 9400
    },
    {
      "epoch": 0.3008770245182767,
      "eval_loss": 4.4929890632629395,
      "eval_runtime": 487.223,
      "eval_samples_per_second": 28.5,
      "eval_steps_per_second": 14.25,
      "num_input_tokens_seen": 12735104,
      "step": 9400
    },
    {
      "epoch": 0.30103706548876513,
      "grad_norm": 18.875,
      "learning_rate": 4.3485490634148375e-05,
      "loss": 3.8633,
      "num_input_tokens_seen": 12741664,
      "step": 9405
    },
    {
      "epoch": 0.30119710645925357,
      "grad_norm": 31.125,
      "learning_rate": 4.347887964318797e-05,
      "loss": 4.1934,
      "num_input_tokens_seen": 12748288,
      "step": 9410
    },
    {
      "epoch": 0.301357147429742,
      "grad_norm": 41.5,
      "learning_rate": 4.34722658025521e-05,
      "loss": 4.721,
      "num_input_tokens_seen": 12755344,
      "step": 9415
    },
    {
      "epoch": 0.30151718840023045,
      "grad_norm": 18.875,
      "learning_rate": 4.346564911326071e-05,
      "loss": 4.1056,
      "num_input_tokens_seen": 12762096,
      "step": 9420
    },
    {
      "epoch": 0.3016772293707189,
      "grad_norm": 18.25,
      "learning_rate": 4.345902957633418e-05,
      "loss": 3.9865,
      "num_input_tokens_seen": 12768928,
      "step": 9425
    },
    {
      "epoch": 0.30183727034120733,
      "grad_norm": 25.25,
      "learning_rate": 4.345240719279331e-05,
      "loss": 4.1249,
      "num_input_tokens_seen": 12775856,
      "step": 9430
    },
    {
      "epoch": 0.30199731131169577,
      "grad_norm": 36.75,
      "learning_rate": 4.3445781963659374e-05,
      "loss": 4.2265,
      "num_input_tokens_seen": 12782160,
      "step": 9435
    },
    {
      "epoch": 0.30215735228218427,
      "grad_norm": 26.375,
      "learning_rate": 4.3439153889954045e-05,
      "loss": 4.3416,
      "num_input_tokens_seen": 12788992,
      "step": 9440
    },
    {
      "epoch": 0.3023173932526727,
      "grad_norm": 28.75,
      "learning_rate": 4.343252297269946e-05,
      "loss": 3.147,
      "num_input_tokens_seen": 12795808,
      "step": 9445
    },
    {
      "epoch": 0.30247743422316115,
      "grad_norm": 32.5,
      "learning_rate": 4.342588921291821e-05,
      "loss": 4.9584,
      "num_input_tokens_seen": 12802640,
      "step": 9450
    },
    {
      "epoch": 0.3026374751936496,
      "grad_norm": 30.625,
      "learning_rate": 4.341925261163328e-05,
      "loss": 4.9101,
      "num_input_tokens_seen": 12809552,
      "step": 9455
    },
    {
      "epoch": 0.302797516164138,
      "grad_norm": 35.25,
      "learning_rate": 4.341261316986813e-05,
      "loss": 3.9812,
      "num_input_tokens_seen": 12816048,
      "step": 9460
    },
    {
      "epoch": 0.30295755713462647,
      "grad_norm": 19.375,
      "learning_rate": 4.340597088864664e-05,
      "loss": 4.2581,
      "num_input_tokens_seen": 12823024,
      "step": 9465
    },
    {
      "epoch": 0.3031175981051149,
      "grad_norm": 36.0,
      "learning_rate": 4.339932576899313e-05,
      "loss": 4.2281,
      "num_input_tokens_seen": 12829904,
      "step": 9470
    },
    {
      "epoch": 0.30327763907560334,
      "grad_norm": 10.9375,
      "learning_rate": 4.3392677811932375e-05,
      "loss": 3.4487,
      "num_input_tokens_seen": 12836592,
      "step": 9475
    },
    {
      "epoch": 0.3034376800460918,
      "grad_norm": 58.5,
      "learning_rate": 4.338602701848956e-05,
      "loss": 4.2542,
      "num_input_tokens_seen": 12843280,
      "step": 9480
    },
    {
      "epoch": 0.3035977210165802,
      "grad_norm": 70.0,
      "learning_rate": 4.337937338969033e-05,
      "loss": 5.0011,
      "num_input_tokens_seen": 12849840,
      "step": 9485
    },
    {
      "epoch": 0.30375776198706866,
      "grad_norm": 41.75,
      "learning_rate": 4.337271692656075e-05,
      "loss": 4.3141,
      "num_input_tokens_seen": 12856896,
      "step": 9490
    },
    {
      "epoch": 0.30391780295755716,
      "grad_norm": 35.0,
      "learning_rate": 4.336605763012733e-05,
      "loss": 4.1219,
      "num_input_tokens_seen": 12864176,
      "step": 9495
    },
    {
      "epoch": 0.3040778439280456,
      "grad_norm": 21.125,
      "learning_rate": 4.3359395501417026e-05,
      "loss": 3.9462,
      "num_input_tokens_seen": 12870704,
      "step": 9500
    },
    {
      "epoch": 0.30423788489853404,
      "grad_norm": 32.75,
      "learning_rate": 4.335273054145722e-05,
      "loss": 5.3049,
      "num_input_tokens_seen": 12877088,
      "step": 9505
    },
    {
      "epoch": 0.3043979258690225,
      "grad_norm": 39.75,
      "learning_rate": 4.334606275127572e-05,
      "loss": 4.617,
      "num_input_tokens_seen": 12883440,
      "step": 9510
    },
    {
      "epoch": 0.3045579668395109,
      "grad_norm": 31.625,
      "learning_rate": 4.33393921319008e-05,
      "loss": 3.4868,
      "num_input_tokens_seen": 12889776,
      "step": 9515
    },
    {
      "epoch": 0.30471800780999936,
      "grad_norm": 20.75,
      "learning_rate": 4.3332718684361146e-05,
      "loss": 4.4098,
      "num_input_tokens_seen": 12896448,
      "step": 9520
    },
    {
      "epoch": 0.3048780487804878,
      "grad_norm": 29.125,
      "learning_rate": 4.332604240968588e-05,
      "loss": 4.8928,
      "num_input_tokens_seen": 12903280,
      "step": 9525
    },
    {
      "epoch": 0.30503808975097624,
      "grad_norm": 21.875,
      "learning_rate": 4.331936330890459e-05,
      "loss": 3.9915,
      "num_input_tokens_seen": 12910208,
      "step": 9530
    },
    {
      "epoch": 0.3051981307214647,
      "grad_norm": 23.5,
      "learning_rate": 4.331268138304725e-05,
      "loss": 4.4186,
      "num_input_tokens_seen": 12916384,
      "step": 9535
    },
    {
      "epoch": 0.3053581716919531,
      "grad_norm": 39.75,
      "learning_rate": 4.330599663314431e-05,
      "loss": 4.0397,
      "num_input_tokens_seen": 12922992,
      "step": 9540
    },
    {
      "epoch": 0.3055182126624416,
      "grad_norm": 50.25,
      "learning_rate": 4.329930906022665e-05,
      "loss": 4.5613,
      "num_input_tokens_seen": 12929664,
      "step": 9545
    },
    {
      "epoch": 0.30567825363293005,
      "grad_norm": 34.5,
      "learning_rate": 4.3292618665325564e-05,
      "loss": 3.3044,
      "num_input_tokens_seen": 12936368,
      "step": 9550
    },
    {
      "epoch": 0.3058382946034185,
      "grad_norm": 52.25,
      "learning_rate": 4.3285925449472796e-05,
      "loss": 4.4146,
      "num_input_tokens_seen": 12943072,
      "step": 9555
    },
    {
      "epoch": 0.30599833557390693,
      "grad_norm": 19.0,
      "learning_rate": 4.327922941370054e-05,
      "loss": 4.5119,
      "num_input_tokens_seen": 12950016,
      "step": 9560
    },
    {
      "epoch": 0.30615837654439537,
      "grad_norm": 24.0,
      "learning_rate": 4.3272530559041384e-05,
      "loss": 4.5064,
      "num_input_tokens_seen": 12956976,
      "step": 9565
    },
    {
      "epoch": 0.3063184175148838,
      "grad_norm": 18.25,
      "learning_rate": 4.32658288865284e-05,
      "loss": 3.3835,
      "num_input_tokens_seen": 12963648,
      "step": 9570
    },
    {
      "epoch": 0.30647845848537225,
      "grad_norm": 24.25,
      "learning_rate": 4.325912439719505e-05,
      "loss": 4.3505,
      "num_input_tokens_seen": 12969936,
      "step": 9575
    },
    {
      "epoch": 0.3066384994558607,
      "grad_norm": 37.0,
      "learning_rate": 4.3252417092075266e-05,
      "loss": 5.1552,
      "num_input_tokens_seen": 12976208,
      "step": 9580
    },
    {
      "epoch": 0.30679854042634913,
      "grad_norm": 33.25,
      "learning_rate": 4.3245706972203385e-05,
      "loss": 4.2765,
      "num_input_tokens_seen": 12982896,
      "step": 9585
    },
    {
      "epoch": 0.30695858139683757,
      "grad_norm": 37.0,
      "learning_rate": 4.323899403861421e-05,
      "loss": 4.3447,
      "num_input_tokens_seen": 12989520,
      "step": 9590
    },
    {
      "epoch": 0.307118622367326,
      "grad_norm": 26.125,
      "learning_rate": 4.3232278292342935e-05,
      "loss": 4.3441,
      "num_input_tokens_seen": 12996544,
      "step": 9595
    },
    {
      "epoch": 0.3072786633378145,
      "grad_norm": 35.0,
      "learning_rate": 4.322555973442524e-05,
      "loss": 4.1695,
      "num_input_tokens_seen": 13003216,
      "step": 9600
    },
    {
      "epoch": 0.3072786633378145,
      "eval_loss": 4.491155624389648,
      "eval_runtime": 488.4075,
      "eval_samples_per_second": 28.431,
      "eval_steps_per_second": 14.216,
      "num_input_tokens_seen": 13003216,
      "step": 9600
    },
    {
      "epoch": 0.30743870430830295,
      "grad_norm": 30.25,
      "learning_rate": 4.3218838365897184e-05,
      "loss": 3.7155,
      "num_input_tokens_seen": 13009808,
      "step": 9605
    },
    {
      "epoch": 0.3075987452787914,
      "grad_norm": 27.875,
      "learning_rate": 4.3212114187795306e-05,
      "loss": 4.2699,
      "num_input_tokens_seen": 13016816,
      "step": 9610
    },
    {
      "epoch": 0.3077587862492798,
      "grad_norm": 21.0,
      "learning_rate": 4.320538720115656e-05,
      "loss": 3.7762,
      "num_input_tokens_seen": 13023472,
      "step": 9615
    },
    {
      "epoch": 0.30791882721976827,
      "grad_norm": 25.875,
      "learning_rate": 4.319865740701831e-05,
      "loss": 5.4013,
      "num_input_tokens_seen": 13030192,
      "step": 9620
    },
    {
      "epoch": 0.3080788681902567,
      "grad_norm": 23.625,
      "learning_rate": 4.3191924806418396e-05,
      "loss": 4.206,
      "num_input_tokens_seen": 13036896,
      "step": 9625
    },
    {
      "epoch": 0.30823890916074514,
      "grad_norm": 24.25,
      "learning_rate": 4.318518940039507e-05,
      "loss": 4.7429,
      "num_input_tokens_seen": 13043280,
      "step": 9630
    },
    {
      "epoch": 0.3083989501312336,
      "grad_norm": 46.75,
      "learning_rate": 4.3178451189987e-05,
      "loss": 4.2431,
      "num_input_tokens_seen": 13050336,
      "step": 9635
    },
    {
      "epoch": 0.308558991101722,
      "grad_norm": 21.125,
      "learning_rate": 4.3171710176233315e-05,
      "loss": 3.5954,
      "num_input_tokens_seen": 13057296,
      "step": 9640
    },
    {
      "epoch": 0.30871903207221046,
      "grad_norm": 19.75,
      "learning_rate": 4.316496636017355e-05,
      "loss": 3.6542,
      "num_input_tokens_seen": 13063984,
      "step": 9645
    },
    {
      "epoch": 0.3088790730426989,
      "grad_norm": 39.25,
      "learning_rate": 4.315821974284771e-05,
      "loss": 5.0347,
      "num_input_tokens_seen": 13070896,
      "step": 9650
    },
    {
      "epoch": 0.3090391140131874,
      "grad_norm": 33.5,
      "learning_rate": 4.315147032529619e-05,
      "loss": 4.365,
      "num_input_tokens_seen": 13077856,
      "step": 9655
    },
    {
      "epoch": 0.30919915498367584,
      "grad_norm": 36.25,
      "learning_rate": 4.3144718108559845e-05,
      "loss": 3.9771,
      "num_input_tokens_seen": 13084800,
      "step": 9660
    },
    {
      "epoch": 0.3093591959541643,
      "grad_norm": 25.375,
      "learning_rate": 4.3137963093679945e-05,
      "loss": 4.3274,
      "num_input_tokens_seen": 13092048,
      "step": 9665
    },
    {
      "epoch": 0.3095192369246527,
      "grad_norm": 31.75,
      "learning_rate": 4.31312052816982e-05,
      "loss": 4.162,
      "num_input_tokens_seen": 13098960,
      "step": 9670
    },
    {
      "epoch": 0.30967927789514116,
      "grad_norm": 39.0,
      "learning_rate": 4.312444467365675e-05,
      "loss": 4.4768,
      "num_input_tokens_seen": 13105968,
      "step": 9675
    },
    {
      "epoch": 0.3098393188656296,
      "grad_norm": 33.75,
      "learning_rate": 4.311768127059816e-05,
      "loss": 4.6342,
      "num_input_tokens_seen": 13112496,
      "step": 9680
    },
    {
      "epoch": 0.30999935983611804,
      "grad_norm": 20.5,
      "learning_rate": 4.3110915073565444e-05,
      "loss": 4.0627,
      "num_input_tokens_seen": 13119056,
      "step": 9685
    },
    {
      "epoch": 0.3101594008066065,
      "grad_norm": 20.5,
      "learning_rate": 4.310414608360203e-05,
      "loss": 4.9956,
      "num_input_tokens_seen": 13125792,
      "step": 9690
    },
    {
      "epoch": 0.3103194417770949,
      "grad_norm": 18.125,
      "learning_rate": 4.309737430175177e-05,
      "loss": 3.7278,
      "num_input_tokens_seen": 13132560,
      "step": 9695
    },
    {
      "epoch": 0.31047948274758336,
      "grad_norm": 46.0,
      "learning_rate": 4.309059972905897e-05,
      "loss": 4.3792,
      "num_input_tokens_seen": 13139248,
      "step": 9700
    },
    {
      "epoch": 0.31063952371807185,
      "grad_norm": 21.625,
      "learning_rate": 4.308382236656836e-05,
      "loss": 4.3372,
      "num_input_tokens_seen": 13146080,
      "step": 9705
    },
    {
      "epoch": 0.3107995646885603,
      "grad_norm": 25.125,
      "learning_rate": 4.307704221532507e-05,
      "loss": 4.6256,
      "num_input_tokens_seen": 13152640,
      "step": 9710
    },
    {
      "epoch": 0.31095960565904873,
      "grad_norm": 43.25,
      "learning_rate": 4.307025927637471e-05,
      "loss": 3.9572,
      "num_input_tokens_seen": 13159552,
      "step": 9715
    },
    {
      "epoch": 0.31111964662953717,
      "grad_norm": 15.375,
      "learning_rate": 4.306347355076328e-05,
      "loss": 4.1025,
      "num_input_tokens_seen": 13166128,
      "step": 9720
    },
    {
      "epoch": 0.3112796876000256,
      "grad_norm": 30.25,
      "learning_rate": 4.305668503953724e-05,
      "loss": 4.633,
      "num_input_tokens_seen": 13172768,
      "step": 9725
    },
    {
      "epoch": 0.31143972857051405,
      "grad_norm": 19.375,
      "learning_rate": 4.3049893743743436e-05,
      "loss": 4.6704,
      "num_input_tokens_seen": 13179888,
      "step": 9730
    },
    {
      "epoch": 0.3115997695410025,
      "grad_norm": 50.25,
      "learning_rate": 4.304309966442919e-05,
      "loss": 4.9508,
      "num_input_tokens_seen": 13186672,
      "step": 9735
    },
    {
      "epoch": 0.31175981051149093,
      "grad_norm": 35.75,
      "learning_rate": 4.303630280264224e-05,
      "loss": 4.4017,
      "num_input_tokens_seen": 13193568,
      "step": 9740
    },
    {
      "epoch": 0.31191985148197937,
      "grad_norm": 25.75,
      "learning_rate": 4.302950315943074e-05,
      "loss": 4.8117,
      "num_input_tokens_seen": 13200256,
      "step": 9745
    },
    {
      "epoch": 0.3120798924524678,
      "grad_norm": 30.625,
      "learning_rate": 4.3022700735843275e-05,
      "loss": 3.8775,
      "num_input_tokens_seen": 13206672,
      "step": 9750
    },
    {
      "epoch": 0.31223993342295625,
      "grad_norm": 21.5,
      "learning_rate": 4.301589553292887e-05,
      "loss": 4.2044,
      "num_input_tokens_seen": 13213168,
      "step": 9755
    },
    {
      "epoch": 0.31239997439344475,
      "grad_norm": 26.75,
      "learning_rate": 4.300908755173697e-05,
      "loss": 3.9018,
      "num_input_tokens_seen": 13220032,
      "step": 9760
    },
    {
      "epoch": 0.3125600153639332,
      "grad_norm": 26.875,
      "learning_rate": 4.300227679331745e-05,
      "loss": 4.8816,
      "num_input_tokens_seen": 13226896,
      "step": 9765
    },
    {
      "epoch": 0.3127200563344216,
      "grad_norm": 27.875,
      "learning_rate": 4.299546325872063e-05,
      "loss": 4.044,
      "num_input_tokens_seen": 13233568,
      "step": 9770
    },
    {
      "epoch": 0.31288009730491007,
      "grad_norm": 20.625,
      "learning_rate": 4.2988646948997225e-05,
      "loss": 3.8415,
      "num_input_tokens_seen": 13240496,
      "step": 9775
    },
    {
      "epoch": 0.3130401382753985,
      "grad_norm": 42.5,
      "learning_rate": 4.29818278651984e-05,
      "loss": 4.3309,
      "num_input_tokens_seen": 13247120,
      "step": 9780
    },
    {
      "epoch": 0.31320017924588694,
      "grad_norm": 30.125,
      "learning_rate": 4.297500600837574e-05,
      "loss": 4.4953,
      "num_input_tokens_seen": 13253952,
      "step": 9785
    },
    {
      "epoch": 0.3133602202163754,
      "grad_norm": 61.25,
      "learning_rate": 4.2968181379581276e-05,
      "loss": 4.8077,
      "num_input_tokens_seen": 13260736,
      "step": 9790
    },
    {
      "epoch": 0.3135202611868638,
      "grad_norm": 22.125,
      "learning_rate": 4.296135397986743e-05,
      "loss": 4.0349,
      "num_input_tokens_seen": 13267120,
      "step": 9795
    },
    {
      "epoch": 0.31368030215735226,
      "grad_norm": 29.875,
      "learning_rate": 4.295452381028709e-05,
      "loss": 4.1089,
      "num_input_tokens_seen": 13273680,
      "step": 9800
    },
    {
      "epoch": 0.31368030215735226,
      "eval_loss": 4.4999213218688965,
      "eval_runtime": 488.249,
      "eval_samples_per_second": 28.44,
      "eval_steps_per_second": 14.22,
      "num_input_tokens_seen": 13273680,
      "step": 9800
    },
    {
      "epoch": 0.3138403431278407,
      "grad_norm": 21.625,
      "learning_rate": 4.294769087189354e-05,
      "loss": 3.2126,
      "num_input_tokens_seen": 13280800,
      "step": 9805
    },
    {
      "epoch": 0.3140003840983292,
      "grad_norm": 23.25,
      "learning_rate": 4.294085516574052e-05,
      "loss": 3.6842,
      "num_input_tokens_seen": 13287824,
      "step": 9810
    },
    {
      "epoch": 0.31416042506881764,
      "grad_norm": 19.875,
      "learning_rate": 4.2934016692882176e-05,
      "loss": 3.7559,
      "num_input_tokens_seen": 13294480,
      "step": 9815
    },
    {
      "epoch": 0.3143204660393061,
      "grad_norm": 25.5,
      "learning_rate": 4.292717545437308e-05,
      "loss": 3.923,
      "num_input_tokens_seen": 13301024,
      "step": 9820
    },
    {
      "epoch": 0.3144805070097945,
      "grad_norm": 32.25,
      "learning_rate": 4.292033145126825e-05,
      "loss": 4.1123,
      "num_input_tokens_seen": 13307424,
      "step": 9825
    },
    {
      "epoch": 0.31464054798028296,
      "grad_norm": 26.75,
      "learning_rate": 4.29134846846231e-05,
      "loss": 4.2192,
      "num_input_tokens_seen": 13313872,
      "step": 9830
    },
    {
      "epoch": 0.3148005889507714,
      "grad_norm": 20.5,
      "learning_rate": 4.29066351554935e-05,
      "loss": 3.2523,
      "num_input_tokens_seen": 13320448,
      "step": 9835
    },
    {
      "epoch": 0.31496062992125984,
      "grad_norm": 47.0,
      "learning_rate": 4.289978286493574e-05,
      "loss": 4.6424,
      "num_input_tokens_seen": 13327040,
      "step": 9840
    },
    {
      "epoch": 0.3151206708917483,
      "grad_norm": 54.0,
      "learning_rate": 4.28929278140065e-05,
      "loss": 4.8366,
      "num_input_tokens_seen": 13333808,
      "step": 9845
    },
    {
      "epoch": 0.3152807118622367,
      "grad_norm": 27.0,
      "learning_rate": 4.288607000376295e-05,
      "loss": 4.8062,
      "num_input_tokens_seen": 13340224,
      "step": 9850
    },
    {
      "epoch": 0.31544075283272516,
      "grad_norm": 33.5,
      "learning_rate": 4.2879209435262624e-05,
      "loss": 4.7117,
      "num_input_tokens_seen": 13347552,
      "step": 9855
    },
    {
      "epoch": 0.3156007938032136,
      "grad_norm": 18.0,
      "learning_rate": 4.287234610956353e-05,
      "loss": 3.8075,
      "num_input_tokens_seen": 13354560,
      "step": 9860
    },
    {
      "epoch": 0.3157608347737021,
      "grad_norm": 31.5,
      "learning_rate": 4.2865480027724056e-05,
      "loss": 3.9405,
      "num_input_tokens_seen": 13361296,
      "step": 9865
    },
    {
      "epoch": 0.31592087574419053,
      "grad_norm": 31.75,
      "learning_rate": 4.285861119080306e-05,
      "loss": 3.9829,
      "num_input_tokens_seen": 13368288,
      "step": 9870
    },
    {
      "epoch": 0.31608091671467897,
      "grad_norm": 34.5,
      "learning_rate": 4.2851739599859784e-05,
      "loss": 4.2664,
      "num_input_tokens_seen": 13374960,
      "step": 9875
    },
    {
      "epoch": 0.3162409576851674,
      "grad_norm": 24.5,
      "learning_rate": 4.2844865255953934e-05,
      "loss": 3.6184,
      "num_input_tokens_seen": 13381888,
      "step": 9880
    },
    {
      "epoch": 0.31640099865565585,
      "grad_norm": 21.125,
      "learning_rate": 4.2837988160145605e-05,
      "loss": 4.9709,
      "num_input_tokens_seen": 13388688,
      "step": 9885
    },
    {
      "epoch": 0.3165610396261443,
      "grad_norm": 28.75,
      "learning_rate": 4.2831108313495336e-05,
      "loss": 4.137,
      "num_input_tokens_seen": 13395888,
      "step": 9890
    },
    {
      "epoch": 0.31672108059663273,
      "grad_norm": 35.75,
      "learning_rate": 4.282422571706408e-05,
      "loss": 3.5552,
      "num_input_tokens_seen": 13403296,
      "step": 9895
    },
    {
      "epoch": 0.31688112156712117,
      "grad_norm": 33.0,
      "learning_rate": 4.281734037191323e-05,
      "loss": 4.3907,
      "num_input_tokens_seen": 13409696,
      "step": 9900
    },
    {
      "epoch": 0.3170411625376096,
      "grad_norm": 24.75,
      "learning_rate": 4.281045227910459e-05,
      "loss": 3.696,
      "num_input_tokens_seen": 13416416,
      "step": 9905
    },
    {
      "epoch": 0.31720120350809805,
      "grad_norm": 28.625,
      "learning_rate": 4.280356143970038e-05,
      "loss": 3.8076,
      "num_input_tokens_seen": 13423152,
      "step": 9910
    },
    {
      "epoch": 0.3173612444785865,
      "grad_norm": 37.75,
      "learning_rate": 4.279666785476327e-05,
      "loss": 4.0574,
      "num_input_tokens_seen": 13430368,
      "step": 9915
    },
    {
      "epoch": 0.317521285449075,
      "grad_norm": 26.25,
      "learning_rate": 4.2789771525356325e-05,
      "loss": 4.6758,
      "num_input_tokens_seen": 13437840,
      "step": 9920
    },
    {
      "epoch": 0.3176813264195634,
      "grad_norm": 29.5,
      "learning_rate": 4.2782872452543056e-05,
      "loss": 4.2802,
      "num_input_tokens_seen": 13444528,
      "step": 9925
    },
    {
      "epoch": 0.31784136739005187,
      "grad_norm": 44.0,
      "learning_rate": 4.2775970637387376e-05,
      "loss": 4.1418,
      "num_input_tokens_seen": 13451376,
      "step": 9930
    },
    {
      "epoch": 0.3180014083605403,
      "grad_norm": 22.875,
      "learning_rate": 4.276906608095363e-05,
      "loss": 3.6144,
      "num_input_tokens_seen": 13458304,
      "step": 9935
    },
    {
      "epoch": 0.31816144933102875,
      "grad_norm": 46.25,
      "learning_rate": 4.276215878430661e-05,
      "loss": 5.524,
      "num_input_tokens_seen": 13464752,
      "step": 9940
    },
    {
      "epoch": 0.3183214903015172,
      "grad_norm": 40.25,
      "learning_rate": 4.275524874851149e-05,
      "loss": 4.5115,
      "num_input_tokens_seen": 13471216,
      "step": 9945
    },
    {
      "epoch": 0.3184815312720056,
      "grad_norm": 25.25,
      "learning_rate": 4.274833597463388e-05,
      "loss": 5.1137,
      "num_input_tokens_seen": 13478080,
      "step": 9950
    },
    {
      "epoch": 0.31864157224249406,
      "grad_norm": 32.25,
      "learning_rate": 4.2741420463739824e-05,
      "loss": 4.3339,
      "num_input_tokens_seen": 13485040,
      "step": 9955
    },
    {
      "epoch": 0.3188016132129825,
      "grad_norm": 47.25,
      "learning_rate": 4.273450221689578e-05,
      "loss": 4.7164,
      "num_input_tokens_seen": 13491808,
      "step": 9960
    },
    {
      "epoch": 0.31896165418347094,
      "grad_norm": 43.75,
      "learning_rate": 4.272758123516863e-05,
      "loss": 4.2931,
      "num_input_tokens_seen": 13498592,
      "step": 9965
    },
    {
      "epoch": 0.31912169515395944,
      "grad_norm": 39.0,
      "learning_rate": 4.272065751962567e-05,
      "loss": 3.8432,
      "num_input_tokens_seen": 13505632,
      "step": 9970
    },
    {
      "epoch": 0.3192817361244479,
      "grad_norm": 49.0,
      "learning_rate": 4.271373107133464e-05,
      "loss": 4.8864,
      "num_input_tokens_seen": 13512288,
      "step": 9975
    },
    {
      "epoch": 0.3194417770949363,
      "grad_norm": 24.875,
      "learning_rate": 4.270680189136366e-05,
      "loss": 3.9445,
      "num_input_tokens_seen": 13519232,
      "step": 9980
    },
    {
      "epoch": 0.31960181806542476,
      "grad_norm": 36.25,
      "learning_rate": 4.269986998078132e-05,
      "loss": 5.5068,
      "num_input_tokens_seen": 13525584,
      "step": 9985
    },
    {
      "epoch": 0.3197618590359132,
      "grad_norm": 46.0,
      "learning_rate": 4.2692935340656595e-05,
      "loss": 4.5589,
      "num_input_tokens_seen": 13532592,
      "step": 9990
    },
    {
      "epoch": 0.31992190000640164,
      "grad_norm": 25.75,
      "learning_rate": 4.26859979720589e-05,
      "loss": 3.9844,
      "num_input_tokens_seen": 13539472,
      "step": 9995
    },
    {
      "epoch": 0.3200819409768901,
      "grad_norm": 47.75,
      "learning_rate": 4.267905787605806e-05,
      "loss": 5.1543,
      "num_input_tokens_seen": 13545840,
      "step": 10000
    },
    {
      "epoch": 0.3200819409768901,
      "eval_loss": 4.505092144012451,
      "eval_runtime": 487.6314,
      "eval_samples_per_second": 28.476,
      "eval_steps_per_second": 14.238,
      "num_input_tokens_seen": 13545840,
      "step": 10000
    },
    {
      "epoch": 0.3202419819473785,
      "grad_norm": 20.875,
      "learning_rate": 4.267211505372433e-05,
      "loss": 4.4304,
      "num_input_tokens_seen": 13552768,
      "step": 10005
    },
    {
      "epoch": 0.32040202291786696,
      "grad_norm": 20.375,
      "learning_rate": 4.266516950612837e-05,
      "loss": 4.2444,
      "num_input_tokens_seen": 13559424,
      "step": 10010
    },
    {
      "epoch": 0.3205620638883554,
      "grad_norm": 19.375,
      "learning_rate": 4.265822123434128e-05,
      "loss": 4.2938,
      "num_input_tokens_seen": 13566384,
      "step": 10015
    },
    {
      "epoch": 0.32072210485884384,
      "grad_norm": 27.5,
      "learning_rate": 4.265127023943457e-05,
      "loss": 4.3405,
      "num_input_tokens_seen": 13573328,
      "step": 10020
    },
    {
      "epoch": 0.32088214582933233,
      "grad_norm": 28.5,
      "learning_rate": 4.2644316522480176e-05,
      "loss": 3.6183,
      "num_input_tokens_seen": 13580288,
      "step": 10025
    },
    {
      "epoch": 0.3210421867998208,
      "grad_norm": 23.0,
      "learning_rate": 4.263736008455044e-05,
      "loss": 4.8474,
      "num_input_tokens_seen": 13587008,
      "step": 10030
    },
    {
      "epoch": 0.3212022277703092,
      "grad_norm": 21.625,
      "learning_rate": 4.2630400926718125e-05,
      "loss": 4.5798,
      "num_input_tokens_seen": 13594608,
      "step": 10035
    },
    {
      "epoch": 0.32136226874079765,
      "grad_norm": 25.75,
      "learning_rate": 4.262343905005644e-05,
      "loss": 4.361,
      "num_input_tokens_seen": 13601232,
      "step": 10040
    },
    {
      "epoch": 0.3215223097112861,
      "grad_norm": 13.625,
      "learning_rate": 4.261647445563897e-05,
      "loss": 4.0213,
      "num_input_tokens_seen": 13608608,
      "step": 10045
    },
    {
      "epoch": 0.32168235068177453,
      "grad_norm": 38.25,
      "learning_rate": 4.260950714453976e-05,
      "loss": 4.1916,
      "num_input_tokens_seen": 13615120,
      "step": 10050
    },
    {
      "epoch": 0.32184239165226297,
      "grad_norm": 17.875,
      "learning_rate": 4.2602537117833266e-05,
      "loss": 4.4405,
      "num_input_tokens_seen": 13621472,
      "step": 10055
    },
    {
      "epoch": 0.3220024326227514,
      "grad_norm": 19.125,
      "learning_rate": 4.259556437659433e-05,
      "loss": 3.7546,
      "num_input_tokens_seen": 13628096,
      "step": 10060
    },
    {
      "epoch": 0.32216247359323985,
      "grad_norm": 32.0,
      "learning_rate": 4.258858892189825e-05,
      "loss": 4.7089,
      "num_input_tokens_seen": 13634784,
      "step": 10065
    },
    {
      "epoch": 0.3223225145637283,
      "grad_norm": 24.125,
      "learning_rate": 4.2581610754820725e-05,
      "loss": 3.875,
      "num_input_tokens_seen": 13641728,
      "step": 10070
    },
    {
      "epoch": 0.3224825555342168,
      "grad_norm": 30.375,
      "learning_rate": 4.2574629876437876e-05,
      "loss": 5.1703,
      "num_input_tokens_seen": 13648224,
      "step": 10075
    },
    {
      "epoch": 0.3226425965047052,
      "grad_norm": 22.125,
      "learning_rate": 4.256764628782625e-05,
      "loss": 4.2297,
      "num_input_tokens_seen": 13654944,
      "step": 10080
    },
    {
      "epoch": 0.32280263747519367,
      "grad_norm": 26.75,
      "learning_rate": 4.256065999006279e-05,
      "loss": 5.0117,
      "num_input_tokens_seen": 13661392,
      "step": 10085
    },
    {
      "epoch": 0.3229626784456821,
      "grad_norm": 30.0,
      "learning_rate": 4.2553670984224885e-05,
      "loss": 4.3485,
      "num_input_tokens_seen": 13667920,
      "step": 10090
    },
    {
      "epoch": 0.32312271941617055,
      "grad_norm": 26.375,
      "learning_rate": 4.254667927139032e-05,
      "loss": 4.3479,
      "num_input_tokens_seen": 13674528,
      "step": 10095
    },
    {
      "epoch": 0.323282760386659,
      "grad_norm": 39.75,
      "learning_rate": 4.2539684852637295e-05,
      "loss": 4.8486,
      "num_input_tokens_seen": 13681568,
      "step": 10100
    },
    {
      "epoch": 0.3234428013571474,
      "grad_norm": 21.25,
      "learning_rate": 4.253268772904446e-05,
      "loss": 3.9177,
      "num_input_tokens_seen": 13688112,
      "step": 10105
    },
    {
      "epoch": 0.32360284232763586,
      "grad_norm": 40.75,
      "learning_rate": 4.252568790169085e-05,
      "loss": 4.7793,
      "num_input_tokens_seen": 13695296,
      "step": 10110
    },
    {
      "epoch": 0.3237628832981243,
      "grad_norm": 31.875,
      "learning_rate": 4.251868537165592e-05,
      "loss": 3.8709,
      "num_input_tokens_seen": 13702224,
      "step": 10115
    },
    {
      "epoch": 0.32392292426861274,
      "grad_norm": 19.25,
      "learning_rate": 4.251168014001955e-05,
      "loss": 3.9761,
      "num_input_tokens_seen": 13708752,
      "step": 10120
    },
    {
      "epoch": 0.3240829652391012,
      "grad_norm": 16.375,
      "learning_rate": 4.250467220786204e-05,
      "loss": 4.0447,
      "num_input_tokens_seen": 13715536,
      "step": 10125
    },
    {
      "epoch": 0.3242430062095897,
      "grad_norm": 24.0,
      "learning_rate": 4.249766157626409e-05,
      "loss": 4.6312,
      "num_input_tokens_seen": 13722544,
      "step": 10130
    },
    {
      "epoch": 0.3244030471800781,
      "grad_norm": 16.375,
      "learning_rate": 4.249064824630684e-05,
      "loss": 3.9,
      "num_input_tokens_seen": 13729056,
      "step": 10135
    },
    {
      "epoch": 0.32456308815056656,
      "grad_norm": 28.75,
      "learning_rate": 4.248363221907183e-05,
      "loss": 3.8494,
      "num_input_tokens_seen": 13735936,
      "step": 10140
    },
    {
      "epoch": 0.324723129121055,
      "grad_norm": 29.375,
      "learning_rate": 4.2476613495641026e-05,
      "loss": 4.0956,
      "num_input_tokens_seen": 13742624,
      "step": 10145
    },
    {
      "epoch": 0.32488317009154344,
      "grad_norm": 20.125,
      "learning_rate": 4.246959207709679e-05,
      "loss": 3.503,
      "num_input_tokens_seen": 13749200,
      "step": 10150
    },
    {
      "epoch": 0.3250432110620319,
      "grad_norm": 37.75,
      "learning_rate": 4.246256796452192e-05,
      "loss": 3.5618,
      "num_input_tokens_seen": 13755760,
      "step": 10155
    },
    {
      "epoch": 0.3252032520325203,
      "grad_norm": 43.5,
      "learning_rate": 4.245554115899962e-05,
      "loss": 4.7532,
      "num_input_tokens_seen": 13762304,
      "step": 10160
    },
    {
      "epoch": 0.32536329300300876,
      "grad_norm": 47.75,
      "learning_rate": 4.2448511661613514e-05,
      "loss": 5.7185,
      "num_input_tokens_seen": 13769120,
      "step": 10165
    },
    {
      "epoch": 0.3255233339734972,
      "grad_norm": 47.5,
      "learning_rate": 4.2441479473447635e-05,
      "loss": 5.4456,
      "num_input_tokens_seen": 13776080,
      "step": 10170
    },
    {
      "epoch": 0.32568337494398564,
      "grad_norm": 25.25,
      "learning_rate": 4.243444459558644e-05,
      "loss": 4.3008,
      "num_input_tokens_seen": 13783280,
      "step": 10175
    },
    {
      "epoch": 0.32584341591447413,
      "grad_norm": 26.375,
      "learning_rate": 4.24274070291148e-05,
      "loss": 4.301,
      "num_input_tokens_seen": 13790000,
      "step": 10180
    },
    {
      "epoch": 0.3260034568849626,
      "grad_norm": 23.375,
      "learning_rate": 4.242036677511798e-05,
      "loss": 4.1613,
      "num_input_tokens_seen": 13796560,
      "step": 10185
    },
    {
      "epoch": 0.326163497855451,
      "grad_norm": 32.0,
      "learning_rate": 4.241332383468169e-05,
      "loss": 4.7978,
      "num_input_tokens_seen": 13803568,
      "step": 10190
    },
    {
      "epoch": 0.32632353882593945,
      "grad_norm": 19.875,
      "learning_rate": 4.2406278208892034e-05,
      "loss": 3.5833,
      "num_input_tokens_seen": 13810368,
      "step": 10195
    },
    {
      "epoch": 0.3264835797964279,
      "grad_norm": 21.625,
      "learning_rate": 4.2399229898835536e-05,
      "loss": 4.2356,
      "num_input_tokens_seen": 13817104,
      "step": 10200
    },
    {
      "epoch": 0.3264835797964279,
      "eval_loss": 4.470882415771484,
      "eval_runtime": 488.2216,
      "eval_samples_per_second": 28.442,
      "eval_steps_per_second": 14.221,
      "num_input_tokens_seen": 13817104,
      "step": 10200
    },
    {
      "epoch": 0.32664362076691633,
      "grad_norm": 21.0,
      "learning_rate": 4.239217890559914e-05,
      "loss": 3.5641,
      "num_input_tokens_seen": 13823952,
      "step": 10205
    },
    {
      "epoch": 0.32680366173740477,
      "grad_norm": 19.25,
      "learning_rate": 4.238512523027019e-05,
      "loss": 4.958,
      "num_input_tokens_seen": 13830816,
      "step": 10210
    },
    {
      "epoch": 0.3269637027078932,
      "grad_norm": 96.5,
      "learning_rate": 4.237806887393645e-05,
      "loss": 3.9075,
      "num_input_tokens_seen": 13837664,
      "step": 10215
    },
    {
      "epoch": 0.32712374367838165,
      "grad_norm": 17.5,
      "learning_rate": 4.237100983768611e-05,
      "loss": 4.3394,
      "num_input_tokens_seen": 13844752,
      "step": 10220
    },
    {
      "epoch": 0.3272837846488701,
      "grad_norm": 25.0,
      "learning_rate": 4.2363948122607756e-05,
      "loss": 3.8675,
      "num_input_tokens_seen": 13851376,
      "step": 10225
    },
    {
      "epoch": 0.32744382561935853,
      "grad_norm": 42.25,
      "learning_rate": 4.235688372979039e-05,
      "loss": 5.524,
      "num_input_tokens_seen": 13857696,
      "step": 10230
    },
    {
      "epoch": 0.327603866589847,
      "grad_norm": 18.5,
      "learning_rate": 4.234981666032343e-05,
      "loss": 3.9425,
      "num_input_tokens_seen": 13864448,
      "step": 10235
    },
    {
      "epoch": 0.32776390756033547,
      "grad_norm": 43.25,
      "learning_rate": 4.2342746915296704e-05,
      "loss": 4.3106,
      "num_input_tokens_seen": 13871584,
      "step": 10240
    },
    {
      "epoch": 0.3279239485308239,
      "grad_norm": 31.625,
      "learning_rate": 4.233567449580047e-05,
      "loss": 3.9562,
      "num_input_tokens_seen": 13878256,
      "step": 10245
    },
    {
      "epoch": 0.32808398950131235,
      "grad_norm": 22.375,
      "learning_rate": 4.232859940292537e-05,
      "loss": 3.3651,
      "num_input_tokens_seen": 13884624,
      "step": 10250
    },
    {
      "epoch": 0.3282440304718008,
      "grad_norm": 18.375,
      "learning_rate": 4.232152163776248e-05,
      "loss": 4.7347,
      "num_input_tokens_seen": 13891792,
      "step": 10255
    },
    {
      "epoch": 0.3284040714422892,
      "grad_norm": 23.25,
      "learning_rate": 4.231444120140328e-05,
      "loss": 5.0549,
      "num_input_tokens_seen": 13898464,
      "step": 10260
    },
    {
      "epoch": 0.32856411241277766,
      "grad_norm": 27.875,
      "learning_rate": 4.230735809493967e-05,
      "loss": 4.5885,
      "num_input_tokens_seen": 13905216,
      "step": 10265
    },
    {
      "epoch": 0.3287241533832661,
      "grad_norm": 39.0,
      "learning_rate": 4.2300272319463926e-05,
      "loss": 4.6248,
      "num_input_tokens_seen": 13912112,
      "step": 10270
    },
    {
      "epoch": 0.32888419435375454,
      "grad_norm": 32.75,
      "learning_rate": 4.2293183876068786e-05,
      "loss": 3.7897,
      "num_input_tokens_seen": 13918416,
      "step": 10275
    },
    {
      "epoch": 0.329044235324243,
      "grad_norm": 16.5,
      "learning_rate": 4.228609276584737e-05,
      "loss": 3.7726,
      "num_input_tokens_seen": 13925440,
      "step": 10280
    },
    {
      "epoch": 0.3292042762947314,
      "grad_norm": 38.5,
      "learning_rate": 4.227899898989323e-05,
      "loss": 4.6912,
      "num_input_tokens_seen": 13932304,
      "step": 10285
    },
    {
      "epoch": 0.3293643172652199,
      "grad_norm": 31.0,
      "learning_rate": 4.2271902549300293e-05,
      "loss": 4.2777,
      "num_input_tokens_seen": 13939328,
      "step": 10290
    },
    {
      "epoch": 0.32952435823570836,
      "grad_norm": 32.75,
      "learning_rate": 4.226480344516294e-05,
      "loss": 3.8145,
      "num_input_tokens_seen": 13946032,
      "step": 10295
    },
    {
      "epoch": 0.3296843992061968,
      "grad_norm": 31.625,
      "learning_rate": 4.2257701678575925e-05,
      "loss": 4.6532,
      "num_input_tokens_seen": 13953072,
      "step": 10300
    },
    {
      "epoch": 0.32984444017668524,
      "grad_norm": 49.5,
      "learning_rate": 4.225059725063444e-05,
      "loss": 4.5356,
      "num_input_tokens_seen": 13959824,
      "step": 10305
    },
    {
      "epoch": 0.3300044811471737,
      "grad_norm": 29.75,
      "learning_rate": 4.2243490162434074e-05,
      "loss": 3.6957,
      "num_input_tokens_seen": 13966496,
      "step": 10310
    },
    {
      "epoch": 0.3301645221176621,
      "grad_norm": 20.75,
      "learning_rate": 4.223638041507083e-05,
      "loss": 3.9029,
      "num_input_tokens_seen": 13973472,
      "step": 10315
    },
    {
      "epoch": 0.33032456308815056,
      "grad_norm": 19.125,
      "learning_rate": 4.2229268009641124e-05,
      "loss": 4.458,
      "num_input_tokens_seen": 13979968,
      "step": 10320
    },
    {
      "epoch": 0.330484604058639,
      "grad_norm": 24.5,
      "learning_rate": 4.222215294724177e-05,
      "loss": 4.3215,
      "num_input_tokens_seen": 13986512,
      "step": 10325
    },
    {
      "epoch": 0.33064464502912744,
      "grad_norm": 29.5,
      "learning_rate": 4.2215035228970005e-05,
      "loss": 4.0054,
      "num_input_tokens_seen": 13993760,
      "step": 10330
    },
    {
      "epoch": 0.3308046859996159,
      "grad_norm": 25.625,
      "learning_rate": 4.2207914855923464e-05,
      "loss": 4.1557,
      "num_input_tokens_seen": 14000400,
      "step": 10335
    },
    {
      "epoch": 0.3309647269701044,
      "grad_norm": 31.75,
      "learning_rate": 4.220079182920021e-05,
      "loss": 4.2615,
      "num_input_tokens_seen": 14006848,
      "step": 10340
    },
    {
      "epoch": 0.3311247679405928,
      "grad_norm": 28.5,
      "learning_rate": 4.2193666149898705e-05,
      "loss": 4.3284,
      "num_input_tokens_seen": 14013360,
      "step": 10345
    },
    {
      "epoch": 0.33128480891108125,
      "grad_norm": 27.625,
      "learning_rate": 4.21865378191178e-05,
      "loss": 4.5525,
      "num_input_tokens_seen": 14020480,
      "step": 10350
    },
    {
      "epoch": 0.3314448498815697,
      "grad_norm": 34.25,
      "learning_rate": 4.217940683795678e-05,
      "loss": 4.4078,
      "num_input_tokens_seen": 14027216,
      "step": 10355
    },
    {
      "epoch": 0.33160489085205813,
      "grad_norm": 32.0,
      "learning_rate": 4.217227320751534e-05,
      "loss": 3.7281,
      "num_input_tokens_seen": 14034032,
      "step": 10360
    },
    {
      "epoch": 0.33176493182254657,
      "grad_norm": 35.5,
      "learning_rate": 4.216513692889358e-05,
      "loss": 3.7562,
      "num_input_tokens_seen": 14040544,
      "step": 10365
    },
    {
      "epoch": 0.331924972793035,
      "grad_norm": 17.625,
      "learning_rate": 4.215799800319199e-05,
      "loss": 3.2161,
      "num_input_tokens_seen": 14047488,
      "step": 10370
    },
    {
      "epoch": 0.33208501376352345,
      "grad_norm": 18.5,
      "learning_rate": 4.2150856431511485e-05,
      "loss": 5.5854,
      "num_input_tokens_seen": 14054128,
      "step": 10375
    },
    {
      "epoch": 0.3322450547340119,
      "grad_norm": 38.5,
      "learning_rate": 4.214371221495339e-05,
      "loss": 4.6139,
      "num_input_tokens_seen": 14060576,
      "step": 10380
    },
    {
      "epoch": 0.33240509570450033,
      "grad_norm": 23.625,
      "learning_rate": 4.213656535461942e-05,
      "loss": 4.2869,
      "num_input_tokens_seen": 14067392,
      "step": 10385
    },
    {
      "epoch": 0.33256513667498877,
      "grad_norm": 25.125,
      "learning_rate": 4.2129415851611734e-05,
      "loss": 4.2137,
      "num_input_tokens_seen": 14074064,
      "step": 10390
    },
    {
      "epoch": 0.33272517764547727,
      "grad_norm": 19.0,
      "learning_rate": 4.2122263707032855e-05,
      "loss": 3.3242,
      "num_input_tokens_seen": 14080896,
      "step": 10395
    },
    {
      "epoch": 0.3328852186159657,
      "grad_norm": 23.375,
      "learning_rate": 4.211510892198574e-05,
      "loss": 4.757,
      "num_input_tokens_seen": 14088032,
      "step": 10400
    },
    {
      "epoch": 0.3328852186159657,
      "eval_loss": 4.485370635986328,
      "eval_runtime": 488.1586,
      "eval_samples_per_second": 28.446,
      "eval_steps_per_second": 14.223,
      "num_input_tokens_seen": 14088032,
      "step": 10400
    },
    {
      "epoch": 0.33304525958645415,
      "grad_norm": 23.875,
      "learning_rate": 4.210795149757375e-05,
      "loss": 4.2488,
      "num_input_tokens_seen": 14094608,
      "step": 10405
    },
    {
      "epoch": 0.3332053005569426,
      "grad_norm": 29.5,
      "learning_rate": 4.210079143490065e-05,
      "loss": 4.8338,
      "num_input_tokens_seen": 14101504,
      "step": 10410
    },
    {
      "epoch": 0.333365341527431,
      "grad_norm": 36.25,
      "learning_rate": 4.2093628735070604e-05,
      "loss": 3.8482,
      "num_input_tokens_seen": 14108496,
      "step": 10415
    },
    {
      "epoch": 0.33352538249791946,
      "grad_norm": 38.75,
      "learning_rate": 4.208646339918819e-05,
      "loss": 3.8637,
      "num_input_tokens_seen": 14115248,
      "step": 10420
    },
    {
      "epoch": 0.3336854234684079,
      "grad_norm": 31.125,
      "learning_rate": 4.2079295428358414e-05,
      "loss": 3.9984,
      "num_input_tokens_seen": 14121952,
      "step": 10425
    },
    {
      "epoch": 0.33384546443889634,
      "grad_norm": 43.75,
      "learning_rate": 4.207212482368664e-05,
      "loss": 4.8772,
      "num_input_tokens_seen": 14128848,
      "step": 10430
    },
    {
      "epoch": 0.3340055054093848,
      "grad_norm": 38.0,
      "learning_rate": 4.206495158627867e-05,
      "loss": 3.812,
      "num_input_tokens_seen": 14135104,
      "step": 10435
    },
    {
      "epoch": 0.3341655463798732,
      "grad_norm": 22.125,
      "learning_rate": 4.205777571724073e-05,
      "loss": 3.5989,
      "num_input_tokens_seen": 14142032,
      "step": 10440
    },
    {
      "epoch": 0.3343255873503617,
      "grad_norm": 39.25,
      "learning_rate": 4.20505972176794e-05,
      "loss": 4.8604,
      "num_input_tokens_seen": 14148816,
      "step": 10445
    },
    {
      "epoch": 0.33448562832085016,
      "grad_norm": 41.0,
      "learning_rate": 4.204341608870171e-05,
      "loss": 4.1104,
      "num_input_tokens_seen": 14155488,
      "step": 10450
    },
    {
      "epoch": 0.3346456692913386,
      "grad_norm": 36.25,
      "learning_rate": 4.203623233141508e-05,
      "loss": 3.7814,
      "num_input_tokens_seen": 14162240,
      "step": 10455
    },
    {
      "epoch": 0.33480571026182704,
      "grad_norm": 17.375,
      "learning_rate": 4.2029045946927334e-05,
      "loss": 4.7236,
      "num_input_tokens_seen": 14168832,
      "step": 10460
    },
    {
      "epoch": 0.3349657512323155,
      "grad_norm": 35.0,
      "learning_rate": 4.20218569363467e-05,
      "loss": 4.6773,
      "num_input_tokens_seen": 14175856,
      "step": 10465
    },
    {
      "epoch": 0.3351257922028039,
      "grad_norm": 41.75,
      "learning_rate": 4.2014665300781834e-05,
      "loss": 4.6282,
      "num_input_tokens_seen": 14183088,
      "step": 10470
    },
    {
      "epoch": 0.33528583317329236,
      "grad_norm": 29.0,
      "learning_rate": 4.200747104134174e-05,
      "loss": 3.7014,
      "num_input_tokens_seen": 14189952,
      "step": 10475
    },
    {
      "epoch": 0.3354458741437808,
      "grad_norm": 16.375,
      "learning_rate": 4.200027415913588e-05,
      "loss": 4.1317,
      "num_input_tokens_seen": 14196672,
      "step": 10480
    },
    {
      "epoch": 0.33560591511426924,
      "grad_norm": 28.625,
      "learning_rate": 4.1993074655274126e-05,
      "loss": 4.367,
      "num_input_tokens_seen": 14203968,
      "step": 10485
    },
    {
      "epoch": 0.3357659560847577,
      "grad_norm": 15.25,
      "learning_rate": 4.198587253086669e-05,
      "loss": 3.6529,
      "num_input_tokens_seen": 14211168,
      "step": 10490
    },
    {
      "epoch": 0.3359259970552461,
      "grad_norm": 25.125,
      "learning_rate": 4.197866778702426e-05,
      "loss": 3.8717,
      "num_input_tokens_seen": 14217920,
      "step": 10495
    },
    {
      "epoch": 0.3360860380257346,
      "grad_norm": 39.75,
      "learning_rate": 4.197146042485789e-05,
      "loss": 4.0233,
      "num_input_tokens_seen": 14224608,
      "step": 10500
    },
    {
      "epoch": 0.33624607899622305,
      "grad_norm": 32.0,
      "learning_rate": 4.1964250445479046e-05,
      "loss": 4.0898,
      "num_input_tokens_seen": 14231488,
      "step": 10505
    },
    {
      "epoch": 0.3364061199667115,
      "grad_norm": 52.25,
      "learning_rate": 4.19570378499996e-05,
      "loss": 4.6669,
      "num_input_tokens_seen": 14238784,
      "step": 10510
    },
    {
      "epoch": 0.33656616093719993,
      "grad_norm": 25.625,
      "learning_rate": 4.194982263953182e-05,
      "loss": 4.6401,
      "num_input_tokens_seen": 14245360,
      "step": 10515
    },
    {
      "epoch": 0.33672620190768837,
      "grad_norm": 33.0,
      "learning_rate": 4.194260481518838e-05,
      "loss": 4.4187,
      "num_input_tokens_seen": 14251808,
      "step": 10520
    },
    {
      "epoch": 0.3368862428781768,
      "grad_norm": 35.0,
      "learning_rate": 4.1935384378082366e-05,
      "loss": 4.2644,
      "num_input_tokens_seen": 14258624,
      "step": 10525
    },
    {
      "epoch": 0.33704628384866525,
      "grad_norm": 23.25,
      "learning_rate": 4.1928161329327267e-05,
      "loss": 4.9709,
      "num_input_tokens_seen": 14265952,
      "step": 10530
    },
    {
      "epoch": 0.3372063248191537,
      "grad_norm": 43.75,
      "learning_rate": 4.1920935670036945e-05,
      "loss": 4.2839,
      "num_input_tokens_seen": 14273264,
      "step": 10535
    },
    {
      "epoch": 0.33736636578964213,
      "grad_norm": 33.5,
      "learning_rate": 4.1913707401325705e-05,
      "loss": 5.2069,
      "num_input_tokens_seen": 14279968,
      "step": 10540
    },
    {
      "epoch": 0.33752640676013057,
      "grad_norm": 22.75,
      "learning_rate": 4.1906476524308235e-05,
      "loss": 3.7931,
      "num_input_tokens_seen": 14286624,
      "step": 10545
    },
    {
      "epoch": 0.33768644773061907,
      "grad_norm": 23.0,
      "learning_rate": 4.189924304009962e-05,
      "loss": 4.7571,
      "num_input_tokens_seen": 14293600,
      "step": 10550
    },
    {
      "epoch": 0.3378464887011075,
      "grad_norm": 30.125,
      "learning_rate": 4.189200694981537e-05,
      "loss": 3.7391,
      "num_input_tokens_seen": 14300272,
      "step": 10555
    },
    {
      "epoch": 0.33800652967159595,
      "grad_norm": 23.375,
      "learning_rate": 4.188476825457136e-05,
      "loss": 4.2054,
      "num_input_tokens_seen": 14307280,
      "step": 10560
    },
    {
      "epoch": 0.3381665706420844,
      "grad_norm": 24.75,
      "learning_rate": 4.18775269554839e-05,
      "loss": 4.5897,
      "num_input_tokens_seen": 14313920,
      "step": 10565
    },
    {
      "epoch": 0.3383266116125728,
      "grad_norm": 27.375,
      "learning_rate": 4.187028305366969e-05,
      "loss": 3.8912,
      "num_input_tokens_seen": 14321040,
      "step": 10570
    },
    {
      "epoch": 0.33848665258306126,
      "grad_norm": 25.875,
      "learning_rate": 4.1863036550245824e-05,
      "loss": 4.4612,
      "num_input_tokens_seen": 14327952,
      "step": 10575
    },
    {
      "epoch": 0.3386466935535497,
      "grad_norm": 38.0,
      "learning_rate": 4.1855787446329806e-05,
      "loss": 3.8502,
      "num_input_tokens_seen": 14334976,
      "step": 10580
    },
    {
      "epoch": 0.33880673452403814,
      "grad_norm": 33.75,
      "learning_rate": 4.184853574303955e-05,
      "loss": 4.5434,
      "num_input_tokens_seen": 14341456,
      "step": 10585
    },
    {
      "epoch": 0.3389667754945266,
      "grad_norm": 43.75,
      "learning_rate": 4.184128144149334e-05,
      "loss": 3.9952,
      "num_input_tokens_seen": 14348384,
      "step": 10590
    },
    {
      "epoch": 0.339126816465015,
      "grad_norm": 21.0,
      "learning_rate": 4.1834024542809896e-05,
      "loss": 4.1159,
      "num_input_tokens_seen": 14354912,
      "step": 10595
    },
    {
      "epoch": 0.33928685743550346,
      "grad_norm": 32.25,
      "learning_rate": 4.1826765048108315e-05,
      "loss": 5.2015,
      "num_input_tokens_seen": 14361280,
      "step": 10600
    },
    {
      "epoch": 0.33928685743550346,
      "eval_loss": 4.496551513671875,
      "eval_runtime": 487.9005,
      "eval_samples_per_second": 28.461,
      "eval_steps_per_second": 14.23,
      "num_input_tokens_seen": 14361280,
      "step": 10600
    },
    {
      "epoch": 0.33944689840599196,
      "grad_norm": 22.75,
      "learning_rate": 4.181950295850811e-05,
      "loss": 3.8899,
      "num_input_tokens_seen": 14367760,
      "step": 10605
    },
    {
      "epoch": 0.3396069393764804,
      "grad_norm": 26.375,
      "learning_rate": 4.181223827512918e-05,
      "loss": 4.1648,
      "num_input_tokens_seen": 14374416,
      "step": 10610
    },
    {
      "epoch": 0.33976698034696884,
      "grad_norm": 39.25,
      "learning_rate": 4.180497099909183e-05,
      "loss": 4.0668,
      "num_input_tokens_seen": 14381184,
      "step": 10615
    },
    {
      "epoch": 0.3399270213174573,
      "grad_norm": 27.125,
      "learning_rate": 4.179770113151677e-05,
      "loss": 5.4074,
      "num_input_tokens_seen": 14387776,
      "step": 10620
    },
    {
      "epoch": 0.3400870622879457,
      "grad_norm": 44.0,
      "learning_rate": 4.179042867352511e-05,
      "loss": 4.3094,
      "num_input_tokens_seen": 14394672,
      "step": 10625
    },
    {
      "epoch": 0.34024710325843416,
      "grad_norm": 24.75,
      "learning_rate": 4.1783153626238334e-05,
      "loss": 3.3363,
      "num_input_tokens_seen": 14401264,
      "step": 10630
    },
    {
      "epoch": 0.3404071442289226,
      "grad_norm": 47.25,
      "learning_rate": 4.177587599077836e-05,
      "loss": 4.5476,
      "num_input_tokens_seen": 14408064,
      "step": 10635
    },
    {
      "epoch": 0.34056718519941104,
      "grad_norm": 29.375,
      "learning_rate": 4.1768595768267494e-05,
      "loss": 4.2288,
      "num_input_tokens_seen": 14415200,
      "step": 10640
    },
    {
      "epoch": 0.3407272261698995,
      "grad_norm": 15.1875,
      "learning_rate": 4.176131295982843e-05,
      "loss": 4.2659,
      "num_input_tokens_seen": 14421600,
      "step": 10645
    },
    {
      "epoch": 0.3408872671403879,
      "grad_norm": 33.0,
      "learning_rate": 4.1754027566584276e-05,
      "loss": 5.1421,
      "num_input_tokens_seen": 14428352,
      "step": 10650
    },
    {
      "epoch": 0.34104730811087636,
      "grad_norm": 30.125,
      "learning_rate": 4.174673958965852e-05,
      "loss": 4.8157,
      "num_input_tokens_seen": 14435536,
      "step": 10655
    },
    {
      "epoch": 0.34120734908136485,
      "grad_norm": 30.875,
      "learning_rate": 4.173944903017507e-05,
      "loss": 4.2859,
      "num_input_tokens_seen": 14442240,
      "step": 10660
    },
    {
      "epoch": 0.3413673900518533,
      "grad_norm": 47.75,
      "learning_rate": 4.173215588925822e-05,
      "loss": 5.6831,
      "num_input_tokens_seen": 14448848,
      "step": 10665
    },
    {
      "epoch": 0.34152743102234173,
      "grad_norm": 30.375,
      "learning_rate": 4.172486016803266e-05,
      "loss": 3.8366,
      "num_input_tokens_seen": 14455696,
      "step": 10670
    },
    {
      "epoch": 0.34168747199283017,
      "grad_norm": 31.0,
      "learning_rate": 4.171756186762349e-05,
      "loss": 4.1709,
      "num_input_tokens_seen": 14462448,
      "step": 10675
    },
    {
      "epoch": 0.3418475129633186,
      "grad_norm": 30.625,
      "learning_rate": 4.171026098915619e-05,
      "loss": 5.2479,
      "num_input_tokens_seen": 14469056,
      "step": 10680
    },
    {
      "epoch": 0.34200755393380705,
      "grad_norm": 37.0,
      "learning_rate": 4.170295753375665e-05,
      "loss": 4.6991,
      "num_input_tokens_seen": 14475840,
      "step": 10685
    },
    {
      "epoch": 0.3421675949042955,
      "grad_norm": 29.25,
      "learning_rate": 4.169565150255117e-05,
      "loss": 3.3249,
      "num_input_tokens_seen": 14482912,
      "step": 10690
    },
    {
      "epoch": 0.34232763587478393,
      "grad_norm": 30.0,
      "learning_rate": 4.16883428966664e-05,
      "loss": 3.3481,
      "num_input_tokens_seen": 14489600,
      "step": 10695
    },
    {
      "epoch": 0.34248767684527237,
      "grad_norm": 24.75,
      "learning_rate": 4.168103171722944e-05,
      "loss": 3.8274,
      "num_input_tokens_seen": 14496256,
      "step": 10700
    },
    {
      "epoch": 0.3426477178157608,
      "grad_norm": 32.75,
      "learning_rate": 4.167371796536777e-05,
      "loss": 4.2651,
      "num_input_tokens_seen": 14503344,
      "step": 10705
    },
    {
      "epoch": 0.3428077587862493,
      "grad_norm": 40.25,
      "learning_rate": 4.166640164220924e-05,
      "loss": 4.2912,
      "num_input_tokens_seen": 14510528,
      "step": 10710
    },
    {
      "epoch": 0.34296779975673775,
      "grad_norm": 37.5,
      "learning_rate": 4.1659082748882144e-05,
      "loss": 4.8085,
      "num_input_tokens_seen": 14517248,
      "step": 10715
    },
    {
      "epoch": 0.3431278407272262,
      "grad_norm": 35.0,
      "learning_rate": 4.1651761286515135e-05,
      "loss": 4.334,
      "num_input_tokens_seen": 14523984,
      "step": 10720
    },
    {
      "epoch": 0.3432878816977146,
      "grad_norm": 24.875,
      "learning_rate": 4.164443725623728e-05,
      "loss": 4.3835,
      "num_input_tokens_seen": 14530448,
      "step": 10725
    },
    {
      "epoch": 0.34344792266820307,
      "grad_norm": 24.0,
      "learning_rate": 4.163711065917802e-05,
      "loss": 4.1092,
      "num_input_tokens_seen": 14537280,
      "step": 10730
    },
    {
      "epoch": 0.3436079636386915,
      "grad_norm": 24.125,
      "learning_rate": 4.1629781496467234e-05,
      "loss": 4.8807,
      "num_input_tokens_seen": 14544400,
      "step": 10735
    },
    {
      "epoch": 0.34376800460917994,
      "grad_norm": 38.5,
      "learning_rate": 4.1622449769235164e-05,
      "loss": 4.3551,
      "num_input_tokens_seen": 14551264,
      "step": 10740
    },
    {
      "epoch": 0.3439280455796684,
      "grad_norm": 29.125,
      "learning_rate": 4.161511547861243e-05,
      "loss": 4.3367,
      "num_input_tokens_seen": 14558176,
      "step": 10745
    },
    {
      "epoch": 0.3440880865501568,
      "grad_norm": 35.25,
      "learning_rate": 4.1607778625730104e-05,
      "loss": 3.6735,
      "num_input_tokens_seen": 14564896,
      "step": 10750
    },
    {
      "epoch": 0.34424812752064526,
      "grad_norm": 30.625,
      "learning_rate": 4.160043921171961e-05,
      "loss": 3.7655,
      "num_input_tokens_seen": 14571904,
      "step": 10755
    },
    {
      "epoch": 0.3444081684911337,
      "grad_norm": 29.75,
      "learning_rate": 4.159309723771276e-05,
      "loss": 3.6747,
      "num_input_tokens_seen": 14578640,
      "step": 10760
    },
    {
      "epoch": 0.3445682094616222,
      "grad_norm": 44.5,
      "learning_rate": 4.158575270484181e-05,
      "loss": 4.921,
      "num_input_tokens_seen": 14584944,
      "step": 10765
    },
    {
      "epoch": 0.34472825043211064,
      "grad_norm": 39.5,
      "learning_rate": 4.157840561423936e-05,
      "loss": 5.5219,
      "num_input_tokens_seen": 14591328,
      "step": 10770
    },
    {
      "epoch": 0.3448882914025991,
      "grad_norm": 32.75,
      "learning_rate": 4.1571055967038416e-05,
      "loss": 4.7034,
      "num_input_tokens_seen": 14597744,
      "step": 10775
    },
    {
      "epoch": 0.3450483323730875,
      "grad_norm": 40.0,
      "learning_rate": 4.156370376437241e-05,
      "loss": 3.9248,
      "num_input_tokens_seen": 14604368,
      "step": 10780
    },
    {
      "epoch": 0.34520837334357596,
      "grad_norm": 28.625,
      "learning_rate": 4.155634900737513e-05,
      "loss": 4.0244,
      "num_input_tokens_seen": 14611040,
      "step": 10785
    },
    {
      "epoch": 0.3453684143140644,
      "grad_norm": 26.75,
      "learning_rate": 4.1548991697180764e-05,
      "loss": 4.525,
      "num_input_tokens_seen": 14617328,
      "step": 10790
    },
    {
      "epoch": 0.34552845528455284,
      "grad_norm": 22.375,
      "learning_rate": 4.1541631834923914e-05,
      "loss": 4.9057,
      "num_input_tokens_seen": 14623888,
      "step": 10795
    },
    {
      "epoch": 0.3456884962550413,
      "grad_norm": 35.25,
      "learning_rate": 4.153426942173956e-05,
      "loss": 4.8125,
      "num_input_tokens_seen": 14631040,
      "step": 10800
    },
    {
      "epoch": 0.3456884962550413,
      "eval_loss": 4.488132953643799,
      "eval_runtime": 488.5019,
      "eval_samples_per_second": 28.426,
      "eval_steps_per_second": 14.213,
      "num_input_tokens_seen": 14631040,
      "step": 10800
    },
    {
      "epoch": 0.3458485372255297,
      "grad_norm": 35.5,
      "learning_rate": 4.152690445876308e-05,
      "loss": 4.1186,
      "num_input_tokens_seen": 14638160,
      "step": 10805
    },
    {
      "epoch": 0.34600857819601816,
      "grad_norm": 46.5,
      "learning_rate": 4.1519536947130245e-05,
      "loss": 4.0844,
      "num_input_tokens_seen": 14645136,
      "step": 10810
    },
    {
      "epoch": 0.34616861916650665,
      "grad_norm": 39.75,
      "learning_rate": 4.151216688797722e-05,
      "loss": 3.9007,
      "num_input_tokens_seen": 14651728,
      "step": 10815
    },
    {
      "epoch": 0.3463286601369951,
      "grad_norm": 27.875,
      "learning_rate": 4.150479428244054e-05,
      "loss": 4.3799,
      "num_input_tokens_seen": 14658192,
      "step": 10820
    },
    {
      "epoch": 0.34648870110748353,
      "grad_norm": 20.25,
      "learning_rate": 4.1497419131657176e-05,
      "loss": 3.4346,
      "num_input_tokens_seen": 14664672,
      "step": 10825
    },
    {
      "epoch": 0.34664874207797197,
      "grad_norm": 31.0,
      "learning_rate": 4.149004143676447e-05,
      "loss": 3.6728,
      "num_input_tokens_seen": 14671280,
      "step": 10830
    },
    {
      "epoch": 0.3468087830484604,
      "grad_norm": 44.5,
      "learning_rate": 4.148266119890015e-05,
      "loss": 4.5085,
      "num_input_tokens_seen": 14678368,
      "step": 10835
    },
    {
      "epoch": 0.34696882401894885,
      "grad_norm": 29.75,
      "learning_rate": 4.1475278419202324e-05,
      "loss": 4.8422,
      "num_input_tokens_seen": 14685216,
      "step": 10840
    },
    {
      "epoch": 0.3471288649894373,
      "grad_norm": 15.875,
      "learning_rate": 4.146789309880953e-05,
      "loss": 3.5247,
      "num_input_tokens_seen": 14691952,
      "step": 10845
    },
    {
      "epoch": 0.34728890595992573,
      "grad_norm": 42.75,
      "learning_rate": 4.146050523886068e-05,
      "loss": 4.7378,
      "num_input_tokens_seen": 14698864,
      "step": 10850
    },
    {
      "epoch": 0.34744894693041417,
      "grad_norm": 27.75,
      "learning_rate": 4.1453114840495055e-05,
      "loss": 5.6802,
      "num_input_tokens_seen": 14705408,
      "step": 10855
    },
    {
      "epoch": 0.3476089879009026,
      "grad_norm": 26.25,
      "learning_rate": 4.1445721904852364e-05,
      "loss": 3.8905,
      "num_input_tokens_seen": 14712304,
      "step": 10860
    },
    {
      "epoch": 0.34776902887139105,
      "grad_norm": 21.625,
      "learning_rate": 4.143832643307269e-05,
      "loss": 4.5949,
      "num_input_tokens_seen": 14718624,
      "step": 10865
    },
    {
      "epoch": 0.34792906984187955,
      "grad_norm": 16.25,
      "learning_rate": 4.1430928426296503e-05,
      "loss": 3.7852,
      "num_input_tokens_seen": 14725920,
      "step": 10870
    },
    {
      "epoch": 0.348089110812368,
      "grad_norm": 23.875,
      "learning_rate": 4.142352788566466e-05,
      "loss": 4.0664,
      "num_input_tokens_seen": 14733008,
      "step": 10875
    },
    {
      "epoch": 0.3482491517828564,
      "grad_norm": 20.625,
      "learning_rate": 4.1416124812318424e-05,
      "loss": 4.4883,
      "num_input_tokens_seen": 14739456,
      "step": 10880
    },
    {
      "epoch": 0.34840919275334487,
      "grad_norm": 39.75,
      "learning_rate": 4.1408719207399453e-05,
      "loss": 4.356,
      "num_input_tokens_seen": 14746320,
      "step": 10885
    },
    {
      "epoch": 0.3485692337238333,
      "grad_norm": 36.5,
      "learning_rate": 4.140131107204978e-05,
      "loss": 4.3959,
      "num_input_tokens_seen": 14753392,
      "step": 10890
    },
    {
      "epoch": 0.34872927469432174,
      "grad_norm": 24.25,
      "learning_rate": 4.139390040741182e-05,
      "loss": 4.1148,
      "num_input_tokens_seen": 14760064,
      "step": 10895
    },
    {
      "epoch": 0.3488893156648102,
      "grad_norm": 13.3125,
      "learning_rate": 4.1386487214628396e-05,
      "loss": 3.3302,
      "num_input_tokens_seen": 14766592,
      "step": 10900
    },
    {
      "epoch": 0.3490493566352986,
      "grad_norm": 26.25,
      "learning_rate": 4.137907149484272e-05,
      "loss": 4.7774,
      "num_input_tokens_seen": 14773504,
      "step": 10905
    },
    {
      "epoch": 0.34920939760578706,
      "grad_norm": 30.625,
      "learning_rate": 4.137165324919839e-05,
      "loss": 4.1173,
      "num_input_tokens_seen": 14780080,
      "step": 10910
    },
    {
      "epoch": 0.3493694385762755,
      "grad_norm": 42.0,
      "learning_rate": 4.136423247883939e-05,
      "loss": 3.9001,
      "num_input_tokens_seen": 14787008,
      "step": 10915
    },
    {
      "epoch": 0.349529479546764,
      "grad_norm": 29.375,
      "learning_rate": 4.135680918491009e-05,
      "loss": 3.314,
      "num_input_tokens_seen": 14793824,
      "step": 10920
    },
    {
      "epoch": 0.34968952051725244,
      "grad_norm": 29.625,
      "learning_rate": 4.1349383368555265e-05,
      "loss": 4.1597,
      "num_input_tokens_seen": 14800672,
      "step": 10925
    },
    {
      "epoch": 0.3498495614877409,
      "grad_norm": 31.0,
      "learning_rate": 4.1341955030920065e-05,
      "loss": 4.5538,
      "num_input_tokens_seen": 14807088,
      "step": 10930
    },
    {
      "epoch": 0.3500096024582293,
      "grad_norm": 23.875,
      "learning_rate": 4.1334524173150036e-05,
      "loss": 3.831,
      "num_input_tokens_seen": 14814224,
      "step": 10935
    },
    {
      "epoch": 0.35016964342871776,
      "grad_norm": 25.375,
      "learning_rate": 4.13270907963911e-05,
      "loss": 3.9419,
      "num_input_tokens_seen": 14820896,
      "step": 10940
    },
    {
      "epoch": 0.3503296843992062,
      "grad_norm": 67.5,
      "learning_rate": 4.131965490178959e-05,
      "loss": 4.7443,
      "num_input_tokens_seen": 14827728,
      "step": 10945
    },
    {
      "epoch": 0.35048972536969464,
      "grad_norm": 41.0,
      "learning_rate": 4.131221649049222e-05,
      "loss": 4.6965,
      "num_input_tokens_seen": 14834432,
      "step": 10950
    },
    {
      "epoch": 0.3506497663401831,
      "grad_norm": 26.0,
      "learning_rate": 4.130477556364606e-05,
      "loss": 5.1084,
      "num_input_tokens_seen": 14841040,
      "step": 10955
    },
    {
      "epoch": 0.3508098073106715,
      "grad_norm": 55.25,
      "learning_rate": 4.129733212239861e-05,
      "loss": 3.9311,
      "num_input_tokens_seen": 14847744,
      "step": 10960
    },
    {
      "epoch": 0.35096984828115996,
      "grad_norm": 28.625,
      "learning_rate": 4.128988616789774e-05,
      "loss": 5.0418,
      "num_input_tokens_seen": 14854768,
      "step": 10965
    },
    {
      "epoch": 0.3511298892516484,
      "grad_norm": 53.25,
      "learning_rate": 4.1282437701291724e-05,
      "loss": 4.2933,
      "num_input_tokens_seen": 14861712,
      "step": 10970
    },
    {
      "epoch": 0.3512899302221369,
      "grad_norm": 43.75,
      "learning_rate": 4.1274986723729184e-05,
      "loss": 4.9593,
      "num_input_tokens_seen": 14868640,
      "step": 10975
    },
    {
      "epoch": 0.35144997119262533,
      "grad_norm": 20.375,
      "learning_rate": 4.126753323635917e-05,
      "loss": 4.4924,
      "num_input_tokens_seen": 14875040,
      "step": 10980
    },
    {
      "epoch": 0.35161001216311377,
      "grad_norm": 26.875,
      "learning_rate": 4.12600772403311e-05,
      "loss": 4.6744,
      "num_input_tokens_seen": 14881728,
      "step": 10985
    },
    {
      "epoch": 0.3517700531336022,
      "grad_norm": 37.25,
      "learning_rate": 4.125261873679479e-05,
      "loss": 4.8144,
      "num_input_tokens_seen": 14888304,
      "step": 10990
    },
    {
      "epoch": 0.35193009410409065,
      "grad_norm": 37.75,
      "learning_rate": 4.124515772690042e-05,
      "loss": 5.205,
      "num_input_tokens_seen": 14895120,
      "step": 10995
    },
    {
      "epoch": 0.3520901350745791,
      "grad_norm": 23.5,
      "learning_rate": 4.123769421179858e-05,
      "loss": 3.7125,
      "num_input_tokens_seen": 14901648,
      "step": 11000
    },
    {
      "epoch": 0.3520901350745791,
      "eval_loss": 4.483687877655029,
      "eval_runtime": 488.581,
      "eval_samples_per_second": 28.421,
      "eval_steps_per_second": 14.211,
      "num_input_tokens_seen": 14901648,
      "step": 11000
    },
    {
      "epoch": 0.35225017604506753,
      "grad_norm": 33.0,
      "learning_rate": 4.1230228192640236e-05,
      "loss": 3.8989,
      "num_input_tokens_seen": 14908368,
      "step": 11005
    },
    {
      "epoch": 0.35241021701555597,
      "grad_norm": 25.75,
      "learning_rate": 4.122275967057675e-05,
      "loss": 3.5951,
      "num_input_tokens_seen": 14915040,
      "step": 11010
    },
    {
      "epoch": 0.3525702579860444,
      "grad_norm": 38.25,
      "learning_rate": 4.1215288646759846e-05,
      "loss": 4.2069,
      "num_input_tokens_seen": 14921872,
      "step": 11015
    },
    {
      "epoch": 0.35273029895653285,
      "grad_norm": 34.25,
      "learning_rate": 4.120781512234166e-05,
      "loss": 3.7943,
      "num_input_tokens_seen": 14928752,
      "step": 11020
    },
    {
      "epoch": 0.3528903399270213,
      "grad_norm": 39.25,
      "learning_rate": 4.120033909847471e-05,
      "loss": 4.7256,
      "num_input_tokens_seen": 14935424,
      "step": 11025
    },
    {
      "epoch": 0.3530503808975098,
      "grad_norm": 40.25,
      "learning_rate": 4.119286057631187e-05,
      "loss": 5.2952,
      "num_input_tokens_seen": 14942528,
      "step": 11030
    },
    {
      "epoch": 0.3532104218679982,
      "grad_norm": 31.875,
      "learning_rate": 4.118537955700646e-05,
      "loss": 4.1934,
      "num_input_tokens_seen": 14949312,
      "step": 11035
    },
    {
      "epoch": 0.35337046283848667,
      "grad_norm": 17.75,
      "learning_rate": 4.11778960417121e-05,
      "loss": 4.4623,
      "num_input_tokens_seen": 14955808,
      "step": 11040
    },
    {
      "epoch": 0.3535305038089751,
      "grad_norm": 31.625,
      "learning_rate": 4.117041003158288e-05,
      "loss": 4.4541,
      "num_input_tokens_seen": 14962448,
      "step": 11045
    },
    {
      "epoch": 0.35369054477946354,
      "grad_norm": 42.75,
      "learning_rate": 4.1162921527773215e-05,
      "loss": 4.7157,
      "num_input_tokens_seen": 14969040,
      "step": 11050
    },
    {
      "epoch": 0.353850585749952,
      "grad_norm": 32.0,
      "learning_rate": 4.115543053143794e-05,
      "loss": 4.6796,
      "num_input_tokens_seen": 14975856,
      "step": 11055
    },
    {
      "epoch": 0.3540106267204404,
      "grad_norm": 52.25,
      "learning_rate": 4.114793704373226e-05,
      "loss": 3.5997,
      "num_input_tokens_seen": 14982496,
      "step": 11060
    },
    {
      "epoch": 0.35417066769092886,
      "grad_norm": 23.5,
      "learning_rate": 4.114044106581175e-05,
      "loss": 4.1922,
      "num_input_tokens_seen": 14989376,
      "step": 11065
    },
    {
      "epoch": 0.3543307086614173,
      "grad_norm": 27.75,
      "learning_rate": 4.11329425988324e-05,
      "loss": 3.4257,
      "num_input_tokens_seen": 14995936,
      "step": 11070
    },
    {
      "epoch": 0.35449074963190574,
      "grad_norm": 23.125,
      "learning_rate": 4.112544164395056e-05,
      "loss": 4.5189,
      "num_input_tokens_seen": 15002160,
      "step": 11075
    },
    {
      "epoch": 0.35465079060239424,
      "grad_norm": 23.125,
      "learning_rate": 4.111793820232297e-05,
      "loss": 4.429,
      "num_input_tokens_seen": 15009376,
      "step": 11080
    },
    {
      "epoch": 0.3548108315728827,
      "grad_norm": 22.125,
      "learning_rate": 4.1110432275106767e-05,
      "loss": 3.6785,
      "num_input_tokens_seen": 15016032,
      "step": 11085
    },
    {
      "epoch": 0.3549708725433711,
      "grad_norm": 29.875,
      "learning_rate": 4.110292386345944e-05,
      "loss": 4.6306,
      "num_input_tokens_seen": 15022816,
      "step": 11090
    },
    {
      "epoch": 0.35513091351385956,
      "grad_norm": 19.625,
      "learning_rate": 4.109541296853891e-05,
      "loss": 4.1938,
      "num_input_tokens_seen": 15029232,
      "step": 11095
    },
    {
      "epoch": 0.355290954484348,
      "grad_norm": 42.75,
      "learning_rate": 4.108789959150341e-05,
      "loss": 4.0917,
      "num_input_tokens_seen": 15035984,
      "step": 11100
    },
    {
      "epoch": 0.35545099545483644,
      "grad_norm": 26.125,
      "learning_rate": 4.108038373351163e-05,
      "loss": 4.4637,
      "num_input_tokens_seen": 15042608,
      "step": 11105
    },
    {
      "epoch": 0.3556110364253249,
      "grad_norm": 26.875,
      "learning_rate": 4.10728653957226e-05,
      "loss": 3.4238,
      "num_input_tokens_seen": 15049472,
      "step": 11110
    },
    {
      "epoch": 0.3557710773958133,
      "grad_norm": 16.75,
      "learning_rate": 4.106534457929575e-05,
      "loss": 3.7855,
      "num_input_tokens_seen": 15056800,
      "step": 11115
    },
    {
      "epoch": 0.35593111836630176,
      "grad_norm": 29.5,
      "learning_rate": 4.105782128539086e-05,
      "loss": 4.415,
      "num_input_tokens_seen": 15063856,
      "step": 11120
    },
    {
      "epoch": 0.3560911593367902,
      "grad_norm": 54.25,
      "learning_rate": 4.1050295515168144e-05,
      "loss": 4.2708,
      "num_input_tokens_seen": 15070336,
      "step": 11125
    },
    {
      "epoch": 0.35625120030727864,
      "grad_norm": 21.0,
      "learning_rate": 4.1042767269788155e-05,
      "loss": 4.3654,
      "num_input_tokens_seen": 15076992,
      "step": 11130
    },
    {
      "epoch": 0.35641124127776713,
      "grad_norm": 21.625,
      "learning_rate": 4.103523655041185e-05,
      "loss": 3.5952,
      "num_input_tokens_seen": 15083408,
      "step": 11135
    },
    {
      "epoch": 0.35657128224825557,
      "grad_norm": 40.25,
      "learning_rate": 4.102770335820055e-05,
      "loss": 4.8828,
      "num_input_tokens_seen": 15090448,
      "step": 11140
    },
    {
      "epoch": 0.356731323218744,
      "grad_norm": 42.5,
      "learning_rate": 4.1020167694315984e-05,
      "loss": 4.4468,
      "num_input_tokens_seen": 15097488,
      "step": 11145
    },
    {
      "epoch": 0.35689136418923245,
      "grad_norm": 24.875,
      "learning_rate": 4.101262955992023e-05,
      "loss": 4.5002,
      "num_input_tokens_seen": 15104272,
      "step": 11150
    },
    {
      "epoch": 0.3570514051597209,
      "grad_norm": 48.5,
      "learning_rate": 4.100508895617578e-05,
      "loss": 4.302,
      "num_input_tokens_seen": 15110928,
      "step": 11155
    },
    {
      "epoch": 0.35721144613020933,
      "grad_norm": 31.75,
      "learning_rate": 4.099754588424547e-05,
      "loss": 4.1955,
      "num_input_tokens_seen": 15117536,
      "step": 11160
    },
    {
      "epoch": 0.35737148710069777,
      "grad_norm": 25.375,
      "learning_rate": 4.0990000345292546e-05,
      "loss": 4.9244,
      "num_input_tokens_seen": 15124416,
      "step": 11165
    },
    {
      "epoch": 0.3575315280711862,
      "grad_norm": 41.0,
      "learning_rate": 4.098245234048064e-05,
      "loss": 4.6156,
      "num_input_tokens_seen": 15130800,
      "step": 11170
    },
    {
      "epoch": 0.35769156904167465,
      "grad_norm": 25.125,
      "learning_rate": 4.0974901870973726e-05,
      "loss": 3.7832,
      "num_input_tokens_seen": 15137408,
      "step": 11175
    },
    {
      "epoch": 0.3578516100121631,
      "grad_norm": 20.125,
      "learning_rate": 4.096734893793619e-05,
      "loss": 5.0863,
      "num_input_tokens_seen": 15144240,
      "step": 11180
    },
    {
      "epoch": 0.3580116509826516,
      "grad_norm": 36.5,
      "learning_rate": 4.095979354253279e-05,
      "loss": 3.4857,
      "num_input_tokens_seen": 15150928,
      "step": 11185
    },
    {
      "epoch": 0.35817169195314,
      "grad_norm": 39.75,
      "learning_rate": 4.0952235685928656e-05,
      "loss": 4.2837,
      "num_input_tokens_seen": 15157552,
      "step": 11190
    },
    {
      "epoch": 0.35833173292362847,
      "grad_norm": 15.3125,
      "learning_rate": 4.094467536928932e-05,
      "loss": 3.5721,
      "num_input_tokens_seen": 15164352,
      "step": 11195
    },
    {
      "epoch": 0.3584917738941169,
      "grad_norm": 31.125,
      "learning_rate": 4.093711259378067e-05,
      "loss": 4.6481,
      "num_input_tokens_seen": 15170800,
      "step": 11200
    },
    {
      "epoch": 0.3584917738941169,
      "eval_loss": 4.487894058227539,
      "eval_runtime": 487.6341,
      "eval_samples_per_second": 28.476,
      "eval_steps_per_second": 14.238,
      "num_input_tokens_seen": 15170800,
      "step": 11200
    },
    {
      "epoch": 0.35865181486460534,
      "grad_norm": 52.5,
      "learning_rate": 4.092954736056897e-05,
      "loss": 4.37,
      "num_input_tokens_seen": 15177744,
      "step": 11205
    },
    {
      "epoch": 0.3588118558350938,
      "grad_norm": 40.25,
      "learning_rate": 4.09219796708209e-05,
      "loss": 3.5599,
      "num_input_tokens_seen": 15184368,
      "step": 11210
    },
    {
      "epoch": 0.3589718968055822,
      "grad_norm": 36.75,
      "learning_rate": 4.0914409525703464e-05,
      "loss": 5.0515,
      "num_input_tokens_seen": 15191296,
      "step": 11215
    },
    {
      "epoch": 0.35913193777607066,
      "grad_norm": 24.125,
      "learning_rate": 4.090683692638408e-05,
      "loss": 5.4642,
      "num_input_tokens_seen": 15197776,
      "step": 11220
    },
    {
      "epoch": 0.3592919787465591,
      "grad_norm": 38.75,
      "learning_rate": 4.089926187403056e-05,
      "loss": 4.7195,
      "num_input_tokens_seen": 15204512,
      "step": 11225
    },
    {
      "epoch": 0.35945201971704754,
      "grad_norm": 29.0,
      "learning_rate": 4.0891684369811044e-05,
      "loss": 3.9507,
      "num_input_tokens_seen": 15210768,
      "step": 11230
    },
    {
      "epoch": 0.359612060687536,
      "grad_norm": 13.6875,
      "learning_rate": 4.0884104414894107e-05,
      "loss": 3.461,
      "num_input_tokens_seen": 15217936,
      "step": 11235
    },
    {
      "epoch": 0.3597721016580245,
      "grad_norm": 34.25,
      "learning_rate": 4.087652201044864e-05,
      "loss": 4.8599,
      "num_input_tokens_seen": 15224528,
      "step": 11240
    },
    {
      "epoch": 0.3599321426285129,
      "grad_norm": 26.0,
      "learning_rate": 4.086893715764397e-05,
      "loss": 5.0471,
      "num_input_tokens_seen": 15231440,
      "step": 11245
    },
    {
      "epoch": 0.36009218359900136,
      "grad_norm": 32.0,
      "learning_rate": 4.086134985764977e-05,
      "loss": 4.4996,
      "num_input_tokens_seen": 15238096,
      "step": 11250
    },
    {
      "epoch": 0.3602522245694898,
      "grad_norm": 52.5,
      "learning_rate": 4.0853760111636085e-05,
      "loss": 4.2011,
      "num_input_tokens_seen": 15244640,
      "step": 11255
    },
    {
      "epoch": 0.36041226553997824,
      "grad_norm": 29.875,
      "learning_rate": 4.084616792077337e-05,
      "loss": 4.2182,
      "num_input_tokens_seen": 15251216,
      "step": 11260
    },
    {
      "epoch": 0.3605723065104667,
      "grad_norm": 42.75,
      "learning_rate": 4.083857328623243e-05,
      "loss": 4.4081,
      "num_input_tokens_seen": 15258432,
      "step": 11265
    },
    {
      "epoch": 0.3607323474809551,
      "grad_norm": 29.5,
      "learning_rate": 4.083097620918444e-05,
      "loss": 4.3564,
      "num_input_tokens_seen": 15265200,
      "step": 11270
    },
    {
      "epoch": 0.36089238845144356,
      "grad_norm": 19.625,
      "learning_rate": 4.082337669080097e-05,
      "loss": 4.0967,
      "num_input_tokens_seen": 15271680,
      "step": 11275
    },
    {
      "epoch": 0.361052429421932,
      "grad_norm": 33.75,
      "learning_rate": 4.081577473225398e-05,
      "loss": 4.1067,
      "num_input_tokens_seen": 15278384,
      "step": 11280
    },
    {
      "epoch": 0.36121247039242044,
      "grad_norm": 14.875,
      "learning_rate": 4.080817033471577e-05,
      "loss": 3.9496,
      "num_input_tokens_seen": 15285248,
      "step": 11285
    },
    {
      "epoch": 0.3613725113629089,
      "grad_norm": 29.875,
      "learning_rate": 4.080056349935903e-05,
      "loss": 3.8458,
      "num_input_tokens_seen": 15292256,
      "step": 11290
    },
    {
      "epoch": 0.3615325523333974,
      "grad_norm": 31.125,
      "learning_rate": 4.079295422735684e-05,
      "loss": 4.323,
      "num_input_tokens_seen": 15298800,
      "step": 11295
    },
    {
      "epoch": 0.3616925933038858,
      "grad_norm": 26.125,
      "learning_rate": 4.078534251988264e-05,
      "loss": 4.6183,
      "num_input_tokens_seen": 15305920,
      "step": 11300
    },
    {
      "epoch": 0.36185263427437425,
      "grad_norm": 26.25,
      "learning_rate": 4.077772837811025e-05,
      "loss": 4.2408,
      "num_input_tokens_seen": 15312544,
      "step": 11305
    },
    {
      "epoch": 0.3620126752448627,
      "grad_norm": 17.25,
      "learning_rate": 4.0770111803213874e-05,
      "loss": 4.1631,
      "num_input_tokens_seen": 15319344,
      "step": 11310
    },
    {
      "epoch": 0.36217271621535113,
      "grad_norm": 23.875,
      "learning_rate": 4.076249279636807e-05,
      "loss": 4.6532,
      "num_input_tokens_seen": 15326192,
      "step": 11315
    },
    {
      "epoch": 0.36233275718583957,
      "grad_norm": 16.125,
      "learning_rate": 4.075487135874781e-05,
      "loss": 4.2461,
      "num_input_tokens_seen": 15332480,
      "step": 11320
    },
    {
      "epoch": 0.362492798156328,
      "grad_norm": 37.75,
      "learning_rate": 4.074724749152837e-05,
      "loss": 3.9842,
      "num_input_tokens_seen": 15339024,
      "step": 11325
    },
    {
      "epoch": 0.36265283912681645,
      "grad_norm": 33.5,
      "learning_rate": 4.07396211958855e-05,
      "loss": 5.2247,
      "num_input_tokens_seen": 15345584,
      "step": 11330
    },
    {
      "epoch": 0.3628128800973049,
      "grad_norm": 31.5,
      "learning_rate": 4.073199247299523e-05,
      "loss": 4.4711,
      "num_input_tokens_seen": 15352176,
      "step": 11335
    },
    {
      "epoch": 0.36297292106779333,
      "grad_norm": 31.75,
      "learning_rate": 4.072436132403403e-05,
      "loss": 3.8989,
      "num_input_tokens_seen": 15359008,
      "step": 11340
    },
    {
      "epoch": 0.3631329620382818,
      "grad_norm": 27.5,
      "learning_rate": 4.0716727750178704e-05,
      "loss": 3.6143,
      "num_input_tokens_seen": 15366096,
      "step": 11345
    },
    {
      "epoch": 0.36329300300877027,
      "grad_norm": 26.25,
      "learning_rate": 4.0709091752606455e-05,
      "loss": 4.105,
      "num_input_tokens_seen": 15373312,
      "step": 11350
    },
    {
      "epoch": 0.3634530439792587,
      "grad_norm": 34.0,
      "learning_rate": 4.070145333249484e-05,
      "loss": 4.8544,
      "num_input_tokens_seen": 15379856,
      "step": 11355
    },
    {
      "epoch": 0.36361308494974715,
      "grad_norm": 27.75,
      "learning_rate": 4.069381249102181e-05,
      "loss": 4.0811,
      "num_input_tokens_seen": 15386880,
      "step": 11360
    },
    {
      "epoch": 0.3637731259202356,
      "grad_norm": 37.5,
      "learning_rate": 4.0686169229365665e-05,
      "loss": 4.4374,
      "num_input_tokens_seen": 15393392,
      "step": 11365
    },
    {
      "epoch": 0.363933166890724,
      "grad_norm": 37.0,
      "learning_rate": 4.067852354870511e-05,
      "loss": 4.6835,
      "num_input_tokens_seen": 15400176,
      "step": 11370
    },
    {
      "epoch": 0.36409320786121246,
      "grad_norm": 29.875,
      "learning_rate": 4.067087545021919e-05,
      "loss": 4.275,
      "num_input_tokens_seen": 15406656,
      "step": 11375
    },
    {
      "epoch": 0.3642532488317009,
      "grad_norm": 25.5,
      "learning_rate": 4.066322493508734e-05,
      "loss": 4.6654,
      "num_input_tokens_seen": 15413296,
      "step": 11380
    },
    {
      "epoch": 0.36441328980218934,
      "grad_norm": 20.25,
      "learning_rate": 4.065557200448937e-05,
      "loss": 3.2861,
      "num_input_tokens_seen": 15419888,
      "step": 11385
    },
    {
      "epoch": 0.3645733307726778,
      "grad_norm": 36.0,
      "learning_rate": 4.064791665960546e-05,
      "loss": 5.1984,
      "num_input_tokens_seen": 15427536,
      "step": 11390
    },
    {
      "epoch": 0.3647333717431662,
      "grad_norm": 18.75,
      "learning_rate": 4.064025890161615e-05,
      "loss": 3.803,
      "num_input_tokens_seen": 15433920,
      "step": 11395
    },
    {
      "epoch": 0.3648934127136547,
      "grad_norm": 25.5,
      "learning_rate": 4.0632598731702373e-05,
      "loss": 4.8716,
      "num_input_tokens_seen": 15440592,
      "step": 11400
    },
    {
      "epoch": 0.3648934127136547,
      "eval_loss": 4.501139163970947,
      "eval_runtime": 488.0206,
      "eval_samples_per_second": 28.454,
      "eval_steps_per_second": 14.227,
      "num_input_tokens_seen": 15440592,
      "step": 11400
    },
    {
      "epoch": 0.36505345368414316,
      "grad_norm": 42.5,
      "learning_rate": 4.0624936151045426e-05,
      "loss": 4.4883,
      "num_input_tokens_seen": 15447136,
      "step": 11405
    },
    {
      "epoch": 0.3652134946546316,
      "grad_norm": 33.0,
      "learning_rate": 4.061727116082696e-05,
      "loss": 4.8656,
      "num_input_tokens_seen": 15453952,
      "step": 11410
    },
    {
      "epoch": 0.36537353562512004,
      "grad_norm": 50.0,
      "learning_rate": 4.060960376222903e-05,
      "loss": 5.2272,
      "num_input_tokens_seen": 15461168,
      "step": 11415
    },
    {
      "epoch": 0.3655335765956085,
      "grad_norm": 30.75,
      "learning_rate": 4.0601933956434034e-05,
      "loss": 4.1602,
      "num_input_tokens_seen": 15468464,
      "step": 11420
    },
    {
      "epoch": 0.3656936175660969,
      "grad_norm": 38.0,
      "learning_rate": 4.059426174462476e-05,
      "loss": 3.8781,
      "num_input_tokens_seen": 15474896,
      "step": 11425
    },
    {
      "epoch": 0.36585365853658536,
      "grad_norm": 24.0,
      "learning_rate": 4.058658712798435e-05,
      "loss": 4.2032,
      "num_input_tokens_seen": 15481616,
      "step": 11430
    },
    {
      "epoch": 0.3660136995070738,
      "grad_norm": 27.625,
      "learning_rate": 4.0578910107696336e-05,
      "loss": 4.0913,
      "num_input_tokens_seen": 15488624,
      "step": 11435
    },
    {
      "epoch": 0.36617374047756224,
      "grad_norm": 35.5,
      "learning_rate": 4.05712306849446e-05,
      "loss": 4.4821,
      "num_input_tokens_seen": 15494960,
      "step": 11440
    },
    {
      "epoch": 0.3663337814480507,
      "grad_norm": 24.25,
      "learning_rate": 4.0563548860913415e-05,
      "loss": 3.909,
      "num_input_tokens_seen": 15501568,
      "step": 11445
    },
    {
      "epoch": 0.3664938224185392,
      "grad_norm": 18.75,
      "learning_rate": 4.0555864636787414e-05,
      "loss": 3.6129,
      "num_input_tokens_seen": 15508416,
      "step": 11450
    },
    {
      "epoch": 0.3666538633890276,
      "grad_norm": 29.0,
      "learning_rate": 4.054817801375159e-05,
      "loss": 5.1117,
      "num_input_tokens_seen": 15515424,
      "step": 11455
    },
    {
      "epoch": 0.36681390435951605,
      "grad_norm": 37.0,
      "learning_rate": 4.054048899299134e-05,
      "loss": 5.1162,
      "num_input_tokens_seen": 15522192,
      "step": 11460
    },
    {
      "epoch": 0.3669739453300045,
      "grad_norm": 32.0,
      "learning_rate": 4.0532797575692385e-05,
      "loss": 4.4252,
      "num_input_tokens_seen": 15529008,
      "step": 11465
    },
    {
      "epoch": 0.36713398630049293,
      "grad_norm": 31.375,
      "learning_rate": 4.052510376304085e-05,
      "loss": 4.0713,
      "num_input_tokens_seen": 15536112,
      "step": 11470
    },
    {
      "epoch": 0.36729402727098137,
      "grad_norm": 17.75,
      "learning_rate": 4.051740755622321e-05,
      "loss": 4.5844,
      "num_input_tokens_seen": 15542544,
      "step": 11475
    },
    {
      "epoch": 0.3674540682414698,
      "grad_norm": 38.75,
      "learning_rate": 4.050970895642632e-05,
      "loss": 5.0277,
      "num_input_tokens_seen": 15549248,
      "step": 11480
    },
    {
      "epoch": 0.36761410921195825,
      "grad_norm": 22.125,
      "learning_rate": 4.050200796483741e-05,
      "loss": 4.2507,
      "num_input_tokens_seen": 15556512,
      "step": 11485
    },
    {
      "epoch": 0.3677741501824467,
      "grad_norm": 20.75,
      "learning_rate": 4.049430458264405e-05,
      "loss": 4.2612,
      "num_input_tokens_seen": 15563296,
      "step": 11490
    },
    {
      "epoch": 0.36793419115293513,
      "grad_norm": 32.5,
      "learning_rate": 4.048659881103422e-05,
      "loss": 4.1794,
      "num_input_tokens_seen": 15569952,
      "step": 11495
    },
    {
      "epoch": 0.36809423212342357,
      "grad_norm": 22.5,
      "learning_rate": 4.0478890651196235e-05,
      "loss": 4.3749,
      "num_input_tokens_seen": 15576656,
      "step": 11500
    },
    {
      "epoch": 0.36825427309391207,
      "grad_norm": 47.0,
      "learning_rate": 4.047118010431879e-05,
      "loss": 4.3978,
      "num_input_tokens_seen": 15583104,
      "step": 11505
    },
    {
      "epoch": 0.3684143140644005,
      "grad_norm": 31.5,
      "learning_rate": 4.046346717159094e-05,
      "loss": 3.8064,
      "num_input_tokens_seen": 15590064,
      "step": 11510
    },
    {
      "epoch": 0.36857435503488895,
      "grad_norm": 39.25,
      "learning_rate": 4.045575185420214e-05,
      "loss": 4.7843,
      "num_input_tokens_seen": 15596816,
      "step": 11515
    },
    {
      "epoch": 0.3687343960053774,
      "grad_norm": 18.625,
      "learning_rate": 4.0448034153342165e-05,
      "loss": 4.7845,
      "num_input_tokens_seen": 15602992,
      "step": 11520
    },
    {
      "epoch": 0.3688944369758658,
      "grad_norm": 51.5,
      "learning_rate": 4.0440314070201194e-05,
      "loss": 3.6832,
      "num_input_tokens_seen": 15609344,
      "step": 11525
    },
    {
      "epoch": 0.36905447794635426,
      "grad_norm": 28.875,
      "learning_rate": 4.043259160596976e-05,
      "loss": 4.6318,
      "num_input_tokens_seen": 15615504,
      "step": 11530
    },
    {
      "epoch": 0.3692145189168427,
      "grad_norm": 32.0,
      "learning_rate": 4.0424866761838767e-05,
      "loss": 5.4511,
      "num_input_tokens_seen": 15621968,
      "step": 11535
    },
    {
      "epoch": 0.36937455988733114,
      "grad_norm": 34.5,
      "learning_rate": 4.041713953899948e-05,
      "loss": 4.204,
      "num_input_tokens_seen": 15628832,
      "step": 11540
    },
    {
      "epoch": 0.3695346008578196,
      "grad_norm": 34.25,
      "learning_rate": 4.0409409938643515e-05,
      "loss": 4.7347,
      "num_input_tokens_seen": 15635552,
      "step": 11545
    },
    {
      "epoch": 0.369694641828308,
      "grad_norm": 28.75,
      "learning_rate": 4.0401677961962904e-05,
      "loss": 4.3147,
      "num_input_tokens_seen": 15642368,
      "step": 11550
    },
    {
      "epoch": 0.3698546827987965,
      "grad_norm": 19.375,
      "learning_rate": 4.039394361015001e-05,
      "loss": 3.6787,
      "num_input_tokens_seen": 15648928,
      "step": 11555
    },
    {
      "epoch": 0.37001472376928496,
      "grad_norm": 22.0,
      "learning_rate": 4.038620688439755e-05,
      "loss": 3.9346,
      "num_input_tokens_seen": 15655632,
      "step": 11560
    },
    {
      "epoch": 0.3701747647397734,
      "grad_norm": 24.125,
      "learning_rate": 4.037846778589862e-05,
      "loss": 4.1953,
      "num_input_tokens_seen": 15662496,
      "step": 11565
    },
    {
      "epoch": 0.37033480571026184,
      "grad_norm": 20.25,
      "learning_rate": 4.0370726315846715e-05,
      "loss": 3.5558,
      "num_input_tokens_seen": 15669248,
      "step": 11570
    },
    {
      "epoch": 0.3704948466807503,
      "grad_norm": 31.75,
      "learning_rate": 4.036298247543565e-05,
      "loss": 4.3782,
      "num_input_tokens_seen": 15675776,
      "step": 11575
    },
    {
      "epoch": 0.3706548876512387,
      "grad_norm": 28.625,
      "learning_rate": 4.035523626585962e-05,
      "loss": 4.6409,
      "num_input_tokens_seen": 15682912,
      "step": 11580
    },
    {
      "epoch": 0.37081492862172716,
      "grad_norm": 23.875,
      "learning_rate": 4.0347487688313194e-05,
      "loss": 5.0929,
      "num_input_tokens_seen": 15689520,
      "step": 11585
    },
    {
      "epoch": 0.3709749695922156,
      "grad_norm": 37.25,
      "learning_rate": 4.0339736743991296e-05,
      "loss": 4.2945,
      "num_input_tokens_seen": 15696224,
      "step": 11590
    },
    {
      "epoch": 0.37113501056270404,
      "grad_norm": 42.0,
      "learning_rate": 4.0331983434089227e-05,
      "loss": 3.3706,
      "num_input_tokens_seen": 15703376,
      "step": 11595
    },
    {
      "epoch": 0.3712950515331925,
      "grad_norm": 35.0,
      "learning_rate": 4.032422775980264e-05,
      "loss": 6.5712,
      "num_input_tokens_seen": 15710608,
      "step": 11600
    },
    {
      "epoch": 0.3712950515331925,
      "eval_loss": 4.4706292152404785,
      "eval_runtime": 488.6708,
      "eval_samples_per_second": 28.416,
      "eval_steps_per_second": 14.208,
      "num_input_tokens_seen": 15710608,
      "step": 11600
    },
    {
      "epoch": 0.3714550925036809,
      "grad_norm": 21.5,
      "learning_rate": 4.031646972232754e-05,
      "loss": 3.9519,
      "num_input_tokens_seen": 15717184,
      "step": 11605
    },
    {
      "epoch": 0.3716151334741694,
      "grad_norm": 30.0,
      "learning_rate": 4.0308709322860344e-05,
      "loss": 4.4013,
      "num_input_tokens_seen": 15723936,
      "step": 11610
    },
    {
      "epoch": 0.37177517444465785,
      "grad_norm": 27.25,
      "learning_rate": 4.0300946562597784e-05,
      "loss": 4.5522,
      "num_input_tokens_seen": 15730912,
      "step": 11615
    },
    {
      "epoch": 0.3719352154151463,
      "grad_norm": 23.75,
      "learning_rate": 4.029318144273698e-05,
      "loss": 4.1371,
      "num_input_tokens_seen": 15737376,
      "step": 11620
    },
    {
      "epoch": 0.37209525638563473,
      "grad_norm": 28.25,
      "learning_rate": 4.0285413964475415e-05,
      "loss": 3.9582,
      "num_input_tokens_seen": 15744080,
      "step": 11625
    },
    {
      "epoch": 0.37225529735612317,
      "grad_norm": 21.25,
      "learning_rate": 4.0277644129010927e-05,
      "loss": 4.958,
      "num_input_tokens_seen": 15750784,
      "step": 11630
    },
    {
      "epoch": 0.3724153383266116,
      "grad_norm": 23.875,
      "learning_rate": 4.0269871937541724e-05,
      "loss": 4.3193,
      "num_input_tokens_seen": 15757520,
      "step": 11635
    },
    {
      "epoch": 0.37257537929710005,
      "grad_norm": 30.375,
      "learning_rate": 4.026209739126637e-05,
      "loss": 3.9897,
      "num_input_tokens_seen": 15764128,
      "step": 11640
    },
    {
      "epoch": 0.3727354202675885,
      "grad_norm": 53.25,
      "learning_rate": 4.025432049138381e-05,
      "loss": 4.4382,
      "num_input_tokens_seen": 15770944,
      "step": 11645
    },
    {
      "epoch": 0.37289546123807693,
      "grad_norm": 30.0,
      "learning_rate": 4.0246541239093325e-05,
      "loss": 4.3669,
      "num_input_tokens_seen": 15777776,
      "step": 11650
    },
    {
      "epoch": 0.37305550220856537,
      "grad_norm": 27.0,
      "learning_rate": 4.023875963559459e-05,
      "loss": 4.9737,
      "num_input_tokens_seen": 15784608,
      "step": 11655
    },
    {
      "epoch": 0.3732155431790538,
      "grad_norm": 49.0,
      "learning_rate": 4.023097568208761e-05,
      "loss": 5.7416,
      "num_input_tokens_seen": 15791328,
      "step": 11660
    },
    {
      "epoch": 0.3733755841495423,
      "grad_norm": 24.0,
      "learning_rate": 4.022318937977277e-05,
      "loss": 4.4528,
      "num_input_tokens_seen": 15798224,
      "step": 11665
    },
    {
      "epoch": 0.37353562512003075,
      "grad_norm": 24.25,
      "learning_rate": 4.021540072985084e-05,
      "loss": 4.1971,
      "num_input_tokens_seen": 15804672,
      "step": 11670
    },
    {
      "epoch": 0.3736956660905192,
      "grad_norm": 19.5,
      "learning_rate": 4.020760973352289e-05,
      "loss": 4.3883,
      "num_input_tokens_seen": 15811760,
      "step": 11675
    },
    {
      "epoch": 0.3738557070610076,
      "grad_norm": 30.25,
      "learning_rate": 4.019981639199042e-05,
      "loss": 5.2803,
      "num_input_tokens_seen": 15818000,
      "step": 11680
    },
    {
      "epoch": 0.37401574803149606,
      "grad_norm": 37.0,
      "learning_rate": 4.0192020706455245e-05,
      "loss": 4.5337,
      "num_input_tokens_seen": 15824784,
      "step": 11685
    },
    {
      "epoch": 0.3741757890019845,
      "grad_norm": 37.25,
      "learning_rate": 4.018422267811956e-05,
      "loss": 7.0054,
      "num_input_tokens_seen": 15831136,
      "step": 11690
    },
    {
      "epoch": 0.37433582997247294,
      "grad_norm": 29.375,
      "learning_rate": 4.017642230818592e-05,
      "loss": 4.5322,
      "num_input_tokens_seen": 15838160,
      "step": 11695
    },
    {
      "epoch": 0.3744958709429614,
      "grad_norm": 40.75,
      "learning_rate": 4.0168619597857246e-05,
      "loss": 4.5625,
      "num_input_tokens_seen": 15844992,
      "step": 11700
    },
    {
      "epoch": 0.3746559119134498,
      "grad_norm": 34.75,
      "learning_rate": 4.016081454833681e-05,
      "loss": 4.3614,
      "num_input_tokens_seen": 15851536,
      "step": 11705
    },
    {
      "epoch": 0.37481595288393826,
      "grad_norm": 19.875,
      "learning_rate": 4.0153007160828245e-05,
      "loss": 4.1328,
      "num_input_tokens_seen": 15858432,
      "step": 11710
    },
    {
      "epoch": 0.37497599385442676,
      "grad_norm": 36.5,
      "learning_rate": 4.0145197436535555e-05,
      "loss": 5.3379,
      "num_input_tokens_seen": 15865104,
      "step": 11715
    },
    {
      "epoch": 0.3751360348249152,
      "grad_norm": 38.0,
      "learning_rate": 4.0137385376663095e-05,
      "loss": 4.4856,
      "num_input_tokens_seen": 15871856,
      "step": 11720
    },
    {
      "epoch": 0.37529607579540364,
      "grad_norm": 28.625,
      "learning_rate": 4.012957098241558e-05,
      "loss": 4.3862,
      "num_input_tokens_seen": 15878976,
      "step": 11725
    },
    {
      "epoch": 0.3754561167658921,
      "grad_norm": 31.375,
      "learning_rate": 4.0121754254998076e-05,
      "loss": 3.9561,
      "num_input_tokens_seen": 15885584,
      "step": 11730
    },
    {
      "epoch": 0.3756161577363805,
      "grad_norm": 47.25,
      "learning_rate": 4.011393519561606e-05,
      "loss": 4.4646,
      "num_input_tokens_seen": 15892336,
      "step": 11735
    },
    {
      "epoch": 0.37577619870686896,
      "grad_norm": 32.0,
      "learning_rate": 4.010611380547529e-05,
      "loss": 3.8593,
      "num_input_tokens_seen": 15899104,
      "step": 11740
    },
    {
      "epoch": 0.3759362396773574,
      "grad_norm": 33.75,
      "learning_rate": 4.009829008578192e-05,
      "loss": 4.6117,
      "num_input_tokens_seen": 15905760,
      "step": 11745
    },
    {
      "epoch": 0.37609628064784584,
      "grad_norm": 24.375,
      "learning_rate": 4.00904640377425e-05,
      "loss": 4.5442,
      "num_input_tokens_seen": 15912496,
      "step": 11750
    },
    {
      "epoch": 0.3762563216183343,
      "grad_norm": 21.5,
      "learning_rate": 4.0082635662563886e-05,
      "loss": 4.382,
      "num_input_tokens_seen": 15919568,
      "step": 11755
    },
    {
      "epoch": 0.3764163625888227,
      "grad_norm": 26.25,
      "learning_rate": 4.007480496145331e-05,
      "loss": 5.5185,
      "num_input_tokens_seen": 15926720,
      "step": 11760
    },
    {
      "epoch": 0.37657640355931116,
      "grad_norm": 21.625,
      "learning_rate": 4.006697193561837e-05,
      "loss": 4.4024,
      "num_input_tokens_seen": 15933264,
      "step": 11765
    },
    {
      "epoch": 0.37673644452979965,
      "grad_norm": 24.125,
      "learning_rate": 4.005913658626701e-05,
      "loss": 5.3652,
      "num_input_tokens_seen": 15939808,
      "step": 11770
    },
    {
      "epoch": 0.3768964855002881,
      "grad_norm": 38.25,
      "learning_rate": 4.005129891460754e-05,
      "loss": 4.1553,
      "num_input_tokens_seen": 15946352,
      "step": 11775
    },
    {
      "epoch": 0.37705652647077653,
      "grad_norm": 33.5,
      "learning_rate": 4.004345892184864e-05,
      "loss": 2.8058,
      "num_input_tokens_seen": 15953312,
      "step": 11780
    },
    {
      "epoch": 0.37721656744126497,
      "grad_norm": 24.75,
      "learning_rate": 4.003561660919932e-05,
      "loss": 4.6632,
      "num_input_tokens_seen": 15960112,
      "step": 11785
    },
    {
      "epoch": 0.3773766084117534,
      "grad_norm": 20.75,
      "learning_rate": 4.002777197786897e-05,
      "loss": 3.5035,
      "num_input_tokens_seen": 15966784,
      "step": 11790
    },
    {
      "epoch": 0.37753664938224185,
      "grad_norm": 28.0,
      "learning_rate": 4.0019925029067326e-05,
      "loss": 4.8026,
      "num_input_tokens_seen": 15973488,
      "step": 11795
    },
    {
      "epoch": 0.3776966903527303,
      "grad_norm": 27.625,
      "learning_rate": 4.0012075764004495e-05,
      "loss": 4.0938,
      "num_input_tokens_seen": 15980176,
      "step": 11800
    },
    {
      "epoch": 0.3776966903527303,
      "eval_loss": 4.476372718811035,
      "eval_runtime": 487.799,
      "eval_samples_per_second": 28.467,
      "eval_steps_per_second": 14.233,
      "num_input_tokens_seen": 15980176,
      "step": 11800
    },
    {
      "epoch": 0.37785673132321873,
      "grad_norm": 45.25,
      "learning_rate": 4.000422418389094e-05,
      "loss": 3.4288,
      "num_input_tokens_seen": 15986800,
      "step": 11805
    },
    {
      "epoch": 0.37801677229370717,
      "grad_norm": 28.875,
      "learning_rate": 3.999637028993744e-05,
      "loss": 4.6669,
      "num_input_tokens_seen": 15993840,
      "step": 11810
    },
    {
      "epoch": 0.3781768132641956,
      "grad_norm": 26.25,
      "learning_rate": 3.99885140833552e-05,
      "loss": 4.3674,
      "num_input_tokens_seen": 16000224,
      "step": 11815
    },
    {
      "epoch": 0.3783368542346841,
      "grad_norm": 21.125,
      "learning_rate": 3.998065556535572e-05,
      "loss": 4.0593,
      "num_input_tokens_seen": 16006752,
      "step": 11820
    },
    {
      "epoch": 0.37849689520517255,
      "grad_norm": 20.375,
      "learning_rate": 3.9972794737150895e-05,
      "loss": 4.2327,
      "num_input_tokens_seen": 16013552,
      "step": 11825
    },
    {
      "epoch": 0.378656936175661,
      "grad_norm": 28.375,
      "learning_rate": 3.996493159995297e-05,
      "loss": 4.6567,
      "num_input_tokens_seen": 16020240,
      "step": 11830
    },
    {
      "epoch": 0.3788169771461494,
      "grad_norm": 26.125,
      "learning_rate": 3.995706615497453e-05,
      "loss": 4.4372,
      "num_input_tokens_seen": 16026896,
      "step": 11835
    },
    {
      "epoch": 0.37897701811663786,
      "grad_norm": 25.5,
      "learning_rate": 3.994919840342852e-05,
      "loss": 4.3887,
      "num_input_tokens_seen": 16033840,
      "step": 11840
    },
    {
      "epoch": 0.3791370590871263,
      "grad_norm": 31.25,
      "learning_rate": 3.994132834652825e-05,
      "loss": 4.2508,
      "num_input_tokens_seen": 16040784,
      "step": 11845
    },
    {
      "epoch": 0.37929710005761474,
      "grad_norm": 30.75,
      "learning_rate": 3.99334559854874e-05,
      "loss": 4.6448,
      "num_input_tokens_seen": 16047648,
      "step": 11850
    },
    {
      "epoch": 0.3794571410281032,
      "grad_norm": 27.875,
      "learning_rate": 3.9925581321519955e-05,
      "loss": 5.3571,
      "num_input_tokens_seen": 16054272,
      "step": 11855
    },
    {
      "epoch": 0.3796171819985916,
      "grad_norm": 30.0,
      "learning_rate": 3.991770435584031e-05,
      "loss": 4.0412,
      "num_input_tokens_seen": 16060896,
      "step": 11860
    },
    {
      "epoch": 0.37977722296908006,
      "grad_norm": 34.75,
      "learning_rate": 3.990982508966319e-05,
      "loss": 3.8498,
      "num_input_tokens_seen": 16067536,
      "step": 11865
    },
    {
      "epoch": 0.3799372639395685,
      "grad_norm": 42.5,
      "learning_rate": 3.990194352420367e-05,
      "loss": 4.7805,
      "num_input_tokens_seen": 16074288,
      "step": 11870
    },
    {
      "epoch": 0.380097304910057,
      "grad_norm": 33.0,
      "learning_rate": 3.9894059660677184e-05,
      "loss": 4.4444,
      "num_input_tokens_seen": 16080640,
      "step": 11875
    },
    {
      "epoch": 0.38025734588054544,
      "grad_norm": 36.25,
      "learning_rate": 3.9886173500299526e-05,
      "loss": 4.7366,
      "num_input_tokens_seen": 16087472,
      "step": 11880
    },
    {
      "epoch": 0.3804173868510339,
      "grad_norm": 40.0,
      "learning_rate": 3.987828504428685e-05,
      "loss": 5.0825,
      "num_input_tokens_seen": 16094048,
      "step": 11885
    },
    {
      "epoch": 0.3805774278215223,
      "grad_norm": 23.5,
      "learning_rate": 3.987039429385565e-05,
      "loss": 4.3544,
      "num_input_tokens_seen": 16100784,
      "step": 11890
    },
    {
      "epoch": 0.38073746879201076,
      "grad_norm": 40.75,
      "learning_rate": 3.986250125022277e-05,
      "loss": 4.0187,
      "num_input_tokens_seen": 16107696,
      "step": 11895
    },
    {
      "epoch": 0.3808975097624992,
      "grad_norm": 25.0,
      "learning_rate": 3.985460591460544e-05,
      "loss": 4.3942,
      "num_input_tokens_seen": 16113920,
      "step": 11900
    },
    {
      "epoch": 0.38105755073298764,
      "grad_norm": 31.375,
      "learning_rate": 3.984670828822118e-05,
      "loss": 3.7935,
      "num_input_tokens_seen": 16120944,
      "step": 11905
    },
    {
      "epoch": 0.3812175917034761,
      "grad_norm": 46.5,
      "learning_rate": 3.983880837228794e-05,
      "loss": 3.9475,
      "num_input_tokens_seen": 16128176,
      "step": 11910
    },
    {
      "epoch": 0.3813776326739645,
      "grad_norm": 29.875,
      "learning_rate": 3.983090616802396e-05,
      "loss": 4.591,
      "num_input_tokens_seen": 16135216,
      "step": 11915
    },
    {
      "epoch": 0.38153767364445296,
      "grad_norm": 33.5,
      "learning_rate": 3.982300167664788e-05,
      "loss": 3.6585,
      "num_input_tokens_seen": 16141712,
      "step": 11920
    },
    {
      "epoch": 0.38169771461494145,
      "grad_norm": 43.5,
      "learning_rate": 3.981509489937868e-05,
      "loss": 3.9593,
      "num_input_tokens_seen": 16148208,
      "step": 11925
    },
    {
      "epoch": 0.3818577555854299,
      "grad_norm": 29.25,
      "learning_rate": 3.9807185837435643e-05,
      "loss": 4.7096,
      "num_input_tokens_seen": 16154992,
      "step": 11930
    },
    {
      "epoch": 0.38201779655591833,
      "grad_norm": 23.75,
      "learning_rate": 3.9799274492038484e-05,
      "loss": 5.0491,
      "num_input_tokens_seen": 16161328,
      "step": 11935
    },
    {
      "epoch": 0.38217783752640677,
      "grad_norm": 28.625,
      "learning_rate": 3.979136086440722e-05,
      "loss": 4.007,
      "num_input_tokens_seen": 16168272,
      "step": 11940
    },
    {
      "epoch": 0.3823378784968952,
      "grad_norm": 40.25,
      "learning_rate": 3.9783444955762226e-05,
      "loss": 4.5433,
      "num_input_tokens_seen": 16175552,
      "step": 11945
    },
    {
      "epoch": 0.38249791946738365,
      "grad_norm": 22.625,
      "learning_rate": 3.977552676732424e-05,
      "loss": 3.7173,
      "num_input_tokens_seen": 16182496,
      "step": 11950
    },
    {
      "epoch": 0.3826579604378721,
      "grad_norm": 24.75,
      "learning_rate": 3.976760630031435e-05,
      "loss": 3.4654,
      "num_input_tokens_seen": 16189040,
      "step": 11955
    },
    {
      "epoch": 0.38281800140836053,
      "grad_norm": 24.125,
      "learning_rate": 3.975968355595398e-05,
      "loss": 4.4816,
      "num_input_tokens_seen": 16195664,
      "step": 11960
    },
    {
      "epoch": 0.38297804237884897,
      "grad_norm": 21.125,
      "learning_rate": 3.9751758535464935e-05,
      "loss": 3.6809,
      "num_input_tokens_seen": 16202192,
      "step": 11965
    },
    {
      "epoch": 0.3831380833493374,
      "grad_norm": 17.0,
      "learning_rate": 3.9743831240069326e-05,
      "loss": 3.708,
      "num_input_tokens_seen": 16209104,
      "step": 11970
    },
    {
      "epoch": 0.38329812431982585,
      "grad_norm": 29.0,
      "learning_rate": 3.9735901670989675e-05,
      "loss": 4.1333,
      "num_input_tokens_seen": 16215504,
      "step": 11975
    },
    {
      "epoch": 0.38345816529031435,
      "grad_norm": 68.5,
      "learning_rate": 3.97279698294488e-05,
      "loss": 4.7064,
      "num_input_tokens_seen": 16222240,
      "step": 11980
    },
    {
      "epoch": 0.3836182062608028,
      "grad_norm": 21.625,
      "learning_rate": 3.9720035716669876e-05,
      "loss": 4.7302,
      "num_input_tokens_seen": 16229040,
      "step": 11985
    },
    {
      "epoch": 0.3837782472312912,
      "grad_norm": 39.25,
      "learning_rate": 3.9712099333876474e-05,
      "loss": 4.2398,
      "num_input_tokens_seen": 16235728,
      "step": 11990
    },
    {
      "epoch": 0.38393828820177966,
      "grad_norm": 33.0,
      "learning_rate": 3.9704160682292475e-05,
      "loss": 3.8947,
      "num_input_tokens_seen": 16242272,
      "step": 11995
    },
    {
      "epoch": 0.3840983291722681,
      "grad_norm": 18.5,
      "learning_rate": 3.9696219763142106e-05,
      "loss": 4.7581,
      "num_input_tokens_seen": 16249072,
      "step": 12000
    },
    {
      "epoch": 0.3840983291722681,
      "eval_loss": 4.492238521575928,
      "eval_runtime": 487.9563,
      "eval_samples_per_second": 28.457,
      "eval_steps_per_second": 14.229,
      "num_input_tokens_seen": 16249072,
      "step": 12000
    },
    {
      "epoch": 0.38425837014275654,
      "grad_norm": 42.75,
      "learning_rate": 3.968827657764997e-05,
      "loss": 3.9953,
      "num_input_tokens_seen": 16255728,
      "step": 12005
    },
    {
      "epoch": 0.384418411113245,
      "grad_norm": 24.5,
      "learning_rate": 3.9680331127041e-05,
      "loss": 4.9059,
      "num_input_tokens_seen": 16262432,
      "step": 12010
    },
    {
      "epoch": 0.3845784520837334,
      "grad_norm": 15.5625,
      "learning_rate": 3.9672383412540495e-05,
      "loss": 4.9785,
      "num_input_tokens_seen": 16268880,
      "step": 12015
    },
    {
      "epoch": 0.38473849305422186,
      "grad_norm": 23.625,
      "learning_rate": 3.966443343537407e-05,
      "loss": 4.574,
      "num_input_tokens_seen": 16275824,
      "step": 12020
    },
    {
      "epoch": 0.3848985340247103,
      "grad_norm": 33.25,
      "learning_rate": 3.965648119676772e-05,
      "loss": 3.6623,
      "num_input_tokens_seen": 16283024,
      "step": 12025
    },
    {
      "epoch": 0.38505857499519874,
      "grad_norm": 42.75,
      "learning_rate": 3.96485266979478e-05,
      "loss": 5.0792,
      "num_input_tokens_seen": 16289616,
      "step": 12030
    },
    {
      "epoch": 0.38521861596568724,
      "grad_norm": 39.25,
      "learning_rate": 3.9640569940140974e-05,
      "loss": 3.5283,
      "num_input_tokens_seen": 16296144,
      "step": 12035
    },
    {
      "epoch": 0.3853786569361757,
      "grad_norm": 20.0,
      "learning_rate": 3.963261092457428e-05,
      "loss": 4.0626,
      "num_input_tokens_seen": 16303200,
      "step": 12040
    },
    {
      "epoch": 0.3855386979066641,
      "grad_norm": 29.125,
      "learning_rate": 3.962464965247509e-05,
      "loss": 5.0934,
      "num_input_tokens_seen": 16310080,
      "step": 12045
    },
    {
      "epoch": 0.38569873887715256,
      "grad_norm": 20.5,
      "learning_rate": 3.9616686125071135e-05,
      "loss": 3.8787,
      "num_input_tokens_seen": 16316944,
      "step": 12050
    },
    {
      "epoch": 0.385858779847641,
      "grad_norm": 43.0,
      "learning_rate": 3.9608720343590506e-05,
      "loss": 3.1216,
      "num_input_tokens_seen": 16323856,
      "step": 12055
    },
    {
      "epoch": 0.38601882081812944,
      "grad_norm": 25.875,
      "learning_rate": 3.960075230926161e-05,
      "loss": 4.3714,
      "num_input_tokens_seen": 16330336,
      "step": 12060
    },
    {
      "epoch": 0.3861788617886179,
      "grad_norm": 19.375,
      "learning_rate": 3.959278202331322e-05,
      "loss": 3.8599,
      "num_input_tokens_seen": 16337280,
      "step": 12065
    },
    {
      "epoch": 0.3863389027591063,
      "grad_norm": 31.5,
      "learning_rate": 3.958480948697446e-05,
      "loss": 4.9855,
      "num_input_tokens_seen": 16344464,
      "step": 12070
    },
    {
      "epoch": 0.38649894372959476,
      "grad_norm": 24.0,
      "learning_rate": 3.95768347014748e-05,
      "loss": 4.6864,
      "num_input_tokens_seen": 16350848,
      "step": 12075
    },
    {
      "epoch": 0.3866589847000832,
      "grad_norm": 19.5,
      "learning_rate": 3.956885766804404e-05,
      "loss": 4.79,
      "num_input_tokens_seen": 16357600,
      "step": 12080
    },
    {
      "epoch": 0.3868190256705717,
      "grad_norm": 28.125,
      "learning_rate": 3.956087838791235e-05,
      "loss": 4.2172,
      "num_input_tokens_seen": 16364192,
      "step": 12085
    },
    {
      "epoch": 0.38697906664106013,
      "grad_norm": 32.25,
      "learning_rate": 3.955289686231022e-05,
      "loss": 4.0017,
      "num_input_tokens_seen": 16370944,
      "step": 12090
    },
    {
      "epoch": 0.38713910761154857,
      "grad_norm": 39.25,
      "learning_rate": 3.9544913092468504e-05,
      "loss": 5.0492,
      "num_input_tokens_seen": 16377280,
      "step": 12095
    },
    {
      "epoch": 0.387299148582037,
      "grad_norm": 33.0,
      "learning_rate": 3.9536927079618425e-05,
      "loss": 3.9817,
      "num_input_tokens_seen": 16383680,
      "step": 12100
    },
    {
      "epoch": 0.38745918955252545,
      "grad_norm": 23.875,
      "learning_rate": 3.9528938824991494e-05,
      "loss": 5.1637,
      "num_input_tokens_seen": 16390464,
      "step": 12105
    },
    {
      "epoch": 0.3876192305230139,
      "grad_norm": 37.0,
      "learning_rate": 3.952094832981962e-05,
      "loss": 4.3041,
      "num_input_tokens_seen": 16397120,
      "step": 12110
    },
    {
      "epoch": 0.38777927149350233,
      "grad_norm": 34.5,
      "learning_rate": 3.951295559533503e-05,
      "loss": 3.8975,
      "num_input_tokens_seen": 16403600,
      "step": 12115
    },
    {
      "epoch": 0.38793931246399077,
      "grad_norm": 16.75,
      "learning_rate": 3.95049606227703e-05,
      "loss": 3.8594,
      "num_input_tokens_seen": 16410144,
      "step": 12120
    },
    {
      "epoch": 0.3880993534344792,
      "grad_norm": 33.5,
      "learning_rate": 3.949696341335838e-05,
      "loss": 3.968,
      "num_input_tokens_seen": 16417184,
      "step": 12125
    },
    {
      "epoch": 0.38825939440496765,
      "grad_norm": 29.375,
      "learning_rate": 3.9488963968332503e-05,
      "loss": 4.0233,
      "num_input_tokens_seen": 16423520,
      "step": 12130
    },
    {
      "epoch": 0.3884194353754561,
      "grad_norm": 32.5,
      "learning_rate": 3.948096228892631e-05,
      "loss": 4.8553,
      "num_input_tokens_seen": 16430416,
      "step": 12135
    },
    {
      "epoch": 0.3885794763459446,
      "grad_norm": 29.5,
      "learning_rate": 3.947295837637375e-05,
      "loss": 4.4405,
      "num_input_tokens_seen": 16437040,
      "step": 12140
    },
    {
      "epoch": 0.388739517316433,
      "grad_norm": 28.375,
      "learning_rate": 3.9464952231909135e-05,
      "loss": 5.5143,
      "num_input_tokens_seen": 16443648,
      "step": 12145
    },
    {
      "epoch": 0.38889955828692147,
      "grad_norm": 25.625,
      "learning_rate": 3.945694385676711e-05,
      "loss": 4.1515,
      "num_input_tokens_seen": 16450624,
      "step": 12150
    },
    {
      "epoch": 0.3890595992574099,
      "grad_norm": 15.1875,
      "learning_rate": 3.944893325218265e-05,
      "loss": 5.0649,
      "num_input_tokens_seen": 16457968,
      "step": 12155
    },
    {
      "epoch": 0.38921964022789834,
      "grad_norm": 30.25,
      "learning_rate": 3.944092041939112e-05,
      "loss": 4.8491,
      "num_input_tokens_seen": 16465120,
      "step": 12160
    },
    {
      "epoch": 0.3893796811983868,
      "grad_norm": 24.5,
      "learning_rate": 3.943290535962818e-05,
      "loss": 4.4237,
      "num_input_tokens_seen": 16471840,
      "step": 12165
    },
    {
      "epoch": 0.3895397221688752,
      "grad_norm": 13.0625,
      "learning_rate": 3.942488807412985e-05,
      "loss": 3.8663,
      "num_input_tokens_seen": 16478480,
      "step": 12170
    },
    {
      "epoch": 0.38969976313936366,
      "grad_norm": 23.125,
      "learning_rate": 3.941686856413251e-05,
      "loss": 3.9112,
      "num_input_tokens_seen": 16485280,
      "step": 12175
    },
    {
      "epoch": 0.3898598041098521,
      "grad_norm": 15.875,
      "learning_rate": 3.9408846830872874e-05,
      "loss": 3.9242,
      "num_input_tokens_seen": 16491968,
      "step": 12180
    },
    {
      "epoch": 0.39001984508034054,
      "grad_norm": 29.5,
      "learning_rate": 3.940082287558798e-05,
      "loss": 4.6384,
      "num_input_tokens_seen": 16498848,
      "step": 12185
    },
    {
      "epoch": 0.39017988605082904,
      "grad_norm": 32.0,
      "learning_rate": 3.939279669951522e-05,
      "loss": 4.7982,
      "num_input_tokens_seen": 16509136,
      "step": 12190
    },
    {
      "epoch": 0.3903399270213175,
      "grad_norm": 41.5,
      "learning_rate": 3.938476830389234e-05,
      "loss": 4.3108,
      "num_input_tokens_seen": 16515952,
      "step": 12195
    },
    {
      "epoch": 0.3904999679918059,
      "grad_norm": 30.625,
      "learning_rate": 3.937673768995742e-05,
      "loss": 4.3731,
      "num_input_tokens_seen": 16522704,
      "step": 12200
    },
    {
      "epoch": 0.3904999679918059,
      "eval_loss": 4.491867542266846,
      "eval_runtime": 488.5159,
      "eval_samples_per_second": 28.425,
      "eval_steps_per_second": 14.212,
      "num_input_tokens_seen": 16522704,
      "step": 12200
    },
    {
      "epoch": 0.39066000896229436,
      "grad_norm": 38.5,
      "learning_rate": 3.936870485894888e-05,
      "loss": 4.8276,
      "num_input_tokens_seen": 16529184,
      "step": 12205
    },
    {
      "epoch": 0.3908200499327828,
      "grad_norm": 25.5,
      "learning_rate": 3.9360669812105475e-05,
      "loss": 4.0837,
      "num_input_tokens_seen": 16535920,
      "step": 12210
    },
    {
      "epoch": 0.39098009090327124,
      "grad_norm": 37.5,
      "learning_rate": 3.9352632550666325e-05,
      "loss": 5.1356,
      "num_input_tokens_seen": 16542736,
      "step": 12215
    },
    {
      "epoch": 0.3911401318737597,
      "grad_norm": 19.375,
      "learning_rate": 3.9344593075870866e-05,
      "loss": 3.9772,
      "num_input_tokens_seen": 16549552,
      "step": 12220
    },
    {
      "epoch": 0.3913001728442481,
      "grad_norm": 33.5,
      "learning_rate": 3.933655138895889e-05,
      "loss": 4.7299,
      "num_input_tokens_seen": 16556400,
      "step": 12225
    },
    {
      "epoch": 0.39146021381473656,
      "grad_norm": 31.125,
      "learning_rate": 3.932850749117053e-05,
      "loss": 4.3899,
      "num_input_tokens_seen": 16563264,
      "step": 12230
    },
    {
      "epoch": 0.391620254785225,
      "grad_norm": 45.0,
      "learning_rate": 3.932046138374624e-05,
      "loss": 4.7039,
      "num_input_tokens_seen": 16569952,
      "step": 12235
    },
    {
      "epoch": 0.39178029575571344,
      "grad_norm": 26.875,
      "learning_rate": 3.9312413067926854e-05,
      "loss": 3.7253,
      "num_input_tokens_seen": 16576480,
      "step": 12240
    },
    {
      "epoch": 0.39194033672620193,
      "grad_norm": 42.5,
      "learning_rate": 3.9304362544953506e-05,
      "loss": 4.98,
      "num_input_tokens_seen": 16583248,
      "step": 12245
    },
    {
      "epoch": 0.39210037769669037,
      "grad_norm": 66.0,
      "learning_rate": 3.929630981606769e-05,
      "loss": 5.0355,
      "num_input_tokens_seen": 16589824,
      "step": 12250
    },
    {
      "epoch": 0.3922604186671788,
      "grad_norm": 24.375,
      "learning_rate": 3.928825488251124e-05,
      "loss": 3.6327,
      "num_input_tokens_seen": 16596800,
      "step": 12255
    },
    {
      "epoch": 0.39242045963766725,
      "grad_norm": 59.75,
      "learning_rate": 3.9280197745526344e-05,
      "loss": 4.8344,
      "num_input_tokens_seen": 16603520,
      "step": 12260
    },
    {
      "epoch": 0.3925805006081557,
      "grad_norm": 76.5,
      "learning_rate": 3.9272138406355495e-05,
      "loss": 3.9102,
      "num_input_tokens_seen": 16610336,
      "step": 12265
    },
    {
      "epoch": 0.39274054157864413,
      "grad_norm": 42.5,
      "learning_rate": 3.926407686624154e-05,
      "loss": 4.3356,
      "num_input_tokens_seen": 16617136,
      "step": 12270
    },
    {
      "epoch": 0.39290058254913257,
      "grad_norm": 19.125,
      "learning_rate": 3.9256013126427684e-05,
      "loss": 4.301,
      "num_input_tokens_seen": 16623520,
      "step": 12275
    },
    {
      "epoch": 0.393060623519621,
      "grad_norm": 19.5,
      "learning_rate": 3.9247947188157455e-05,
      "loss": 3.9394,
      "num_input_tokens_seen": 16630832,
      "step": 12280
    },
    {
      "epoch": 0.39322066449010945,
      "grad_norm": 17.125,
      "learning_rate": 3.9239879052674715e-05,
      "loss": 4.9893,
      "num_input_tokens_seen": 16637728,
      "step": 12285
    },
    {
      "epoch": 0.3933807054605979,
      "grad_norm": 25.625,
      "learning_rate": 3.9231808721223673e-05,
      "loss": 4.2118,
      "num_input_tokens_seen": 16644608,
      "step": 12290
    },
    {
      "epoch": 0.3935407464310864,
      "grad_norm": 28.75,
      "learning_rate": 3.9223736195048886e-05,
      "loss": 3.8443,
      "num_input_tokens_seen": 16651536,
      "step": 12295
    },
    {
      "epoch": 0.3937007874015748,
      "grad_norm": 36.25,
      "learning_rate": 3.921566147539523e-05,
      "loss": 3.736,
      "num_input_tokens_seen": 16658224,
      "step": 12300
    },
    {
      "epoch": 0.39386082837206327,
      "grad_norm": 41.25,
      "learning_rate": 3.920758456350792e-05,
      "loss": 3.8734,
      "num_input_tokens_seen": 16665056,
      "step": 12305
    },
    {
      "epoch": 0.3940208693425517,
      "grad_norm": 50.5,
      "learning_rate": 3.919950546063253e-05,
      "loss": 5.0215,
      "num_input_tokens_seen": 16671984,
      "step": 12310
    },
    {
      "epoch": 0.39418091031304014,
      "grad_norm": 28.25,
      "learning_rate": 3.919142416801496e-05,
      "loss": 3.5229,
      "num_input_tokens_seen": 16678624,
      "step": 12315
    },
    {
      "epoch": 0.3943409512835286,
      "grad_norm": 21.625,
      "learning_rate": 3.918334068690144e-05,
      "loss": 3.8074,
      "num_input_tokens_seen": 16685712,
      "step": 12320
    },
    {
      "epoch": 0.394500992254017,
      "grad_norm": 31.875,
      "learning_rate": 3.917525501853855e-05,
      "loss": 3.4057,
      "num_input_tokens_seen": 16692464,
      "step": 12325
    },
    {
      "epoch": 0.39466103322450546,
      "grad_norm": 31.0,
      "learning_rate": 3.916716716417319e-05,
      "loss": 3.8259,
      "num_input_tokens_seen": 16698992,
      "step": 12330
    },
    {
      "epoch": 0.3948210741949939,
      "grad_norm": 32.75,
      "learning_rate": 3.915907712505263e-05,
      "loss": 4.9308,
      "num_input_tokens_seen": 16705456,
      "step": 12335
    },
    {
      "epoch": 0.39498111516548234,
      "grad_norm": 44.25,
      "learning_rate": 3.915098490242444e-05,
      "loss": 4.0601,
      "num_input_tokens_seen": 16711936,
      "step": 12340
    },
    {
      "epoch": 0.3951411561359708,
      "grad_norm": 37.25,
      "learning_rate": 3.914289049753654e-05,
      "loss": 4.7607,
      "num_input_tokens_seen": 16718704,
      "step": 12345
    },
    {
      "epoch": 0.3953011971064593,
      "grad_norm": 25.0,
      "learning_rate": 3.913479391163719e-05,
      "loss": 4.6485,
      "num_input_tokens_seen": 16725952,
      "step": 12350
    },
    {
      "epoch": 0.3954612380769477,
      "grad_norm": 20.625,
      "learning_rate": 3.9126695145975e-05,
      "loss": 3.443,
      "num_input_tokens_seen": 16732960,
      "step": 12355
    },
    {
      "epoch": 0.39562127904743616,
      "grad_norm": 27.0,
      "learning_rate": 3.911859420179889e-05,
      "loss": 5.4642,
      "num_input_tokens_seen": 16739472,
      "step": 12360
    },
    {
      "epoch": 0.3957813200179246,
      "grad_norm": 14.1875,
      "learning_rate": 3.911049108035813e-05,
      "loss": 4.4528,
      "num_input_tokens_seen": 16745984,
      "step": 12365
    },
    {
      "epoch": 0.39594136098841304,
      "grad_norm": 21.125,
      "learning_rate": 3.910238578290232e-05,
      "loss": 3.1048,
      "num_input_tokens_seen": 16752448,
      "step": 12370
    },
    {
      "epoch": 0.3961014019589015,
      "grad_norm": 20.0,
      "learning_rate": 3.90942783106814e-05,
      "loss": 3.7899,
      "num_input_tokens_seen": 16759216,
      "step": 12375
    },
    {
      "epoch": 0.3962614429293899,
      "grad_norm": 15.4375,
      "learning_rate": 3.908616866494564e-05,
      "loss": 4.1383,
      "num_input_tokens_seen": 16766160,
      "step": 12380
    },
    {
      "epoch": 0.39642148389987836,
      "grad_norm": 29.25,
      "learning_rate": 3.907805684694566e-05,
      "loss": 4.6322,
      "num_input_tokens_seen": 16773520,
      "step": 12385
    },
    {
      "epoch": 0.3965815248703668,
      "grad_norm": 19.875,
      "learning_rate": 3.90699428579324e-05,
      "loss": 4.5199,
      "num_input_tokens_seen": 16780480,
      "step": 12390
    },
    {
      "epoch": 0.39674156584085524,
      "grad_norm": 24.375,
      "learning_rate": 3.906182669915713e-05,
      "loss": 4.4492,
      "num_input_tokens_seen": 16787568,
      "step": 12395
    },
    {
      "epoch": 0.3969016068113437,
      "grad_norm": 34.5,
      "learning_rate": 3.9053708371871476e-05,
      "loss": 4.5962,
      "num_input_tokens_seen": 16794064,
      "step": 12400
    },
    {
      "epoch": 0.3969016068113437,
      "eval_loss": 4.491331100463867,
      "eval_runtime": 488.1707,
      "eval_samples_per_second": 28.445,
      "eval_steps_per_second": 14.222,
      "num_input_tokens_seen": 16794064,
      "step": 12400
    },
    {
      "epoch": 0.39706164778183217,
      "grad_norm": 45.5,
      "learning_rate": 3.904558787732738e-05,
      "loss": 4.3302,
      "num_input_tokens_seen": 16801264,
      "step": 12405
    },
    {
      "epoch": 0.3972216887523206,
      "grad_norm": 35.25,
      "learning_rate": 3.9037465216777135e-05,
      "loss": 3.5064,
      "num_input_tokens_seen": 16807936,
      "step": 12410
    },
    {
      "epoch": 0.39738172972280905,
      "grad_norm": 20.0,
      "learning_rate": 3.902934039147334e-05,
      "loss": 4.3428,
      "num_input_tokens_seen": 16814384,
      "step": 12415
    },
    {
      "epoch": 0.3975417706932975,
      "grad_norm": 31.875,
      "learning_rate": 3.902121340266894e-05,
      "loss": 3.3463,
      "num_input_tokens_seen": 16821088,
      "step": 12420
    },
    {
      "epoch": 0.39770181166378593,
      "grad_norm": 46.25,
      "learning_rate": 3.9013084251617246e-05,
      "loss": 4.0403,
      "num_input_tokens_seen": 16827552,
      "step": 12425
    },
    {
      "epoch": 0.39786185263427437,
      "grad_norm": 23.375,
      "learning_rate": 3.9004952939571865e-05,
      "loss": 3.6556,
      "num_input_tokens_seen": 16834304,
      "step": 12430
    },
    {
      "epoch": 0.3980218936047628,
      "grad_norm": 29.125,
      "learning_rate": 3.899681946778673e-05,
      "loss": 4.6712,
      "num_input_tokens_seen": 16841008,
      "step": 12435
    },
    {
      "epoch": 0.39818193457525125,
      "grad_norm": 24.625,
      "learning_rate": 3.898868383751615e-05,
      "loss": 4.1423,
      "num_input_tokens_seen": 16847872,
      "step": 12440
    },
    {
      "epoch": 0.3983419755457397,
      "grad_norm": 35.0,
      "learning_rate": 3.8980546050014724e-05,
      "loss": 4.3497,
      "num_input_tokens_seen": 16854960,
      "step": 12445
    },
    {
      "epoch": 0.39850201651622813,
      "grad_norm": 30.75,
      "learning_rate": 3.897240610653741e-05,
      "loss": 5.4661,
      "num_input_tokens_seen": 16861152,
      "step": 12450
    },
    {
      "epoch": 0.3986620574867166,
      "grad_norm": 21.25,
      "learning_rate": 3.896426400833948e-05,
      "loss": 3.499,
      "num_input_tokens_seen": 16867648,
      "step": 12455
    },
    {
      "epoch": 0.39882209845720507,
      "grad_norm": 20.875,
      "learning_rate": 3.895611975667656e-05,
      "loss": 4.6597,
      "num_input_tokens_seen": 16874496,
      "step": 12460
    },
    {
      "epoch": 0.3989821394276935,
      "grad_norm": 54.75,
      "learning_rate": 3.8947973352804584e-05,
      "loss": 4.7254,
      "num_input_tokens_seen": 16881632,
      "step": 12465
    },
    {
      "epoch": 0.39914218039818194,
      "grad_norm": 25.25,
      "learning_rate": 3.893982479797984e-05,
      "loss": 4.5558,
      "num_input_tokens_seen": 16888656,
      "step": 12470
    },
    {
      "epoch": 0.3993022213686704,
      "grad_norm": 21.875,
      "learning_rate": 3.8931674093458926e-05,
      "loss": 3.8743,
      "num_input_tokens_seen": 16895248,
      "step": 12475
    },
    {
      "epoch": 0.3994622623391588,
      "grad_norm": 59.75,
      "learning_rate": 3.89235212404988e-05,
      "loss": 5.0509,
      "num_input_tokens_seen": 16901744,
      "step": 12480
    },
    {
      "epoch": 0.39962230330964726,
      "grad_norm": 32.0,
      "learning_rate": 3.891536624035672e-05,
      "loss": 3.1617,
      "num_input_tokens_seen": 16908640,
      "step": 12485
    },
    {
      "epoch": 0.3997823442801357,
      "grad_norm": 19.25,
      "learning_rate": 3.8907209094290295e-05,
      "loss": 3.8759,
      "num_input_tokens_seen": 16916208,
      "step": 12490
    },
    {
      "epoch": 0.39994238525062414,
      "grad_norm": 32.75,
      "learning_rate": 3.8899049803557466e-05,
      "loss": 3.9861,
      "num_input_tokens_seen": 16923136,
      "step": 12495
    },
    {
      "epoch": 0.4001024262211126,
      "grad_norm": 51.0,
      "learning_rate": 3.889088836941648e-05,
      "loss": 4.6367,
      "num_input_tokens_seen": 16929552,
      "step": 12500
    },
    {
      "epoch": 0.400262467191601,
      "grad_norm": 41.0,
      "learning_rate": 3.8882724793125946e-05,
      "loss": 3.3251,
      "num_input_tokens_seen": 16936240,
      "step": 12505
    },
    {
      "epoch": 0.4004225081620895,
      "grad_norm": 52.75,
      "learning_rate": 3.8874559075944794e-05,
      "loss": 4.6176,
      "num_input_tokens_seen": 16942928,
      "step": 12510
    },
    {
      "epoch": 0.40058254913257796,
      "grad_norm": 25.75,
      "learning_rate": 3.886639121913227e-05,
      "loss": 4.3152,
      "num_input_tokens_seen": 16949104,
      "step": 12515
    },
    {
      "epoch": 0.4007425901030664,
      "grad_norm": 25.125,
      "learning_rate": 3.885822122394797e-05,
      "loss": 3.8602,
      "num_input_tokens_seen": 16955936,
      "step": 12520
    },
    {
      "epoch": 0.40090263107355484,
      "grad_norm": 26.0,
      "learning_rate": 3.8850049091651794e-05,
      "loss": 3.907,
      "num_input_tokens_seen": 16962848,
      "step": 12525
    },
    {
      "epoch": 0.4010626720440433,
      "grad_norm": 28.125,
      "learning_rate": 3.8841874823504e-05,
      "loss": 3.8651,
      "num_input_tokens_seen": 16969440,
      "step": 12530
    },
    {
      "epoch": 0.4012227130145317,
      "grad_norm": 31.0,
      "learning_rate": 3.8833698420765157e-05,
      "loss": 3.7358,
      "num_input_tokens_seen": 16976080,
      "step": 12535
    },
    {
      "epoch": 0.40138275398502016,
      "grad_norm": 44.25,
      "learning_rate": 3.882551988469618e-05,
      "loss": 4.3821,
      "num_input_tokens_seen": 16982528,
      "step": 12540
    },
    {
      "epoch": 0.4015427949555086,
      "grad_norm": 22.0,
      "learning_rate": 3.881733921655829e-05,
      "loss": 4.3763,
      "num_input_tokens_seen": 16988864,
      "step": 12545
    },
    {
      "epoch": 0.40170283592599704,
      "grad_norm": 21.375,
      "learning_rate": 3.8809156417613054e-05,
      "loss": 3.9668,
      "num_input_tokens_seen": 16995952,
      "step": 12550
    },
    {
      "epoch": 0.4018628768964855,
      "grad_norm": 31.875,
      "learning_rate": 3.8800971489122364e-05,
      "loss": 3.9274,
      "num_input_tokens_seen": 17002816,
      "step": 12555
    },
    {
      "epoch": 0.402022917866974,
      "grad_norm": 16.5,
      "learning_rate": 3.8792784432348434e-05,
      "loss": 3.1817,
      "num_input_tokens_seen": 17009312,
      "step": 12560
    },
    {
      "epoch": 0.4021829588374624,
      "grad_norm": 16.75,
      "learning_rate": 3.878459524855381e-05,
      "loss": 3.9661,
      "num_input_tokens_seen": 17016000,
      "step": 12565
    },
    {
      "epoch": 0.40234299980795085,
      "grad_norm": 41.5,
      "learning_rate": 3.8776403939001384e-05,
      "loss": 3.8502,
      "num_input_tokens_seen": 17022736,
      "step": 12570
    },
    {
      "epoch": 0.4025030407784393,
      "grad_norm": 28.0,
      "learning_rate": 3.876821050495433e-05,
      "loss": 4.2034,
      "num_input_tokens_seen": 17029104,
      "step": 12575
    },
    {
      "epoch": 0.40266308174892773,
      "grad_norm": 21.75,
      "learning_rate": 3.87600149476762e-05,
      "loss": 3.639,
      "num_input_tokens_seen": 17035696,
      "step": 12580
    },
    {
      "epoch": 0.40282312271941617,
      "grad_norm": 29.125,
      "learning_rate": 3.8751817268430843e-05,
      "loss": 4.3936,
      "num_input_tokens_seen": 17042272,
      "step": 12585
    },
    {
      "epoch": 0.4029831636899046,
      "grad_norm": 27.0,
      "learning_rate": 3.8743617468482464e-05,
      "loss": 3.9635,
      "num_input_tokens_seen": 17048848,
      "step": 12590
    },
    {
      "epoch": 0.40314320466039305,
      "grad_norm": 40.0,
      "learning_rate": 3.8735415549095535e-05,
      "loss": 4.4147,
      "num_input_tokens_seen": 17055744,
      "step": 12595
    },
    {
      "epoch": 0.4033032456308815,
      "grad_norm": 31.375,
      "learning_rate": 3.8727211511534934e-05,
      "loss": 3.8551,
      "num_input_tokens_seen": 17062288,
      "step": 12600
    },
    {
      "epoch": 0.4033032456308815,
      "eval_loss": 4.491095542907715,
      "eval_runtime": 488.0966,
      "eval_samples_per_second": 28.449,
      "eval_steps_per_second": 14.225,
      "num_input_tokens_seen": 17062288,
      "step": 12600
    },
    {
      "epoch": 0.40346328660136993,
      "grad_norm": 31.75,
      "learning_rate": 3.8719005357065804e-05,
      "loss": 4.3304,
      "num_input_tokens_seen": 17068912,
      "step": 12605
    },
    {
      "epoch": 0.40362332757185837,
      "grad_norm": 35.0,
      "learning_rate": 3.8710797086953645e-05,
      "loss": 4.6216,
      "num_input_tokens_seen": 17075600,
      "step": 12610
    },
    {
      "epoch": 0.40378336854234687,
      "grad_norm": 25.625,
      "learning_rate": 3.870258670246427e-05,
      "loss": 3.9299,
      "num_input_tokens_seen": 17082128,
      "step": 12615
    },
    {
      "epoch": 0.4039434095128353,
      "grad_norm": 28.125,
      "learning_rate": 3.869437420486384e-05,
      "loss": 3.8977,
      "num_input_tokens_seen": 17088640,
      "step": 12620
    },
    {
      "epoch": 0.40410345048332375,
      "grad_norm": 29.75,
      "learning_rate": 3.8686159595418805e-05,
      "loss": 3.8232,
      "num_input_tokens_seen": 17095648,
      "step": 12625
    },
    {
      "epoch": 0.4042634914538122,
      "grad_norm": 50.75,
      "learning_rate": 3.867794287539597e-05,
      "loss": 5.207,
      "num_input_tokens_seen": 17102768,
      "step": 12630
    },
    {
      "epoch": 0.4044235324243006,
      "grad_norm": 33.75,
      "learning_rate": 3.866972404606245e-05,
      "loss": 4.0631,
      "num_input_tokens_seen": 17109584,
      "step": 12635
    },
    {
      "epoch": 0.40458357339478906,
      "grad_norm": 12.1875,
      "learning_rate": 3.866150310868571e-05,
      "loss": 3.9987,
      "num_input_tokens_seen": 17115968,
      "step": 12640
    },
    {
      "epoch": 0.4047436143652775,
      "grad_norm": 71.5,
      "learning_rate": 3.8653280064533506e-05,
      "loss": 4.4792,
      "num_input_tokens_seen": 17122672,
      "step": 12645
    },
    {
      "epoch": 0.40490365533576594,
      "grad_norm": 16.375,
      "learning_rate": 3.864505491487394e-05,
      "loss": 3.8881,
      "num_input_tokens_seen": 17129440,
      "step": 12650
    },
    {
      "epoch": 0.4050636963062544,
      "grad_norm": 25.5,
      "learning_rate": 3.8636827660975414e-05,
      "loss": 3.3678,
      "num_input_tokens_seen": 17136064,
      "step": 12655
    },
    {
      "epoch": 0.4052237372767428,
      "grad_norm": 15.0625,
      "learning_rate": 3.862859830410671e-05,
      "loss": 4.2388,
      "num_input_tokens_seen": 17142544,
      "step": 12660
    },
    {
      "epoch": 0.4053837782472313,
      "grad_norm": 33.25,
      "learning_rate": 3.862036684553688e-05,
      "loss": 5.5139,
      "num_input_tokens_seen": 17149120,
      "step": 12665
    },
    {
      "epoch": 0.40554381921771976,
      "grad_norm": 35.5,
      "learning_rate": 3.8612133286535314e-05,
      "loss": 5.0123,
      "num_input_tokens_seen": 17155680,
      "step": 12670
    },
    {
      "epoch": 0.4057038601882082,
      "grad_norm": 36.5,
      "learning_rate": 3.860389762837173e-05,
      "loss": 4.4016,
      "num_input_tokens_seen": 17162432,
      "step": 12675
    },
    {
      "epoch": 0.40586390115869664,
      "grad_norm": 38.75,
      "learning_rate": 3.859565987231618e-05,
      "loss": 3.6422,
      "num_input_tokens_seen": 17169376,
      "step": 12680
    },
    {
      "epoch": 0.4060239421291851,
      "grad_norm": 27.75,
      "learning_rate": 3.858742001963902e-05,
      "loss": 4.9104,
      "num_input_tokens_seen": 17176096,
      "step": 12685
    },
    {
      "epoch": 0.4061839830996735,
      "grad_norm": 25.625,
      "learning_rate": 3.857917807161094e-05,
      "loss": 4.0444,
      "num_input_tokens_seen": 17182480,
      "step": 12690
    },
    {
      "epoch": 0.40634402407016196,
      "grad_norm": 51.75,
      "learning_rate": 3.857093402950296e-05,
      "loss": 5.2532,
      "num_input_tokens_seen": 17189440,
      "step": 12695
    },
    {
      "epoch": 0.4065040650406504,
      "grad_norm": 22.125,
      "learning_rate": 3.8562687894586414e-05,
      "loss": 3.7475,
      "num_input_tokens_seen": 17196640,
      "step": 12700
    },
    {
      "epoch": 0.40666410601113884,
      "grad_norm": 29.0,
      "learning_rate": 3.8554439668132946e-05,
      "loss": 4.3304,
      "num_input_tokens_seen": 17203440,
      "step": 12705
    },
    {
      "epoch": 0.4068241469816273,
      "grad_norm": 16.375,
      "learning_rate": 3.854618935141455e-05,
      "loss": 4.703,
      "num_input_tokens_seen": 17210032,
      "step": 12710
    },
    {
      "epoch": 0.4069841879521157,
      "grad_norm": 21.625,
      "learning_rate": 3.8537936945703525e-05,
      "loss": 4.603,
      "num_input_tokens_seen": 17216880,
      "step": 12715
    },
    {
      "epoch": 0.4071442289226042,
      "grad_norm": 25.5,
      "learning_rate": 3.852968245227249e-05,
      "loss": 4.5408,
      "num_input_tokens_seen": 17223440,
      "step": 12720
    },
    {
      "epoch": 0.40730426989309265,
      "grad_norm": 31.375,
      "learning_rate": 3.85214258723944e-05,
      "loss": 4.0416,
      "num_input_tokens_seen": 17230432,
      "step": 12725
    },
    {
      "epoch": 0.4074643108635811,
      "grad_norm": 26.75,
      "learning_rate": 3.8513167207342524e-05,
      "loss": 4.4795,
      "num_input_tokens_seen": 17236880,
      "step": 12730
    },
    {
      "epoch": 0.40762435183406953,
      "grad_norm": 27.0,
      "learning_rate": 3.850490645839044e-05,
      "loss": 5.011,
      "num_input_tokens_seen": 17243680,
      "step": 12735
    },
    {
      "epoch": 0.40778439280455797,
      "grad_norm": 25.125,
      "learning_rate": 3.849664362681207e-05,
      "loss": 4.7807,
      "num_input_tokens_seen": 17250464,
      "step": 12740
    },
    {
      "epoch": 0.4079444337750464,
      "grad_norm": 53.25,
      "learning_rate": 3.848837871388165e-05,
      "loss": 5.4862,
      "num_input_tokens_seen": 17257168,
      "step": 12745
    },
    {
      "epoch": 0.40810447474553485,
      "grad_norm": 34.0,
      "learning_rate": 3.848011172087371e-05,
      "loss": 3.9312,
      "num_input_tokens_seen": 17264096,
      "step": 12750
    },
    {
      "epoch": 0.4082645157160233,
      "grad_norm": 18.875,
      "learning_rate": 3.847184264906315e-05,
      "loss": 3.7816,
      "num_input_tokens_seen": 17271088,
      "step": 12755
    },
    {
      "epoch": 0.40842455668651173,
      "grad_norm": 30.375,
      "learning_rate": 3.846357149972516e-05,
      "loss": 4.5497,
      "num_input_tokens_seen": 17277680,
      "step": 12760
    },
    {
      "epoch": 0.40858459765700017,
      "grad_norm": 31.625,
      "learning_rate": 3.8455298274135246e-05,
      "loss": 4.0048,
      "num_input_tokens_seen": 17284272,
      "step": 12765
    },
    {
      "epoch": 0.4087446386274886,
      "grad_norm": 37.25,
      "learning_rate": 3.8447022973569254e-05,
      "loss": 4.4227,
      "num_input_tokens_seen": 17291120,
      "step": 12770
    },
    {
      "epoch": 0.4089046795979771,
      "grad_norm": 20.125,
      "learning_rate": 3.843874559930332e-05,
      "loss": 4.1644,
      "num_input_tokens_seen": 17297472,
      "step": 12775
    },
    {
      "epoch": 0.40906472056846555,
      "grad_norm": 28.625,
      "learning_rate": 3.843046615261394e-05,
      "loss": 4.6297,
      "num_input_tokens_seen": 17304224,
      "step": 12780
    },
    {
      "epoch": 0.409224761538954,
      "grad_norm": 45.25,
      "learning_rate": 3.842218463477791e-05,
      "loss": 5.094,
      "num_input_tokens_seen": 17311216,
      "step": 12785
    },
    {
      "epoch": 0.4093848025094424,
      "grad_norm": 35.5,
      "learning_rate": 3.841390104707233e-05,
      "loss": 4.2838,
      "num_input_tokens_seen": 17317936,
      "step": 12790
    },
    {
      "epoch": 0.40954484347993086,
      "grad_norm": 39.5,
      "learning_rate": 3.8405615390774643e-05,
      "loss": 4.3253,
      "num_input_tokens_seen": 17324512,
      "step": 12795
    },
    {
      "epoch": 0.4097048844504193,
      "grad_norm": 24.875,
      "learning_rate": 3.839732766716259e-05,
      "loss": 4.5508,
      "num_input_tokens_seen": 17331072,
      "step": 12800
    },
    {
      "epoch": 0.4097048844504193,
      "eval_loss": 4.49737548828125,
      "eval_runtime": 488.2059,
      "eval_samples_per_second": 28.443,
      "eval_steps_per_second": 14.221,
      "num_input_tokens_seen": 17331072,
      "step": 12800
    },
    {
      "epoch": 0.40986492542090774,
      "grad_norm": 37.75,
      "learning_rate": 3.838903787751425e-05,
      "loss": 4.5876,
      "num_input_tokens_seen": 17338080,
      "step": 12805
    },
    {
      "epoch": 0.4100249663913962,
      "grad_norm": 24.125,
      "learning_rate": 3.838074602310802e-05,
      "loss": 2.9951,
      "num_input_tokens_seen": 17344368,
      "step": 12810
    },
    {
      "epoch": 0.4101850073618846,
      "grad_norm": 28.125,
      "learning_rate": 3.837245210522258e-05,
      "loss": 3.7434,
      "num_input_tokens_seen": 17351152,
      "step": 12815
    },
    {
      "epoch": 0.41034504833237306,
      "grad_norm": 28.125,
      "learning_rate": 3.8364156125136996e-05,
      "loss": 3.9632,
      "num_input_tokens_seen": 17357616,
      "step": 12820
    },
    {
      "epoch": 0.41050508930286156,
      "grad_norm": 24.0,
      "learning_rate": 3.835585808413059e-05,
      "loss": 4.6952,
      "num_input_tokens_seen": 17364352,
      "step": 12825
    },
    {
      "epoch": 0.41066513027335,
      "grad_norm": 25.875,
      "learning_rate": 3.8347557983483024e-05,
      "loss": 4.4028,
      "num_input_tokens_seen": 17371056,
      "step": 12830
    },
    {
      "epoch": 0.41082517124383844,
      "grad_norm": 29.75,
      "learning_rate": 3.833925582447428e-05,
      "loss": 4.9428,
      "num_input_tokens_seen": 17378032,
      "step": 12835
    },
    {
      "epoch": 0.4109852122143269,
      "grad_norm": 25.25,
      "learning_rate": 3.8330951608384656e-05,
      "loss": 4.6215,
      "num_input_tokens_seen": 17385104,
      "step": 12840
    },
    {
      "epoch": 0.4111452531848153,
      "grad_norm": 28.625,
      "learning_rate": 3.832264533649477e-05,
      "loss": 4.0839,
      "num_input_tokens_seen": 17391616,
      "step": 12845
    },
    {
      "epoch": 0.41130529415530376,
      "grad_norm": 26.125,
      "learning_rate": 3.8314337010085555e-05,
      "loss": 5.0279,
      "num_input_tokens_seen": 17398336,
      "step": 12850
    },
    {
      "epoch": 0.4114653351257922,
      "grad_norm": 35.5,
      "learning_rate": 3.830602663043824e-05,
      "loss": 4.2563,
      "num_input_tokens_seen": 17404848,
      "step": 12855
    },
    {
      "epoch": 0.41162537609628064,
      "grad_norm": 44.0,
      "learning_rate": 3.8297714198834414e-05,
      "loss": 3.7576,
      "num_input_tokens_seen": 17411312,
      "step": 12860
    },
    {
      "epoch": 0.4117854170667691,
      "grad_norm": 24.75,
      "learning_rate": 3.828939971655595e-05,
      "loss": 3.937,
      "num_input_tokens_seen": 17418128,
      "step": 12865
    },
    {
      "epoch": 0.4119454580372575,
      "grad_norm": 36.5,
      "learning_rate": 3.828108318488505e-05,
      "loss": 5.1682,
      "num_input_tokens_seen": 17424672,
      "step": 12870
    },
    {
      "epoch": 0.41210549900774596,
      "grad_norm": 14.0,
      "learning_rate": 3.8272764605104216e-05,
      "loss": 2.8847,
      "num_input_tokens_seen": 17431472,
      "step": 12875
    },
    {
      "epoch": 0.41226553997823445,
      "grad_norm": 16.75,
      "learning_rate": 3.826444397849628e-05,
      "loss": 3.3049,
      "num_input_tokens_seen": 17438480,
      "step": 12880
    },
    {
      "epoch": 0.4124255809487229,
      "grad_norm": 17.875,
      "learning_rate": 3.825612130634439e-05,
      "loss": 3.8797,
      "num_input_tokens_seen": 17444800,
      "step": 12885
    },
    {
      "epoch": 0.41258562191921133,
      "grad_norm": 17.375,
      "learning_rate": 3.824779658993202e-05,
      "loss": 4.2569,
      "num_input_tokens_seen": 17451568,
      "step": 12890
    },
    {
      "epoch": 0.41274566288969977,
      "grad_norm": 22.0,
      "learning_rate": 3.823946983054292e-05,
      "loss": 3.9473,
      "num_input_tokens_seen": 17458144,
      "step": 12895
    },
    {
      "epoch": 0.4129057038601882,
      "grad_norm": 31.375,
      "learning_rate": 3.82311410294612e-05,
      "loss": 3.6175,
      "num_input_tokens_seen": 17464736,
      "step": 12900
    },
    {
      "epoch": 0.41306574483067665,
      "grad_norm": 16.125,
      "learning_rate": 3.822281018797127e-05,
      "loss": 3.9776,
      "num_input_tokens_seen": 17471760,
      "step": 12905
    },
    {
      "epoch": 0.4132257858011651,
      "grad_norm": 29.25,
      "learning_rate": 3.821447730735783e-05,
      "loss": 4.8531,
      "num_input_tokens_seen": 17478176,
      "step": 12910
    },
    {
      "epoch": 0.41338582677165353,
      "grad_norm": 25.75,
      "learning_rate": 3.820614238890592e-05,
      "loss": 4.0697,
      "num_input_tokens_seen": 17484400,
      "step": 12915
    },
    {
      "epoch": 0.41354586774214197,
      "grad_norm": 30.75,
      "learning_rate": 3.819780543390091e-05,
      "loss": 4.4503,
      "num_input_tokens_seen": 17491488,
      "step": 12920
    },
    {
      "epoch": 0.4137059087126304,
      "grad_norm": 32.0,
      "learning_rate": 3.818946644362844e-05,
      "loss": 4.0353,
      "num_input_tokens_seen": 17497776,
      "step": 12925
    },
    {
      "epoch": 0.4138659496831189,
      "grad_norm": 25.875,
      "learning_rate": 3.81811254193745e-05,
      "loss": 4.2662,
      "num_input_tokens_seen": 17504960,
      "step": 12930
    },
    {
      "epoch": 0.41402599065360735,
      "grad_norm": 32.5,
      "learning_rate": 3.8172782362425366e-05,
      "loss": 4.3457,
      "num_input_tokens_seen": 17511824,
      "step": 12935
    },
    {
      "epoch": 0.4141860316240958,
      "grad_norm": 28.5,
      "learning_rate": 3.816443727406765e-05,
      "loss": 3.8832,
      "num_input_tokens_seen": 17518544,
      "step": 12940
    },
    {
      "epoch": 0.4143460725945842,
      "grad_norm": 23.75,
      "learning_rate": 3.815609015558829e-05,
      "loss": 4.8658,
      "num_input_tokens_seen": 17525440,
      "step": 12945
    },
    {
      "epoch": 0.41450611356507266,
      "grad_norm": 21.875,
      "learning_rate": 3.814774100827448e-05,
      "loss": 4.7209,
      "num_input_tokens_seen": 17532128,
      "step": 12950
    },
    {
      "epoch": 0.4146661545355611,
      "grad_norm": 28.5,
      "learning_rate": 3.813938983341379e-05,
      "loss": 3.8453,
      "num_input_tokens_seen": 17538976,
      "step": 12955
    },
    {
      "epoch": 0.41482619550604954,
      "grad_norm": 18.625,
      "learning_rate": 3.813103663229407e-05,
      "loss": 3.9504,
      "num_input_tokens_seen": 17545936,
      "step": 12960
    },
    {
      "epoch": 0.414986236476538,
      "grad_norm": 35.25,
      "learning_rate": 3.812268140620349e-05,
      "loss": 4.3523,
      "num_input_tokens_seen": 17552816,
      "step": 12965
    },
    {
      "epoch": 0.4151462774470264,
      "grad_norm": 33.0,
      "learning_rate": 3.811432415643051e-05,
      "loss": 4.6358,
      "num_input_tokens_seen": 17559616,
      "step": 12970
    },
    {
      "epoch": 0.41530631841751486,
      "grad_norm": 20.375,
      "learning_rate": 3.8105964884263954e-05,
      "loss": 4.4518,
      "num_input_tokens_seen": 17566400,
      "step": 12975
    },
    {
      "epoch": 0.4154663593880033,
      "grad_norm": 26.5,
      "learning_rate": 3.809760359099291e-05,
      "loss": 4.3277,
      "num_input_tokens_seen": 17573024,
      "step": 12980
    },
    {
      "epoch": 0.4156264003584918,
      "grad_norm": 28.875,
      "learning_rate": 3.8089240277906804e-05,
      "loss": 5.136,
      "num_input_tokens_seen": 17579616,
      "step": 12985
    },
    {
      "epoch": 0.41578644132898024,
      "grad_norm": 32.25,
      "learning_rate": 3.808087494629535e-05,
      "loss": 3.9378,
      "num_input_tokens_seen": 17586624,
      "step": 12990
    },
    {
      "epoch": 0.4159464822994687,
      "grad_norm": 28.375,
      "learning_rate": 3.8072507597448595e-05,
      "loss": 4.9933,
      "num_input_tokens_seen": 17592944,
      "step": 12995
    },
    {
      "epoch": 0.4161065232699571,
      "grad_norm": 39.0,
      "learning_rate": 3.806413823265689e-05,
      "loss": 4.2258,
      "num_input_tokens_seen": 17599616,
      "step": 13000
    },
    {
      "epoch": 0.4161065232699571,
      "eval_loss": 4.500463485717773,
      "eval_runtime": 488.2227,
      "eval_samples_per_second": 28.442,
      "eval_steps_per_second": 14.221,
      "num_input_tokens_seen": 17599616,
      "step": 13000
    },
    {
      "epoch": 0.41626656424044556,
      "grad_norm": 31.75,
      "learning_rate": 3.805576685321089e-05,
      "loss": 3.9311,
      "num_input_tokens_seen": 17606688,
      "step": 13005
    },
    {
      "epoch": 0.416426605210934,
      "grad_norm": 31.0,
      "learning_rate": 3.804739346040158e-05,
      "loss": 4.5251,
      "num_input_tokens_seen": 17613568,
      "step": 13010
    },
    {
      "epoch": 0.41658664618142244,
      "grad_norm": 38.25,
      "learning_rate": 3.8039018055520234e-05,
      "loss": 4.9309,
      "num_input_tokens_seen": 17620896,
      "step": 13015
    },
    {
      "epoch": 0.4167466871519109,
      "grad_norm": 27.25,
      "learning_rate": 3.803064063985844e-05,
      "loss": 3.3536,
      "num_input_tokens_seen": 17627536,
      "step": 13020
    },
    {
      "epoch": 0.4169067281223993,
      "grad_norm": 12.75,
      "learning_rate": 3.802226121470811e-05,
      "loss": 3.9067,
      "num_input_tokens_seen": 17634000,
      "step": 13025
    },
    {
      "epoch": 0.41706676909288776,
      "grad_norm": 25.875,
      "learning_rate": 3.801387978136145e-05,
      "loss": 3.9913,
      "num_input_tokens_seen": 17640688,
      "step": 13030
    },
    {
      "epoch": 0.4172268100633762,
      "grad_norm": 22.5,
      "learning_rate": 3.800549634111099e-05,
      "loss": 4.4529,
      "num_input_tokens_seen": 17647648,
      "step": 13035
    },
    {
      "epoch": 0.4173868510338647,
      "grad_norm": 33.0,
      "learning_rate": 3.799711089524955e-05,
      "loss": 4.6483,
      "num_input_tokens_seen": 17654752,
      "step": 13040
    },
    {
      "epoch": 0.41754689200435313,
      "grad_norm": 27.5,
      "learning_rate": 3.7988723445070285e-05,
      "loss": 5.0597,
      "num_input_tokens_seen": 17661568,
      "step": 13045
    },
    {
      "epoch": 0.41770693297484157,
      "grad_norm": 17.5,
      "learning_rate": 3.798033399186663e-05,
      "loss": 3.3627,
      "num_input_tokens_seen": 17668224,
      "step": 13050
    },
    {
      "epoch": 0.41786697394533,
      "grad_norm": 33.5,
      "learning_rate": 3.797194253693237e-05,
      "loss": 4.3766,
      "num_input_tokens_seen": 17675392,
      "step": 13055
    },
    {
      "epoch": 0.41802701491581845,
      "grad_norm": 38.25,
      "learning_rate": 3.796354908156153e-05,
      "loss": 4.2864,
      "num_input_tokens_seen": 17682688,
      "step": 13060
    },
    {
      "epoch": 0.4181870558863069,
      "grad_norm": 23.5,
      "learning_rate": 3.795515362704853e-05,
      "loss": 3.8495,
      "num_input_tokens_seen": 17689568,
      "step": 13065
    },
    {
      "epoch": 0.41834709685679533,
      "grad_norm": 34.25,
      "learning_rate": 3.794675617468803e-05,
      "loss": 4.2627,
      "num_input_tokens_seen": 17696032,
      "step": 13070
    },
    {
      "epoch": 0.41850713782728377,
      "grad_norm": 26.5,
      "learning_rate": 3.793835672577503e-05,
      "loss": 3.8876,
      "num_input_tokens_seen": 17702512,
      "step": 13075
    },
    {
      "epoch": 0.4186671787977722,
      "grad_norm": 19.75,
      "learning_rate": 3.7929955281604826e-05,
      "loss": 4.976,
      "num_input_tokens_seen": 17709168,
      "step": 13080
    },
    {
      "epoch": 0.41882721976826065,
      "grad_norm": 32.5,
      "learning_rate": 3.7921551843473036e-05,
      "loss": 4.7406,
      "num_input_tokens_seen": 17715776,
      "step": 13085
    },
    {
      "epoch": 0.41898726073874915,
      "grad_norm": 47.25,
      "learning_rate": 3.791314641267557e-05,
      "loss": 4.339,
      "num_input_tokens_seen": 17722432,
      "step": 13090
    },
    {
      "epoch": 0.4191473017092376,
      "grad_norm": 17.375,
      "learning_rate": 3.790473899050864e-05,
      "loss": 4.9165,
      "num_input_tokens_seen": 17728752,
      "step": 13095
    },
    {
      "epoch": 0.419307342679726,
      "grad_norm": 43.25,
      "learning_rate": 3.7896329578268794e-05,
      "loss": 4.3966,
      "num_input_tokens_seen": 17735808,
      "step": 13100
    },
    {
      "epoch": 0.41946738365021446,
      "grad_norm": 26.625,
      "learning_rate": 3.7887918177252855e-05,
      "loss": 5.1509,
      "num_input_tokens_seen": 17742512,
      "step": 13105
    },
    {
      "epoch": 0.4196274246207029,
      "grad_norm": 38.25,
      "learning_rate": 3.787950478875798e-05,
      "loss": 3.7238,
      "num_input_tokens_seen": 17749408,
      "step": 13110
    },
    {
      "epoch": 0.41978746559119134,
      "grad_norm": 29.0,
      "learning_rate": 3.787108941408162e-05,
      "loss": 4.4628,
      "num_input_tokens_seen": 17756224,
      "step": 13115
    },
    {
      "epoch": 0.4199475065616798,
      "grad_norm": 30.75,
      "learning_rate": 3.786267205452151e-05,
      "loss": 3.3507,
      "num_input_tokens_seen": 17763152,
      "step": 13120
    },
    {
      "epoch": 0.4201075475321682,
      "grad_norm": 47.0,
      "learning_rate": 3.785425271137573e-05,
      "loss": 5.216,
      "num_input_tokens_seen": 17769840,
      "step": 13125
    },
    {
      "epoch": 0.42026758850265666,
      "grad_norm": 42.75,
      "learning_rate": 3.7845831385942655e-05,
      "loss": 4.1568,
      "num_input_tokens_seen": 17776416,
      "step": 13130
    },
    {
      "epoch": 0.4204276294731451,
      "grad_norm": 54.0,
      "learning_rate": 3.7837408079520944e-05,
      "loss": 5.4837,
      "num_input_tokens_seen": 17782992,
      "step": 13135
    },
    {
      "epoch": 0.42058767044363354,
      "grad_norm": 27.25,
      "learning_rate": 3.782898279340957e-05,
      "loss": 5.737,
      "num_input_tokens_seen": 17789360,
      "step": 13140
    },
    {
      "epoch": 0.42074771141412204,
      "grad_norm": 21.75,
      "learning_rate": 3.782055552890784e-05,
      "loss": 5.6039,
      "num_input_tokens_seen": 17795760,
      "step": 13145
    },
    {
      "epoch": 0.4209077523846105,
      "grad_norm": 17.0,
      "learning_rate": 3.781212628731534e-05,
      "loss": 4.1072,
      "num_input_tokens_seen": 17802112,
      "step": 13150
    },
    {
      "epoch": 0.4210677933550989,
      "grad_norm": 22.875,
      "learning_rate": 3.7803695069931946e-05,
      "loss": 4.1806,
      "num_input_tokens_seen": 17808880,
      "step": 13155
    },
    {
      "epoch": 0.42122783432558736,
      "grad_norm": 23.375,
      "learning_rate": 3.779526187805789e-05,
      "loss": 3.9791,
      "num_input_tokens_seen": 17815696,
      "step": 13160
    },
    {
      "epoch": 0.4213878752960758,
      "grad_norm": 22.875,
      "learning_rate": 3.778682671299364e-05,
      "loss": 3.9204,
      "num_input_tokens_seen": 17822528,
      "step": 13165
    },
    {
      "epoch": 0.42154791626656424,
      "grad_norm": 19.875,
      "learning_rate": 3.777838957604003e-05,
      "loss": 5.0726,
      "num_input_tokens_seen": 17829328,
      "step": 13170
    },
    {
      "epoch": 0.4217079572370527,
      "grad_norm": 26.75,
      "learning_rate": 3.776995046849816e-05,
      "loss": 3.8203,
      "num_input_tokens_seen": 17836000,
      "step": 13175
    },
    {
      "epoch": 0.4218679982075411,
      "grad_norm": 34.25,
      "learning_rate": 3.776150939166945e-05,
      "loss": 4.5373,
      "num_input_tokens_seen": 17842800,
      "step": 13180
    },
    {
      "epoch": 0.42202803917802956,
      "grad_norm": 20.75,
      "learning_rate": 3.775306634685562e-05,
      "loss": 3.8599,
      "num_input_tokens_seen": 17849680,
      "step": 13185
    },
    {
      "epoch": 0.422188080148518,
      "grad_norm": 15.1875,
      "learning_rate": 3.7744621335358696e-05,
      "loss": 4.6432,
      "num_input_tokens_seen": 17856400,
      "step": 13190
    },
    {
      "epoch": 0.4223481211190065,
      "grad_norm": 21.875,
      "learning_rate": 3.7736174358481e-05,
      "loss": 4.0733,
      "num_input_tokens_seen": 17862784,
      "step": 13195
    },
    {
      "epoch": 0.42250816208949493,
      "grad_norm": 33.75,
      "learning_rate": 3.7727725417525175e-05,
      "loss": 3.7439,
      "num_input_tokens_seen": 17869424,
      "step": 13200
    },
    {
      "epoch": 0.42250816208949493,
      "eval_loss": 4.490839004516602,
      "eval_runtime": 488.0343,
      "eval_samples_per_second": 28.453,
      "eval_steps_per_second": 14.226,
      "num_input_tokens_seen": 17869424,
      "step": 13200
    },
    {
      "epoch": 0.42266820305998337,
      "grad_norm": 24.5,
      "learning_rate": 3.771927451379414e-05,
      "loss": 4.1337,
      "num_input_tokens_seen": 17876800,
      "step": 13205
    },
    {
      "epoch": 0.4228282440304718,
      "grad_norm": 52.0,
      "learning_rate": 3.7710821648591135e-05,
      "loss": 4.6403,
      "num_input_tokens_seen": 17883200,
      "step": 13210
    },
    {
      "epoch": 0.42298828500096025,
      "grad_norm": 19.375,
      "learning_rate": 3.7702366823219694e-05,
      "loss": 3.3833,
      "num_input_tokens_seen": 17890144,
      "step": 13215
    },
    {
      "epoch": 0.4231483259714487,
      "grad_norm": 42.0,
      "learning_rate": 3.769391003898366e-05,
      "loss": 4.1933,
      "num_input_tokens_seen": 17896928,
      "step": 13220
    },
    {
      "epoch": 0.42330836694193713,
      "grad_norm": 16.5,
      "learning_rate": 3.768545129718718e-05,
      "loss": 4.7351,
      "num_input_tokens_seen": 17903728,
      "step": 13225
    },
    {
      "epoch": 0.42346840791242557,
      "grad_norm": 22.125,
      "learning_rate": 3.7676990599134686e-05,
      "loss": 4.6987,
      "num_input_tokens_seen": 17910816,
      "step": 13230
    },
    {
      "epoch": 0.423628448882914,
      "grad_norm": 28.625,
      "learning_rate": 3.766852794613095e-05,
      "loss": 4.9052,
      "num_input_tokens_seen": 17917504,
      "step": 13235
    },
    {
      "epoch": 0.42378848985340245,
      "grad_norm": 36.5,
      "learning_rate": 3.766006333948099e-05,
      "loss": 4.9077,
      "num_input_tokens_seen": 17924544,
      "step": 13240
    },
    {
      "epoch": 0.4239485308238909,
      "grad_norm": 20.125,
      "learning_rate": 3.765159678049017e-05,
      "loss": 4.8182,
      "num_input_tokens_seen": 17931136,
      "step": 13245
    },
    {
      "epoch": 0.4241085717943794,
      "grad_norm": 24.25,
      "learning_rate": 3.7643128270464134e-05,
      "loss": 4.5902,
      "num_input_tokens_seen": 17938416,
      "step": 13250
    },
    {
      "epoch": 0.4242686127648678,
      "grad_norm": 30.0,
      "learning_rate": 3.763465781070884e-05,
      "loss": 4.2677,
      "num_input_tokens_seen": 17945632,
      "step": 13255
    },
    {
      "epoch": 0.42442865373535626,
      "grad_norm": 28.75,
      "learning_rate": 3.762618540253052e-05,
      "loss": 3.4851,
      "num_input_tokens_seen": 17953104,
      "step": 13260
    },
    {
      "epoch": 0.4245886947058447,
      "grad_norm": 37.75,
      "learning_rate": 3.761771104723576e-05,
      "loss": 5.0072,
      "num_input_tokens_seen": 17959760,
      "step": 13265
    },
    {
      "epoch": 0.42474873567633314,
      "grad_norm": 32.75,
      "learning_rate": 3.7609234746131386e-05,
      "loss": 4.021,
      "num_input_tokens_seen": 17966576,
      "step": 13270
    },
    {
      "epoch": 0.4249087766468216,
      "grad_norm": 18.625,
      "learning_rate": 3.7600756500524556e-05,
      "loss": 4.6762,
      "num_input_tokens_seen": 17973264,
      "step": 13275
    },
    {
      "epoch": 0.42506881761731,
      "grad_norm": 27.125,
      "learning_rate": 3.759227631172271e-05,
      "loss": 5.1786,
      "num_input_tokens_seen": 17979824,
      "step": 13280
    },
    {
      "epoch": 0.42522885858779846,
      "grad_norm": 27.25,
      "learning_rate": 3.758379418103363e-05,
      "loss": 4.2756,
      "num_input_tokens_seen": 17986592,
      "step": 13285
    },
    {
      "epoch": 0.4253888995582869,
      "grad_norm": 38.0,
      "learning_rate": 3.757531010976534e-05,
      "loss": 3.7386,
      "num_input_tokens_seen": 17993648,
      "step": 13290
    },
    {
      "epoch": 0.42554894052877534,
      "grad_norm": 19.0,
      "learning_rate": 3.75668240992262e-05,
      "loss": 4.1228,
      "num_input_tokens_seen": 18000528,
      "step": 13295
    },
    {
      "epoch": 0.42570898149926384,
      "grad_norm": 31.875,
      "learning_rate": 3.7558336150724865e-05,
      "loss": 4.678,
      "num_input_tokens_seen": 18007792,
      "step": 13300
    },
    {
      "epoch": 0.4258690224697523,
      "grad_norm": 25.25,
      "learning_rate": 3.754984626557028e-05,
      "loss": 4.851,
      "num_input_tokens_seen": 18014448,
      "step": 13305
    },
    {
      "epoch": 0.4260290634402407,
      "grad_norm": 35.25,
      "learning_rate": 3.754135444507168e-05,
      "loss": 5.4659,
      "num_input_tokens_seen": 18021120,
      "step": 13310
    },
    {
      "epoch": 0.42618910441072916,
      "grad_norm": 22.875,
      "learning_rate": 3.753286069053863e-05,
      "loss": 4.4692,
      "num_input_tokens_seen": 18027536,
      "step": 13315
    },
    {
      "epoch": 0.4263491453812176,
      "grad_norm": 34.5,
      "learning_rate": 3.7524365003280945e-05,
      "loss": 4.4111,
      "num_input_tokens_seen": 18034576,
      "step": 13320
    },
    {
      "epoch": 0.42650918635170604,
      "grad_norm": 23.0,
      "learning_rate": 3.75158673846088e-05,
      "loss": 3.6454,
      "num_input_tokens_seen": 18040896,
      "step": 13325
    },
    {
      "epoch": 0.4266692273221945,
      "grad_norm": 18.875,
      "learning_rate": 3.750736783583262e-05,
      "loss": 3.8775,
      "num_input_tokens_seen": 18047776,
      "step": 13330
    },
    {
      "epoch": 0.4268292682926829,
      "grad_norm": 23.375,
      "learning_rate": 3.7498866358263144e-05,
      "loss": 4.0045,
      "num_input_tokens_seen": 18054544,
      "step": 13335
    },
    {
      "epoch": 0.42698930926317136,
      "grad_norm": 47.75,
      "learning_rate": 3.74903629532114e-05,
      "loss": 4.3524,
      "num_input_tokens_seen": 18060992,
      "step": 13340
    },
    {
      "epoch": 0.4271493502336598,
      "grad_norm": 17.875,
      "learning_rate": 3.748185762198873e-05,
      "loss": 4.2091,
      "num_input_tokens_seen": 18067424,
      "step": 13345
    },
    {
      "epoch": 0.42730939120414824,
      "grad_norm": 25.625,
      "learning_rate": 3.747335036590676e-05,
      "loss": 4.2932,
      "num_input_tokens_seen": 18074032,
      "step": 13350
    },
    {
      "epoch": 0.42746943217463673,
      "grad_norm": 30.375,
      "learning_rate": 3.7464841186277405e-05,
      "loss": 5.0038,
      "num_input_tokens_seen": 18080544,
      "step": 13355
    },
    {
      "epoch": 0.42762947314512517,
      "grad_norm": 27.75,
      "learning_rate": 3.7456330084412896e-05,
      "loss": 3.5619,
      "num_input_tokens_seen": 18087376,
      "step": 13360
    },
    {
      "epoch": 0.4277895141156136,
      "grad_norm": 23.75,
      "learning_rate": 3.744781706162576e-05,
      "loss": 3.5608,
      "num_input_tokens_seen": 18094208,
      "step": 13365
    },
    {
      "epoch": 0.42794955508610205,
      "grad_norm": 51.75,
      "learning_rate": 3.743930211922879e-05,
      "loss": 4.2505,
      "num_input_tokens_seen": 18100752,
      "step": 13370
    },
    {
      "epoch": 0.4281095960565905,
      "grad_norm": 32.75,
      "learning_rate": 3.743078525853513e-05,
      "loss": 4.6643,
      "num_input_tokens_seen": 18107248,
      "step": 13375
    },
    {
      "epoch": 0.42826963702707893,
      "grad_norm": 34.25,
      "learning_rate": 3.7422266480858154e-05,
      "loss": 3.6598,
      "num_input_tokens_seen": 18113856,
      "step": 13380
    },
    {
      "epoch": 0.42842967799756737,
      "grad_norm": 33.0,
      "learning_rate": 3.741374578751158e-05,
      "loss": 5.287,
      "num_input_tokens_seen": 18120384,
      "step": 13385
    },
    {
      "epoch": 0.4285897189680558,
      "grad_norm": 24.75,
      "learning_rate": 3.740522317980941e-05,
      "loss": 4.2657,
      "num_input_tokens_seen": 18127472,
      "step": 13390
    },
    {
      "epoch": 0.42874975993854425,
      "grad_norm": 36.25,
      "learning_rate": 3.739669865906593e-05,
      "loss": 4.5449,
      "num_input_tokens_seen": 18134224,
      "step": 13395
    },
    {
      "epoch": 0.4289098009090327,
      "grad_norm": 26.875,
      "learning_rate": 3.738817222659573e-05,
      "loss": 3.3628,
      "num_input_tokens_seen": 18141136,
      "step": 13400
    },
    {
      "epoch": 0.4289098009090327,
      "eval_loss": 4.512414932250977,
      "eval_runtime": 488.3129,
      "eval_samples_per_second": 28.437,
      "eval_steps_per_second": 14.218,
      "num_input_tokens_seen": 18141136,
      "step": 13400
    },
    {
      "epoch": 0.42906984187952113,
      "grad_norm": 32.75,
      "learning_rate": 3.73796438837137e-05,
      "loss": 3.7886,
      "num_input_tokens_seen": 18148160,
      "step": 13405
    },
    {
      "epoch": 0.4292298828500096,
      "grad_norm": 42.25,
      "learning_rate": 3.7371113631735e-05,
      "loss": 4.7449,
      "num_input_tokens_seen": 18154768,
      "step": 13410
    },
    {
      "epoch": 0.42938992382049807,
      "grad_norm": 19.375,
      "learning_rate": 3.736258147197512e-05,
      "loss": 4.0225,
      "num_input_tokens_seen": 18161632,
      "step": 13415
    },
    {
      "epoch": 0.4295499647909865,
      "grad_norm": 21.625,
      "learning_rate": 3.735404740574981e-05,
      "loss": 3.7567,
      "num_input_tokens_seen": 18168688,
      "step": 13420
    },
    {
      "epoch": 0.42971000576147494,
      "grad_norm": 14.9375,
      "learning_rate": 3.7345511434375145e-05,
      "loss": 4.0078,
      "num_input_tokens_seen": 18175296,
      "step": 13425
    },
    {
      "epoch": 0.4298700467319634,
      "grad_norm": 45.25,
      "learning_rate": 3.733697355916748e-05,
      "loss": 4.5916,
      "num_input_tokens_seen": 18182544,
      "step": 13430
    },
    {
      "epoch": 0.4300300877024518,
      "grad_norm": 25.5,
      "learning_rate": 3.732843378144345e-05,
      "loss": 3.7781,
      "num_input_tokens_seen": 18189584,
      "step": 13435
    },
    {
      "epoch": 0.43019012867294026,
      "grad_norm": 19.625,
      "learning_rate": 3.7319892102519995e-05,
      "loss": 4.6439,
      "num_input_tokens_seen": 18196736,
      "step": 13440
    },
    {
      "epoch": 0.4303501696434287,
      "grad_norm": 31.375,
      "learning_rate": 3.731134852371436e-05,
      "loss": 4.0857,
      "num_input_tokens_seen": 18203424,
      "step": 13445
    },
    {
      "epoch": 0.43051021061391714,
      "grad_norm": 37.75,
      "learning_rate": 3.730280304634408e-05,
      "loss": 4.174,
      "num_input_tokens_seen": 18210256,
      "step": 13450
    },
    {
      "epoch": 0.4306702515844056,
      "grad_norm": 30.875,
      "learning_rate": 3.729425567172696e-05,
      "loss": 4.2074,
      "num_input_tokens_seen": 18217008,
      "step": 13455
    },
    {
      "epoch": 0.4308302925548941,
      "grad_norm": 39.0,
      "learning_rate": 3.728570640118111e-05,
      "loss": 4.7871,
      "num_input_tokens_seen": 18223712,
      "step": 13460
    },
    {
      "epoch": 0.4309903335253825,
      "grad_norm": 24.0,
      "learning_rate": 3.727715523602494e-05,
      "loss": 4.1555,
      "num_input_tokens_seen": 18230192,
      "step": 13465
    },
    {
      "epoch": 0.43115037449587096,
      "grad_norm": 65.5,
      "learning_rate": 3.726860217757715e-05,
      "loss": 4.0112,
      "num_input_tokens_seen": 18236912,
      "step": 13470
    },
    {
      "epoch": 0.4313104154663594,
      "grad_norm": 27.875,
      "learning_rate": 3.726004722715673e-05,
      "loss": 5.3264,
      "num_input_tokens_seen": 18244176,
      "step": 13475
    },
    {
      "epoch": 0.43147045643684784,
      "grad_norm": 31.375,
      "learning_rate": 3.725149038608296e-05,
      "loss": 3.4342,
      "num_input_tokens_seen": 18251184,
      "step": 13480
    },
    {
      "epoch": 0.4316304974073363,
      "grad_norm": 32.75,
      "learning_rate": 3.7242931655675404e-05,
      "loss": 4.0453,
      "num_input_tokens_seen": 18258496,
      "step": 13485
    },
    {
      "epoch": 0.4317905383778247,
      "grad_norm": 37.75,
      "learning_rate": 3.7234371037253937e-05,
      "loss": 3.6232,
      "num_input_tokens_seen": 18265328,
      "step": 13490
    },
    {
      "epoch": 0.43195057934831316,
      "grad_norm": 40.25,
      "learning_rate": 3.7225808532138705e-05,
      "loss": 4.8556,
      "num_input_tokens_seen": 18272480,
      "step": 13495
    },
    {
      "epoch": 0.4321106203188016,
      "grad_norm": 33.5,
      "learning_rate": 3.721724414165016e-05,
      "loss": 3.9692,
      "num_input_tokens_seen": 18278976,
      "step": 13500
    },
    {
      "epoch": 0.43227066128929004,
      "grad_norm": 41.0,
      "learning_rate": 3.720867786710904e-05,
      "loss": 4.2231,
      "num_input_tokens_seen": 18285680,
      "step": 13505
    },
    {
      "epoch": 0.4324307022597785,
      "grad_norm": 36.5,
      "learning_rate": 3.7200109709836366e-05,
      "loss": 3.6343,
      "num_input_tokens_seen": 18292192,
      "step": 13510
    },
    {
      "epoch": 0.43259074323026697,
      "grad_norm": 29.375,
      "learning_rate": 3.7191539671153465e-05,
      "loss": 4.3478,
      "num_input_tokens_seen": 18298960,
      "step": 13515
    },
    {
      "epoch": 0.4327507842007554,
      "grad_norm": 33.0,
      "learning_rate": 3.718296775238193e-05,
      "loss": 3.4724,
      "num_input_tokens_seen": 18305904,
      "step": 13520
    },
    {
      "epoch": 0.43291082517124385,
      "grad_norm": 31.125,
      "learning_rate": 3.7174393954843675e-05,
      "loss": 5.0747,
      "num_input_tokens_seen": 18312720,
      "step": 13525
    },
    {
      "epoch": 0.4330708661417323,
      "grad_norm": 24.625,
      "learning_rate": 3.716581827986087e-05,
      "loss": 4.0159,
      "num_input_tokens_seen": 18319568,
      "step": 13530
    },
    {
      "epoch": 0.43323090711222073,
      "grad_norm": 27.375,
      "learning_rate": 3.7157240728756004e-05,
      "loss": 4.0774,
      "num_input_tokens_seen": 18326144,
      "step": 13535
    },
    {
      "epoch": 0.43339094808270917,
      "grad_norm": 23.75,
      "learning_rate": 3.714866130285184e-05,
      "loss": 4.5613,
      "num_input_tokens_seen": 18333216,
      "step": 13540
    },
    {
      "epoch": 0.4335509890531976,
      "grad_norm": 21.75,
      "learning_rate": 3.714008000347143e-05,
      "loss": 3.9571,
      "num_input_tokens_seen": 18340368,
      "step": 13545
    },
    {
      "epoch": 0.43371103002368605,
      "grad_norm": 39.5,
      "learning_rate": 3.7131496831938126e-05,
      "loss": 4.7329,
      "num_input_tokens_seen": 18347264,
      "step": 13550
    },
    {
      "epoch": 0.4338710709941745,
      "grad_norm": 36.0,
      "learning_rate": 3.7122911789575565e-05,
      "loss": 4.2714,
      "num_input_tokens_seen": 18354304,
      "step": 13555
    },
    {
      "epoch": 0.43403111196466293,
      "grad_norm": 29.375,
      "learning_rate": 3.711432487770765e-05,
      "loss": 4.4964,
      "num_input_tokens_seen": 18360992,
      "step": 13560
    },
    {
      "epoch": 0.4341911529351514,
      "grad_norm": 31.5,
      "learning_rate": 3.710573609765861e-05,
      "loss": 5.22,
      "num_input_tokens_seen": 18367312,
      "step": 13565
    },
    {
      "epoch": 0.43435119390563987,
      "grad_norm": 39.5,
      "learning_rate": 3.709714545075292e-05,
      "loss": 4.4102,
      "num_input_tokens_seen": 18373984,
      "step": 13570
    },
    {
      "epoch": 0.4345112348761283,
      "grad_norm": 18.875,
      "learning_rate": 3.708855293831538e-05,
      "loss": 4.8608,
      "num_input_tokens_seen": 18380720,
      "step": 13575
    },
    {
      "epoch": 0.43467127584661674,
      "grad_norm": 59.25,
      "learning_rate": 3.707995856167107e-05,
      "loss": 4.5243,
      "num_input_tokens_seen": 18387552,
      "step": 13580
    },
    {
      "epoch": 0.4348313168171052,
      "grad_norm": 20.625,
      "learning_rate": 3.707136232214534e-05,
      "loss": 3.9415,
      "num_input_tokens_seen": 18394272,
      "step": 13585
    },
    {
      "epoch": 0.4349913577875936,
      "grad_norm": 31.25,
      "learning_rate": 3.7062764221063844e-05,
      "loss": 4.3943,
      "num_input_tokens_seen": 18400928,
      "step": 13590
    },
    {
      "epoch": 0.43515139875808206,
      "grad_norm": 22.5,
      "learning_rate": 3.705416425975252e-05,
      "loss": 4.2346,
      "num_input_tokens_seen": 18407472,
      "step": 13595
    },
    {
      "epoch": 0.4353114397285705,
      "grad_norm": 17.0,
      "learning_rate": 3.704556243953758e-05,
      "loss": 3.783,
      "num_input_tokens_seen": 18414272,
      "step": 13600
    },
    {
      "epoch": 0.4353114397285705,
      "eval_loss": 4.494548797607422,
      "eval_runtime": 488.2993,
      "eval_samples_per_second": 28.437,
      "eval_steps_per_second": 14.219,
      "num_input_tokens_seen": 18414272,
      "step": 13600
    },
    {
      "epoch": 0.43547148069905894,
      "grad_norm": 59.75,
      "learning_rate": 3.7036958761745535e-05,
      "loss": 4.0511,
      "num_input_tokens_seen": 18420864,
      "step": 13605
    },
    {
      "epoch": 0.4356315216695474,
      "grad_norm": 31.25,
      "learning_rate": 3.702835322770318e-05,
      "loss": 4.1746,
      "num_input_tokens_seen": 18427136,
      "step": 13610
    },
    {
      "epoch": 0.4357915626400358,
      "grad_norm": 18.375,
      "learning_rate": 3.701974583873761e-05,
      "loss": 4.4565,
      "num_input_tokens_seen": 18433920,
      "step": 13615
    },
    {
      "epoch": 0.4359516036105243,
      "grad_norm": 28.625,
      "learning_rate": 3.701113659617618e-05,
      "loss": 4.2203,
      "num_input_tokens_seen": 18440528,
      "step": 13620
    },
    {
      "epoch": 0.43611164458101276,
      "grad_norm": 35.25,
      "learning_rate": 3.7002525501346535e-05,
      "loss": 5.6182,
      "num_input_tokens_seen": 18447088,
      "step": 13625
    },
    {
      "epoch": 0.4362716855515012,
      "grad_norm": 62.0,
      "learning_rate": 3.699391255557664e-05,
      "loss": 4.528,
      "num_input_tokens_seen": 18453488,
      "step": 13630
    },
    {
      "epoch": 0.43643172652198964,
      "grad_norm": 23.375,
      "learning_rate": 3.69852977601947e-05,
      "loss": 4.0334,
      "num_input_tokens_seen": 18460608,
      "step": 13635
    },
    {
      "epoch": 0.4365917674924781,
      "grad_norm": 35.5,
      "learning_rate": 3.697668111652922e-05,
      "loss": 3.6373,
      "num_input_tokens_seen": 18467536,
      "step": 13640
    },
    {
      "epoch": 0.4367518084629665,
      "grad_norm": 19.625,
      "learning_rate": 3.6968062625909005e-05,
      "loss": 4.3952,
      "num_input_tokens_seen": 18474880,
      "step": 13645
    },
    {
      "epoch": 0.43691184943345496,
      "grad_norm": 39.75,
      "learning_rate": 3.6959442289663135e-05,
      "loss": 4.9581,
      "num_input_tokens_seen": 18482048,
      "step": 13650
    },
    {
      "epoch": 0.4370718904039434,
      "grad_norm": 25.0,
      "learning_rate": 3.695082010912098e-05,
      "loss": 3.7616,
      "num_input_tokens_seen": 18488752,
      "step": 13655
    },
    {
      "epoch": 0.43723193137443184,
      "grad_norm": 28.5,
      "learning_rate": 3.694219608561217e-05,
      "loss": 4.2434,
      "num_input_tokens_seen": 18495360,
      "step": 13660
    },
    {
      "epoch": 0.4373919723449203,
      "grad_norm": 35.0,
      "learning_rate": 3.693357022046665e-05,
      "loss": 4.6666,
      "num_input_tokens_seen": 18502208,
      "step": 13665
    },
    {
      "epoch": 0.43755201331540877,
      "grad_norm": 17.625,
      "learning_rate": 3.6924942515014644e-05,
      "loss": 3.5591,
      "num_input_tokens_seen": 18509232,
      "step": 13670
    },
    {
      "epoch": 0.4377120542858972,
      "grad_norm": 36.0,
      "learning_rate": 3.691631297058664e-05,
      "loss": 4.886,
      "num_input_tokens_seen": 18515888,
      "step": 13675
    },
    {
      "epoch": 0.43787209525638565,
      "grad_norm": 36.25,
      "learning_rate": 3.6907681588513424e-05,
      "loss": 4.6631,
      "num_input_tokens_seen": 18522784,
      "step": 13680
    },
    {
      "epoch": 0.4380321362268741,
      "grad_norm": 20.375,
      "learning_rate": 3.689904837012606e-05,
      "loss": 3.895,
      "num_input_tokens_seen": 18529552,
      "step": 13685
    },
    {
      "epoch": 0.43819217719736253,
      "grad_norm": 21.75,
      "learning_rate": 3.689041331675591e-05,
      "loss": 4.9718,
      "num_input_tokens_seen": 18536512,
      "step": 13690
    },
    {
      "epoch": 0.43835221816785097,
      "grad_norm": 21.125,
      "learning_rate": 3.688177642973461e-05,
      "loss": 3.3981,
      "num_input_tokens_seen": 18543312,
      "step": 13695
    },
    {
      "epoch": 0.4385122591383394,
      "grad_norm": 34.25,
      "learning_rate": 3.687313771039406e-05,
      "loss": 4.263,
      "num_input_tokens_seen": 18550048,
      "step": 13700
    },
    {
      "epoch": 0.43867230010882785,
      "grad_norm": 37.0,
      "learning_rate": 3.686449716006647e-05,
      "loss": 5.2306,
      "num_input_tokens_seen": 18556864,
      "step": 13705
    },
    {
      "epoch": 0.4388323410793163,
      "grad_norm": 31.875,
      "learning_rate": 3.685585478008432e-05,
      "loss": 4.569,
      "num_input_tokens_seen": 18563936,
      "step": 13710
    },
    {
      "epoch": 0.43899238204980473,
      "grad_norm": 35.5,
      "learning_rate": 3.6847210571780364e-05,
      "loss": 3.8738,
      "num_input_tokens_seen": 18570688,
      "step": 13715
    },
    {
      "epoch": 0.43915242302029317,
      "grad_norm": 41.75,
      "learning_rate": 3.683856453648767e-05,
      "loss": 4.1407,
      "num_input_tokens_seen": 18577520,
      "step": 13720
    },
    {
      "epoch": 0.43931246399078167,
      "grad_norm": 26.625,
      "learning_rate": 3.682991667553954e-05,
      "loss": 4.1161,
      "num_input_tokens_seen": 18584192,
      "step": 13725
    },
    {
      "epoch": 0.4394725049612701,
      "grad_norm": 20.125,
      "learning_rate": 3.6821266990269606e-05,
      "loss": 3.5023,
      "num_input_tokens_seen": 18591104,
      "step": 13730
    },
    {
      "epoch": 0.43963254593175854,
      "grad_norm": 22.375,
      "learning_rate": 3.681261548201174e-05,
      "loss": 4.5095,
      "num_input_tokens_seen": 18597792,
      "step": 13735
    },
    {
      "epoch": 0.439792586902247,
      "grad_norm": 29.875,
      "learning_rate": 3.6803962152100125e-05,
      "loss": 4.1145,
      "num_input_tokens_seen": 18604704,
      "step": 13740
    },
    {
      "epoch": 0.4399526278727354,
      "grad_norm": 19.375,
      "learning_rate": 3.67953070018692e-05,
      "loss": 3.6806,
      "num_input_tokens_seen": 18611440,
      "step": 13745
    },
    {
      "epoch": 0.44011266884322386,
      "grad_norm": 19.5,
      "learning_rate": 3.678665003265371e-05,
      "loss": 5.2914,
      "num_input_tokens_seen": 18617840,
      "step": 13750
    },
    {
      "epoch": 0.4402727098137123,
      "grad_norm": 20.375,
      "learning_rate": 3.677799124578867e-05,
      "loss": 4.9694,
      "num_input_tokens_seen": 18624304,
      "step": 13755
    },
    {
      "epoch": 0.44043275078420074,
      "grad_norm": 27.5,
      "learning_rate": 3.676933064260937e-05,
      "loss": 4.0326,
      "num_input_tokens_seen": 18631328,
      "step": 13760
    },
    {
      "epoch": 0.4405927917546892,
      "grad_norm": 24.375,
      "learning_rate": 3.6760668224451365e-05,
      "loss": 3.5556,
      "num_input_tokens_seen": 18638192,
      "step": 13765
    },
    {
      "epoch": 0.4407528327251776,
      "grad_norm": 23.875,
      "learning_rate": 3.675200399265054e-05,
      "loss": 3.8562,
      "num_input_tokens_seen": 18645120,
      "step": 13770
    },
    {
      "epoch": 0.44091287369566606,
      "grad_norm": 29.25,
      "learning_rate": 3.6743337948543014e-05,
      "loss": 4.5957,
      "num_input_tokens_seen": 18651776,
      "step": 13775
    },
    {
      "epoch": 0.44107291466615456,
      "grad_norm": 23.5,
      "learning_rate": 3.6734670093465204e-05,
      "loss": 3.5828,
      "num_input_tokens_seen": 18658432,
      "step": 13780
    },
    {
      "epoch": 0.441232955636643,
      "grad_norm": 28.375,
      "learning_rate": 3.672600042875379e-05,
      "loss": 3.7729,
      "num_input_tokens_seen": 18665200,
      "step": 13785
    },
    {
      "epoch": 0.44139299660713144,
      "grad_norm": 30.125,
      "learning_rate": 3.671732895574575e-05,
      "loss": 4.5053,
      "num_input_tokens_seen": 18672016,
      "step": 13790
    },
    {
      "epoch": 0.4415530375776199,
      "grad_norm": 34.0,
      "learning_rate": 3.670865567577834e-05,
      "loss": 4.978,
      "num_input_tokens_seen": 18678608,
      "step": 13795
    },
    {
      "epoch": 0.4417130785481083,
      "grad_norm": 17.0,
      "learning_rate": 3.669998059018909e-05,
      "loss": 5.2144,
      "num_input_tokens_seen": 18685264,
      "step": 13800
    },
    {
      "epoch": 0.4417130785481083,
      "eval_loss": 4.500638484954834,
      "eval_runtime": 488.7042,
      "eval_samples_per_second": 28.414,
      "eval_steps_per_second": 14.207,
      "num_input_tokens_seen": 18685264,
      "step": 13800
    },
    {
      "epoch": 0.44187311951859676,
      "grad_norm": 30.0,
      "learning_rate": 3.6691303700315796e-05,
      "loss": 4.2627,
      "num_input_tokens_seen": 18692656,
      "step": 13805
    },
    {
      "epoch": 0.4420331604890852,
      "grad_norm": 22.125,
      "learning_rate": 3.668262500749655e-05,
      "loss": 4.4872,
      "num_input_tokens_seen": 18699376,
      "step": 13810
    },
    {
      "epoch": 0.44219320145957364,
      "grad_norm": 29.875,
      "learning_rate": 3.667394451306971e-05,
      "loss": 3.2453,
      "num_input_tokens_seen": 18706224,
      "step": 13815
    },
    {
      "epoch": 0.4423532424300621,
      "grad_norm": 63.0,
      "learning_rate": 3.666526221837393e-05,
      "loss": 3.7071,
      "num_input_tokens_seen": 18713232,
      "step": 13820
    },
    {
      "epoch": 0.4425132834005505,
      "grad_norm": 45.0,
      "learning_rate": 3.665657812474812e-05,
      "loss": 4.7439,
      "num_input_tokens_seen": 18719680,
      "step": 13825
    },
    {
      "epoch": 0.442673324371039,
      "grad_norm": 46.25,
      "learning_rate": 3.664789223353147e-05,
      "loss": 4.3676,
      "num_input_tokens_seen": 18726496,
      "step": 13830
    },
    {
      "epoch": 0.44283336534152745,
      "grad_norm": 32.0,
      "learning_rate": 3.663920454606347e-05,
      "loss": 3.6703,
      "num_input_tokens_seen": 18733216,
      "step": 13835
    },
    {
      "epoch": 0.4429934063120159,
      "grad_norm": 21.25,
      "learning_rate": 3.6630515063683856e-05,
      "loss": 5.2655,
      "num_input_tokens_seen": 18740144,
      "step": 13840
    },
    {
      "epoch": 0.44315344728250433,
      "grad_norm": 27.5,
      "learning_rate": 3.662182378773267e-05,
      "loss": 3.642,
      "num_input_tokens_seen": 18746864,
      "step": 13845
    },
    {
      "epoch": 0.44331348825299277,
      "grad_norm": 31.375,
      "learning_rate": 3.66131307195502e-05,
      "loss": 4.9505,
      "num_input_tokens_seen": 18753744,
      "step": 13850
    },
    {
      "epoch": 0.4434735292234812,
      "grad_norm": 17.0,
      "learning_rate": 3.6604435860477034e-05,
      "loss": 4.4113,
      "num_input_tokens_seen": 18760624,
      "step": 13855
    },
    {
      "epoch": 0.44363357019396965,
      "grad_norm": 33.75,
      "learning_rate": 3.6595739211854025e-05,
      "loss": 4.3632,
      "num_input_tokens_seen": 18767264,
      "step": 13860
    },
    {
      "epoch": 0.4437936111644581,
      "grad_norm": 34.25,
      "learning_rate": 3.658704077502231e-05,
      "loss": 4.9559,
      "num_input_tokens_seen": 18773952,
      "step": 13865
    },
    {
      "epoch": 0.44395365213494653,
      "grad_norm": 46.0,
      "learning_rate": 3.65783405513233e-05,
      "loss": 4.3137,
      "num_input_tokens_seen": 18780336,
      "step": 13870
    },
    {
      "epoch": 0.44411369310543497,
      "grad_norm": 42.5,
      "learning_rate": 3.656963854209867e-05,
      "loss": 4.9218,
      "num_input_tokens_seen": 18786848,
      "step": 13875
    },
    {
      "epoch": 0.4442737340759234,
      "grad_norm": 43.75,
      "learning_rate": 3.656093474869038e-05,
      "loss": 5.7851,
      "num_input_tokens_seen": 18793632,
      "step": 13880
    },
    {
      "epoch": 0.4444337750464119,
      "grad_norm": 21.5,
      "learning_rate": 3.655222917244068e-05,
      "loss": 4.7122,
      "num_input_tokens_seen": 18800512,
      "step": 13885
    },
    {
      "epoch": 0.44459381601690035,
      "grad_norm": 67.5,
      "learning_rate": 3.6543521814692054e-05,
      "loss": 4.2578,
      "num_input_tokens_seen": 18807200,
      "step": 13890
    },
    {
      "epoch": 0.4447538569873888,
      "grad_norm": 31.5,
      "learning_rate": 3.653481267678731e-05,
      "loss": 3.9478,
      "num_input_tokens_seen": 18813856,
      "step": 13895
    },
    {
      "epoch": 0.4449138979578772,
      "grad_norm": 42.5,
      "learning_rate": 3.652610176006949e-05,
      "loss": 3.8594,
      "num_input_tokens_seen": 18820576,
      "step": 13900
    },
    {
      "epoch": 0.44507393892836566,
      "grad_norm": 16.5,
      "learning_rate": 3.6517389065881925e-05,
      "loss": 4.0967,
      "num_input_tokens_seen": 18827184,
      "step": 13905
    },
    {
      "epoch": 0.4452339798988541,
      "grad_norm": 41.25,
      "learning_rate": 3.650867459556824e-05,
      "loss": 5.2302,
      "num_input_tokens_seen": 18833856,
      "step": 13910
    },
    {
      "epoch": 0.44539402086934254,
      "grad_norm": 22.0,
      "learning_rate": 3.64999583504723e-05,
      "loss": 3.3747,
      "num_input_tokens_seen": 18840656,
      "step": 13915
    },
    {
      "epoch": 0.445554061839831,
      "grad_norm": 24.25,
      "learning_rate": 3.649124033193827e-05,
      "loss": 4.576,
      "num_input_tokens_seen": 18847552,
      "step": 13920
    },
    {
      "epoch": 0.4457141028103194,
      "grad_norm": 29.125,
      "learning_rate": 3.648252054131057e-05,
      "loss": 4.7346,
      "num_input_tokens_seen": 18854464,
      "step": 13925
    },
    {
      "epoch": 0.44587414378080786,
      "grad_norm": 19.375,
      "learning_rate": 3.647379897993391e-05,
      "loss": 4.9242,
      "num_input_tokens_seen": 18861568,
      "step": 13930
    },
    {
      "epoch": 0.44603418475129636,
      "grad_norm": 14.5625,
      "learning_rate": 3.646507564915325e-05,
      "loss": 3.5067,
      "num_input_tokens_seen": 18868240,
      "step": 13935
    },
    {
      "epoch": 0.4461942257217848,
      "grad_norm": 16.0,
      "learning_rate": 3.645635055031385e-05,
      "loss": 4.7163,
      "num_input_tokens_seen": 18875376,
      "step": 13940
    },
    {
      "epoch": 0.44635426669227324,
      "grad_norm": 19.5,
      "learning_rate": 3.6447623684761224e-05,
      "loss": 4.2815,
      "num_input_tokens_seen": 18881904,
      "step": 13945
    },
    {
      "epoch": 0.4465143076627617,
      "grad_norm": 31.125,
      "learning_rate": 3.643889505384117e-05,
      "loss": 4.5879,
      "num_input_tokens_seen": 18888608,
      "step": 13950
    },
    {
      "epoch": 0.4466743486332501,
      "grad_norm": 40.25,
      "learning_rate": 3.6430164658899744e-05,
      "loss": 4.5674,
      "num_input_tokens_seen": 18895744,
      "step": 13955
    },
    {
      "epoch": 0.44683438960373856,
      "grad_norm": 37.75,
      "learning_rate": 3.642143250128329e-05,
      "loss": 4.353,
      "num_input_tokens_seen": 18902688,
      "step": 13960
    },
    {
      "epoch": 0.446994430574227,
      "grad_norm": 31.625,
      "learning_rate": 3.641269858233841e-05,
      "loss": 4.6201,
      "num_input_tokens_seen": 18909280,
      "step": 13965
    },
    {
      "epoch": 0.44715447154471544,
      "grad_norm": 32.75,
      "learning_rate": 3.640396290341199e-05,
      "loss": 3.7696,
      "num_input_tokens_seen": 18916208,
      "step": 13970
    },
    {
      "epoch": 0.4473145125152039,
      "grad_norm": 29.625,
      "learning_rate": 3.639522546585118e-05,
      "loss": 4.2905,
      "num_input_tokens_seen": 18923120,
      "step": 13975
    },
    {
      "epoch": 0.4474745534856923,
      "grad_norm": 27.875,
      "learning_rate": 3.6386486271003404e-05,
      "loss": 4.1303,
      "num_input_tokens_seen": 18930256,
      "step": 13980
    },
    {
      "epoch": 0.44763459445618076,
      "grad_norm": 20.625,
      "learning_rate": 3.6377745320216346e-05,
      "loss": 4.4566,
      "num_input_tokens_seen": 18936880,
      "step": 13985
    },
    {
      "epoch": 0.44779463542666925,
      "grad_norm": 16.875,
      "learning_rate": 3.636900261483798e-05,
      "loss": 3.7134,
      "num_input_tokens_seen": 18943552,
      "step": 13990
    },
    {
      "epoch": 0.4479546763971577,
      "grad_norm": 35.5,
      "learning_rate": 3.636025815621654e-05,
      "loss": 4.1333,
      "num_input_tokens_seen": 18950176,
      "step": 13995
    },
    {
      "epoch": 0.44811471736764613,
      "grad_norm": 34.75,
      "learning_rate": 3.635151194570054e-05,
      "loss": 4.1907,
      "num_input_tokens_seen": 18957072,
      "step": 14000
    },
    {
      "epoch": 0.44811471736764613,
      "eval_loss": 4.47998046875,
      "eval_runtime": 489.0,
      "eval_samples_per_second": 28.397,
      "eval_steps_per_second": 14.198,
      "num_input_tokens_seen": 18957072,
      "step": 14000
    },
    {
      "epoch": 0.44827475833813457,
      "grad_norm": 20.5,
      "learning_rate": 3.634276398463873e-05,
      "loss": 4.338,
      "num_input_tokens_seen": 18963792,
      "step": 14005
    },
    {
      "epoch": 0.448434799308623,
      "grad_norm": 30.5,
      "learning_rate": 3.633401427438018e-05,
      "loss": 3.5303,
      "num_input_tokens_seen": 18970752,
      "step": 14010
    },
    {
      "epoch": 0.44859484027911145,
      "grad_norm": 23.0,
      "learning_rate": 3.63252628162742e-05,
      "loss": 4.0246,
      "num_input_tokens_seen": 18977424,
      "step": 14015
    },
    {
      "epoch": 0.4487548812495999,
      "grad_norm": 34.25,
      "learning_rate": 3.6316509611670364e-05,
      "loss": 4.7756,
      "num_input_tokens_seen": 18984224,
      "step": 14020
    },
    {
      "epoch": 0.44891492222008833,
      "grad_norm": 31.875,
      "learning_rate": 3.630775466191854e-05,
      "loss": 4.8522,
      "num_input_tokens_seen": 18990784,
      "step": 14025
    },
    {
      "epoch": 0.44907496319057677,
      "grad_norm": 31.875,
      "learning_rate": 3.629899796836884e-05,
      "loss": 4.3628,
      "num_input_tokens_seen": 18997488,
      "step": 14030
    },
    {
      "epoch": 0.4492350041610652,
      "grad_norm": 23.375,
      "learning_rate": 3.6290239532371666e-05,
      "loss": 4.2516,
      "num_input_tokens_seen": 19004416,
      "step": 14035
    },
    {
      "epoch": 0.4493950451315537,
      "grad_norm": 18.625,
      "learning_rate": 3.628147935527767e-05,
      "loss": 5.0091,
      "num_input_tokens_seen": 19011600,
      "step": 14040
    },
    {
      "epoch": 0.44955508610204215,
      "grad_norm": 22.375,
      "learning_rate": 3.627271743843779e-05,
      "loss": 3.5649,
      "num_input_tokens_seen": 19018304,
      "step": 14045
    },
    {
      "epoch": 0.4497151270725306,
      "grad_norm": 21.0,
      "learning_rate": 3.626395378320321e-05,
      "loss": 4.3021,
      "num_input_tokens_seen": 19024992,
      "step": 14050
    },
    {
      "epoch": 0.449875168043019,
      "grad_norm": 18.625,
      "learning_rate": 3.625518839092541e-05,
      "loss": 4.1199,
      "num_input_tokens_seen": 19031856,
      "step": 14055
    },
    {
      "epoch": 0.45003520901350746,
      "grad_norm": 44.0,
      "learning_rate": 3.624642126295612e-05,
      "loss": 4.5845,
      "num_input_tokens_seen": 19038432,
      "step": 14060
    },
    {
      "epoch": 0.4501952499839959,
      "grad_norm": 33.0,
      "learning_rate": 3.6237652400647345e-05,
      "loss": 4.4019,
      "num_input_tokens_seen": 19045264,
      "step": 14065
    },
    {
      "epoch": 0.45035529095448434,
      "grad_norm": 48.5,
      "learning_rate": 3.622888180535134e-05,
      "loss": 4.5204,
      "num_input_tokens_seen": 19052496,
      "step": 14070
    },
    {
      "epoch": 0.4505153319249728,
      "grad_norm": 28.25,
      "learning_rate": 3.6220109478420655e-05,
      "loss": 3.3667,
      "num_input_tokens_seen": 19059232,
      "step": 14075
    },
    {
      "epoch": 0.4506753728954612,
      "grad_norm": 26.5,
      "learning_rate": 3.6211335421208084e-05,
      "loss": 4.5394,
      "num_input_tokens_seen": 19065952,
      "step": 14080
    },
    {
      "epoch": 0.45083541386594966,
      "grad_norm": 25.0,
      "learning_rate": 3.62025596350667e-05,
      "loss": 4.1287,
      "num_input_tokens_seen": 19072384,
      "step": 14085
    },
    {
      "epoch": 0.4509954548364381,
      "grad_norm": 34.25,
      "learning_rate": 3.619378212134984e-05,
      "loss": 3.8279,
      "num_input_tokens_seen": 19079440,
      "step": 14090
    },
    {
      "epoch": 0.4511554958069266,
      "grad_norm": 19.0,
      "learning_rate": 3.618500288141111e-05,
      "loss": 3.7192,
      "num_input_tokens_seen": 19086112,
      "step": 14095
    },
    {
      "epoch": 0.45131553677741504,
      "grad_norm": 35.5,
      "learning_rate": 3.617622191660438e-05,
      "loss": 5.0445,
      "num_input_tokens_seen": 19093264,
      "step": 14100
    },
    {
      "epoch": 0.4514755777479035,
      "grad_norm": 22.5,
      "learning_rate": 3.616743922828377e-05,
      "loss": 3.2116,
      "num_input_tokens_seen": 19100560,
      "step": 14105
    },
    {
      "epoch": 0.4516356187183919,
      "grad_norm": 26.75,
      "learning_rate": 3.615865481780371e-05,
      "loss": 3.6621,
      "num_input_tokens_seen": 19107344,
      "step": 14110
    },
    {
      "epoch": 0.45179565968888036,
      "grad_norm": 41.0,
      "learning_rate": 3.614986868651883e-05,
      "loss": 4.8295,
      "num_input_tokens_seen": 19114064,
      "step": 14115
    },
    {
      "epoch": 0.4519557006593688,
      "grad_norm": 26.5,
      "learning_rate": 3.614108083578409e-05,
      "loss": 4.9366,
      "num_input_tokens_seen": 19120512,
      "step": 14120
    },
    {
      "epoch": 0.45211574162985724,
      "grad_norm": 28.125,
      "learning_rate": 3.613229126695467e-05,
      "loss": 4.1651,
      "num_input_tokens_seen": 19127584,
      "step": 14125
    },
    {
      "epoch": 0.4522757826003457,
      "grad_norm": 25.375,
      "learning_rate": 3.612349998138605e-05,
      "loss": 3.959,
      "num_input_tokens_seen": 19134320,
      "step": 14130
    },
    {
      "epoch": 0.4524358235708341,
      "grad_norm": 40.0,
      "learning_rate": 3.6114706980433946e-05,
      "loss": 4.6354,
      "num_input_tokens_seen": 19141056,
      "step": 14135
    },
    {
      "epoch": 0.45259586454132256,
      "grad_norm": 35.75,
      "learning_rate": 3.610591226545435e-05,
      "loss": 4.0044,
      "num_input_tokens_seen": 19148144,
      "step": 14140
    },
    {
      "epoch": 0.452755905511811,
      "grad_norm": 33.75,
      "learning_rate": 3.6097115837803505e-05,
      "loss": 4.6913,
      "num_input_tokens_seen": 19154912,
      "step": 14145
    },
    {
      "epoch": 0.4529159464822995,
      "grad_norm": 35.25,
      "learning_rate": 3.608831769883795e-05,
      "loss": 4.9325,
      "num_input_tokens_seen": 19162224,
      "step": 14150
    },
    {
      "epoch": 0.45307598745278793,
      "grad_norm": 38.5,
      "learning_rate": 3.607951784991446e-05,
      "loss": 4.365,
      "num_input_tokens_seen": 19169152,
      "step": 14155
    },
    {
      "epoch": 0.45323602842327637,
      "grad_norm": 17.625,
      "learning_rate": 3.6070716292390085e-05,
      "loss": 3.3995,
      "num_input_tokens_seen": 19175648,
      "step": 14160
    },
    {
      "epoch": 0.4533960693937648,
      "grad_norm": 22.625,
      "learning_rate": 3.606191302762213e-05,
      "loss": 5.3084,
      "num_input_tokens_seen": 19182048,
      "step": 14165
    },
    {
      "epoch": 0.45355611036425325,
      "grad_norm": 57.25,
      "learning_rate": 3.605310805696818e-05,
      "loss": 4.0265,
      "num_input_tokens_seen": 19188704,
      "step": 14170
    },
    {
      "epoch": 0.4537161513347417,
      "grad_norm": 18.625,
      "learning_rate": 3.6044301381786067e-05,
      "loss": 3.6661,
      "num_input_tokens_seen": 19195408,
      "step": 14175
    },
    {
      "epoch": 0.45387619230523013,
      "grad_norm": 29.25,
      "learning_rate": 3.6035493003433883e-05,
      "loss": 4.5601,
      "num_input_tokens_seen": 19202256,
      "step": 14180
    },
    {
      "epoch": 0.45403623327571857,
      "grad_norm": 46.25,
      "learning_rate": 3.6026682923269994e-05,
      "loss": 4.9834,
      "num_input_tokens_seen": 19208992,
      "step": 14185
    },
    {
      "epoch": 0.454196274246207,
      "grad_norm": 16.875,
      "learning_rate": 3.6017871142653034e-05,
      "loss": 4.3118,
      "num_input_tokens_seen": 19216208,
      "step": 14190
    },
    {
      "epoch": 0.45435631521669545,
      "grad_norm": 31.25,
      "learning_rate": 3.600905766294189e-05,
      "loss": 3.8185,
      "num_input_tokens_seen": 19223328,
      "step": 14195
    },
    {
      "epoch": 0.45451635618718395,
      "grad_norm": 15.5,
      "learning_rate": 3.60002424854957e-05,
      "loss": 3.1513,
      "num_input_tokens_seen": 19230480,
      "step": 14200
    },
    {
      "epoch": 0.45451635618718395,
      "eval_loss": 4.492216110229492,
      "eval_runtime": 488.1645,
      "eval_samples_per_second": 28.445,
      "eval_steps_per_second": 14.223,
      "num_input_tokens_seen": 19230480,
      "step": 14200
    },
    {
      "epoch": 0.4546763971576724,
      "grad_norm": 58.25,
      "learning_rate": 3.5991425611673876e-05,
      "loss": 4.5467,
      "num_input_tokens_seen": 19236976,
      "step": 14205
    },
    {
      "epoch": 0.4548364381281608,
      "grad_norm": 39.0,
      "learning_rate": 3.5982607042836105e-05,
      "loss": 4.8839,
      "num_input_tokens_seen": 19243616,
      "step": 14210
    },
    {
      "epoch": 0.45499647909864926,
      "grad_norm": 30.0,
      "learning_rate": 3.597378678034231e-05,
      "loss": 4.39,
      "num_input_tokens_seen": 19250352,
      "step": 14215
    },
    {
      "epoch": 0.4551565200691377,
      "grad_norm": 22.125,
      "learning_rate": 3.596496482555269e-05,
      "loss": 3.8356,
      "num_input_tokens_seen": 19257328,
      "step": 14220
    },
    {
      "epoch": 0.45531656103962614,
      "grad_norm": 42.5,
      "learning_rate": 3.595614117982769e-05,
      "loss": 4.2979,
      "num_input_tokens_seen": 19264160,
      "step": 14225
    },
    {
      "epoch": 0.4554766020101146,
      "grad_norm": 21.0,
      "learning_rate": 3.594731584452805e-05,
      "loss": 4.0614,
      "num_input_tokens_seen": 19271216,
      "step": 14230
    },
    {
      "epoch": 0.455636642980603,
      "grad_norm": 33.25,
      "learning_rate": 3.593848882101472e-05,
      "loss": 4.3749,
      "num_input_tokens_seen": 19277984,
      "step": 14235
    },
    {
      "epoch": 0.45579668395109146,
      "grad_norm": 33.25,
      "learning_rate": 3.592966011064896e-05,
      "loss": 4.3798,
      "num_input_tokens_seen": 19285040,
      "step": 14240
    },
    {
      "epoch": 0.4559567249215799,
      "grad_norm": 28.625,
      "learning_rate": 3.592082971479226e-05,
      "loss": 3.5438,
      "num_input_tokens_seen": 19292352,
      "step": 14245
    },
    {
      "epoch": 0.45611676589206834,
      "grad_norm": 21.375,
      "learning_rate": 3.5911997634806385e-05,
      "loss": 4.9574,
      "num_input_tokens_seen": 19299104,
      "step": 14250
    },
    {
      "epoch": 0.45627680686255684,
      "grad_norm": 52.25,
      "learning_rate": 3.5903163872053336e-05,
      "loss": 5.2022,
      "num_input_tokens_seen": 19305936,
      "step": 14255
    },
    {
      "epoch": 0.4564368478330453,
      "grad_norm": 40.5,
      "learning_rate": 3.58943284278954e-05,
      "loss": 4.4944,
      "num_input_tokens_seen": 19312208,
      "step": 14260
    },
    {
      "epoch": 0.4565968888035337,
      "grad_norm": 19.625,
      "learning_rate": 3.588549130369512e-05,
      "loss": 3.6944,
      "num_input_tokens_seen": 19319232,
      "step": 14265
    },
    {
      "epoch": 0.45675692977402216,
      "grad_norm": 18.5,
      "learning_rate": 3.5876652500815274e-05,
      "loss": 3.7342,
      "num_input_tokens_seen": 19326144,
      "step": 14270
    },
    {
      "epoch": 0.4569169707445106,
      "grad_norm": 31.625,
      "learning_rate": 3.586781202061894e-05,
      "loss": 4.2798,
      "num_input_tokens_seen": 19332928,
      "step": 14275
    },
    {
      "epoch": 0.45707701171499904,
      "grad_norm": 19.5,
      "learning_rate": 3.585896986446942e-05,
      "loss": 3.6308,
      "num_input_tokens_seen": 19339472,
      "step": 14280
    },
    {
      "epoch": 0.4572370526854875,
      "grad_norm": 38.25,
      "learning_rate": 3.585012603373028e-05,
      "loss": 5.3554,
      "num_input_tokens_seen": 19346144,
      "step": 14285
    },
    {
      "epoch": 0.4573970936559759,
      "grad_norm": 35.0,
      "learning_rate": 3.584128052976535e-05,
      "loss": 6.2308,
      "num_input_tokens_seen": 19352704,
      "step": 14290
    },
    {
      "epoch": 0.45755713462646436,
      "grad_norm": 12.9375,
      "learning_rate": 3.5832433353938724e-05,
      "loss": 3.545,
      "num_input_tokens_seen": 19359440,
      "step": 14295
    },
    {
      "epoch": 0.4577171755969528,
      "grad_norm": 19.875,
      "learning_rate": 3.5823584507614746e-05,
      "loss": 4.2241,
      "num_input_tokens_seen": 19366320,
      "step": 14300
    },
    {
      "epoch": 0.4578772165674413,
      "grad_norm": 18.25,
      "learning_rate": 3.581473399215802e-05,
      "loss": 3.4645,
      "num_input_tokens_seen": 19373264,
      "step": 14305
    },
    {
      "epoch": 0.45803725753792973,
      "grad_norm": 29.0,
      "learning_rate": 3.580588180893341e-05,
      "loss": 4.6629,
      "num_input_tokens_seen": 19380272,
      "step": 14310
    },
    {
      "epoch": 0.45819729850841817,
      "grad_norm": 20.875,
      "learning_rate": 3.579702795930602e-05,
      "loss": 4.3165,
      "num_input_tokens_seen": 19387360,
      "step": 14315
    },
    {
      "epoch": 0.4583573394789066,
      "grad_norm": 51.75,
      "learning_rate": 3.578817244464125e-05,
      "loss": 5.0749,
      "num_input_tokens_seen": 19394272,
      "step": 14320
    },
    {
      "epoch": 0.45851738044939505,
      "grad_norm": 24.875,
      "learning_rate": 3.577931526630471e-05,
      "loss": 3.986,
      "num_input_tokens_seen": 19401664,
      "step": 14325
    },
    {
      "epoch": 0.4586774214198835,
      "grad_norm": 36.5,
      "learning_rate": 3.577045642566229e-05,
      "loss": 4.3603,
      "num_input_tokens_seen": 19408144,
      "step": 14330
    },
    {
      "epoch": 0.45883746239037193,
      "grad_norm": 21.75,
      "learning_rate": 3.576159592408014e-05,
      "loss": 4.6064,
      "num_input_tokens_seen": 19414864,
      "step": 14335
    },
    {
      "epoch": 0.45899750336086037,
      "grad_norm": 14.75,
      "learning_rate": 3.575273376292466e-05,
      "loss": 4.2312,
      "num_input_tokens_seen": 19421104,
      "step": 14340
    },
    {
      "epoch": 0.4591575443313488,
      "grad_norm": 19.75,
      "learning_rate": 3.574386994356251e-05,
      "loss": 4.2995,
      "num_input_tokens_seen": 19427888,
      "step": 14345
    },
    {
      "epoch": 0.45931758530183725,
      "grad_norm": 34.75,
      "learning_rate": 3.573500446736059e-05,
      "loss": 4.8787,
      "num_input_tokens_seen": 19434288,
      "step": 14350
    },
    {
      "epoch": 0.4594776262723257,
      "grad_norm": 46.0,
      "learning_rate": 3.5726137335686094e-05,
      "loss": 4.4157,
      "num_input_tokens_seen": 19441536,
      "step": 14355
    },
    {
      "epoch": 0.4596376672428142,
      "grad_norm": 33.25,
      "learning_rate": 3.571726854990642e-05,
      "loss": 4.1717,
      "num_input_tokens_seen": 19448624,
      "step": 14360
    },
    {
      "epoch": 0.4597977082133026,
      "grad_norm": 23.875,
      "learning_rate": 3.570839811138925e-05,
      "loss": 4.0295,
      "num_input_tokens_seen": 19455296,
      "step": 14365
    },
    {
      "epoch": 0.45995774918379106,
      "grad_norm": 12.8125,
      "learning_rate": 3.569952602150252e-05,
      "loss": 4.5776,
      "num_input_tokens_seen": 19462112,
      "step": 14370
    },
    {
      "epoch": 0.4601177901542795,
      "grad_norm": 26.625,
      "learning_rate": 3.569065228161442e-05,
      "loss": 4.3348,
      "num_input_tokens_seen": 19468816,
      "step": 14375
    },
    {
      "epoch": 0.46027783112476794,
      "grad_norm": 20.5,
      "learning_rate": 3.5681776893093395e-05,
      "loss": 3.9178,
      "num_input_tokens_seen": 19475776,
      "step": 14380
    },
    {
      "epoch": 0.4604378720952564,
      "grad_norm": 41.0,
      "learning_rate": 3.5672899857308134e-05,
      "loss": 5.0605,
      "num_input_tokens_seen": 19482224,
      "step": 14385
    },
    {
      "epoch": 0.4605979130657448,
      "grad_norm": 25.5,
      "learning_rate": 3.566402117562759e-05,
      "loss": 3.7848,
      "num_input_tokens_seen": 19489312,
      "step": 14390
    },
    {
      "epoch": 0.46075795403623326,
      "grad_norm": 15.375,
      "learning_rate": 3.565514084942097e-05,
      "loss": 4.3443,
      "num_input_tokens_seen": 19496576,
      "step": 14395
    },
    {
      "epoch": 0.4609179950067217,
      "grad_norm": 16.25,
      "learning_rate": 3.564625888005773e-05,
      "loss": 3.8682,
      "num_input_tokens_seen": 19503472,
      "step": 14400
    },
    {
      "epoch": 0.4609179950067217,
      "eval_loss": 4.497103691101074,
      "eval_runtime": 488.8126,
      "eval_samples_per_second": 28.408,
      "eval_steps_per_second": 14.204,
      "num_input_tokens_seen": 19503472,
      "step": 14400
    },
    {
      "epoch": 0.46107803597721014,
      "grad_norm": 29.25,
      "learning_rate": 3.563737526890759e-05,
      "loss": 3.6065,
      "num_input_tokens_seen": 19510016,
      "step": 14405
    },
    {
      "epoch": 0.4612380769476986,
      "grad_norm": 29.75,
      "learning_rate": 3.562849001734049e-05,
      "loss": 4.1929,
      "num_input_tokens_seen": 19516832,
      "step": 14410
    },
    {
      "epoch": 0.4613981179181871,
      "grad_norm": 47.75,
      "learning_rate": 3.561960312672667e-05,
      "loss": 5.1266,
      "num_input_tokens_seen": 19523152,
      "step": 14415
    },
    {
      "epoch": 0.4615581588886755,
      "grad_norm": 26.5,
      "learning_rate": 3.5610714598436596e-05,
      "loss": 4.4146,
      "num_input_tokens_seen": 19529904,
      "step": 14420
    },
    {
      "epoch": 0.46171819985916396,
      "grad_norm": 28.375,
      "learning_rate": 3.5601824433840986e-05,
      "loss": 4.7417,
      "num_input_tokens_seen": 19536592,
      "step": 14425
    },
    {
      "epoch": 0.4618782408296524,
      "grad_norm": 35.0,
      "learning_rate": 3.559293263431082e-05,
      "loss": 5.2575,
      "num_input_tokens_seen": 19543104,
      "step": 14430
    },
    {
      "epoch": 0.46203828180014084,
      "grad_norm": 38.25,
      "learning_rate": 3.558403920121732e-05,
      "loss": 4.038,
      "num_input_tokens_seen": 19549824,
      "step": 14435
    },
    {
      "epoch": 0.4621983227706293,
      "grad_norm": 29.0,
      "learning_rate": 3.557514413593197e-05,
      "loss": 4.3776,
      "num_input_tokens_seen": 19556560,
      "step": 14440
    },
    {
      "epoch": 0.4623583637411177,
      "grad_norm": 25.75,
      "learning_rate": 3.55662474398265e-05,
      "loss": 4.4046,
      "num_input_tokens_seen": 19563600,
      "step": 14445
    },
    {
      "epoch": 0.46251840471160616,
      "grad_norm": 27.5,
      "learning_rate": 3.555734911427288e-05,
      "loss": 3.8154,
      "num_input_tokens_seen": 19570272,
      "step": 14450
    },
    {
      "epoch": 0.4626784456820946,
      "grad_norm": 27.375,
      "learning_rate": 3.5548449160643363e-05,
      "loss": 3.5611,
      "num_input_tokens_seen": 19577136,
      "step": 14455
    },
    {
      "epoch": 0.46283848665258304,
      "grad_norm": 38.0,
      "learning_rate": 3.553954758031043e-05,
      "loss": 4.7058,
      "num_input_tokens_seen": 19584016,
      "step": 14460
    },
    {
      "epoch": 0.46299852762307153,
      "grad_norm": 25.625,
      "learning_rate": 3.5530644374646815e-05,
      "loss": 4.7671,
      "num_input_tokens_seen": 19590640,
      "step": 14465
    },
    {
      "epoch": 0.46315856859355997,
      "grad_norm": 24.125,
      "learning_rate": 3.552173954502549e-05,
      "loss": 3.4247,
      "num_input_tokens_seen": 19597296,
      "step": 14470
    },
    {
      "epoch": 0.4633186095640484,
      "grad_norm": 34.5,
      "learning_rate": 3.55128330928197e-05,
      "loss": 5.3097,
      "num_input_tokens_seen": 19603776,
      "step": 14475
    },
    {
      "epoch": 0.46347865053453685,
      "grad_norm": 49.5,
      "learning_rate": 3.550392501940294e-05,
      "loss": 4.2384,
      "num_input_tokens_seen": 19610320,
      "step": 14480
    },
    {
      "epoch": 0.4636386915050253,
      "grad_norm": 31.25,
      "learning_rate": 3.5495015326148945e-05,
      "loss": 3.7788,
      "num_input_tokens_seen": 19617168,
      "step": 14485
    },
    {
      "epoch": 0.46379873247551373,
      "grad_norm": 23.875,
      "learning_rate": 3.548610401443169e-05,
      "loss": 3.7297,
      "num_input_tokens_seen": 19623936,
      "step": 14490
    },
    {
      "epoch": 0.46395877344600217,
      "grad_norm": 32.25,
      "learning_rate": 3.547719108562543e-05,
      "loss": 4.5306,
      "num_input_tokens_seen": 19630512,
      "step": 14495
    },
    {
      "epoch": 0.4641188144164906,
      "grad_norm": 30.25,
      "learning_rate": 3.546827654110464e-05,
      "loss": 3.7587,
      "num_input_tokens_seen": 19637008,
      "step": 14500
    },
    {
      "epoch": 0.46427885538697905,
      "grad_norm": 26.625,
      "learning_rate": 3.545936038224405e-05,
      "loss": 3.2092,
      "num_input_tokens_seen": 19643712,
      "step": 14505
    },
    {
      "epoch": 0.4644388963574675,
      "grad_norm": 41.75,
      "learning_rate": 3.545044261041864e-05,
      "loss": 3.729,
      "num_input_tokens_seen": 19650496,
      "step": 14510
    },
    {
      "epoch": 0.46459893732795593,
      "grad_norm": 15.875,
      "learning_rate": 3.5441523227003657e-05,
      "loss": 4.0368,
      "num_input_tokens_seen": 19657840,
      "step": 14515
    },
    {
      "epoch": 0.4647589782984444,
      "grad_norm": 40.0,
      "learning_rate": 3.543260223337459e-05,
      "loss": 4.6196,
      "num_input_tokens_seen": 19664528,
      "step": 14520
    },
    {
      "epoch": 0.46491901926893286,
      "grad_norm": 38.5,
      "learning_rate": 3.542367963090714e-05,
      "loss": 4.9372,
      "num_input_tokens_seen": 19671488,
      "step": 14525
    },
    {
      "epoch": 0.4650790602394213,
      "grad_norm": 21.5,
      "learning_rate": 3.5414755420977295e-05,
      "loss": 4.7042,
      "num_input_tokens_seen": 19678384,
      "step": 14530
    },
    {
      "epoch": 0.46523910120990974,
      "grad_norm": 36.0,
      "learning_rate": 3.54058296049613e-05,
      "loss": 4.4965,
      "num_input_tokens_seen": 19685504,
      "step": 14535
    },
    {
      "epoch": 0.4653991421803982,
      "grad_norm": 33.75,
      "learning_rate": 3.53969021842356e-05,
      "loss": 4.2514,
      "num_input_tokens_seen": 19692192,
      "step": 14540
    },
    {
      "epoch": 0.4655591831508866,
      "grad_norm": 23.0,
      "learning_rate": 3.5387973160176926e-05,
      "loss": 3.9897,
      "num_input_tokens_seen": 19698880,
      "step": 14545
    },
    {
      "epoch": 0.46571922412137506,
      "grad_norm": 10.25,
      "learning_rate": 3.537904253416224e-05,
      "loss": 3.798,
      "num_input_tokens_seen": 19705776,
      "step": 14550
    },
    {
      "epoch": 0.4658792650918635,
      "grad_norm": 26.0,
      "learning_rate": 3.537011030756878e-05,
      "loss": 5.3979,
      "num_input_tokens_seen": 19715952,
      "step": 14555
    },
    {
      "epoch": 0.46603930606235194,
      "grad_norm": 31.25,
      "learning_rate": 3.536117648177399e-05,
      "loss": 5.0579,
      "num_input_tokens_seen": 19722896,
      "step": 14560
    },
    {
      "epoch": 0.4661993470328404,
      "grad_norm": 25.75,
      "learning_rate": 3.535224105815558e-05,
      "loss": 4.2862,
      "num_input_tokens_seen": 19729648,
      "step": 14565
    },
    {
      "epoch": 0.4663593880033289,
      "grad_norm": 27.125,
      "learning_rate": 3.5343304038091494e-05,
      "loss": 4.2568,
      "num_input_tokens_seen": 19736368,
      "step": 14570
    },
    {
      "epoch": 0.4665194289738173,
      "grad_norm": 26.5,
      "learning_rate": 3.5334365422959955e-05,
      "loss": 5.0616,
      "num_input_tokens_seen": 19743152,
      "step": 14575
    },
    {
      "epoch": 0.46667946994430576,
      "grad_norm": 19.625,
      "learning_rate": 3.5325425214139396e-05,
      "loss": 4.9137,
      "num_input_tokens_seen": 19749872,
      "step": 14580
    },
    {
      "epoch": 0.4668395109147942,
      "grad_norm": 17.875,
      "learning_rate": 3.531648341300851e-05,
      "loss": 5.1983,
      "num_input_tokens_seen": 19756864,
      "step": 14585
    },
    {
      "epoch": 0.46699955188528264,
      "grad_norm": 23.25,
      "learning_rate": 3.530754002094623e-05,
      "loss": 4.469,
      "num_input_tokens_seen": 19763648,
      "step": 14590
    },
    {
      "epoch": 0.4671595928557711,
      "grad_norm": 28.375,
      "learning_rate": 3.529859503933175e-05,
      "loss": 4.1055,
      "num_input_tokens_seen": 19770544,
      "step": 14595
    },
    {
      "epoch": 0.4673196338262595,
      "grad_norm": 21.0,
      "learning_rate": 3.52896484695445e-05,
      "loss": 4.9699,
      "num_input_tokens_seen": 19777344,
      "step": 14600
    },
    {
      "epoch": 0.4673196338262595,
      "eval_loss": 4.495926856994629,
      "eval_runtime": 489.682,
      "eval_samples_per_second": 28.357,
      "eval_steps_per_second": 14.179,
      "num_input_tokens_seen": 19777344,
      "step": 14600
    },
    {
      "epoch": 0.46747967479674796,
      "grad_norm": 25.125,
      "learning_rate": 3.528070031296414e-05,
      "loss": 4.7956,
      "num_input_tokens_seen": 19784496,
      "step": 14605
    },
    {
      "epoch": 0.4676397157672364,
      "grad_norm": 23.0,
      "learning_rate": 3.5271750570970605e-05,
      "loss": 4.3659,
      "num_input_tokens_seen": 19791520,
      "step": 14610
    },
    {
      "epoch": 0.46779975673772484,
      "grad_norm": 20.25,
      "learning_rate": 3.526279924494405e-05,
      "loss": 4.7454,
      "num_input_tokens_seen": 19798368,
      "step": 14615
    },
    {
      "epoch": 0.4679597977082133,
      "grad_norm": 25.5,
      "learning_rate": 3.5253846336264874e-05,
      "loss": 3.5013,
      "num_input_tokens_seen": 19805152,
      "step": 14620
    },
    {
      "epoch": 0.46811983867870177,
      "grad_norm": 22.125,
      "learning_rate": 3.5244891846313736e-05,
      "loss": 4.2612,
      "num_input_tokens_seen": 19811792,
      "step": 14625
    },
    {
      "epoch": 0.4682798796491902,
      "grad_norm": 22.625,
      "learning_rate": 3.5235935776471527e-05,
      "loss": 4.0235,
      "num_input_tokens_seen": 19818384,
      "step": 14630
    },
    {
      "epoch": 0.46843992061967865,
      "grad_norm": 25.875,
      "learning_rate": 3.522697812811939e-05,
      "loss": 3.6373,
      "num_input_tokens_seen": 19825360,
      "step": 14635
    },
    {
      "epoch": 0.4685999615901671,
      "grad_norm": 36.25,
      "learning_rate": 3.521801890263871e-05,
      "loss": 5.1555,
      "num_input_tokens_seen": 19831968,
      "step": 14640
    },
    {
      "epoch": 0.46876000256065553,
      "grad_norm": 56.5,
      "learning_rate": 3.5209058101411114e-05,
      "loss": 5.1393,
      "num_input_tokens_seen": 19838928,
      "step": 14645
    },
    {
      "epoch": 0.46892004353114397,
      "grad_norm": 20.0,
      "learning_rate": 3.520009572581845e-05,
      "loss": 4.2217,
      "num_input_tokens_seen": 19845696,
      "step": 14650
    },
    {
      "epoch": 0.4690800845016324,
      "grad_norm": 31.375,
      "learning_rate": 3.519113177724285e-05,
      "loss": 2.9596,
      "num_input_tokens_seen": 19852704,
      "step": 14655
    },
    {
      "epoch": 0.46924012547212085,
      "grad_norm": 23.5,
      "learning_rate": 3.5182166257066656e-05,
      "loss": 4.3596,
      "num_input_tokens_seen": 19859872,
      "step": 14660
    },
    {
      "epoch": 0.4694001664426093,
      "grad_norm": 33.25,
      "learning_rate": 3.517319916667247e-05,
      "loss": 4.4291,
      "num_input_tokens_seen": 19866464,
      "step": 14665
    },
    {
      "epoch": 0.46956020741309773,
      "grad_norm": 20.0,
      "learning_rate": 3.516423050744313e-05,
      "loss": 3.8319,
      "num_input_tokens_seen": 19873328,
      "step": 14670
    },
    {
      "epoch": 0.4697202483835862,
      "grad_norm": 28.75,
      "learning_rate": 3.5155260280761704e-05,
      "loss": 5.2626,
      "num_input_tokens_seen": 19879936,
      "step": 14675
    },
    {
      "epoch": 0.46988028935407467,
      "grad_norm": 29.0,
      "learning_rate": 3.514628848801154e-05,
      "loss": 4.0089,
      "num_input_tokens_seen": 19886640,
      "step": 14680
    },
    {
      "epoch": 0.4700403303245631,
      "grad_norm": 39.75,
      "learning_rate": 3.5137315130576174e-05,
      "loss": 4.1235,
      "num_input_tokens_seen": 19893696,
      "step": 14685
    },
    {
      "epoch": 0.47020037129505154,
      "grad_norm": 30.875,
      "learning_rate": 3.512834020983942e-05,
      "loss": 3.663,
      "num_input_tokens_seen": 19900224,
      "step": 14690
    },
    {
      "epoch": 0.47036041226554,
      "grad_norm": 33.0,
      "learning_rate": 3.5119363727185334e-05,
      "loss": 4.7133,
      "num_input_tokens_seen": 19906688,
      "step": 14695
    },
    {
      "epoch": 0.4705204532360284,
      "grad_norm": 32.5,
      "learning_rate": 3.511038568399819e-05,
      "loss": 4.7823,
      "num_input_tokens_seen": 19913728,
      "step": 14700
    },
    {
      "epoch": 0.47068049420651686,
      "grad_norm": 17.75,
      "learning_rate": 3.510140608166251e-05,
      "loss": 4.9328,
      "num_input_tokens_seen": 19920896,
      "step": 14705
    },
    {
      "epoch": 0.4708405351770053,
      "grad_norm": 30.5,
      "learning_rate": 3.509242492156308e-05,
      "loss": 4.379,
      "num_input_tokens_seen": 19927584,
      "step": 14710
    },
    {
      "epoch": 0.47100057614749374,
      "grad_norm": 20.875,
      "learning_rate": 3.5083442205084896e-05,
      "loss": 3.8958,
      "num_input_tokens_seen": 19933904,
      "step": 14715
    },
    {
      "epoch": 0.4711606171179822,
      "grad_norm": 22.25,
      "learning_rate": 3.507445793361321e-05,
      "loss": 3.0661,
      "num_input_tokens_seen": 19940752,
      "step": 14720
    },
    {
      "epoch": 0.4713206580884706,
      "grad_norm": 23.75,
      "learning_rate": 3.5065472108533505e-05,
      "loss": 3.9015,
      "num_input_tokens_seen": 19947408,
      "step": 14725
    },
    {
      "epoch": 0.4714806990589591,
      "grad_norm": 21.25,
      "learning_rate": 3.5056484731231504e-05,
      "loss": 3.6214,
      "num_input_tokens_seen": 19954304,
      "step": 14730
    },
    {
      "epoch": 0.47164074002944756,
      "grad_norm": 27.875,
      "learning_rate": 3.504749580309319e-05,
      "loss": 4.6266,
      "num_input_tokens_seen": 19961440,
      "step": 14735
    },
    {
      "epoch": 0.471800780999936,
      "grad_norm": 38.25,
      "learning_rate": 3.5038505325504753e-05,
      "loss": 3.6919,
      "num_input_tokens_seen": 19968208,
      "step": 14740
    },
    {
      "epoch": 0.47196082197042444,
      "grad_norm": 51.5,
      "learning_rate": 3.502951329985264e-05,
      "loss": 4.2439,
      "num_input_tokens_seen": 19974960,
      "step": 14745
    },
    {
      "epoch": 0.4721208629409129,
      "grad_norm": 34.0,
      "learning_rate": 3.502051972752354e-05,
      "loss": 3.8076,
      "num_input_tokens_seen": 19981760,
      "step": 14750
    },
    {
      "epoch": 0.4722809039114013,
      "grad_norm": 42.5,
      "learning_rate": 3.5011524609904374e-05,
      "loss": 4.6054,
      "num_input_tokens_seen": 19988736,
      "step": 14755
    },
    {
      "epoch": 0.47244094488188976,
      "grad_norm": 36.0,
      "learning_rate": 3.50025279483823e-05,
      "loss": 4.1132,
      "num_input_tokens_seen": 19995728,
      "step": 14760
    },
    {
      "epoch": 0.4726009858523782,
      "grad_norm": 21.75,
      "learning_rate": 3.499352974434472e-05,
      "loss": 3.8504,
      "num_input_tokens_seen": 20002672,
      "step": 14765
    },
    {
      "epoch": 0.47276102682286664,
      "grad_norm": 23.125,
      "learning_rate": 3.498452999917926e-05,
      "loss": 4.0207,
      "num_input_tokens_seen": 20009408,
      "step": 14770
    },
    {
      "epoch": 0.4729210677933551,
      "grad_norm": 41.25,
      "learning_rate": 3.4975528714273795e-05,
      "loss": 4.8948,
      "num_input_tokens_seen": 20016080,
      "step": 14775
    },
    {
      "epoch": 0.4730811087638435,
      "grad_norm": 45.0,
      "learning_rate": 3.4966525891016454e-05,
      "loss": 4.8415,
      "num_input_tokens_seen": 20022816,
      "step": 14780
    },
    {
      "epoch": 0.473241149734332,
      "grad_norm": 28.375,
      "learning_rate": 3.495752153079557e-05,
      "loss": 4.0168,
      "num_input_tokens_seen": 20029136,
      "step": 14785
    },
    {
      "epoch": 0.47340119070482045,
      "grad_norm": 26.0,
      "learning_rate": 3.494851563499974e-05,
      "loss": 3.9468,
      "num_input_tokens_seen": 20035648,
      "step": 14790
    },
    {
      "epoch": 0.4735612316753089,
      "grad_norm": 22.5,
      "learning_rate": 3.493950820501777e-05,
      "loss": 5.4633,
      "num_input_tokens_seen": 20042240,
      "step": 14795
    },
    {
      "epoch": 0.47372127264579733,
      "grad_norm": 27.125,
      "learning_rate": 3.493049924223872e-05,
      "loss": 3.1368,
      "num_input_tokens_seen": 20049328,
      "step": 14800
    },
    {
      "epoch": 0.47372127264579733,
      "eval_loss": 4.4739813804626465,
      "eval_runtime": 490.2404,
      "eval_samples_per_second": 28.325,
      "eval_steps_per_second": 14.162,
      "num_input_tokens_seen": 20049328,
      "step": 14800
    },
    {
      "epoch": 0.47388131361628577,
      "grad_norm": 17.375,
      "learning_rate": 3.49214887480519e-05,
      "loss": 3.3529,
      "num_input_tokens_seen": 20056064,
      "step": 14805
    },
    {
      "epoch": 0.4740413545867742,
      "grad_norm": 37.75,
      "learning_rate": 3.4912476723846834e-05,
      "loss": 3.4789,
      "num_input_tokens_seen": 20063104,
      "step": 14810
    },
    {
      "epoch": 0.47420139555726265,
      "grad_norm": 41.5,
      "learning_rate": 3.490346317101328e-05,
      "loss": 4.4208,
      "num_input_tokens_seen": 20069680,
      "step": 14815
    },
    {
      "epoch": 0.4743614365277511,
      "grad_norm": 22.5,
      "learning_rate": 3.4894448090941266e-05,
      "loss": 3.8545,
      "num_input_tokens_seen": 20076512,
      "step": 14820
    },
    {
      "epoch": 0.47452147749823953,
      "grad_norm": 17.5,
      "learning_rate": 3.488543148502101e-05,
      "loss": 5.0149,
      "num_input_tokens_seen": 20082992,
      "step": 14825
    },
    {
      "epoch": 0.47468151846872797,
      "grad_norm": 29.875,
      "learning_rate": 3.487641335464299e-05,
      "loss": 3.405,
      "num_input_tokens_seen": 20089536,
      "step": 14830
    },
    {
      "epoch": 0.47484155943921647,
      "grad_norm": 16.375,
      "learning_rate": 3.4867393701197914e-05,
      "loss": 4.0971,
      "num_input_tokens_seen": 20096288,
      "step": 14835
    },
    {
      "epoch": 0.4750016004097049,
      "grad_norm": 37.5,
      "learning_rate": 3.485837252607673e-05,
      "loss": 4.8311,
      "num_input_tokens_seen": 20102992,
      "step": 14840
    },
    {
      "epoch": 0.47516164138019334,
      "grad_norm": 22.75,
      "learning_rate": 3.4849349830670615e-05,
      "loss": 3.8818,
      "num_input_tokens_seen": 20109904,
      "step": 14845
    },
    {
      "epoch": 0.4753216823506818,
      "grad_norm": 36.0,
      "learning_rate": 3.4840325616370976e-05,
      "loss": 4.2456,
      "num_input_tokens_seen": 20116608,
      "step": 14850
    },
    {
      "epoch": 0.4754817233211702,
      "grad_norm": 28.375,
      "learning_rate": 3.483129988456947e-05,
      "loss": 4.1511,
      "num_input_tokens_seen": 20123552,
      "step": 14855
    },
    {
      "epoch": 0.47564176429165866,
      "grad_norm": 51.75,
      "learning_rate": 3.482227263665797e-05,
      "loss": 4.4303,
      "num_input_tokens_seen": 20130560,
      "step": 14860
    },
    {
      "epoch": 0.4758018052621471,
      "grad_norm": 23.5,
      "learning_rate": 3.48132438740286e-05,
      "loss": 3.6198,
      "num_input_tokens_seen": 20137248,
      "step": 14865
    },
    {
      "epoch": 0.47596184623263554,
      "grad_norm": 23.375,
      "learning_rate": 3.48042135980737e-05,
      "loss": 3.1948,
      "num_input_tokens_seen": 20143760,
      "step": 14870
    },
    {
      "epoch": 0.476121887203124,
      "grad_norm": 20.25,
      "learning_rate": 3.479518181018586e-05,
      "loss": 4.2625,
      "num_input_tokens_seen": 20150512,
      "step": 14875
    },
    {
      "epoch": 0.4762819281736124,
      "grad_norm": 42.75,
      "learning_rate": 3.4786148511757886e-05,
      "loss": 5.1241,
      "num_input_tokens_seen": 20157248,
      "step": 14880
    },
    {
      "epoch": 0.47644196914410086,
      "grad_norm": 35.75,
      "learning_rate": 3.477711370418284e-05,
      "loss": 4.5746,
      "num_input_tokens_seen": 20164192,
      "step": 14885
    },
    {
      "epoch": 0.47660201011458936,
      "grad_norm": 24.375,
      "learning_rate": 3.476807738885399e-05,
      "loss": 4.5289,
      "num_input_tokens_seen": 20170976,
      "step": 14890
    },
    {
      "epoch": 0.4767620510850778,
      "grad_norm": 18.25,
      "learning_rate": 3.475903956716485e-05,
      "loss": 4.0441,
      "num_input_tokens_seen": 20177488,
      "step": 14895
    },
    {
      "epoch": 0.47692209205556624,
      "grad_norm": 33.25,
      "learning_rate": 3.475000024050917e-05,
      "loss": 4.2818,
      "num_input_tokens_seen": 20183904,
      "step": 14900
    },
    {
      "epoch": 0.4770821330260547,
      "grad_norm": 28.5,
      "learning_rate": 3.4740959410280926e-05,
      "loss": 4.4815,
      "num_input_tokens_seen": 20190528,
      "step": 14905
    },
    {
      "epoch": 0.4772421739965431,
      "grad_norm": 27.875,
      "learning_rate": 3.4731917077874324e-05,
      "loss": 4.3037,
      "num_input_tokens_seen": 20197216,
      "step": 14910
    },
    {
      "epoch": 0.47740221496703156,
      "grad_norm": 32.25,
      "learning_rate": 3.4722873244683816e-05,
      "loss": 3.9474,
      "num_input_tokens_seen": 20203792,
      "step": 14915
    },
    {
      "epoch": 0.47756225593752,
      "grad_norm": 31.75,
      "learning_rate": 3.4713827912104065e-05,
      "loss": 5.0694,
      "num_input_tokens_seen": 20210608,
      "step": 14920
    },
    {
      "epoch": 0.47772229690800844,
      "grad_norm": 34.0,
      "learning_rate": 3.470478108152998e-05,
      "loss": 3.5747,
      "num_input_tokens_seen": 20217456,
      "step": 14925
    },
    {
      "epoch": 0.4778823378784969,
      "grad_norm": 27.375,
      "learning_rate": 3.4695732754356695e-05,
      "loss": 4.8969,
      "num_input_tokens_seen": 20224384,
      "step": 14930
    },
    {
      "epoch": 0.4780423788489853,
      "grad_norm": 41.0,
      "learning_rate": 3.4686682931979576e-05,
      "loss": 3.9549,
      "num_input_tokens_seen": 20230848,
      "step": 14935
    },
    {
      "epoch": 0.4782024198194738,
      "grad_norm": 31.5,
      "learning_rate": 3.467763161579422e-05,
      "loss": 4.821,
      "num_input_tokens_seen": 20237856,
      "step": 14940
    },
    {
      "epoch": 0.47836246078996225,
      "grad_norm": 25.75,
      "learning_rate": 3.466857880719645e-05,
      "loss": 3.2874,
      "num_input_tokens_seen": 20245040,
      "step": 14945
    },
    {
      "epoch": 0.4785225017604507,
      "grad_norm": 32.75,
      "learning_rate": 3.465952450758233e-05,
      "loss": 3.9554,
      "num_input_tokens_seen": 20251504,
      "step": 14950
    },
    {
      "epoch": 0.47868254273093913,
      "grad_norm": 19.75,
      "learning_rate": 3.4650468718348126e-05,
      "loss": 5.2293,
      "num_input_tokens_seen": 20258528,
      "step": 14955
    },
    {
      "epoch": 0.47884258370142757,
      "grad_norm": 45.75,
      "learning_rate": 3.464141144089038e-05,
      "loss": 5.2078,
      "num_input_tokens_seen": 20265200,
      "step": 14960
    },
    {
      "epoch": 0.479002624671916,
      "grad_norm": 33.75,
      "learning_rate": 3.463235267660583e-05,
      "loss": 3.6405,
      "num_input_tokens_seen": 20271840,
      "step": 14965
    },
    {
      "epoch": 0.47916266564240445,
      "grad_norm": 37.0,
      "learning_rate": 3.462329242689145e-05,
      "loss": 4.4337,
      "num_input_tokens_seen": 20278592,
      "step": 14970
    },
    {
      "epoch": 0.4793227066128929,
      "grad_norm": 31.625,
      "learning_rate": 3.461423069314444e-05,
      "loss": 4.3477,
      "num_input_tokens_seen": 20285280,
      "step": 14975
    },
    {
      "epoch": 0.47948274758338133,
      "grad_norm": 25.0,
      "learning_rate": 3.460516747676224e-05,
      "loss": 4.2874,
      "num_input_tokens_seen": 20292032,
      "step": 14980
    },
    {
      "epoch": 0.47964278855386977,
      "grad_norm": 35.75,
      "learning_rate": 3.459610277914251e-05,
      "loss": 4.611,
      "num_input_tokens_seen": 20298768,
      "step": 14985
    },
    {
      "epoch": 0.4798028295243582,
      "grad_norm": 30.5,
      "learning_rate": 3.458703660168314e-05,
      "loss": 4.6302,
      "num_input_tokens_seen": 20305728,
      "step": 14990
    },
    {
      "epoch": 0.4799628704948467,
      "grad_norm": 50.75,
      "learning_rate": 3.457796894578224e-05,
      "loss": 3.6424,
      "num_input_tokens_seen": 20312432,
      "step": 14995
    },
    {
      "epoch": 0.48012291146533514,
      "grad_norm": 31.0,
      "learning_rate": 3.456889981283817e-05,
      "loss": 4.6479,
      "num_input_tokens_seen": 20319488,
      "step": 15000
    },
    {
      "epoch": 0.48012291146533514,
      "eval_loss": 4.496302127838135,
      "eval_runtime": 489.7827,
      "eval_samples_per_second": 28.351,
      "eval_steps_per_second": 14.176,
      "num_input_tokens_seen": 20319488,
      "step": 15000
    },
    {
      "epoch": 0.4802829524358236,
      "grad_norm": 19.875,
      "learning_rate": 3.45598292042495e-05,
      "loss": 4.3164,
      "num_input_tokens_seen": 20326416,
      "step": 15005
    },
    {
      "epoch": 0.480442993406312,
      "grad_norm": 35.75,
      "learning_rate": 3.4550757121415035e-05,
      "loss": 4.7858,
      "num_input_tokens_seen": 20332816,
      "step": 15010
    },
    {
      "epoch": 0.48060303437680046,
      "grad_norm": 34.0,
      "learning_rate": 3.454168356573378e-05,
      "loss": 4.7943,
      "num_input_tokens_seen": 20339072,
      "step": 15015
    },
    {
      "epoch": 0.4807630753472889,
      "grad_norm": 32.5,
      "learning_rate": 3.453260853860503e-05,
      "loss": 4.0636,
      "num_input_tokens_seen": 20346176,
      "step": 15020
    },
    {
      "epoch": 0.48092311631777734,
      "grad_norm": 52.75,
      "learning_rate": 3.452353204142824e-05,
      "loss": 5.1709,
      "num_input_tokens_seen": 20352800,
      "step": 15025
    },
    {
      "epoch": 0.4810831572882658,
      "grad_norm": 38.0,
      "learning_rate": 3.4514454075603136e-05,
      "loss": 4.3051,
      "num_input_tokens_seen": 20359520,
      "step": 15030
    },
    {
      "epoch": 0.4812431982587542,
      "grad_norm": 31.625,
      "learning_rate": 3.450537464252964e-05,
      "loss": 4.6424,
      "num_input_tokens_seen": 20366720,
      "step": 15035
    },
    {
      "epoch": 0.48140323922924266,
      "grad_norm": 26.625,
      "learning_rate": 3.4496293743607925e-05,
      "loss": 4.9163,
      "num_input_tokens_seen": 20373216,
      "step": 15040
    },
    {
      "epoch": 0.48156328019973116,
      "grad_norm": 37.25,
      "learning_rate": 3.448721138023838e-05,
      "loss": 3.8504,
      "num_input_tokens_seen": 20380016,
      "step": 15045
    },
    {
      "epoch": 0.4817233211702196,
      "grad_norm": 46.25,
      "learning_rate": 3.447812755382162e-05,
      "loss": 5.049,
      "num_input_tokens_seen": 20386624,
      "step": 15050
    },
    {
      "epoch": 0.48188336214070804,
      "grad_norm": 49.5,
      "learning_rate": 3.446904226575847e-05,
      "loss": 4.7674,
      "num_input_tokens_seen": 20392960,
      "step": 15055
    },
    {
      "epoch": 0.4820434031111965,
      "grad_norm": 27.25,
      "learning_rate": 3.445995551745002e-05,
      "loss": 5.4063,
      "num_input_tokens_seen": 20400032,
      "step": 15060
    },
    {
      "epoch": 0.4822034440816849,
      "grad_norm": 19.625,
      "learning_rate": 3.445086731029753e-05,
      "loss": 4.1496,
      "num_input_tokens_seen": 20406800,
      "step": 15065
    },
    {
      "epoch": 0.48236348505217336,
      "grad_norm": 25.125,
      "learning_rate": 3.444177764570255e-05,
      "loss": 4.4104,
      "num_input_tokens_seen": 20413536,
      "step": 15070
    },
    {
      "epoch": 0.4825235260226618,
      "grad_norm": 26.375,
      "learning_rate": 3.44326865250668e-05,
      "loss": 5.194,
      "num_input_tokens_seen": 20420144,
      "step": 15075
    },
    {
      "epoch": 0.48268356699315024,
      "grad_norm": 33.75,
      "learning_rate": 3.442359394979225e-05,
      "loss": 4.7306,
      "num_input_tokens_seen": 20426784,
      "step": 15080
    },
    {
      "epoch": 0.4828436079636387,
      "grad_norm": 25.125,
      "learning_rate": 3.441449992128108e-05,
      "loss": 5.004,
      "num_input_tokens_seen": 20433824,
      "step": 15085
    },
    {
      "epoch": 0.4830036489341271,
      "grad_norm": 28.375,
      "learning_rate": 3.440540444093573e-05,
      "loss": 4.4823,
      "num_input_tokens_seen": 20440464,
      "step": 15090
    },
    {
      "epoch": 0.48316368990461556,
      "grad_norm": 91.5,
      "learning_rate": 3.43963075101588e-05,
      "loss": 4.1053,
      "num_input_tokens_seen": 20447376,
      "step": 15095
    },
    {
      "epoch": 0.48332373087510405,
      "grad_norm": 28.875,
      "learning_rate": 3.438720913035318e-05,
      "loss": 3.6255,
      "num_input_tokens_seen": 20454096,
      "step": 15100
    },
    {
      "epoch": 0.4834837718455925,
      "grad_norm": 41.75,
      "learning_rate": 3.437810930292195e-05,
      "loss": 3.8799,
      "num_input_tokens_seen": 20461104,
      "step": 15105
    },
    {
      "epoch": 0.48364381281608093,
      "grad_norm": 26.125,
      "learning_rate": 3.43690080292684e-05,
      "loss": 4.0941,
      "num_input_tokens_seen": 20467728,
      "step": 15110
    },
    {
      "epoch": 0.48380385378656937,
      "grad_norm": 47.25,
      "learning_rate": 3.435990531079608e-05,
      "loss": 4.2563,
      "num_input_tokens_seen": 20474160,
      "step": 15115
    },
    {
      "epoch": 0.4839638947570578,
      "grad_norm": 30.5,
      "learning_rate": 3.435080114890874e-05,
      "loss": 3.9961,
      "num_input_tokens_seen": 20480896,
      "step": 15120
    },
    {
      "epoch": 0.48412393572754625,
      "grad_norm": 22.0,
      "learning_rate": 3.434169554501035e-05,
      "loss": 3.6518,
      "num_input_tokens_seen": 20487664,
      "step": 15125
    },
    {
      "epoch": 0.4842839766980347,
      "grad_norm": 40.75,
      "learning_rate": 3.433258850050511e-05,
      "loss": 4.3056,
      "num_input_tokens_seen": 20494624,
      "step": 15130
    },
    {
      "epoch": 0.48444401766852313,
      "grad_norm": 24.75,
      "learning_rate": 3.4323480016797446e-05,
      "loss": 3.8412,
      "num_input_tokens_seen": 20501520,
      "step": 15135
    },
    {
      "epoch": 0.48460405863901157,
      "grad_norm": 54.25,
      "learning_rate": 3.4314370095291995e-05,
      "loss": 4.6066,
      "num_input_tokens_seen": 20508560,
      "step": 15140
    },
    {
      "epoch": 0.4847640996095,
      "grad_norm": 15.625,
      "learning_rate": 3.430525873739363e-05,
      "loss": 5.0863,
      "num_input_tokens_seen": 20515504,
      "step": 15145
    },
    {
      "epoch": 0.48492414057998845,
      "grad_norm": 34.25,
      "learning_rate": 3.429614594450743e-05,
      "loss": 3.7662,
      "num_input_tokens_seen": 20522640,
      "step": 15150
    },
    {
      "epoch": 0.48508418155047694,
      "grad_norm": 35.0,
      "learning_rate": 3.428703171803869e-05,
      "loss": 4.8879,
      "num_input_tokens_seen": 20529648,
      "step": 15155
    },
    {
      "epoch": 0.4852442225209654,
      "grad_norm": 28.75,
      "learning_rate": 3.4277916059392964e-05,
      "loss": 3.9943,
      "num_input_tokens_seen": 20536496,
      "step": 15160
    },
    {
      "epoch": 0.4854042634914538,
      "grad_norm": 37.5,
      "learning_rate": 3.426879896997598e-05,
      "loss": 3.896,
      "num_input_tokens_seen": 20543552,
      "step": 15165
    },
    {
      "epoch": 0.48556430446194226,
      "grad_norm": 24.0,
      "learning_rate": 3.425968045119372e-05,
      "loss": 4.7714,
      "num_input_tokens_seen": 20550176,
      "step": 15170
    },
    {
      "epoch": 0.4857243454324307,
      "grad_norm": 22.0,
      "learning_rate": 3.425056050445237e-05,
      "loss": 3.4859,
      "num_input_tokens_seen": 20556960,
      "step": 15175
    },
    {
      "epoch": 0.48588438640291914,
      "grad_norm": 27.75,
      "learning_rate": 3.4241439131158336e-05,
      "loss": 3.6976,
      "num_input_tokens_seen": 20563808,
      "step": 15180
    },
    {
      "epoch": 0.4860444273734076,
      "grad_norm": 22.125,
      "learning_rate": 3.423231633271825e-05,
      "loss": 3.5774,
      "num_input_tokens_seen": 20570304,
      "step": 15185
    },
    {
      "epoch": 0.486204468343896,
      "grad_norm": 21.375,
      "learning_rate": 3.4223192110538985e-05,
      "loss": 4.0536,
      "num_input_tokens_seen": 20576656,
      "step": 15190
    },
    {
      "epoch": 0.48636450931438446,
      "grad_norm": 29.25,
      "learning_rate": 3.4214066466027575e-05,
      "loss": 4.1977,
      "num_input_tokens_seen": 20583184,
      "step": 15195
    },
    {
      "epoch": 0.4865245502848729,
      "grad_norm": 28.125,
      "learning_rate": 3.4204939400591325e-05,
      "loss": 4.9356,
      "num_input_tokens_seen": 20589760,
      "step": 15200
    },
    {
      "epoch": 0.4865245502848729,
      "eval_loss": 4.482365131378174,
      "eval_runtime": 489.8489,
      "eval_samples_per_second": 28.348,
      "eval_steps_per_second": 14.174,
      "num_input_tokens_seen": 20589760,
      "step": 15200
    },
    {
      "epoch": 0.4866845912553614,
      "grad_norm": 25.5,
      "learning_rate": 3.419581091563775e-05,
      "loss": 4.4281,
      "num_input_tokens_seen": 20596368,
      "step": 15205
    },
    {
      "epoch": 0.48684463222584984,
      "grad_norm": 50.0,
      "learning_rate": 3.418668101257456e-05,
      "loss": 3.9968,
      "num_input_tokens_seen": 20603328,
      "step": 15210
    },
    {
      "epoch": 0.4870046731963383,
      "grad_norm": 25.75,
      "learning_rate": 3.417754969280971e-05,
      "loss": 4.1925,
      "num_input_tokens_seen": 20610368,
      "step": 15215
    },
    {
      "epoch": 0.4871647141668267,
      "grad_norm": 31.25,
      "learning_rate": 3.416841695775137e-05,
      "loss": 4.7933,
      "num_input_tokens_seen": 20616992,
      "step": 15220
    },
    {
      "epoch": 0.48732475513731516,
      "grad_norm": 44.0,
      "learning_rate": 3.415928280880792e-05,
      "loss": 4.6825,
      "num_input_tokens_seen": 20623696,
      "step": 15225
    },
    {
      "epoch": 0.4874847961078036,
      "grad_norm": 30.625,
      "learning_rate": 3.4150147247387965e-05,
      "loss": 4.8716,
      "num_input_tokens_seen": 20630224,
      "step": 15230
    },
    {
      "epoch": 0.48764483707829204,
      "grad_norm": 23.0,
      "learning_rate": 3.4141010274900306e-05,
      "loss": 4.0482,
      "num_input_tokens_seen": 20636944,
      "step": 15235
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 23.0,
      "learning_rate": 3.413187189275399e-05,
      "loss": 3.9967,
      "num_input_tokens_seen": 20643632,
      "step": 15240
    },
    {
      "epoch": 0.4879649190192689,
      "grad_norm": 48.5,
      "learning_rate": 3.4122732102358265e-05,
      "loss": 4.0123,
      "num_input_tokens_seen": 20650352,
      "step": 15245
    },
    {
      "epoch": 0.48812495998975736,
      "grad_norm": 31.375,
      "learning_rate": 3.411359090512261e-05,
      "loss": 4.6729,
      "num_input_tokens_seen": 20656864,
      "step": 15250
    },
    {
      "epoch": 0.4882850009602458,
      "grad_norm": 24.875,
      "learning_rate": 3.410444830245672e-05,
      "loss": 3.9172,
      "num_input_tokens_seen": 20663584,
      "step": 15255
    },
    {
      "epoch": 0.4884450419307343,
      "grad_norm": 35.0,
      "learning_rate": 3.409530429577048e-05,
      "loss": 4.4877,
      "num_input_tokens_seen": 20670256,
      "step": 15260
    },
    {
      "epoch": 0.48860508290122273,
      "grad_norm": 14.25,
      "learning_rate": 3.408615888647402e-05,
      "loss": 4.4081,
      "num_input_tokens_seen": 20676800,
      "step": 15265
    },
    {
      "epoch": 0.48876512387171117,
      "grad_norm": 25.625,
      "learning_rate": 3.4077012075977675e-05,
      "loss": 4.0676,
      "num_input_tokens_seen": 20683632,
      "step": 15270
    },
    {
      "epoch": 0.4889251648421996,
      "grad_norm": 40.25,
      "learning_rate": 3.4067863865692e-05,
      "loss": 5.2325,
      "num_input_tokens_seen": 20690448,
      "step": 15275
    },
    {
      "epoch": 0.48908520581268805,
      "grad_norm": 35.5,
      "learning_rate": 3.4058714257027755e-05,
      "loss": 5.0518,
      "num_input_tokens_seen": 20697456,
      "step": 15280
    },
    {
      "epoch": 0.4892452467831765,
      "grad_norm": 17.5,
      "learning_rate": 3.404956325139594e-05,
      "loss": 4.2506,
      "num_input_tokens_seen": 20704112,
      "step": 15285
    },
    {
      "epoch": 0.48940528775366493,
      "grad_norm": 43.75,
      "learning_rate": 3.404041085020775e-05,
      "loss": 4.8942,
      "num_input_tokens_seen": 20710928,
      "step": 15290
    },
    {
      "epoch": 0.48956532872415337,
      "grad_norm": 24.625,
      "learning_rate": 3.403125705487459e-05,
      "loss": 4.9057,
      "num_input_tokens_seen": 20717504,
      "step": 15295
    },
    {
      "epoch": 0.4897253696946418,
      "grad_norm": 19.625,
      "learning_rate": 3.402210186680811e-05,
      "loss": 3.3983,
      "num_input_tokens_seen": 20724640,
      "step": 15300
    },
    {
      "epoch": 0.48988541066513025,
      "grad_norm": 20.125,
      "learning_rate": 3.4012945287420137e-05,
      "loss": 4.862,
      "num_input_tokens_seen": 20731472,
      "step": 15305
    },
    {
      "epoch": 0.49004545163561875,
      "grad_norm": 42.5,
      "learning_rate": 3.400378731812274e-05,
      "loss": 4.571,
      "num_input_tokens_seen": 20739040,
      "step": 15310
    },
    {
      "epoch": 0.4902054926061072,
      "grad_norm": 22.0,
      "learning_rate": 3.399462796032817e-05,
      "loss": 4.3381,
      "num_input_tokens_seen": 20745344,
      "step": 15315
    },
    {
      "epoch": 0.4903655335765956,
      "grad_norm": 33.25,
      "learning_rate": 3.3985467215448954e-05,
      "loss": 4.4104,
      "num_input_tokens_seen": 20751904,
      "step": 15320
    },
    {
      "epoch": 0.49052557454708406,
      "grad_norm": 25.875,
      "learning_rate": 3.3976305084897776e-05,
      "loss": 3.9376,
      "num_input_tokens_seen": 20758256,
      "step": 15325
    },
    {
      "epoch": 0.4906856155175725,
      "grad_norm": 19.0,
      "learning_rate": 3.3967141570087544e-05,
      "loss": 3.1591,
      "num_input_tokens_seen": 20765040,
      "step": 15330
    },
    {
      "epoch": 0.49084565648806094,
      "grad_norm": 21.125,
      "learning_rate": 3.39579766724314e-05,
      "loss": 4.1707,
      "num_input_tokens_seen": 20771600,
      "step": 15335
    },
    {
      "epoch": 0.4910056974585494,
      "grad_norm": 21.5,
      "learning_rate": 3.3948810393342677e-05,
      "loss": 4.1782,
      "num_input_tokens_seen": 20778128,
      "step": 15340
    },
    {
      "epoch": 0.4911657384290378,
      "grad_norm": 46.25,
      "learning_rate": 3.3939642734234936e-05,
      "loss": 4.9812,
      "num_input_tokens_seen": 20785376,
      "step": 15345
    },
    {
      "epoch": 0.49132577939952626,
      "grad_norm": 21.25,
      "learning_rate": 3.393047369652194e-05,
      "loss": 3.3467,
      "num_input_tokens_seen": 20792496,
      "step": 15350
    },
    {
      "epoch": 0.4914858203700147,
      "grad_norm": 24.25,
      "learning_rate": 3.3921303281617664e-05,
      "loss": 3.9671,
      "num_input_tokens_seen": 20799248,
      "step": 15355
    },
    {
      "epoch": 0.49164586134050314,
      "grad_norm": 19.5,
      "learning_rate": 3.391213149093632e-05,
      "loss": 4.5953,
      "num_input_tokens_seen": 20806304,
      "step": 15360
    },
    {
      "epoch": 0.49180590231099164,
      "grad_norm": 42.5,
      "learning_rate": 3.3902958325892303e-05,
      "loss": 4.5078,
      "num_input_tokens_seen": 20813104,
      "step": 15365
    },
    {
      "epoch": 0.4919659432814801,
      "grad_norm": 25.0,
      "learning_rate": 3.389378378790023e-05,
      "loss": 4.9647,
      "num_input_tokens_seen": 20820368,
      "step": 15370
    },
    {
      "epoch": 0.4921259842519685,
      "grad_norm": 27.375,
      "learning_rate": 3.388460787837493e-05,
      "loss": 5.2711,
      "num_input_tokens_seen": 20826944,
      "step": 15375
    },
    {
      "epoch": 0.49228602522245696,
      "grad_norm": 28.375,
      "learning_rate": 3.387543059873145e-05,
      "loss": 3.8003,
      "num_input_tokens_seen": 20834000,
      "step": 15380
    },
    {
      "epoch": 0.4924460661929454,
      "grad_norm": 30.5,
      "learning_rate": 3.386625195038503e-05,
      "loss": 3.9814,
      "num_input_tokens_seen": 20841248,
      "step": 15385
    },
    {
      "epoch": 0.49260610716343384,
      "grad_norm": 38.5,
      "learning_rate": 3.3857071934751136e-05,
      "loss": 5.1291,
      "num_input_tokens_seen": 20847744,
      "step": 15390
    },
    {
      "epoch": 0.4927661481339223,
      "grad_norm": 18.75,
      "learning_rate": 3.384789055324544e-05,
      "loss": 4.6433,
      "num_input_tokens_seen": 20854192,
      "step": 15395
    },
    {
      "epoch": 0.4929261891044107,
      "grad_norm": 26.375,
      "learning_rate": 3.3838707807283843e-05,
      "loss": 4.012,
      "num_input_tokens_seen": 20860624,
      "step": 15400
    },
    {
      "epoch": 0.4929261891044107,
      "eval_loss": 4.48311185836792,
      "eval_runtime": 489.7287,
      "eval_samples_per_second": 28.354,
      "eval_steps_per_second": 14.177,
      "num_input_tokens_seen": 20860624,
      "step": 15400
    },
    {
      "epoch": 0.49308623007489916,
      "grad_norm": 42.0,
      "learning_rate": 3.382952369828243e-05,
      "loss": 4.4695,
      "num_input_tokens_seen": 20867536,
      "step": 15405
    },
    {
      "epoch": 0.4932462710453876,
      "grad_norm": 24.75,
      "learning_rate": 3.38203382276575e-05,
      "loss": 4.3077,
      "num_input_tokens_seen": 20874000,
      "step": 15410
    },
    {
      "epoch": 0.4934063120158761,
      "grad_norm": 21.875,
      "learning_rate": 3.381115139682557e-05,
      "loss": 3.8214,
      "num_input_tokens_seen": 20881280,
      "step": 15415
    },
    {
      "epoch": 0.49356635298636453,
      "grad_norm": 23.5,
      "learning_rate": 3.3801963207203366e-05,
      "loss": 5.2584,
      "num_input_tokens_seen": 20887936,
      "step": 15420
    },
    {
      "epoch": 0.49372639395685297,
      "grad_norm": 16.0,
      "learning_rate": 3.379277366020782e-05,
      "loss": 4.4507,
      "num_input_tokens_seen": 20895312,
      "step": 15425
    },
    {
      "epoch": 0.4938864349273414,
      "grad_norm": 32.25,
      "learning_rate": 3.3783582757256085e-05,
      "loss": 5.2354,
      "num_input_tokens_seen": 20902048,
      "step": 15430
    },
    {
      "epoch": 0.49404647589782985,
      "grad_norm": 24.0,
      "learning_rate": 3.3774390499765504e-05,
      "loss": 4.3554,
      "num_input_tokens_seen": 20909072,
      "step": 15435
    },
    {
      "epoch": 0.4942065168683183,
      "grad_norm": 44.75,
      "learning_rate": 3.376519688915364e-05,
      "loss": 4.4093,
      "num_input_tokens_seen": 20915520,
      "step": 15440
    },
    {
      "epoch": 0.49436655783880673,
      "grad_norm": 34.25,
      "learning_rate": 3.3756001926838273e-05,
      "loss": 5.0422,
      "num_input_tokens_seen": 20922208,
      "step": 15445
    },
    {
      "epoch": 0.49452659880929517,
      "grad_norm": 36.25,
      "learning_rate": 3.374680561423737e-05,
      "loss": 4.0871,
      "num_input_tokens_seen": 20928752,
      "step": 15450
    },
    {
      "epoch": 0.4946866397797836,
      "grad_norm": 19.25,
      "learning_rate": 3.373760795276912e-05,
      "loss": 4.2002,
      "num_input_tokens_seen": 20935312,
      "step": 15455
    },
    {
      "epoch": 0.49484668075027205,
      "grad_norm": 21.125,
      "learning_rate": 3.372840894385192e-05,
      "loss": 3.7467,
      "num_input_tokens_seen": 20942208,
      "step": 15460
    },
    {
      "epoch": 0.4950067217207605,
      "grad_norm": 32.75,
      "learning_rate": 3.3719208588904375e-05,
      "loss": 5.0267,
      "num_input_tokens_seen": 20948592,
      "step": 15465
    },
    {
      "epoch": 0.495166762691249,
      "grad_norm": 15.625,
      "learning_rate": 3.371000688934529e-05,
      "loss": 3.6187,
      "num_input_tokens_seen": 20955440,
      "step": 15470
    },
    {
      "epoch": 0.4953268036617374,
      "grad_norm": 37.0,
      "learning_rate": 3.370080384659369e-05,
      "loss": 4.3276,
      "num_input_tokens_seen": 20962352,
      "step": 15475
    },
    {
      "epoch": 0.49548684463222586,
      "grad_norm": 40.0,
      "learning_rate": 3.36915994620688e-05,
      "loss": 4.5747,
      "num_input_tokens_seen": 20969312,
      "step": 15480
    },
    {
      "epoch": 0.4956468856027143,
      "grad_norm": 27.25,
      "learning_rate": 3.3682393737190035e-05,
      "loss": 4.9813,
      "num_input_tokens_seen": 20976128,
      "step": 15485
    },
    {
      "epoch": 0.49580692657320274,
      "grad_norm": 34.0,
      "learning_rate": 3.3673186673377054e-05,
      "loss": 4.5547,
      "num_input_tokens_seen": 20982752,
      "step": 15490
    },
    {
      "epoch": 0.4959669675436912,
      "grad_norm": 24.0,
      "learning_rate": 3.366397827204969e-05,
      "loss": 4.389,
      "num_input_tokens_seen": 20989872,
      "step": 15495
    },
    {
      "epoch": 0.4961270085141796,
      "grad_norm": 25.875,
      "learning_rate": 3.3654768534628e-05,
      "loss": 4.2937,
      "num_input_tokens_seen": 20996256,
      "step": 15500
    },
    {
      "epoch": 0.49628704948466806,
      "grad_norm": 27.375,
      "learning_rate": 3.3645557462532245e-05,
      "loss": 4.814,
      "num_input_tokens_seen": 21002560,
      "step": 15505
    },
    {
      "epoch": 0.4964470904551565,
      "grad_norm": 41.25,
      "learning_rate": 3.363634505718288e-05,
      "loss": 4.6516,
      "num_input_tokens_seen": 21009248,
      "step": 15510
    },
    {
      "epoch": 0.49660713142564494,
      "grad_norm": 33.25,
      "learning_rate": 3.362713132000057e-05,
      "loss": 4.0934,
      "num_input_tokens_seen": 21016048,
      "step": 15515
    },
    {
      "epoch": 0.4967671723961334,
      "grad_norm": 15.0,
      "learning_rate": 3.36179162524062e-05,
      "loss": 4.31,
      "num_input_tokens_seen": 21023136,
      "step": 15520
    },
    {
      "epoch": 0.4969272133666219,
      "grad_norm": 33.75,
      "learning_rate": 3.3608699855820846e-05,
      "loss": 3.8571,
      "num_input_tokens_seen": 21030192,
      "step": 15525
    },
    {
      "epoch": 0.4970872543371103,
      "grad_norm": 19.5,
      "learning_rate": 3.359948213166578e-05,
      "loss": 4.3738,
      "num_input_tokens_seen": 21037456,
      "step": 15530
    },
    {
      "epoch": 0.49724729530759876,
      "grad_norm": 20.75,
      "learning_rate": 3.359026308136252e-05,
      "loss": 4.4786,
      "num_input_tokens_seen": 21044304,
      "step": 15535
    },
    {
      "epoch": 0.4974073362780872,
      "grad_norm": 34.5,
      "learning_rate": 3.358104270633272e-05,
      "loss": 4.0106,
      "num_input_tokens_seen": 21051232,
      "step": 15540
    },
    {
      "epoch": 0.49756737724857564,
      "grad_norm": 18.375,
      "learning_rate": 3.357182100799831e-05,
      "loss": 3.5214,
      "num_input_tokens_seen": 21057968,
      "step": 15545
    },
    {
      "epoch": 0.4977274182190641,
      "grad_norm": 20.0,
      "learning_rate": 3.3562597987781384e-05,
      "loss": 3.8682,
      "num_input_tokens_seen": 21064624,
      "step": 15550
    },
    {
      "epoch": 0.4978874591895525,
      "grad_norm": 15.9375,
      "learning_rate": 3.355337364710424e-05,
      "loss": 4.318,
      "num_input_tokens_seen": 21071680,
      "step": 15555
    },
    {
      "epoch": 0.49804750016004096,
      "grad_norm": 20.375,
      "learning_rate": 3.354414798738939e-05,
      "loss": 3.9067,
      "num_input_tokens_seen": 21078672,
      "step": 15560
    },
    {
      "epoch": 0.4982075411305294,
      "grad_norm": 18.625,
      "learning_rate": 3.353492101005955e-05,
      "loss": 4.1223,
      "num_input_tokens_seen": 21085328,
      "step": 15565
    },
    {
      "epoch": 0.49836758210101784,
      "grad_norm": 40.75,
      "learning_rate": 3.352569271653763e-05,
      "loss": 4.183,
      "num_input_tokens_seen": 21092272,
      "step": 15570
    },
    {
      "epoch": 0.49852762307150633,
      "grad_norm": 21.375,
      "learning_rate": 3.351646310824675e-05,
      "loss": 4.444,
      "num_input_tokens_seen": 21099008,
      "step": 15575
    },
    {
      "epoch": 0.49868766404199477,
      "grad_norm": 44.25,
      "learning_rate": 3.350723218661023e-05,
      "loss": 3.8416,
      "num_input_tokens_seen": 21106080,
      "step": 15580
    },
    {
      "epoch": 0.4988477050124832,
      "grad_norm": 29.5,
      "learning_rate": 3.349799995305162e-05,
      "loss": 4.9059,
      "num_input_tokens_seen": 21112928,
      "step": 15585
    },
    {
      "epoch": 0.49900774598297165,
      "grad_norm": 32.75,
      "learning_rate": 3.348876640899461e-05,
      "loss": 4.4599,
      "num_input_tokens_seen": 21119664,
      "step": 15590
    },
    {
      "epoch": 0.4991677869534601,
      "grad_norm": 15.8125,
      "learning_rate": 3.3479531555863144e-05,
      "loss": 3.8963,
      "num_input_tokens_seen": 21126576,
      "step": 15595
    },
    {
      "epoch": 0.49932782792394853,
      "grad_norm": 73.0,
      "learning_rate": 3.3470295395081344e-05,
      "loss": 4.1237,
      "num_input_tokens_seen": 21133104,
      "step": 15600
    },
    {
      "epoch": 0.49932782792394853,
      "eval_loss": 4.482914924621582,
      "eval_runtime": 489.8511,
      "eval_samples_per_second": 28.347,
      "eval_steps_per_second": 14.174,
      "num_input_tokens_seen": 21133104,
      "step": 15600
    },
    {
      "epoch": 0.49948786889443697,
      "grad_norm": 46.0,
      "learning_rate": 3.3461057928073556e-05,
      "loss": 4.8807,
      "num_input_tokens_seen": 21140128,
      "step": 15605
    },
    {
      "epoch": 0.4996479098649254,
      "grad_norm": 27.0,
      "learning_rate": 3.345181915626431e-05,
      "loss": 4.4365,
      "num_input_tokens_seen": 21147040,
      "step": 15610
    },
    {
      "epoch": 0.49980795083541385,
      "grad_norm": 35.25,
      "learning_rate": 3.344257908107834e-05,
      "loss": 4.4764,
      "num_input_tokens_seen": 21153696,
      "step": 15615
    },
    {
      "epoch": 0.4999679918059023,
      "grad_norm": 43.0,
      "learning_rate": 3.343333770394058e-05,
      "loss": 4.2115,
      "num_input_tokens_seen": 21160176,
      "step": 15620
    },
    {
      "epoch": 0.5001280327763907,
      "grad_norm": 41.75,
      "learning_rate": 3.342409502627616e-05,
      "loss": 3.9658,
      "num_input_tokens_seen": 21166672,
      "step": 15625
    },
    {
      "epoch": 0.5002880737468792,
      "grad_norm": 29.5,
      "learning_rate": 3.341485104951043e-05,
      "loss": 4.1648,
      "num_input_tokens_seen": 21173712,
      "step": 15630
    },
    {
      "epoch": 0.5004481147173676,
      "grad_norm": 42.25,
      "learning_rate": 3.340560577506892e-05,
      "loss": 4.5801,
      "num_input_tokens_seen": 21179984,
      "step": 15635
    },
    {
      "epoch": 0.5006081556878561,
      "grad_norm": 33.75,
      "learning_rate": 3.339635920437735e-05,
      "loss": 4.8386,
      "num_input_tokens_seen": 21186528,
      "step": 15640
    },
    {
      "epoch": 0.5007681966583445,
      "grad_norm": 28.0,
      "learning_rate": 3.338711133886169e-05,
      "loss": 4.2957,
      "num_input_tokens_seen": 21193248,
      "step": 15645
    },
    {
      "epoch": 0.500928237628833,
      "grad_norm": 27.875,
      "learning_rate": 3.3377862179948064e-05,
      "loss": 5.2322,
      "num_input_tokens_seen": 21199904,
      "step": 15650
    },
    {
      "epoch": 0.5010882785993215,
      "grad_norm": 32.75,
      "learning_rate": 3.336861172906281e-05,
      "loss": 4.941,
      "num_input_tokens_seen": 21206528,
      "step": 15655
    },
    {
      "epoch": 0.5012483195698099,
      "grad_norm": 19.625,
      "learning_rate": 3.335935998763245e-05,
      "loss": 4.9174,
      "num_input_tokens_seen": 21213520,
      "step": 15660
    },
    {
      "epoch": 0.5014083605402984,
      "grad_norm": 29.875,
      "learning_rate": 3.3350106957083744e-05,
      "loss": 3.9965,
      "num_input_tokens_seen": 21220384,
      "step": 15665
    },
    {
      "epoch": 0.5015684015107867,
      "grad_norm": 38.75,
      "learning_rate": 3.33408526388436e-05,
      "loss": 4.7896,
      "num_input_tokens_seen": 21227136,
      "step": 15670
    },
    {
      "epoch": 0.5017284424812752,
      "grad_norm": 30.25,
      "learning_rate": 3.3331597034339166e-05,
      "loss": 4.6826,
      "num_input_tokens_seen": 21233968,
      "step": 15675
    },
    {
      "epoch": 0.5018884834517636,
      "grad_norm": 37.25,
      "learning_rate": 3.3322340144997764e-05,
      "loss": 4.4761,
      "num_input_tokens_seen": 21240560,
      "step": 15680
    },
    {
      "epoch": 0.5020485244222521,
      "grad_norm": 16.125,
      "learning_rate": 3.331308197224693e-05,
      "loss": 4.4938,
      "num_input_tokens_seen": 21247888,
      "step": 15685
    },
    {
      "epoch": 0.5022085653927405,
      "grad_norm": 41.5,
      "learning_rate": 3.330382251751438e-05,
      "loss": 4.7966,
      "num_input_tokens_seen": 21254624,
      "step": 15690
    },
    {
      "epoch": 0.502368606363229,
      "grad_norm": 26.125,
      "learning_rate": 3.3294561782228054e-05,
      "loss": 3.9778,
      "num_input_tokens_seen": 21261344,
      "step": 15695
    },
    {
      "epoch": 0.5025286473337174,
      "grad_norm": 27.75,
      "learning_rate": 3.328529976781607e-05,
      "loss": 3.5222,
      "num_input_tokens_seen": 21268560,
      "step": 15700
    },
    {
      "epoch": 0.5026886883042059,
      "grad_norm": 22.25,
      "learning_rate": 3.327603647570673e-05,
      "loss": 3.9547,
      "num_input_tokens_seen": 21275408,
      "step": 15705
    },
    {
      "epoch": 0.5028487292746944,
      "grad_norm": 35.0,
      "learning_rate": 3.326677190732857e-05,
      "loss": 4.6481,
      "num_input_tokens_seen": 21282192,
      "step": 15710
    },
    {
      "epoch": 0.5030087702451828,
      "grad_norm": 20.25,
      "learning_rate": 3.325750606411029e-05,
      "loss": 3.8369,
      "num_input_tokens_seen": 21289360,
      "step": 15715
    },
    {
      "epoch": 0.5031688112156713,
      "grad_norm": 44.5,
      "learning_rate": 3.3248238947480804e-05,
      "loss": 3.8788,
      "num_input_tokens_seen": 21296096,
      "step": 15720
    },
    {
      "epoch": 0.5033288521861596,
      "grad_norm": 42.75,
      "learning_rate": 3.323897055886922e-05,
      "loss": 3.7632,
      "num_input_tokens_seen": 21302944,
      "step": 15725
    },
    {
      "epoch": 0.5034888931566481,
      "grad_norm": 46.25,
      "learning_rate": 3.322970089970484e-05,
      "loss": 3.7139,
      "num_input_tokens_seen": 21309584,
      "step": 15730
    },
    {
      "epoch": 0.5036489341271365,
      "grad_norm": 28.0,
      "learning_rate": 3.3220429971417165e-05,
      "loss": 4.1425,
      "num_input_tokens_seen": 21316288,
      "step": 15735
    },
    {
      "epoch": 0.503808975097625,
      "grad_norm": 23.375,
      "learning_rate": 3.321115777543588e-05,
      "loss": 3.4657,
      "num_input_tokens_seen": 21322736,
      "step": 15740
    },
    {
      "epoch": 0.5039690160681134,
      "grad_norm": 46.75,
      "learning_rate": 3.320188431319088e-05,
      "loss": 5.0824,
      "num_input_tokens_seen": 21329936,
      "step": 15745
    },
    {
      "epoch": 0.5041290570386019,
      "grad_norm": 31.125,
      "learning_rate": 3.319260958611224e-05,
      "loss": 4.2647,
      "num_input_tokens_seen": 21336288,
      "step": 15750
    },
    {
      "epoch": 0.5042890980090903,
      "grad_norm": 30.25,
      "learning_rate": 3.3183333595630256e-05,
      "loss": 4.2538,
      "num_input_tokens_seen": 21342832,
      "step": 15755
    },
    {
      "epoch": 0.5044491389795788,
      "grad_norm": 24.0,
      "learning_rate": 3.317405634317538e-05,
      "loss": 4.4367,
      "num_input_tokens_seen": 21349632,
      "step": 15760
    },
    {
      "epoch": 0.5046091799500673,
      "grad_norm": 32.5,
      "learning_rate": 3.3164777830178315e-05,
      "loss": 4.1974,
      "num_input_tokens_seen": 21355904,
      "step": 15765
    },
    {
      "epoch": 0.5047692209205557,
      "grad_norm": 24.5,
      "learning_rate": 3.315549805806989e-05,
      "loss": 4.1742,
      "num_input_tokens_seen": 21362384,
      "step": 15770
    },
    {
      "epoch": 0.5049292618910441,
      "grad_norm": 38.75,
      "learning_rate": 3.314621702828118e-05,
      "loss": 4.3704,
      "num_input_tokens_seen": 21369200,
      "step": 15775
    },
    {
      "epoch": 0.5050893028615325,
      "grad_norm": 25.375,
      "learning_rate": 3.313693474224342e-05,
      "loss": 3.8431,
      "num_input_tokens_seen": 21376176,
      "step": 15780
    },
    {
      "epoch": 0.505249343832021,
      "grad_norm": 18.75,
      "learning_rate": 3.312765120138809e-05,
      "loss": 3.952,
      "num_input_tokens_seen": 21382976,
      "step": 15785
    },
    {
      "epoch": 0.5054093848025094,
      "grad_norm": 28.625,
      "learning_rate": 3.311836640714679e-05,
      "loss": 5.4367,
      "num_input_tokens_seen": 21389472,
      "step": 15790
    },
    {
      "epoch": 0.5055694257729979,
      "grad_norm": 28.75,
      "learning_rate": 3.310908036095137e-05,
      "loss": 3.014,
      "num_input_tokens_seen": 21396192,
      "step": 15795
    },
    {
      "epoch": 0.5057294667434863,
      "grad_norm": 23.5,
      "learning_rate": 3.309979306423386e-05,
      "loss": 5.6766,
      "num_input_tokens_seen": 21403072,
      "step": 15800
    },
    {
      "epoch": 0.5057294667434863,
      "eval_loss": 4.474778652191162,
      "eval_runtime": 490.1542,
      "eval_samples_per_second": 28.33,
      "eval_steps_per_second": 14.165,
      "num_input_tokens_seen": 21403072,
      "step": 15800
    },
    {
      "epoch": 0.5058895077139748,
      "grad_norm": 39.75,
      "learning_rate": 3.309050451842647e-05,
      "loss": 4.3501,
      "num_input_tokens_seen": 21410096,
      "step": 15805
    },
    {
      "epoch": 0.5060495486844632,
      "grad_norm": 24.125,
      "learning_rate": 3.3081214724961604e-05,
      "loss": 3.9878,
      "num_input_tokens_seen": 21416768,
      "step": 15810
    },
    {
      "epoch": 0.5062095896549517,
      "grad_norm": 16.875,
      "learning_rate": 3.307192368527188e-05,
      "loss": 3.4224,
      "num_input_tokens_seen": 21423856,
      "step": 15815
    },
    {
      "epoch": 0.5063696306254402,
      "grad_norm": 29.375,
      "learning_rate": 3.306263140079008e-05,
      "loss": 4.3502,
      "num_input_tokens_seen": 21430464,
      "step": 15820
    },
    {
      "epoch": 0.5065296715959285,
      "grad_norm": 29.25,
      "learning_rate": 3.30533378729492e-05,
      "loss": 3.4573,
      "num_input_tokens_seen": 21437568,
      "step": 15825
    },
    {
      "epoch": 0.506689712566417,
      "grad_norm": 33.5,
      "learning_rate": 3.304404310318242e-05,
      "loss": 4.3715,
      "num_input_tokens_seen": 21443888,
      "step": 15830
    },
    {
      "epoch": 0.5068497535369054,
      "grad_norm": 28.125,
      "learning_rate": 3.3034747092923105e-05,
      "loss": 3.6209,
      "num_input_tokens_seen": 21450896,
      "step": 15835
    },
    {
      "epoch": 0.5070097945073939,
      "grad_norm": 22.125,
      "learning_rate": 3.3025449843604806e-05,
      "loss": 4.8076,
      "num_input_tokens_seen": 21457600,
      "step": 15840
    },
    {
      "epoch": 0.5071698354778823,
      "grad_norm": 40.0,
      "learning_rate": 3.30161513566613e-05,
      "loss": 3.6641,
      "num_input_tokens_seen": 21464880,
      "step": 15845
    },
    {
      "epoch": 0.5073298764483708,
      "grad_norm": 36.5,
      "learning_rate": 3.3006851633526506e-05,
      "loss": 4.5667,
      "num_input_tokens_seen": 21471744,
      "step": 15850
    },
    {
      "epoch": 0.5074899174188592,
      "grad_norm": 40.0,
      "learning_rate": 3.2997550675634584e-05,
      "loss": 4.7316,
      "num_input_tokens_seen": 21478368,
      "step": 15855
    },
    {
      "epoch": 0.5076499583893477,
      "grad_norm": 33.25,
      "learning_rate": 3.2988248484419825e-05,
      "loss": 3.3191,
      "num_input_tokens_seen": 21485136,
      "step": 15860
    },
    {
      "epoch": 0.5078099993598362,
      "grad_norm": 62.0,
      "learning_rate": 3.2978945061316776e-05,
      "loss": 5.6877,
      "num_input_tokens_seen": 21491456,
      "step": 15865
    },
    {
      "epoch": 0.5079700403303246,
      "grad_norm": 17.125,
      "learning_rate": 3.296964040776013e-05,
      "loss": 3.9468,
      "num_input_tokens_seen": 21497760,
      "step": 15870
    },
    {
      "epoch": 0.508130081300813,
      "grad_norm": 21.5,
      "learning_rate": 3.296033452518478e-05,
      "loss": 4.6138,
      "num_input_tokens_seen": 21504240,
      "step": 15875
    },
    {
      "epoch": 0.5082901222713014,
      "grad_norm": 23.875,
      "learning_rate": 3.2951027415025806e-05,
      "loss": 4.6426,
      "num_input_tokens_seen": 21511024,
      "step": 15880
    },
    {
      "epoch": 0.5084501632417899,
      "grad_norm": 39.75,
      "learning_rate": 3.294171907871849e-05,
      "loss": 3.8751,
      "num_input_tokens_seen": 21517760,
      "step": 15885
    },
    {
      "epoch": 0.5086102042122783,
      "grad_norm": 36.75,
      "learning_rate": 3.293240951769828e-05,
      "loss": 3.7971,
      "num_input_tokens_seen": 21524576,
      "step": 15890
    },
    {
      "epoch": 0.5087702451827668,
      "grad_norm": 30.125,
      "learning_rate": 3.2923098733400846e-05,
      "loss": 4.1373,
      "num_input_tokens_seen": 21531984,
      "step": 15895
    },
    {
      "epoch": 0.5089302861532552,
      "grad_norm": 26.125,
      "learning_rate": 3.291378672726202e-05,
      "loss": 3.9855,
      "num_input_tokens_seen": 21538576,
      "step": 15900
    },
    {
      "epoch": 0.5090903271237437,
      "grad_norm": 36.75,
      "learning_rate": 3.2904473500717824e-05,
      "loss": 4.2044,
      "num_input_tokens_seen": 21545728,
      "step": 15905
    },
    {
      "epoch": 0.5092503680942321,
      "grad_norm": 22.625,
      "learning_rate": 3.289515905520449e-05,
      "loss": 3.905,
      "num_input_tokens_seen": 21552432,
      "step": 15910
    },
    {
      "epoch": 0.5094104090647206,
      "grad_norm": 39.25,
      "learning_rate": 3.288584339215841e-05,
      "loss": 4.4094,
      "num_input_tokens_seen": 21559040,
      "step": 15915
    },
    {
      "epoch": 0.5095704500352091,
      "grad_norm": 23.875,
      "learning_rate": 3.287652651301617e-05,
      "loss": 4.8724,
      "num_input_tokens_seen": 21565792,
      "step": 15920
    },
    {
      "epoch": 0.5097304910056975,
      "grad_norm": 65.5,
      "learning_rate": 3.286720841921457e-05,
      "loss": 4.8237,
      "num_input_tokens_seen": 21572400,
      "step": 15925
    },
    {
      "epoch": 0.509890531976186,
      "grad_norm": 28.125,
      "learning_rate": 3.285788911219056e-05,
      "loss": 4.6497,
      "num_input_tokens_seen": 21579488,
      "step": 15930
    },
    {
      "epoch": 0.5100505729466743,
      "grad_norm": 31.5,
      "learning_rate": 3.284856859338131e-05,
      "loss": 4.689,
      "num_input_tokens_seen": 21586560,
      "step": 15935
    },
    {
      "epoch": 0.5102106139171628,
      "grad_norm": 20.625,
      "learning_rate": 3.283924686422414e-05,
      "loss": 4.1299,
      "num_input_tokens_seen": 21593376,
      "step": 15940
    },
    {
      "epoch": 0.5103706548876512,
      "grad_norm": 24.375,
      "learning_rate": 3.282992392615659e-05,
      "loss": 3.8933,
      "num_input_tokens_seen": 21600352,
      "step": 15945
    },
    {
      "epoch": 0.5105306958581397,
      "grad_norm": 27.5,
      "learning_rate": 3.282059978061638e-05,
      "loss": 4.4301,
      "num_input_tokens_seen": 21607136,
      "step": 15950
    },
    {
      "epoch": 0.5106907368286281,
      "grad_norm": 33.25,
      "learning_rate": 3.28112744290414e-05,
      "loss": 4.0016,
      "num_input_tokens_seen": 21614656,
      "step": 15955
    },
    {
      "epoch": 0.5108507777991166,
      "grad_norm": 33.75,
      "learning_rate": 3.280194787286974e-05,
      "loss": 4.464,
      "num_input_tokens_seen": 21621744,
      "step": 15960
    },
    {
      "epoch": 0.511010818769605,
      "grad_norm": 39.0,
      "learning_rate": 3.2792620113539674e-05,
      "loss": 4.353,
      "num_input_tokens_seen": 21628224,
      "step": 15965
    },
    {
      "epoch": 0.5111708597400935,
      "grad_norm": 28.125,
      "learning_rate": 3.278329115248966e-05,
      "loss": 4.2976,
      "num_input_tokens_seen": 21635152,
      "step": 15970
    },
    {
      "epoch": 0.511330900710582,
      "grad_norm": 27.375,
      "learning_rate": 3.277396099115834e-05,
      "loss": 3.5623,
      "num_input_tokens_seen": 21641952,
      "step": 15975
    },
    {
      "epoch": 0.5114909416810703,
      "grad_norm": 23.125,
      "learning_rate": 3.276462963098454e-05,
      "loss": 4.0227,
      "num_input_tokens_seen": 21648736,
      "step": 15980
    },
    {
      "epoch": 0.5116509826515588,
      "grad_norm": 46.5,
      "learning_rate": 3.275529707340728e-05,
      "loss": 4.0929,
      "num_input_tokens_seen": 21655296,
      "step": 15985
    },
    {
      "epoch": 0.5118110236220472,
      "grad_norm": 27.375,
      "learning_rate": 3.274596331986574e-05,
      "loss": 4.7883,
      "num_input_tokens_seen": 21661648,
      "step": 15990
    },
    {
      "epoch": 0.5119710645925357,
      "grad_norm": 27.0,
      "learning_rate": 3.273662837179932e-05,
      "loss": 3.3194,
      "num_input_tokens_seen": 21668688,
      "step": 15995
    },
    {
      "epoch": 0.5121311055630241,
      "grad_norm": 18.875,
      "learning_rate": 3.272729223064758e-05,
      "loss": 4.0869,
      "num_input_tokens_seen": 21675712,
      "step": 16000
    },
    {
      "epoch": 0.5121311055630241,
      "eval_loss": 4.493227958679199,
      "eval_runtime": 489.7003,
      "eval_samples_per_second": 28.356,
      "eval_steps_per_second": 14.178,
      "num_input_tokens_seen": 21675712,
      "step": 16000
    },
    {
      "epoch": 0.5122911465335126,
      "grad_norm": 34.0,
      "learning_rate": 3.2717954897850264e-05,
      "loss": 4.3506,
      "num_input_tokens_seen": 21682272,
      "step": 16005
    },
    {
      "epoch": 0.512451187504001,
      "grad_norm": 33.5,
      "learning_rate": 3.270861637484733e-05,
      "loss": 3.786,
      "num_input_tokens_seen": 21689088,
      "step": 16010
    },
    {
      "epoch": 0.5126112284744895,
      "grad_norm": 28.375,
      "learning_rate": 3.2699276663078867e-05,
      "loss": 3.2936,
      "num_input_tokens_seen": 21696672,
      "step": 16015
    },
    {
      "epoch": 0.5127712694449779,
      "grad_norm": 25.25,
      "learning_rate": 3.268993576398519e-05,
      "loss": 3.3801,
      "num_input_tokens_seen": 21703504,
      "step": 16020
    },
    {
      "epoch": 0.5129313104154664,
      "grad_norm": 19.5,
      "learning_rate": 3.268059367900678e-05,
      "loss": 3.9711,
      "num_input_tokens_seen": 21710048,
      "step": 16025
    },
    {
      "epoch": 0.5130913513859549,
      "grad_norm": 31.5,
      "learning_rate": 3.26712504095843e-05,
      "loss": 5.1043,
      "num_input_tokens_seen": 21716432,
      "step": 16030
    },
    {
      "epoch": 0.5132513923564432,
      "grad_norm": 46.0,
      "learning_rate": 3.2661905957158615e-05,
      "loss": 3.9718,
      "num_input_tokens_seen": 21722864,
      "step": 16035
    },
    {
      "epoch": 0.5134114333269317,
      "grad_norm": 27.875,
      "learning_rate": 3.2652560323170734e-05,
      "loss": 4.6701,
      "num_input_tokens_seen": 21729776,
      "step": 16040
    },
    {
      "epoch": 0.5135714742974201,
      "grad_norm": 23.875,
      "learning_rate": 3.264321350906189e-05,
      "loss": 3.9333,
      "num_input_tokens_seen": 21736608,
      "step": 16045
    },
    {
      "epoch": 0.5137315152679086,
      "grad_norm": 29.625,
      "learning_rate": 3.263386551627346e-05,
      "loss": 4.1595,
      "num_input_tokens_seen": 21743024,
      "step": 16050
    },
    {
      "epoch": 0.513891556238397,
      "grad_norm": 23.75,
      "learning_rate": 3.2624516346247055e-05,
      "loss": 4.1178,
      "num_input_tokens_seen": 21749712,
      "step": 16055
    },
    {
      "epoch": 0.5140515972088855,
      "grad_norm": 18.0,
      "learning_rate": 3.2615166000424404e-05,
      "loss": 3.7591,
      "num_input_tokens_seen": 21756848,
      "step": 16060
    },
    {
      "epoch": 0.5142116381793739,
      "grad_norm": 60.5,
      "learning_rate": 3.260581448024745e-05,
      "loss": 5.4104,
      "num_input_tokens_seen": 21763008,
      "step": 16065
    },
    {
      "epoch": 0.5143716791498624,
      "grad_norm": 28.0,
      "learning_rate": 3.2596461787158335e-05,
      "loss": 4.8119,
      "num_input_tokens_seen": 21769504,
      "step": 16070
    },
    {
      "epoch": 0.5145317201203508,
      "grad_norm": 25.75,
      "learning_rate": 3.258710792259934e-05,
      "loss": 5.1325,
      "num_input_tokens_seen": 21776624,
      "step": 16075
    },
    {
      "epoch": 0.5146917610908393,
      "grad_norm": 30.0,
      "learning_rate": 3.257775288801296e-05,
      "loss": 4.3629,
      "num_input_tokens_seen": 21783680,
      "step": 16080
    },
    {
      "epoch": 0.5148518020613277,
      "grad_norm": 28.75,
      "learning_rate": 3.256839668484186e-05,
      "loss": 4.2558,
      "num_input_tokens_seen": 21790048,
      "step": 16085
    },
    {
      "epoch": 0.5150118430318161,
      "grad_norm": 36.25,
      "learning_rate": 3.255903931452888e-05,
      "loss": 4.7677,
      "num_input_tokens_seen": 21796624,
      "step": 16090
    },
    {
      "epoch": 0.5151718840023046,
      "grad_norm": 26.375,
      "learning_rate": 3.2549680778517045e-05,
      "loss": 4.4093,
      "num_input_tokens_seen": 21803568,
      "step": 16095
    },
    {
      "epoch": 0.515331924972793,
      "grad_norm": 27.75,
      "learning_rate": 3.2540321078249556e-05,
      "loss": 5.4519,
      "num_input_tokens_seen": 21810000,
      "step": 16100
    },
    {
      "epoch": 0.5154919659432815,
      "grad_norm": 27.0,
      "learning_rate": 3.2530960215169795e-05,
      "loss": 3.6859,
      "num_input_tokens_seen": 21816560,
      "step": 16105
    },
    {
      "epoch": 0.5156520069137699,
      "grad_norm": 45.75,
      "learning_rate": 3.2521598190721345e-05,
      "loss": 4.6513,
      "num_input_tokens_seen": 21823904,
      "step": 16110
    },
    {
      "epoch": 0.5158120478842584,
      "grad_norm": 40.0,
      "learning_rate": 3.251223500634792e-05,
      "loss": 5.1631,
      "num_input_tokens_seen": 21830944,
      "step": 16115
    },
    {
      "epoch": 0.5159720888547468,
      "grad_norm": 23.875,
      "learning_rate": 3.2502870663493445e-05,
      "loss": 4.3584,
      "num_input_tokens_seen": 21837792,
      "step": 16120
    },
    {
      "epoch": 0.5161321298252353,
      "grad_norm": 46.75,
      "learning_rate": 3.249350516360203e-05,
      "loss": 4.4239,
      "num_input_tokens_seen": 21844224,
      "step": 16125
    },
    {
      "epoch": 0.5162921707957238,
      "grad_norm": 33.75,
      "learning_rate": 3.248413850811797e-05,
      "loss": 4.6363,
      "num_input_tokens_seen": 21851024,
      "step": 16130
    },
    {
      "epoch": 0.5164522117662121,
      "grad_norm": 33.25,
      "learning_rate": 3.2474770698485677e-05,
      "loss": 4.8979,
      "num_input_tokens_seen": 21857920,
      "step": 16135
    },
    {
      "epoch": 0.5166122527367006,
      "grad_norm": 11.5,
      "learning_rate": 3.246540173614983e-05,
      "loss": 4.2833,
      "num_input_tokens_seen": 21864432,
      "step": 16140
    },
    {
      "epoch": 0.516772293707189,
      "grad_norm": 21.875,
      "learning_rate": 3.2456031622555197e-05,
      "loss": 4.6249,
      "num_input_tokens_seen": 21871072,
      "step": 16145
    },
    {
      "epoch": 0.5169323346776775,
      "grad_norm": 35.0,
      "learning_rate": 3.2446660359146794e-05,
      "loss": 4.8288,
      "num_input_tokens_seen": 21877680,
      "step": 16150
    },
    {
      "epoch": 0.5170923756481659,
      "grad_norm": 20.0,
      "learning_rate": 3.2437287947369786e-05,
      "loss": 3.7325,
      "num_input_tokens_seen": 21884304,
      "step": 16155
    },
    {
      "epoch": 0.5172524166186544,
      "grad_norm": 25.875,
      "learning_rate": 3.2427914388669525e-05,
      "loss": 4.1051,
      "num_input_tokens_seen": 21891440,
      "step": 16160
    },
    {
      "epoch": 0.5174124575891428,
      "grad_norm": 29.125,
      "learning_rate": 3.241853968449151e-05,
      "loss": 4.4893,
      "num_input_tokens_seen": 21899056,
      "step": 16165
    },
    {
      "epoch": 0.5175724985596313,
      "grad_norm": 29.5,
      "learning_rate": 3.240916383628144e-05,
      "loss": 3.8993,
      "num_input_tokens_seen": 21906192,
      "step": 16170
    },
    {
      "epoch": 0.5177325395301197,
      "grad_norm": 23.625,
      "learning_rate": 3.239978684548521e-05,
      "loss": 4.4224,
      "num_input_tokens_seen": 21912768,
      "step": 16175
    },
    {
      "epoch": 0.5178925805006082,
      "grad_norm": 32.0,
      "learning_rate": 3.239040871354885e-05,
      "loss": 4.7463,
      "num_input_tokens_seen": 21919088,
      "step": 16180
    },
    {
      "epoch": 0.5180526214710967,
      "grad_norm": 27.125,
      "learning_rate": 3.2381029441918596e-05,
      "loss": 4.9444,
      "num_input_tokens_seen": 21926416,
      "step": 16185
    },
    {
      "epoch": 0.518212662441585,
      "grad_norm": 29.125,
      "learning_rate": 3.2371649032040845e-05,
      "loss": 4.3724,
      "num_input_tokens_seen": 21933008,
      "step": 16190
    },
    {
      "epoch": 0.5183727034120735,
      "grad_norm": 35.75,
      "learning_rate": 3.2362267485362174e-05,
      "loss": 4.2736,
      "num_input_tokens_seen": 21940000,
      "step": 16195
    },
    {
      "epoch": 0.5185327443825619,
      "grad_norm": 43.0,
      "learning_rate": 3.235288480332934e-05,
      "loss": 4.3651,
      "num_input_tokens_seen": 21946528,
      "step": 16200
    },
    {
      "epoch": 0.5185327443825619,
      "eval_loss": 4.465407848358154,
      "eval_runtime": 489.6847,
      "eval_samples_per_second": 28.357,
      "eval_steps_per_second": 14.179,
      "num_input_tokens_seen": 21946528,
      "step": 16200
    },
    {
      "epoch": 0.5186927853530504,
      "grad_norm": 26.125,
      "learning_rate": 3.234350098738927e-05,
      "loss": 4.2114,
      "num_input_tokens_seen": 21953568,
      "step": 16205
    },
    {
      "epoch": 0.5188528263235388,
      "grad_norm": 40.25,
      "learning_rate": 3.233411603898906e-05,
      "loss": 4.2564,
      "num_input_tokens_seen": 21960160,
      "step": 16210
    },
    {
      "epoch": 0.5190128672940273,
      "grad_norm": 20.25,
      "learning_rate": 3.232472995957599e-05,
      "loss": 3.9461,
      "num_input_tokens_seen": 21967104,
      "step": 16215
    },
    {
      "epoch": 0.5191729082645157,
      "grad_norm": 21.625,
      "learning_rate": 3.231534275059751e-05,
      "loss": 4.3691,
      "num_input_tokens_seen": 21973968,
      "step": 16220
    },
    {
      "epoch": 0.5193329492350042,
      "grad_norm": 21.75,
      "learning_rate": 3.230595441350125e-05,
      "loss": 3.3833,
      "num_input_tokens_seen": 21980848,
      "step": 16225
    },
    {
      "epoch": 0.5194929902054926,
      "grad_norm": 21.125,
      "learning_rate": 3.2296564949735e-05,
      "loss": 4.0948,
      "num_input_tokens_seen": 21987392,
      "step": 16230
    },
    {
      "epoch": 0.519653031175981,
      "grad_norm": 27.0,
      "learning_rate": 3.228717436074675e-05,
      "loss": 4.5182,
      "num_input_tokens_seen": 21994528,
      "step": 16235
    },
    {
      "epoch": 0.5198130721464695,
      "grad_norm": 23.125,
      "learning_rate": 3.227778264798463e-05,
      "loss": 4.5695,
      "num_input_tokens_seen": 22001280,
      "step": 16240
    },
    {
      "epoch": 0.5199731131169579,
      "grad_norm": 27.0,
      "learning_rate": 3.226838981289698e-05,
      "loss": 4.7461,
      "num_input_tokens_seen": 22008128,
      "step": 16245
    },
    {
      "epoch": 0.5201331540874464,
      "grad_norm": 41.0,
      "learning_rate": 3.225899585693227e-05,
      "loss": 4.4245,
      "num_input_tokens_seen": 22014656,
      "step": 16250
    },
    {
      "epoch": 0.5202931950579348,
      "grad_norm": 28.25,
      "learning_rate": 3.224960078153918e-05,
      "loss": 4.9428,
      "num_input_tokens_seen": 22021440,
      "step": 16255
    },
    {
      "epoch": 0.5204532360284233,
      "grad_norm": 43.0,
      "learning_rate": 3.224020458816655e-05,
      "loss": 4.761,
      "num_input_tokens_seen": 22028176,
      "step": 16260
    },
    {
      "epoch": 0.5206132769989117,
      "grad_norm": 24.75,
      "learning_rate": 3.223080727826337e-05,
      "loss": 4.22,
      "num_input_tokens_seen": 22034560,
      "step": 16265
    },
    {
      "epoch": 0.5207733179694002,
      "grad_norm": 50.0,
      "learning_rate": 3.222140885327885e-05,
      "loss": 4.4947,
      "num_input_tokens_seen": 22041424,
      "step": 16270
    },
    {
      "epoch": 0.5209333589398886,
      "grad_norm": 18.125,
      "learning_rate": 3.221200931466234e-05,
      "loss": 4.793,
      "num_input_tokens_seen": 22048528,
      "step": 16275
    },
    {
      "epoch": 0.5210933999103771,
      "grad_norm": 34.5,
      "learning_rate": 3.220260866386336e-05,
      "loss": 5.4567,
      "num_input_tokens_seen": 22055344,
      "step": 16280
    },
    {
      "epoch": 0.5212534408808654,
      "grad_norm": 26.25,
      "learning_rate": 3.21932069023316e-05,
      "loss": 4.118,
      "num_input_tokens_seen": 22061968,
      "step": 16285
    },
    {
      "epoch": 0.5214134818513539,
      "grad_norm": 20.875,
      "learning_rate": 3.218380403151695e-05,
      "loss": 3.6845,
      "num_input_tokens_seen": 22068496,
      "step": 16290
    },
    {
      "epoch": 0.5215735228218424,
      "grad_norm": 40.75,
      "learning_rate": 3.217440005286943e-05,
      "loss": 3.4416,
      "num_input_tokens_seen": 22075536,
      "step": 16295
    },
    {
      "epoch": 0.5217335637923308,
      "grad_norm": 22.5,
      "learning_rate": 3.216499496783928e-05,
      "loss": 3.8704,
      "num_input_tokens_seen": 22082592,
      "step": 16300
    },
    {
      "epoch": 0.5218936047628193,
      "grad_norm": 23.5,
      "learning_rate": 3.2155588777876856e-05,
      "loss": 4.235,
      "num_input_tokens_seen": 22089248,
      "step": 16305
    },
    {
      "epoch": 0.5220536457333077,
      "grad_norm": 22.25,
      "learning_rate": 3.214618148443273e-05,
      "loss": 5.5924,
      "num_input_tokens_seen": 22095856,
      "step": 16310
    },
    {
      "epoch": 0.5222136867037962,
      "grad_norm": 37.0,
      "learning_rate": 3.2136773088957595e-05,
      "loss": 5.1155,
      "num_input_tokens_seen": 22102416,
      "step": 16315
    },
    {
      "epoch": 0.5223737276742846,
      "grad_norm": 31.5,
      "learning_rate": 3.2127363592902374e-05,
      "loss": 5.1699,
      "num_input_tokens_seen": 22109280,
      "step": 16320
    },
    {
      "epoch": 0.5225337686447731,
      "grad_norm": 34.75,
      "learning_rate": 3.211795299771812e-05,
      "loss": 4.7334,
      "num_input_tokens_seen": 22116784,
      "step": 16325
    },
    {
      "epoch": 0.5226938096152615,
      "grad_norm": 27.25,
      "learning_rate": 3.210854130485605e-05,
      "loss": 3.8686,
      "num_input_tokens_seen": 22123232,
      "step": 16330
    },
    {
      "epoch": 0.52285385058575,
      "grad_norm": 21.375,
      "learning_rate": 3.209912851576759e-05,
      "loss": 4.035,
      "num_input_tokens_seen": 22130016,
      "step": 16335
    },
    {
      "epoch": 0.5230138915562383,
      "grad_norm": 42.0,
      "learning_rate": 3.208971463190431e-05,
      "loss": 4.9065,
      "num_input_tokens_seen": 22136768,
      "step": 16340
    },
    {
      "epoch": 0.5231739325267268,
      "grad_norm": 49.5,
      "learning_rate": 3.208029965471793e-05,
      "loss": 4.3945,
      "num_input_tokens_seen": 22142992,
      "step": 16345
    },
    {
      "epoch": 0.5233339734972153,
      "grad_norm": 32.25,
      "learning_rate": 3.2070883585660364e-05,
      "loss": 4.8508,
      "num_input_tokens_seen": 22149440,
      "step": 16350
    },
    {
      "epoch": 0.5234940144677037,
      "grad_norm": 25.0,
      "learning_rate": 3.20614664261837e-05,
      "loss": 3.9184,
      "num_input_tokens_seen": 22155872,
      "step": 16355
    },
    {
      "epoch": 0.5236540554381922,
      "grad_norm": 17.875,
      "learning_rate": 3.205204817774016e-05,
      "loss": 3.4352,
      "num_input_tokens_seen": 22162752,
      "step": 16360
    },
    {
      "epoch": 0.5238140964086806,
      "grad_norm": 26.25,
      "learning_rate": 3.204262884178218e-05,
      "loss": 4.0107,
      "num_input_tokens_seen": 22169184,
      "step": 16365
    },
    {
      "epoch": 0.5239741373791691,
      "grad_norm": 23.375,
      "learning_rate": 3.2033208419762314e-05,
      "loss": 3.8485,
      "num_input_tokens_seen": 22175872,
      "step": 16370
    },
    {
      "epoch": 0.5241341783496575,
      "grad_norm": 28.5,
      "learning_rate": 3.2023786913133344e-05,
      "loss": 4.9041,
      "num_input_tokens_seen": 22183216,
      "step": 16375
    },
    {
      "epoch": 0.524294219320146,
      "grad_norm": 26.25,
      "learning_rate": 3.201436432334816e-05,
      "loss": 4.369,
      "num_input_tokens_seen": 22190112,
      "step": 16380
    },
    {
      "epoch": 0.5244542602906344,
      "grad_norm": 22.25,
      "learning_rate": 3.2004940651859844e-05,
      "loss": 3.9576,
      "num_input_tokens_seen": 22197440,
      "step": 16385
    },
    {
      "epoch": 0.5246143012611229,
      "grad_norm": 24.875,
      "learning_rate": 3.1995515900121655e-05,
      "loss": 3.9191,
      "num_input_tokens_seen": 22204160,
      "step": 16390
    },
    {
      "epoch": 0.5247743422316113,
      "grad_norm": 26.0,
      "learning_rate": 3.1986090069587e-05,
      "loss": 4.8431,
      "num_input_tokens_seen": 22210992,
      "step": 16395
    },
    {
      "epoch": 0.5249343832020997,
      "grad_norm": 28.875,
      "learning_rate": 3.1976663161709466e-05,
      "loss": 4.2341,
      "num_input_tokens_seen": 22217936,
      "step": 16400
    },
    {
      "epoch": 0.5249343832020997,
      "eval_loss": 4.463719844818115,
      "eval_runtime": 489.556,
      "eval_samples_per_second": 28.364,
      "eval_steps_per_second": 14.182,
      "num_input_tokens_seen": 22217936,
      "step": 16400
    },
    {
      "epoch": 0.5250944241725882,
      "grad_norm": 41.75,
      "learning_rate": 3.196723517794279e-05,
      "loss": 4.7982,
      "num_input_tokens_seen": 22224864,
      "step": 16405
    },
    {
      "epoch": 0.5252544651430766,
      "grad_norm": 40.0,
      "learning_rate": 3.19578061197409e-05,
      "loss": 4.2153,
      "num_input_tokens_seen": 22231200,
      "step": 16410
    },
    {
      "epoch": 0.5254145061135651,
      "grad_norm": 23.75,
      "learning_rate": 3.194837598855787e-05,
      "loss": 4.8304,
      "num_input_tokens_seen": 22238064,
      "step": 16415
    },
    {
      "epoch": 0.5255745470840535,
      "grad_norm": 33.0,
      "learning_rate": 3.193894478584794e-05,
      "loss": 4.3837,
      "num_input_tokens_seen": 22244768,
      "step": 16420
    },
    {
      "epoch": 0.525734588054542,
      "grad_norm": 23.625,
      "learning_rate": 3.192951251306553e-05,
      "loss": 4.0981,
      "num_input_tokens_seen": 22251824,
      "step": 16425
    },
    {
      "epoch": 0.5258946290250304,
      "grad_norm": 19.625,
      "learning_rate": 3.192007917166521e-05,
      "loss": 4.251,
      "num_input_tokens_seen": 22258752,
      "step": 16430
    },
    {
      "epoch": 0.5260546699955189,
      "grad_norm": 20.75,
      "learning_rate": 3.191064476310171e-05,
      "loss": 5.3041,
      "num_input_tokens_seen": 22265472,
      "step": 16435
    },
    {
      "epoch": 0.5262147109660072,
      "grad_norm": 30.25,
      "learning_rate": 3.1901209288829944e-05,
      "loss": 3.4025,
      "num_input_tokens_seen": 22272192,
      "step": 16440
    },
    {
      "epoch": 0.5263747519364957,
      "grad_norm": 20.125,
      "learning_rate": 3.1891772750304985e-05,
      "loss": 4.3609,
      "num_input_tokens_seen": 22278832,
      "step": 16445
    },
    {
      "epoch": 0.5265347929069842,
      "grad_norm": 21.0,
      "learning_rate": 3.188233514898206e-05,
      "loss": 4.0639,
      "num_input_tokens_seen": 22286208,
      "step": 16450
    },
    {
      "epoch": 0.5266948338774726,
      "grad_norm": 35.0,
      "learning_rate": 3.187289648631657e-05,
      "loss": 4.2026,
      "num_input_tokens_seen": 22292848,
      "step": 16455
    },
    {
      "epoch": 0.5268548748479611,
      "grad_norm": 36.75,
      "learning_rate": 3.186345676376406e-05,
      "loss": 3.9446,
      "num_input_tokens_seen": 22299760,
      "step": 16460
    },
    {
      "epoch": 0.5270149158184495,
      "grad_norm": 36.0,
      "learning_rate": 3.1854015982780275e-05,
      "loss": 4.2884,
      "num_input_tokens_seen": 22306464,
      "step": 16465
    },
    {
      "epoch": 0.527174956788938,
      "grad_norm": 20.0,
      "learning_rate": 3.1844574144821084e-05,
      "loss": 4.4002,
      "num_input_tokens_seen": 22313328,
      "step": 16470
    },
    {
      "epoch": 0.5273349977594264,
      "grad_norm": 27.75,
      "learning_rate": 3.1835131251342554e-05,
      "loss": 4.9165,
      "num_input_tokens_seen": 22320016,
      "step": 16475
    },
    {
      "epoch": 0.5274950387299149,
      "grad_norm": 25.0,
      "learning_rate": 3.182568730380089e-05,
      "loss": 4.0926,
      "num_input_tokens_seen": 22327088,
      "step": 16480
    },
    {
      "epoch": 0.5276550797004033,
      "grad_norm": 26.875,
      "learning_rate": 3.181624230365245e-05,
      "loss": 4.1966,
      "num_input_tokens_seen": 22334224,
      "step": 16485
    },
    {
      "epoch": 0.5278151206708918,
      "grad_norm": 28.5,
      "learning_rate": 3.180679625235381e-05,
      "loss": 4.5757,
      "num_input_tokens_seen": 22340816,
      "step": 16490
    },
    {
      "epoch": 0.5279751616413801,
      "grad_norm": 17.375,
      "learning_rate": 3.1797349151361646e-05,
      "loss": 4.3434,
      "num_input_tokens_seen": 22347440,
      "step": 16495
    },
    {
      "epoch": 0.5281352026118686,
      "grad_norm": 24.125,
      "learning_rate": 3.178790100213281e-05,
      "loss": 3.9167,
      "num_input_tokens_seen": 22354064,
      "step": 16500
    },
    {
      "epoch": 0.5282952435823571,
      "grad_norm": 30.875,
      "learning_rate": 3.1778451806124346e-05,
      "loss": 4.4216,
      "num_input_tokens_seen": 22360928,
      "step": 16505
    },
    {
      "epoch": 0.5284552845528455,
      "grad_norm": 20.0,
      "learning_rate": 3.176900156479342e-05,
      "loss": 4.302,
      "num_input_tokens_seen": 22367824,
      "step": 16510
    },
    {
      "epoch": 0.528615325523334,
      "grad_norm": 18.125,
      "learning_rate": 3.17595502795974e-05,
      "loss": 4.0098,
      "num_input_tokens_seen": 22374896,
      "step": 16515
    },
    {
      "epoch": 0.5287753664938224,
      "grad_norm": 16.125,
      "learning_rate": 3.175009795199377e-05,
      "loss": 4.3981,
      "num_input_tokens_seen": 22381824,
      "step": 16520
    },
    {
      "epoch": 0.5289354074643109,
      "grad_norm": 28.875,
      "learning_rate": 3.1740644583440224e-05,
      "loss": 4.2709,
      "num_input_tokens_seen": 22388272,
      "step": 16525
    },
    {
      "epoch": 0.5290954484347993,
      "grad_norm": 38.0,
      "learning_rate": 3.173119017539457e-05,
      "loss": 4.7923,
      "num_input_tokens_seen": 22394448,
      "step": 16530
    },
    {
      "epoch": 0.5292554894052878,
      "grad_norm": 34.25,
      "learning_rate": 3.172173472931479e-05,
      "loss": 4.8309,
      "num_input_tokens_seen": 22401424,
      "step": 16535
    },
    {
      "epoch": 0.5294155303757762,
      "grad_norm": 37.75,
      "learning_rate": 3.1712278246659055e-05,
      "loss": 4.3772,
      "num_input_tokens_seen": 22408528,
      "step": 16540
    },
    {
      "epoch": 0.5295755713462647,
      "grad_norm": 22.625,
      "learning_rate": 3.170282072888566e-05,
      "loss": 4.3032,
      "num_input_tokens_seen": 22415536,
      "step": 16545
    },
    {
      "epoch": 0.529735612316753,
      "grad_norm": 30.5,
      "learning_rate": 3.169336217745307e-05,
      "loss": 4.4086,
      "num_input_tokens_seen": 22422240,
      "step": 16550
    },
    {
      "epoch": 0.5298956532872415,
      "grad_norm": 21.375,
      "learning_rate": 3.1683902593819924e-05,
      "loss": 4.5571,
      "num_input_tokens_seen": 22429152,
      "step": 16555
    },
    {
      "epoch": 0.53005569425773,
      "grad_norm": 21.125,
      "learning_rate": 3.1674441979445e-05,
      "loss": 3.5718,
      "num_input_tokens_seen": 22435904,
      "step": 16560
    },
    {
      "epoch": 0.5302157352282184,
      "grad_norm": 32.25,
      "learning_rate": 3.166498033578725e-05,
      "loss": 4.1239,
      "num_input_tokens_seen": 22442976,
      "step": 16565
    },
    {
      "epoch": 0.5303757761987069,
      "grad_norm": 34.5,
      "learning_rate": 3.165551766430578e-05,
      "loss": 5.2896,
      "num_input_tokens_seen": 22449360,
      "step": 16570
    },
    {
      "epoch": 0.5305358171691953,
      "grad_norm": 28.0,
      "learning_rate": 3.164605396645984e-05,
      "loss": 3.677,
      "num_input_tokens_seen": 22455984,
      "step": 16575
    },
    {
      "epoch": 0.5306958581396838,
      "grad_norm": 38.75,
      "learning_rate": 3.163658924370886e-05,
      "loss": 4.6467,
      "num_input_tokens_seen": 22462304,
      "step": 16580
    },
    {
      "epoch": 0.5308558991101722,
      "grad_norm": 17.75,
      "learning_rate": 3.1627123497512415e-05,
      "loss": 4.7058,
      "num_input_tokens_seen": 22469312,
      "step": 16585
    },
    {
      "epoch": 0.5310159400806607,
      "grad_norm": 36.5,
      "learning_rate": 3.1617656729330245e-05,
      "loss": 4.7391,
      "num_input_tokens_seen": 22475776,
      "step": 16590
    },
    {
      "epoch": 0.531175981051149,
      "grad_norm": 38.0,
      "learning_rate": 3.1608188940622255e-05,
      "loss": 5.36,
      "num_input_tokens_seen": 22482512,
      "step": 16595
    },
    {
      "epoch": 0.5313360220216375,
      "grad_norm": 22.75,
      "learning_rate": 3.159872013284847e-05,
      "loss": 4.0066,
      "num_input_tokens_seen": 22489168,
      "step": 16600
    },
    {
      "epoch": 0.5313360220216375,
      "eval_loss": 4.488368511199951,
      "eval_runtime": 489.5613,
      "eval_samples_per_second": 28.364,
      "eval_steps_per_second": 14.182,
      "num_input_tokens_seen": 22489168,
      "step": 16600
    },
    {
      "epoch": 0.531496062992126,
      "grad_norm": 29.125,
      "learning_rate": 3.1589250307469134e-05,
      "loss": 4.5506,
      "num_input_tokens_seen": 22495904,
      "step": 16605
    },
    {
      "epoch": 0.5316561039626144,
      "grad_norm": 21.5,
      "learning_rate": 3.1579779465944586e-05,
      "loss": 4.5966,
      "num_input_tokens_seen": 22502608,
      "step": 16610
    },
    {
      "epoch": 0.5318161449331029,
      "grad_norm": 25.375,
      "learning_rate": 3.1570307609735363e-05,
      "loss": 4.9443,
      "num_input_tokens_seen": 22509968,
      "step": 16615
    },
    {
      "epoch": 0.5319761859035913,
      "grad_norm": 26.5,
      "learning_rate": 3.156083474030213e-05,
      "loss": 3.9617,
      "num_input_tokens_seen": 22516624,
      "step": 16620
    },
    {
      "epoch": 0.5321362268740798,
      "grad_norm": 36.75,
      "learning_rate": 3.155136085910573e-05,
      "loss": 5.0082,
      "num_input_tokens_seen": 22523328,
      "step": 16625
    },
    {
      "epoch": 0.5322962678445682,
      "grad_norm": 18.375,
      "learning_rate": 3.154188596760717e-05,
      "loss": 4.2732,
      "num_input_tokens_seen": 22530160,
      "step": 16630
    },
    {
      "epoch": 0.5324563088150567,
      "grad_norm": 24.25,
      "learning_rate": 3.153241006726757e-05,
      "loss": 4.0898,
      "num_input_tokens_seen": 22536912,
      "step": 16635
    },
    {
      "epoch": 0.5326163497855451,
      "grad_norm": 10.25,
      "learning_rate": 3.152293315954825e-05,
      "loss": 3.0444,
      "num_input_tokens_seen": 22543744,
      "step": 16640
    },
    {
      "epoch": 0.5327763907560336,
      "grad_norm": 42.25,
      "learning_rate": 3.1513455245910666e-05,
      "loss": 5.7733,
      "num_input_tokens_seen": 22550656,
      "step": 16645
    },
    {
      "epoch": 0.5329364317265219,
      "grad_norm": 23.875,
      "learning_rate": 3.150397632781643e-05,
      "loss": 4.1205,
      "num_input_tokens_seen": 22557376,
      "step": 16650
    },
    {
      "epoch": 0.5330964726970104,
      "grad_norm": 28.25,
      "learning_rate": 3.149449640672731e-05,
      "loss": 4.2041,
      "num_input_tokens_seen": 22564272,
      "step": 16655
    },
    {
      "epoch": 0.5332565136674989,
      "grad_norm": 26.625,
      "learning_rate": 3.148501548410523e-05,
      "loss": 3.9003,
      "num_input_tokens_seen": 22570896,
      "step": 16660
    },
    {
      "epoch": 0.5334165546379873,
      "grad_norm": 20.25,
      "learning_rate": 3.1475533561412256e-05,
      "loss": 4.0309,
      "num_input_tokens_seen": 22577520,
      "step": 16665
    },
    {
      "epoch": 0.5335765956084758,
      "grad_norm": 38.0,
      "learning_rate": 3.146605064011065e-05,
      "loss": 3.6333,
      "num_input_tokens_seen": 22583952,
      "step": 16670
    },
    {
      "epoch": 0.5337366365789642,
      "grad_norm": 30.625,
      "learning_rate": 3.145656672166277e-05,
      "loss": 3.8694,
      "num_input_tokens_seen": 22590576,
      "step": 16675
    },
    {
      "epoch": 0.5338966775494527,
      "grad_norm": 29.875,
      "learning_rate": 3.144708180753116e-05,
      "loss": 4.4339,
      "num_input_tokens_seen": 22597440,
      "step": 16680
    },
    {
      "epoch": 0.5340567185199411,
      "grad_norm": 43.0,
      "learning_rate": 3.143759589917851e-05,
      "loss": 4.2861,
      "num_input_tokens_seen": 22604096,
      "step": 16685
    },
    {
      "epoch": 0.5342167594904296,
      "grad_norm": 42.0,
      "learning_rate": 3.142810899806768e-05,
      "loss": 5.7505,
      "num_input_tokens_seen": 22610912,
      "step": 16690
    },
    {
      "epoch": 0.534376800460918,
      "grad_norm": 25.25,
      "learning_rate": 3.141862110566166e-05,
      "loss": 3.5456,
      "num_input_tokens_seen": 22618032,
      "step": 16695
    },
    {
      "epoch": 0.5345368414314065,
      "grad_norm": 18.875,
      "learning_rate": 3.1409132223423606e-05,
      "loss": 3.9942,
      "num_input_tokens_seen": 22624816,
      "step": 16700
    },
    {
      "epoch": 0.5346968824018948,
      "grad_norm": 22.875,
      "learning_rate": 3.139964235281682e-05,
      "loss": 4.7415,
      "num_input_tokens_seen": 22631680,
      "step": 16705
    },
    {
      "epoch": 0.5348569233723833,
      "grad_norm": 22.25,
      "learning_rate": 3.139015149530476e-05,
      "loss": 4.0926,
      "num_input_tokens_seen": 22638208,
      "step": 16710
    },
    {
      "epoch": 0.5350169643428718,
      "grad_norm": 28.375,
      "learning_rate": 3.1380659652351034e-05,
      "loss": 3.992,
      "num_input_tokens_seen": 22644704,
      "step": 16715
    },
    {
      "epoch": 0.5351770053133602,
      "grad_norm": 27.5,
      "learning_rate": 3.137116682541941e-05,
      "loss": 3.7794,
      "num_input_tokens_seen": 22651568,
      "step": 16720
    },
    {
      "epoch": 0.5353370462838487,
      "grad_norm": 41.25,
      "learning_rate": 3.136167301597379e-05,
      "loss": 4.3451,
      "num_input_tokens_seen": 22658576,
      "step": 16725
    },
    {
      "epoch": 0.5354970872543371,
      "grad_norm": 20.0,
      "learning_rate": 3.1352178225478254e-05,
      "loss": 3.6667,
      "num_input_tokens_seen": 22665152,
      "step": 16730
    },
    {
      "epoch": 0.5356571282248256,
      "grad_norm": 26.125,
      "learning_rate": 3.1342682455396996e-05,
      "loss": 4.9628,
      "num_input_tokens_seen": 22671632,
      "step": 16735
    },
    {
      "epoch": 0.535817169195314,
      "grad_norm": 26.375,
      "learning_rate": 3.133318570719441e-05,
      "loss": 3.79,
      "num_input_tokens_seen": 22678720,
      "step": 16740
    },
    {
      "epoch": 0.5359772101658025,
      "grad_norm": 19.25,
      "learning_rate": 3.132368798233499e-05,
      "loss": 3.5625,
      "num_input_tokens_seen": 22684944,
      "step": 16745
    },
    {
      "epoch": 0.5361372511362908,
      "grad_norm": 15.875,
      "learning_rate": 3.131418928228342e-05,
      "loss": 4.1593,
      "num_input_tokens_seen": 22691392,
      "step": 16750
    },
    {
      "epoch": 0.5362972921067793,
      "grad_norm": 19.375,
      "learning_rate": 3.1304689608504514e-05,
      "loss": 3.4137,
      "num_input_tokens_seen": 22697984,
      "step": 16755
    },
    {
      "epoch": 0.5364573330772677,
      "grad_norm": 30.75,
      "learning_rate": 3.129518896246324e-05,
      "loss": 4.4734,
      "num_input_tokens_seen": 22704736,
      "step": 16760
    },
    {
      "epoch": 0.5366173740477562,
      "grad_norm": 24.125,
      "learning_rate": 3.128568734562472e-05,
      "loss": 4.1412,
      "num_input_tokens_seen": 22711552,
      "step": 16765
    },
    {
      "epoch": 0.5367774150182447,
      "grad_norm": 15.5,
      "learning_rate": 3.127618475945421e-05,
      "loss": 4.7095,
      "num_input_tokens_seen": 22718240,
      "step": 16770
    },
    {
      "epoch": 0.5369374559887331,
      "grad_norm": 20.375,
      "learning_rate": 3.126668120541715e-05,
      "loss": 4.7401,
      "num_input_tokens_seen": 22724944,
      "step": 16775
    },
    {
      "epoch": 0.5370974969592216,
      "grad_norm": 28.875,
      "learning_rate": 3.1257176684979096e-05,
      "loss": 4.5988,
      "num_input_tokens_seen": 22732096,
      "step": 16780
    },
    {
      "epoch": 0.53725753792971,
      "grad_norm": 32.75,
      "learning_rate": 3.124767119960576e-05,
      "loss": 4.5462,
      "num_input_tokens_seen": 22739040,
      "step": 16785
    },
    {
      "epoch": 0.5374175789001985,
      "grad_norm": 32.25,
      "learning_rate": 3.123816475076301e-05,
      "loss": 3.8541,
      "num_input_tokens_seen": 22745744,
      "step": 16790
    },
    {
      "epoch": 0.5375776198706869,
      "grad_norm": 41.75,
      "learning_rate": 3.122865733991687e-05,
      "loss": 3.898,
      "num_input_tokens_seen": 22752592,
      "step": 16795
    },
    {
      "epoch": 0.5377376608411754,
      "grad_norm": 19.75,
      "learning_rate": 3.1219148968533486e-05,
      "loss": 4.5365,
      "num_input_tokens_seen": 22759200,
      "step": 16800
    },
    {
      "epoch": 0.5377376608411754,
      "eval_loss": 4.4925689697265625,
      "eval_runtime": 489.5289,
      "eval_samples_per_second": 28.366,
      "eval_steps_per_second": 14.183,
      "num_input_tokens_seen": 22759200,
      "step": 16800
    },
    {
      "epoch": 0.5378977018116637,
      "grad_norm": 37.0,
      "learning_rate": 3.120963963807918e-05,
      "loss": 4.0922,
      "num_input_tokens_seen": 22765728,
      "step": 16805
    },
    {
      "epoch": 0.5380577427821522,
      "grad_norm": 36.5,
      "learning_rate": 3.12001293500204e-05,
      "loss": 3.81,
      "num_input_tokens_seen": 22772240,
      "step": 16810
    },
    {
      "epoch": 0.5382177837526406,
      "grad_norm": 29.625,
      "learning_rate": 3.1190618105823765e-05,
      "loss": 3.7666,
      "num_input_tokens_seen": 22778800,
      "step": 16815
    },
    {
      "epoch": 0.5383778247231291,
      "grad_norm": 25.5,
      "learning_rate": 3.118110590695603e-05,
      "loss": 3.251,
      "num_input_tokens_seen": 22785408,
      "step": 16820
    },
    {
      "epoch": 0.5385378656936176,
      "grad_norm": 24.625,
      "learning_rate": 3.117159275488407e-05,
      "loss": 3.2859,
      "num_input_tokens_seen": 22792032,
      "step": 16825
    },
    {
      "epoch": 0.538697906664106,
      "grad_norm": 16.75,
      "learning_rate": 3.1162078651074956e-05,
      "loss": 3.4272,
      "num_input_tokens_seen": 22798992,
      "step": 16830
    },
    {
      "epoch": 0.5388579476345945,
      "grad_norm": 24.875,
      "learning_rate": 3.1152563596995885e-05,
      "loss": 4.2917,
      "num_input_tokens_seen": 22805904,
      "step": 16835
    },
    {
      "epoch": 0.5390179886050829,
      "grad_norm": 38.25,
      "learning_rate": 3.1143047594114186e-05,
      "loss": 4.3568,
      "num_input_tokens_seen": 22812592,
      "step": 16840
    },
    {
      "epoch": 0.5391780295755714,
      "grad_norm": 33.0,
      "learning_rate": 3.113353064389734e-05,
      "loss": 4.9394,
      "num_input_tokens_seen": 22819312,
      "step": 16845
    },
    {
      "epoch": 0.5393380705460598,
      "grad_norm": 35.75,
      "learning_rate": 3.1124012747812993e-05,
      "loss": 4.0076,
      "num_input_tokens_seen": 22825568,
      "step": 16850
    },
    {
      "epoch": 0.5394981115165483,
      "grad_norm": 25.25,
      "learning_rate": 3.1114493907328936e-05,
      "loss": 3.7971,
      "num_input_tokens_seen": 22832480,
      "step": 16855
    },
    {
      "epoch": 0.5396581524870366,
      "grad_norm": 20.875,
      "learning_rate": 3.110497412391306e-05,
      "loss": 4.8621,
      "num_input_tokens_seen": 22839616,
      "step": 16860
    },
    {
      "epoch": 0.5398181934575251,
      "grad_norm": 24.625,
      "learning_rate": 3.1095453399033466e-05,
      "loss": 3.8844,
      "num_input_tokens_seen": 22846096,
      "step": 16865
    },
    {
      "epoch": 0.5399782344280136,
      "grad_norm": 21.25,
      "learning_rate": 3.108593173415835e-05,
      "loss": 3.6366,
      "num_input_tokens_seen": 22852912,
      "step": 16870
    },
    {
      "epoch": 0.540138275398502,
      "grad_norm": 28.0,
      "learning_rate": 3.107640913075609e-05,
      "loss": 4.1113,
      "num_input_tokens_seen": 22860416,
      "step": 16875
    },
    {
      "epoch": 0.5402983163689905,
      "grad_norm": 26.125,
      "learning_rate": 3.106688559029517e-05,
      "loss": 4.2421,
      "num_input_tokens_seen": 22867392,
      "step": 16880
    },
    {
      "epoch": 0.5404583573394789,
      "grad_norm": 25.625,
      "learning_rate": 3.105736111424425e-05,
      "loss": 4.2191,
      "num_input_tokens_seen": 22873872,
      "step": 16885
    },
    {
      "epoch": 0.5406183983099674,
      "grad_norm": 26.875,
      "learning_rate": 3.1047835704072136e-05,
      "loss": 4.1104,
      "num_input_tokens_seen": 22881152,
      "step": 16890
    },
    {
      "epoch": 0.5407784392804558,
      "grad_norm": 21.125,
      "learning_rate": 3.103830936124775e-05,
      "loss": 4.3938,
      "num_input_tokens_seen": 22887872,
      "step": 16895
    },
    {
      "epoch": 0.5409384802509443,
      "grad_norm": 36.75,
      "learning_rate": 3.102878208724018e-05,
      "loss": 4.3878,
      "num_input_tokens_seen": 22894560,
      "step": 16900
    },
    {
      "epoch": 0.5410985212214326,
      "grad_norm": 32.0,
      "learning_rate": 3.101925388351865e-05,
      "loss": 4.3696,
      "num_input_tokens_seen": 22900992,
      "step": 16905
    },
    {
      "epoch": 0.5412585621919211,
      "grad_norm": 40.75,
      "learning_rate": 3.1009724751552515e-05,
      "loss": 4.1348,
      "num_input_tokens_seen": 22907536,
      "step": 16910
    },
    {
      "epoch": 0.5414186031624095,
      "grad_norm": 32.0,
      "learning_rate": 3.100019469281131e-05,
      "loss": 4.319,
      "num_input_tokens_seen": 22914288,
      "step": 16915
    },
    {
      "epoch": 0.541578644132898,
      "grad_norm": 30.5,
      "learning_rate": 3.0990663708764685e-05,
      "loss": 4.0497,
      "num_input_tokens_seen": 22920928,
      "step": 16920
    },
    {
      "epoch": 0.5417386851033865,
      "grad_norm": 44.75,
      "learning_rate": 3.098113180088243e-05,
      "loss": 3.5132,
      "num_input_tokens_seen": 22927456,
      "step": 16925
    },
    {
      "epoch": 0.5418987260738749,
      "grad_norm": 30.25,
      "learning_rate": 3.097159897063448e-05,
      "loss": 4.2238,
      "num_input_tokens_seen": 22934096,
      "step": 16930
    },
    {
      "epoch": 0.5420587670443634,
      "grad_norm": 19.5,
      "learning_rate": 3.096206521949094e-05,
      "loss": 4.432,
      "num_input_tokens_seen": 22940704,
      "step": 16935
    },
    {
      "epoch": 0.5422188080148518,
      "grad_norm": 27.5,
      "learning_rate": 3.0952530548922006e-05,
      "loss": 4.2829,
      "num_input_tokens_seen": 22947584,
      "step": 16940
    },
    {
      "epoch": 0.5423788489853403,
      "grad_norm": 33.0,
      "learning_rate": 3.0942994960398064e-05,
      "loss": 3.1362,
      "num_input_tokens_seen": 22954272,
      "step": 16945
    },
    {
      "epoch": 0.5425388899558287,
      "grad_norm": 29.25,
      "learning_rate": 3.093345845538961e-05,
      "loss": 3.7613,
      "num_input_tokens_seen": 22961152,
      "step": 16950
    },
    {
      "epoch": 0.5426989309263172,
      "grad_norm": 30.375,
      "learning_rate": 3.09239210353673e-05,
      "loss": 4.059,
      "num_input_tokens_seen": 22967936,
      "step": 16955
    },
    {
      "epoch": 0.5428589718968055,
      "grad_norm": 25.0,
      "learning_rate": 3.0914382701801926e-05,
      "loss": 4.6801,
      "num_input_tokens_seen": 22974848,
      "step": 16960
    },
    {
      "epoch": 0.543019012867294,
      "grad_norm": 15.25,
      "learning_rate": 3.090484345616441e-05,
      "loss": 3.4041,
      "num_input_tokens_seen": 22981408,
      "step": 16965
    },
    {
      "epoch": 0.5431790538377824,
      "grad_norm": 17.5,
      "learning_rate": 3.0895303299925825e-05,
      "loss": 4.5401,
      "num_input_tokens_seen": 22987968,
      "step": 16970
    },
    {
      "epoch": 0.5433390948082709,
      "grad_norm": 21.25,
      "learning_rate": 3.0885762234557393e-05,
      "loss": 4.0864,
      "num_input_tokens_seen": 22994736,
      "step": 16975
    },
    {
      "epoch": 0.5434991357787594,
      "grad_norm": 24.375,
      "learning_rate": 3.087622026153045e-05,
      "loss": 4.5626,
      "num_input_tokens_seen": 23001664,
      "step": 16980
    },
    {
      "epoch": 0.5436591767492478,
      "grad_norm": 23.375,
      "learning_rate": 3.086667738231651e-05,
      "loss": 5.1516,
      "num_input_tokens_seen": 23008480,
      "step": 16985
    },
    {
      "epoch": 0.5438192177197363,
      "grad_norm": 27.625,
      "learning_rate": 3.085713359838718e-05,
      "loss": 3.3581,
      "num_input_tokens_seen": 23015168,
      "step": 16990
    },
    {
      "epoch": 0.5439792586902247,
      "grad_norm": 18.625,
      "learning_rate": 3.084758891121425e-05,
      "loss": 4.1467,
      "num_input_tokens_seen": 23021552,
      "step": 16995
    },
    {
      "epoch": 0.5441392996607132,
      "grad_norm": 32.0,
      "learning_rate": 3.083804332226963e-05,
      "loss": 4.2455,
      "num_input_tokens_seen": 23028128,
      "step": 17000
    },
    {
      "epoch": 0.5441392996607132,
      "eval_loss": 4.49400520324707,
      "eval_runtime": 489.7114,
      "eval_samples_per_second": 28.355,
      "eval_steps_per_second": 14.178,
      "num_input_tokens_seen": 23028128,
      "step": 17000
    },
    {
      "epoch": 0.5442993406312016,
      "grad_norm": 32.25,
      "learning_rate": 3.082849683302536e-05,
      "loss": 4.3491,
      "num_input_tokens_seen": 23034880,
      "step": 17005
    },
    {
      "epoch": 0.54445938160169,
      "grad_norm": 20.375,
      "learning_rate": 3.081894944495363e-05,
      "loss": 5.0644,
      "num_input_tokens_seen": 23041856,
      "step": 17010
    },
    {
      "epoch": 0.5446194225721784,
      "grad_norm": 43.25,
      "learning_rate": 3.080940115952677e-05,
      "loss": 5.102,
      "num_input_tokens_seen": 23048848,
      "step": 17015
    },
    {
      "epoch": 0.5447794635426669,
      "grad_norm": 38.25,
      "learning_rate": 3.0799851978217245e-05,
      "loss": 4.8121,
      "num_input_tokens_seen": 23055776,
      "step": 17020
    },
    {
      "epoch": 0.5449395045131553,
      "grad_norm": 29.25,
      "learning_rate": 3.0790301902497666e-05,
      "loss": 4.291,
      "num_input_tokens_seen": 23062976,
      "step": 17025
    },
    {
      "epoch": 0.5450995454836438,
      "grad_norm": 38.0,
      "learning_rate": 3.078075093384076e-05,
      "loss": 4.6813,
      "num_input_tokens_seen": 23069920,
      "step": 17030
    },
    {
      "epoch": 0.5452595864541323,
      "grad_norm": 28.25,
      "learning_rate": 3.077119907371942e-05,
      "loss": 4.311,
      "num_input_tokens_seen": 23076640,
      "step": 17035
    },
    {
      "epoch": 0.5454196274246207,
      "grad_norm": 28.625,
      "learning_rate": 3.076164632360666e-05,
      "loss": 4.4499,
      "num_input_tokens_seen": 23083600,
      "step": 17040
    },
    {
      "epoch": 0.5455796683951092,
      "grad_norm": 40.25,
      "learning_rate": 3.075209268497563e-05,
      "loss": 4.7146,
      "num_input_tokens_seen": 23090592,
      "step": 17045
    },
    {
      "epoch": 0.5457397093655976,
      "grad_norm": 29.375,
      "learning_rate": 3.074253815929961e-05,
      "loss": 3.7067,
      "num_input_tokens_seen": 23097408,
      "step": 17050
    },
    {
      "epoch": 0.5458997503360861,
      "grad_norm": 35.5,
      "learning_rate": 3.0732982748052054e-05,
      "loss": 4.765,
      "num_input_tokens_seen": 23104176,
      "step": 17055
    },
    {
      "epoch": 0.5460597913065744,
      "grad_norm": 31.0,
      "learning_rate": 3.072342645270651e-05,
      "loss": 5.6158,
      "num_input_tokens_seen": 23110880,
      "step": 17060
    },
    {
      "epoch": 0.5462198322770629,
      "grad_norm": 33.0,
      "learning_rate": 3.071386927473668e-05,
      "loss": 4.0811,
      "num_input_tokens_seen": 23117728,
      "step": 17065
    },
    {
      "epoch": 0.5463798732475513,
      "grad_norm": 26.875,
      "learning_rate": 3.0704311215616404e-05,
      "loss": 4.4385,
      "num_input_tokens_seen": 23124240,
      "step": 17070
    },
    {
      "epoch": 0.5465399142180398,
      "grad_norm": 24.625,
      "learning_rate": 3.0694752276819656e-05,
      "loss": 4.0845,
      "num_input_tokens_seen": 23130560,
      "step": 17075
    },
    {
      "epoch": 0.5466999551885282,
      "grad_norm": 44.5,
      "learning_rate": 3.068519245982054e-05,
      "loss": 4.8855,
      "num_input_tokens_seen": 23137104,
      "step": 17080
    },
    {
      "epoch": 0.5468599961590167,
      "grad_norm": 35.75,
      "learning_rate": 3.0675631766093304e-05,
      "loss": 4.5031,
      "num_input_tokens_seen": 23144144,
      "step": 17085
    },
    {
      "epoch": 0.5470200371295052,
      "grad_norm": 16.25,
      "learning_rate": 3.066607019711232e-05,
      "loss": 4.194,
      "num_input_tokens_seen": 23150704,
      "step": 17090
    },
    {
      "epoch": 0.5471800780999936,
      "grad_norm": 28.625,
      "learning_rate": 3.065650775435211e-05,
      "loss": 4.3641,
      "num_input_tokens_seen": 23157312,
      "step": 17095
    },
    {
      "epoch": 0.5473401190704821,
      "grad_norm": 32.75,
      "learning_rate": 3.0646944439287326e-05,
      "loss": 5.8172,
      "num_input_tokens_seen": 23163920,
      "step": 17100
    },
    {
      "epoch": 0.5475001600409705,
      "grad_norm": 43.0,
      "learning_rate": 3.0637380253392736e-05,
      "loss": 4.8583,
      "num_input_tokens_seen": 23170480,
      "step": 17105
    },
    {
      "epoch": 0.547660201011459,
      "grad_norm": 18.0,
      "learning_rate": 3.062781519814327e-05,
      "loss": 3.6427,
      "num_input_tokens_seen": 23176720,
      "step": 17110
    },
    {
      "epoch": 0.5478202419819473,
      "grad_norm": 37.25,
      "learning_rate": 3.0618249275013985e-05,
      "loss": 3.7506,
      "num_input_tokens_seen": 23183280,
      "step": 17115
    },
    {
      "epoch": 0.5479802829524358,
      "grad_norm": 22.0,
      "learning_rate": 3.060868248548005e-05,
      "loss": 4.8536,
      "num_input_tokens_seen": 23190128,
      "step": 17120
    },
    {
      "epoch": 0.5481403239229242,
      "grad_norm": 44.25,
      "learning_rate": 3.0599114831016796e-05,
      "loss": 4.474,
      "num_input_tokens_seen": 23196944,
      "step": 17125
    },
    {
      "epoch": 0.5483003648934127,
      "grad_norm": 38.75,
      "learning_rate": 3.0589546313099666e-05,
      "loss": 5.0298,
      "num_input_tokens_seen": 23203936,
      "step": 17130
    },
    {
      "epoch": 0.5484604058639012,
      "grad_norm": 18.625,
      "learning_rate": 3.0579976933204255e-05,
      "loss": 4.2342,
      "num_input_tokens_seen": 23210400,
      "step": 17135
    },
    {
      "epoch": 0.5486204468343896,
      "grad_norm": 14.75,
      "learning_rate": 3.0570406692806284e-05,
      "loss": 4.408,
      "num_input_tokens_seen": 23217568,
      "step": 17140
    },
    {
      "epoch": 0.5487804878048781,
      "grad_norm": 30.75,
      "learning_rate": 3.05608355933816e-05,
      "loss": 4.663,
      "num_input_tokens_seen": 23224592,
      "step": 17145
    },
    {
      "epoch": 0.5489405287753665,
      "grad_norm": 32.5,
      "learning_rate": 3.055126363640618e-05,
      "loss": 4.2929,
      "num_input_tokens_seen": 23231264,
      "step": 17150
    },
    {
      "epoch": 0.549100569745855,
      "grad_norm": 34.25,
      "learning_rate": 3.0541690823356146e-05,
      "loss": 4.0822,
      "num_input_tokens_seen": 23238320,
      "step": 17155
    },
    {
      "epoch": 0.5492606107163434,
      "grad_norm": 31.5,
      "learning_rate": 3.053211715570775e-05,
      "loss": 3.0802,
      "num_input_tokens_seen": 23245472,
      "step": 17160
    },
    {
      "epoch": 0.5494206516868319,
      "grad_norm": 33.5,
      "learning_rate": 3.052254263493736e-05,
      "loss": 4.242,
      "num_input_tokens_seen": 23252736,
      "step": 17165
    },
    {
      "epoch": 0.5495806926573202,
      "grad_norm": 39.75,
      "learning_rate": 3.0512967262521498e-05,
      "loss": 3.6534,
      "num_input_tokens_seen": 23260016,
      "step": 17170
    },
    {
      "epoch": 0.5497407336278087,
      "grad_norm": 25.125,
      "learning_rate": 3.0503391039936803e-05,
      "loss": 4.7271,
      "num_input_tokens_seen": 23266592,
      "step": 17175
    },
    {
      "epoch": 0.5499007745982971,
      "grad_norm": 31.0,
      "learning_rate": 3.0493813968660056e-05,
      "loss": 3.5825,
      "num_input_tokens_seen": 23273504,
      "step": 17180
    },
    {
      "epoch": 0.5500608155687856,
      "grad_norm": 33.5,
      "learning_rate": 3.0484236050168153e-05,
      "loss": 5.2696,
      "num_input_tokens_seen": 23280480,
      "step": 17185
    },
    {
      "epoch": 0.5502208565392741,
      "grad_norm": 27.125,
      "learning_rate": 3.0474657285938123e-05,
      "loss": 4.0492,
      "num_input_tokens_seen": 23287328,
      "step": 17190
    },
    {
      "epoch": 0.5503808975097625,
      "grad_norm": 28.375,
      "learning_rate": 3.046507767744715e-05,
      "loss": 5.205,
      "num_input_tokens_seen": 23293888,
      "step": 17195
    },
    {
      "epoch": 0.550540938480251,
      "grad_norm": 20.0,
      "learning_rate": 3.045549722617252e-05,
      "loss": 4.2426,
      "num_input_tokens_seen": 23300528,
      "step": 17200
    },
    {
      "epoch": 0.550540938480251,
      "eval_loss": 4.498417377471924,
      "eval_runtime": 489.7404,
      "eval_samples_per_second": 28.354,
      "eval_steps_per_second": 14.177,
      "num_input_tokens_seen": 23300528,
      "step": 17200
    },
    {
      "epoch": 0.5507009794507394,
      "grad_norm": 25.25,
      "learning_rate": 3.0445915933591658e-05,
      "loss": 4.0044,
      "num_input_tokens_seen": 23307360,
      "step": 17205
    },
    {
      "epoch": 0.5508610204212279,
      "grad_norm": 33.75,
      "learning_rate": 3.0436333801182114e-05,
      "loss": 4.8427,
      "num_input_tokens_seen": 23314272,
      "step": 17210
    },
    {
      "epoch": 0.5510210613917162,
      "grad_norm": 49.0,
      "learning_rate": 3.0426750830421596e-05,
      "loss": 5.2844,
      "num_input_tokens_seen": 23320848,
      "step": 17215
    },
    {
      "epoch": 0.5511811023622047,
      "grad_norm": 42.0,
      "learning_rate": 3.0417167022787897e-05,
      "loss": 4.4725,
      "num_input_tokens_seen": 23327744,
      "step": 17220
    },
    {
      "epoch": 0.5513411433326931,
      "grad_norm": 32.25,
      "learning_rate": 3.0407582379758966e-05,
      "loss": 5.0125,
      "num_input_tokens_seen": 23334320,
      "step": 17225
    },
    {
      "epoch": 0.5515011843031816,
      "grad_norm": 30.375,
      "learning_rate": 3.039799690281287e-05,
      "loss": 5.7251,
      "num_input_tokens_seen": 23340720,
      "step": 17230
    },
    {
      "epoch": 0.55166122527367,
      "grad_norm": 39.0,
      "learning_rate": 3.0388410593427823e-05,
      "loss": 5.7473,
      "num_input_tokens_seen": 23347360,
      "step": 17235
    },
    {
      "epoch": 0.5518212662441585,
      "grad_norm": 39.5,
      "learning_rate": 3.0378823453082146e-05,
      "loss": 4.3653,
      "num_input_tokens_seen": 23353952,
      "step": 17240
    },
    {
      "epoch": 0.551981307214647,
      "grad_norm": 32.75,
      "learning_rate": 3.03692354832543e-05,
      "loss": 3.238,
      "num_input_tokens_seen": 23360480,
      "step": 17245
    },
    {
      "epoch": 0.5521413481851354,
      "grad_norm": 28.125,
      "learning_rate": 3.0359646685422865e-05,
      "loss": 4.6648,
      "num_input_tokens_seen": 23367120,
      "step": 17250
    },
    {
      "epoch": 0.5523013891556239,
      "grad_norm": 22.25,
      "learning_rate": 3.035005706106656e-05,
      "loss": 4.2088,
      "num_input_tokens_seen": 23373856,
      "step": 17255
    },
    {
      "epoch": 0.5524614301261123,
      "grad_norm": 17.875,
      "learning_rate": 3.034046661166422e-05,
      "loss": 4.9552,
      "num_input_tokens_seen": 23380224,
      "step": 17260
    },
    {
      "epoch": 0.5526214710966008,
      "grad_norm": 25.0,
      "learning_rate": 3.033087533869482e-05,
      "loss": 3.968,
      "num_input_tokens_seen": 23386848,
      "step": 17265
    },
    {
      "epoch": 0.5527815120670891,
      "grad_norm": 21.375,
      "learning_rate": 3.0321283243637444e-05,
      "loss": 4.1038,
      "num_input_tokens_seen": 23393680,
      "step": 17270
    },
    {
      "epoch": 0.5529415530375776,
      "grad_norm": 25.375,
      "learning_rate": 3.0311690327971326e-05,
      "loss": 4.9157,
      "num_input_tokens_seen": 23400448,
      "step": 17275
    },
    {
      "epoch": 0.553101594008066,
      "grad_norm": 35.5,
      "learning_rate": 3.030209659317581e-05,
      "loss": 4.7206,
      "num_input_tokens_seen": 23407248,
      "step": 17280
    },
    {
      "epoch": 0.5532616349785545,
      "grad_norm": 25.375,
      "learning_rate": 3.0292502040730362e-05,
      "loss": 4.1529,
      "num_input_tokens_seen": 23413824,
      "step": 17285
    },
    {
      "epoch": 0.5534216759490429,
      "grad_norm": 39.5,
      "learning_rate": 3.0282906672114597e-05,
      "loss": 3.9653,
      "num_input_tokens_seen": 23420624,
      "step": 17290
    },
    {
      "epoch": 0.5535817169195314,
      "grad_norm": 19.5,
      "learning_rate": 3.027331048880823e-05,
      "loss": 4.5759,
      "num_input_tokens_seen": 23427440,
      "step": 17295
    },
    {
      "epoch": 0.5537417578900199,
      "grad_norm": 23.625,
      "learning_rate": 3.0263713492291123e-05,
      "loss": 3.2744,
      "num_input_tokens_seen": 23434672,
      "step": 17300
    },
    {
      "epoch": 0.5539017988605083,
      "grad_norm": 58.0,
      "learning_rate": 3.0254115684043242e-05,
      "loss": 5.0272,
      "num_input_tokens_seen": 23441472,
      "step": 17305
    },
    {
      "epoch": 0.5540618398309968,
      "grad_norm": 21.5,
      "learning_rate": 3.024451706554469e-05,
      "loss": 4.4765,
      "num_input_tokens_seen": 23448352,
      "step": 17310
    },
    {
      "epoch": 0.5542218808014852,
      "grad_norm": 34.0,
      "learning_rate": 3.0234917638275705e-05,
      "loss": 4.7852,
      "num_input_tokens_seen": 23454992,
      "step": 17315
    },
    {
      "epoch": 0.5543819217719737,
      "grad_norm": 19.125,
      "learning_rate": 3.0225317403716635e-05,
      "loss": 4.6618,
      "num_input_tokens_seen": 23461632,
      "step": 17320
    },
    {
      "epoch": 0.554541962742462,
      "grad_norm": 16.125,
      "learning_rate": 3.0215716363347956e-05,
      "loss": 4.1134,
      "num_input_tokens_seen": 23468272,
      "step": 17325
    },
    {
      "epoch": 0.5547020037129505,
      "grad_norm": 24.5,
      "learning_rate": 3.0206114518650275e-05,
      "loss": 3.9834,
      "num_input_tokens_seen": 23474832,
      "step": 17330
    },
    {
      "epoch": 0.5548620446834389,
      "grad_norm": 36.0,
      "learning_rate": 3.0196511871104304e-05,
      "loss": 5.2037,
      "num_input_tokens_seen": 23481648,
      "step": 17335
    },
    {
      "epoch": 0.5550220856539274,
      "grad_norm": 39.25,
      "learning_rate": 3.01869084221909e-05,
      "loss": 4.574,
      "num_input_tokens_seen": 23488160,
      "step": 17340
    },
    {
      "epoch": 0.5551821266244159,
      "grad_norm": 21.625,
      "learning_rate": 3.0177304173391037e-05,
      "loss": 3.9678,
      "num_input_tokens_seen": 23495280,
      "step": 17345
    },
    {
      "epoch": 0.5553421675949043,
      "grad_norm": 19.0,
      "learning_rate": 3.01676991261858e-05,
      "loss": 4.7617,
      "num_input_tokens_seen": 23501776,
      "step": 17350
    },
    {
      "epoch": 0.5555022085653928,
      "grad_norm": 26.125,
      "learning_rate": 3.015809328205642e-05,
      "loss": 3.94,
      "num_input_tokens_seen": 23508400,
      "step": 17355
    },
    {
      "epoch": 0.5556622495358812,
      "grad_norm": 33.25,
      "learning_rate": 3.0148486642484248e-05,
      "loss": 4.0245,
      "num_input_tokens_seen": 23515312,
      "step": 17360
    },
    {
      "epoch": 0.5558222905063697,
      "grad_norm": 23.0,
      "learning_rate": 3.0138879208950722e-05,
      "loss": 3.5694,
      "num_input_tokens_seen": 23521712,
      "step": 17365
    },
    {
      "epoch": 0.555982331476858,
      "grad_norm": 29.25,
      "learning_rate": 3.012927098293744e-05,
      "loss": 4.9814,
      "num_input_tokens_seen": 23528592,
      "step": 17370
    },
    {
      "epoch": 0.5561423724473465,
      "grad_norm": 31.25,
      "learning_rate": 3.0119661965926123e-05,
      "loss": 5.0729,
      "num_input_tokens_seen": 23535744,
      "step": 17375
    },
    {
      "epoch": 0.5563024134178349,
      "grad_norm": 28.0,
      "learning_rate": 3.0110052159398587e-05,
      "loss": 4.7976,
      "num_input_tokens_seen": 23542400,
      "step": 17380
    },
    {
      "epoch": 0.5564624543883234,
      "grad_norm": 26.625,
      "learning_rate": 3.0100441564836802e-05,
      "loss": 3.9302,
      "num_input_tokens_seen": 23548912,
      "step": 17385
    },
    {
      "epoch": 0.5566224953588118,
      "grad_norm": 16.625,
      "learning_rate": 3.0090830183722817e-05,
      "loss": 4.071,
      "num_input_tokens_seen": 23555888,
      "step": 17390
    },
    {
      "epoch": 0.5567825363293003,
      "grad_norm": 23.125,
      "learning_rate": 3.0081218017538852e-05,
      "loss": 3.1822,
      "num_input_tokens_seen": 23562736,
      "step": 17395
    },
    {
      "epoch": 0.5569425772997888,
      "grad_norm": 30.25,
      "learning_rate": 3.0071605067767212e-05,
      "loss": 4.291,
      "num_input_tokens_seen": 23569728,
      "step": 17400
    },
    {
      "epoch": 0.5569425772997888,
      "eval_loss": 4.473456859588623,
      "eval_runtime": 490.3686,
      "eval_samples_per_second": 28.317,
      "eval_steps_per_second": 14.159,
      "num_input_tokens_seen": 23569728,
      "step": 17400
    },
    {
      "epoch": 0.5571026182702772,
      "grad_norm": 16.625,
      "learning_rate": 3.006199133589034e-05,
      "loss": 3.4311,
      "num_input_tokens_seen": 23576656,
      "step": 17405
    },
    {
      "epoch": 0.5572626592407657,
      "grad_norm": 25.0,
      "learning_rate": 3.005237682339079e-05,
      "loss": 4.7978,
      "num_input_tokens_seen": 23583280,
      "step": 17410
    },
    {
      "epoch": 0.5574227002112541,
      "grad_norm": 32.25,
      "learning_rate": 3.0042761531751228e-05,
      "loss": 4.2701,
      "num_input_tokens_seen": 23589952,
      "step": 17415
    },
    {
      "epoch": 0.5575827411817426,
      "grad_norm": 21.25,
      "learning_rate": 3.0033145462454482e-05,
      "loss": 4.5736,
      "num_input_tokens_seen": 23596608,
      "step": 17420
    },
    {
      "epoch": 0.5577427821522309,
      "grad_norm": 33.25,
      "learning_rate": 3.002352861698345e-05,
      "loss": 5.1429,
      "num_input_tokens_seen": 23603120,
      "step": 17425
    },
    {
      "epoch": 0.5579028231227194,
      "grad_norm": 18.5,
      "learning_rate": 3.0013910996821178e-05,
      "loss": 3.4266,
      "num_input_tokens_seen": 23609712,
      "step": 17430
    },
    {
      "epoch": 0.5580628640932078,
      "grad_norm": 33.75,
      "learning_rate": 3.0004292603450817e-05,
      "loss": 3.3803,
      "num_input_tokens_seen": 23616352,
      "step": 17435
    },
    {
      "epoch": 0.5582229050636963,
      "grad_norm": 36.25,
      "learning_rate": 2.9994673438355653e-05,
      "loss": 5.0169,
      "num_input_tokens_seen": 23623200,
      "step": 17440
    },
    {
      "epoch": 0.5583829460341847,
      "grad_norm": 44.0,
      "learning_rate": 2.9985053503019078e-05,
      "loss": 5.0234,
      "num_input_tokens_seen": 23629568,
      "step": 17445
    },
    {
      "epoch": 0.5585429870046732,
      "grad_norm": 21.0,
      "learning_rate": 2.99754327989246e-05,
      "loss": 3.8647,
      "num_input_tokens_seen": 23636400,
      "step": 17450
    },
    {
      "epoch": 0.5587030279751617,
      "grad_norm": 42.5,
      "learning_rate": 2.9965811327555864e-05,
      "loss": 4.2667,
      "num_input_tokens_seen": 23642928,
      "step": 17455
    },
    {
      "epoch": 0.5588630689456501,
      "grad_norm": 41.0,
      "learning_rate": 2.995618909039662e-05,
      "loss": 4.4443,
      "num_input_tokens_seen": 23649744,
      "step": 17460
    },
    {
      "epoch": 0.5590231099161386,
      "grad_norm": 24.5,
      "learning_rate": 2.9946566088930727e-05,
      "loss": 4.1369,
      "num_input_tokens_seen": 23656416,
      "step": 17465
    },
    {
      "epoch": 0.559183150886627,
      "grad_norm": 27.375,
      "learning_rate": 2.9936942324642192e-05,
      "loss": 3.976,
      "num_input_tokens_seen": 23662736,
      "step": 17470
    },
    {
      "epoch": 0.5593431918571155,
      "grad_norm": 27.25,
      "learning_rate": 2.9927317799015097e-05,
      "loss": 4.4859,
      "num_input_tokens_seen": 23669600,
      "step": 17475
    },
    {
      "epoch": 0.5595032328276038,
      "grad_norm": 44.25,
      "learning_rate": 2.9917692513533685e-05,
      "loss": 4.6686,
      "num_input_tokens_seen": 23676400,
      "step": 17480
    },
    {
      "epoch": 0.5596632737980923,
      "grad_norm": 45.75,
      "learning_rate": 2.990806646968229e-05,
      "loss": 5.1429,
      "num_input_tokens_seen": 23683168,
      "step": 17485
    },
    {
      "epoch": 0.5598233147685807,
      "grad_norm": 24.75,
      "learning_rate": 2.989843966894536e-05,
      "loss": 3.964,
      "num_input_tokens_seen": 23689984,
      "step": 17490
    },
    {
      "epoch": 0.5599833557390692,
      "grad_norm": 26.5,
      "learning_rate": 2.9888812112807472e-05,
      "loss": 3.8681,
      "num_input_tokens_seen": 23697200,
      "step": 17495
    },
    {
      "epoch": 0.5601433967095576,
      "grad_norm": 48.0,
      "learning_rate": 2.987918380275333e-05,
      "loss": 4.1815,
      "num_input_tokens_seen": 23703952,
      "step": 17500
    },
    {
      "epoch": 0.5603034376800461,
      "grad_norm": 33.75,
      "learning_rate": 2.9869554740267724e-05,
      "loss": 4.8174,
      "num_input_tokens_seen": 23711136,
      "step": 17505
    },
    {
      "epoch": 0.5604634786505346,
      "grad_norm": 31.625,
      "learning_rate": 2.9859924926835585e-05,
      "loss": 3.3985,
      "num_input_tokens_seen": 23718304,
      "step": 17510
    },
    {
      "epoch": 0.560623519621023,
      "grad_norm": 39.25,
      "learning_rate": 2.9850294363941944e-05,
      "loss": 3.7835,
      "num_input_tokens_seen": 23725216,
      "step": 17515
    },
    {
      "epoch": 0.5607835605915115,
      "grad_norm": 18.375,
      "learning_rate": 2.9840663053071967e-05,
      "loss": 3.8581,
      "num_input_tokens_seen": 23731696,
      "step": 17520
    },
    {
      "epoch": 0.5609436015619998,
      "grad_norm": 39.25,
      "learning_rate": 2.983103099571091e-05,
      "loss": 4.0738,
      "num_input_tokens_seen": 23738544,
      "step": 17525
    },
    {
      "epoch": 0.5611036425324883,
      "grad_norm": 36.5,
      "learning_rate": 2.9821398193344164e-05,
      "loss": 4.0837,
      "num_input_tokens_seen": 23745840,
      "step": 17530
    },
    {
      "epoch": 0.5612636835029767,
      "grad_norm": 18.0,
      "learning_rate": 2.9811764647457226e-05,
      "loss": 4.649,
      "num_input_tokens_seen": 23752400,
      "step": 17535
    },
    {
      "epoch": 0.5614237244734652,
      "grad_norm": 43.75,
      "learning_rate": 2.9802130359535714e-05,
      "loss": 5.4228,
      "num_input_tokens_seen": 23759040,
      "step": 17540
    },
    {
      "epoch": 0.5615837654439536,
      "grad_norm": 15.5,
      "learning_rate": 2.979249533106535e-05,
      "loss": 3.5914,
      "num_input_tokens_seen": 23765872,
      "step": 17545
    },
    {
      "epoch": 0.5617438064144421,
      "grad_norm": 23.0,
      "learning_rate": 2.9782859563531986e-05,
      "loss": 4.2148,
      "num_input_tokens_seen": 23772336,
      "step": 17550
    },
    {
      "epoch": 0.5619038473849305,
      "grad_norm": 29.375,
      "learning_rate": 2.977322305842156e-05,
      "loss": 3.9354,
      "num_input_tokens_seen": 23778864,
      "step": 17555
    },
    {
      "epoch": 0.562063888355419,
      "grad_norm": 32.75,
      "learning_rate": 2.9763585817220162e-05,
      "loss": 4.9584,
      "num_input_tokens_seen": 23785136,
      "step": 17560
    },
    {
      "epoch": 0.5622239293259075,
      "grad_norm": 39.25,
      "learning_rate": 2.975394784141397e-05,
      "loss": 4.0998,
      "num_input_tokens_seen": 23791664,
      "step": 17565
    },
    {
      "epoch": 0.5623839702963959,
      "grad_norm": 22.125,
      "learning_rate": 2.974430913248928e-05,
      "loss": 5.0354,
      "num_input_tokens_seen": 23797904,
      "step": 17570
    },
    {
      "epoch": 0.5625440112668844,
      "grad_norm": 27.875,
      "learning_rate": 2.9734669691932497e-05,
      "loss": 4.2184,
      "num_input_tokens_seen": 23804560,
      "step": 17575
    },
    {
      "epoch": 0.5627040522373727,
      "grad_norm": 23.75,
      "learning_rate": 2.9725029521230147e-05,
      "loss": 3.8962,
      "num_input_tokens_seen": 23811456,
      "step": 17580
    },
    {
      "epoch": 0.5628640932078612,
      "grad_norm": 20.75,
      "learning_rate": 2.9715388621868873e-05,
      "loss": 4.6774,
      "num_input_tokens_seen": 23818256,
      "step": 17585
    },
    {
      "epoch": 0.5630241341783496,
      "grad_norm": 35.5,
      "learning_rate": 2.970574699533541e-05,
      "loss": 4.47,
      "num_input_tokens_seen": 23825120,
      "step": 17590
    },
    {
      "epoch": 0.5631841751488381,
      "grad_norm": 34.75,
      "learning_rate": 2.969610464311662e-05,
      "loss": 3.6332,
      "num_input_tokens_seen": 23831984,
      "step": 17595
    },
    {
      "epoch": 0.5633442161193265,
      "grad_norm": 50.75,
      "learning_rate": 2.9686461566699487e-05,
      "loss": 4.815,
      "num_input_tokens_seen": 23838464,
      "step": 17600
    },
    {
      "epoch": 0.5633442161193265,
      "eval_loss": 4.505002975463867,
      "eval_runtime": 489.8355,
      "eval_samples_per_second": 28.348,
      "eval_steps_per_second": 14.174,
      "num_input_tokens_seen": 23838464,
      "step": 17600
    },
    {
      "epoch": 0.563504257089815,
      "grad_norm": 24.0,
      "learning_rate": 2.9676817767571086e-05,
      "loss": 4.6713,
      "num_input_tokens_seen": 23845088,
      "step": 17605
    },
    {
      "epoch": 0.5636642980603035,
      "grad_norm": 29.375,
      "learning_rate": 2.966717324721861e-05,
      "loss": 4.5056,
      "num_input_tokens_seen": 23851952,
      "step": 17610
    },
    {
      "epoch": 0.5638243390307919,
      "grad_norm": 38.5,
      "learning_rate": 2.9657528007129366e-05,
      "loss": 4.3296,
      "num_input_tokens_seen": 23858576,
      "step": 17615
    },
    {
      "epoch": 0.5639843800012804,
      "grad_norm": 29.25,
      "learning_rate": 2.9647882048790777e-05,
      "loss": 3.1909,
      "num_input_tokens_seen": 23865664,
      "step": 17620
    },
    {
      "epoch": 0.5641444209717688,
      "grad_norm": 19.0,
      "learning_rate": 2.963823537369037e-05,
      "loss": 3.4574,
      "num_input_tokens_seen": 23872240,
      "step": 17625
    },
    {
      "epoch": 0.5643044619422573,
      "grad_norm": 19.5,
      "learning_rate": 2.9628587983315775e-05,
      "loss": 4.2919,
      "num_input_tokens_seen": 23878928,
      "step": 17630
    },
    {
      "epoch": 0.5644645029127456,
      "grad_norm": 24.125,
      "learning_rate": 2.9618939879154746e-05,
      "loss": 4.3935,
      "num_input_tokens_seen": 23885504,
      "step": 17635
    },
    {
      "epoch": 0.5646245438832341,
      "grad_norm": 44.0,
      "learning_rate": 2.9609291062695143e-05,
      "loss": 4.7955,
      "num_input_tokens_seen": 23892240,
      "step": 17640
    },
    {
      "epoch": 0.5647845848537225,
      "grad_norm": 40.75,
      "learning_rate": 2.9599641535424938e-05,
      "loss": 4.0465,
      "num_input_tokens_seen": 23898912,
      "step": 17645
    },
    {
      "epoch": 0.564944625824211,
      "grad_norm": 22.875,
      "learning_rate": 2.9589991298832202e-05,
      "loss": 4.8511,
      "num_input_tokens_seen": 23905872,
      "step": 17650
    },
    {
      "epoch": 0.5651046667946994,
      "grad_norm": 21.125,
      "learning_rate": 2.958034035440513e-05,
      "loss": 4.3474,
      "num_input_tokens_seen": 23912816,
      "step": 17655
    },
    {
      "epoch": 0.5652647077651879,
      "grad_norm": 34.25,
      "learning_rate": 2.957068870363201e-05,
      "loss": 4.837,
      "num_input_tokens_seen": 23919264,
      "step": 17660
    },
    {
      "epoch": 0.5654247487356764,
      "grad_norm": 20.875,
      "learning_rate": 2.956103634800126e-05,
      "loss": 4.3524,
      "num_input_tokens_seen": 23925936,
      "step": 17665
    },
    {
      "epoch": 0.5655847897061648,
      "grad_norm": 35.5,
      "learning_rate": 2.9551383289001384e-05,
      "loss": 5.0356,
      "num_input_tokens_seen": 23932688,
      "step": 17670
    },
    {
      "epoch": 0.5657448306766533,
      "grad_norm": 49.75,
      "learning_rate": 2.9541729528121005e-05,
      "loss": 4.3867,
      "num_input_tokens_seen": 23940032,
      "step": 17675
    },
    {
      "epoch": 0.5659048716471416,
      "grad_norm": 33.25,
      "learning_rate": 2.9532075066848856e-05,
      "loss": 5.0845,
      "num_input_tokens_seen": 23946768,
      "step": 17680
    },
    {
      "epoch": 0.5660649126176301,
      "grad_norm": 16.5,
      "learning_rate": 2.9522419906673786e-05,
      "loss": 4.7979,
      "num_input_tokens_seen": 23953728,
      "step": 17685
    },
    {
      "epoch": 0.5662249535881185,
      "grad_norm": 18.125,
      "learning_rate": 2.951276404908474e-05,
      "loss": 3.9513,
      "num_input_tokens_seen": 23960928,
      "step": 17690
    },
    {
      "epoch": 0.566384994558607,
      "grad_norm": 23.25,
      "learning_rate": 2.9503107495570752e-05,
      "loss": 3.652,
      "num_input_tokens_seen": 23967440,
      "step": 17695
    },
    {
      "epoch": 0.5665450355290954,
      "grad_norm": 29.5,
      "learning_rate": 2.9493450247621003e-05,
      "loss": 4.0876,
      "num_input_tokens_seen": 23974160,
      "step": 17700
    },
    {
      "epoch": 0.5667050764995839,
      "grad_norm": 41.0,
      "learning_rate": 2.948379230672476e-05,
      "loss": 4.0627,
      "num_input_tokens_seen": 23981248,
      "step": 17705
    },
    {
      "epoch": 0.5668651174700723,
      "grad_norm": 35.25,
      "learning_rate": 2.9474133674371396e-05,
      "loss": 4.2979,
      "num_input_tokens_seen": 23988064,
      "step": 17710
    },
    {
      "epoch": 0.5670251584405608,
      "grad_norm": 40.75,
      "learning_rate": 2.9464474352050387e-05,
      "loss": 3.8763,
      "num_input_tokens_seen": 23994896,
      "step": 17715
    },
    {
      "epoch": 0.5671851994110493,
      "grad_norm": 23.375,
      "learning_rate": 2.9454814341251336e-05,
      "loss": 4.4895,
      "num_input_tokens_seen": 24001584,
      "step": 17720
    },
    {
      "epoch": 0.5673452403815377,
      "grad_norm": 33.75,
      "learning_rate": 2.9445153643463942e-05,
      "loss": 5.3402,
      "num_input_tokens_seen": 24008544,
      "step": 17725
    },
    {
      "epoch": 0.5675052813520262,
      "grad_norm": 22.25,
      "learning_rate": 2.943549226017798e-05,
      "loss": 4.2495,
      "num_input_tokens_seen": 24015232,
      "step": 17730
    },
    {
      "epoch": 0.5676653223225145,
      "grad_norm": 35.0,
      "learning_rate": 2.942583019288337e-05,
      "loss": 4.5723,
      "num_input_tokens_seen": 24022048,
      "step": 17735
    },
    {
      "epoch": 0.567825363293003,
      "grad_norm": 30.625,
      "learning_rate": 2.9416167443070132e-05,
      "loss": 4.1993,
      "num_input_tokens_seen": 24028512,
      "step": 17740
    },
    {
      "epoch": 0.5679854042634914,
      "grad_norm": 27.375,
      "learning_rate": 2.9406504012228375e-05,
      "loss": 4.5038,
      "num_input_tokens_seen": 24035168,
      "step": 17745
    },
    {
      "epoch": 0.5681454452339799,
      "grad_norm": 45.5,
      "learning_rate": 2.939683990184832e-05,
      "loss": 5.0298,
      "num_input_tokens_seen": 24041664,
      "step": 17750
    },
    {
      "epoch": 0.5683054862044683,
      "grad_norm": 25.0,
      "learning_rate": 2.93871751134203e-05,
      "loss": 3.6194,
      "num_input_tokens_seen": 24048528,
      "step": 17755
    },
    {
      "epoch": 0.5684655271749568,
      "grad_norm": 59.25,
      "learning_rate": 2.9377509648434752e-05,
      "loss": 4.7844,
      "num_input_tokens_seen": 24054864,
      "step": 17760
    },
    {
      "epoch": 0.5686255681454452,
      "grad_norm": 37.5,
      "learning_rate": 2.9367843508382203e-05,
      "loss": 3.7066,
      "num_input_tokens_seen": 24061536,
      "step": 17765
    },
    {
      "epoch": 0.5687856091159337,
      "grad_norm": 32.0,
      "learning_rate": 2.9358176694753293e-05,
      "loss": 4.109,
      "num_input_tokens_seen": 24068848,
      "step": 17770
    },
    {
      "epoch": 0.5689456500864222,
      "grad_norm": 28.25,
      "learning_rate": 2.9348509209038766e-05,
      "loss": 4.5913,
      "num_input_tokens_seen": 24075680,
      "step": 17775
    },
    {
      "epoch": 0.5691056910569106,
      "grad_norm": 22.875,
      "learning_rate": 2.933884105272947e-05,
      "loss": 4.5379,
      "num_input_tokens_seen": 24082912,
      "step": 17780
    },
    {
      "epoch": 0.569265732027399,
      "grad_norm": 14.5625,
      "learning_rate": 2.9329172227316366e-05,
      "loss": 3.8792,
      "num_input_tokens_seen": 24089648,
      "step": 17785
    },
    {
      "epoch": 0.5694257729978874,
      "grad_norm": 25.625,
      "learning_rate": 2.93195027342905e-05,
      "loss": 4.0687,
      "num_input_tokens_seen": 24096160,
      "step": 17790
    },
    {
      "epoch": 0.5695858139683759,
      "grad_norm": 21.125,
      "learning_rate": 2.9309832575143024e-05,
      "loss": 4.1441,
      "num_input_tokens_seen": 24102832,
      "step": 17795
    },
    {
      "epoch": 0.5697458549388643,
      "grad_norm": 47.25,
      "learning_rate": 2.930016175136521e-05,
      "loss": 4.839,
      "num_input_tokens_seen": 24109808,
      "step": 17800
    },
    {
      "epoch": 0.5697458549388643,
      "eval_loss": 4.498166084289551,
      "eval_runtime": 490.1686,
      "eval_samples_per_second": 28.329,
      "eval_steps_per_second": 14.165,
      "num_input_tokens_seen": 24109808,
      "step": 17800
    },
    {
      "epoch": 0.5699058959093528,
      "grad_norm": 33.0,
      "learning_rate": 2.9290490264448412e-05,
      "loss": 4.4893,
      "num_input_tokens_seen": 24116448,
      "step": 17805
    },
    {
      "epoch": 0.5700659368798412,
      "grad_norm": 25.125,
      "learning_rate": 2.9280818115884094e-05,
      "loss": 4.4605,
      "num_input_tokens_seen": 24123440,
      "step": 17810
    },
    {
      "epoch": 0.5702259778503297,
      "grad_norm": 64.5,
      "learning_rate": 2.9271145307163828e-05,
      "loss": 3.8761,
      "num_input_tokens_seen": 24130112,
      "step": 17815
    },
    {
      "epoch": 0.5703860188208181,
      "grad_norm": 22.375,
      "learning_rate": 2.9261471839779287e-05,
      "loss": 3.7916,
      "num_input_tokens_seen": 24136944,
      "step": 17820
    },
    {
      "epoch": 0.5705460597913066,
      "grad_norm": 37.5,
      "learning_rate": 2.925179771522223e-05,
      "loss": 3.9432,
      "num_input_tokens_seen": 24143072,
      "step": 17825
    },
    {
      "epoch": 0.5707061007617951,
      "grad_norm": 31.125,
      "learning_rate": 2.9242122934984535e-05,
      "loss": 4.1157,
      "num_input_tokens_seen": 24149664,
      "step": 17830
    },
    {
      "epoch": 0.5708661417322834,
      "grad_norm": 27.375,
      "learning_rate": 2.9232447500558176e-05,
      "loss": 4.1411,
      "num_input_tokens_seen": 24156896,
      "step": 17835
    },
    {
      "epoch": 0.571026182702772,
      "grad_norm": 52.25,
      "learning_rate": 2.9222771413435225e-05,
      "loss": 4.7827,
      "num_input_tokens_seen": 24163648,
      "step": 17840
    },
    {
      "epoch": 0.5711862236732603,
      "grad_norm": 22.75,
      "learning_rate": 2.9213094675107848e-05,
      "loss": 5.1111,
      "num_input_tokens_seen": 24170464,
      "step": 17845
    },
    {
      "epoch": 0.5713462646437488,
      "grad_norm": 45.5,
      "learning_rate": 2.9203417287068335e-05,
      "loss": 5.0798,
      "num_input_tokens_seen": 24177136,
      "step": 17850
    },
    {
      "epoch": 0.5715063056142372,
      "grad_norm": 24.125,
      "learning_rate": 2.9193739250809042e-05,
      "loss": 3.5392,
      "num_input_tokens_seen": 24183744,
      "step": 17855
    },
    {
      "epoch": 0.5716663465847257,
      "grad_norm": 22.375,
      "learning_rate": 2.9184060567822463e-05,
      "loss": 4.5146,
      "num_input_tokens_seen": 24190336,
      "step": 17860
    },
    {
      "epoch": 0.5718263875552141,
      "grad_norm": 29.0,
      "learning_rate": 2.9174381239601166e-05,
      "loss": 4.3817,
      "num_input_tokens_seen": 24197472,
      "step": 17865
    },
    {
      "epoch": 0.5719864285257026,
      "grad_norm": 16.625,
      "learning_rate": 2.916470126763783e-05,
      "loss": 3.4293,
      "num_input_tokens_seen": 24204080,
      "step": 17870
    },
    {
      "epoch": 0.5721464694961911,
      "grad_norm": 34.25,
      "learning_rate": 2.9155020653425203e-05,
      "loss": 4.2068,
      "num_input_tokens_seen": 24210784,
      "step": 17875
    },
    {
      "epoch": 0.5723065104666795,
      "grad_norm": 19.875,
      "learning_rate": 2.9145339398456184e-05,
      "loss": 4.4056,
      "num_input_tokens_seen": 24217472,
      "step": 17880
    },
    {
      "epoch": 0.572466551437168,
      "grad_norm": 26.25,
      "learning_rate": 2.913565750422374e-05,
      "loss": 3.1782,
      "num_input_tokens_seen": 24224032,
      "step": 17885
    },
    {
      "epoch": 0.5726265924076563,
      "grad_norm": 39.5,
      "learning_rate": 2.9125974972220938e-05,
      "loss": 4.4215,
      "num_input_tokens_seen": 24230624,
      "step": 17890
    },
    {
      "epoch": 0.5727866333781448,
      "grad_norm": 33.75,
      "learning_rate": 2.9116291803940932e-05,
      "loss": 5.3269,
      "num_input_tokens_seen": 24237584,
      "step": 17895
    },
    {
      "epoch": 0.5729466743486332,
      "grad_norm": 30.875,
      "learning_rate": 2.910660800087701e-05,
      "loss": 4.9768,
      "num_input_tokens_seen": 24244048,
      "step": 17900
    },
    {
      "epoch": 0.5731067153191217,
      "grad_norm": 14.125,
      "learning_rate": 2.909692356452254e-05,
      "loss": 3.9971,
      "num_input_tokens_seen": 24250960,
      "step": 17905
    },
    {
      "epoch": 0.5732667562896101,
      "grad_norm": 37.25,
      "learning_rate": 2.9087238496370962e-05,
      "loss": 4.3037,
      "num_input_tokens_seen": 24257856,
      "step": 17910
    },
    {
      "epoch": 0.5734267972600986,
      "grad_norm": 26.125,
      "learning_rate": 2.907755279791583e-05,
      "loss": 4.1064,
      "num_input_tokens_seen": 24264768,
      "step": 17915
    },
    {
      "epoch": 0.573586838230587,
      "grad_norm": 20.0,
      "learning_rate": 2.906786647065083e-05,
      "loss": 4.5606,
      "num_input_tokens_seen": 24271728,
      "step": 17920
    },
    {
      "epoch": 0.5737468792010755,
      "grad_norm": 54.25,
      "learning_rate": 2.9058179516069695e-05,
      "loss": 4.4054,
      "num_input_tokens_seen": 24278480,
      "step": 17925
    },
    {
      "epoch": 0.573906920171564,
      "grad_norm": 30.25,
      "learning_rate": 2.9048491935666282e-05,
      "loss": 4.0803,
      "num_input_tokens_seen": 24285360,
      "step": 17930
    },
    {
      "epoch": 0.5740669611420524,
      "grad_norm": 28.5,
      "learning_rate": 2.9038803730934534e-05,
      "loss": 4.1422,
      "num_input_tokens_seen": 24292192,
      "step": 17935
    },
    {
      "epoch": 0.5742270021125409,
      "grad_norm": 33.0,
      "learning_rate": 2.9029114903368503e-05,
      "loss": 3.9805,
      "num_input_tokens_seen": 24299136,
      "step": 17940
    },
    {
      "epoch": 0.5743870430830292,
      "grad_norm": 34.75,
      "learning_rate": 2.9019425454462318e-05,
      "loss": 4.6325,
      "num_input_tokens_seen": 24305680,
      "step": 17945
    },
    {
      "epoch": 0.5745470840535177,
      "grad_norm": 27.25,
      "learning_rate": 2.9009735385710212e-05,
      "loss": 4.3192,
      "num_input_tokens_seen": 24313248,
      "step": 17950
    },
    {
      "epoch": 0.5747071250240061,
      "grad_norm": 50.0,
      "learning_rate": 2.900004469860652e-05,
      "loss": 4.623,
      "num_input_tokens_seen": 24319968,
      "step": 17955
    },
    {
      "epoch": 0.5748671659944946,
      "grad_norm": 46.0,
      "learning_rate": 2.8990353394645668e-05,
      "loss": 5.2443,
      "num_input_tokens_seen": 24326912,
      "step": 17960
    },
    {
      "epoch": 0.575027206964983,
      "grad_norm": 26.375,
      "learning_rate": 2.8980661475322186e-05,
      "loss": 4.3526,
      "num_input_tokens_seen": 24333200,
      "step": 17965
    },
    {
      "epoch": 0.5751872479354715,
      "grad_norm": 27.0,
      "learning_rate": 2.897096894213067e-05,
      "loss": 4.1288,
      "num_input_tokens_seen": 24339872,
      "step": 17970
    },
    {
      "epoch": 0.5753472889059599,
      "grad_norm": 20.875,
      "learning_rate": 2.8961275796565845e-05,
      "loss": 3.6772,
      "num_input_tokens_seen": 24346448,
      "step": 17975
    },
    {
      "epoch": 0.5755073298764484,
      "grad_norm": 39.25,
      "learning_rate": 2.8951582040122517e-05,
      "loss": 4.5286,
      "num_input_tokens_seen": 24353040,
      "step": 17980
    },
    {
      "epoch": 0.5756673708469369,
      "grad_norm": 24.75,
      "learning_rate": 2.894188767429557e-05,
      "loss": 4.2685,
      "num_input_tokens_seen": 24359600,
      "step": 17985
    },
    {
      "epoch": 0.5758274118174252,
      "grad_norm": 25.375,
      "learning_rate": 2.8932192700580014e-05,
      "loss": 4.5751,
      "num_input_tokens_seen": 24366576,
      "step": 17990
    },
    {
      "epoch": 0.5759874527879137,
      "grad_norm": 30.25,
      "learning_rate": 2.8922497120470916e-05,
      "loss": 4.1597,
      "num_input_tokens_seen": 24373584,
      "step": 17995
    },
    {
      "epoch": 0.5761474937584021,
      "grad_norm": 28.875,
      "learning_rate": 2.891280093546348e-05,
      "loss": 4.8494,
      "num_input_tokens_seen": 24380336,
      "step": 18000
    },
    {
      "epoch": 0.5761474937584021,
      "eval_loss": 4.490935325622559,
      "eval_runtime": 490.294,
      "eval_samples_per_second": 28.322,
      "eval_steps_per_second": 14.161,
      "num_input_tokens_seen": 24380336,
      "step": 18000
    },
    {
      "epoch": 0.5763075347288906,
      "grad_norm": 34.75,
      "learning_rate": 2.890310414705297e-05,
      "loss": 4.2363,
      "num_input_tokens_seen": 24387856,
      "step": 18005
    },
    {
      "epoch": 0.576467575699379,
      "grad_norm": 37.25,
      "learning_rate": 2.8893406756734742e-05,
      "loss": 4.035,
      "num_input_tokens_seen": 24394976,
      "step": 18010
    },
    {
      "epoch": 0.5766276166698675,
      "grad_norm": 27.125,
      "learning_rate": 2.888370876600427e-05,
      "loss": 4.4929,
      "num_input_tokens_seen": 24402256,
      "step": 18015
    },
    {
      "epoch": 0.5767876576403559,
      "grad_norm": 20.625,
      "learning_rate": 2.8874010176357104e-05,
      "loss": 4.5386,
      "num_input_tokens_seen": 24409536,
      "step": 18020
    },
    {
      "epoch": 0.5769476986108444,
      "grad_norm": 18.5,
      "learning_rate": 2.886431098928888e-05,
      "loss": 3.3821,
      "num_input_tokens_seen": 24416016,
      "step": 18025
    },
    {
      "epoch": 0.5771077395813328,
      "grad_norm": 25.375,
      "learning_rate": 2.885461120629534e-05,
      "loss": 3.631,
      "num_input_tokens_seen": 24422656,
      "step": 18030
    },
    {
      "epoch": 0.5772677805518213,
      "grad_norm": 23.25,
      "learning_rate": 2.8844910828872317e-05,
      "loss": 4.0612,
      "num_input_tokens_seen": 24430112,
      "step": 18035
    },
    {
      "epoch": 0.5774278215223098,
      "grad_norm": 68.5,
      "learning_rate": 2.8835209858515715e-05,
      "loss": 5.128,
      "num_input_tokens_seen": 24437280,
      "step": 18040
    },
    {
      "epoch": 0.5775878624927981,
      "grad_norm": 46.75,
      "learning_rate": 2.8825508296721566e-05,
      "loss": 5.2967,
      "num_input_tokens_seen": 24444096,
      "step": 18045
    },
    {
      "epoch": 0.5777479034632866,
      "grad_norm": 35.5,
      "learning_rate": 2.881580614498596e-05,
      "loss": 3.9074,
      "num_input_tokens_seen": 24450704,
      "step": 18050
    },
    {
      "epoch": 0.577907944433775,
      "grad_norm": 39.5,
      "learning_rate": 2.8806103404805103e-05,
      "loss": 4.425,
      "num_input_tokens_seen": 24457168,
      "step": 18055
    },
    {
      "epoch": 0.5780679854042635,
      "grad_norm": 35.5,
      "learning_rate": 2.8796400077675257e-05,
      "loss": 4.0646,
      "num_input_tokens_seen": 24463968,
      "step": 18060
    },
    {
      "epoch": 0.5782280263747519,
      "grad_norm": 11.75,
      "learning_rate": 2.8786696165092812e-05,
      "loss": 4.4878,
      "num_input_tokens_seen": 24470576,
      "step": 18065
    },
    {
      "epoch": 0.5783880673452404,
      "grad_norm": 19.875,
      "learning_rate": 2.8776991668554236e-05,
      "loss": 4.3976,
      "num_input_tokens_seen": 24477440,
      "step": 18070
    },
    {
      "epoch": 0.5785481083157288,
      "grad_norm": 28.25,
      "learning_rate": 2.876728658955608e-05,
      "loss": 3.8647,
      "num_input_tokens_seen": 24484400,
      "step": 18075
    },
    {
      "epoch": 0.5787081492862173,
      "grad_norm": 25.125,
      "learning_rate": 2.8757580929594986e-05,
      "loss": 3.9524,
      "num_input_tokens_seen": 24490992,
      "step": 18080
    },
    {
      "epoch": 0.5788681902567057,
      "grad_norm": 41.0,
      "learning_rate": 2.87478746901677e-05,
      "loss": 3.825,
      "num_input_tokens_seen": 24497952,
      "step": 18085
    },
    {
      "epoch": 0.5790282312271942,
      "grad_norm": 24.125,
      "learning_rate": 2.873816787277103e-05,
      "loss": 4.9794,
      "num_input_tokens_seen": 24504112,
      "step": 18090
    },
    {
      "epoch": 0.5791882721976827,
      "grad_norm": 30.375,
      "learning_rate": 2.8728460478901903e-05,
      "loss": 5.4194,
      "num_input_tokens_seen": 24510704,
      "step": 18095
    },
    {
      "epoch": 0.579348313168171,
      "grad_norm": 28.75,
      "learning_rate": 2.8718752510057307e-05,
      "loss": 3.1706,
      "num_input_tokens_seen": 24517376,
      "step": 18100
    },
    {
      "epoch": 0.5795083541386595,
      "grad_norm": 22.875,
      "learning_rate": 2.870904396773435e-05,
      "loss": 4.0907,
      "num_input_tokens_seen": 24524304,
      "step": 18105
    },
    {
      "epoch": 0.5796683951091479,
      "grad_norm": 40.25,
      "learning_rate": 2.86993348534302e-05,
      "loss": 4.7945,
      "num_input_tokens_seen": 24530624,
      "step": 18110
    },
    {
      "epoch": 0.5798284360796364,
      "grad_norm": 30.625,
      "learning_rate": 2.868962516864212e-05,
      "loss": 3.6497,
      "num_input_tokens_seen": 24537120,
      "step": 18115
    },
    {
      "epoch": 0.5799884770501248,
      "grad_norm": 21.875,
      "learning_rate": 2.8679914914867477e-05,
      "loss": 4.5341,
      "num_input_tokens_seen": 24543904,
      "step": 18120
    },
    {
      "epoch": 0.5801485180206133,
      "grad_norm": 31.5,
      "learning_rate": 2.8670204093603713e-05,
      "loss": 3.9386,
      "num_input_tokens_seen": 24550800,
      "step": 18125
    },
    {
      "epoch": 0.5803085589911017,
      "grad_norm": 31.75,
      "learning_rate": 2.8660492706348357e-05,
      "loss": 3.7212,
      "num_input_tokens_seen": 24557552,
      "step": 18130
    },
    {
      "epoch": 0.5804685999615902,
      "grad_norm": 26.5,
      "learning_rate": 2.8650780754599022e-05,
      "loss": 3.8079,
      "num_input_tokens_seen": 24564192,
      "step": 18135
    },
    {
      "epoch": 0.5806286409320787,
      "grad_norm": 24.0,
      "learning_rate": 2.8641068239853407e-05,
      "loss": 4.1396,
      "num_input_tokens_seen": 24570752,
      "step": 18140
    },
    {
      "epoch": 0.580788681902567,
      "grad_norm": 24.625,
      "learning_rate": 2.863135516360932e-05,
      "loss": 4.4674,
      "num_input_tokens_seen": 24577392,
      "step": 18145
    },
    {
      "epoch": 0.5809487228730555,
      "grad_norm": 19.625,
      "learning_rate": 2.8621641527364633e-05,
      "loss": 4.1912,
      "num_input_tokens_seen": 24584016,
      "step": 18150
    },
    {
      "epoch": 0.5811087638435439,
      "grad_norm": 44.75,
      "learning_rate": 2.8611927332617313e-05,
      "loss": 4.8951,
      "num_input_tokens_seen": 24590960,
      "step": 18155
    },
    {
      "epoch": 0.5812688048140324,
      "grad_norm": 30.625,
      "learning_rate": 2.8602212580865405e-05,
      "loss": 5.2277,
      "num_input_tokens_seen": 24598384,
      "step": 18160
    },
    {
      "epoch": 0.5814288457845208,
      "grad_norm": 34.0,
      "learning_rate": 2.859249727360705e-05,
      "loss": 4.2271,
      "num_input_tokens_seen": 24605568,
      "step": 18165
    },
    {
      "epoch": 0.5815888867550093,
      "grad_norm": 27.25,
      "learning_rate": 2.8582781412340465e-05,
      "loss": 4.0643,
      "num_input_tokens_seen": 24611968,
      "step": 18170
    },
    {
      "epoch": 0.5817489277254977,
      "grad_norm": 16.375,
      "learning_rate": 2.857306499856397e-05,
      "loss": 4.9567,
      "num_input_tokens_seen": 24619184,
      "step": 18175
    },
    {
      "epoch": 0.5819089686959862,
      "grad_norm": 22.625,
      "learning_rate": 2.856334803377594e-05,
      "loss": 3.2428,
      "num_input_tokens_seen": 24626400,
      "step": 18180
    },
    {
      "epoch": 0.5820690096664746,
      "grad_norm": 41.5,
      "learning_rate": 2.8553630519474867e-05,
      "loss": 4.3475,
      "num_input_tokens_seen": 24632720,
      "step": 18185
    },
    {
      "epoch": 0.5822290506369631,
      "grad_norm": 23.125,
      "learning_rate": 2.8543912457159317e-05,
      "loss": 5.2265,
      "num_input_tokens_seen": 24639536,
      "step": 18190
    },
    {
      "epoch": 0.5823890916074516,
      "grad_norm": 16.375,
      "learning_rate": 2.853419384832792e-05,
      "loss": 3.8446,
      "num_input_tokens_seen": 24646304,
      "step": 18195
    },
    {
      "epoch": 0.5825491325779399,
      "grad_norm": 17.625,
      "learning_rate": 2.8524474694479423e-05,
      "loss": 4.3989,
      "num_input_tokens_seen": 24653072,
      "step": 18200
    },
    {
      "epoch": 0.5825491325779399,
      "eval_loss": 4.4915924072265625,
      "eval_runtime": 490.5394,
      "eval_samples_per_second": 28.308,
      "eval_steps_per_second": 14.154,
      "num_input_tokens_seen": 24653072,
      "step": 18200
    },
    {
      "epoch": 0.5827091735484284,
      "grad_norm": 28.625,
      "learning_rate": 2.851475499711264e-05,
      "loss": 4.7738,
      "num_input_tokens_seen": 24660000,
      "step": 18205
    },
    {
      "epoch": 0.5828692145189168,
      "grad_norm": 36.25,
      "learning_rate": 2.8505034757726468e-05,
      "loss": 4.1577,
      "num_input_tokens_seen": 24666544,
      "step": 18210
    },
    {
      "epoch": 0.5830292554894053,
      "grad_norm": 26.625,
      "learning_rate": 2.8495313977819886e-05,
      "loss": 4.914,
      "num_input_tokens_seen": 24673168,
      "step": 18215
    },
    {
      "epoch": 0.5831892964598937,
      "grad_norm": 33.25,
      "learning_rate": 2.8485592658891956e-05,
      "loss": 3.8408,
      "num_input_tokens_seen": 24680144,
      "step": 18220
    },
    {
      "epoch": 0.5833493374303822,
      "grad_norm": 40.5,
      "learning_rate": 2.8475870802441844e-05,
      "loss": 4.5991,
      "num_input_tokens_seen": 24687648,
      "step": 18225
    },
    {
      "epoch": 0.5835093784008706,
      "grad_norm": 48.75,
      "learning_rate": 2.8466148409968774e-05,
      "loss": 3.9428,
      "num_input_tokens_seen": 24694336,
      "step": 18230
    },
    {
      "epoch": 0.5836694193713591,
      "grad_norm": 44.75,
      "learning_rate": 2.8456425482972067e-05,
      "loss": 5.2271,
      "num_input_tokens_seen": 24701008,
      "step": 18235
    },
    {
      "epoch": 0.5838294603418475,
      "grad_norm": 17.5,
      "learning_rate": 2.84467020229511e-05,
      "loss": 3.8654,
      "num_input_tokens_seen": 24707872,
      "step": 18240
    },
    {
      "epoch": 0.583989501312336,
      "grad_norm": 29.5,
      "learning_rate": 2.8436978031405375e-05,
      "loss": 4.1428,
      "num_input_tokens_seen": 24714944,
      "step": 18245
    },
    {
      "epoch": 0.5841495422828245,
      "grad_norm": 29.25,
      "learning_rate": 2.842725350983445e-05,
      "loss": 3.903,
      "num_input_tokens_seen": 24721616,
      "step": 18250
    },
    {
      "epoch": 0.5843095832533128,
      "grad_norm": 38.75,
      "learning_rate": 2.8417528459737957e-05,
      "loss": 3.8857,
      "num_input_tokens_seen": 24728048,
      "step": 18255
    },
    {
      "epoch": 0.5844696242238013,
      "grad_norm": 43.5,
      "learning_rate": 2.8407802882615624e-05,
      "loss": 4.2292,
      "num_input_tokens_seen": 24734480,
      "step": 18260
    },
    {
      "epoch": 0.5846296651942897,
      "grad_norm": 14.8125,
      "learning_rate": 2.8398076779967277e-05,
      "loss": 4.6004,
      "num_input_tokens_seen": 24741296,
      "step": 18265
    },
    {
      "epoch": 0.5847897061647782,
      "grad_norm": 26.875,
      "learning_rate": 2.8388350153292774e-05,
      "loss": 4.9506,
      "num_input_tokens_seen": 24747632,
      "step": 18270
    },
    {
      "epoch": 0.5849497471352666,
      "grad_norm": 30.625,
      "learning_rate": 2.8378623004092103e-05,
      "loss": 5.3581,
      "num_input_tokens_seen": 24754416,
      "step": 18275
    },
    {
      "epoch": 0.5851097881057551,
      "grad_norm": 35.0,
      "learning_rate": 2.8368895333865302e-05,
      "loss": 4.3547,
      "num_input_tokens_seen": 24760976,
      "step": 18280
    },
    {
      "epoch": 0.5852698290762435,
      "grad_norm": 37.75,
      "learning_rate": 2.835916714411251e-05,
      "loss": 4.1525,
      "num_input_tokens_seen": 24767648,
      "step": 18285
    },
    {
      "epoch": 0.585429870046732,
      "grad_norm": 22.25,
      "learning_rate": 2.8349438436333926e-05,
      "loss": 5.0304,
      "num_input_tokens_seen": 24774112,
      "step": 18290
    },
    {
      "epoch": 0.5855899110172204,
      "grad_norm": 29.125,
      "learning_rate": 2.833970921202984e-05,
      "loss": 4.1451,
      "num_input_tokens_seen": 24780752,
      "step": 18295
    },
    {
      "epoch": 0.5857499519877089,
      "grad_norm": 21.0,
      "learning_rate": 2.8329979472700628e-05,
      "loss": 3.7592,
      "num_input_tokens_seen": 24787504,
      "step": 18300
    },
    {
      "epoch": 0.5859099929581973,
      "grad_norm": 24.75,
      "learning_rate": 2.832024921984674e-05,
      "loss": 4.0731,
      "num_input_tokens_seen": 24794480,
      "step": 18305
    },
    {
      "epoch": 0.5860700339286857,
      "grad_norm": 34.0,
      "learning_rate": 2.8310518454968693e-05,
      "loss": 5.2818,
      "num_input_tokens_seen": 24801280,
      "step": 18310
    },
    {
      "epoch": 0.5862300748991742,
      "grad_norm": 31.375,
      "learning_rate": 2.8300787179567095e-05,
      "loss": 5.049,
      "num_input_tokens_seen": 24808080,
      "step": 18315
    },
    {
      "epoch": 0.5863901158696626,
      "grad_norm": 31.875,
      "learning_rate": 2.8291055395142636e-05,
      "loss": 3.7517,
      "num_input_tokens_seen": 24815088,
      "step": 18320
    },
    {
      "epoch": 0.5865501568401511,
      "grad_norm": 23.625,
      "learning_rate": 2.8281323103196073e-05,
      "loss": 4.4089,
      "num_input_tokens_seen": 24821520,
      "step": 18325
    },
    {
      "epoch": 0.5867101978106395,
      "grad_norm": 26.5,
      "learning_rate": 2.8271590305228256e-05,
      "loss": 4.9444,
      "num_input_tokens_seen": 24828592,
      "step": 18330
    },
    {
      "epoch": 0.586870238781128,
      "grad_norm": 29.625,
      "learning_rate": 2.82618570027401e-05,
      "loss": 4.5738,
      "num_input_tokens_seen": 24835424,
      "step": 18335
    },
    {
      "epoch": 0.5870302797516164,
      "grad_norm": 52.5,
      "learning_rate": 2.8252123197232604e-05,
      "loss": 3.6269,
      "num_input_tokens_seen": 24842448,
      "step": 18340
    },
    {
      "epoch": 0.5871903207221049,
      "grad_norm": 30.25,
      "learning_rate": 2.8242388890206843e-05,
      "loss": 4.1857,
      "num_input_tokens_seen": 24849408,
      "step": 18345
    },
    {
      "epoch": 0.5873503616925934,
      "grad_norm": 26.5,
      "learning_rate": 2.8232654083163967e-05,
      "loss": 5.0594,
      "num_input_tokens_seen": 24856416,
      "step": 18350
    },
    {
      "epoch": 0.5875104026630817,
      "grad_norm": 34.5,
      "learning_rate": 2.822291877760521e-05,
      "loss": 4.4283,
      "num_input_tokens_seen": 24863136,
      "step": 18355
    },
    {
      "epoch": 0.5876704436335702,
      "grad_norm": 39.25,
      "learning_rate": 2.8213182975031864e-05,
      "loss": 3.0284,
      "num_input_tokens_seen": 24869904,
      "step": 18360
    },
    {
      "epoch": 0.5878304846040586,
      "grad_norm": 31.625,
      "learning_rate": 2.8203446676945337e-05,
      "loss": 4.113,
      "num_input_tokens_seen": 24877248,
      "step": 18365
    },
    {
      "epoch": 0.5879905255745471,
      "grad_norm": 16.25,
      "learning_rate": 2.8193709884847075e-05,
      "loss": 3.8779,
      "num_input_tokens_seen": 24883472,
      "step": 18370
    },
    {
      "epoch": 0.5881505665450355,
      "grad_norm": 43.25,
      "learning_rate": 2.8183972600238605e-05,
      "loss": 4.2734,
      "num_input_tokens_seen": 24890432,
      "step": 18375
    },
    {
      "epoch": 0.588310607515524,
      "grad_norm": 48.25,
      "learning_rate": 2.817423482462156e-05,
      "loss": 4.341,
      "num_input_tokens_seen": 24896944,
      "step": 18380
    },
    {
      "epoch": 0.5884706484860124,
      "grad_norm": 21.5,
      "learning_rate": 2.8164496559497605e-05,
      "loss": 4.6865,
      "num_input_tokens_seen": 24904016,
      "step": 18385
    },
    {
      "epoch": 0.5886306894565009,
      "grad_norm": 31.25,
      "learning_rate": 2.815475780636852e-05,
      "loss": 3.7818,
      "num_input_tokens_seen": 24911008,
      "step": 18390
    },
    {
      "epoch": 0.5887907304269893,
      "grad_norm": 30.375,
      "learning_rate": 2.814501856673613e-05,
      "loss": 4.697,
      "num_input_tokens_seen": 24918032,
      "step": 18395
    },
    {
      "epoch": 0.5889507713974778,
      "grad_norm": 33.0,
      "learning_rate": 2.8135278842102353e-05,
      "loss": 3.9486,
      "num_input_tokens_seen": 24924912,
      "step": 18400
    },
    {
      "epoch": 0.5889507713974778,
      "eval_loss": 4.457133769989014,
      "eval_runtime": 489.7077,
      "eval_samples_per_second": 28.356,
      "eval_steps_per_second": 14.178,
      "num_input_tokens_seen": 24924912,
      "step": 18400
    },
    {
      "epoch": 0.5891108123679663,
      "grad_norm": 24.0,
      "learning_rate": 2.8125538633969183e-05,
      "loss": 4.4112,
      "num_input_tokens_seen": 24931712,
      "step": 18405
    },
    {
      "epoch": 0.5892708533384546,
      "grad_norm": 14.5625,
      "learning_rate": 2.8115797943838677e-05,
      "loss": 4.4409,
      "num_input_tokens_seen": 24938288,
      "step": 18410
    },
    {
      "epoch": 0.5894308943089431,
      "grad_norm": 33.75,
      "learning_rate": 2.810605677321298e-05,
      "loss": 4.5197,
      "num_input_tokens_seen": 24945264,
      "step": 18415
    },
    {
      "epoch": 0.5895909352794315,
      "grad_norm": 25.875,
      "learning_rate": 2.809631512359428e-05,
      "loss": 3.9637,
      "num_input_tokens_seen": 24951920,
      "step": 18420
    },
    {
      "epoch": 0.58975097624992,
      "grad_norm": 28.5,
      "learning_rate": 2.8086572996484884e-05,
      "loss": 3.7622,
      "num_input_tokens_seen": 24958752,
      "step": 18425
    },
    {
      "epoch": 0.5899110172204084,
      "grad_norm": 38.75,
      "learning_rate": 2.8076830393387143e-05,
      "loss": 4.5991,
      "num_input_tokens_seen": 24965408,
      "step": 18430
    },
    {
      "epoch": 0.5900710581908969,
      "grad_norm": 36.25,
      "learning_rate": 2.8067087315803497e-05,
      "loss": 4.3487,
      "num_input_tokens_seen": 24972208,
      "step": 18435
    },
    {
      "epoch": 0.5902310991613853,
      "grad_norm": 37.25,
      "learning_rate": 2.8057343765236433e-05,
      "loss": 3.8818,
      "num_input_tokens_seen": 24978896,
      "step": 18440
    },
    {
      "epoch": 0.5903911401318738,
      "grad_norm": 46.25,
      "learning_rate": 2.804759974318854e-05,
      "loss": 3.997,
      "num_input_tokens_seen": 24986144,
      "step": 18445
    },
    {
      "epoch": 0.5905511811023622,
      "grad_norm": 18.5,
      "learning_rate": 2.8037855251162482e-05,
      "loss": 3.5372,
      "num_input_tokens_seen": 24992896,
      "step": 18450
    },
    {
      "epoch": 0.5907112220728507,
      "grad_norm": 31.375,
      "learning_rate": 2.802811029066096e-05,
      "loss": 4.0088,
      "num_input_tokens_seen": 24999552,
      "step": 18455
    },
    {
      "epoch": 0.5908712630433391,
      "grad_norm": 32.75,
      "learning_rate": 2.8018364863186764e-05,
      "loss": 4.4126,
      "num_input_tokens_seen": 25006752,
      "step": 18460
    },
    {
      "epoch": 0.5910313040138275,
      "grad_norm": 29.375,
      "learning_rate": 2.800861897024279e-05,
      "loss": 5.7409,
      "num_input_tokens_seen": 25013472,
      "step": 18465
    },
    {
      "epoch": 0.591191344984316,
      "grad_norm": 54.75,
      "learning_rate": 2.799887261333196e-05,
      "loss": 4.4253,
      "num_input_tokens_seen": 25020176,
      "step": 18470
    },
    {
      "epoch": 0.5913513859548044,
      "grad_norm": 32.75,
      "learning_rate": 2.798912579395728e-05,
      "loss": 3.9002,
      "num_input_tokens_seen": 25026992,
      "step": 18475
    },
    {
      "epoch": 0.5915114269252929,
      "grad_norm": 35.5,
      "learning_rate": 2.797937851362185e-05,
      "loss": 4.6968,
      "num_input_tokens_seen": 25033920,
      "step": 18480
    },
    {
      "epoch": 0.5916714678957813,
      "grad_norm": 30.625,
      "learning_rate": 2.7969630773828802e-05,
      "loss": 5.1654,
      "num_input_tokens_seen": 25040800,
      "step": 18485
    },
    {
      "epoch": 0.5918315088662698,
      "grad_norm": 29.125,
      "learning_rate": 2.7959882576081382e-05,
      "loss": 4.5823,
      "num_input_tokens_seen": 25047248,
      "step": 18490
    },
    {
      "epoch": 0.5919915498367582,
      "grad_norm": 21.25,
      "learning_rate": 2.795013392188286e-05,
      "loss": 4.3173,
      "num_input_tokens_seen": 25053792,
      "step": 18495
    },
    {
      "epoch": 0.5921515908072467,
      "grad_norm": 22.375,
      "learning_rate": 2.7940384812736614e-05,
      "loss": 4.4449,
      "num_input_tokens_seen": 25060688,
      "step": 18500
    },
    {
      "epoch": 0.592311631777735,
      "grad_norm": 34.0,
      "learning_rate": 2.7930635250146087e-05,
      "loss": 4.1871,
      "num_input_tokens_seen": 25067376,
      "step": 18505
    },
    {
      "epoch": 0.5924716727482235,
      "grad_norm": 32.25,
      "learning_rate": 2.792088523561477e-05,
      "loss": 3.9323,
      "num_input_tokens_seen": 25074064,
      "step": 18510
    },
    {
      "epoch": 0.592631713718712,
      "grad_norm": 28.875,
      "learning_rate": 2.7911134770646246e-05,
      "loss": 4.7506,
      "num_input_tokens_seen": 25081344,
      "step": 18515
    },
    {
      "epoch": 0.5927917546892004,
      "grad_norm": 25.0,
      "learning_rate": 2.7901383856744157e-05,
      "loss": 4.4248,
      "num_input_tokens_seen": 25088208,
      "step": 18520
    },
    {
      "epoch": 0.5929517956596889,
      "grad_norm": 23.375,
      "learning_rate": 2.7891632495412217e-05,
      "loss": 4.6213,
      "num_input_tokens_seen": 25095008,
      "step": 18525
    },
    {
      "epoch": 0.5931118366301773,
      "grad_norm": 28.5,
      "learning_rate": 2.7881880688154205e-05,
      "loss": 4.1406,
      "num_input_tokens_seen": 25101632,
      "step": 18530
    },
    {
      "epoch": 0.5932718776006658,
      "grad_norm": 32.0,
      "learning_rate": 2.7872128436473977e-05,
      "loss": 3.8143,
      "num_input_tokens_seen": 25108432,
      "step": 18535
    },
    {
      "epoch": 0.5934319185711542,
      "grad_norm": 32.0,
      "learning_rate": 2.7862375741875448e-05,
      "loss": 3.5826,
      "num_input_tokens_seen": 25114928,
      "step": 18540
    },
    {
      "epoch": 0.5935919595416427,
      "grad_norm": 28.75,
      "learning_rate": 2.785262260586261e-05,
      "loss": 5.0656,
      "num_input_tokens_seen": 25121920,
      "step": 18545
    },
    {
      "epoch": 0.5937520005121311,
      "grad_norm": 28.0,
      "learning_rate": 2.7842869029939517e-05,
      "loss": 4.9672,
      "num_input_tokens_seen": 25128304,
      "step": 18550
    },
    {
      "epoch": 0.5939120414826196,
      "grad_norm": 30.875,
      "learning_rate": 2.7833115015610296e-05,
      "loss": 4.8189,
      "num_input_tokens_seen": 25135168,
      "step": 18555
    },
    {
      "epoch": 0.5940720824531079,
      "grad_norm": 40.75,
      "learning_rate": 2.7823360564379136e-05,
      "loss": 4.1267,
      "num_input_tokens_seen": 25141632,
      "step": 18560
    },
    {
      "epoch": 0.5942321234235964,
      "grad_norm": 21.625,
      "learning_rate": 2.7813605677750297e-05,
      "loss": 4.5082,
      "num_input_tokens_seen": 25148352,
      "step": 18565
    },
    {
      "epoch": 0.5943921643940849,
      "grad_norm": 37.75,
      "learning_rate": 2.7803850357228102e-05,
      "loss": 4.5379,
      "num_input_tokens_seen": 25154960,
      "step": 18570
    },
    {
      "epoch": 0.5945522053645733,
      "grad_norm": 29.375,
      "learning_rate": 2.779409460431695e-05,
      "loss": 4.0392,
      "num_input_tokens_seen": 25161616,
      "step": 18575
    },
    {
      "epoch": 0.5947122463350618,
      "grad_norm": 30.25,
      "learning_rate": 2.778433842052129e-05,
      "loss": 4.1426,
      "num_input_tokens_seen": 25168656,
      "step": 18580
    },
    {
      "epoch": 0.5948722873055502,
      "grad_norm": 35.0,
      "learning_rate": 2.7774581807345664e-05,
      "loss": 4.9384,
      "num_input_tokens_seen": 25175984,
      "step": 18585
    },
    {
      "epoch": 0.5950323282760387,
      "grad_norm": 48.5,
      "learning_rate": 2.776482476629465e-05,
      "loss": 5.2535,
      "num_input_tokens_seen": 25182848,
      "step": 18590
    },
    {
      "epoch": 0.5951923692465271,
      "grad_norm": 32.0,
      "learning_rate": 2.7755067298872924e-05,
      "loss": 5.2518,
      "num_input_tokens_seen": 25189824,
      "step": 18595
    },
    {
      "epoch": 0.5953524102170156,
      "grad_norm": 52.25,
      "learning_rate": 2.774530940658518e-05,
      "loss": 4.4004,
      "num_input_tokens_seen": 25196400,
      "step": 18600
    },
    {
      "epoch": 0.5953524102170156,
      "eval_loss": 4.5108418464660645,
      "eval_runtime": 490.3027,
      "eval_samples_per_second": 28.321,
      "eval_steps_per_second": 14.161,
      "num_input_tokens_seen": 25196400,
      "step": 18600
    },
    {
      "epoch": 0.595512451187504,
      "grad_norm": 31.125,
      "learning_rate": 2.7735551090936236e-05,
      "loss": 4.5818,
      "num_input_tokens_seen": 25203536,
      "step": 18605
    },
    {
      "epoch": 0.5956724921579925,
      "grad_norm": 17.875,
      "learning_rate": 2.7725792353430934e-05,
      "loss": 3.7179,
      "num_input_tokens_seen": 25210224,
      "step": 18610
    },
    {
      "epoch": 0.595832533128481,
      "grad_norm": 33.5,
      "learning_rate": 2.77160331955742e-05,
      "loss": 4.1938,
      "num_input_tokens_seen": 25217296,
      "step": 18615
    },
    {
      "epoch": 0.5959925740989693,
      "grad_norm": 28.875,
      "learning_rate": 2.7706273618871008e-05,
      "loss": 3.9513,
      "num_input_tokens_seen": 25224096,
      "step": 18620
    },
    {
      "epoch": 0.5961526150694578,
      "grad_norm": 43.0,
      "learning_rate": 2.769651362482642e-05,
      "loss": 4.6589,
      "num_input_tokens_seen": 25230896,
      "step": 18625
    },
    {
      "epoch": 0.5963126560399462,
      "grad_norm": 31.75,
      "learning_rate": 2.768675321494555e-05,
      "loss": 4.9298,
      "num_input_tokens_seen": 25238080,
      "step": 18630
    },
    {
      "epoch": 0.5964726970104347,
      "grad_norm": 21.875,
      "learning_rate": 2.7676992390733565e-05,
      "loss": 4.8137,
      "num_input_tokens_seen": 25244720,
      "step": 18635
    },
    {
      "epoch": 0.5966327379809231,
      "grad_norm": 43.0,
      "learning_rate": 2.766723115369571e-05,
      "loss": 4.3686,
      "num_input_tokens_seen": 25251104,
      "step": 18640
    },
    {
      "epoch": 0.5967927789514116,
      "grad_norm": 39.0,
      "learning_rate": 2.765746950533729e-05,
      "loss": 4.9355,
      "num_input_tokens_seen": 25257664,
      "step": 18645
    },
    {
      "epoch": 0.5969528199219,
      "grad_norm": 18.5,
      "learning_rate": 2.7647707447163684e-05,
      "loss": 4.4404,
      "num_input_tokens_seen": 25264128,
      "step": 18650
    },
    {
      "epoch": 0.5971128608923885,
      "grad_norm": 45.0,
      "learning_rate": 2.7637944980680315e-05,
      "loss": 4.1205,
      "num_input_tokens_seen": 25270704,
      "step": 18655
    },
    {
      "epoch": 0.5972729018628768,
      "grad_norm": 26.5,
      "learning_rate": 2.762818210739268e-05,
      "loss": 4.6498,
      "num_input_tokens_seen": 25277248,
      "step": 18660
    },
    {
      "epoch": 0.5974329428333653,
      "grad_norm": 28.875,
      "learning_rate": 2.7618418828806332e-05,
      "loss": 3.4685,
      "num_input_tokens_seen": 25283936,
      "step": 18665
    },
    {
      "epoch": 0.5975929838038538,
      "grad_norm": 30.375,
      "learning_rate": 2.76086551464269e-05,
      "loss": 4.9342,
      "num_input_tokens_seen": 25290400,
      "step": 18670
    },
    {
      "epoch": 0.5977530247743422,
      "grad_norm": 46.5,
      "learning_rate": 2.759889106176006e-05,
      "loss": 5.0711,
      "num_input_tokens_seen": 25297264,
      "step": 18675
    },
    {
      "epoch": 0.5979130657448307,
      "grad_norm": 21.875,
      "learning_rate": 2.758912657631156e-05,
      "loss": 3.5644,
      "num_input_tokens_seen": 25304176,
      "step": 18680
    },
    {
      "epoch": 0.5980731067153191,
      "grad_norm": 15.125,
      "learning_rate": 2.7579361691587198e-05,
      "loss": 4.0528,
      "num_input_tokens_seen": 25310928,
      "step": 18685
    },
    {
      "epoch": 0.5982331476858076,
      "grad_norm": 32.75,
      "learning_rate": 2.756959640909285e-05,
      "loss": 4.2539,
      "num_input_tokens_seen": 25317664,
      "step": 18690
    },
    {
      "epoch": 0.598393188656296,
      "grad_norm": 35.5,
      "learning_rate": 2.7559830730334452e-05,
      "loss": 4.0074,
      "num_input_tokens_seen": 25324512,
      "step": 18695
    },
    {
      "epoch": 0.5985532296267845,
      "grad_norm": 24.0,
      "learning_rate": 2.7550064656817988e-05,
      "loss": 4.1045,
      "num_input_tokens_seen": 25331552,
      "step": 18700
    },
    {
      "epoch": 0.5987132705972729,
      "grad_norm": 33.25,
      "learning_rate": 2.7540298190049503e-05,
      "loss": 4.5471,
      "num_input_tokens_seen": 25338048,
      "step": 18705
    },
    {
      "epoch": 0.5988733115677614,
      "grad_norm": 28.25,
      "learning_rate": 2.7530531331535107e-05,
      "loss": 4.1506,
      "num_input_tokens_seen": 25344736,
      "step": 18710
    },
    {
      "epoch": 0.5990333525382497,
      "grad_norm": 23.125,
      "learning_rate": 2.752076408278099e-05,
      "loss": 4.0018,
      "num_input_tokens_seen": 25352400,
      "step": 18715
    },
    {
      "epoch": 0.5991933935087382,
      "grad_norm": 22.0,
      "learning_rate": 2.751099644529337e-05,
      "loss": 4.1535,
      "num_input_tokens_seen": 25359328,
      "step": 18720
    },
    {
      "epoch": 0.5993534344792267,
      "grad_norm": 42.5,
      "learning_rate": 2.7501228420578533e-05,
      "loss": 4.9728,
      "num_input_tokens_seen": 25366224,
      "step": 18725
    },
    {
      "epoch": 0.5995134754497151,
      "grad_norm": 42.25,
      "learning_rate": 2.7491460010142857e-05,
      "loss": 4.9363,
      "num_input_tokens_seen": 25372816,
      "step": 18730
    },
    {
      "epoch": 0.5996735164202036,
      "grad_norm": 36.5,
      "learning_rate": 2.7481691215492727e-05,
      "loss": 4.5644,
      "num_input_tokens_seen": 25379808,
      "step": 18735
    },
    {
      "epoch": 0.599833557390692,
      "grad_norm": 43.25,
      "learning_rate": 2.747192203813463e-05,
      "loss": 5.1675,
      "num_input_tokens_seen": 25386560,
      "step": 18740
    },
    {
      "epoch": 0.5999935983611805,
      "grad_norm": 20.875,
      "learning_rate": 2.7462152479575087e-05,
      "loss": 4.9083,
      "num_input_tokens_seen": 25393024,
      "step": 18745
    },
    {
      "epoch": 0.6001536393316689,
      "grad_norm": 15.9375,
      "learning_rate": 2.7452382541320697e-05,
      "loss": 3.367,
      "num_input_tokens_seen": 25399888,
      "step": 18750
    },
    {
      "epoch": 0.6003136803021574,
      "grad_norm": 17.75,
      "learning_rate": 2.7442612224878096e-05,
      "loss": 4.3743,
      "num_input_tokens_seen": 25406544,
      "step": 18755
    },
    {
      "epoch": 0.6004737212726458,
      "grad_norm": 32.25,
      "learning_rate": 2.7432841531753994e-05,
      "loss": 4.1889,
      "num_input_tokens_seen": 25413520,
      "step": 18760
    },
    {
      "epoch": 0.6006337622431343,
      "grad_norm": 17.125,
      "learning_rate": 2.7423070463455147e-05,
      "loss": 3.2069,
      "num_input_tokens_seen": 25420432,
      "step": 18765
    },
    {
      "epoch": 0.6007938032136226,
      "grad_norm": 14.3125,
      "learning_rate": 2.7413299021488397e-05,
      "loss": 4.4063,
      "num_input_tokens_seen": 25427024,
      "step": 18770
    },
    {
      "epoch": 0.6009538441841111,
      "grad_norm": 37.75,
      "learning_rate": 2.7403527207360615e-05,
      "loss": 4.6539,
      "num_input_tokens_seen": 25433968,
      "step": 18775
    },
    {
      "epoch": 0.6011138851545996,
      "grad_norm": 25.125,
      "learning_rate": 2.7393755022578722e-05,
      "loss": 4.5752,
      "num_input_tokens_seen": 25441072,
      "step": 18780
    },
    {
      "epoch": 0.601273926125088,
      "grad_norm": 34.5,
      "learning_rate": 2.7383982468649714e-05,
      "loss": 4.6699,
      "num_input_tokens_seen": 25448320,
      "step": 18785
    },
    {
      "epoch": 0.6014339670955765,
      "grad_norm": 49.25,
      "learning_rate": 2.7374209547080665e-05,
      "loss": 4.2449,
      "num_input_tokens_seen": 25455744,
      "step": 18790
    },
    {
      "epoch": 0.6015940080660649,
      "grad_norm": 17.375,
      "learning_rate": 2.7364436259378663e-05,
      "loss": 4.5572,
      "num_input_tokens_seen": 25462080,
      "step": 18795
    },
    {
      "epoch": 0.6017540490365534,
      "grad_norm": 26.375,
      "learning_rate": 2.735466260705088e-05,
      "loss": 4.237,
      "num_input_tokens_seen": 25468816,
      "step": 18800
    },
    {
      "epoch": 0.6017540490365534,
      "eval_loss": 4.479493141174316,
      "eval_runtime": 489.584,
      "eval_samples_per_second": 28.363,
      "eval_steps_per_second": 14.181,
      "num_input_tokens_seen": 25468816,
      "step": 18800
    },
    {
      "epoch": 0.6019140900070418,
      "grad_norm": 23.375,
      "learning_rate": 2.7344888591604524e-05,
      "loss": 5.3278,
      "num_input_tokens_seen": 25475648,
      "step": 18805
    },
    {
      "epoch": 0.6020741309775303,
      "grad_norm": 44.25,
      "learning_rate": 2.7335114214546893e-05,
      "loss": 4.6309,
      "num_input_tokens_seen": 25482336,
      "step": 18810
    },
    {
      "epoch": 0.6022341719480186,
      "grad_norm": 24.625,
      "learning_rate": 2.7325339477385293e-05,
      "loss": 3.6203,
      "num_input_tokens_seen": 25488960,
      "step": 18815
    },
    {
      "epoch": 0.6023942129185071,
      "grad_norm": 43.0,
      "learning_rate": 2.7315564381627128e-05,
      "loss": 4.1708,
      "num_input_tokens_seen": 25495776,
      "step": 18820
    },
    {
      "epoch": 0.6025542538889955,
      "grad_norm": 30.875,
      "learning_rate": 2.7305788928779835e-05,
      "loss": 4.058,
      "num_input_tokens_seen": 25502480,
      "step": 18825
    },
    {
      "epoch": 0.602714294859484,
      "grad_norm": 17.875,
      "learning_rate": 2.729601312035091e-05,
      "loss": 3.7471,
      "num_input_tokens_seen": 25509184,
      "step": 18830
    },
    {
      "epoch": 0.6028743358299725,
      "grad_norm": 42.25,
      "learning_rate": 2.7286236957847915e-05,
      "loss": 4.3105,
      "num_input_tokens_seen": 25516032,
      "step": 18835
    },
    {
      "epoch": 0.6030343768004609,
      "grad_norm": 19.625,
      "learning_rate": 2.7276460442778446e-05,
      "loss": 3.9014,
      "num_input_tokens_seen": 25523104,
      "step": 18840
    },
    {
      "epoch": 0.6031944177709494,
      "grad_norm": 18.5,
      "learning_rate": 2.726668357665017e-05,
      "loss": 4.4499,
      "num_input_tokens_seen": 25530272,
      "step": 18845
    },
    {
      "epoch": 0.6033544587414378,
      "grad_norm": 37.25,
      "learning_rate": 2.7256906360970808e-05,
      "loss": 4.9061,
      "num_input_tokens_seen": 25536704,
      "step": 18850
    },
    {
      "epoch": 0.6035144997119263,
      "grad_norm": 25.125,
      "learning_rate": 2.7247128797248117e-05,
      "loss": 4.6751,
      "num_input_tokens_seen": 25543504,
      "step": 18855
    },
    {
      "epoch": 0.6036745406824147,
      "grad_norm": 35.75,
      "learning_rate": 2.7237350886989925e-05,
      "loss": 4.8012,
      "num_input_tokens_seen": 25550752,
      "step": 18860
    },
    {
      "epoch": 0.6038345816529032,
      "grad_norm": 34.0,
      "learning_rate": 2.7227572631704107e-05,
      "loss": 4.0528,
      "num_input_tokens_seen": 25557504,
      "step": 18865
    },
    {
      "epoch": 0.6039946226233915,
      "grad_norm": 18.5,
      "learning_rate": 2.7217794032898596e-05,
      "loss": 4.5857,
      "num_input_tokens_seen": 25564272,
      "step": 18870
    },
    {
      "epoch": 0.60415466359388,
      "grad_norm": 30.75,
      "learning_rate": 2.7208015092081384e-05,
      "loss": 3.9289,
      "num_input_tokens_seen": 25570704,
      "step": 18875
    },
    {
      "epoch": 0.6043147045643685,
      "grad_norm": 17.25,
      "learning_rate": 2.719823581076049e-05,
      "loss": 4.3447,
      "num_input_tokens_seen": 25577424,
      "step": 18880
    },
    {
      "epoch": 0.6044747455348569,
      "grad_norm": 25.625,
      "learning_rate": 2.718845619044401e-05,
      "loss": 3.5212,
      "num_input_tokens_seen": 25584224,
      "step": 18885
    },
    {
      "epoch": 0.6046347865053454,
      "grad_norm": 24.625,
      "learning_rate": 2.7178676232640088e-05,
      "loss": 4.3103,
      "num_input_tokens_seen": 25591120,
      "step": 18890
    },
    {
      "epoch": 0.6047948274758338,
      "grad_norm": 50.25,
      "learning_rate": 2.716889593885691e-05,
      "loss": 3.7284,
      "num_input_tokens_seen": 25598032,
      "step": 18895
    },
    {
      "epoch": 0.6049548684463223,
      "grad_norm": 36.25,
      "learning_rate": 2.7159115310602716e-05,
      "loss": 4.5867,
      "num_input_tokens_seen": 25605360,
      "step": 18900
    },
    {
      "epoch": 0.6051149094168107,
      "grad_norm": 27.375,
      "learning_rate": 2.7149334349385814e-05,
      "loss": 4.0319,
      "num_input_tokens_seen": 25612128,
      "step": 18905
    },
    {
      "epoch": 0.6052749503872992,
      "grad_norm": 32.5,
      "learning_rate": 2.713955305671454e-05,
      "loss": 4.3204,
      "num_input_tokens_seen": 25618592,
      "step": 18910
    },
    {
      "epoch": 0.6054349913577876,
      "grad_norm": 64.5,
      "learning_rate": 2.71297714340973e-05,
      "loss": 5.6071,
      "num_input_tokens_seen": 25625280,
      "step": 18915
    },
    {
      "epoch": 0.605595032328276,
      "grad_norm": 47.25,
      "learning_rate": 2.7119989483042545e-05,
      "loss": 5.3055,
      "num_input_tokens_seen": 25631920,
      "step": 18920
    },
    {
      "epoch": 0.6057550732987644,
      "grad_norm": 27.125,
      "learning_rate": 2.7110207205058768e-05,
      "loss": 5.1671,
      "num_input_tokens_seen": 25638784,
      "step": 18925
    },
    {
      "epoch": 0.6059151142692529,
      "grad_norm": 35.5,
      "learning_rate": 2.7100424601654517e-05,
      "loss": 4.2465,
      "num_input_tokens_seen": 25645536,
      "step": 18930
    },
    {
      "epoch": 0.6060751552397414,
      "grad_norm": 20.75,
      "learning_rate": 2.7090641674338403e-05,
      "loss": 3.334,
      "num_input_tokens_seen": 25652400,
      "step": 18935
    },
    {
      "epoch": 0.6062351962102298,
      "grad_norm": 19.75,
      "learning_rate": 2.7080858424619072e-05,
      "loss": 4.2823,
      "num_input_tokens_seen": 25659248,
      "step": 18940
    },
    {
      "epoch": 0.6063952371807183,
      "grad_norm": 31.875,
      "learning_rate": 2.707107485400521e-05,
      "loss": 4.9956,
      "num_input_tokens_seen": 25666128,
      "step": 18945
    },
    {
      "epoch": 0.6065552781512067,
      "grad_norm": 28.375,
      "learning_rate": 2.7061290964005586e-05,
      "loss": 4.7646,
      "num_input_tokens_seen": 25672976,
      "step": 18950
    },
    {
      "epoch": 0.6067153191216952,
      "grad_norm": 35.5,
      "learning_rate": 2.7051506756129e-05,
      "loss": 4.4713,
      "num_input_tokens_seen": 25679552,
      "step": 18955
    },
    {
      "epoch": 0.6068753600921836,
      "grad_norm": 23.5,
      "learning_rate": 2.704172223188428e-05,
      "loss": 4.8142,
      "num_input_tokens_seen": 25686864,
      "step": 18960
    },
    {
      "epoch": 0.6070354010626721,
      "grad_norm": 63.75,
      "learning_rate": 2.7031937392780334e-05,
      "loss": 4.2548,
      "num_input_tokens_seen": 25693472,
      "step": 18965
    },
    {
      "epoch": 0.6071954420331604,
      "grad_norm": 22.5,
      "learning_rate": 2.702215224032611e-05,
      "loss": 5.1693,
      "num_input_tokens_seen": 25700976,
      "step": 18970
    },
    {
      "epoch": 0.6073554830036489,
      "grad_norm": 18.0,
      "learning_rate": 2.70123667760306e-05,
      "loss": 3.5753,
      "num_input_tokens_seen": 25707792,
      "step": 18975
    },
    {
      "epoch": 0.6075155239741373,
      "grad_norm": 16.25,
      "learning_rate": 2.7002581001402845e-05,
      "loss": 4.2508,
      "num_input_tokens_seen": 25714768,
      "step": 18980
    },
    {
      "epoch": 0.6076755649446258,
      "grad_norm": 20.0,
      "learning_rate": 2.6992794917951923e-05,
      "loss": 4.2972,
      "num_input_tokens_seen": 25721248,
      "step": 18985
    },
    {
      "epoch": 0.6078356059151143,
      "grad_norm": 27.875,
      "learning_rate": 2.6983008527187e-05,
      "loss": 4.1447,
      "num_input_tokens_seen": 25728320,
      "step": 18990
    },
    {
      "epoch": 0.6079956468856027,
      "grad_norm": 29.0,
      "learning_rate": 2.697322183061723e-05,
      "loss": 4.729,
      "num_input_tokens_seen": 25734704,
      "step": 18995
    },
    {
      "epoch": 0.6081556878560912,
      "grad_norm": 32.25,
      "learning_rate": 2.696343482975186e-05,
      "loss": 4.0165,
      "num_input_tokens_seen": 25741776,
      "step": 19000
    },
    {
      "epoch": 0.6081556878560912,
      "eval_loss": 4.492568492889404,
      "eval_runtime": 490.1834,
      "eval_samples_per_second": 28.328,
      "eval_steps_per_second": 14.164,
      "num_input_tokens_seen": 25741776,
      "step": 19000
    },
    {
      "epoch": 0.6083157288265796,
      "grad_norm": 39.0,
      "learning_rate": 2.695364752610016e-05,
      "loss": 3.546,
      "num_input_tokens_seen": 25748336,
      "step": 19005
    },
    {
      "epoch": 0.6084757697970681,
      "grad_norm": 21.75,
      "learning_rate": 2.6943859921171467e-05,
      "loss": 4.0876,
      "num_input_tokens_seen": 25755024,
      "step": 19010
    },
    {
      "epoch": 0.6086358107675565,
      "grad_norm": 28.25,
      "learning_rate": 2.6934072016475143e-05,
      "loss": 5.3501,
      "num_input_tokens_seen": 25762512,
      "step": 19015
    },
    {
      "epoch": 0.608795851738045,
      "grad_norm": 21.5,
      "learning_rate": 2.6924283813520606e-05,
      "loss": 4.4406,
      "num_input_tokens_seen": 25769808,
      "step": 19020
    },
    {
      "epoch": 0.6089558927085333,
      "grad_norm": 31.75,
      "learning_rate": 2.691449531381733e-05,
      "loss": 4.7447,
      "num_input_tokens_seen": 25776656,
      "step": 19025
    },
    {
      "epoch": 0.6091159336790218,
      "grad_norm": 28.75,
      "learning_rate": 2.6904706518874816e-05,
      "loss": 4.7866,
      "num_input_tokens_seen": 25783248,
      "step": 19030
    },
    {
      "epoch": 0.6092759746495102,
      "grad_norm": 22.75,
      "learning_rate": 2.6894917430202615e-05,
      "loss": 4.1068,
      "num_input_tokens_seen": 25789984,
      "step": 19035
    },
    {
      "epoch": 0.6094360156199987,
      "grad_norm": 18.25,
      "learning_rate": 2.6885128049310343e-05,
      "loss": 4.5761,
      "num_input_tokens_seen": 25796704,
      "step": 19040
    },
    {
      "epoch": 0.6095960565904872,
      "grad_norm": 24.5,
      "learning_rate": 2.687533837770762e-05,
      "loss": 3.8813,
      "num_input_tokens_seen": 25803472,
      "step": 19045
    },
    {
      "epoch": 0.6097560975609756,
      "grad_norm": 25.25,
      "learning_rate": 2.6865548416904162e-05,
      "loss": 4.2524,
      "num_input_tokens_seen": 25810384,
      "step": 19050
    },
    {
      "epoch": 0.6099161385314641,
      "grad_norm": 28.125,
      "learning_rate": 2.68557581684097e-05,
      "loss": 2.9722,
      "num_input_tokens_seen": 25817248,
      "step": 19055
    },
    {
      "epoch": 0.6100761795019525,
      "grad_norm": 31.75,
      "learning_rate": 2.6845967633733998e-05,
      "loss": 3.6143,
      "num_input_tokens_seen": 25824000,
      "step": 19060
    },
    {
      "epoch": 0.610236220472441,
      "grad_norm": 17.5,
      "learning_rate": 2.683617681438689e-05,
      "loss": 3.9302,
      "num_input_tokens_seen": 25830624,
      "step": 19065
    },
    {
      "epoch": 0.6103962614429294,
      "grad_norm": 28.875,
      "learning_rate": 2.682638571187825e-05,
      "loss": 4.5542,
      "num_input_tokens_seen": 25837488,
      "step": 19070
    },
    {
      "epoch": 0.6105563024134179,
      "grad_norm": 24.625,
      "learning_rate": 2.6816594327717976e-05,
      "loss": 4.2746,
      "num_input_tokens_seen": 25843984,
      "step": 19075
    },
    {
      "epoch": 0.6107163433839062,
      "grad_norm": 31.125,
      "learning_rate": 2.680680266341603e-05,
      "loss": 3.8191,
      "num_input_tokens_seen": 25850752,
      "step": 19080
    },
    {
      "epoch": 0.6108763843543947,
      "grad_norm": 52.75,
      "learning_rate": 2.67970107204824e-05,
      "loss": 4.6745,
      "num_input_tokens_seen": 25857296,
      "step": 19085
    },
    {
      "epoch": 0.6110364253248832,
      "grad_norm": 37.5,
      "learning_rate": 2.6787218500427142e-05,
      "loss": 5.1542,
      "num_input_tokens_seen": 25864368,
      "step": 19090
    },
    {
      "epoch": 0.6111964662953716,
      "grad_norm": 16.375,
      "learning_rate": 2.6777426004760332e-05,
      "loss": 4.1244,
      "num_input_tokens_seen": 25871744,
      "step": 19095
    },
    {
      "epoch": 0.6113565072658601,
      "grad_norm": 23.25,
      "learning_rate": 2.6767633234992094e-05,
      "loss": 4.0154,
      "num_input_tokens_seen": 25878992,
      "step": 19100
    },
    {
      "epoch": 0.6115165482363485,
      "grad_norm": 37.5,
      "learning_rate": 2.6757840192632598e-05,
      "loss": 4.1328,
      "num_input_tokens_seen": 25885872,
      "step": 19105
    },
    {
      "epoch": 0.611676589206837,
      "grad_norm": 35.25,
      "learning_rate": 2.6748046879192052e-05,
      "loss": 4.2282,
      "num_input_tokens_seen": 25892608,
      "step": 19110
    },
    {
      "epoch": 0.6118366301773254,
      "grad_norm": 28.875,
      "learning_rate": 2.673825329618071e-05,
      "loss": 4.8144,
      "num_input_tokens_seen": 25899344,
      "step": 19115
    },
    {
      "epoch": 0.6119966711478139,
      "grad_norm": 42.75,
      "learning_rate": 2.6728459445108866e-05,
      "loss": 4.2896,
      "num_input_tokens_seen": 25906400,
      "step": 19120
    },
    {
      "epoch": 0.6121567121183022,
      "grad_norm": 25.0,
      "learning_rate": 2.6718665327486854e-05,
      "loss": 4.0125,
      "num_input_tokens_seen": 25913472,
      "step": 19125
    },
    {
      "epoch": 0.6123167530887907,
      "grad_norm": 27.25,
      "learning_rate": 2.6708870944825048e-05,
      "loss": 5.276,
      "num_input_tokens_seen": 25920448,
      "step": 19130
    },
    {
      "epoch": 0.6124767940592791,
      "grad_norm": 28.875,
      "learning_rate": 2.6699076298633874e-05,
      "loss": 4.6395,
      "num_input_tokens_seen": 25927408,
      "step": 19135
    },
    {
      "epoch": 0.6126368350297676,
      "grad_norm": 33.25,
      "learning_rate": 2.6689281390423788e-05,
      "loss": 3.9592,
      "num_input_tokens_seen": 25934480,
      "step": 19140
    },
    {
      "epoch": 0.6127968760002561,
      "grad_norm": 30.75,
      "learning_rate": 2.667948622170527e-05,
      "loss": 6.1733,
      "num_input_tokens_seen": 25941424,
      "step": 19145
    },
    {
      "epoch": 0.6129569169707445,
      "grad_norm": 19.375,
      "learning_rate": 2.6669690793988873e-05,
      "loss": 4.2209,
      "num_input_tokens_seen": 25947968,
      "step": 19150
    },
    {
      "epoch": 0.613116957941233,
      "grad_norm": 17.25,
      "learning_rate": 2.665989510878518e-05,
      "loss": 3.4974,
      "num_input_tokens_seen": 25954816,
      "step": 19155
    },
    {
      "epoch": 0.6132769989117214,
      "grad_norm": 34.25,
      "learning_rate": 2.6650099167604793e-05,
      "loss": 4.2978,
      "num_input_tokens_seen": 25962128,
      "step": 19160
    },
    {
      "epoch": 0.6134370398822099,
      "grad_norm": 39.75,
      "learning_rate": 2.6640302971958376e-05,
      "loss": 4.1687,
      "num_input_tokens_seen": 25968352,
      "step": 19165
    },
    {
      "epoch": 0.6135970808526983,
      "grad_norm": 34.75,
      "learning_rate": 2.6630506523356635e-05,
      "loss": 4.2815,
      "num_input_tokens_seen": 25975104,
      "step": 19170
    },
    {
      "epoch": 0.6137571218231868,
      "grad_norm": 45.75,
      "learning_rate": 2.6620709823310297e-05,
      "loss": 4.4026,
      "num_input_tokens_seen": 25982816,
      "step": 19175
    },
    {
      "epoch": 0.6139171627936751,
      "grad_norm": 39.75,
      "learning_rate": 2.661091287333014e-05,
      "loss": 3.1859,
      "num_input_tokens_seen": 25989520,
      "step": 19180
    },
    {
      "epoch": 0.6140772037641636,
      "grad_norm": 35.5,
      "learning_rate": 2.660111567492696e-05,
      "loss": 4.3055,
      "num_input_tokens_seen": 25996560,
      "step": 19185
    },
    {
      "epoch": 0.614237244734652,
      "grad_norm": 20.625,
      "learning_rate": 2.6591318229611635e-05,
      "loss": 4.2177,
      "num_input_tokens_seen": 26003152,
      "step": 19190
    },
    {
      "epoch": 0.6143972857051405,
      "grad_norm": 26.75,
      "learning_rate": 2.6581520538895037e-05,
      "loss": 4.6302,
      "num_input_tokens_seen": 26010080,
      "step": 19195
    },
    {
      "epoch": 0.614557326675629,
      "grad_norm": 46.5,
      "learning_rate": 2.6571722604288102e-05,
      "loss": 4.9951,
      "num_input_tokens_seen": 26017088,
      "step": 19200
    },
    {
      "epoch": 0.614557326675629,
      "eval_loss": 4.478407382965088,
      "eval_runtime": 490.3593,
      "eval_samples_per_second": 28.318,
      "eval_steps_per_second": 14.159,
      "num_input_tokens_seen": 26017088,
      "step": 19200
    },
    {
      "epoch": 0.6147173676461174,
      "grad_norm": 38.25,
      "learning_rate": 2.656192442730179e-05,
      "loss": 4.0643,
      "num_input_tokens_seen": 26023952,
      "step": 19205
    },
    {
      "epoch": 0.6148774086166059,
      "grad_norm": 25.25,
      "learning_rate": 2.6552126009447098e-05,
      "loss": 3.9333,
      "num_input_tokens_seen": 26030848,
      "step": 19210
    },
    {
      "epoch": 0.6150374495870943,
      "grad_norm": 32.75,
      "learning_rate": 2.654232735223507e-05,
      "loss": 4.0659,
      "num_input_tokens_seen": 26037760,
      "step": 19215
    },
    {
      "epoch": 0.6151974905575828,
      "grad_norm": 33.25,
      "learning_rate": 2.6532528457176787e-05,
      "loss": 4.1896,
      "num_input_tokens_seen": 26044608,
      "step": 19220
    },
    {
      "epoch": 0.6153575315280712,
      "grad_norm": 31.625,
      "learning_rate": 2.6522729325783348e-05,
      "loss": 4.9028,
      "num_input_tokens_seen": 26051440,
      "step": 19225
    },
    {
      "epoch": 0.6155175724985597,
      "grad_norm": 29.75,
      "learning_rate": 2.6512929959565914e-05,
      "loss": 4.5477,
      "num_input_tokens_seen": 26057776,
      "step": 19230
    },
    {
      "epoch": 0.615677613469048,
      "grad_norm": 18.25,
      "learning_rate": 2.6503130360035673e-05,
      "loss": 4.9965,
      "num_input_tokens_seen": 26064368,
      "step": 19235
    },
    {
      "epoch": 0.6158376544395365,
      "grad_norm": 32.5,
      "learning_rate": 2.6493330528703835e-05,
      "loss": 5.1177,
      "num_input_tokens_seen": 26070800,
      "step": 19240
    },
    {
      "epoch": 0.6159976954100249,
      "grad_norm": 18.625,
      "learning_rate": 2.648353046708167e-05,
      "loss": 4.3937,
      "num_input_tokens_seen": 26077520,
      "step": 19245
    },
    {
      "epoch": 0.6161577363805134,
      "grad_norm": 54.5,
      "learning_rate": 2.647373017668046e-05,
      "loss": 5.1936,
      "num_input_tokens_seen": 26084000,
      "step": 19250
    },
    {
      "epoch": 0.6163177773510019,
      "grad_norm": 47.0,
      "learning_rate": 2.6463929659011537e-05,
      "loss": 5.1543,
      "num_input_tokens_seen": 26090832,
      "step": 19255
    },
    {
      "epoch": 0.6164778183214903,
      "grad_norm": 50.75,
      "learning_rate": 2.6454128915586262e-05,
      "loss": 4.0833,
      "num_input_tokens_seen": 26097440,
      "step": 19260
    },
    {
      "epoch": 0.6166378592919788,
      "grad_norm": 62.0,
      "learning_rate": 2.6444327947916036e-05,
      "loss": 4.6664,
      "num_input_tokens_seen": 26104288,
      "step": 19265
    },
    {
      "epoch": 0.6167979002624672,
      "grad_norm": 19.75,
      "learning_rate": 2.6434526757512292e-05,
      "loss": 4.7508,
      "num_input_tokens_seen": 26110864,
      "step": 19270
    },
    {
      "epoch": 0.6169579412329557,
      "grad_norm": 28.25,
      "learning_rate": 2.6424725345886486e-05,
      "loss": 4.259,
      "num_input_tokens_seen": 26117536,
      "step": 19275
    },
    {
      "epoch": 0.617117982203444,
      "grad_norm": 32.0,
      "learning_rate": 2.641492371455014e-05,
      "loss": 4.3163,
      "num_input_tokens_seen": 26124592,
      "step": 19280
    },
    {
      "epoch": 0.6172780231739325,
      "grad_norm": 29.125,
      "learning_rate": 2.640512186501477e-05,
      "loss": 4.3672,
      "num_input_tokens_seen": 26131008,
      "step": 19285
    },
    {
      "epoch": 0.6174380641444209,
      "grad_norm": 17.875,
      "learning_rate": 2.639531979879195e-05,
      "loss": 3.8607,
      "num_input_tokens_seen": 26137600,
      "step": 19290
    },
    {
      "epoch": 0.6175981051149094,
      "grad_norm": 23.625,
      "learning_rate": 2.638551751739328e-05,
      "loss": 4.8656,
      "num_input_tokens_seen": 26144064,
      "step": 19295
    },
    {
      "epoch": 0.6177581460853978,
      "grad_norm": 26.125,
      "learning_rate": 2.6375715022330404e-05,
      "loss": 4.1825,
      "num_input_tokens_seen": 26150528,
      "step": 19300
    },
    {
      "epoch": 0.6179181870558863,
      "grad_norm": 24.375,
      "learning_rate": 2.6365912315114976e-05,
      "loss": 3.6151,
      "num_input_tokens_seen": 26157680,
      "step": 19305
    },
    {
      "epoch": 0.6180782280263748,
      "grad_norm": 23.375,
      "learning_rate": 2.6356109397258704e-05,
      "loss": 4.2728,
      "num_input_tokens_seen": 26164816,
      "step": 19310
    },
    {
      "epoch": 0.6182382689968632,
      "grad_norm": 29.875,
      "learning_rate": 2.6346306270273325e-05,
      "loss": 4.1321,
      "num_input_tokens_seen": 26172224,
      "step": 19315
    },
    {
      "epoch": 0.6183983099673517,
      "grad_norm": 34.25,
      "learning_rate": 2.6336502935670608e-05,
      "loss": 4.6851,
      "num_input_tokens_seen": 26179168,
      "step": 19320
    },
    {
      "epoch": 0.6185583509378401,
      "grad_norm": 26.0,
      "learning_rate": 2.6326699394962333e-05,
      "loss": 4.2563,
      "num_input_tokens_seen": 26186336,
      "step": 19325
    },
    {
      "epoch": 0.6187183919083286,
      "grad_norm": 34.5,
      "learning_rate": 2.6316895649660334e-05,
      "loss": 4.7849,
      "num_input_tokens_seen": 26193120,
      "step": 19330
    },
    {
      "epoch": 0.6188784328788169,
      "grad_norm": 25.625,
      "learning_rate": 2.6307091701276486e-05,
      "loss": 3.3015,
      "num_input_tokens_seen": 26200368,
      "step": 19335
    },
    {
      "epoch": 0.6190384738493054,
      "grad_norm": 25.0,
      "learning_rate": 2.629728755132267e-05,
      "loss": 4.4619,
      "num_input_tokens_seen": 26207152,
      "step": 19340
    },
    {
      "epoch": 0.6191985148197938,
      "grad_norm": 26.25,
      "learning_rate": 2.628748320131081e-05,
      "loss": 5.2661,
      "num_input_tokens_seen": 26214272,
      "step": 19345
    },
    {
      "epoch": 0.6193585557902823,
      "grad_norm": 29.0,
      "learning_rate": 2.6277678652752856e-05,
      "loss": 4.01,
      "num_input_tokens_seen": 26221312,
      "step": 19350
    },
    {
      "epoch": 0.6195185967607708,
      "grad_norm": 42.25,
      "learning_rate": 2.6267873907160807e-05,
      "loss": 4.7592,
      "num_input_tokens_seen": 26227920,
      "step": 19355
    },
    {
      "epoch": 0.6196786377312592,
      "grad_norm": 17.125,
      "learning_rate": 2.6258068966046668e-05,
      "loss": 4.3226,
      "num_input_tokens_seen": 26234432,
      "step": 19360
    },
    {
      "epoch": 0.6198386787017477,
      "grad_norm": 34.0,
      "learning_rate": 2.6248263830922475e-05,
      "loss": 5.118,
      "num_input_tokens_seen": 26240752,
      "step": 19365
    },
    {
      "epoch": 0.6199987196722361,
      "grad_norm": 31.5,
      "learning_rate": 2.6238458503300318e-05,
      "loss": 4.1206,
      "num_input_tokens_seen": 26247056,
      "step": 19370
    },
    {
      "epoch": 0.6201587606427246,
      "grad_norm": 25.625,
      "learning_rate": 2.6228652984692292e-05,
      "loss": 4.3041,
      "num_input_tokens_seen": 26253568,
      "step": 19375
    },
    {
      "epoch": 0.620318801613213,
      "grad_norm": 25.0,
      "learning_rate": 2.621884727661054e-05,
      "loss": 3.9615,
      "num_input_tokens_seen": 26260112,
      "step": 19380
    },
    {
      "epoch": 0.6204788425837015,
      "grad_norm": 43.25,
      "learning_rate": 2.6209041380567222e-05,
      "loss": 4.7302,
      "num_input_tokens_seen": 26266656,
      "step": 19385
    },
    {
      "epoch": 0.6206388835541898,
      "grad_norm": 18.125,
      "learning_rate": 2.6199235298074527e-05,
      "loss": 4.1815,
      "num_input_tokens_seen": 26273280,
      "step": 19390
    },
    {
      "epoch": 0.6207989245246783,
      "grad_norm": 40.25,
      "learning_rate": 2.618942903064468e-05,
      "loss": 4.5413,
      "num_input_tokens_seen": 26279808,
      "step": 19395
    },
    {
      "epoch": 0.6209589654951667,
      "grad_norm": 24.375,
      "learning_rate": 2.6179622579789932e-05,
      "loss": 3.5959,
      "num_input_tokens_seen": 26286480,
      "step": 19400
    },
    {
      "epoch": 0.6209589654951667,
      "eval_loss": 4.4900360107421875,
      "eval_runtime": 490.2543,
      "eval_samples_per_second": 28.324,
      "eval_steps_per_second": 14.162,
      "num_input_tokens_seen": 26286480,
      "step": 19400
    },
    {
      "epoch": 0.6211190064656552,
      "grad_norm": 16.25,
      "learning_rate": 2.6169815947022553e-05,
      "loss": 4.5888,
      "num_input_tokens_seen": 26293328,
      "step": 19405
    },
    {
      "epoch": 0.6212790474361437,
      "grad_norm": 29.75,
      "learning_rate": 2.6160009133854853e-05,
      "loss": 4.1687,
      "num_input_tokens_seen": 26300496,
      "step": 19410
    },
    {
      "epoch": 0.6214390884066321,
      "grad_norm": 39.25,
      "learning_rate": 2.6150202141799168e-05,
      "loss": 3.7687,
      "num_input_tokens_seen": 26307200,
      "step": 19415
    },
    {
      "epoch": 0.6215991293771206,
      "grad_norm": 27.25,
      "learning_rate": 2.614039497236786e-05,
      "loss": 4.3224,
      "num_input_tokens_seen": 26314000,
      "step": 19420
    },
    {
      "epoch": 0.621759170347609,
      "grad_norm": 24.625,
      "learning_rate": 2.6130587627073315e-05,
      "loss": 4.1658,
      "num_input_tokens_seen": 26320800,
      "step": 19425
    },
    {
      "epoch": 0.6219192113180975,
      "grad_norm": 20.375,
      "learning_rate": 2.6120780107427956e-05,
      "loss": 3.8354,
      "num_input_tokens_seen": 26327712,
      "step": 19430
    },
    {
      "epoch": 0.6220792522885858,
      "grad_norm": 47.25,
      "learning_rate": 2.6110972414944214e-05,
      "loss": 3.6834,
      "num_input_tokens_seen": 26334432,
      "step": 19435
    },
    {
      "epoch": 0.6222392932590743,
      "grad_norm": 47.25,
      "learning_rate": 2.6101164551134565e-05,
      "loss": 4.9955,
      "num_input_tokens_seen": 26341056,
      "step": 19440
    },
    {
      "epoch": 0.6223993342295627,
      "grad_norm": 23.625,
      "learning_rate": 2.6091356517511505e-05,
      "loss": 4.1973,
      "num_input_tokens_seen": 26347408,
      "step": 19445
    },
    {
      "epoch": 0.6225593752000512,
      "grad_norm": 22.5,
      "learning_rate": 2.608154831558755e-05,
      "loss": 3.3955,
      "num_input_tokens_seen": 26354256,
      "step": 19450
    },
    {
      "epoch": 0.6227194161705396,
      "grad_norm": 34.25,
      "learning_rate": 2.607173994687526e-05,
      "loss": 4.7934,
      "num_input_tokens_seen": 26361024,
      "step": 19455
    },
    {
      "epoch": 0.6228794571410281,
      "grad_norm": 27.75,
      "learning_rate": 2.6061931412887196e-05,
      "loss": 4.6881,
      "num_input_tokens_seen": 26367488,
      "step": 19460
    },
    {
      "epoch": 0.6230394981115166,
      "grad_norm": 31.625,
      "learning_rate": 2.6052122715135973e-05,
      "loss": 4.1478,
      "num_input_tokens_seen": 26374416,
      "step": 19465
    },
    {
      "epoch": 0.623199539082005,
      "grad_norm": 29.25,
      "learning_rate": 2.60423138551342e-05,
      "loss": 4.0959,
      "num_input_tokens_seen": 26381264,
      "step": 19470
    },
    {
      "epoch": 0.6233595800524935,
      "grad_norm": 26.625,
      "learning_rate": 2.6032504834394527e-05,
      "loss": 4.5826,
      "num_input_tokens_seen": 26388096,
      "step": 19475
    },
    {
      "epoch": 0.6235196210229819,
      "grad_norm": 24.0,
      "learning_rate": 2.602269565442964e-05,
      "loss": 4.9463,
      "num_input_tokens_seen": 26394864,
      "step": 19480
    },
    {
      "epoch": 0.6236796619934704,
      "grad_norm": 25.25,
      "learning_rate": 2.6012886316752227e-05,
      "loss": 4.4617,
      "num_input_tokens_seen": 26401456,
      "step": 19485
    },
    {
      "epoch": 0.6238397029639587,
      "grad_norm": 23.25,
      "learning_rate": 2.6003076822875018e-05,
      "loss": 3.1805,
      "num_input_tokens_seen": 26408160,
      "step": 19490
    },
    {
      "epoch": 0.6239997439344472,
      "grad_norm": 47.0,
      "learning_rate": 2.5993267174310755e-05,
      "loss": 4.5383,
      "num_input_tokens_seen": 26414608,
      "step": 19495
    },
    {
      "epoch": 0.6241597849049356,
      "grad_norm": 24.0,
      "learning_rate": 2.5983457372572218e-05,
      "loss": 3.9353,
      "num_input_tokens_seen": 26421248,
      "step": 19500
    },
    {
      "epoch": 0.6243198258754241,
      "grad_norm": 37.25,
      "learning_rate": 2.597364741917219e-05,
      "loss": 4.1434,
      "num_input_tokens_seen": 26428192,
      "step": 19505
    },
    {
      "epoch": 0.6244798668459125,
      "grad_norm": 19.125,
      "learning_rate": 2.5963837315623492e-05,
      "loss": 4.7019,
      "num_input_tokens_seen": 26434384,
      "step": 19510
    },
    {
      "epoch": 0.624639907816401,
      "grad_norm": 25.125,
      "learning_rate": 2.595402706343897e-05,
      "loss": 3.59,
      "num_input_tokens_seen": 26441360,
      "step": 19515
    },
    {
      "epoch": 0.6247999487868895,
      "grad_norm": 23.375,
      "learning_rate": 2.594421666413148e-05,
      "loss": 3.5683,
      "num_input_tokens_seen": 26447664,
      "step": 19520
    },
    {
      "epoch": 0.6249599897573779,
      "grad_norm": 35.5,
      "learning_rate": 2.5934406119213928e-05,
      "loss": 4.1619,
      "num_input_tokens_seen": 26454992,
      "step": 19525
    },
    {
      "epoch": 0.6251200307278664,
      "grad_norm": 25.625,
      "learning_rate": 2.5924595430199193e-05,
      "loss": 4.082,
      "num_input_tokens_seen": 26461920,
      "step": 19530
    },
    {
      "epoch": 0.6252800716983548,
      "grad_norm": 48.25,
      "learning_rate": 2.5914784598600238e-05,
      "loss": 5.0876,
      "num_input_tokens_seen": 26468608,
      "step": 19535
    },
    {
      "epoch": 0.6254401126688433,
      "grad_norm": 34.75,
      "learning_rate": 2.5904973625930002e-05,
      "loss": 4.2259,
      "num_input_tokens_seen": 26476064,
      "step": 19540
    },
    {
      "epoch": 0.6256001536393316,
      "grad_norm": 39.25,
      "learning_rate": 2.5895162513701456e-05,
      "loss": 4.389,
      "num_input_tokens_seen": 26482560,
      "step": 19545
    },
    {
      "epoch": 0.6257601946098201,
      "grad_norm": 40.0,
      "learning_rate": 2.5885351263427593e-05,
      "loss": 3.8679,
      "num_input_tokens_seen": 26489584,
      "step": 19550
    },
    {
      "epoch": 0.6259202355803085,
      "grad_norm": 20.0,
      "learning_rate": 2.5875539876621448e-05,
      "loss": 4.0948,
      "num_input_tokens_seen": 26496480,
      "step": 19555
    },
    {
      "epoch": 0.626080276550797,
      "grad_norm": 16.75,
      "learning_rate": 2.586572835479605e-05,
      "loss": 3.9922,
      "num_input_tokens_seen": 26502944,
      "step": 19560
    },
    {
      "epoch": 0.6262403175212854,
      "grad_norm": 33.0,
      "learning_rate": 2.585591669946446e-05,
      "loss": 3.4076,
      "num_input_tokens_seen": 26509776,
      "step": 19565
    },
    {
      "epoch": 0.6264003584917739,
      "grad_norm": 27.5,
      "learning_rate": 2.5846104912139756e-05,
      "loss": 4.606,
      "num_input_tokens_seen": 26516736,
      "step": 19570
    },
    {
      "epoch": 0.6265603994622624,
      "grad_norm": 19.875,
      "learning_rate": 2.583629299433505e-05,
      "loss": 3.6727,
      "num_input_tokens_seen": 26523408,
      "step": 19575
    },
    {
      "epoch": 0.6267204404327508,
      "grad_norm": 30.0,
      "learning_rate": 2.582648094756345e-05,
      "loss": 4.7163,
      "num_input_tokens_seen": 26529776,
      "step": 19580
    },
    {
      "epoch": 0.6268804814032393,
      "grad_norm": 17.0,
      "learning_rate": 2.5816668773338098e-05,
      "loss": 4.3409,
      "num_input_tokens_seen": 26536528,
      "step": 19585
    },
    {
      "epoch": 0.6270405223737276,
      "grad_norm": 40.5,
      "learning_rate": 2.580685647317216e-05,
      "loss": 4.5772,
      "num_input_tokens_seen": 26543520,
      "step": 19590
    },
    {
      "epoch": 0.6272005633442161,
      "grad_norm": 48.5,
      "learning_rate": 2.5797044048578818e-05,
      "loss": 3.7843,
      "num_input_tokens_seen": 26550352,
      "step": 19595
    },
    {
      "epoch": 0.6273606043147045,
      "grad_norm": 34.0,
      "learning_rate": 2.5787231501071262e-05,
      "loss": 4.5629,
      "num_input_tokens_seen": 26557200,
      "step": 19600
    },
    {
      "epoch": 0.6273606043147045,
      "eval_loss": 4.48704719543457,
      "eval_runtime": 489.7578,
      "eval_samples_per_second": 28.353,
      "eval_steps_per_second": 14.176,
      "num_input_tokens_seen": 26557200,
      "step": 19600
    },
    {
      "epoch": 0.627520645285193,
      "grad_norm": 34.0,
      "learning_rate": 2.577741883216272e-05,
      "loss": 4.6467,
      "num_input_tokens_seen": 26563824,
      "step": 19605
    },
    {
      "epoch": 0.6276806862556814,
      "grad_norm": 27.625,
      "learning_rate": 2.576760604336642e-05,
      "loss": 4.7734,
      "num_input_tokens_seen": 26570384,
      "step": 19610
    },
    {
      "epoch": 0.6278407272261699,
      "grad_norm": 26.0,
      "learning_rate": 2.575779313619563e-05,
      "loss": 4.4607,
      "num_input_tokens_seen": 26576816,
      "step": 19615
    },
    {
      "epoch": 0.6280007681966584,
      "grad_norm": 46.75,
      "learning_rate": 2.5747980112163605e-05,
      "loss": 4.4754,
      "num_input_tokens_seen": 26583360,
      "step": 19620
    },
    {
      "epoch": 0.6281608091671468,
      "grad_norm": 31.5,
      "learning_rate": 2.5738166972783656e-05,
      "loss": 3.971,
      "num_input_tokens_seen": 26590160,
      "step": 19625
    },
    {
      "epoch": 0.6283208501376353,
      "grad_norm": 29.75,
      "learning_rate": 2.5728353719569075e-05,
      "loss": 5.7564,
      "num_input_tokens_seen": 26597168,
      "step": 19630
    },
    {
      "epoch": 0.6284808911081237,
      "grad_norm": 22.75,
      "learning_rate": 2.57185403540332e-05,
      "loss": 4.3172,
      "num_input_tokens_seen": 26603856,
      "step": 19635
    },
    {
      "epoch": 0.6286409320786122,
      "grad_norm": 22.5,
      "learning_rate": 2.5708726877689375e-05,
      "loss": 4.741,
      "num_input_tokens_seen": 26610368,
      "step": 19640
    },
    {
      "epoch": 0.6288009730491005,
      "grad_norm": 22.875,
      "learning_rate": 2.5698913292050964e-05,
      "loss": 4.2314,
      "num_input_tokens_seen": 26616928,
      "step": 19645
    },
    {
      "epoch": 0.628961014019589,
      "grad_norm": 35.0,
      "learning_rate": 2.568909959863133e-05,
      "loss": 4.2819,
      "num_input_tokens_seen": 26623600,
      "step": 19650
    },
    {
      "epoch": 0.6291210549900774,
      "grad_norm": 33.5,
      "learning_rate": 2.5679285798943887e-05,
      "loss": 4.5734,
      "num_input_tokens_seen": 26630560,
      "step": 19655
    },
    {
      "epoch": 0.6292810959605659,
      "grad_norm": 64.5,
      "learning_rate": 2.5669471894502035e-05,
      "loss": 5.7896,
      "num_input_tokens_seen": 26636704,
      "step": 19660
    },
    {
      "epoch": 0.6294411369310543,
      "grad_norm": 29.5,
      "learning_rate": 2.56596578868192e-05,
      "loss": 4.9265,
      "num_input_tokens_seen": 26643344,
      "step": 19665
    },
    {
      "epoch": 0.6296011779015428,
      "grad_norm": 40.25,
      "learning_rate": 2.564984377740883e-05,
      "loss": 6.0277,
      "num_input_tokens_seen": 26650400,
      "step": 19670
    },
    {
      "epoch": 0.6297612188720313,
      "grad_norm": 27.625,
      "learning_rate": 2.564002956778438e-05,
      "loss": 3.7003,
      "num_input_tokens_seen": 26657104,
      "step": 19675
    },
    {
      "epoch": 0.6299212598425197,
      "grad_norm": 27.5,
      "learning_rate": 2.563021525945934e-05,
      "loss": 3.7817,
      "num_input_tokens_seen": 26663552,
      "step": 19680
    },
    {
      "epoch": 0.6300813008130082,
      "grad_norm": 15.25,
      "learning_rate": 2.562040085394718e-05,
      "loss": 4.4856,
      "num_input_tokens_seen": 26670048,
      "step": 19685
    },
    {
      "epoch": 0.6302413417834966,
      "grad_norm": 31.625,
      "learning_rate": 2.56105863527614e-05,
      "loss": 3.9178,
      "num_input_tokens_seen": 26676848,
      "step": 19690
    },
    {
      "epoch": 0.630401382753985,
      "grad_norm": 24.125,
      "learning_rate": 2.5600771757415548e-05,
      "loss": 4.3568,
      "num_input_tokens_seen": 26683568,
      "step": 19695
    },
    {
      "epoch": 0.6305614237244734,
      "grad_norm": 22.625,
      "learning_rate": 2.5590957069423134e-05,
      "loss": 3.7257,
      "num_input_tokens_seen": 26690160,
      "step": 19700
    },
    {
      "epoch": 0.6307214646949619,
      "grad_norm": 30.75,
      "learning_rate": 2.5581142290297716e-05,
      "loss": 3.4489,
      "num_input_tokens_seen": 26696896,
      "step": 19705
    },
    {
      "epoch": 0.6308815056654503,
      "grad_norm": 19.5,
      "learning_rate": 2.557132742155285e-05,
      "loss": 3.8549,
      "num_input_tokens_seen": 26703520,
      "step": 19710
    },
    {
      "epoch": 0.6310415466359388,
      "grad_norm": 13.375,
      "learning_rate": 2.556151246470212e-05,
      "loss": 4.0886,
      "num_input_tokens_seen": 26710640,
      "step": 19715
    },
    {
      "epoch": 0.6312015876064272,
      "grad_norm": 20.75,
      "learning_rate": 2.5551697421259114e-05,
      "loss": 3.982,
      "num_input_tokens_seen": 26717536,
      "step": 19720
    },
    {
      "epoch": 0.6313616285769157,
      "grad_norm": 36.25,
      "learning_rate": 2.554188229273743e-05,
      "loss": 4.3457,
      "num_input_tokens_seen": 26724368,
      "step": 19725
    },
    {
      "epoch": 0.6315216695474042,
      "grad_norm": 27.875,
      "learning_rate": 2.5532067080650678e-05,
      "loss": 4.5733,
      "num_input_tokens_seen": 26731168,
      "step": 19730
    },
    {
      "epoch": 0.6316817105178926,
      "grad_norm": 24.0,
      "learning_rate": 2.55222517865125e-05,
      "loss": 4.8652,
      "num_input_tokens_seen": 26738160,
      "step": 19735
    },
    {
      "epoch": 0.6318417514883811,
      "grad_norm": 12.3125,
      "learning_rate": 2.5512436411836538e-05,
      "loss": 2.7559,
      "num_input_tokens_seen": 26745840,
      "step": 19740
    },
    {
      "epoch": 0.6320017924588694,
      "grad_norm": 31.5,
      "learning_rate": 2.5502620958136443e-05,
      "loss": 4.4285,
      "num_input_tokens_seen": 26752448,
      "step": 19745
    },
    {
      "epoch": 0.6321618334293579,
      "grad_norm": 44.5,
      "learning_rate": 2.5492805426925874e-05,
      "loss": 4.6568,
      "num_input_tokens_seen": 26759088,
      "step": 19750
    },
    {
      "epoch": 0.6323218743998463,
      "grad_norm": 40.5,
      "learning_rate": 2.5482989819718523e-05,
      "loss": 4.4337,
      "num_input_tokens_seen": 26766288,
      "step": 19755
    },
    {
      "epoch": 0.6324819153703348,
      "grad_norm": 20.125,
      "learning_rate": 2.5473174138028065e-05,
      "loss": 3.7159,
      "num_input_tokens_seen": 26773024,
      "step": 19760
    },
    {
      "epoch": 0.6326419563408232,
      "grad_norm": 27.5,
      "learning_rate": 2.5463358383368212e-05,
      "loss": 5.0612,
      "num_input_tokens_seen": 26779552,
      "step": 19765
    },
    {
      "epoch": 0.6328019973113117,
      "grad_norm": 27.75,
      "learning_rate": 2.545354255725267e-05,
      "loss": 3.52,
      "num_input_tokens_seen": 26786272,
      "step": 19770
    },
    {
      "epoch": 0.6329620382818001,
      "grad_norm": 28.75,
      "learning_rate": 2.5443726661195165e-05,
      "loss": 4.8438,
      "num_input_tokens_seen": 26792992,
      "step": 19775
    },
    {
      "epoch": 0.6331220792522886,
      "grad_norm": 19.875,
      "learning_rate": 2.543391069670944e-05,
      "loss": 4.2775,
      "num_input_tokens_seen": 26799920,
      "step": 19780
    },
    {
      "epoch": 0.6332821202227771,
      "grad_norm": 38.0,
      "learning_rate": 2.5424094665309228e-05,
      "loss": 4.9284,
      "num_input_tokens_seen": 26806800,
      "step": 19785
    },
    {
      "epoch": 0.6334421611932655,
      "grad_norm": 35.5,
      "learning_rate": 2.5414278568508292e-05,
      "loss": 4.7304,
      "num_input_tokens_seen": 26813968,
      "step": 19790
    },
    {
      "epoch": 0.633602202163754,
      "grad_norm": 22.625,
      "learning_rate": 2.540446240782039e-05,
      "loss": 3.1681,
      "num_input_tokens_seen": 26820928,
      "step": 19795
    },
    {
      "epoch": 0.6337622431342423,
      "grad_norm": 35.0,
      "learning_rate": 2.5394646184759307e-05,
      "loss": 4.0744,
      "num_input_tokens_seen": 26827696,
      "step": 19800
    },
    {
      "epoch": 0.6337622431342423,
      "eval_loss": 4.497011184692383,
      "eval_runtime": 489.6003,
      "eval_samples_per_second": 28.362,
      "eval_steps_per_second": 14.181,
      "num_input_tokens_seen": 26827696,
      "step": 19800
    },
    {
      "epoch": 0.6339222841047308,
      "grad_norm": 41.0,
      "learning_rate": 2.538482990083882e-05,
      "loss": 4.942,
      "num_input_tokens_seen": 26834288,
      "step": 19805
    },
    {
      "epoch": 0.6340823250752192,
      "grad_norm": 32.25,
      "learning_rate": 2.5375013557572725e-05,
      "loss": 4.1314,
      "num_input_tokens_seen": 26840544,
      "step": 19810
    },
    {
      "epoch": 0.6342423660457077,
      "grad_norm": 17.125,
      "learning_rate": 2.536519715647483e-05,
      "loss": 3.3579,
      "num_input_tokens_seen": 26847456,
      "step": 19815
    },
    {
      "epoch": 0.6344024070161961,
      "grad_norm": 42.0,
      "learning_rate": 2.535538069905894e-05,
      "loss": 4.1846,
      "num_input_tokens_seen": 26854080,
      "step": 19820
    },
    {
      "epoch": 0.6345624479866846,
      "grad_norm": 27.875,
      "learning_rate": 2.534556418683888e-05,
      "loss": 4.0624,
      "num_input_tokens_seen": 26860480,
      "step": 19825
    },
    {
      "epoch": 0.634722488957173,
      "grad_norm": 28.75,
      "learning_rate": 2.5335747621328486e-05,
      "loss": 3.5442,
      "num_input_tokens_seen": 26867584,
      "step": 19830
    },
    {
      "epoch": 0.6348825299276615,
      "grad_norm": 22.125,
      "learning_rate": 2.5325931004041586e-05,
      "loss": 4.0804,
      "num_input_tokens_seen": 26874144,
      "step": 19835
    },
    {
      "epoch": 0.63504257089815,
      "grad_norm": 20.625,
      "learning_rate": 2.5316114336492032e-05,
      "loss": 4.8123,
      "num_input_tokens_seen": 26880528,
      "step": 19840
    },
    {
      "epoch": 0.6352026118686384,
      "grad_norm": 19.125,
      "learning_rate": 2.530629762019367e-05,
      "loss": 3.8692,
      "num_input_tokens_seen": 26887584,
      "step": 19845
    },
    {
      "epoch": 0.6353626528391269,
      "grad_norm": 35.5,
      "learning_rate": 2.5296480856660364e-05,
      "loss": 4.254,
      "num_input_tokens_seen": 26894144,
      "step": 19850
    },
    {
      "epoch": 0.6355226938096152,
      "grad_norm": 39.75,
      "learning_rate": 2.528666404740599e-05,
      "loss": 4.8495,
      "num_input_tokens_seen": 26900944,
      "step": 19855
    },
    {
      "epoch": 0.6356827347801037,
      "grad_norm": 47.75,
      "learning_rate": 2.527684719394442e-05,
      "loss": 4.1581,
      "num_input_tokens_seen": 26907456,
      "step": 19860
    },
    {
      "epoch": 0.6358427757505921,
      "grad_norm": 38.5,
      "learning_rate": 2.526703029778953e-05,
      "loss": 3.7633,
      "num_input_tokens_seen": 26914976,
      "step": 19865
    },
    {
      "epoch": 0.6360028167210806,
      "grad_norm": 32.75,
      "learning_rate": 2.5257213360455208e-05,
      "loss": 4.4858,
      "num_input_tokens_seen": 26921392,
      "step": 19870
    },
    {
      "epoch": 0.636162857691569,
      "grad_norm": 26.125,
      "learning_rate": 2.5247396383455353e-05,
      "loss": 5.9976,
      "num_input_tokens_seen": 26927760,
      "step": 19875
    },
    {
      "epoch": 0.6363228986620575,
      "grad_norm": 18.25,
      "learning_rate": 2.523757936830387e-05,
      "loss": 4.3799,
      "num_input_tokens_seen": 26934560,
      "step": 19880
    },
    {
      "epoch": 0.636482939632546,
      "grad_norm": 19.875,
      "learning_rate": 2.5227762316514662e-05,
      "loss": 4.4403,
      "num_input_tokens_seen": 26941136,
      "step": 19885
    },
    {
      "epoch": 0.6366429806030344,
      "grad_norm": 19.875,
      "learning_rate": 2.5217945229601648e-05,
      "loss": 5.1491,
      "num_input_tokens_seen": 26947808,
      "step": 19890
    },
    {
      "epoch": 0.6368030215735229,
      "grad_norm": 28.375,
      "learning_rate": 2.5208128109078738e-05,
      "loss": 4.0944,
      "num_input_tokens_seen": 26954608,
      "step": 19895
    },
    {
      "epoch": 0.6369630625440112,
      "grad_norm": 26.875,
      "learning_rate": 2.5198310956459853e-05,
      "loss": 4.9284,
      "num_input_tokens_seen": 26961280,
      "step": 19900
    },
    {
      "epoch": 0.6371231035144997,
      "grad_norm": 56.0,
      "learning_rate": 2.518849377325893e-05,
      "loss": 5.1222,
      "num_input_tokens_seen": 26968192,
      "step": 19905
    },
    {
      "epoch": 0.6372831444849881,
      "grad_norm": 98.5,
      "learning_rate": 2.51786765609899e-05,
      "loss": 4.9522,
      "num_input_tokens_seen": 26975024,
      "step": 19910
    },
    {
      "epoch": 0.6374431854554766,
      "grad_norm": 17.0,
      "learning_rate": 2.5168859321166694e-05,
      "loss": 4.0077,
      "num_input_tokens_seen": 26981616,
      "step": 19915
    },
    {
      "epoch": 0.637603226425965,
      "grad_norm": 24.25,
      "learning_rate": 2.515904205530326e-05,
      "loss": 3.8158,
      "num_input_tokens_seen": 26988592,
      "step": 19920
    },
    {
      "epoch": 0.6377632673964535,
      "grad_norm": 28.5,
      "learning_rate": 2.514922476491355e-05,
      "loss": 4.3323,
      "num_input_tokens_seen": 26995264,
      "step": 19925
    },
    {
      "epoch": 0.6379233083669419,
      "grad_norm": 36.0,
      "learning_rate": 2.51394074515115e-05,
      "loss": 3.8243,
      "num_input_tokens_seen": 27002128,
      "step": 19930
    },
    {
      "epoch": 0.6380833493374304,
      "grad_norm": 24.5,
      "learning_rate": 2.5129590116611067e-05,
      "loss": 3.9887,
      "num_input_tokens_seen": 27009120,
      "step": 19935
    },
    {
      "epoch": 0.6382433903079189,
      "grad_norm": 21.25,
      "learning_rate": 2.5119772761726212e-05,
      "loss": 5.0444,
      "num_input_tokens_seen": 27015824,
      "step": 19940
    },
    {
      "epoch": 0.6384034312784073,
      "grad_norm": 22.25,
      "learning_rate": 2.5109955388370893e-05,
      "loss": 2.7733,
      "num_input_tokens_seen": 27022992,
      "step": 19945
    },
    {
      "epoch": 0.6385634722488958,
      "grad_norm": 26.5,
      "learning_rate": 2.510013799805907e-05,
      "loss": 3.8318,
      "num_input_tokens_seen": 27029712,
      "step": 19950
    },
    {
      "epoch": 0.6387235132193841,
      "grad_norm": 32.25,
      "learning_rate": 2.5090320592304706e-05,
      "loss": 4.6313,
      "num_input_tokens_seen": 27036784,
      "step": 19955
    },
    {
      "epoch": 0.6388835541898726,
      "grad_norm": 30.25,
      "learning_rate": 2.5080503172621777e-05,
      "loss": 4.191,
      "num_input_tokens_seen": 27043392,
      "step": 19960
    },
    {
      "epoch": 0.639043595160361,
      "grad_norm": 30.5,
      "learning_rate": 2.5070685740524246e-05,
      "loss": 3.9424,
      "num_input_tokens_seen": 27050112,
      "step": 19965
    },
    {
      "epoch": 0.6392036361308495,
      "grad_norm": 53.25,
      "learning_rate": 2.5060868297526084e-05,
      "loss": 4.1264,
      "num_input_tokens_seen": 27056752,
      "step": 19970
    },
    {
      "epoch": 0.6393636771013379,
      "grad_norm": 33.75,
      "learning_rate": 2.5051050845141267e-05,
      "loss": 4.4723,
      "num_input_tokens_seen": 27064176,
      "step": 19975
    },
    {
      "epoch": 0.6395237180718264,
      "grad_norm": 35.0,
      "learning_rate": 2.5041233384883765e-05,
      "loss": 4.7935,
      "num_input_tokens_seen": 27071008,
      "step": 19980
    },
    {
      "epoch": 0.6396837590423148,
      "grad_norm": 28.375,
      "learning_rate": 2.5031415918267564e-05,
      "loss": 4.1089,
      "num_input_tokens_seen": 27077744,
      "step": 19985
    },
    {
      "epoch": 0.6398438000128033,
      "grad_norm": 28.75,
      "learning_rate": 2.5021598446806626e-05,
      "loss": 4.536,
      "num_input_tokens_seen": 27084304,
      "step": 19990
    },
    {
      "epoch": 0.6400038409832918,
      "grad_norm": 40.5,
      "learning_rate": 2.5011780972014937e-05,
      "loss": 4.5801,
      "num_input_tokens_seen": 27091248,
      "step": 19995
    },
    {
      "epoch": 0.6401638819537802,
      "grad_norm": 13.9375,
      "learning_rate": 2.5001963495406478e-05,
      "loss": 3.8765,
      "num_input_tokens_seen": 27098112,
      "step": 20000
    },
    {
      "epoch": 0.6401638819537802,
      "eval_loss": 4.486845970153809,
      "eval_runtime": 489.9391,
      "eval_samples_per_second": 28.342,
      "eval_steps_per_second": 14.171,
      "num_input_tokens_seen": 27098112,
      "step": 20000
    },
    {
      "epoch": 0.6403239229242687,
      "grad_norm": 28.125,
      "learning_rate": 2.499214601849522e-05,
      "loss": 3.9299,
      "num_input_tokens_seen": 27104848,
      "step": 20005
    },
    {
      "epoch": 0.640483963894757,
      "grad_norm": 24.625,
      "learning_rate": 2.4982328542795148e-05,
      "loss": 4.4514,
      "num_input_tokens_seen": 27111344,
      "step": 20010
    },
    {
      "epoch": 0.6406440048652455,
      "grad_norm": 34.0,
      "learning_rate": 2.497251106982024e-05,
      "loss": 4.2652,
      "num_input_tokens_seen": 27118112,
      "step": 20015
    },
    {
      "epoch": 0.6408040458357339,
      "grad_norm": 35.5,
      "learning_rate": 2.4962693601084458e-05,
      "loss": 4.0966,
      "num_input_tokens_seen": 27124624,
      "step": 20020
    },
    {
      "epoch": 0.6409640868062224,
      "grad_norm": 29.625,
      "learning_rate": 2.4952876138101794e-05,
      "loss": 3.6119,
      "num_input_tokens_seen": 27131312,
      "step": 20025
    },
    {
      "epoch": 0.6411241277767108,
      "grad_norm": 24.875,
      "learning_rate": 2.4943058682386233e-05,
      "loss": 4.2405,
      "num_input_tokens_seen": 27138160,
      "step": 20030
    },
    {
      "epoch": 0.6412841687471993,
      "grad_norm": 35.0,
      "learning_rate": 2.493324123545173e-05,
      "loss": 4.5934,
      "num_input_tokens_seen": 27144784,
      "step": 20035
    },
    {
      "epoch": 0.6414442097176877,
      "grad_norm": 28.375,
      "learning_rate": 2.4923423798812272e-05,
      "loss": 5.4396,
      "num_input_tokens_seen": 27151264,
      "step": 20040
    },
    {
      "epoch": 0.6416042506881762,
      "grad_norm": 14.0625,
      "learning_rate": 2.4913606373981825e-05,
      "loss": 3.4023,
      "num_input_tokens_seen": 27158112,
      "step": 20045
    },
    {
      "epoch": 0.6417642916586647,
      "grad_norm": 33.75,
      "learning_rate": 2.4903788962474357e-05,
      "loss": 4.5731,
      "num_input_tokens_seen": 27164752,
      "step": 20050
    },
    {
      "epoch": 0.641924332629153,
      "grad_norm": 32.75,
      "learning_rate": 2.489397156580385e-05,
      "loss": 4.1385,
      "num_input_tokens_seen": 27171424,
      "step": 20055
    },
    {
      "epoch": 0.6420843735996415,
      "grad_norm": 18.375,
      "learning_rate": 2.4884154185484246e-05,
      "loss": 4.1279,
      "num_input_tokens_seen": 27177920,
      "step": 20060
    },
    {
      "epoch": 0.6422444145701299,
      "grad_norm": 32.5,
      "learning_rate": 2.4874336823029526e-05,
      "loss": 5.3196,
      "num_input_tokens_seen": 27185088,
      "step": 20065
    },
    {
      "epoch": 0.6424044555406184,
      "grad_norm": 47.25,
      "learning_rate": 2.4864519479953656e-05,
      "loss": 4.6356,
      "num_input_tokens_seen": 27191824,
      "step": 20070
    },
    {
      "epoch": 0.6425644965111068,
      "grad_norm": 17.75,
      "learning_rate": 2.485470215777058e-05,
      "loss": 4.3882,
      "num_input_tokens_seen": 27198384,
      "step": 20075
    },
    {
      "epoch": 0.6427245374815953,
      "grad_norm": 48.25,
      "learning_rate": 2.4844884857994258e-05,
      "loss": 4.1775,
      "num_input_tokens_seen": 27205216,
      "step": 20080
    },
    {
      "epoch": 0.6428845784520837,
      "grad_norm": 24.625,
      "learning_rate": 2.4835067582138638e-05,
      "loss": 4.4676,
      "num_input_tokens_seen": 27212240,
      "step": 20085
    },
    {
      "epoch": 0.6430446194225722,
      "grad_norm": 11.875,
      "learning_rate": 2.4825250331717666e-05,
      "loss": 5.0826,
      "num_input_tokens_seen": 27219184,
      "step": 20090
    },
    {
      "epoch": 0.6432046603930607,
      "grad_norm": 24.875,
      "learning_rate": 2.4815433108245298e-05,
      "loss": 4.8161,
      "num_input_tokens_seen": 27225952,
      "step": 20095
    },
    {
      "epoch": 0.6433647013635491,
      "grad_norm": 32.75,
      "learning_rate": 2.4805615913235456e-05,
      "loss": 4.8506,
      "num_input_tokens_seen": 27232720,
      "step": 20100
    },
    {
      "epoch": 0.6435247423340376,
      "grad_norm": 28.75,
      "learning_rate": 2.479579874820208e-05,
      "loss": 5.056,
      "num_input_tokens_seen": 27239424,
      "step": 20105
    },
    {
      "epoch": 0.6436847833045259,
      "grad_norm": 36.75,
      "learning_rate": 2.4785981614659115e-05,
      "loss": 4.3995,
      "num_input_tokens_seen": 27245904,
      "step": 20110
    },
    {
      "epoch": 0.6438448242750144,
      "grad_norm": 39.25,
      "learning_rate": 2.477616451412047e-05,
      "loss": 4.3976,
      "num_input_tokens_seen": 27253504,
      "step": 20115
    },
    {
      "epoch": 0.6440048652455028,
      "grad_norm": 35.25,
      "learning_rate": 2.476634744810007e-05,
      "loss": 4.4836,
      "num_input_tokens_seen": 27260192,
      "step": 20120
    },
    {
      "epoch": 0.6441649062159913,
      "grad_norm": 30.875,
      "learning_rate": 2.475653041811183e-05,
      "loss": 4.1575,
      "num_input_tokens_seen": 27266864,
      "step": 20125
    },
    {
      "epoch": 0.6443249471864797,
      "grad_norm": 21.125,
      "learning_rate": 2.4746713425669652e-05,
      "loss": 3.4745,
      "num_input_tokens_seen": 27274096,
      "step": 20130
    },
    {
      "epoch": 0.6444849881569682,
      "grad_norm": 20.75,
      "learning_rate": 2.4736896472287458e-05,
      "loss": 4.6588,
      "num_input_tokens_seen": 27280784,
      "step": 20135
    },
    {
      "epoch": 0.6446450291274566,
      "grad_norm": 18.25,
      "learning_rate": 2.4727079559479124e-05,
      "loss": 3.1253,
      "num_input_tokens_seen": 27288208,
      "step": 20140
    },
    {
      "epoch": 0.6448050700979451,
      "grad_norm": 47.0,
      "learning_rate": 2.4717262688758557e-05,
      "loss": 4.7557,
      "num_input_tokens_seen": 27295200,
      "step": 20145
    },
    {
      "epoch": 0.6449651110684336,
      "grad_norm": 40.75,
      "learning_rate": 2.4707445861639637e-05,
      "loss": 4.5329,
      "num_input_tokens_seen": 27301968,
      "step": 20150
    },
    {
      "epoch": 0.645125152038922,
      "grad_norm": 46.25,
      "learning_rate": 2.4697629079636244e-05,
      "loss": 5.1081,
      "num_input_tokens_seen": 27308544,
      "step": 20155
    },
    {
      "epoch": 0.6452851930094105,
      "grad_norm": 43.0,
      "learning_rate": 2.4687812344262244e-05,
      "loss": 4.8151,
      "num_input_tokens_seen": 27315792,
      "step": 20160
    },
    {
      "epoch": 0.6454452339798988,
      "grad_norm": 16.0,
      "learning_rate": 2.46779956570315e-05,
      "loss": 3.5827,
      "num_input_tokens_seen": 27322848,
      "step": 20165
    },
    {
      "epoch": 0.6456052749503873,
      "grad_norm": 30.0,
      "learning_rate": 2.466817901945787e-05,
      "loss": 4.0748,
      "num_input_tokens_seen": 27329296,
      "step": 20170
    },
    {
      "epoch": 0.6457653159208757,
      "grad_norm": 35.75,
      "learning_rate": 2.4658362433055217e-05,
      "loss": 4.7791,
      "num_input_tokens_seen": 27335712,
      "step": 20175
    },
    {
      "epoch": 0.6459253568913642,
      "grad_norm": 17.0,
      "learning_rate": 2.4648545899337356e-05,
      "loss": 3.9463,
      "num_input_tokens_seen": 27342288,
      "step": 20180
    },
    {
      "epoch": 0.6460853978618526,
      "grad_norm": 30.625,
      "learning_rate": 2.4638729419818143e-05,
      "loss": 4.1265,
      "num_input_tokens_seen": 27348848,
      "step": 20185
    },
    {
      "epoch": 0.6462454388323411,
      "grad_norm": 21.875,
      "learning_rate": 2.46289129960114e-05,
      "loss": 4.2915,
      "num_input_tokens_seen": 27355696,
      "step": 20190
    },
    {
      "epoch": 0.6464054798028295,
      "grad_norm": 27.25,
      "learning_rate": 2.4619096629430924e-05,
      "loss": 4.5646,
      "num_input_tokens_seen": 27363056,
      "step": 20195
    },
    {
      "epoch": 0.646565520773318,
      "grad_norm": 31.125,
      "learning_rate": 2.4609280321590543e-05,
      "loss": 4.1175,
      "num_input_tokens_seen": 27369984,
      "step": 20200
    },
    {
      "epoch": 0.646565520773318,
      "eval_loss": 4.498813629150391,
      "eval_runtime": 489.701,
      "eval_samples_per_second": 28.356,
      "eval_steps_per_second": 14.178,
      "num_input_tokens_seen": 27369984,
      "step": 20200
    },
    {
      "epoch": 0.6467255617438065,
      "grad_norm": 27.0,
      "learning_rate": 2.4599464074004037e-05,
      "loss": 4.6931,
      "num_input_tokens_seen": 27376736,
      "step": 20205
    },
    {
      "epoch": 0.6468856027142948,
      "grad_norm": 23.875,
      "learning_rate": 2.4589647888185204e-05,
      "loss": 4.6939,
      "num_input_tokens_seen": 27383952,
      "step": 20210
    },
    {
      "epoch": 0.6470456436847833,
      "grad_norm": 21.875,
      "learning_rate": 2.4579831765647836e-05,
      "loss": 4.5555,
      "num_input_tokens_seen": 27390496,
      "step": 20215
    },
    {
      "epoch": 0.6472056846552717,
      "grad_norm": 21.0,
      "learning_rate": 2.4570015707905676e-05,
      "loss": 4.6989,
      "num_input_tokens_seen": 27397456,
      "step": 20220
    },
    {
      "epoch": 0.6473657256257602,
      "grad_norm": 23.5,
      "learning_rate": 2.4560199716472508e-05,
      "loss": 4.3586,
      "num_input_tokens_seen": 27404144,
      "step": 20225
    },
    {
      "epoch": 0.6475257665962486,
      "grad_norm": 27.5,
      "learning_rate": 2.455038379286207e-05,
      "loss": 5.4169,
      "num_input_tokens_seen": 27411024,
      "step": 20230
    },
    {
      "epoch": 0.6476858075667371,
      "grad_norm": 44.75,
      "learning_rate": 2.4540567938588095e-05,
      "loss": 5.0422,
      "num_input_tokens_seen": 27417824,
      "step": 20235
    },
    {
      "epoch": 0.6478458485372255,
      "grad_norm": 38.75,
      "learning_rate": 2.4530752155164328e-05,
      "loss": 4.6108,
      "num_input_tokens_seen": 27424896,
      "step": 20240
    },
    {
      "epoch": 0.648005889507714,
      "grad_norm": 21.875,
      "learning_rate": 2.4520936444104463e-05,
      "loss": 4.085,
      "num_input_tokens_seen": 27431360,
      "step": 20245
    },
    {
      "epoch": 0.6481659304782024,
      "grad_norm": 25.125,
      "learning_rate": 2.4511120806922218e-05,
      "loss": 3.903,
      "num_input_tokens_seen": 27438416,
      "step": 20250
    },
    {
      "epoch": 0.6483259714486909,
      "grad_norm": 22.875,
      "learning_rate": 2.45013052451313e-05,
      "loss": 4.268,
      "num_input_tokens_seen": 27445248,
      "step": 20255
    },
    {
      "epoch": 0.6484860124191794,
      "grad_norm": 32.25,
      "learning_rate": 2.4491489760245376e-05,
      "loss": 5.1302,
      "num_input_tokens_seen": 27452128,
      "step": 20260
    },
    {
      "epoch": 0.6486460533896677,
      "grad_norm": 27.375,
      "learning_rate": 2.4481674353778115e-05,
      "loss": 4.1165,
      "num_input_tokens_seen": 27458880,
      "step": 20265
    },
    {
      "epoch": 0.6488060943601562,
      "grad_norm": 40.0,
      "learning_rate": 2.447185902724319e-05,
      "loss": 3.755,
      "num_input_tokens_seen": 27465808,
      "step": 20270
    },
    {
      "epoch": 0.6489661353306446,
      "grad_norm": 42.0,
      "learning_rate": 2.4462043782154233e-05,
      "loss": 4.5112,
      "num_input_tokens_seen": 27472256,
      "step": 20275
    },
    {
      "epoch": 0.6491261763011331,
      "grad_norm": 19.125,
      "learning_rate": 2.4452228620024895e-05,
      "loss": 3.7719,
      "num_input_tokens_seen": 27478736,
      "step": 20280
    },
    {
      "epoch": 0.6492862172716215,
      "grad_norm": 18.75,
      "learning_rate": 2.4442413542368776e-05,
      "loss": 4.6731,
      "num_input_tokens_seen": 27485344,
      "step": 20285
    },
    {
      "epoch": 0.64944625824211,
      "grad_norm": 24.75,
      "learning_rate": 2.4432598550699502e-05,
      "loss": 4.0872,
      "num_input_tokens_seen": 27492336,
      "step": 20290
    },
    {
      "epoch": 0.6496062992125984,
      "grad_norm": 23.625,
      "learning_rate": 2.4422783646530663e-05,
      "loss": 4.1832,
      "num_input_tokens_seen": 27499072,
      "step": 20295
    },
    {
      "epoch": 0.6497663401830869,
      "grad_norm": 28.5,
      "learning_rate": 2.441296883137584e-05,
      "loss": 4.7051,
      "num_input_tokens_seen": 27505728,
      "step": 20300
    },
    {
      "epoch": 0.6499263811535753,
      "grad_norm": 18.25,
      "learning_rate": 2.4403154106748592e-05,
      "loss": 4.6135,
      "num_input_tokens_seen": 27512432,
      "step": 20305
    },
    {
      "epoch": 0.6500864221240638,
      "grad_norm": 30.375,
      "learning_rate": 2.4393339474162494e-05,
      "loss": 3.9498,
      "num_input_tokens_seen": 27519328,
      "step": 20310
    },
    {
      "epoch": 0.6502464630945523,
      "grad_norm": 31.625,
      "learning_rate": 2.4383524935131062e-05,
      "loss": 4.2541,
      "num_input_tokens_seen": 27526448,
      "step": 20315
    },
    {
      "epoch": 0.6504065040650406,
      "grad_norm": 28.75,
      "learning_rate": 2.437371049116784e-05,
      "loss": 4.3636,
      "num_input_tokens_seen": 27533376,
      "step": 20320
    },
    {
      "epoch": 0.6505665450355291,
      "grad_norm": 25.0,
      "learning_rate": 2.436389614378632e-05,
      "loss": 4.6231,
      "num_input_tokens_seen": 27539968,
      "step": 20325
    },
    {
      "epoch": 0.6507265860060175,
      "grad_norm": 11.75,
      "learning_rate": 2.435408189450002e-05,
      "loss": 3.7856,
      "num_input_tokens_seen": 27546624,
      "step": 20330
    },
    {
      "epoch": 0.650886626976506,
      "grad_norm": 68.0,
      "learning_rate": 2.4344267744822406e-05,
      "loss": 5.3375,
      "num_input_tokens_seen": 27553408,
      "step": 20335
    },
    {
      "epoch": 0.6510466679469944,
      "grad_norm": 20.25,
      "learning_rate": 2.4334453696266944e-05,
      "loss": 3.79,
      "num_input_tokens_seen": 27559888,
      "step": 20340
    },
    {
      "epoch": 0.6512067089174829,
      "grad_norm": 17.125,
      "learning_rate": 2.432463975034708e-05,
      "loss": 4.0935,
      "num_input_tokens_seen": 27566368,
      "step": 20345
    },
    {
      "epoch": 0.6513667498879713,
      "grad_norm": 28.875,
      "learning_rate": 2.4314825908576265e-05,
      "loss": 4.1668,
      "num_input_tokens_seen": 27572624,
      "step": 20350
    },
    {
      "epoch": 0.6515267908584598,
      "grad_norm": 27.0,
      "learning_rate": 2.4305012172467897e-05,
      "loss": 4.8891,
      "num_input_tokens_seen": 27579328,
      "step": 20355
    },
    {
      "epoch": 0.6516868318289483,
      "grad_norm": 35.0,
      "learning_rate": 2.4295198543535393e-05,
      "loss": 4.0925,
      "num_input_tokens_seen": 27586064,
      "step": 20360
    },
    {
      "epoch": 0.6518468727994366,
      "grad_norm": 32.75,
      "learning_rate": 2.4285385023292124e-05,
      "loss": 3.698,
      "num_input_tokens_seen": 27592832,
      "step": 20365
    },
    {
      "epoch": 0.6520069137699251,
      "grad_norm": 46.5,
      "learning_rate": 2.427557161325147e-05,
      "loss": 4.8773,
      "num_input_tokens_seen": 27599728,
      "step": 20370
    },
    {
      "epoch": 0.6521669547404135,
      "grad_norm": 22.75,
      "learning_rate": 2.4265758314926778e-05,
      "loss": 4.4109,
      "num_input_tokens_seen": 27606640,
      "step": 20375
    },
    {
      "epoch": 0.652326995710902,
      "grad_norm": 23.75,
      "learning_rate": 2.4255945129831373e-05,
      "loss": 4.8002,
      "num_input_tokens_seen": 27613360,
      "step": 20380
    },
    {
      "epoch": 0.6524870366813904,
      "grad_norm": 41.5,
      "learning_rate": 2.4246132059478578e-05,
      "loss": 4.2736,
      "num_input_tokens_seen": 27620048,
      "step": 20385
    },
    {
      "epoch": 0.6526470776518789,
      "grad_norm": 34.5,
      "learning_rate": 2.4236319105381706e-05,
      "loss": 4.7593,
      "num_input_tokens_seen": 27627680,
      "step": 20390
    },
    {
      "epoch": 0.6528071186223673,
      "grad_norm": 57.75,
      "learning_rate": 2.422650626905401e-05,
      "loss": 4.6654,
      "num_input_tokens_seen": 27634560,
      "step": 20395
    },
    {
      "epoch": 0.6529671595928558,
      "grad_norm": 39.5,
      "learning_rate": 2.4216693552008785e-05,
      "loss": 5.3047,
      "num_input_tokens_seen": 27640768,
      "step": 20400
    },
    {
      "epoch": 0.6529671595928558,
      "eval_loss": 4.470890522003174,
      "eval_runtime": 490.2192,
      "eval_samples_per_second": 28.326,
      "eval_steps_per_second": 14.163,
      "num_input_tokens_seen": 27640768,
      "step": 20400
    },
    {
      "epoch": 0.6531272005633442,
      "grad_norm": 28.125,
      "learning_rate": 2.4206880955759247e-05,
      "loss": 5.5776,
      "num_input_tokens_seen": 27647424,
      "step": 20405
    },
    {
      "epoch": 0.6532872415338327,
      "grad_norm": 25.75,
      "learning_rate": 2.419706848181863e-05,
      "loss": 4.3699,
      "num_input_tokens_seen": 27654304,
      "step": 20410
    },
    {
      "epoch": 0.6534472825043212,
      "grad_norm": 14.6875,
      "learning_rate": 2.4187256131700153e-05,
      "loss": 4.6863,
      "num_input_tokens_seen": 27661328,
      "step": 20415
    },
    {
      "epoch": 0.6536073234748095,
      "grad_norm": 34.25,
      "learning_rate": 2.4177443906916985e-05,
      "loss": 4.6109,
      "num_input_tokens_seen": 27669104,
      "step": 20420
    },
    {
      "epoch": 0.653767364445298,
      "grad_norm": 31.75,
      "learning_rate": 2.4167631808982303e-05,
      "loss": 4.2934,
      "num_input_tokens_seen": 27676496,
      "step": 20425
    },
    {
      "epoch": 0.6539274054157864,
      "grad_norm": 29.875,
      "learning_rate": 2.4157819839409264e-05,
      "loss": 4.4907,
      "num_input_tokens_seen": 27683200,
      "step": 20430
    },
    {
      "epoch": 0.6540874463862749,
      "grad_norm": 35.0,
      "learning_rate": 2.414800799971098e-05,
      "loss": 4.9613,
      "num_input_tokens_seen": 27690272,
      "step": 20435
    },
    {
      "epoch": 0.6542474873567633,
      "grad_norm": 26.875,
      "learning_rate": 2.4138196291400582e-05,
      "loss": 4.5333,
      "num_input_tokens_seen": 27696816,
      "step": 20440
    },
    {
      "epoch": 0.6544075283272518,
      "grad_norm": 42.0,
      "learning_rate": 2.412838471599114e-05,
      "loss": 4.0232,
      "num_input_tokens_seen": 27703856,
      "step": 20445
    },
    {
      "epoch": 0.6545675692977402,
      "grad_norm": 36.0,
      "learning_rate": 2.411857327499572e-05,
      "loss": 4.392,
      "num_input_tokens_seen": 27710528,
      "step": 20450
    },
    {
      "epoch": 0.6547276102682287,
      "grad_norm": 44.25,
      "learning_rate": 2.410876196992739e-05,
      "loss": 4.4242,
      "num_input_tokens_seen": 27717056,
      "step": 20455
    },
    {
      "epoch": 0.6548876512387171,
      "grad_norm": 39.25,
      "learning_rate": 2.4098950802299156e-05,
      "loss": 5.2596,
      "num_input_tokens_seen": 27723856,
      "step": 20460
    },
    {
      "epoch": 0.6550476922092056,
      "grad_norm": 24.0,
      "learning_rate": 2.4089139773624027e-05,
      "loss": 4.5648,
      "num_input_tokens_seen": 27730416,
      "step": 20465
    },
    {
      "epoch": 0.655207733179694,
      "grad_norm": 25.875,
      "learning_rate": 2.4079328885415007e-05,
      "loss": 4.3519,
      "num_input_tokens_seen": 27737328,
      "step": 20470
    },
    {
      "epoch": 0.6553677741501824,
      "grad_norm": 26.5,
      "learning_rate": 2.4069518139185036e-05,
      "loss": 3.9828,
      "num_input_tokens_seen": 27743792,
      "step": 20475
    },
    {
      "epoch": 0.6555278151206709,
      "grad_norm": 31.375,
      "learning_rate": 2.405970753644706e-05,
      "loss": 4.4205,
      "num_input_tokens_seen": 27750400,
      "step": 20480
    },
    {
      "epoch": 0.6556878560911593,
      "grad_norm": 34.0,
      "learning_rate": 2.4049897078714e-05,
      "loss": 4.5511,
      "num_input_tokens_seen": 27757360,
      "step": 20485
    },
    {
      "epoch": 0.6558478970616478,
      "grad_norm": 24.875,
      "learning_rate": 2.404008676749874e-05,
      "loss": 4.6054,
      "num_input_tokens_seen": 27763840,
      "step": 20490
    },
    {
      "epoch": 0.6560079380321362,
      "grad_norm": 22.0,
      "learning_rate": 2.403027660431418e-05,
      "loss": 4.5038,
      "num_input_tokens_seen": 27770336,
      "step": 20495
    },
    {
      "epoch": 0.6561679790026247,
      "grad_norm": 26.375,
      "learning_rate": 2.402046659067314e-05,
      "loss": 5.0474,
      "num_input_tokens_seen": 27776960,
      "step": 20500
    },
    {
      "epoch": 0.6563280199731131,
      "grad_norm": 22.375,
      "learning_rate": 2.401065672808847e-05,
      "loss": 3.32,
      "num_input_tokens_seen": 27783472,
      "step": 20505
    },
    {
      "epoch": 0.6564880609436016,
      "grad_norm": 29.5,
      "learning_rate": 2.400084701807296e-05,
      "loss": 3.5848,
      "num_input_tokens_seen": 27789984,
      "step": 20510
    },
    {
      "epoch": 0.65664810191409,
      "grad_norm": 33.0,
      "learning_rate": 2.39910374621394e-05,
      "loss": 4.3595,
      "num_input_tokens_seen": 27796576,
      "step": 20515
    },
    {
      "epoch": 0.6568081428845784,
      "grad_norm": 24.25,
      "learning_rate": 2.3981228061800544e-05,
      "loss": 4.2404,
      "num_input_tokens_seen": 27802960,
      "step": 20520
    },
    {
      "epoch": 0.656968183855067,
      "grad_norm": 23.875,
      "learning_rate": 2.3971418818569115e-05,
      "loss": 4.75,
      "num_input_tokens_seen": 27809680,
      "step": 20525
    },
    {
      "epoch": 0.6571282248255553,
      "grad_norm": 28.25,
      "learning_rate": 2.3961609733957832e-05,
      "loss": 3.7925,
      "num_input_tokens_seen": 27816272,
      "step": 20530
    },
    {
      "epoch": 0.6572882657960438,
      "grad_norm": 30.625,
      "learning_rate": 2.395180080947939e-05,
      "loss": 3.5241,
      "num_input_tokens_seen": 27822992,
      "step": 20535
    },
    {
      "epoch": 0.6574483067665322,
      "grad_norm": 26.625,
      "learning_rate": 2.394199204664642e-05,
      "loss": 4.3123,
      "num_input_tokens_seen": 27829568,
      "step": 20540
    },
    {
      "epoch": 0.6576083477370207,
      "grad_norm": 22.25,
      "learning_rate": 2.3932183446971583e-05,
      "loss": 4.6374,
      "num_input_tokens_seen": 27836448,
      "step": 20545
    },
    {
      "epoch": 0.6577683887075091,
      "grad_norm": 36.5,
      "learning_rate": 2.3922375011967473e-05,
      "loss": 4.7935,
      "num_input_tokens_seen": 27843504,
      "step": 20550
    },
    {
      "epoch": 0.6579284296779976,
      "grad_norm": 29.25,
      "learning_rate": 2.3912566743146676e-05,
      "loss": 4.5706,
      "num_input_tokens_seen": 27850240,
      "step": 20555
    },
    {
      "epoch": 0.658088470648486,
      "grad_norm": 19.5,
      "learning_rate": 2.390275864202176e-05,
      "loss": 4.1221,
      "num_input_tokens_seen": 27856912,
      "step": 20560
    },
    {
      "epoch": 0.6582485116189745,
      "grad_norm": 39.5,
      "learning_rate": 2.3892950710105243e-05,
      "loss": 3.892,
      "num_input_tokens_seen": 27863504,
      "step": 20565
    },
    {
      "epoch": 0.6584085525894628,
      "grad_norm": 29.5,
      "learning_rate": 2.3883142948909635e-05,
      "loss": 4.6041,
      "num_input_tokens_seen": 27870096,
      "step": 20570
    },
    {
      "epoch": 0.6585685935599513,
      "grad_norm": 17.875,
      "learning_rate": 2.3873335359947433e-05,
      "loss": 4.4362,
      "num_input_tokens_seen": 27876656,
      "step": 20575
    },
    {
      "epoch": 0.6587286345304398,
      "grad_norm": 24.125,
      "learning_rate": 2.3863527944731066e-05,
      "loss": 4.0426,
      "num_input_tokens_seen": 27883392,
      "step": 20580
    },
    {
      "epoch": 0.6588886755009282,
      "grad_norm": 41.25,
      "learning_rate": 2.385372070477298e-05,
      "loss": 4.5007,
      "num_input_tokens_seen": 27890032,
      "step": 20585
    },
    {
      "epoch": 0.6590487164714167,
      "grad_norm": 44.75,
      "learning_rate": 2.384391364158556e-05,
      "loss": 4.4126,
      "num_input_tokens_seen": 27896976,
      "step": 20590
    },
    {
      "epoch": 0.6592087574419051,
      "grad_norm": 32.25,
      "learning_rate": 2.3834106756681185e-05,
      "loss": 4.035,
      "num_input_tokens_seen": 27903600,
      "step": 20595
    },
    {
      "epoch": 0.6593687984123936,
      "grad_norm": 25.625,
      "learning_rate": 2.3824300051572206e-05,
      "loss": 4.0005,
      "num_input_tokens_seen": 27910480,
      "step": 20600
    },
    {
      "epoch": 0.6593687984123936,
      "eval_loss": 4.489149570465088,
      "eval_runtime": 490.1838,
      "eval_samples_per_second": 28.328,
      "eval_steps_per_second": 14.164,
      "num_input_tokens_seen": 27910480,
      "step": 20600
    },
    {
      "epoch": 0.659528839382882,
      "grad_norm": 30.25,
      "learning_rate": 2.3814493527770923e-05,
      "loss": 4.8134,
      "num_input_tokens_seen": 27917056,
      "step": 20605
    },
    {
      "epoch": 0.6596888803533705,
      "grad_norm": 38.0,
      "learning_rate": 2.3804687186789637e-05,
      "loss": 3.3758,
      "num_input_tokens_seen": 27924048,
      "step": 20610
    },
    {
      "epoch": 0.6598489213238589,
      "grad_norm": 44.25,
      "learning_rate": 2.379488103014062e-05,
      "loss": 3.9824,
      "num_input_tokens_seen": 27930736,
      "step": 20615
    },
    {
      "epoch": 0.6600089622943474,
      "grad_norm": 20.25,
      "learning_rate": 2.3785075059336086e-05,
      "loss": 3.7035,
      "num_input_tokens_seen": 27937856,
      "step": 20620
    },
    {
      "epoch": 0.6601690032648359,
      "grad_norm": 40.75,
      "learning_rate": 2.3775269275888248e-05,
      "loss": 4.7106,
      "num_input_tokens_seen": 27944816,
      "step": 20625
    },
    {
      "epoch": 0.6603290442353242,
      "grad_norm": 28.875,
      "learning_rate": 2.3765463681309274e-05,
      "loss": 5.115,
      "num_input_tokens_seen": 27951504,
      "step": 20630
    },
    {
      "epoch": 0.6604890852058127,
      "grad_norm": 38.25,
      "learning_rate": 2.3755658277111313e-05,
      "loss": 4.7237,
      "num_input_tokens_seen": 27958160,
      "step": 20635
    },
    {
      "epoch": 0.6606491261763011,
      "grad_norm": 35.5,
      "learning_rate": 2.374585306480649e-05,
      "loss": 5.4064,
      "num_input_tokens_seen": 27964880,
      "step": 20640
    },
    {
      "epoch": 0.6608091671467896,
      "grad_norm": 34.0,
      "learning_rate": 2.3736048045906877e-05,
      "loss": 4.7055,
      "num_input_tokens_seen": 27971424,
      "step": 20645
    },
    {
      "epoch": 0.660969208117278,
      "grad_norm": 36.0,
      "learning_rate": 2.372624322192454e-05,
      "loss": 4.5432,
      "num_input_tokens_seen": 27977632,
      "step": 20650
    },
    {
      "epoch": 0.6611292490877665,
      "grad_norm": 34.25,
      "learning_rate": 2.3716438594371516e-05,
      "loss": 4.4653,
      "num_input_tokens_seen": 27984512,
      "step": 20655
    },
    {
      "epoch": 0.6612892900582549,
      "grad_norm": 21.125,
      "learning_rate": 2.3706634164759784e-05,
      "loss": 4.3277,
      "num_input_tokens_seen": 27991184,
      "step": 20660
    },
    {
      "epoch": 0.6614493310287434,
      "grad_norm": 23.5,
      "learning_rate": 2.3696829934601323e-05,
      "loss": 4.8435,
      "num_input_tokens_seen": 27997840,
      "step": 20665
    },
    {
      "epoch": 0.6616093719992318,
      "grad_norm": 21.375,
      "learning_rate": 2.3687025905408053e-05,
      "loss": 3.5216,
      "num_input_tokens_seen": 28004832,
      "step": 20670
    },
    {
      "epoch": 0.6617694129697202,
      "grad_norm": 33.5,
      "learning_rate": 2.3677222078691886e-05,
      "loss": 4.9462,
      "num_input_tokens_seen": 28011344,
      "step": 20675
    },
    {
      "epoch": 0.6619294539402087,
      "grad_norm": 25.0,
      "learning_rate": 2.366741845596471e-05,
      "loss": 4.7009,
      "num_input_tokens_seen": 28018112,
      "step": 20680
    },
    {
      "epoch": 0.6620894949106971,
      "grad_norm": 29.0,
      "learning_rate": 2.3657615038738343e-05,
      "loss": 4.8271,
      "num_input_tokens_seen": 28024976,
      "step": 20685
    },
    {
      "epoch": 0.6622495358811856,
      "grad_norm": 27.0,
      "learning_rate": 2.3647811828524614e-05,
      "loss": 4.3242,
      "num_input_tokens_seen": 28031632,
      "step": 20690
    },
    {
      "epoch": 0.662409576851674,
      "grad_norm": 41.0,
      "learning_rate": 2.363800882683529e-05,
      "loss": 4.5723,
      "num_input_tokens_seen": 28038464,
      "step": 20695
    },
    {
      "epoch": 0.6625696178221625,
      "grad_norm": 29.375,
      "learning_rate": 2.3628206035182125e-05,
      "loss": 4.0399,
      "num_input_tokens_seen": 28044928,
      "step": 20700
    },
    {
      "epoch": 0.6627296587926509,
      "grad_norm": 19.0,
      "learning_rate": 2.361840345507683e-05,
      "loss": 4.0119,
      "num_input_tokens_seen": 28051552,
      "step": 20705
    },
    {
      "epoch": 0.6628896997631394,
      "grad_norm": 28.0,
      "learning_rate": 2.3608601088031073e-05,
      "loss": 4.083,
      "num_input_tokens_seen": 28058560,
      "step": 20710
    },
    {
      "epoch": 0.6630497407336278,
      "grad_norm": 42.5,
      "learning_rate": 2.3598798935556516e-05,
      "loss": 4.5337,
      "num_input_tokens_seen": 28065536,
      "step": 20715
    },
    {
      "epoch": 0.6632097817041163,
      "grad_norm": 73.0,
      "learning_rate": 2.3588996999164784e-05,
      "loss": 3.8027,
      "num_input_tokens_seen": 28072512,
      "step": 20720
    },
    {
      "epoch": 0.6633698226746046,
      "grad_norm": 22.5,
      "learning_rate": 2.3579195280367434e-05,
      "loss": 4.088,
      "num_input_tokens_seen": 28079360,
      "step": 20725
    },
    {
      "epoch": 0.6635298636450931,
      "grad_norm": 31.0,
      "learning_rate": 2.356939378067603e-05,
      "loss": 3.6265,
      "num_input_tokens_seen": 28086304,
      "step": 20730
    },
    {
      "epoch": 0.6636899046155816,
      "grad_norm": 41.75,
      "learning_rate": 2.3559592501602092e-05,
      "loss": 3.9395,
      "num_input_tokens_seen": 28092832,
      "step": 20735
    },
    {
      "epoch": 0.66384994558607,
      "grad_norm": 25.0,
      "learning_rate": 2.3549791444657076e-05,
      "loss": 3.9322,
      "num_input_tokens_seen": 28099328,
      "step": 20740
    },
    {
      "epoch": 0.6640099865565585,
      "grad_norm": 27.875,
      "learning_rate": 2.353999061135246e-05,
      "loss": 4.4263,
      "num_input_tokens_seen": 28106192,
      "step": 20745
    },
    {
      "epoch": 0.6641700275270469,
      "grad_norm": 46.75,
      "learning_rate": 2.3530190003199626e-05,
      "loss": 4.5338,
      "num_input_tokens_seen": 28112928,
      "step": 20750
    },
    {
      "epoch": 0.6643300684975354,
      "grad_norm": 30.25,
      "learning_rate": 2.3520389621709965e-05,
      "loss": 4.6199,
      "num_input_tokens_seen": 28119968,
      "step": 20755
    },
    {
      "epoch": 0.6644901094680238,
      "grad_norm": 30.75,
      "learning_rate": 2.351058946839483e-05,
      "loss": 4.5746,
      "num_input_tokens_seen": 28126800,
      "step": 20760
    },
    {
      "epoch": 0.6646501504385123,
      "grad_norm": 39.75,
      "learning_rate": 2.350078954476551e-05,
      "loss": 3.8685,
      "num_input_tokens_seen": 28133040,
      "step": 20765
    },
    {
      "epoch": 0.6648101914090007,
      "grad_norm": 28.25,
      "learning_rate": 2.3490989852333272e-05,
      "loss": 3.4555,
      "num_input_tokens_seen": 28140016,
      "step": 20770
    },
    {
      "epoch": 0.6649702323794892,
      "grad_norm": 30.5,
      "learning_rate": 2.3481190392609377e-05,
      "loss": 3.748,
      "num_input_tokens_seen": 28146512,
      "step": 20775
    },
    {
      "epoch": 0.6651302733499775,
      "grad_norm": 32.5,
      "learning_rate": 2.3471391167105e-05,
      "loss": 3.3569,
      "num_input_tokens_seen": 28153648,
      "step": 20780
    },
    {
      "epoch": 0.665290314320466,
      "grad_norm": 23.5,
      "learning_rate": 2.3461592177331325e-05,
      "loss": 4.5724,
      "num_input_tokens_seen": 28160304,
      "step": 20785
    },
    {
      "epoch": 0.6654503552909545,
      "grad_norm": 30.875,
      "learning_rate": 2.345179342479946e-05,
      "loss": 3.781,
      "num_input_tokens_seen": 28167008,
      "step": 20790
    },
    {
      "epoch": 0.6656103962614429,
      "grad_norm": 22.25,
      "learning_rate": 2.3441994911020503e-05,
      "loss": 4.0536,
      "num_input_tokens_seen": 28173936,
      "step": 20795
    },
    {
      "epoch": 0.6657704372319314,
      "grad_norm": 22.0,
      "learning_rate": 2.3432196637505522e-05,
      "loss": 4.4139,
      "num_input_tokens_seen": 28180240,
      "step": 20800
    },
    {
      "epoch": 0.6657704372319314,
      "eval_loss": 4.480555057525635,
      "eval_runtime": 489.6897,
      "eval_samples_per_second": 28.357,
      "eval_steps_per_second": 14.178,
      "num_input_tokens_seen": 28180240,
      "step": 20800
    },
    {
      "epoch": 0.6659304782024198,
      "grad_norm": 20.0,
      "learning_rate": 2.3422398605765515e-05,
      "loss": 4.2598,
      "num_input_tokens_seen": 28186832,
      "step": 20805
    },
    {
      "epoch": 0.6660905191729083,
      "grad_norm": 23.5,
      "learning_rate": 2.3412600817311462e-05,
      "loss": 4.2564,
      "num_input_tokens_seen": 28194368,
      "step": 20810
    },
    {
      "epoch": 0.6662505601433967,
      "grad_norm": 29.875,
      "learning_rate": 2.3402803273654326e-05,
      "loss": 4.2622,
      "num_input_tokens_seen": 28200976,
      "step": 20815
    },
    {
      "epoch": 0.6664106011138852,
      "grad_norm": 42.25,
      "learning_rate": 2.3393005976304983e-05,
      "loss": 4.1066,
      "num_input_tokens_seen": 28207984,
      "step": 20820
    },
    {
      "epoch": 0.6665706420843736,
      "grad_norm": 25.75,
      "learning_rate": 2.338320892677432e-05,
      "loss": 4.9422,
      "num_input_tokens_seen": 28214720,
      "step": 20825
    },
    {
      "epoch": 0.666730683054862,
      "grad_norm": 47.25,
      "learning_rate": 2.3373412126573155e-05,
      "loss": 5.0894,
      "num_input_tokens_seen": 28221312,
      "step": 20830
    },
    {
      "epoch": 0.6668907240253504,
      "grad_norm": 29.0,
      "learning_rate": 2.3363615577212285e-05,
      "loss": 3.7317,
      "num_input_tokens_seen": 28228176,
      "step": 20835
    },
    {
      "epoch": 0.6670507649958389,
      "grad_norm": 36.0,
      "learning_rate": 2.3353819280202455e-05,
      "loss": 4.4459,
      "num_input_tokens_seen": 28235056,
      "step": 20840
    },
    {
      "epoch": 0.6672108059663274,
      "grad_norm": 27.5,
      "learning_rate": 2.334402323705438e-05,
      "loss": 4.1646,
      "num_input_tokens_seen": 28241840,
      "step": 20845
    },
    {
      "epoch": 0.6673708469368158,
      "grad_norm": 16.0,
      "learning_rate": 2.3334227449278725e-05,
      "loss": 4.1028,
      "num_input_tokens_seen": 28248560,
      "step": 20850
    },
    {
      "epoch": 0.6675308879073043,
      "grad_norm": 26.375,
      "learning_rate": 2.3324431918386143e-05,
      "loss": 5.4166,
      "num_input_tokens_seen": 28255440,
      "step": 20855
    },
    {
      "epoch": 0.6676909288777927,
      "grad_norm": 56.25,
      "learning_rate": 2.3314636645887207e-05,
      "loss": 4.8347,
      "num_input_tokens_seen": 28262336,
      "step": 20860
    },
    {
      "epoch": 0.6678509698482812,
      "grad_norm": 21.0,
      "learning_rate": 2.3304841633292487e-05,
      "loss": 4.8466,
      "num_input_tokens_seen": 28268768,
      "step": 20865
    },
    {
      "epoch": 0.6680110108187696,
      "grad_norm": 22.75,
      "learning_rate": 2.329504688211248e-05,
      "loss": 3.9244,
      "num_input_tokens_seen": 28275344,
      "step": 20870
    },
    {
      "epoch": 0.6681710517892581,
      "grad_norm": 33.75,
      "learning_rate": 2.3285252393857677e-05,
      "loss": 3.8984,
      "num_input_tokens_seen": 28282192,
      "step": 20875
    },
    {
      "epoch": 0.6683310927597464,
      "grad_norm": 29.0,
      "learning_rate": 2.327545817003851e-05,
      "loss": 4.8095,
      "num_input_tokens_seen": 28288720,
      "step": 20880
    },
    {
      "epoch": 0.6684911337302349,
      "grad_norm": 47.75,
      "learning_rate": 2.326566421216535e-05,
      "loss": 4.3271,
      "num_input_tokens_seen": 28295232,
      "step": 20885
    },
    {
      "epoch": 0.6686511747007234,
      "grad_norm": 25.625,
      "learning_rate": 2.3255870521748565e-05,
      "loss": 4.4239,
      "num_input_tokens_seen": 28301872,
      "step": 20890
    },
    {
      "epoch": 0.6688112156712118,
      "grad_norm": 27.375,
      "learning_rate": 2.3246077100298474e-05,
      "loss": 4.2418,
      "num_input_tokens_seen": 28308576,
      "step": 20895
    },
    {
      "epoch": 0.6689712566417003,
      "grad_norm": 39.5,
      "learning_rate": 2.3236283949325328e-05,
      "loss": 4.588,
      "num_input_tokens_seen": 28314880,
      "step": 20900
    },
    {
      "epoch": 0.6691312976121887,
      "grad_norm": 27.75,
      "learning_rate": 2.3226491070339368e-05,
      "loss": 4.6864,
      "num_input_tokens_seen": 28321872,
      "step": 20905
    },
    {
      "epoch": 0.6692913385826772,
      "grad_norm": 41.75,
      "learning_rate": 2.3216698464850762e-05,
      "loss": 3.5664,
      "num_input_tokens_seen": 28328848,
      "step": 20910
    },
    {
      "epoch": 0.6694513795531656,
      "grad_norm": 21.375,
      "learning_rate": 2.320690613436967e-05,
      "loss": 4.0809,
      "num_input_tokens_seen": 28335536,
      "step": 20915
    },
    {
      "epoch": 0.6696114205236541,
      "grad_norm": 34.0,
      "learning_rate": 2.3197114080406192e-05,
      "loss": 4.2397,
      "num_input_tokens_seen": 28342192,
      "step": 20920
    },
    {
      "epoch": 0.6697714614941425,
      "grad_norm": 26.375,
      "learning_rate": 2.3187322304470365e-05,
      "loss": 4.7514,
      "num_input_tokens_seen": 28349200,
      "step": 20925
    },
    {
      "epoch": 0.669931502464631,
      "grad_norm": 20.25,
      "learning_rate": 2.3177530808072222e-05,
      "loss": 4.3491,
      "num_input_tokens_seen": 28356064,
      "step": 20930
    },
    {
      "epoch": 0.6700915434351193,
      "grad_norm": 29.0,
      "learning_rate": 2.316773959272174e-05,
      "loss": 4.0153,
      "num_input_tokens_seen": 28362944,
      "step": 20935
    },
    {
      "epoch": 0.6702515844056078,
      "grad_norm": 22.5,
      "learning_rate": 2.3157948659928823e-05,
      "loss": 4.3984,
      "num_input_tokens_seen": 28369696,
      "step": 20940
    },
    {
      "epoch": 0.6704116253760963,
      "grad_norm": 21.125,
      "learning_rate": 2.3148158011203388e-05,
      "loss": 3.5142,
      "num_input_tokens_seen": 28376384,
      "step": 20945
    },
    {
      "epoch": 0.6705716663465847,
      "grad_norm": 31.5,
      "learning_rate": 2.3138367648055253e-05,
      "loss": 4.5897,
      "num_input_tokens_seen": 28383200,
      "step": 20950
    },
    {
      "epoch": 0.6707317073170732,
      "grad_norm": 28.125,
      "learning_rate": 2.312857757199422e-05,
      "loss": 4.6661,
      "num_input_tokens_seen": 28390112,
      "step": 20955
    },
    {
      "epoch": 0.6708917482875616,
      "grad_norm": 31.25,
      "learning_rate": 2.3118787784530048e-05,
      "loss": 4.3849,
      "num_input_tokens_seen": 28397072,
      "step": 20960
    },
    {
      "epoch": 0.6710517892580501,
      "grad_norm": 25.25,
      "learning_rate": 2.310899828717243e-05,
      "loss": 3.8615,
      "num_input_tokens_seen": 28404048,
      "step": 20965
    },
    {
      "epoch": 0.6712118302285385,
      "grad_norm": 42.75,
      "learning_rate": 2.309920908143104e-05,
      "loss": 4.282,
      "num_input_tokens_seen": 28411072,
      "step": 20970
    },
    {
      "epoch": 0.671371871199027,
      "grad_norm": 37.5,
      "learning_rate": 2.308942016881551e-05,
      "loss": 4.5415,
      "num_input_tokens_seen": 28418016,
      "step": 20975
    },
    {
      "epoch": 0.6715319121695154,
      "grad_norm": 31.75,
      "learning_rate": 2.307963155083539e-05,
      "loss": 4.4887,
      "num_input_tokens_seen": 28424480,
      "step": 20980
    },
    {
      "epoch": 0.6716919531400039,
      "grad_norm": 21.75,
      "learning_rate": 2.306984322900022e-05,
      "loss": 4.1856,
      "num_input_tokens_seen": 28431584,
      "step": 20985
    },
    {
      "epoch": 0.6718519941104922,
      "grad_norm": 31.875,
      "learning_rate": 2.3060055204819482e-05,
      "loss": 4.5656,
      "num_input_tokens_seen": 28438336,
      "step": 20990
    },
    {
      "epoch": 0.6720120350809807,
      "grad_norm": 27.5,
      "learning_rate": 2.3050267479802604e-05,
      "loss": 3.6513,
      "num_input_tokens_seen": 28445312,
      "step": 20995
    },
    {
      "epoch": 0.6721720760514692,
      "grad_norm": 26.25,
      "learning_rate": 2.304048005545899e-05,
      "loss": 4.7914,
      "num_input_tokens_seen": 28451984,
      "step": 21000
    },
    {
      "epoch": 0.6721720760514692,
      "eval_loss": 4.479284286499023,
      "eval_runtime": 489.6094,
      "eval_samples_per_second": 28.361,
      "eval_steps_per_second": 14.181,
      "num_input_tokens_seen": 28451984,
      "step": 21000
    },
    {
      "epoch": 0.6723321170219576,
      "grad_norm": 11.8125,
      "learning_rate": 2.3030692933297972e-05,
      "loss": 2.9018,
      "num_input_tokens_seen": 28458928,
      "step": 21005
    },
    {
      "epoch": 0.6724921579924461,
      "grad_norm": 30.875,
      "learning_rate": 2.3020906114828843e-05,
      "loss": 4.3125,
      "num_input_tokens_seen": 28465648,
      "step": 21010
    },
    {
      "epoch": 0.6726521989629345,
      "grad_norm": 25.0,
      "learning_rate": 2.301111960156088e-05,
      "loss": 3.219,
      "num_input_tokens_seen": 28473008,
      "step": 21015
    },
    {
      "epoch": 0.672812239933423,
      "grad_norm": 31.25,
      "learning_rate": 2.300133339500326e-05,
      "loss": 4.9582,
      "num_input_tokens_seen": 28479952,
      "step": 21020
    },
    {
      "epoch": 0.6729722809039114,
      "grad_norm": 14.9375,
      "learning_rate": 2.2991547496665148e-05,
      "loss": 3.6036,
      "num_input_tokens_seen": 28486288,
      "step": 21025
    },
    {
      "epoch": 0.6731323218743999,
      "grad_norm": 29.875,
      "learning_rate": 2.298176190805565e-05,
      "loss": 5.115,
      "num_input_tokens_seen": 28492912,
      "step": 21030
    },
    {
      "epoch": 0.6732923628448882,
      "grad_norm": 22.875,
      "learning_rate": 2.2971976630683826e-05,
      "loss": 3.7549,
      "num_input_tokens_seen": 28499872,
      "step": 21035
    },
    {
      "epoch": 0.6734524038153767,
      "grad_norm": 26.5,
      "learning_rate": 2.29621916660587e-05,
      "loss": 4.5947,
      "num_input_tokens_seen": 28506560,
      "step": 21040
    },
    {
      "epoch": 0.6736124447858651,
      "grad_norm": 21.625,
      "learning_rate": 2.295240701568922e-05,
      "loss": 3.493,
      "num_input_tokens_seen": 28513360,
      "step": 21045
    },
    {
      "epoch": 0.6737724857563536,
      "grad_norm": 19.125,
      "learning_rate": 2.2942622681084312e-05,
      "loss": 3.9703,
      "num_input_tokens_seen": 28520320,
      "step": 21050
    },
    {
      "epoch": 0.6739325267268421,
      "grad_norm": 33.5,
      "learning_rate": 2.293283866375284e-05,
      "loss": 4.6793,
      "num_input_tokens_seen": 28526800,
      "step": 21055
    },
    {
      "epoch": 0.6740925676973305,
      "grad_norm": 32.25,
      "learning_rate": 2.2923054965203627e-05,
      "loss": 5.0674,
      "num_input_tokens_seen": 28534080,
      "step": 21060
    },
    {
      "epoch": 0.674252608667819,
      "grad_norm": 44.0,
      "learning_rate": 2.2913271586945443e-05,
      "loss": 4.8226,
      "num_input_tokens_seen": 28540544,
      "step": 21065
    },
    {
      "epoch": 0.6744126496383074,
      "grad_norm": 16.0,
      "learning_rate": 2.290348853048699e-05,
      "loss": 3.9933,
      "num_input_tokens_seen": 28547568,
      "step": 21070
    },
    {
      "epoch": 0.6745726906087959,
      "grad_norm": 23.125,
      "learning_rate": 2.2893705797336956e-05,
      "loss": 4.2238,
      "num_input_tokens_seen": 28554288,
      "step": 21075
    },
    {
      "epoch": 0.6747327315792843,
      "grad_norm": 31.375,
      "learning_rate": 2.288392338900397e-05,
      "loss": 4.3033,
      "num_input_tokens_seen": 28561216,
      "step": 21080
    },
    {
      "epoch": 0.6748927725497728,
      "grad_norm": 29.625,
      "learning_rate": 2.2874141306996576e-05,
      "loss": 5.05,
      "num_input_tokens_seen": 28568304,
      "step": 21085
    },
    {
      "epoch": 0.6750528135202611,
      "grad_norm": 32.0,
      "learning_rate": 2.2864359552823312e-05,
      "loss": 3.4805,
      "num_input_tokens_seen": 28575024,
      "step": 21090
    },
    {
      "epoch": 0.6752128544907496,
      "grad_norm": 33.75,
      "learning_rate": 2.2854578127992648e-05,
      "loss": 3.8566,
      "num_input_tokens_seen": 28582096,
      "step": 21095
    },
    {
      "epoch": 0.6753728954612381,
      "grad_norm": 45.5,
      "learning_rate": 2.2844797034012988e-05,
      "loss": 3.1942,
      "num_input_tokens_seen": 28588704,
      "step": 21100
    },
    {
      "epoch": 0.6755329364317265,
      "grad_norm": 23.625,
      "learning_rate": 2.2835016272392722e-05,
      "loss": 3.7847,
      "num_input_tokens_seen": 28595904,
      "step": 21105
    },
    {
      "epoch": 0.675692977402215,
      "grad_norm": 26.75,
      "learning_rate": 2.2825235844640142e-05,
      "loss": 4.2137,
      "num_input_tokens_seen": 28602304,
      "step": 21110
    },
    {
      "epoch": 0.6758530183727034,
      "grad_norm": 30.5,
      "learning_rate": 2.2815455752263522e-05,
      "loss": 4.979,
      "num_input_tokens_seen": 28609344,
      "step": 21115
    },
    {
      "epoch": 0.6760130593431919,
      "grad_norm": 37.25,
      "learning_rate": 2.2805675996771092e-05,
      "loss": 4.2062,
      "num_input_tokens_seen": 28616192,
      "step": 21120
    },
    {
      "epoch": 0.6761731003136803,
      "grad_norm": 34.25,
      "learning_rate": 2.2795896579670987e-05,
      "loss": 4.7709,
      "num_input_tokens_seen": 28623104,
      "step": 21125
    },
    {
      "epoch": 0.6763331412841688,
      "grad_norm": 30.0,
      "learning_rate": 2.2786117502471337e-05,
      "loss": 4.8276,
      "num_input_tokens_seen": 28630144,
      "step": 21130
    },
    {
      "epoch": 0.6764931822546572,
      "grad_norm": 22.25,
      "learning_rate": 2.2776338766680185e-05,
      "loss": 4.7218,
      "num_input_tokens_seen": 28636704,
      "step": 21135
    },
    {
      "epoch": 0.6766532232251457,
      "grad_norm": 42.25,
      "learning_rate": 2.2766560373805533e-05,
      "loss": 5.3689,
      "num_input_tokens_seen": 28643408,
      "step": 21140
    },
    {
      "epoch": 0.676813264195634,
      "grad_norm": 46.25,
      "learning_rate": 2.2756782325355353e-05,
      "loss": 4.374,
      "num_input_tokens_seen": 28650624,
      "step": 21145
    },
    {
      "epoch": 0.6769733051661225,
      "grad_norm": 33.0,
      "learning_rate": 2.2747004622837514e-05,
      "loss": 4.0412,
      "num_input_tokens_seen": 28657072,
      "step": 21150
    },
    {
      "epoch": 0.677133346136611,
      "grad_norm": 36.75,
      "learning_rate": 2.2737227267759878e-05,
      "loss": 5.3646,
      "num_input_tokens_seen": 28663536,
      "step": 21155
    },
    {
      "epoch": 0.6772933871070994,
      "grad_norm": 10.125,
      "learning_rate": 2.272745026163024e-05,
      "loss": 4.6865,
      "num_input_tokens_seen": 28670192,
      "step": 21160
    },
    {
      "epoch": 0.6774534280775879,
      "grad_norm": 35.5,
      "learning_rate": 2.271767360595633e-05,
      "loss": 4.3282,
      "num_input_tokens_seen": 28676640,
      "step": 21165
    },
    {
      "epoch": 0.6776134690480763,
      "grad_norm": 52.0,
      "learning_rate": 2.270789730224583e-05,
      "loss": 4.6352,
      "num_input_tokens_seen": 28683264,
      "step": 21170
    },
    {
      "epoch": 0.6777735100185648,
      "grad_norm": 26.5,
      "learning_rate": 2.2698121352006367e-05,
      "loss": 3.8716,
      "num_input_tokens_seen": 28689696,
      "step": 21175
    },
    {
      "epoch": 0.6779335509890532,
      "grad_norm": 41.25,
      "learning_rate": 2.2688345756745517e-05,
      "loss": 4.2159,
      "num_input_tokens_seen": 28696032,
      "step": 21180
    },
    {
      "epoch": 0.6780935919595417,
      "grad_norm": 34.5,
      "learning_rate": 2.267857051797081e-05,
      "loss": 4.1809,
      "num_input_tokens_seen": 28703088,
      "step": 21185
    },
    {
      "epoch": 0.67825363293003,
      "grad_norm": 36.0,
      "learning_rate": 2.2668795637189695e-05,
      "loss": 4.2084,
      "num_input_tokens_seen": 28710304,
      "step": 21190
    },
    {
      "epoch": 0.6784136739005185,
      "grad_norm": 32.5,
      "learning_rate": 2.2659021115909586e-05,
      "loss": 4.6422,
      "num_input_tokens_seen": 28717136,
      "step": 21195
    },
    {
      "epoch": 0.6785737148710069,
      "grad_norm": 23.625,
      "learning_rate": 2.2649246955637847e-05,
      "loss": 4.6492,
      "num_input_tokens_seen": 28723904,
      "step": 21200
    },
    {
      "epoch": 0.6785737148710069,
      "eval_loss": 4.47866678237915,
      "eval_runtime": 490.2586,
      "eval_samples_per_second": 28.324,
      "eval_steps_per_second": 14.162,
      "num_input_tokens_seen": 28723904,
      "step": 21200
    },
    {
      "epoch": 0.6787337558414954,
      "grad_norm": 22.0,
      "learning_rate": 2.2639473157881766e-05,
      "loss": 4.0668,
      "num_input_tokens_seen": 28730512,
      "step": 21205
    },
    {
      "epoch": 0.6788937968119839,
      "grad_norm": 27.375,
      "learning_rate": 2.2629699724148594e-05,
      "loss": 5.1983,
      "num_input_tokens_seen": 28737232,
      "step": 21210
    },
    {
      "epoch": 0.6790538377824723,
      "grad_norm": 22.375,
      "learning_rate": 2.26199266559455e-05,
      "loss": 4.3256,
      "num_input_tokens_seen": 28744608,
      "step": 21215
    },
    {
      "epoch": 0.6792138787529608,
      "grad_norm": 13.5,
      "learning_rate": 2.2610153954779625e-05,
      "loss": 4.5592,
      "num_input_tokens_seen": 28751776,
      "step": 21220
    },
    {
      "epoch": 0.6793739197234492,
      "grad_norm": 25.625,
      "learning_rate": 2.2600381622158056e-05,
      "loss": 4.7331,
      "num_input_tokens_seen": 28758672,
      "step": 21225
    },
    {
      "epoch": 0.6795339606939377,
      "grad_norm": 39.75,
      "learning_rate": 2.2590609659587783e-05,
      "loss": 4.9925,
      "num_input_tokens_seen": 28765440,
      "step": 21230
    },
    {
      "epoch": 0.6796940016644261,
      "grad_norm": 27.875,
      "learning_rate": 2.2580838068575787e-05,
      "loss": 3.958,
      "num_input_tokens_seen": 28772240,
      "step": 21235
    },
    {
      "epoch": 0.6798540426349146,
      "grad_norm": 23.0,
      "learning_rate": 2.257106685062896e-05,
      "loss": 4.374,
      "num_input_tokens_seen": 28779072,
      "step": 21240
    },
    {
      "epoch": 0.6800140836054029,
      "grad_norm": 24.375,
      "learning_rate": 2.256129600725415e-05,
      "loss": 4.8412,
      "num_input_tokens_seen": 28785904,
      "step": 21245
    },
    {
      "epoch": 0.6801741245758914,
      "grad_norm": 21.625,
      "learning_rate": 2.2551525539958145e-05,
      "loss": 3.9055,
      "num_input_tokens_seen": 28792528,
      "step": 21250
    },
    {
      "epoch": 0.6803341655463798,
      "grad_norm": 29.125,
      "learning_rate": 2.2541755450247663e-05,
      "loss": 4.2623,
      "num_input_tokens_seen": 28799344,
      "step": 21255
    },
    {
      "epoch": 0.6804942065168683,
      "grad_norm": 23.25,
      "learning_rate": 2.2531985739629382e-05,
      "loss": 4.6604,
      "num_input_tokens_seen": 28806016,
      "step": 21260
    },
    {
      "epoch": 0.6806542474873568,
      "grad_norm": 21.5,
      "learning_rate": 2.2522216409609924e-05,
      "loss": 3.7103,
      "num_input_tokens_seen": 28813056,
      "step": 21265
    },
    {
      "epoch": 0.6808142884578452,
      "grad_norm": 31.125,
      "learning_rate": 2.2512447461695826e-05,
      "loss": 4.7916,
      "num_input_tokens_seen": 28819904,
      "step": 21270
    },
    {
      "epoch": 0.6809743294283337,
      "grad_norm": 36.75,
      "learning_rate": 2.2502678897393593e-05,
      "loss": 4.5229,
      "num_input_tokens_seen": 28826656,
      "step": 21275
    },
    {
      "epoch": 0.6811343703988221,
      "grad_norm": 50.0,
      "learning_rate": 2.2492910718209665e-05,
      "loss": 5.0106,
      "num_input_tokens_seen": 28833392,
      "step": 21280
    },
    {
      "epoch": 0.6812944113693106,
      "grad_norm": 27.875,
      "learning_rate": 2.2483142925650398e-05,
      "loss": 4.8775,
      "num_input_tokens_seen": 28840016,
      "step": 21285
    },
    {
      "epoch": 0.681454452339799,
      "grad_norm": 47.75,
      "learning_rate": 2.247337552122213e-05,
      "loss": 6.3062,
      "num_input_tokens_seen": 28846304,
      "step": 21290
    },
    {
      "epoch": 0.6816144933102875,
      "grad_norm": 36.0,
      "learning_rate": 2.24636085064311e-05,
      "loss": 4.209,
      "num_input_tokens_seen": 28853040,
      "step": 21295
    },
    {
      "epoch": 0.6817745342807758,
      "grad_norm": 30.0,
      "learning_rate": 2.245384188278351e-05,
      "loss": 4.7826,
      "num_input_tokens_seen": 28859584,
      "step": 21300
    },
    {
      "epoch": 0.6819345752512643,
      "grad_norm": 24.875,
      "learning_rate": 2.2444075651785513e-05,
      "loss": 3.8475,
      "num_input_tokens_seen": 28865952,
      "step": 21305
    },
    {
      "epoch": 0.6820946162217527,
      "grad_norm": 20.375,
      "learning_rate": 2.243430981494316e-05,
      "loss": 3.6104,
      "num_input_tokens_seen": 28872640,
      "step": 21310
    },
    {
      "epoch": 0.6822546571922412,
      "grad_norm": 27.5,
      "learning_rate": 2.2424544373762475e-05,
      "loss": 4.0333,
      "num_input_tokens_seen": 28879360,
      "step": 21315
    },
    {
      "epoch": 0.6824146981627297,
      "grad_norm": 28.625,
      "learning_rate": 2.2414779329749418e-05,
      "loss": 3.5596,
      "num_input_tokens_seen": 28885952,
      "step": 21320
    },
    {
      "epoch": 0.6825747391332181,
      "grad_norm": 21.125,
      "learning_rate": 2.2405014684409873e-05,
      "loss": 4.1618,
      "num_input_tokens_seen": 28892896,
      "step": 21325
    },
    {
      "epoch": 0.6827347801037066,
      "grad_norm": 28.625,
      "learning_rate": 2.239525043924968e-05,
      "loss": 4.2218,
      "num_input_tokens_seen": 28899488,
      "step": 21330
    },
    {
      "epoch": 0.682894821074195,
      "grad_norm": 19.375,
      "learning_rate": 2.2385486595774592e-05,
      "loss": 3.8399,
      "num_input_tokens_seen": 28906320,
      "step": 21335
    },
    {
      "epoch": 0.6830548620446835,
      "grad_norm": 35.0,
      "learning_rate": 2.237572315549033e-05,
      "loss": 4.3602,
      "num_input_tokens_seen": 28912912,
      "step": 21340
    },
    {
      "epoch": 0.6832149030151718,
      "grad_norm": 30.625,
      "learning_rate": 2.2365960119902545e-05,
      "loss": 4.3848,
      "num_input_tokens_seen": 28919328,
      "step": 21345
    },
    {
      "epoch": 0.6833749439856603,
      "grad_norm": 19.125,
      "learning_rate": 2.2356197490516806e-05,
      "loss": 4.6888,
      "num_input_tokens_seen": 28926016,
      "step": 21350
    },
    {
      "epoch": 0.6835349849561487,
      "grad_norm": 27.375,
      "learning_rate": 2.234643526883863e-05,
      "loss": 4.1073,
      "num_input_tokens_seen": 28932784,
      "step": 21355
    },
    {
      "epoch": 0.6836950259266372,
      "grad_norm": 24.875,
      "learning_rate": 2.2336673456373497e-05,
      "loss": 3.8212,
      "num_input_tokens_seen": 28939776,
      "step": 21360
    },
    {
      "epoch": 0.6838550668971257,
      "grad_norm": 25.125,
      "learning_rate": 2.2326912054626772e-05,
      "loss": 4.5836,
      "num_input_tokens_seen": 28946512,
      "step": 21365
    },
    {
      "epoch": 0.6840151078676141,
      "grad_norm": 17.75,
      "learning_rate": 2.2317151065103813e-05,
      "loss": 3.2659,
      "num_input_tokens_seen": 28953408,
      "step": 21370
    },
    {
      "epoch": 0.6841751488381026,
      "grad_norm": 31.25,
      "learning_rate": 2.2307390489309865e-05,
      "loss": 4.3065,
      "num_input_tokens_seen": 28960032,
      "step": 21375
    },
    {
      "epoch": 0.684335189808591,
      "grad_norm": 34.25,
      "learning_rate": 2.2297630328750146e-05,
      "loss": 4.3545,
      "num_input_tokens_seen": 28966992,
      "step": 21380
    },
    {
      "epoch": 0.6844952307790795,
      "grad_norm": 32.75,
      "learning_rate": 2.228787058492979e-05,
      "loss": 3.8956,
      "num_input_tokens_seen": 28973680,
      "step": 21385
    },
    {
      "epoch": 0.6846552717495679,
      "grad_norm": 45.25,
      "learning_rate": 2.2278111259353875e-05,
      "loss": 3.7233,
      "num_input_tokens_seen": 28980624,
      "step": 21390
    },
    {
      "epoch": 0.6848153127200564,
      "grad_norm": 27.875,
      "learning_rate": 2.2268352353527395e-05,
      "loss": 3.5216,
      "num_input_tokens_seen": 28987408,
      "step": 21395
    },
    {
      "epoch": 0.6849753536905447,
      "grad_norm": 32.75,
      "learning_rate": 2.225859386895533e-05,
      "loss": 5.6937,
      "num_input_tokens_seen": 28994096,
      "step": 21400
    },
    {
      "epoch": 0.6849753536905447,
      "eval_loss": 4.488027095794678,
      "eval_runtime": 490.0776,
      "eval_samples_per_second": 28.334,
      "eval_steps_per_second": 14.167,
      "num_input_tokens_seen": 28994096,
      "step": 21400
    },
    {
      "epoch": 0.6851353946610332,
      "grad_norm": 34.25,
      "learning_rate": 2.2248835807142525e-05,
      "loss": 4.1882,
      "num_input_tokens_seen": 29001456,
      "step": 21405
    },
    {
      "epoch": 0.6852954356315216,
      "grad_norm": 22.0,
      "learning_rate": 2.2239078169593826e-05,
      "loss": 4.0299,
      "num_input_tokens_seen": 29008448,
      "step": 21410
    },
    {
      "epoch": 0.6854554766020101,
      "grad_norm": 38.5,
      "learning_rate": 2.222932095781396e-05,
      "loss": 3.0542,
      "num_input_tokens_seen": 29016000,
      "step": 21415
    },
    {
      "epoch": 0.6856155175724986,
      "grad_norm": 53.25,
      "learning_rate": 2.221956417330762e-05,
      "loss": 2.802,
      "num_input_tokens_seen": 29022784,
      "step": 21420
    },
    {
      "epoch": 0.685775558542987,
      "grad_norm": 21.0,
      "learning_rate": 2.2209807817579438e-05,
      "loss": 4.2885,
      "num_input_tokens_seen": 29030080,
      "step": 21425
    },
    {
      "epoch": 0.6859355995134755,
      "grad_norm": 47.0,
      "learning_rate": 2.220005189213394e-05,
      "loss": 3.8184,
      "num_input_tokens_seen": 29036752,
      "step": 21430
    },
    {
      "epoch": 0.6860956404839639,
      "grad_norm": 38.5,
      "learning_rate": 2.2190296398475624e-05,
      "loss": 3.8493,
      "num_input_tokens_seen": 29043616,
      "step": 21435
    },
    {
      "epoch": 0.6862556814544524,
      "grad_norm": 31.75,
      "learning_rate": 2.2180541338108926e-05,
      "loss": 4.2786,
      "num_input_tokens_seen": 29050288,
      "step": 21440
    },
    {
      "epoch": 0.6864157224249408,
      "grad_norm": 44.5,
      "learning_rate": 2.2170786712538176e-05,
      "loss": 4.2462,
      "num_input_tokens_seen": 29056960,
      "step": 21445
    },
    {
      "epoch": 0.6865757633954293,
      "grad_norm": 54.5,
      "learning_rate": 2.216103252326768e-05,
      "loss": 4.8255,
      "num_input_tokens_seen": 29063872,
      "step": 21450
    },
    {
      "epoch": 0.6867358043659176,
      "grad_norm": 38.75,
      "learning_rate": 2.2151278771801635e-05,
      "loss": 4.2036,
      "num_input_tokens_seen": 29070400,
      "step": 21455
    },
    {
      "epoch": 0.6868958453364061,
      "grad_norm": 32.25,
      "learning_rate": 2.21415254596442e-05,
      "loss": 3.7846,
      "num_input_tokens_seen": 29077344,
      "step": 21460
    },
    {
      "epoch": 0.6870558863068945,
      "grad_norm": 26.75,
      "learning_rate": 2.213177258829947e-05,
      "loss": 4.309,
      "num_input_tokens_seen": 29084048,
      "step": 21465
    },
    {
      "epoch": 0.687215927277383,
      "grad_norm": 26.125,
      "learning_rate": 2.2122020159271445e-05,
      "loss": 4.3798,
      "num_input_tokens_seen": 29091008,
      "step": 21470
    },
    {
      "epoch": 0.6873759682478715,
      "grad_norm": 26.625,
      "learning_rate": 2.2112268174064075e-05,
      "loss": 3.8512,
      "num_input_tokens_seen": 29097792,
      "step": 21475
    },
    {
      "epoch": 0.6875360092183599,
      "grad_norm": 36.25,
      "learning_rate": 2.2102516634181253e-05,
      "loss": 3.5871,
      "num_input_tokens_seen": 29104640,
      "step": 21480
    },
    {
      "epoch": 0.6876960501888484,
      "grad_norm": 15.125,
      "learning_rate": 2.209276554112677e-05,
      "loss": 4.5274,
      "num_input_tokens_seen": 29111568,
      "step": 21485
    },
    {
      "epoch": 0.6878560911593368,
      "grad_norm": 18.125,
      "learning_rate": 2.2083014896404384e-05,
      "loss": 4.504,
      "num_input_tokens_seen": 29118432,
      "step": 21490
    },
    {
      "epoch": 0.6880161321298253,
      "grad_norm": 24.5,
      "learning_rate": 2.207326470151775e-05,
      "loss": 4.0491,
      "num_input_tokens_seen": 29124976,
      "step": 21495
    },
    {
      "epoch": 0.6881761731003136,
      "grad_norm": 18.75,
      "learning_rate": 2.2063514957970477e-05,
      "loss": 3.6447,
      "num_input_tokens_seen": 29131488,
      "step": 21500
    },
    {
      "epoch": 0.6883362140708021,
      "grad_norm": 40.5,
      "learning_rate": 2.205376566726611e-05,
      "loss": 5.4693,
      "num_input_tokens_seen": 29138672,
      "step": 21505
    },
    {
      "epoch": 0.6884962550412905,
      "grad_norm": 40.25,
      "learning_rate": 2.204401683090809e-05,
      "loss": 4.8444,
      "num_input_tokens_seen": 29145440,
      "step": 21510
    },
    {
      "epoch": 0.688656296011779,
      "grad_norm": 23.875,
      "learning_rate": 2.203426845039982e-05,
      "loss": 4.0556,
      "num_input_tokens_seen": 29152192,
      "step": 21515
    },
    {
      "epoch": 0.6888163369822674,
      "grad_norm": 51.5,
      "learning_rate": 2.202452052724464e-05,
      "loss": 4.4757,
      "num_input_tokens_seen": 29159088,
      "step": 21520
    },
    {
      "epoch": 0.6889763779527559,
      "grad_norm": 38.25,
      "learning_rate": 2.2014773062945777e-05,
      "loss": 4.0315,
      "num_input_tokens_seen": 29165632,
      "step": 21525
    },
    {
      "epoch": 0.6891364189232444,
      "grad_norm": 35.75,
      "learning_rate": 2.2005026059006427e-05,
      "loss": 4.7741,
      "num_input_tokens_seen": 29172432,
      "step": 21530
    },
    {
      "epoch": 0.6892964598937328,
      "grad_norm": 26.5,
      "learning_rate": 2.1995279516929695e-05,
      "loss": 4.1579,
      "num_input_tokens_seen": 29178992,
      "step": 21535
    },
    {
      "epoch": 0.6894565008642213,
      "grad_norm": 42.5,
      "learning_rate": 2.1985533438218613e-05,
      "loss": 4.0359,
      "num_input_tokens_seen": 29185872,
      "step": 21540
    },
    {
      "epoch": 0.6896165418347097,
      "grad_norm": 35.5,
      "learning_rate": 2.197578782437617e-05,
      "loss": 4.7044,
      "num_input_tokens_seen": 29192544,
      "step": 21545
    },
    {
      "epoch": 0.6897765828051982,
      "grad_norm": 51.5,
      "learning_rate": 2.196604267690524e-05,
      "loss": 4.9122,
      "num_input_tokens_seen": 29199040,
      "step": 21550
    },
    {
      "epoch": 0.6899366237756865,
      "grad_norm": 27.75,
      "learning_rate": 2.195629799730865e-05,
      "loss": 4.9946,
      "num_input_tokens_seen": 29205840,
      "step": 21555
    },
    {
      "epoch": 0.690096664746175,
      "grad_norm": 30.0,
      "learning_rate": 2.1946553787089173e-05,
      "loss": 4.3781,
      "num_input_tokens_seen": 29212336,
      "step": 21560
    },
    {
      "epoch": 0.6902567057166634,
      "grad_norm": 45.25,
      "learning_rate": 2.193681004774947e-05,
      "loss": 4.4088,
      "num_input_tokens_seen": 29219136,
      "step": 21565
    },
    {
      "epoch": 0.6904167466871519,
      "grad_norm": 30.625,
      "learning_rate": 2.1927066780792154e-05,
      "loss": 4.9303,
      "num_input_tokens_seen": 29226480,
      "step": 21570
    },
    {
      "epoch": 0.6905767876576403,
      "grad_norm": 21.5,
      "learning_rate": 2.191732398771975e-05,
      "loss": 4.3234,
      "num_input_tokens_seen": 29233376,
      "step": 21575
    },
    {
      "epoch": 0.6907368286281288,
      "grad_norm": 46.75,
      "learning_rate": 2.1907581670034725e-05,
      "loss": 4.6716,
      "num_input_tokens_seen": 29240464,
      "step": 21580
    },
    {
      "epoch": 0.6908968695986173,
      "grad_norm": 38.0,
      "learning_rate": 2.189783982923948e-05,
      "loss": 4.0196,
      "num_input_tokens_seen": 29247360,
      "step": 21585
    },
    {
      "epoch": 0.6910569105691057,
      "grad_norm": 18.25,
      "learning_rate": 2.1888098466836303e-05,
      "loss": 3.5901,
      "num_input_tokens_seen": 29254112,
      "step": 21590
    },
    {
      "epoch": 0.6912169515395942,
      "grad_norm": 21.5,
      "learning_rate": 2.1878357584327457e-05,
      "loss": 4.0865,
      "num_input_tokens_seen": 29260960,
      "step": 21595
    },
    {
      "epoch": 0.6913769925100826,
      "grad_norm": 24.875,
      "learning_rate": 2.1868617183215103e-05,
      "loss": 3.4293,
      "num_input_tokens_seen": 29267904,
      "step": 21600
    },
    {
      "epoch": 0.6913769925100826,
      "eval_loss": 4.491370677947998,
      "eval_runtime": 490.2248,
      "eval_samples_per_second": 28.326,
      "eval_steps_per_second": 14.163,
      "num_input_tokens_seen": 29267904,
      "step": 21600
    },
    {
      "epoch": 0.691537033480571,
      "grad_norm": 60.0,
      "learning_rate": 2.1858877265001327e-05,
      "loss": 3.9118,
      "num_input_tokens_seen": 29274496,
      "step": 21605
    },
    {
      "epoch": 0.6916970744510594,
      "grad_norm": 25.75,
      "learning_rate": 2.184913783118816e-05,
      "loss": 4.3149,
      "num_input_tokens_seen": 29281120,
      "step": 21610
    },
    {
      "epoch": 0.6918571154215479,
      "grad_norm": 26.5,
      "learning_rate": 2.1839398883277522e-05,
      "loss": 4.225,
      "num_input_tokens_seen": 29287808,
      "step": 21615
    },
    {
      "epoch": 0.6920171563920363,
      "grad_norm": 20.75,
      "learning_rate": 2.182966042277129e-05,
      "loss": 4.4242,
      "num_input_tokens_seen": 29294640,
      "step": 21620
    },
    {
      "epoch": 0.6921771973625248,
      "grad_norm": 21.375,
      "learning_rate": 2.181992245117128e-05,
      "loss": 3.7991,
      "num_input_tokens_seen": 29301536,
      "step": 21625
    },
    {
      "epoch": 0.6923372383330133,
      "grad_norm": 26.0,
      "learning_rate": 2.181018496997918e-05,
      "loss": 3.9781,
      "num_input_tokens_seen": 29308432,
      "step": 21630
    },
    {
      "epoch": 0.6924972793035017,
      "grad_norm": 29.125,
      "learning_rate": 2.1800447980696648e-05,
      "loss": 4.5544,
      "num_input_tokens_seen": 29315152,
      "step": 21635
    },
    {
      "epoch": 0.6926573202739902,
      "grad_norm": 42.25,
      "learning_rate": 2.1790711484825248e-05,
      "loss": 4.2278,
      "num_input_tokens_seen": 29322032,
      "step": 21640
    },
    {
      "epoch": 0.6928173612444786,
      "grad_norm": 22.0,
      "learning_rate": 2.178097548386646e-05,
      "loss": 3.9235,
      "num_input_tokens_seen": 29328848,
      "step": 21645
    },
    {
      "epoch": 0.6929774022149671,
      "grad_norm": 33.25,
      "learning_rate": 2.1771239979321712e-05,
      "loss": 3.9434,
      "num_input_tokens_seen": 29335504,
      "step": 21650
    },
    {
      "epoch": 0.6931374431854554,
      "grad_norm": 22.875,
      "learning_rate": 2.1761504972692327e-05,
      "loss": 4.6136,
      "num_input_tokens_seen": 29342368,
      "step": 21655
    },
    {
      "epoch": 0.6932974841559439,
      "grad_norm": 26.125,
      "learning_rate": 2.1751770465479572e-05,
      "loss": 3.2883,
      "num_input_tokens_seen": 29349296,
      "step": 21660
    },
    {
      "epoch": 0.6934575251264323,
      "grad_norm": 39.0,
      "learning_rate": 2.174203645918464e-05,
      "loss": 4.8188,
      "num_input_tokens_seen": 29355904,
      "step": 21665
    },
    {
      "epoch": 0.6936175660969208,
      "grad_norm": 30.25,
      "learning_rate": 2.1732302955308624e-05,
      "loss": 4.553,
      "num_input_tokens_seen": 29362624,
      "step": 21670
    },
    {
      "epoch": 0.6937776070674092,
      "grad_norm": 32.25,
      "learning_rate": 2.172256995535255e-05,
      "loss": 4.0835,
      "num_input_tokens_seen": 29369680,
      "step": 21675
    },
    {
      "epoch": 0.6939376480378977,
      "grad_norm": 22.125,
      "learning_rate": 2.171283746081739e-05,
      "loss": 3.8608,
      "num_input_tokens_seen": 29376480,
      "step": 21680
    },
    {
      "epoch": 0.6940976890083862,
      "grad_norm": 46.25,
      "learning_rate": 2.1703105473203988e-05,
      "loss": 5.445,
      "num_input_tokens_seen": 29383056,
      "step": 21685
    },
    {
      "epoch": 0.6942577299788746,
      "grad_norm": 39.25,
      "learning_rate": 2.1693373994013168e-05,
      "loss": 4.2596,
      "num_input_tokens_seen": 29390032,
      "step": 21690
    },
    {
      "epoch": 0.6944177709493631,
      "grad_norm": 36.0,
      "learning_rate": 2.168364302474562e-05,
      "loss": 4.65,
      "num_input_tokens_seen": 29396688,
      "step": 21695
    },
    {
      "epoch": 0.6945778119198515,
      "grad_norm": 31.25,
      "learning_rate": 2.167391256690199e-05,
      "loss": 4.3424,
      "num_input_tokens_seen": 29403408,
      "step": 21700
    },
    {
      "epoch": 0.69473785289034,
      "grad_norm": 38.75,
      "learning_rate": 2.1664182621982855e-05,
      "loss": 3.4961,
      "num_input_tokens_seen": 29410048,
      "step": 21705
    },
    {
      "epoch": 0.6948978938608283,
      "grad_norm": 37.75,
      "learning_rate": 2.1654453191488673e-05,
      "loss": 5.3125,
      "num_input_tokens_seen": 29416576,
      "step": 21710
    },
    {
      "epoch": 0.6950579348313168,
      "grad_norm": 18.625,
      "learning_rate": 2.1644724276919846e-05,
      "loss": 3.7049,
      "num_input_tokens_seen": 29423120,
      "step": 21715
    },
    {
      "epoch": 0.6952179758018052,
      "grad_norm": 28.375,
      "learning_rate": 2.1634995879776715e-05,
      "loss": 4.0029,
      "num_input_tokens_seen": 29429568,
      "step": 21720
    },
    {
      "epoch": 0.6953780167722937,
      "grad_norm": 42.75,
      "learning_rate": 2.162526800155949e-05,
      "loss": 4.5111,
      "num_input_tokens_seen": 29436176,
      "step": 21725
    },
    {
      "epoch": 0.6955380577427821,
      "grad_norm": 33.5,
      "learning_rate": 2.1615540643768363e-05,
      "loss": 3.5619,
      "num_input_tokens_seen": 29443872,
      "step": 21730
    },
    {
      "epoch": 0.6956980987132706,
      "grad_norm": 44.5,
      "learning_rate": 2.160581380790339e-05,
      "loss": 4.8413,
      "num_input_tokens_seen": 29450944,
      "step": 21735
    },
    {
      "epoch": 0.6958581396837591,
      "grad_norm": 33.5,
      "learning_rate": 2.1596087495464586e-05,
      "loss": 4.3217,
      "num_input_tokens_seen": 29458240,
      "step": 21740
    },
    {
      "epoch": 0.6960181806542475,
      "grad_norm": 39.25,
      "learning_rate": 2.1586361707951866e-05,
      "loss": 4.0528,
      "num_input_tokens_seen": 29465376,
      "step": 21745
    },
    {
      "epoch": 0.696178221624736,
      "grad_norm": 42.0,
      "learning_rate": 2.157663644686507e-05,
      "loss": 4.4412,
      "num_input_tokens_seen": 29472080,
      "step": 21750
    },
    {
      "epoch": 0.6963382625952244,
      "grad_norm": 28.125,
      "learning_rate": 2.156691171370396e-05,
      "loss": 4.2142,
      "num_input_tokens_seen": 29479168,
      "step": 21755
    },
    {
      "epoch": 0.6964983035657129,
      "grad_norm": 27.375,
      "learning_rate": 2.1557187509968195e-05,
      "loss": 4.2009,
      "num_input_tokens_seen": 29485984,
      "step": 21760
    },
    {
      "epoch": 0.6966583445362012,
      "grad_norm": 43.5,
      "learning_rate": 2.1547463837157382e-05,
      "loss": 5.0509,
      "num_input_tokens_seen": 29492624,
      "step": 21765
    },
    {
      "epoch": 0.6968183855066897,
      "grad_norm": 27.625,
      "learning_rate": 2.1537740696771045e-05,
      "loss": 4.0605,
      "num_input_tokens_seen": 29499504,
      "step": 21770
    },
    {
      "epoch": 0.6969784264771781,
      "grad_norm": 21.625,
      "learning_rate": 2.1528018090308587e-05,
      "loss": 4.673,
      "num_input_tokens_seen": 29506672,
      "step": 21775
    },
    {
      "epoch": 0.6971384674476666,
      "grad_norm": 20.875,
      "learning_rate": 2.151829601926938e-05,
      "loss": 4.2396,
      "num_input_tokens_seen": 29513488,
      "step": 21780
    },
    {
      "epoch": 0.697298508418155,
      "grad_norm": 31.625,
      "learning_rate": 2.1508574485152684e-05,
      "loss": 4.4549,
      "num_input_tokens_seen": 29520112,
      "step": 21785
    },
    {
      "epoch": 0.6974585493886435,
      "grad_norm": 34.5,
      "learning_rate": 2.1498853489457667e-05,
      "loss": 4.5601,
      "num_input_tokens_seen": 29526768,
      "step": 21790
    },
    {
      "epoch": 0.697618590359132,
      "grad_norm": 27.625,
      "learning_rate": 2.1489133033683455e-05,
      "loss": 4.4645,
      "num_input_tokens_seen": 29533664,
      "step": 21795
    },
    {
      "epoch": 0.6977786313296204,
      "grad_norm": 40.0,
      "learning_rate": 2.1479413119329038e-05,
      "loss": 3.8018,
      "num_input_tokens_seen": 29540768,
      "step": 21800
    },
    {
      "epoch": 0.6977786313296204,
      "eval_loss": 4.4885640144348145,
      "eval_runtime": 490.2328,
      "eval_samples_per_second": 28.325,
      "eval_steps_per_second": 14.163,
      "num_input_tokens_seen": 29540768,
      "step": 21800
    },
    {
      "epoch": 0.6979386723001089,
      "grad_norm": 25.5,
      "learning_rate": 2.1469693747893355e-05,
      "loss": 4.5884,
      "num_input_tokens_seen": 29547536,
      "step": 21805
    },
    {
      "epoch": 0.6980987132705972,
      "grad_norm": 32.25,
      "learning_rate": 2.1459974920875274e-05,
      "loss": 5.1451,
      "num_input_tokens_seen": 29554128,
      "step": 21810
    },
    {
      "epoch": 0.6982587542410857,
      "grad_norm": 24.375,
      "learning_rate": 2.145025663977354e-05,
      "loss": 4.5866,
      "num_input_tokens_seen": 29560864,
      "step": 21815
    },
    {
      "epoch": 0.6984187952115741,
      "grad_norm": 12.3125,
      "learning_rate": 2.1440538906086844e-05,
      "loss": 4.4479,
      "num_input_tokens_seen": 29567088,
      "step": 21820
    },
    {
      "epoch": 0.6985788361820626,
      "grad_norm": 32.25,
      "learning_rate": 2.1430821721313782e-05,
      "loss": 4.5558,
      "num_input_tokens_seen": 29573984,
      "step": 21825
    },
    {
      "epoch": 0.698738877152551,
      "grad_norm": 24.125,
      "learning_rate": 2.142110508695286e-05,
      "loss": 4.9153,
      "num_input_tokens_seen": 29581104,
      "step": 21830
    },
    {
      "epoch": 0.6988989181230395,
      "grad_norm": 35.0,
      "learning_rate": 2.1411389004502515e-05,
      "loss": 4.8935,
      "num_input_tokens_seen": 29587968,
      "step": 21835
    },
    {
      "epoch": 0.699058959093528,
      "grad_norm": 19.75,
      "learning_rate": 2.140167347546107e-05,
      "loss": 3.9586,
      "num_input_tokens_seen": 29594720,
      "step": 21840
    },
    {
      "epoch": 0.6992190000640164,
      "grad_norm": 24.0,
      "learning_rate": 2.1391958501326793e-05,
      "loss": 5.3243,
      "num_input_tokens_seen": 29601648,
      "step": 21845
    },
    {
      "epoch": 0.6993790410345049,
      "grad_norm": 22.125,
      "learning_rate": 2.1382244083597873e-05,
      "loss": 3.7651,
      "num_input_tokens_seen": 29608032,
      "step": 21850
    },
    {
      "epoch": 0.6995390820049933,
      "grad_norm": 18.125,
      "learning_rate": 2.137253022377237e-05,
      "loss": 4.5648,
      "num_input_tokens_seen": 29614336,
      "step": 21855
    },
    {
      "epoch": 0.6996991229754818,
      "grad_norm": 29.5,
      "learning_rate": 2.136281692334829e-05,
      "loss": 4.8861,
      "num_input_tokens_seen": 29621808,
      "step": 21860
    },
    {
      "epoch": 0.6998591639459701,
      "grad_norm": 22.25,
      "learning_rate": 2.135310418382356e-05,
      "loss": 3.2626,
      "num_input_tokens_seen": 29628624,
      "step": 21865
    },
    {
      "epoch": 0.7000192049164586,
      "grad_norm": 28.125,
      "learning_rate": 2.134339200669598e-05,
      "loss": 3.9274,
      "num_input_tokens_seen": 29635376,
      "step": 21870
    },
    {
      "epoch": 0.700179245886947,
      "grad_norm": 32.75,
      "learning_rate": 2.133368039346331e-05,
      "loss": 3.8375,
      "num_input_tokens_seen": 29642368,
      "step": 21875
    },
    {
      "epoch": 0.7003392868574355,
      "grad_norm": 45.5,
      "learning_rate": 2.1323969345623195e-05,
      "loss": 4.9148,
      "num_input_tokens_seen": 29649264,
      "step": 21880
    },
    {
      "epoch": 0.7004993278279239,
      "grad_norm": 34.0,
      "learning_rate": 2.1314258864673207e-05,
      "loss": 3.6816,
      "num_input_tokens_seen": 29655664,
      "step": 21885
    },
    {
      "epoch": 0.7006593687984124,
      "grad_norm": 24.875,
      "learning_rate": 2.130454895211082e-05,
      "loss": 4.636,
      "num_input_tokens_seen": 29662336,
      "step": 21890
    },
    {
      "epoch": 0.7008194097689009,
      "grad_norm": 32.75,
      "learning_rate": 2.129483960943342e-05,
      "loss": 3.9966,
      "num_input_tokens_seen": 29669072,
      "step": 21895
    },
    {
      "epoch": 0.7009794507393893,
      "grad_norm": 26.125,
      "learning_rate": 2.128513083813831e-05,
      "loss": 4.8207,
      "num_input_tokens_seen": 29676336,
      "step": 21900
    },
    {
      "epoch": 0.7011394917098778,
      "grad_norm": 20.125,
      "learning_rate": 2.1275422639722724e-05,
      "loss": 4.8375,
      "num_input_tokens_seen": 29683104,
      "step": 21905
    },
    {
      "epoch": 0.7012995326803662,
      "grad_norm": 24.25,
      "learning_rate": 2.126571501568376e-05,
      "loss": 3.8648,
      "num_input_tokens_seen": 29690256,
      "step": 21910
    },
    {
      "epoch": 0.7014595736508547,
      "grad_norm": 21.875,
      "learning_rate": 2.1256007967518478e-05,
      "loss": 4.7088,
      "num_input_tokens_seen": 29697040,
      "step": 21915
    },
    {
      "epoch": 0.701619614621343,
      "grad_norm": 38.5,
      "learning_rate": 2.124630149672381e-05,
      "loss": 4.3598,
      "num_input_tokens_seen": 29703984,
      "step": 21920
    },
    {
      "epoch": 0.7017796555918315,
      "grad_norm": 17.0,
      "learning_rate": 2.1236595604796624e-05,
      "loss": 3.8699,
      "num_input_tokens_seen": 29710736,
      "step": 21925
    },
    {
      "epoch": 0.7019396965623199,
      "grad_norm": 22.625,
      "learning_rate": 2.1226890293233693e-05,
      "loss": 3.8064,
      "num_input_tokens_seen": 29717040,
      "step": 21930
    },
    {
      "epoch": 0.7020997375328084,
      "grad_norm": 26.25,
      "learning_rate": 2.1217185563531694e-05,
      "loss": 4.4751,
      "num_input_tokens_seen": 29723152,
      "step": 21935
    },
    {
      "epoch": 0.7022597785032968,
      "grad_norm": 22.75,
      "learning_rate": 2.120748141718721e-05,
      "loss": 3.8367,
      "num_input_tokens_seen": 29729936,
      "step": 21940
    },
    {
      "epoch": 0.7024198194737853,
      "grad_norm": 26.125,
      "learning_rate": 2.1197777855696765e-05,
      "loss": 3.8753,
      "num_input_tokens_seen": 29736736,
      "step": 21945
    },
    {
      "epoch": 0.7025798604442738,
      "grad_norm": 33.5,
      "learning_rate": 2.1188074880556746e-05,
      "loss": 4.8992,
      "num_input_tokens_seen": 29743392,
      "step": 21950
    },
    {
      "epoch": 0.7027399014147622,
      "grad_norm": 27.875,
      "learning_rate": 2.1178372493263495e-05,
      "loss": 3.7438,
      "num_input_tokens_seen": 29750096,
      "step": 21955
    },
    {
      "epoch": 0.7028999423852507,
      "grad_norm": 31.375,
      "learning_rate": 2.116867069531322e-05,
      "loss": 3.9151,
      "num_input_tokens_seen": 29757184,
      "step": 21960
    },
    {
      "epoch": 0.703059983355739,
      "grad_norm": 25.0,
      "learning_rate": 2.1158969488202073e-05,
      "loss": 3.5565,
      "num_input_tokens_seen": 29763600,
      "step": 21965
    },
    {
      "epoch": 0.7032200243262275,
      "grad_norm": 47.5,
      "learning_rate": 2.114926887342611e-05,
      "loss": 4.1813,
      "num_input_tokens_seen": 29770304,
      "step": 21970
    },
    {
      "epoch": 0.7033800652967159,
      "grad_norm": 31.0,
      "learning_rate": 2.113956885248127e-05,
      "loss": 5.429,
      "num_input_tokens_seen": 29777424,
      "step": 21975
    },
    {
      "epoch": 0.7035401062672044,
      "grad_norm": 25.0,
      "learning_rate": 2.112986942686342e-05,
      "loss": 4.8762,
      "num_input_tokens_seen": 29784912,
      "step": 21980
    },
    {
      "epoch": 0.7037001472376928,
      "grad_norm": 26.5,
      "learning_rate": 2.112017059806835e-05,
      "loss": 4.5304,
      "num_input_tokens_seen": 29791616,
      "step": 21985
    },
    {
      "epoch": 0.7038601882081813,
      "grad_norm": 21.125,
      "learning_rate": 2.1110472367591724e-05,
      "loss": 5.3793,
      "num_input_tokens_seen": 29798432,
      "step": 21990
    },
    {
      "epoch": 0.7040202291786697,
      "grad_norm": 20.5,
      "learning_rate": 2.1100774736929145e-05,
      "loss": 4.5331,
      "num_input_tokens_seen": 29805168,
      "step": 21995
    },
    {
      "epoch": 0.7041802701491582,
      "grad_norm": 39.25,
      "learning_rate": 2.10910777075761e-05,
      "loss": 4.4407,
      "num_input_tokens_seen": 29812480,
      "step": 22000
    },
    {
      "epoch": 0.7041802701491582,
      "eval_loss": 4.493778228759766,
      "eval_runtime": 490.7143,
      "eval_samples_per_second": 28.298,
      "eval_steps_per_second": 14.149,
      "num_input_tokens_seen": 29812480,
      "step": 22000
    },
    {
      "epoch": 0.7043403111196467,
      "grad_norm": 21.75,
      "learning_rate": 2.108138128102799e-05,
      "loss": 4.8384,
      "num_input_tokens_seen": 29819264,
      "step": 22005
    },
    {
      "epoch": 0.7045003520901351,
      "grad_norm": 30.125,
      "learning_rate": 2.107168545878014e-05,
      "loss": 4.4927,
      "num_input_tokens_seen": 29825808,
      "step": 22010
    },
    {
      "epoch": 0.7046603930606236,
      "grad_norm": 40.0,
      "learning_rate": 2.106199024232775e-05,
      "loss": 3.9689,
      "num_input_tokens_seen": 29832640,
      "step": 22015
    },
    {
      "epoch": 0.7048204340311119,
      "grad_norm": 16.25,
      "learning_rate": 2.105229563316595e-05,
      "loss": 4.1939,
      "num_input_tokens_seen": 29839904,
      "step": 22020
    },
    {
      "epoch": 0.7049804750016004,
      "grad_norm": 29.5,
      "learning_rate": 2.1042601632789784e-05,
      "loss": 4.4228,
      "num_input_tokens_seen": 29847248,
      "step": 22025
    },
    {
      "epoch": 0.7051405159720888,
      "grad_norm": 23.0,
      "learning_rate": 2.103290824269417e-05,
      "loss": 3.6007,
      "num_input_tokens_seen": 29853840,
      "step": 22030
    },
    {
      "epoch": 0.7053005569425773,
      "grad_norm": 32.75,
      "learning_rate": 2.1023215464373965e-05,
      "loss": 4.8999,
      "num_input_tokens_seen": 29860352,
      "step": 22035
    },
    {
      "epoch": 0.7054605979130657,
      "grad_norm": 31.5,
      "learning_rate": 2.1013523299323908e-05,
      "loss": 4.3115,
      "num_input_tokens_seen": 29866512,
      "step": 22040
    },
    {
      "epoch": 0.7056206388835542,
      "grad_norm": 22.0,
      "learning_rate": 2.1003831749038654e-05,
      "loss": 4.2904,
      "num_input_tokens_seen": 29873424,
      "step": 22045
    },
    {
      "epoch": 0.7057806798540426,
      "grad_norm": 30.625,
      "learning_rate": 2.099414081501277e-05,
      "loss": 4.49,
      "num_input_tokens_seen": 29879856,
      "step": 22050
    },
    {
      "epoch": 0.7059407208245311,
      "grad_norm": 33.25,
      "learning_rate": 2.09844504987407e-05,
      "loss": 3.7149,
      "num_input_tokens_seen": 29886816,
      "step": 22055
    },
    {
      "epoch": 0.7061007617950196,
      "grad_norm": 24.125,
      "learning_rate": 2.097476080171683e-05,
      "loss": 4.6867,
      "num_input_tokens_seen": 29893728,
      "step": 22060
    },
    {
      "epoch": 0.706260802765508,
      "grad_norm": 21.625,
      "learning_rate": 2.0965071725435436e-05,
      "loss": 4.514,
      "num_input_tokens_seen": 29900096,
      "step": 22065
    },
    {
      "epoch": 0.7064208437359965,
      "grad_norm": 30.75,
      "learning_rate": 2.0955383271390684e-05,
      "loss": 4.4345,
      "num_input_tokens_seen": 29907280,
      "step": 22070
    },
    {
      "epoch": 0.7065808847064848,
      "grad_norm": 25.875,
      "learning_rate": 2.094569544107666e-05,
      "loss": 3.5933,
      "num_input_tokens_seen": 29913904,
      "step": 22075
    },
    {
      "epoch": 0.7067409256769733,
      "grad_norm": 42.75,
      "learning_rate": 2.093600823598735e-05,
      "loss": 5.036,
      "num_input_tokens_seen": 29920608,
      "step": 22080
    },
    {
      "epoch": 0.7069009666474617,
      "grad_norm": 44.25,
      "learning_rate": 2.092632165761663e-05,
      "loss": 4.8412,
      "num_input_tokens_seen": 29927424,
      "step": 22085
    },
    {
      "epoch": 0.7070610076179502,
      "grad_norm": 24.375,
      "learning_rate": 2.091663570745832e-05,
      "loss": 4.615,
      "num_input_tokens_seen": 29934944,
      "step": 22090
    },
    {
      "epoch": 0.7072210485884386,
      "grad_norm": 47.75,
      "learning_rate": 2.0906950387006086e-05,
      "loss": 3.6576,
      "num_input_tokens_seen": 29941744,
      "step": 22095
    },
    {
      "epoch": 0.7073810895589271,
      "grad_norm": 41.0,
      "learning_rate": 2.0897265697753543e-05,
      "loss": 3.7226,
      "num_input_tokens_seen": 29948560,
      "step": 22100
    },
    {
      "epoch": 0.7075411305294156,
      "grad_norm": 89.5,
      "learning_rate": 2.088758164119419e-05,
      "loss": 3.1088,
      "num_input_tokens_seen": 29955088,
      "step": 22105
    },
    {
      "epoch": 0.707701171499904,
      "grad_norm": 29.125,
      "learning_rate": 2.0877898218821428e-05,
      "loss": 3.3781,
      "num_input_tokens_seen": 29961824,
      "step": 22110
    },
    {
      "epoch": 0.7078612124703925,
      "grad_norm": 21.25,
      "learning_rate": 2.0868215432128565e-05,
      "loss": 5.1172,
      "num_input_tokens_seen": 29968240,
      "step": 22115
    },
    {
      "epoch": 0.7080212534408808,
      "grad_norm": 31.375,
      "learning_rate": 2.0858533282608796e-05,
      "loss": 4.074,
      "num_input_tokens_seen": 29974784,
      "step": 22120
    },
    {
      "epoch": 0.7081812944113693,
      "grad_norm": 36.0,
      "learning_rate": 2.084885177175524e-05,
      "loss": 5.0892,
      "num_input_tokens_seen": 29981328,
      "step": 22125
    },
    {
      "epoch": 0.7083413353818577,
      "grad_norm": 21.375,
      "learning_rate": 2.0839170901060917e-05,
      "loss": 5.1704,
      "num_input_tokens_seen": 29988272,
      "step": 22130
    },
    {
      "epoch": 0.7085013763523462,
      "grad_norm": 32.5,
      "learning_rate": 2.082949067201872e-05,
      "loss": 3.9709,
      "num_input_tokens_seen": 29994992,
      "step": 22135
    },
    {
      "epoch": 0.7086614173228346,
      "grad_norm": 20.125,
      "learning_rate": 2.0819811086121475e-05,
      "loss": 3.8984,
      "num_input_tokens_seen": 30001616,
      "step": 22140
    },
    {
      "epoch": 0.7088214582933231,
      "grad_norm": 45.0,
      "learning_rate": 2.08101321448619e-05,
      "loss": 3.9608,
      "num_input_tokens_seen": 30008112,
      "step": 22145
    },
    {
      "epoch": 0.7089814992638115,
      "grad_norm": 23.25,
      "learning_rate": 2.080045384973259e-05,
      "loss": 4.1116,
      "num_input_tokens_seen": 30014592,
      "step": 22150
    },
    {
      "epoch": 0.7091415402343,
      "grad_norm": 19.25,
      "learning_rate": 2.0790776202226082e-05,
      "loss": 4.0605,
      "num_input_tokens_seen": 30021440,
      "step": 22155
    },
    {
      "epoch": 0.7093015812047885,
      "grad_norm": 31.25,
      "learning_rate": 2.078109920383477e-05,
      "loss": 3.8148,
      "num_input_tokens_seen": 30027984,
      "step": 22160
    },
    {
      "epoch": 0.7094616221752769,
      "grad_norm": 47.25,
      "learning_rate": 2.0771422856050978e-05,
      "loss": 4.7055,
      "num_input_tokens_seen": 30035008,
      "step": 22165
    },
    {
      "epoch": 0.7096216631457654,
      "grad_norm": 61.0,
      "learning_rate": 2.076174716036693e-05,
      "loss": 4.2396,
      "num_input_tokens_seen": 30041488,
      "step": 22170
    },
    {
      "epoch": 0.7097817041162537,
      "grad_norm": 38.0,
      "learning_rate": 2.075207211827472e-05,
      "loss": 4.3995,
      "num_input_tokens_seen": 30048064,
      "step": 22175
    },
    {
      "epoch": 0.7099417450867422,
      "grad_norm": 31.875,
      "learning_rate": 2.074239773126638e-05,
      "loss": 3.9091,
      "num_input_tokens_seen": 30054928,
      "step": 22180
    },
    {
      "epoch": 0.7101017860572306,
      "grad_norm": 54.5,
      "learning_rate": 2.073272400083382e-05,
      "loss": 4.7248,
      "num_input_tokens_seen": 30061424,
      "step": 22185
    },
    {
      "epoch": 0.7102618270277191,
      "grad_norm": 23.375,
      "learning_rate": 2.072305092846883e-05,
      "loss": 3.8313,
      "num_input_tokens_seen": 30067776,
      "step": 22190
    },
    {
      "epoch": 0.7104218679982075,
      "grad_norm": 14.9375,
      "learning_rate": 2.0713378515663152e-05,
      "loss": 4.0975,
      "num_input_tokens_seen": 30074304,
      "step": 22195
    },
    {
      "epoch": 0.710581908968696,
      "grad_norm": 31.625,
      "learning_rate": 2.070370676390836e-05,
      "loss": 4.5343,
      "num_input_tokens_seen": 30080624,
      "step": 22200
    },
    {
      "epoch": 0.710581908968696,
      "eval_loss": 4.500586986541748,
      "eval_runtime": 490.2277,
      "eval_samples_per_second": 28.326,
      "eval_steps_per_second": 14.163,
      "num_input_tokens_seen": 30080624,
      "step": 22200
    },
    {
      "epoch": 0.7107419499391844,
      "grad_norm": 31.0,
      "learning_rate": 2.0694035674695974e-05,
      "loss": 4.2389,
      "num_input_tokens_seen": 30087520,
      "step": 22205
    },
    {
      "epoch": 0.7109019909096729,
      "grad_norm": 27.875,
      "learning_rate": 2.0684365249517416e-05,
      "loss": 4.8777,
      "num_input_tokens_seen": 30093904,
      "step": 22210
    },
    {
      "epoch": 0.7110620318801614,
      "grad_norm": 33.75,
      "learning_rate": 2.067469548986396e-05,
      "loss": 4.5937,
      "num_input_tokens_seen": 30100576,
      "step": 22215
    },
    {
      "epoch": 0.7112220728506498,
      "grad_norm": 35.75,
      "learning_rate": 2.066502639722681e-05,
      "loss": 4.0035,
      "num_input_tokens_seen": 30107312,
      "step": 22220
    },
    {
      "epoch": 0.7113821138211383,
      "grad_norm": 37.5,
      "learning_rate": 2.065535797309708e-05,
      "loss": 5.0541,
      "num_input_tokens_seen": 30114304,
      "step": 22225
    },
    {
      "epoch": 0.7115421547916266,
      "grad_norm": 28.125,
      "learning_rate": 2.0645690218965736e-05,
      "loss": 5.1816,
      "num_input_tokens_seen": 30121120,
      "step": 22230
    },
    {
      "epoch": 0.7117021957621151,
      "grad_norm": 40.5,
      "learning_rate": 2.063602313632369e-05,
      "loss": 4.4374,
      "num_input_tokens_seen": 30127664,
      "step": 22235
    },
    {
      "epoch": 0.7118622367326035,
      "grad_norm": 29.625,
      "learning_rate": 2.0626356726661704e-05,
      "loss": 4.5463,
      "num_input_tokens_seen": 30135008,
      "step": 22240
    },
    {
      "epoch": 0.712022277703092,
      "grad_norm": 18.875,
      "learning_rate": 2.0616690991470477e-05,
      "loss": 3.931,
      "num_input_tokens_seen": 30141504,
      "step": 22245
    },
    {
      "epoch": 0.7121823186735804,
      "grad_norm": 23.25,
      "learning_rate": 2.0607025932240595e-05,
      "loss": 4.6295,
      "num_input_tokens_seen": 30147792,
      "step": 22250
    },
    {
      "epoch": 0.7123423596440689,
      "grad_norm": 26.875,
      "learning_rate": 2.059736155046251e-05,
      "loss": 4.5557,
      "num_input_tokens_seen": 30154672,
      "step": 22255
    },
    {
      "epoch": 0.7125024006145573,
      "grad_norm": 16.125,
      "learning_rate": 2.0587697847626603e-05,
      "loss": 4.0653,
      "num_input_tokens_seen": 30161216,
      "step": 22260
    },
    {
      "epoch": 0.7126624415850458,
      "grad_norm": 37.0,
      "learning_rate": 2.057803482522314e-05,
      "loss": 3.9234,
      "num_input_tokens_seen": 30168016,
      "step": 22265
    },
    {
      "epoch": 0.7128224825555343,
      "grad_norm": 11.75,
      "learning_rate": 2.056837248474227e-05,
      "loss": 4.584,
      "num_input_tokens_seen": 30175312,
      "step": 22270
    },
    {
      "epoch": 0.7129825235260226,
      "grad_norm": 32.75,
      "learning_rate": 2.0558710827674064e-05,
      "loss": 3.8243,
      "num_input_tokens_seen": 30181904,
      "step": 22275
    },
    {
      "epoch": 0.7131425644965111,
      "grad_norm": 42.0,
      "learning_rate": 2.054904985550845e-05,
      "loss": 4.0371,
      "num_input_tokens_seen": 30188544,
      "step": 22280
    },
    {
      "epoch": 0.7133026054669995,
      "grad_norm": 16.625,
      "learning_rate": 2.0539389569735287e-05,
      "loss": 4.1023,
      "num_input_tokens_seen": 30195232,
      "step": 22285
    },
    {
      "epoch": 0.713462646437488,
      "grad_norm": 44.75,
      "learning_rate": 2.052972997184431e-05,
      "loss": 4.6593,
      "num_input_tokens_seen": 30201680,
      "step": 22290
    },
    {
      "epoch": 0.7136226874079764,
      "grad_norm": 40.25,
      "learning_rate": 2.0520071063325146e-05,
      "loss": 3.817,
      "num_input_tokens_seen": 30208336,
      "step": 22295
    },
    {
      "epoch": 0.7137827283784649,
      "grad_norm": 35.0,
      "learning_rate": 2.051041284566732e-05,
      "loss": 5.1682,
      "num_input_tokens_seen": 30215008,
      "step": 22300
    },
    {
      "epoch": 0.7139427693489533,
      "grad_norm": 20.75,
      "learning_rate": 2.050075532036026e-05,
      "loss": 4.3504,
      "num_input_tokens_seen": 30221600,
      "step": 22305
    },
    {
      "epoch": 0.7141028103194418,
      "grad_norm": 16.375,
      "learning_rate": 2.0491098488893264e-05,
      "loss": 3.6786,
      "num_input_tokens_seen": 30228768,
      "step": 22310
    },
    {
      "epoch": 0.7142628512899302,
      "grad_norm": 14.1875,
      "learning_rate": 2.0481442352755546e-05,
      "loss": 4.1246,
      "num_input_tokens_seen": 30235424,
      "step": 22315
    },
    {
      "epoch": 0.7144228922604187,
      "grad_norm": 27.375,
      "learning_rate": 2.0471786913436198e-05,
      "loss": 4.0148,
      "num_input_tokens_seen": 30242160,
      "step": 22320
    },
    {
      "epoch": 0.7145829332309072,
      "grad_norm": 20.875,
      "learning_rate": 2.0462132172424218e-05,
      "loss": 3.3731,
      "num_input_tokens_seen": 30248928,
      "step": 22325
    },
    {
      "epoch": 0.7147429742013955,
      "grad_norm": 29.375,
      "learning_rate": 2.0452478131208484e-05,
      "loss": 4.46,
      "num_input_tokens_seen": 30255536,
      "step": 22330
    },
    {
      "epoch": 0.714903015171884,
      "grad_norm": 32.5,
      "learning_rate": 2.0442824791277765e-05,
      "loss": 4.9042,
      "num_input_tokens_seen": 30262016,
      "step": 22335
    },
    {
      "epoch": 0.7150630561423724,
      "grad_norm": 29.125,
      "learning_rate": 2.0433172154120727e-05,
      "loss": 4.1949,
      "num_input_tokens_seen": 30269200,
      "step": 22340
    },
    {
      "epoch": 0.7152230971128609,
      "grad_norm": 21.75,
      "learning_rate": 2.0423520221225947e-05,
      "loss": 4.6976,
      "num_input_tokens_seen": 30275792,
      "step": 22345
    },
    {
      "epoch": 0.7153831380833493,
      "grad_norm": 29.0,
      "learning_rate": 2.0413868994081848e-05,
      "loss": 4.2436,
      "num_input_tokens_seen": 30282720,
      "step": 22350
    },
    {
      "epoch": 0.7155431790538378,
      "grad_norm": 33.0,
      "learning_rate": 2.0404218474176795e-05,
      "loss": 4.6499,
      "num_input_tokens_seen": 30289936,
      "step": 22355
    },
    {
      "epoch": 0.7157032200243262,
      "grad_norm": 34.0,
      "learning_rate": 2.0394568662999002e-05,
      "loss": 4.8729,
      "num_input_tokens_seen": 30296368,
      "step": 22360
    },
    {
      "epoch": 0.7158632609948147,
      "grad_norm": 25.625,
      "learning_rate": 2.0384919562036593e-05,
      "loss": 3.8073,
      "num_input_tokens_seen": 30303488,
      "step": 22365
    },
    {
      "epoch": 0.7160233019653032,
      "grad_norm": 22.75,
      "learning_rate": 2.0375271172777593e-05,
      "loss": 3.5203,
      "num_input_tokens_seen": 30310816,
      "step": 22370
    },
    {
      "epoch": 0.7161833429357916,
      "grad_norm": 41.25,
      "learning_rate": 2.0365623496709885e-05,
      "loss": 4.5957,
      "num_input_tokens_seen": 30317440,
      "step": 22375
    },
    {
      "epoch": 0.71634338390628,
      "grad_norm": 53.75,
      "learning_rate": 2.0355976535321283e-05,
      "loss": 4.3272,
      "num_input_tokens_seen": 30324512,
      "step": 22380
    },
    {
      "epoch": 0.7165034248767684,
      "grad_norm": 31.0,
      "learning_rate": 2.034633029009945e-05,
      "loss": 3.7375,
      "num_input_tokens_seen": 30331136,
      "step": 22385
    },
    {
      "epoch": 0.7166634658472569,
      "grad_norm": 21.25,
      "learning_rate": 2.0336684762531972e-05,
      "loss": 3.5939,
      "num_input_tokens_seen": 30338480,
      "step": 22390
    },
    {
      "epoch": 0.7168235068177453,
      "grad_norm": 30.75,
      "learning_rate": 2.032703995410631e-05,
      "loss": 4.2735,
      "num_input_tokens_seen": 30345424,
      "step": 22395
    },
    {
      "epoch": 0.7169835477882338,
      "grad_norm": 30.75,
      "learning_rate": 2.031739586630981e-05,
      "loss": 4.3453,
      "num_input_tokens_seen": 30352256,
      "step": 22400
    },
    {
      "epoch": 0.7169835477882338,
      "eval_loss": 4.502237319946289,
      "eval_runtime": 489.8868,
      "eval_samples_per_second": 28.345,
      "eval_steps_per_second": 14.173,
      "num_input_tokens_seen": 30352256,
      "step": 22400
    },
    {
      "epoch": 0.7171435887587222,
      "grad_norm": 24.75,
      "learning_rate": 2.0307752500629707e-05,
      "loss": 4.5542,
      "num_input_tokens_seen": 30359232,
      "step": 22405
    },
    {
      "epoch": 0.7173036297292107,
      "grad_norm": 19.5,
      "learning_rate": 2.0298109858553144e-05,
      "loss": 3.9688,
      "num_input_tokens_seen": 30366128,
      "step": 22410
    },
    {
      "epoch": 0.7174636706996991,
      "grad_norm": 27.0,
      "learning_rate": 2.028846794156712e-05,
      "loss": 5.043,
      "num_input_tokens_seen": 30373488,
      "step": 22415
    },
    {
      "epoch": 0.7176237116701876,
      "grad_norm": 28.25,
      "learning_rate": 2.027882675115856e-05,
      "loss": 4.637,
      "num_input_tokens_seen": 30380560,
      "step": 22420
    },
    {
      "epoch": 0.7177837526406761,
      "grad_norm": 30.875,
      "learning_rate": 2.026918628881423e-05,
      "loss": 5.0349,
      "num_input_tokens_seen": 30387504,
      "step": 22425
    },
    {
      "epoch": 0.7179437936111644,
      "grad_norm": 17.5,
      "learning_rate": 2.0259546556020833e-05,
      "loss": 3.8542,
      "num_input_tokens_seen": 30394016,
      "step": 22430
    },
    {
      "epoch": 0.7181038345816529,
      "grad_norm": 24.75,
      "learning_rate": 2.024990755426493e-05,
      "loss": 3.662,
      "num_input_tokens_seen": 30400912,
      "step": 22435
    },
    {
      "epoch": 0.7182638755521413,
      "grad_norm": 23.5,
      "learning_rate": 2.0240269285032975e-05,
      "loss": 4.5828,
      "num_input_tokens_seen": 30407552,
      "step": 22440
    },
    {
      "epoch": 0.7184239165226298,
      "grad_norm": 34.25,
      "learning_rate": 2.0230631749811306e-05,
      "loss": 3.3532,
      "num_input_tokens_seen": 30414528,
      "step": 22445
    },
    {
      "epoch": 0.7185839574931182,
      "grad_norm": 42.75,
      "learning_rate": 2.0220994950086162e-05,
      "loss": 4.9467,
      "num_input_tokens_seen": 30421264,
      "step": 22450
    },
    {
      "epoch": 0.7187439984636067,
      "grad_norm": 24.375,
      "learning_rate": 2.021135888734365e-05,
      "loss": 4.6436,
      "num_input_tokens_seen": 30428160,
      "step": 22455
    },
    {
      "epoch": 0.7189040394340951,
      "grad_norm": 54.25,
      "learning_rate": 2.0201723563069783e-05,
      "loss": 4.3803,
      "num_input_tokens_seen": 30434784,
      "step": 22460
    },
    {
      "epoch": 0.7190640804045836,
      "grad_norm": 27.25,
      "learning_rate": 2.0192088978750433e-05,
      "loss": 4.4147,
      "num_input_tokens_seen": 30441184,
      "step": 22465
    },
    {
      "epoch": 0.719224121375072,
      "grad_norm": 27.875,
      "learning_rate": 2.0182455135871385e-05,
      "loss": 3.9712,
      "num_input_tokens_seen": 30447760,
      "step": 22470
    },
    {
      "epoch": 0.7193841623455605,
      "grad_norm": 17.375,
      "learning_rate": 2.0172822035918305e-05,
      "loss": 3.4543,
      "num_input_tokens_seen": 30454544,
      "step": 22475
    },
    {
      "epoch": 0.719544203316049,
      "grad_norm": 30.75,
      "learning_rate": 2.016318968037671e-05,
      "loss": 4.4581,
      "num_input_tokens_seen": 30461136,
      "step": 22480
    },
    {
      "epoch": 0.7197042442865373,
      "grad_norm": 20.25,
      "learning_rate": 2.015355807073206e-05,
      "loss": 5.015,
      "num_input_tokens_seen": 30468096,
      "step": 22485
    },
    {
      "epoch": 0.7198642852570258,
      "grad_norm": 21.0,
      "learning_rate": 2.0143927208469664e-05,
      "loss": 3.1802,
      "num_input_tokens_seen": 30474512,
      "step": 22490
    },
    {
      "epoch": 0.7200243262275142,
      "grad_norm": 27.875,
      "learning_rate": 2.0134297095074708e-05,
      "loss": 3.685,
      "num_input_tokens_seen": 30481376,
      "step": 22495
    },
    {
      "epoch": 0.7201843671980027,
      "grad_norm": 43.25,
      "learning_rate": 2.0124667732032297e-05,
      "loss": 4.7962,
      "num_input_tokens_seen": 30488368,
      "step": 22500
    },
    {
      "epoch": 0.7203444081684911,
      "grad_norm": 16.875,
      "learning_rate": 2.011503912082738e-05,
      "loss": 4.6563,
      "num_input_tokens_seen": 30494960,
      "step": 22505
    },
    {
      "epoch": 0.7205044491389796,
      "grad_norm": 22.0,
      "learning_rate": 2.0105411262944823e-05,
      "loss": 4.3536,
      "num_input_tokens_seen": 30501648,
      "step": 22510
    },
    {
      "epoch": 0.720664490109468,
      "grad_norm": 31.5,
      "learning_rate": 2.0095784159869366e-05,
      "loss": 4.5078,
      "num_input_tokens_seen": 30508064,
      "step": 22515
    },
    {
      "epoch": 0.7208245310799565,
      "grad_norm": 26.5,
      "learning_rate": 2.0086157813085608e-05,
      "loss": 3.5527,
      "num_input_tokens_seen": 30514880,
      "step": 22520
    },
    {
      "epoch": 0.7209845720504449,
      "grad_norm": 21.375,
      "learning_rate": 2.0076532224078068e-05,
      "loss": 3.3457,
      "num_input_tokens_seen": 30521520,
      "step": 22525
    },
    {
      "epoch": 0.7211446130209334,
      "grad_norm": 20.375,
      "learning_rate": 2.0066907394331142e-05,
      "loss": 4.0437,
      "num_input_tokens_seen": 30528272,
      "step": 22530
    },
    {
      "epoch": 0.7213046539914219,
      "grad_norm": 24.875,
      "learning_rate": 2.0057283325329077e-05,
      "loss": 4.0732,
      "num_input_tokens_seen": 30534768,
      "step": 22535
    },
    {
      "epoch": 0.7214646949619102,
      "grad_norm": 37.25,
      "learning_rate": 2.0047660018556047e-05,
      "loss": 4.5266,
      "num_input_tokens_seen": 30541152,
      "step": 22540
    },
    {
      "epoch": 0.7216247359323987,
      "grad_norm": 27.125,
      "learning_rate": 2.0038037475496075e-05,
      "loss": 4.8626,
      "num_input_tokens_seen": 30547936,
      "step": 22545
    },
    {
      "epoch": 0.7217847769028871,
      "grad_norm": 34.75,
      "learning_rate": 2.0028415697633073e-05,
      "loss": 4.7355,
      "num_input_tokens_seen": 30554208,
      "step": 22550
    },
    {
      "epoch": 0.7219448178733756,
      "grad_norm": 18.125,
      "learning_rate": 2.0018794686450858e-05,
      "loss": 4.5325,
      "num_input_tokens_seen": 30560480,
      "step": 22555
    },
    {
      "epoch": 0.722104858843864,
      "grad_norm": 24.375,
      "learning_rate": 2.0009174443433088e-05,
      "loss": 3.6242,
      "num_input_tokens_seen": 30567232,
      "step": 22560
    },
    {
      "epoch": 0.7222648998143525,
      "grad_norm": 39.5,
      "learning_rate": 1.999955497006334e-05,
      "loss": 4.724,
      "num_input_tokens_seen": 30574144,
      "step": 22565
    },
    {
      "epoch": 0.7224249407848409,
      "grad_norm": 21.125,
      "learning_rate": 1.9989936267825067e-05,
      "loss": 5.0224,
      "num_input_tokens_seen": 30580896,
      "step": 22570
    },
    {
      "epoch": 0.7225849817553294,
      "grad_norm": 21.375,
      "learning_rate": 1.9980318338201572e-05,
      "loss": 3.868,
      "num_input_tokens_seen": 30588096,
      "step": 22575
    },
    {
      "epoch": 0.7227450227258178,
      "grad_norm": 35.0,
      "learning_rate": 1.997070118267607e-05,
      "loss": 3.5445,
      "num_input_tokens_seen": 30594800,
      "step": 22580
    },
    {
      "epoch": 0.7229050636963062,
      "grad_norm": 29.625,
      "learning_rate": 1.9961084802731654e-05,
      "loss": 5.833,
      "num_input_tokens_seen": 30601632,
      "step": 22585
    },
    {
      "epoch": 0.7230651046667947,
      "grad_norm": 43.5,
      "learning_rate": 1.9951469199851273e-05,
      "loss": 3.6214,
      "num_input_tokens_seen": 30608096,
      "step": 22590
    },
    {
      "epoch": 0.7232251456372831,
      "grad_norm": 25.5,
      "learning_rate": 1.99418543755178e-05,
      "loss": 3.8639,
      "num_input_tokens_seen": 30615088,
      "step": 22595
    },
    {
      "epoch": 0.7233851866077716,
      "grad_norm": 34.75,
      "learning_rate": 1.9932240331213936e-05,
      "loss": 4.6245,
      "num_input_tokens_seen": 30622032,
      "step": 22600
    },
    {
      "epoch": 0.7233851866077716,
      "eval_loss": 4.496070861816406,
      "eval_runtime": 489.8207,
      "eval_samples_per_second": 28.349,
      "eval_steps_per_second": 14.175,
      "num_input_tokens_seen": 30622032,
      "step": 22600
    },
    {
      "epoch": 0.72354522757826,
      "grad_norm": 19.125,
      "learning_rate": 1.9922627068422297e-05,
      "loss": 4.018,
      "num_input_tokens_seen": 30628816,
      "step": 22605
    },
    {
      "epoch": 0.7237052685487485,
      "grad_norm": 27.875,
      "learning_rate": 1.991301458862538e-05,
      "loss": 4.227,
      "num_input_tokens_seen": 30635984,
      "step": 22610
    },
    {
      "epoch": 0.7238653095192369,
      "grad_norm": 28.625,
      "learning_rate": 1.9903402893305536e-05,
      "loss": 4.4456,
      "num_input_tokens_seen": 30642512,
      "step": 22615
    },
    {
      "epoch": 0.7240253504897254,
      "grad_norm": 68.0,
      "learning_rate": 1.9893791983945016e-05,
      "loss": 5.4762,
      "num_input_tokens_seen": 30649408,
      "step": 22620
    },
    {
      "epoch": 0.7241853914602138,
      "grad_norm": 25.5,
      "learning_rate": 1.988418186202594e-05,
      "loss": 5.3315,
      "num_input_tokens_seen": 30656096,
      "step": 22625
    },
    {
      "epoch": 0.7243454324307023,
      "grad_norm": 49.25,
      "learning_rate": 1.98745725290303e-05,
      "loss": 5.9195,
      "num_input_tokens_seen": 30662832,
      "step": 22630
    },
    {
      "epoch": 0.7245054734011908,
      "grad_norm": 44.75,
      "learning_rate": 1.986496398644e-05,
      "loss": 4.401,
      "num_input_tokens_seen": 30669744,
      "step": 22635
    },
    {
      "epoch": 0.7246655143716791,
      "grad_norm": 29.25,
      "learning_rate": 1.9855356235736777e-05,
      "loss": 5.1976,
      "num_input_tokens_seen": 30676752,
      "step": 22640
    },
    {
      "epoch": 0.7248255553421676,
      "grad_norm": 34.25,
      "learning_rate": 1.9845749278402277e-05,
      "loss": 5.3236,
      "num_input_tokens_seen": 30683136,
      "step": 22645
    },
    {
      "epoch": 0.724985596312656,
      "grad_norm": 38.0,
      "learning_rate": 1.9836143115918006e-05,
      "loss": 4.7444,
      "num_input_tokens_seen": 30690112,
      "step": 22650
    },
    {
      "epoch": 0.7251456372831445,
      "grad_norm": 44.5,
      "learning_rate": 1.9826537749765367e-05,
      "loss": 4.3829,
      "num_input_tokens_seen": 30696896,
      "step": 22655
    },
    {
      "epoch": 0.7253056782536329,
      "grad_norm": 33.75,
      "learning_rate": 1.9816933181425625e-05,
      "loss": 4.8044,
      "num_input_tokens_seen": 30703488,
      "step": 22660
    },
    {
      "epoch": 0.7254657192241214,
      "grad_norm": 32.5,
      "learning_rate": 1.9807329412379903e-05,
      "loss": 4.4577,
      "num_input_tokens_seen": 30710144,
      "step": 22665
    },
    {
      "epoch": 0.7256257601946098,
      "grad_norm": 30.375,
      "learning_rate": 1.9797726444109247e-05,
      "loss": 4.7956,
      "num_input_tokens_seen": 30716832,
      "step": 22670
    },
    {
      "epoch": 0.7257858011650983,
      "grad_norm": 47.75,
      "learning_rate": 1.9788124278094557e-05,
      "loss": 4.5499,
      "num_input_tokens_seen": 30723488,
      "step": 22675
    },
    {
      "epoch": 0.7259458421355867,
      "grad_norm": 26.5,
      "learning_rate": 1.9778522915816594e-05,
      "loss": 4.4085,
      "num_input_tokens_seen": 30730160,
      "step": 22680
    },
    {
      "epoch": 0.7261058831060752,
      "grad_norm": 18.75,
      "learning_rate": 1.9768922358756014e-05,
      "loss": 3.9212,
      "num_input_tokens_seen": 30736672,
      "step": 22685
    },
    {
      "epoch": 0.7262659240765637,
      "grad_norm": 22.375,
      "learning_rate": 1.9759322608393353e-05,
      "loss": 4.2978,
      "num_input_tokens_seen": 30743504,
      "step": 22690
    },
    {
      "epoch": 0.726425965047052,
      "grad_norm": 16.0,
      "learning_rate": 1.9749723666208992e-05,
      "loss": 4.4707,
      "num_input_tokens_seen": 30750256,
      "step": 22695
    },
    {
      "epoch": 0.7265860060175405,
      "grad_norm": 26.0,
      "learning_rate": 1.9740125533683235e-05,
      "loss": 4.4007,
      "num_input_tokens_seen": 30757056,
      "step": 22700
    },
    {
      "epoch": 0.7267460469880289,
      "grad_norm": 40.25,
      "learning_rate": 1.9730528212296208e-05,
      "loss": 4.8823,
      "num_input_tokens_seen": 30763744,
      "step": 22705
    },
    {
      "epoch": 0.7269060879585174,
      "grad_norm": 41.25,
      "learning_rate": 1.9720931703527945e-05,
      "loss": 4.1204,
      "num_input_tokens_seen": 30770432,
      "step": 22710
    },
    {
      "epoch": 0.7270661289290058,
      "grad_norm": 29.375,
      "learning_rate": 1.9711336008858373e-05,
      "loss": 5.3084,
      "num_input_tokens_seen": 30777328,
      "step": 22715
    },
    {
      "epoch": 0.7272261698994943,
      "grad_norm": 28.5,
      "learning_rate": 1.9701741129767233e-05,
      "loss": 4.9222,
      "num_input_tokens_seen": 30784048,
      "step": 22720
    },
    {
      "epoch": 0.7273862108699827,
      "grad_norm": 26.0,
      "learning_rate": 1.9692147067734202e-05,
      "loss": 4.2591,
      "num_input_tokens_seen": 30790704,
      "step": 22725
    },
    {
      "epoch": 0.7275462518404712,
      "grad_norm": 23.5,
      "learning_rate": 1.96825538242388e-05,
      "loss": 3.9468,
      "num_input_tokens_seen": 30798016,
      "step": 22730
    },
    {
      "epoch": 0.7277062928109596,
      "grad_norm": 22.625,
      "learning_rate": 1.967296140076041e-05,
      "loss": 3.9805,
      "num_input_tokens_seen": 30805088,
      "step": 22735
    },
    {
      "epoch": 0.727866333781448,
      "grad_norm": 48.0,
      "learning_rate": 1.966336979877833e-05,
      "loss": 4.0846,
      "num_input_tokens_seen": 30811744,
      "step": 22740
    },
    {
      "epoch": 0.7280263747519365,
      "grad_norm": 34.25,
      "learning_rate": 1.9653779019771678e-05,
      "loss": 3.4363,
      "num_input_tokens_seen": 30818496,
      "step": 22745
    },
    {
      "epoch": 0.7281864157224249,
      "grad_norm": 53.0,
      "learning_rate": 1.9644189065219488e-05,
      "loss": 4.7743,
      "num_input_tokens_seen": 30825408,
      "step": 22750
    },
    {
      "epoch": 0.7283464566929134,
      "grad_norm": 29.0,
      "learning_rate": 1.9634599936600655e-05,
      "loss": 4.7418,
      "num_input_tokens_seen": 30831920,
      "step": 22755
    },
    {
      "epoch": 0.7285064976634018,
      "grad_norm": 22.875,
      "learning_rate": 1.9625011635393935e-05,
      "loss": 4.5244,
      "num_input_tokens_seen": 30838896,
      "step": 22760
    },
    {
      "epoch": 0.7286665386338903,
      "grad_norm": 28.25,
      "learning_rate": 1.9615424163077963e-05,
      "loss": 4.4289,
      "num_input_tokens_seen": 30845392,
      "step": 22765
    },
    {
      "epoch": 0.7288265796043787,
      "grad_norm": 27.25,
      "learning_rate": 1.9605837521131263e-05,
      "loss": 4.0128,
      "num_input_tokens_seen": 30852528,
      "step": 22770
    },
    {
      "epoch": 0.7289866205748672,
      "grad_norm": 24.0,
      "learning_rate": 1.9596251711032192e-05,
      "loss": 4.3591,
      "num_input_tokens_seen": 30859232,
      "step": 22775
    },
    {
      "epoch": 0.7291466615453556,
      "grad_norm": 23.75,
      "learning_rate": 1.958666673425903e-05,
      "loss": 5.0677,
      "num_input_tokens_seen": 30866592,
      "step": 22780
    },
    {
      "epoch": 0.7293067025158441,
      "grad_norm": 30.125,
      "learning_rate": 1.957708259228987e-05,
      "loss": 3.9685,
      "num_input_tokens_seen": 30873152,
      "step": 22785
    },
    {
      "epoch": 0.7294667434863324,
      "grad_norm": 34.75,
      "learning_rate": 1.956749928660273e-05,
      "loss": 4.4859,
      "num_input_tokens_seen": 30880384,
      "step": 22790
    },
    {
      "epoch": 0.7296267844568209,
      "grad_norm": 20.625,
      "learning_rate": 1.955791681867547e-05,
      "loss": 3.2916,
      "num_input_tokens_seen": 30887168,
      "step": 22795
    },
    {
      "epoch": 0.7297868254273094,
      "grad_norm": 55.0,
      "learning_rate": 1.9548335189985824e-05,
      "loss": 4.41,
      "num_input_tokens_seen": 30894016,
      "step": 22800
    },
    {
      "epoch": 0.7297868254273094,
      "eval_loss": 4.486946105957031,
      "eval_runtime": 489.9148,
      "eval_samples_per_second": 28.344,
      "eval_steps_per_second": 14.172,
      "num_input_tokens_seen": 30894016,
      "step": 22800
    },
    {
      "epoch": 0.7299468663977978,
      "grad_norm": 47.0,
      "learning_rate": 1.9538754402011396e-05,
      "loss": 5.1439,
      "num_input_tokens_seen": 30900768,
      "step": 22805
    },
    {
      "epoch": 0.7301069073682863,
      "grad_norm": 26.625,
      "learning_rate": 1.952917445622968e-05,
      "loss": 4.4985,
      "num_input_tokens_seen": 30907744,
      "step": 22810
    },
    {
      "epoch": 0.7302669483387747,
      "grad_norm": 26.25,
      "learning_rate": 1.9519595354118005e-05,
      "loss": 3.7659,
      "num_input_tokens_seen": 30913904,
      "step": 22815
    },
    {
      "epoch": 0.7304269893092632,
      "grad_norm": 37.75,
      "learning_rate": 1.951001709715361e-05,
      "loss": 4.4566,
      "num_input_tokens_seen": 30920256,
      "step": 22820
    },
    {
      "epoch": 0.7305870302797516,
      "grad_norm": 16.125,
      "learning_rate": 1.9500439686813556e-05,
      "loss": 3.4655,
      "num_input_tokens_seen": 30926688,
      "step": 22825
    },
    {
      "epoch": 0.7307470712502401,
      "grad_norm": 27.0,
      "learning_rate": 1.949086312457482e-05,
      "loss": 4.1025,
      "num_input_tokens_seen": 30933392,
      "step": 22830
    },
    {
      "epoch": 0.7309071122207285,
      "grad_norm": 25.0,
      "learning_rate": 1.9481287411914223e-05,
      "loss": 4.5893,
      "num_input_tokens_seen": 30940032,
      "step": 22835
    },
    {
      "epoch": 0.731067153191217,
      "grad_norm": 24.125,
      "learning_rate": 1.9471712550308457e-05,
      "loss": 4.7704,
      "num_input_tokens_seen": 30947008,
      "step": 22840
    },
    {
      "epoch": 0.7312271941617055,
      "grad_norm": 33.5,
      "learning_rate": 1.946213854123409e-05,
      "loss": 3.7088,
      "num_input_tokens_seen": 30953968,
      "step": 22845
    },
    {
      "epoch": 0.7313872351321938,
      "grad_norm": 39.5,
      "learning_rate": 1.9452565386167554e-05,
      "loss": 4.3542,
      "num_input_tokens_seen": 30960656,
      "step": 22850
    },
    {
      "epoch": 0.7315472761026823,
      "grad_norm": 25.0,
      "learning_rate": 1.9442993086585142e-05,
      "loss": 4.5393,
      "num_input_tokens_seen": 30967168,
      "step": 22855
    },
    {
      "epoch": 0.7317073170731707,
      "grad_norm": 39.5,
      "learning_rate": 1.9433421643963043e-05,
      "loss": 3.9824,
      "num_input_tokens_seen": 30974128,
      "step": 22860
    },
    {
      "epoch": 0.7318673580436592,
      "grad_norm": 27.125,
      "learning_rate": 1.942385105977727e-05,
      "loss": 4.465,
      "num_input_tokens_seen": 30980656,
      "step": 22865
    },
    {
      "epoch": 0.7320273990141476,
      "grad_norm": 24.625,
      "learning_rate": 1.9414281335503743e-05,
      "loss": 4.0331,
      "num_input_tokens_seen": 30987360,
      "step": 22870
    },
    {
      "epoch": 0.7321874399846361,
      "grad_norm": 22.875,
      "learning_rate": 1.9404712472618232e-05,
      "loss": 4.3307,
      "num_input_tokens_seen": 30994224,
      "step": 22875
    },
    {
      "epoch": 0.7323474809551245,
      "grad_norm": 43.25,
      "learning_rate": 1.939514447259636e-05,
      "loss": 4.093,
      "num_input_tokens_seen": 31001024,
      "step": 22880
    },
    {
      "epoch": 0.732507521925613,
      "grad_norm": 35.75,
      "learning_rate": 1.938557733691365e-05,
      "loss": 4.7375,
      "num_input_tokens_seen": 31007552,
      "step": 22885
    },
    {
      "epoch": 0.7326675628961014,
      "grad_norm": 29.5,
      "learning_rate": 1.9376011067045476e-05,
      "loss": 4.5424,
      "num_input_tokens_seen": 31014304,
      "step": 22890
    },
    {
      "epoch": 0.7328276038665898,
      "grad_norm": 33.5,
      "learning_rate": 1.9366445664467065e-05,
      "loss": 4.7925,
      "num_input_tokens_seen": 31021104,
      "step": 22895
    },
    {
      "epoch": 0.7329876448370783,
      "grad_norm": 21.0,
      "learning_rate": 1.9356881130653533e-05,
      "loss": 4.3464,
      "num_input_tokens_seen": 31028032,
      "step": 22900
    },
    {
      "epoch": 0.7331476858075667,
      "grad_norm": 42.25,
      "learning_rate": 1.9347317467079846e-05,
      "loss": 3.8322,
      "num_input_tokens_seen": 31034496,
      "step": 22905
    },
    {
      "epoch": 0.7333077267780552,
      "grad_norm": 58.5,
      "learning_rate": 1.9337754675220836e-05,
      "loss": 4.0879,
      "num_input_tokens_seen": 31041152,
      "step": 22910
    },
    {
      "epoch": 0.7334677677485436,
      "grad_norm": 34.25,
      "learning_rate": 1.9328192756551218e-05,
      "loss": 3.2751,
      "num_input_tokens_seen": 31047760,
      "step": 22915
    },
    {
      "epoch": 0.7336278087190321,
      "grad_norm": 24.125,
      "learning_rate": 1.931863171254555e-05,
      "loss": 4.2503,
      "num_input_tokens_seen": 31054768,
      "step": 22920
    },
    {
      "epoch": 0.7337878496895205,
      "grad_norm": 27.0,
      "learning_rate": 1.930907154467826e-05,
      "loss": 4.1871,
      "num_input_tokens_seen": 31061344,
      "step": 22925
    },
    {
      "epoch": 0.733947890660009,
      "grad_norm": 21.875,
      "learning_rate": 1.9299512254423673e-05,
      "loss": 4.2461,
      "num_input_tokens_seen": 31068080,
      "step": 22930
    },
    {
      "epoch": 0.7341079316304974,
      "grad_norm": 26.125,
      "learning_rate": 1.9289953843255914e-05,
      "loss": 4.7822,
      "num_input_tokens_seen": 31074880,
      "step": 22935
    },
    {
      "epoch": 0.7342679726009859,
      "grad_norm": 34.0,
      "learning_rate": 1.9280396312649048e-05,
      "loss": 4.9401,
      "num_input_tokens_seen": 31081792,
      "step": 22940
    },
    {
      "epoch": 0.7344280135714742,
      "grad_norm": 30.625,
      "learning_rate": 1.9270839664076936e-05,
      "loss": 4.166,
      "num_input_tokens_seen": 31088544,
      "step": 22945
    },
    {
      "epoch": 0.7345880545419627,
      "grad_norm": 14.4375,
      "learning_rate": 1.9261283899013345e-05,
      "loss": 4.2981,
      "num_input_tokens_seen": 31095408,
      "step": 22950
    },
    {
      "epoch": 0.7347480955124512,
      "grad_norm": 14.8125,
      "learning_rate": 1.92517290189319e-05,
      "loss": 4.1413,
      "num_input_tokens_seen": 31102272,
      "step": 22955
    },
    {
      "epoch": 0.7349081364829396,
      "grad_norm": 46.75,
      "learning_rate": 1.924217502530607e-05,
      "loss": 5.1193,
      "num_input_tokens_seen": 31108640,
      "step": 22960
    },
    {
      "epoch": 0.7350681774534281,
      "grad_norm": 39.75,
      "learning_rate": 1.9232621919609207e-05,
      "loss": 4.6167,
      "num_input_tokens_seen": 31115168,
      "step": 22965
    },
    {
      "epoch": 0.7352282184239165,
      "grad_norm": 21.875,
      "learning_rate": 1.9223069703314534e-05,
      "loss": 4.3098,
      "num_input_tokens_seen": 31121744,
      "step": 22970
    },
    {
      "epoch": 0.735388259394405,
      "grad_norm": 32.5,
      "learning_rate": 1.92135183778951e-05,
      "loss": 4.5397,
      "num_input_tokens_seen": 31128560,
      "step": 22975
    },
    {
      "epoch": 0.7355483003648934,
      "grad_norm": 29.125,
      "learning_rate": 1.9203967944823857e-05,
      "loss": 3.7594,
      "num_input_tokens_seen": 31135408,
      "step": 22980
    },
    {
      "epoch": 0.7357083413353819,
      "grad_norm": 46.75,
      "learning_rate": 1.9194418405573588e-05,
      "loss": 5.2865,
      "num_input_tokens_seen": 31142112,
      "step": 22985
    },
    {
      "epoch": 0.7358683823058703,
      "grad_norm": 42.75,
      "learning_rate": 1.9184869761616954e-05,
      "loss": 4.3379,
      "num_input_tokens_seen": 31149232,
      "step": 22990
    },
    {
      "epoch": 0.7360284232763588,
      "grad_norm": 39.75,
      "learning_rate": 1.9175322014426495e-05,
      "loss": 3.9211,
      "num_input_tokens_seen": 31156368,
      "step": 22995
    },
    {
      "epoch": 0.7361884642468471,
      "grad_norm": 25.125,
      "learning_rate": 1.9165775165474565e-05,
      "loss": 4.678,
      "num_input_tokens_seen": 31162736,
      "step": 23000
    },
    {
      "epoch": 0.7361884642468471,
      "eval_loss": 4.488949298858643,
      "eval_runtime": 490.491,
      "eval_samples_per_second": 28.31,
      "eval_steps_per_second": 14.155,
      "num_input_tokens_seen": 31162736,
      "step": 23000
    },
    {
      "epoch": 0.7363485052173356,
      "grad_norm": 39.75,
      "learning_rate": 1.9156229216233434e-05,
      "loss": 4.0741,
      "num_input_tokens_seen": 31169120,
      "step": 23005
    },
    {
      "epoch": 0.7365085461878241,
      "grad_norm": 21.125,
      "learning_rate": 1.9146684168175184e-05,
      "loss": 4.7488,
      "num_input_tokens_seen": 31175760,
      "step": 23010
    },
    {
      "epoch": 0.7366685871583125,
      "grad_norm": 15.375,
      "learning_rate": 1.9137140022771796e-05,
      "loss": 4.4164,
      "num_input_tokens_seen": 31182720,
      "step": 23015
    },
    {
      "epoch": 0.736828628128801,
      "grad_norm": 22.5,
      "learning_rate": 1.9127596781495103e-05,
      "loss": 4.3488,
      "num_input_tokens_seen": 31189808,
      "step": 23020
    },
    {
      "epoch": 0.7369886690992894,
      "grad_norm": 31.5,
      "learning_rate": 1.9118054445816767e-05,
      "loss": 4.2308,
      "num_input_tokens_seen": 31197008,
      "step": 23025
    },
    {
      "epoch": 0.7371487100697779,
      "grad_norm": 24.875,
      "learning_rate": 1.9108513017208356e-05,
      "loss": 5.02,
      "num_input_tokens_seen": 31203552,
      "step": 23030
    },
    {
      "epoch": 0.7373087510402663,
      "grad_norm": 25.125,
      "learning_rate": 1.9098972497141287e-05,
      "loss": 4.2656,
      "num_input_tokens_seen": 31211408,
      "step": 23035
    },
    {
      "epoch": 0.7374687920107548,
      "grad_norm": 28.25,
      "learning_rate": 1.9089432887086806e-05,
      "loss": 3.7865,
      "num_input_tokens_seen": 31218016,
      "step": 23040
    },
    {
      "epoch": 0.7376288329812432,
      "grad_norm": 25.875,
      "learning_rate": 1.9079894188516056e-05,
      "loss": 5.033,
      "num_input_tokens_seen": 31225120,
      "step": 23045
    },
    {
      "epoch": 0.7377888739517316,
      "grad_norm": 30.125,
      "learning_rate": 1.907035640290002e-05,
      "loss": 4.2169,
      "num_input_tokens_seen": 31231744,
      "step": 23050
    },
    {
      "epoch": 0.73794891492222,
      "grad_norm": 45.0,
      "learning_rate": 1.9060819531709534e-05,
      "loss": 4.9522,
      "num_input_tokens_seen": 31238176,
      "step": 23055
    },
    {
      "epoch": 0.7381089558927085,
      "grad_norm": 16.25,
      "learning_rate": 1.9051283576415325e-05,
      "loss": 4.2145,
      "num_input_tokens_seen": 31244720,
      "step": 23060
    },
    {
      "epoch": 0.738268996863197,
      "grad_norm": 45.75,
      "learning_rate": 1.904174853848793e-05,
      "loss": 5.131,
      "num_input_tokens_seen": 31251568,
      "step": 23065
    },
    {
      "epoch": 0.7384290378336854,
      "grad_norm": 35.25,
      "learning_rate": 1.903221441939779e-05,
      "loss": 4.1637,
      "num_input_tokens_seen": 31258096,
      "step": 23070
    },
    {
      "epoch": 0.7385890788041739,
      "grad_norm": 27.375,
      "learning_rate": 1.9022681220615194e-05,
      "loss": 3.6289,
      "num_input_tokens_seen": 31264992,
      "step": 23075
    },
    {
      "epoch": 0.7387491197746623,
      "grad_norm": 55.75,
      "learning_rate": 1.9013148943610255e-05,
      "loss": 4.2146,
      "num_input_tokens_seen": 31271616,
      "step": 23080
    },
    {
      "epoch": 0.7389091607451508,
      "grad_norm": 14.3125,
      "learning_rate": 1.9003617589852998e-05,
      "loss": 3.4797,
      "num_input_tokens_seen": 31278272,
      "step": 23085
    },
    {
      "epoch": 0.7390692017156392,
      "grad_norm": 31.5,
      "learning_rate": 1.899408716081326e-05,
      "loss": 4.6528,
      "num_input_tokens_seen": 31285040,
      "step": 23090
    },
    {
      "epoch": 0.7392292426861277,
      "grad_norm": 21.125,
      "learning_rate": 1.898455765796075e-05,
      "loss": 4.6194,
      "num_input_tokens_seen": 31292160,
      "step": 23095
    },
    {
      "epoch": 0.739389283656616,
      "grad_norm": 20.0,
      "learning_rate": 1.8975029082765053e-05,
      "loss": 5.0545,
      "num_input_tokens_seen": 31298384,
      "step": 23100
    },
    {
      "epoch": 0.7395493246271045,
      "grad_norm": 25.875,
      "learning_rate": 1.8965501436695577e-05,
      "loss": 4.5,
      "num_input_tokens_seen": 31304944,
      "step": 23105
    },
    {
      "epoch": 0.739709365597593,
      "grad_norm": 33.25,
      "learning_rate": 1.895597472122161e-05,
      "loss": 3.8414,
      "num_input_tokens_seen": 31311520,
      "step": 23110
    },
    {
      "epoch": 0.7398694065680814,
      "grad_norm": 45.5,
      "learning_rate": 1.894644893781231e-05,
      "loss": 4.4038,
      "num_input_tokens_seen": 31318128,
      "step": 23115
    },
    {
      "epoch": 0.7400294475385699,
      "grad_norm": 19.625,
      "learning_rate": 1.893692408793665e-05,
      "loss": 4.3366,
      "num_input_tokens_seen": 31324544,
      "step": 23120
    },
    {
      "epoch": 0.7401894885090583,
      "grad_norm": 17.875,
      "learning_rate": 1.8927400173063493e-05,
      "loss": 4.2826,
      "num_input_tokens_seen": 31331328,
      "step": 23125
    },
    {
      "epoch": 0.7403495294795468,
      "grad_norm": 22.375,
      "learning_rate": 1.891787719466154e-05,
      "loss": 3.5501,
      "num_input_tokens_seen": 31337936,
      "step": 23130
    },
    {
      "epoch": 0.7405095704500352,
      "grad_norm": 20.75,
      "learning_rate": 1.8908355154199346e-05,
      "loss": 4.3242,
      "num_input_tokens_seen": 31344736,
      "step": 23135
    },
    {
      "epoch": 0.7406696114205237,
      "grad_norm": 26.5,
      "learning_rate": 1.8898834053145357e-05,
      "loss": 4.1937,
      "num_input_tokens_seen": 31351712,
      "step": 23140
    },
    {
      "epoch": 0.7408296523910121,
      "grad_norm": 21.5,
      "learning_rate": 1.8889313892967813e-05,
      "loss": 3.3144,
      "num_input_tokens_seen": 31358528,
      "step": 23145
    },
    {
      "epoch": 0.7409896933615006,
      "grad_norm": 27.875,
      "learning_rate": 1.8879794675134863e-05,
      "loss": 4.3211,
      "num_input_tokens_seen": 31365456,
      "step": 23150
    },
    {
      "epoch": 0.7411497343319889,
      "grad_norm": 35.0,
      "learning_rate": 1.8870276401114494e-05,
      "loss": 4.8535,
      "num_input_tokens_seen": 31372080,
      "step": 23155
    },
    {
      "epoch": 0.7413097753024774,
      "grad_norm": 31.125,
      "learning_rate": 1.886075907237453e-05,
      "loss": 5.0812,
      "num_input_tokens_seen": 31378416,
      "step": 23160
    },
    {
      "epoch": 0.7414698162729659,
      "grad_norm": 37.25,
      "learning_rate": 1.8851242690382672e-05,
      "loss": 5.1511,
      "num_input_tokens_seen": 31385168,
      "step": 23165
    },
    {
      "epoch": 0.7416298572434543,
      "grad_norm": 44.75,
      "learning_rate": 1.884172725660645e-05,
      "loss": 5.2598,
      "num_input_tokens_seen": 31392144,
      "step": 23170
    },
    {
      "epoch": 0.7417898982139428,
      "grad_norm": 28.125,
      "learning_rate": 1.8832212772513277e-05,
      "loss": 3.9086,
      "num_input_tokens_seen": 31398816,
      "step": 23175
    },
    {
      "epoch": 0.7419499391844312,
      "grad_norm": 12.9375,
      "learning_rate": 1.8822699239570414e-05,
      "loss": 3.8311,
      "num_input_tokens_seen": 31405840,
      "step": 23180
    },
    {
      "epoch": 0.7421099801549197,
      "grad_norm": 20.625,
      "learning_rate": 1.8813186659244943e-05,
      "loss": 4.0955,
      "num_input_tokens_seen": 31412720,
      "step": 23185
    },
    {
      "epoch": 0.7422700211254081,
      "grad_norm": 24.125,
      "learning_rate": 1.880367503300385e-05,
      "loss": 4.425,
      "num_input_tokens_seen": 31419440,
      "step": 23190
    },
    {
      "epoch": 0.7424300620958966,
      "grad_norm": 42.5,
      "learning_rate": 1.8794164362313927e-05,
      "loss": 4.0916,
      "num_input_tokens_seen": 31426560,
      "step": 23195
    },
    {
      "epoch": 0.742590103066385,
      "grad_norm": 39.25,
      "learning_rate": 1.878465464864185e-05,
      "loss": 5.3991,
      "num_input_tokens_seen": 31433344,
      "step": 23200
    },
    {
      "epoch": 0.742590103066385,
      "eval_loss": 4.491044998168945,
      "eval_runtime": 490.0237,
      "eval_samples_per_second": 28.337,
      "eval_steps_per_second": 14.169,
      "num_input_tokens_seen": 31433344,
      "step": 23200
    },
    {
      "epoch": 0.7427501440368734,
      "grad_norm": 35.75,
      "learning_rate": 1.877514589345414e-05,
      "loss": 4.0014,
      "num_input_tokens_seen": 31440400,
      "step": 23205
    },
    {
      "epoch": 0.7429101850073618,
      "grad_norm": 56.0,
      "learning_rate": 1.876563809821715e-05,
      "loss": 5.2851,
      "num_input_tokens_seen": 31447136,
      "step": 23210
    },
    {
      "epoch": 0.7430702259778503,
      "grad_norm": 25.5,
      "learning_rate": 1.8756131264397106e-05,
      "loss": 3.9256,
      "num_input_tokens_seen": 31453904,
      "step": 23215
    },
    {
      "epoch": 0.7432302669483388,
      "grad_norm": 34.0,
      "learning_rate": 1.87466253934601e-05,
      "loss": 4.9204,
      "num_input_tokens_seen": 31460976,
      "step": 23220
    },
    {
      "epoch": 0.7433903079188272,
      "grad_norm": 27.125,
      "learning_rate": 1.8737120486872033e-05,
      "loss": 4.4722,
      "num_input_tokens_seen": 31468000,
      "step": 23225
    },
    {
      "epoch": 0.7435503488893157,
      "grad_norm": 17.5,
      "learning_rate": 1.8727616546098696e-05,
      "loss": 4.405,
      "num_input_tokens_seen": 31474576,
      "step": 23230
    },
    {
      "epoch": 0.7437103898598041,
      "grad_norm": 44.0,
      "learning_rate": 1.8718113572605716e-05,
      "loss": 4.8878,
      "num_input_tokens_seen": 31481456,
      "step": 23235
    },
    {
      "epoch": 0.7438704308302926,
      "grad_norm": 36.5,
      "learning_rate": 1.8708611567858554e-05,
      "loss": 4.3604,
      "num_input_tokens_seen": 31488208,
      "step": 23240
    },
    {
      "epoch": 0.744030471800781,
      "grad_norm": 41.75,
      "learning_rate": 1.8699110533322565e-05,
      "loss": 4.1698,
      "num_input_tokens_seen": 31495712,
      "step": 23245
    },
    {
      "epoch": 0.7441905127712695,
      "grad_norm": 27.25,
      "learning_rate": 1.8689610470462897e-05,
      "loss": 4.6081,
      "num_input_tokens_seen": 31502576,
      "step": 23250
    },
    {
      "epoch": 0.7443505537417578,
      "grad_norm": 32.0,
      "learning_rate": 1.8680111380744604e-05,
      "loss": 3.6207,
      "num_input_tokens_seen": 31509216,
      "step": 23255
    },
    {
      "epoch": 0.7445105947122463,
      "grad_norm": 20.875,
      "learning_rate": 1.8670613265632564e-05,
      "loss": 4.9795,
      "num_input_tokens_seen": 31515424,
      "step": 23260
    },
    {
      "epoch": 0.7446706356827347,
      "grad_norm": 15.9375,
      "learning_rate": 1.866111612659149e-05,
      "loss": 3.7693,
      "num_input_tokens_seen": 31522016,
      "step": 23265
    },
    {
      "epoch": 0.7448306766532232,
      "grad_norm": 17.625,
      "learning_rate": 1.8651619965085967e-05,
      "loss": 4.2864,
      "num_input_tokens_seen": 31528848,
      "step": 23270
    },
    {
      "epoch": 0.7449907176237117,
      "grad_norm": 47.0,
      "learning_rate": 1.8642124782580433e-05,
      "loss": 3.8888,
      "num_input_tokens_seen": 31535936,
      "step": 23275
    },
    {
      "epoch": 0.7451507585942001,
      "grad_norm": 17.25,
      "learning_rate": 1.8632630580539144e-05,
      "loss": 3.767,
      "num_input_tokens_seen": 31543008,
      "step": 23280
    },
    {
      "epoch": 0.7453107995646886,
      "grad_norm": 38.0,
      "learning_rate": 1.862313736042625e-05,
      "loss": 3.3875,
      "num_input_tokens_seen": 31550208,
      "step": 23285
    },
    {
      "epoch": 0.745470840535177,
      "grad_norm": 43.5,
      "learning_rate": 1.8613645123705703e-05,
      "loss": 4.681,
      "num_input_tokens_seen": 31556752,
      "step": 23290
    },
    {
      "epoch": 0.7456308815056655,
      "grad_norm": 22.25,
      "learning_rate": 1.8604153871841328e-05,
      "loss": 4.7461,
      "num_input_tokens_seen": 31563680,
      "step": 23295
    },
    {
      "epoch": 0.7457909224761539,
      "grad_norm": 33.75,
      "learning_rate": 1.859466360629682e-05,
      "loss": 4.7378,
      "num_input_tokens_seen": 31570384,
      "step": 23300
    },
    {
      "epoch": 0.7459509634466424,
      "grad_norm": 31.5,
      "learning_rate": 1.8585174328535666e-05,
      "loss": 4.0922,
      "num_input_tokens_seen": 31577440,
      "step": 23305
    },
    {
      "epoch": 0.7461110044171307,
      "grad_norm": 22.375,
      "learning_rate": 1.857568604002124e-05,
      "loss": 4.2712,
      "num_input_tokens_seen": 31584512,
      "step": 23310
    },
    {
      "epoch": 0.7462710453876192,
      "grad_norm": 26.375,
      "learning_rate": 1.8566198742216774e-05,
      "loss": 4.3134,
      "num_input_tokens_seen": 31591344,
      "step": 23315
    },
    {
      "epoch": 0.7464310863581076,
      "grad_norm": 35.5,
      "learning_rate": 1.85567124365853e-05,
      "loss": 4.1804,
      "num_input_tokens_seen": 31598160,
      "step": 23320
    },
    {
      "epoch": 0.7465911273285961,
      "grad_norm": 26.5,
      "learning_rate": 1.854722712458975e-05,
      "loss": 4.0935,
      "num_input_tokens_seen": 31605072,
      "step": 23325
    },
    {
      "epoch": 0.7467511682990846,
      "grad_norm": 16.25,
      "learning_rate": 1.853774280769286e-05,
      "loss": 3.9031,
      "num_input_tokens_seen": 31611760,
      "step": 23330
    },
    {
      "epoch": 0.746911209269573,
      "grad_norm": 22.75,
      "learning_rate": 1.852825948735724e-05,
      "loss": 4.0942,
      "num_input_tokens_seen": 31619264,
      "step": 23335
    },
    {
      "epoch": 0.7470712502400615,
      "grad_norm": 31.875,
      "learning_rate": 1.851877716504534e-05,
      "loss": 4.5316,
      "num_input_tokens_seen": 31626384,
      "step": 23340
    },
    {
      "epoch": 0.7472312912105499,
      "grad_norm": 31.75,
      "learning_rate": 1.8509295842219448e-05,
      "loss": 5.9473,
      "num_input_tokens_seen": 31632624,
      "step": 23345
    },
    {
      "epoch": 0.7473913321810384,
      "grad_norm": 20.375,
      "learning_rate": 1.8499815520341697e-05,
      "loss": 4.3876,
      "num_input_tokens_seen": 31639584,
      "step": 23350
    },
    {
      "epoch": 0.7475513731515268,
      "grad_norm": 22.875,
      "learning_rate": 1.8490336200874094e-05,
      "loss": 3.9249,
      "num_input_tokens_seen": 31646176,
      "step": 23355
    },
    {
      "epoch": 0.7477114141220152,
      "grad_norm": 14.9375,
      "learning_rate": 1.848085788527844e-05,
      "loss": 4.0646,
      "num_input_tokens_seen": 31652896,
      "step": 23360
    },
    {
      "epoch": 0.7478714550925036,
      "grad_norm": 26.25,
      "learning_rate": 1.847138057501644e-05,
      "loss": 4.7258,
      "num_input_tokens_seen": 31659472,
      "step": 23365
    },
    {
      "epoch": 0.7480314960629921,
      "grad_norm": 15.3125,
      "learning_rate": 1.8461904271549582e-05,
      "loss": 4.2277,
      "num_input_tokens_seen": 31666672,
      "step": 23370
    },
    {
      "epoch": 0.7481915370334806,
      "grad_norm": 30.25,
      "learning_rate": 1.845242897633926e-05,
      "loss": 4.1675,
      "num_input_tokens_seen": 31673472,
      "step": 23375
    },
    {
      "epoch": 0.748351578003969,
      "grad_norm": 20.0,
      "learning_rate": 1.844295469084667e-05,
      "loss": 4.357,
      "num_input_tokens_seen": 31680464,
      "step": 23380
    },
    {
      "epoch": 0.7485116189744575,
      "grad_norm": 37.25,
      "learning_rate": 1.843348141653286e-05,
      "loss": 4.608,
      "num_input_tokens_seen": 31687264,
      "step": 23385
    },
    {
      "epoch": 0.7486716599449459,
      "grad_norm": 32.75,
      "learning_rate": 1.842400915485874e-05,
      "loss": 4.1811,
      "num_input_tokens_seen": 31694128,
      "step": 23390
    },
    {
      "epoch": 0.7488317009154344,
      "grad_norm": 18.75,
      "learning_rate": 1.8414537907285053e-05,
      "loss": 3.8715,
      "num_input_tokens_seen": 31701168,
      "step": 23395
    },
    {
      "epoch": 0.7489917418859228,
      "grad_norm": 16.75,
      "learning_rate": 1.840506767527237e-05,
      "loss": 4.5956,
      "num_input_tokens_seen": 31708288,
      "step": 23400
    },
    {
      "epoch": 0.7489917418859228,
      "eval_loss": 4.492280960083008,
      "eval_runtime": 490.0292,
      "eval_samples_per_second": 28.337,
      "eval_steps_per_second": 14.169,
      "num_input_tokens_seen": 31708288,
      "step": 23400
    },
    {
      "epoch": 0.7491517828564113,
      "grad_norm": 23.5,
      "learning_rate": 1.8395598460281137e-05,
      "loss": 4.5207,
      "num_input_tokens_seen": 31715040,
      "step": 23405
    },
    {
      "epoch": 0.7493118238268996,
      "grad_norm": 21.375,
      "learning_rate": 1.838613026377161e-05,
      "loss": 3.8892,
      "num_input_tokens_seen": 31721920,
      "step": 23410
    },
    {
      "epoch": 0.7494718647973881,
      "grad_norm": 41.25,
      "learning_rate": 1.8376663087203917e-05,
      "loss": 4.5771,
      "num_input_tokens_seen": 31728352,
      "step": 23415
    },
    {
      "epoch": 0.7496319057678765,
      "grad_norm": 28.625,
      "learning_rate": 1.8367196932038014e-05,
      "loss": 4.9176,
      "num_input_tokens_seen": 31735488,
      "step": 23420
    },
    {
      "epoch": 0.749791946738365,
      "grad_norm": 27.25,
      "learning_rate": 1.8357731799733686e-05,
      "loss": 4.8568,
      "num_input_tokens_seen": 31742336,
      "step": 23425
    },
    {
      "epoch": 0.7499519877088535,
      "grad_norm": 18.375,
      "learning_rate": 1.8348267691750586e-05,
      "loss": 5.2845,
      "num_input_tokens_seen": 31748752,
      "step": 23430
    },
    {
      "epoch": 0.7501120286793419,
      "grad_norm": 22.5,
      "learning_rate": 1.833880460954821e-05,
      "loss": 4.5836,
      "num_input_tokens_seen": 31755376,
      "step": 23435
    },
    {
      "epoch": 0.7502720696498304,
      "grad_norm": 31.875,
      "learning_rate": 1.8329342554585866e-05,
      "loss": 4.177,
      "num_input_tokens_seen": 31762176,
      "step": 23440
    },
    {
      "epoch": 0.7504321106203188,
      "grad_norm": 30.0,
      "learning_rate": 1.8319881528322735e-05,
      "loss": 3.7358,
      "num_input_tokens_seen": 31768896,
      "step": 23445
    },
    {
      "epoch": 0.7505921515908073,
      "grad_norm": 59.0,
      "learning_rate": 1.8310421532217815e-05,
      "loss": 4.1153,
      "num_input_tokens_seen": 31775648,
      "step": 23450
    },
    {
      "epoch": 0.7507521925612957,
      "grad_norm": 25.125,
      "learning_rate": 1.8300962567729958e-05,
      "loss": 4.5617,
      "num_input_tokens_seen": 31782880,
      "step": 23455
    },
    {
      "epoch": 0.7509122335317842,
      "grad_norm": 33.75,
      "learning_rate": 1.8291504636317866e-05,
      "loss": 4.4258,
      "num_input_tokens_seen": 31789120,
      "step": 23460
    },
    {
      "epoch": 0.7510722745022725,
      "grad_norm": 25.625,
      "learning_rate": 1.8282047739440055e-05,
      "loss": 4.0633,
      "num_input_tokens_seen": 31796192,
      "step": 23465
    },
    {
      "epoch": 0.751232315472761,
      "grad_norm": 45.0,
      "learning_rate": 1.8272591878554903e-05,
      "loss": 3.0538,
      "num_input_tokens_seen": 31803184,
      "step": 23470
    },
    {
      "epoch": 0.7513923564432494,
      "grad_norm": 38.0,
      "learning_rate": 1.8263137055120638e-05,
      "loss": 4.3138,
      "num_input_tokens_seen": 31810176,
      "step": 23475
    },
    {
      "epoch": 0.7515523974137379,
      "grad_norm": 34.5,
      "learning_rate": 1.8253683270595295e-05,
      "loss": 3.8936,
      "num_input_tokens_seen": 31817088,
      "step": 23480
    },
    {
      "epoch": 0.7517124383842264,
      "grad_norm": 40.5,
      "learning_rate": 1.824423052643677e-05,
      "loss": 4.9222,
      "num_input_tokens_seen": 31823728,
      "step": 23485
    },
    {
      "epoch": 0.7518724793547148,
      "grad_norm": 26.75,
      "learning_rate": 1.82347788241028e-05,
      "loss": 4.5623,
      "num_input_tokens_seen": 31830704,
      "step": 23490
    },
    {
      "epoch": 0.7520325203252033,
      "grad_norm": 20.25,
      "learning_rate": 1.8225328165050942e-05,
      "loss": 3.8782,
      "num_input_tokens_seen": 31837936,
      "step": 23495
    },
    {
      "epoch": 0.7521925612956917,
      "grad_norm": 21.5,
      "learning_rate": 1.821587855073863e-05,
      "loss": 3.4768,
      "num_input_tokens_seen": 31845488,
      "step": 23500
    },
    {
      "epoch": 0.7523526022661802,
      "grad_norm": 23.875,
      "learning_rate": 1.8206429982623086e-05,
      "loss": 3.9333,
      "num_input_tokens_seen": 31852128,
      "step": 23505
    },
    {
      "epoch": 0.7525126432366686,
      "grad_norm": 32.5,
      "learning_rate": 1.8196982462161416e-05,
      "loss": 5.3423,
      "num_input_tokens_seen": 31858624,
      "step": 23510
    },
    {
      "epoch": 0.752672684207157,
      "grad_norm": 30.25,
      "learning_rate": 1.818753599081055e-05,
      "loss": 5.2463,
      "num_input_tokens_seen": 31865360,
      "step": 23515
    },
    {
      "epoch": 0.7528327251776454,
      "grad_norm": 29.375,
      "learning_rate": 1.817809057002724e-05,
      "loss": 4.1818,
      "num_input_tokens_seen": 31871712,
      "step": 23520
    },
    {
      "epoch": 0.7529927661481339,
      "grad_norm": 31.625,
      "learning_rate": 1.8168646201268096e-05,
      "loss": 4.1005,
      "num_input_tokens_seen": 31878688,
      "step": 23525
    },
    {
      "epoch": 0.7531528071186223,
      "grad_norm": 48.5,
      "learning_rate": 1.8159202885989557e-05,
      "loss": 3.663,
      "num_input_tokens_seen": 31885536,
      "step": 23530
    },
    {
      "epoch": 0.7533128480891108,
      "grad_norm": 38.5,
      "learning_rate": 1.814976062564789e-05,
      "loss": 4.8665,
      "num_input_tokens_seen": 31892912,
      "step": 23535
    },
    {
      "epoch": 0.7534728890595993,
      "grad_norm": 49.25,
      "learning_rate": 1.8140319421699234e-05,
      "loss": 4.0965,
      "num_input_tokens_seen": 31899984,
      "step": 23540
    },
    {
      "epoch": 0.7536329300300877,
      "grad_norm": 25.25,
      "learning_rate": 1.8130879275599515e-05,
      "loss": 4.4098,
      "num_input_tokens_seen": 31906784,
      "step": 23545
    },
    {
      "epoch": 0.7537929710005762,
      "grad_norm": 23.5,
      "learning_rate": 1.8121440188804544e-05,
      "loss": 4.6794,
      "num_input_tokens_seen": 31913488,
      "step": 23550
    },
    {
      "epoch": 0.7539530119710646,
      "grad_norm": 24.75,
      "learning_rate": 1.811200216276993e-05,
      "loss": 4.5128,
      "num_input_tokens_seen": 31920672,
      "step": 23555
    },
    {
      "epoch": 0.7541130529415531,
      "grad_norm": 35.75,
      "learning_rate": 1.810256519895115e-05,
      "loss": 4.1667,
      "num_input_tokens_seen": 31927424,
      "step": 23560
    },
    {
      "epoch": 0.7542730939120414,
      "grad_norm": 37.0,
      "learning_rate": 1.8093129298803494e-05,
      "loss": 4.9029,
      "num_input_tokens_seen": 31934496,
      "step": 23565
    },
    {
      "epoch": 0.7544331348825299,
      "grad_norm": 28.0,
      "learning_rate": 1.808369446378209e-05,
      "loss": 4.1518,
      "num_input_tokens_seen": 31941392,
      "step": 23570
    },
    {
      "epoch": 0.7545931758530183,
      "grad_norm": 32.75,
      "learning_rate": 1.8074260695341914e-05,
      "loss": 4.4956,
      "num_input_tokens_seen": 31948144,
      "step": 23575
    },
    {
      "epoch": 0.7547532168235068,
      "grad_norm": 41.25,
      "learning_rate": 1.8064827994937782e-05,
      "loss": 5.4114,
      "num_input_tokens_seen": 31954976,
      "step": 23580
    },
    {
      "epoch": 0.7549132577939953,
      "grad_norm": 19.25,
      "learning_rate": 1.8055396364024317e-05,
      "loss": 4.1798,
      "num_input_tokens_seen": 31961632,
      "step": 23585
    },
    {
      "epoch": 0.7550732987644837,
      "grad_norm": 21.125,
      "learning_rate": 1.804596580405601e-05,
      "loss": 4.3663,
      "num_input_tokens_seen": 31968656,
      "step": 23590
    },
    {
      "epoch": 0.7552333397349722,
      "grad_norm": 34.5,
      "learning_rate": 1.8036536316487174e-05,
      "loss": 4.2796,
      "num_input_tokens_seen": 31975184,
      "step": 23595
    },
    {
      "epoch": 0.7553933807054606,
      "grad_norm": 26.5,
      "learning_rate": 1.802710790277193e-05,
      "loss": 4.255,
      "num_input_tokens_seen": 31982128,
      "step": 23600
    },
    {
      "epoch": 0.7553933807054606,
      "eval_loss": 4.468667507171631,
      "eval_runtime": 490.4769,
      "eval_samples_per_second": 28.311,
      "eval_steps_per_second": 14.156,
      "num_input_tokens_seen": 31982128,
      "step": 23600
    },
    {
      "epoch": 0.7555534216759491,
      "grad_norm": 38.5,
      "learning_rate": 1.801768056436429e-05,
      "loss": 4.5063,
      "num_input_tokens_seen": 31988560,
      "step": 23605
    },
    {
      "epoch": 0.7557134626464375,
      "grad_norm": 20.875,
      "learning_rate": 1.8008254302718035e-05,
      "loss": 4.8233,
      "num_input_tokens_seen": 31995552,
      "step": 23610
    },
    {
      "epoch": 0.755873503616926,
      "grad_norm": 29.5,
      "learning_rate": 1.7998829119286837e-05,
      "loss": 4.2147,
      "num_input_tokens_seen": 32002496,
      "step": 23615
    },
    {
      "epoch": 0.7560335445874143,
      "grad_norm": 34.5,
      "learning_rate": 1.798940501552418e-05,
      "loss": 3.8849,
      "num_input_tokens_seen": 32009216,
      "step": 23620
    },
    {
      "epoch": 0.7561935855579028,
      "grad_norm": 19.625,
      "learning_rate": 1.797998199288336e-05,
      "loss": 3.7632,
      "num_input_tokens_seen": 32015456,
      "step": 23625
    },
    {
      "epoch": 0.7563536265283912,
      "grad_norm": 34.75,
      "learning_rate": 1.7970560052817543e-05,
      "loss": 4.4551,
      "num_input_tokens_seen": 32022832,
      "step": 23630
    },
    {
      "epoch": 0.7565136674988797,
      "grad_norm": 11.8125,
      "learning_rate": 1.7961139196779702e-05,
      "loss": 3.008,
      "num_input_tokens_seen": 32029328,
      "step": 23635
    },
    {
      "epoch": 0.7566737084693682,
      "grad_norm": 49.0,
      "learning_rate": 1.7951719426222647e-05,
      "loss": 3.2548,
      "num_input_tokens_seen": 32036016,
      "step": 23640
    },
    {
      "epoch": 0.7568337494398566,
      "grad_norm": 34.75,
      "learning_rate": 1.794230074259904e-05,
      "loss": 4.4658,
      "num_input_tokens_seen": 32042608,
      "step": 23645
    },
    {
      "epoch": 0.7569937904103451,
      "grad_norm": 22.5,
      "learning_rate": 1.7932883147361336e-05,
      "loss": 4.3477,
      "num_input_tokens_seen": 32049392,
      "step": 23650
    },
    {
      "epoch": 0.7571538313808335,
      "grad_norm": 39.0,
      "learning_rate": 1.7923466641961865e-05,
      "loss": 4.2534,
      "num_input_tokens_seen": 32056176,
      "step": 23655
    },
    {
      "epoch": 0.757313872351322,
      "grad_norm": 85.0,
      "learning_rate": 1.791405122785278e-05,
      "loss": 4.0922,
      "num_input_tokens_seen": 32062992,
      "step": 23660
    },
    {
      "epoch": 0.7574739133218104,
      "grad_norm": 33.75,
      "learning_rate": 1.7904636906486037e-05,
      "loss": 5.3236,
      "num_input_tokens_seen": 32069552,
      "step": 23665
    },
    {
      "epoch": 0.7576339542922989,
      "grad_norm": 32.25,
      "learning_rate": 1.7895223679313448e-05,
      "loss": 3.9588,
      "num_input_tokens_seen": 32076304,
      "step": 23670
    },
    {
      "epoch": 0.7577939952627872,
      "grad_norm": 27.125,
      "learning_rate": 1.7885811547786653e-05,
      "loss": 4.425,
      "num_input_tokens_seen": 32082928,
      "step": 23675
    },
    {
      "epoch": 0.7579540362332757,
      "grad_norm": 18.875,
      "learning_rate": 1.7876400513357115e-05,
      "loss": 3.6045,
      "num_input_tokens_seen": 32089888,
      "step": 23680
    },
    {
      "epoch": 0.7581140772037641,
      "grad_norm": 19.75,
      "learning_rate": 1.7866990577476146e-05,
      "loss": 3.7091,
      "num_input_tokens_seen": 32096432,
      "step": 23685
    },
    {
      "epoch": 0.7582741181742526,
      "grad_norm": 42.0,
      "learning_rate": 1.7857581741594863e-05,
      "loss": 4.8085,
      "num_input_tokens_seen": 32103088,
      "step": 23690
    },
    {
      "epoch": 0.7584341591447411,
      "grad_norm": 22.75,
      "learning_rate": 1.7848174007164237e-05,
      "loss": 4.5367,
      "num_input_tokens_seen": 32109744,
      "step": 23695
    },
    {
      "epoch": 0.7585942001152295,
      "grad_norm": 24.125,
      "learning_rate": 1.7838767375635052e-05,
      "loss": 4.0806,
      "num_input_tokens_seen": 32116512,
      "step": 23700
    },
    {
      "epoch": 0.758754241085718,
      "grad_norm": 22.75,
      "learning_rate": 1.782936184845793e-05,
      "loss": 3.95,
      "num_input_tokens_seen": 32123120,
      "step": 23705
    },
    {
      "epoch": 0.7589142820562064,
      "grad_norm": 34.5,
      "learning_rate": 1.7819957427083334e-05,
      "loss": 4.4859,
      "num_input_tokens_seen": 32129840,
      "step": 23710
    },
    {
      "epoch": 0.7590743230266949,
      "grad_norm": 34.5,
      "learning_rate": 1.7810554112961516e-05,
      "loss": 4.5104,
      "num_input_tokens_seen": 32136544,
      "step": 23715
    },
    {
      "epoch": 0.7592343639971832,
      "grad_norm": 18.875,
      "learning_rate": 1.7801151907542607e-05,
      "loss": 4.1474,
      "num_input_tokens_seen": 32143072,
      "step": 23720
    },
    {
      "epoch": 0.7593944049676717,
      "grad_norm": 35.75,
      "learning_rate": 1.7791750812276547e-05,
      "loss": 5.7989,
      "num_input_tokens_seen": 32149888,
      "step": 23725
    },
    {
      "epoch": 0.7595544459381601,
      "grad_norm": 22.125,
      "learning_rate": 1.778235082861309e-05,
      "loss": 3.5256,
      "num_input_tokens_seen": 32157088,
      "step": 23730
    },
    {
      "epoch": 0.7597144869086486,
      "grad_norm": 21.5,
      "learning_rate": 1.777295195800184e-05,
      "loss": 4.3745,
      "num_input_tokens_seen": 32163424,
      "step": 23735
    },
    {
      "epoch": 0.759874527879137,
      "grad_norm": 36.75,
      "learning_rate": 1.7763554201892215e-05,
      "loss": 4.3203,
      "num_input_tokens_seen": 32170352,
      "step": 23740
    },
    {
      "epoch": 0.7600345688496255,
      "grad_norm": 29.125,
      "learning_rate": 1.7754157561733476e-05,
      "loss": 3.5602,
      "num_input_tokens_seen": 32177312,
      "step": 23745
    },
    {
      "epoch": 0.760194609820114,
      "grad_norm": 33.75,
      "learning_rate": 1.7744762038974702e-05,
      "loss": 3.4377,
      "num_input_tokens_seen": 32184336,
      "step": 23750
    },
    {
      "epoch": 0.7603546507906024,
      "grad_norm": 20.0,
      "learning_rate": 1.7735367635064788e-05,
      "loss": 3.8074,
      "num_input_tokens_seen": 32191200,
      "step": 23755
    },
    {
      "epoch": 0.7605146917610909,
      "grad_norm": 36.0,
      "learning_rate": 1.7725974351452474e-05,
      "loss": 4.48,
      "num_input_tokens_seen": 32197904,
      "step": 23760
    },
    {
      "epoch": 0.7606747327315793,
      "grad_norm": 33.0,
      "learning_rate": 1.771658218958634e-05,
      "loss": 5.4793,
      "num_input_tokens_seen": 32204768,
      "step": 23765
    },
    {
      "epoch": 0.7608347737020678,
      "grad_norm": 18.75,
      "learning_rate": 1.770719115091475e-05,
      "loss": 4.5996,
      "num_input_tokens_seen": 32211728,
      "step": 23770
    },
    {
      "epoch": 0.7609948146725561,
      "grad_norm": 16.25,
      "learning_rate": 1.7697801236885935e-05,
      "loss": 5.1383,
      "num_input_tokens_seen": 32218784,
      "step": 23775
    },
    {
      "epoch": 0.7611548556430446,
      "grad_norm": 24.625,
      "learning_rate": 1.7688412448947944e-05,
      "loss": 4.0443,
      "num_input_tokens_seen": 32225488,
      "step": 23780
    },
    {
      "epoch": 0.761314896613533,
      "grad_norm": 42.0,
      "learning_rate": 1.767902478854862e-05,
      "loss": 4.0843,
      "num_input_tokens_seen": 32232944,
      "step": 23785
    },
    {
      "epoch": 0.7614749375840215,
      "grad_norm": 36.25,
      "learning_rate": 1.766963825713569e-05,
      "loss": 3.6852,
      "num_input_tokens_seen": 32239504,
      "step": 23790
    },
    {
      "epoch": 0.7616349785545099,
      "grad_norm": 21.25,
      "learning_rate": 1.766025285615665e-05,
      "loss": 4.1686,
      "num_input_tokens_seen": 32246304,
      "step": 23795
    },
    {
      "epoch": 0.7617950195249984,
      "grad_norm": 26.5,
      "learning_rate": 1.7650868587058854e-05,
      "loss": 4.7721,
      "num_input_tokens_seen": 32253040,
      "step": 23800
    },
    {
      "epoch": 0.7617950195249984,
      "eval_loss": 4.484765529632568,
      "eval_runtime": 490.4348,
      "eval_samples_per_second": 28.314,
      "eval_steps_per_second": 14.157,
      "num_input_tokens_seen": 32253040,
      "step": 23800
    },
    {
      "epoch": 0.7619550604954869,
      "grad_norm": 26.0,
      "learning_rate": 1.7641485451289484e-05,
      "loss": 4.1868,
      "num_input_tokens_seen": 32259824,
      "step": 23805
    },
    {
      "epoch": 0.7621151014659753,
      "grad_norm": 33.0,
      "learning_rate": 1.7632103450295534e-05,
      "loss": 3.9867,
      "num_input_tokens_seen": 32266896,
      "step": 23810
    },
    {
      "epoch": 0.7622751424364638,
      "grad_norm": 20.375,
      "learning_rate": 1.762272258552381e-05,
      "loss": 3.8677,
      "num_input_tokens_seen": 32273920,
      "step": 23815
    },
    {
      "epoch": 0.7624351834069522,
      "grad_norm": 19.0,
      "learning_rate": 1.7613342858420988e-05,
      "loss": 3.7831,
      "num_input_tokens_seen": 32280560,
      "step": 23820
    },
    {
      "epoch": 0.7625952243774407,
      "grad_norm": 27.375,
      "learning_rate": 1.760396427043351e-05,
      "loss": 4.3204,
      "num_input_tokens_seen": 32287248,
      "step": 23825
    },
    {
      "epoch": 0.762755265347929,
      "grad_norm": 28.75,
      "learning_rate": 1.7594586823007696e-05,
      "loss": 4.3727,
      "num_input_tokens_seen": 32293616,
      "step": 23830
    },
    {
      "epoch": 0.7629153063184175,
      "grad_norm": 26.75,
      "learning_rate": 1.7585210517589646e-05,
      "loss": 4.006,
      "num_input_tokens_seen": 32300704,
      "step": 23835
    },
    {
      "epoch": 0.7630753472889059,
      "grad_norm": 17.0,
      "learning_rate": 1.7575835355625314e-05,
      "loss": 4.2761,
      "num_input_tokens_seen": 32307840,
      "step": 23840
    },
    {
      "epoch": 0.7632353882593944,
      "grad_norm": 18.5,
      "learning_rate": 1.756646133856048e-05,
      "loss": 4.0515,
      "num_input_tokens_seen": 32314800,
      "step": 23845
    },
    {
      "epoch": 0.7633954292298829,
      "grad_norm": 28.5,
      "learning_rate": 1.7557088467840714e-05,
      "loss": 4.3088,
      "num_input_tokens_seen": 32321952,
      "step": 23850
    },
    {
      "epoch": 0.7635554702003713,
      "grad_norm": 33.25,
      "learning_rate": 1.7547716744911438e-05,
      "loss": 5.1488,
      "num_input_tokens_seen": 32328704,
      "step": 23855
    },
    {
      "epoch": 0.7637155111708598,
      "grad_norm": 22.875,
      "learning_rate": 1.7538346171217902e-05,
      "loss": 3.742,
      "num_input_tokens_seen": 32335536,
      "step": 23860
    },
    {
      "epoch": 0.7638755521413482,
      "grad_norm": 40.0,
      "learning_rate": 1.7528976748205146e-05,
      "loss": 4.4952,
      "num_input_tokens_seen": 32342432,
      "step": 23865
    },
    {
      "epoch": 0.7640355931118367,
      "grad_norm": 31.0,
      "learning_rate": 1.751960847731807e-05,
      "loss": 4.4899,
      "num_input_tokens_seen": 32349024,
      "step": 23870
    },
    {
      "epoch": 0.764195634082325,
      "grad_norm": 44.5,
      "learning_rate": 1.7510241360001362e-05,
      "loss": 4.7459,
      "num_input_tokens_seen": 32355520,
      "step": 23875
    },
    {
      "epoch": 0.7643556750528135,
      "grad_norm": 21.0,
      "learning_rate": 1.7500875397699562e-05,
      "loss": 3.8407,
      "num_input_tokens_seen": 32362352,
      "step": 23880
    },
    {
      "epoch": 0.7645157160233019,
      "grad_norm": 33.0,
      "learning_rate": 1.7491510591857015e-05,
      "loss": 4.4884,
      "num_input_tokens_seen": 32369392,
      "step": 23885
    },
    {
      "epoch": 0.7646757569937904,
      "grad_norm": 25.625,
      "learning_rate": 1.7482146943917896e-05,
      "loss": 4.5406,
      "num_input_tokens_seen": 32376672,
      "step": 23890
    },
    {
      "epoch": 0.7648357979642788,
      "grad_norm": 37.0,
      "learning_rate": 1.7472784455326185e-05,
      "loss": 5.088,
      "num_input_tokens_seen": 32383312,
      "step": 23895
    },
    {
      "epoch": 0.7649958389347673,
      "grad_norm": 18.875,
      "learning_rate": 1.746342312752572e-05,
      "loss": 4.0672,
      "num_input_tokens_seen": 32390208,
      "step": 23900
    },
    {
      "epoch": 0.7651558799052558,
      "grad_norm": 28.25,
      "learning_rate": 1.74540629619601e-05,
      "loss": 4.4805,
      "num_input_tokens_seen": 32397040,
      "step": 23905
    },
    {
      "epoch": 0.7653159208757442,
      "grad_norm": 23.875,
      "learning_rate": 1.7444703960072815e-05,
      "loss": 5.2587,
      "num_input_tokens_seen": 32403584,
      "step": 23910
    },
    {
      "epoch": 0.7654759618462327,
      "grad_norm": 31.0,
      "learning_rate": 1.7435346123307118e-05,
      "loss": 5.5275,
      "num_input_tokens_seen": 32410320,
      "step": 23915
    },
    {
      "epoch": 0.7656360028167211,
      "grad_norm": 37.5,
      "learning_rate": 1.742598945310611e-05,
      "loss": 4.1573,
      "num_input_tokens_seen": 32417168,
      "step": 23920
    },
    {
      "epoch": 0.7657960437872096,
      "grad_norm": 21.875,
      "learning_rate": 1.741663395091272e-05,
      "loss": 4.4448,
      "num_input_tokens_seen": 32423536,
      "step": 23925
    },
    {
      "epoch": 0.7659560847576979,
      "grad_norm": 45.75,
      "learning_rate": 1.7407279618169657e-05,
      "loss": 5.1914,
      "num_input_tokens_seen": 32430080,
      "step": 23930
    },
    {
      "epoch": 0.7661161257281864,
      "grad_norm": 41.0,
      "learning_rate": 1.73979264563195e-05,
      "loss": 4.3739,
      "num_input_tokens_seen": 32436608,
      "step": 23935
    },
    {
      "epoch": 0.7662761666986748,
      "grad_norm": 28.625,
      "learning_rate": 1.7388574466804625e-05,
      "loss": 4.7276,
      "num_input_tokens_seen": 32443424,
      "step": 23940
    },
    {
      "epoch": 0.7664362076691633,
      "grad_norm": 28.125,
      "learning_rate": 1.7379223651067207e-05,
      "loss": 3.6476,
      "num_input_tokens_seen": 32450096,
      "step": 23945
    },
    {
      "epoch": 0.7665962486396517,
      "grad_norm": 31.875,
      "learning_rate": 1.736987401054928e-05,
      "loss": 4.3716,
      "num_input_tokens_seen": 32456960,
      "step": 23950
    },
    {
      "epoch": 0.7667562896101402,
      "grad_norm": 24.125,
      "learning_rate": 1.736052554669266e-05,
      "loss": 4.0941,
      "num_input_tokens_seen": 32463552,
      "step": 23955
    },
    {
      "epoch": 0.7669163305806287,
      "grad_norm": 24.0,
      "learning_rate": 1.7351178260939007e-05,
      "loss": 5.0591,
      "num_input_tokens_seen": 32470288,
      "step": 23960
    },
    {
      "epoch": 0.7670763715511171,
      "grad_norm": 14.3125,
      "learning_rate": 1.7341832154729794e-05,
      "loss": 4.1446,
      "num_input_tokens_seen": 32477088,
      "step": 23965
    },
    {
      "epoch": 0.7672364125216056,
      "grad_norm": 29.75,
      "learning_rate": 1.7332487229506286e-05,
      "loss": 3.6924,
      "num_input_tokens_seen": 32484240,
      "step": 23970
    },
    {
      "epoch": 0.767396453492094,
      "grad_norm": 27.75,
      "learning_rate": 1.732314348670961e-05,
      "loss": 5.0777,
      "num_input_tokens_seen": 32491152,
      "step": 23975
    },
    {
      "epoch": 0.7675564944625825,
      "grad_norm": 25.0,
      "learning_rate": 1.7313800927780686e-05,
      "loss": 4.5904,
      "num_input_tokens_seen": 32497680,
      "step": 23980
    },
    {
      "epoch": 0.7677165354330708,
      "grad_norm": 47.75,
      "learning_rate": 1.7304459554160245e-05,
      "loss": 5.1902,
      "num_input_tokens_seen": 32504576,
      "step": 23985
    },
    {
      "epoch": 0.7678765764035593,
      "grad_norm": 29.0,
      "learning_rate": 1.7295119367288853e-05,
      "loss": 3.9352,
      "num_input_tokens_seen": 32511136,
      "step": 23990
    },
    {
      "epoch": 0.7680366173740477,
      "grad_norm": 24.625,
      "learning_rate": 1.728578036860688e-05,
      "loss": 4.1208,
      "num_input_tokens_seen": 32517792,
      "step": 23995
    },
    {
      "epoch": 0.7681966583445362,
      "grad_norm": 20.25,
      "learning_rate": 1.7276442559554513e-05,
      "loss": 3.9138,
      "num_input_tokens_seen": 32524464,
      "step": 24000
    },
    {
      "epoch": 0.7681966583445362,
      "eval_loss": 4.508407115936279,
      "eval_runtime": 489.9416,
      "eval_samples_per_second": 28.342,
      "eval_steps_per_second": 14.171,
      "num_input_tokens_seen": 32524464,
      "step": 24000
    },
    {
      "epoch": 0.7683566993150246,
      "grad_norm": 43.5,
      "learning_rate": 1.726710594157177e-05,
      "loss": 4.531,
      "num_input_tokens_seen": 32531040,
      "step": 24005
    },
    {
      "epoch": 0.7685167402855131,
      "grad_norm": 32.5,
      "learning_rate": 1.725777051609846e-05,
      "loss": 4.1588,
      "num_input_tokens_seen": 32537664,
      "step": 24010
    },
    {
      "epoch": 0.7686767812560016,
      "grad_norm": 36.25,
      "learning_rate": 1.7248436284574228e-05,
      "loss": 4.5029,
      "num_input_tokens_seen": 32544336,
      "step": 24015
    },
    {
      "epoch": 0.76883682222649,
      "grad_norm": 23.875,
      "learning_rate": 1.723910324843855e-05,
      "loss": 3.8376,
      "num_input_tokens_seen": 32551280,
      "step": 24020
    },
    {
      "epoch": 0.7689968631969785,
      "grad_norm": 32.25,
      "learning_rate": 1.722977140913067e-05,
      "loss": 4.4938,
      "num_input_tokens_seen": 32558304,
      "step": 24025
    },
    {
      "epoch": 0.7691569041674668,
      "grad_norm": 35.0,
      "learning_rate": 1.7220440768089688e-05,
      "loss": 4.7495,
      "num_input_tokens_seen": 32564720,
      "step": 24030
    },
    {
      "epoch": 0.7693169451379553,
      "grad_norm": 44.5,
      "learning_rate": 1.7211111326754505e-05,
      "loss": 4.3367,
      "num_input_tokens_seen": 32571760,
      "step": 24035
    },
    {
      "epoch": 0.7694769861084437,
      "grad_norm": 20.0,
      "learning_rate": 1.720178308656383e-05,
      "loss": 4.8222,
      "num_input_tokens_seen": 32578160,
      "step": 24040
    },
    {
      "epoch": 0.7696370270789322,
      "grad_norm": 23.375,
      "learning_rate": 1.719245604895621e-05,
      "loss": 4.3458,
      "num_input_tokens_seen": 32584832,
      "step": 24045
    },
    {
      "epoch": 0.7697970680494206,
      "grad_norm": 43.5,
      "learning_rate": 1.7183130215369972e-05,
      "loss": 4.4693,
      "num_input_tokens_seen": 32591296,
      "step": 24050
    },
    {
      "epoch": 0.7699571090199091,
      "grad_norm": 19.125,
      "learning_rate": 1.7173805587243292e-05,
      "loss": 4.1551,
      "num_input_tokens_seen": 32597920,
      "step": 24055
    },
    {
      "epoch": 0.7701171499903975,
      "grad_norm": 30.75,
      "learning_rate": 1.7164482166014147e-05,
      "loss": 4.245,
      "num_input_tokens_seen": 32604704,
      "step": 24060
    },
    {
      "epoch": 0.770277190960886,
      "grad_norm": 27.125,
      "learning_rate": 1.7155159953120313e-05,
      "loss": 4.9423,
      "num_input_tokens_seen": 32611456,
      "step": 24065
    },
    {
      "epoch": 0.7704372319313745,
      "grad_norm": 58.25,
      "learning_rate": 1.714583894999941e-05,
      "loss": 3.739,
      "num_input_tokens_seen": 32618064,
      "step": 24070
    },
    {
      "epoch": 0.7705972729018629,
      "grad_norm": 44.75,
      "learning_rate": 1.7136519158088826e-05,
      "loss": 4.915,
      "num_input_tokens_seen": 32624672,
      "step": 24075
    },
    {
      "epoch": 0.7707573138723514,
      "grad_norm": 26.0,
      "learning_rate": 1.712720057882581e-05,
      "loss": 4.341,
      "num_input_tokens_seen": 32631248,
      "step": 24080
    },
    {
      "epoch": 0.7709173548428397,
      "grad_norm": 13.875,
      "learning_rate": 1.7117883213647413e-05,
      "loss": 4.4295,
      "num_input_tokens_seen": 32637856,
      "step": 24085
    },
    {
      "epoch": 0.7710773958133282,
      "grad_norm": 32.75,
      "learning_rate": 1.710856706399046e-05,
      "loss": 4.121,
      "num_input_tokens_seen": 32644768,
      "step": 24090
    },
    {
      "epoch": 0.7712374367838166,
      "grad_norm": 42.0,
      "learning_rate": 1.7099252131291648e-05,
      "loss": 5.1849,
      "num_input_tokens_seen": 32651744,
      "step": 24095
    },
    {
      "epoch": 0.7713974777543051,
      "grad_norm": 33.25,
      "learning_rate": 1.708993841698744e-05,
      "loss": 4.1038,
      "num_input_tokens_seen": 32658176,
      "step": 24100
    },
    {
      "epoch": 0.7715575187247935,
      "grad_norm": 50.0,
      "learning_rate": 1.7080625922514132e-05,
      "loss": 4.8901,
      "num_input_tokens_seen": 32664944,
      "step": 24105
    },
    {
      "epoch": 0.771717559695282,
      "grad_norm": 21.875,
      "learning_rate": 1.7071314649307836e-05,
      "loss": 4.4256,
      "num_input_tokens_seen": 32671456,
      "step": 24110
    },
    {
      "epoch": 0.7718776006657705,
      "grad_norm": 29.0,
      "learning_rate": 1.7062004598804448e-05,
      "loss": 4.0772,
      "num_input_tokens_seen": 32678368,
      "step": 24115
    },
    {
      "epoch": 0.7720376416362589,
      "grad_norm": 20.0,
      "learning_rate": 1.7052695772439702e-05,
      "loss": 4.0481,
      "num_input_tokens_seen": 32685248,
      "step": 24120
    },
    {
      "epoch": 0.7721976826067474,
      "grad_norm": 33.5,
      "learning_rate": 1.7043388171649154e-05,
      "loss": 4.4583,
      "num_input_tokens_seen": 32691952,
      "step": 24125
    },
    {
      "epoch": 0.7723577235772358,
      "grad_norm": 37.25,
      "learning_rate": 1.7034081797868127e-05,
      "loss": 3.9139,
      "num_input_tokens_seen": 32699600,
      "step": 24130
    },
    {
      "epoch": 0.7725177645477243,
      "grad_norm": 28.25,
      "learning_rate": 1.70247766525318e-05,
      "loss": 4.3339,
      "num_input_tokens_seen": 32706576,
      "step": 24135
    },
    {
      "epoch": 0.7726778055182126,
      "grad_norm": 21.375,
      "learning_rate": 1.701547273707514e-05,
      "loss": 5.0842,
      "num_input_tokens_seen": 32713488,
      "step": 24140
    },
    {
      "epoch": 0.7728378464887011,
      "grad_norm": 23.25,
      "learning_rate": 1.7006170052932916e-05,
      "loss": 5.1878,
      "num_input_tokens_seen": 32720496,
      "step": 24145
    },
    {
      "epoch": 0.7729978874591895,
      "grad_norm": 38.0,
      "learning_rate": 1.6996868601539735e-05,
      "loss": 5.1716,
      "num_input_tokens_seen": 32726896,
      "step": 24150
    },
    {
      "epoch": 0.773157928429678,
      "grad_norm": 42.0,
      "learning_rate": 1.6987568384329977e-05,
      "loss": 4.3372,
      "num_input_tokens_seen": 32733840,
      "step": 24155
    },
    {
      "epoch": 0.7733179694001664,
      "grad_norm": 29.0,
      "learning_rate": 1.6978269402737866e-05,
      "loss": 4.3507,
      "num_input_tokens_seen": 32740336,
      "step": 24160
    },
    {
      "epoch": 0.7734780103706549,
      "grad_norm": 50.5,
      "learning_rate": 1.696897165819743e-05,
      "loss": 5.4219,
      "num_input_tokens_seen": 32747232,
      "step": 24165
    },
    {
      "epoch": 0.7736380513411434,
      "grad_norm": 27.875,
      "learning_rate": 1.6959675152142487e-05,
      "loss": 4.629,
      "num_input_tokens_seen": 32754000,
      "step": 24170
    },
    {
      "epoch": 0.7737980923116318,
      "grad_norm": 31.625,
      "learning_rate": 1.6950379886006667e-05,
      "loss": 3.9584,
      "num_input_tokens_seen": 32760448,
      "step": 24175
    },
    {
      "epoch": 0.7739581332821203,
      "grad_norm": 24.5,
      "learning_rate": 1.6941085861223438e-05,
      "loss": 5.1275,
      "num_input_tokens_seen": 32767296,
      "step": 24180
    },
    {
      "epoch": 0.7741181742526086,
      "grad_norm": 24.0,
      "learning_rate": 1.6931793079226034e-05,
      "loss": 3.444,
      "num_input_tokens_seen": 32774672,
      "step": 24185
    },
    {
      "epoch": 0.7742782152230971,
      "grad_norm": 24.5,
      "learning_rate": 1.692250154144754e-05,
      "loss": 4.0935,
      "num_input_tokens_seen": 32781072,
      "step": 24190
    },
    {
      "epoch": 0.7744382561935855,
      "grad_norm": 14.125,
      "learning_rate": 1.6913211249320807e-05,
      "loss": 4.1219,
      "num_input_tokens_seen": 32787936,
      "step": 24195
    },
    {
      "epoch": 0.774598297164074,
      "grad_norm": 19.75,
      "learning_rate": 1.6903922204278522e-05,
      "loss": 3.8332,
      "num_input_tokens_seen": 32794928,
      "step": 24200
    },
    {
      "epoch": 0.774598297164074,
      "eval_loss": 4.489377021789551,
      "eval_runtime": 490.5299,
      "eval_samples_per_second": 28.308,
      "eval_steps_per_second": 14.154,
      "num_input_tokens_seen": 32794928,
      "step": 24200
    },
    {
      "epoch": 0.7747583381345624,
      "grad_norm": 43.5,
      "learning_rate": 1.6894634407753186e-05,
      "loss": 5.3154,
      "num_input_tokens_seen": 32801632,
      "step": 24205
    },
    {
      "epoch": 0.7749183791050509,
      "grad_norm": 19.125,
      "learning_rate": 1.6885347861177077e-05,
      "loss": 4.4735,
      "num_input_tokens_seen": 32808704,
      "step": 24210
    },
    {
      "epoch": 0.7750784200755393,
      "grad_norm": 34.25,
      "learning_rate": 1.6876062565982298e-05,
      "loss": 3.4194,
      "num_input_tokens_seen": 32815184,
      "step": 24215
    },
    {
      "epoch": 0.7752384610460278,
      "grad_norm": 39.5,
      "learning_rate": 1.6866778523600774e-05,
      "loss": 4.5794,
      "num_input_tokens_seen": 32821888,
      "step": 24220
    },
    {
      "epoch": 0.7753985020165163,
      "grad_norm": 14.0,
      "learning_rate": 1.6857495735464195e-05,
      "loss": 3.8671,
      "num_input_tokens_seen": 32828480,
      "step": 24225
    },
    {
      "epoch": 0.7755585429870047,
      "grad_norm": 46.75,
      "learning_rate": 1.6848214203004115e-05,
      "loss": 4.5937,
      "num_input_tokens_seen": 32835056,
      "step": 24230
    },
    {
      "epoch": 0.7757185839574932,
      "grad_norm": 24.875,
      "learning_rate": 1.6838933927651835e-05,
      "loss": 3.6823,
      "num_input_tokens_seen": 32842080,
      "step": 24235
    },
    {
      "epoch": 0.7758786249279815,
      "grad_norm": 27.125,
      "learning_rate": 1.6829654910838506e-05,
      "loss": 4.5538,
      "num_input_tokens_seen": 32848960,
      "step": 24240
    },
    {
      "epoch": 0.77603866589847,
      "grad_norm": 48.75,
      "learning_rate": 1.6820377153995065e-05,
      "loss": 4.4424,
      "num_input_tokens_seen": 32855728,
      "step": 24245
    },
    {
      "epoch": 0.7761987068689584,
      "grad_norm": 14.75,
      "learning_rate": 1.681110065855226e-05,
      "loss": 4.3714,
      "num_input_tokens_seen": 32862304,
      "step": 24250
    },
    {
      "epoch": 0.7763587478394469,
      "grad_norm": 32.5,
      "learning_rate": 1.6801825425940642e-05,
      "loss": 3.7456,
      "num_input_tokens_seen": 32868848,
      "step": 24255
    },
    {
      "epoch": 0.7765187888099353,
      "grad_norm": 30.875,
      "learning_rate": 1.679255145759056e-05,
      "loss": 3.6366,
      "num_input_tokens_seen": 32876096,
      "step": 24260
    },
    {
      "epoch": 0.7766788297804238,
      "grad_norm": 30.0,
      "learning_rate": 1.6783278754932187e-05,
      "loss": 3.7123,
      "num_input_tokens_seen": 32883072,
      "step": 24265
    },
    {
      "epoch": 0.7768388707509122,
      "grad_norm": 18.25,
      "learning_rate": 1.6774007319395496e-05,
      "loss": 4.154,
      "num_input_tokens_seen": 32890256,
      "step": 24270
    },
    {
      "epoch": 0.7769989117214007,
      "grad_norm": 42.0,
      "learning_rate": 1.6764737152410243e-05,
      "loss": 4.345,
      "num_input_tokens_seen": 32897200,
      "step": 24275
    },
    {
      "epoch": 0.7771589526918892,
      "grad_norm": 43.25,
      "learning_rate": 1.6755468255406016e-05,
      "loss": 3.7037,
      "num_input_tokens_seen": 32903792,
      "step": 24280
    },
    {
      "epoch": 0.7773189936623776,
      "grad_norm": 26.125,
      "learning_rate": 1.674620062981219e-05,
      "loss": 4.293,
      "num_input_tokens_seen": 32910592,
      "step": 24285
    },
    {
      "epoch": 0.777479034632866,
      "grad_norm": 38.25,
      "learning_rate": 1.6736934277057947e-05,
      "loss": 4.1731,
      "num_input_tokens_seen": 32917200,
      "step": 24290
    },
    {
      "epoch": 0.7776390756033544,
      "grad_norm": 28.125,
      "learning_rate": 1.6727669198572286e-05,
      "loss": 4.6072,
      "num_input_tokens_seen": 32924016,
      "step": 24295
    },
    {
      "epoch": 0.7777991165738429,
      "grad_norm": 27.75,
      "learning_rate": 1.6718405395783984e-05,
      "loss": 4.1843,
      "num_input_tokens_seen": 32930880,
      "step": 24300
    },
    {
      "epoch": 0.7779591575443313,
      "grad_norm": 17.875,
      "learning_rate": 1.6709142870121643e-05,
      "loss": 4.1987,
      "num_input_tokens_seen": 32937584,
      "step": 24305
    },
    {
      "epoch": 0.7781191985148198,
      "grad_norm": 41.0,
      "learning_rate": 1.669988162301367e-05,
      "loss": 5.4235,
      "num_input_tokens_seen": 32944384,
      "step": 24310
    },
    {
      "epoch": 0.7782792394853082,
      "grad_norm": 38.25,
      "learning_rate": 1.6690621655888243e-05,
      "loss": 4.8893,
      "num_input_tokens_seen": 32951584,
      "step": 24315
    },
    {
      "epoch": 0.7784392804557967,
      "grad_norm": 39.0,
      "learning_rate": 1.6681362970173386e-05,
      "loss": 4.9689,
      "num_input_tokens_seen": 32958320,
      "step": 24320
    },
    {
      "epoch": 0.7785993214262851,
      "grad_norm": 38.75,
      "learning_rate": 1.6672105567296904e-05,
      "loss": 4.4919,
      "num_input_tokens_seen": 32965360,
      "step": 24325
    },
    {
      "epoch": 0.7787593623967736,
      "grad_norm": 21.0,
      "learning_rate": 1.666284944868639e-05,
      "loss": 3.7867,
      "num_input_tokens_seen": 32972176,
      "step": 24330
    },
    {
      "epoch": 0.7789194033672621,
      "grad_norm": 20.875,
      "learning_rate": 1.665359461576927e-05,
      "loss": 4.1573,
      "num_input_tokens_seen": 32978672,
      "step": 24335
    },
    {
      "epoch": 0.7790794443377504,
      "grad_norm": 27.125,
      "learning_rate": 1.6644341069972736e-05,
      "loss": 4.1823,
      "num_input_tokens_seen": 32986080,
      "step": 24340
    },
    {
      "epoch": 0.7792394853082389,
      "grad_norm": 24.625,
      "learning_rate": 1.6635088812723813e-05,
      "loss": 4.6674,
      "num_input_tokens_seen": 32992816,
      "step": 24345
    },
    {
      "epoch": 0.7793995262787273,
      "grad_norm": 25.75,
      "learning_rate": 1.6625837845449328e-05,
      "loss": 5.0351,
      "num_input_tokens_seen": 32999504,
      "step": 24350
    },
    {
      "epoch": 0.7795595672492158,
      "grad_norm": 18.5,
      "learning_rate": 1.6616588169575874e-05,
      "loss": 4.0372,
      "num_input_tokens_seen": 33006192,
      "step": 24355
    },
    {
      "epoch": 0.7797196082197042,
      "grad_norm": 24.75,
      "learning_rate": 1.6607339786529878e-05,
      "loss": 4.1895,
      "num_input_tokens_seen": 33012656,
      "step": 24360
    },
    {
      "epoch": 0.7798796491901927,
      "grad_norm": 28.5,
      "learning_rate": 1.659809269773756e-05,
      "loss": 4.8806,
      "num_input_tokens_seen": 33020080,
      "step": 24365
    },
    {
      "epoch": 0.7800396901606811,
      "grad_norm": 29.25,
      "learning_rate": 1.658884690462493e-05,
      "loss": 3.705,
      "num_input_tokens_seen": 33026736,
      "step": 24370
    },
    {
      "epoch": 0.7801997311311696,
      "grad_norm": 39.25,
      "learning_rate": 1.6579602408617813e-05,
      "loss": 4.4407,
      "num_input_tokens_seen": 33033312,
      "step": 24375
    },
    {
      "epoch": 0.7803597721016581,
      "grad_norm": 33.75,
      "learning_rate": 1.657035921114181e-05,
      "loss": 4.3036,
      "num_input_tokens_seen": 33040656,
      "step": 24380
    },
    {
      "epoch": 0.7805198130721465,
      "grad_norm": 44.75,
      "learning_rate": 1.656111731362236e-05,
      "loss": 4.0028,
      "num_input_tokens_seen": 33047600,
      "step": 24385
    },
    {
      "epoch": 0.780679854042635,
      "grad_norm": 40.0,
      "learning_rate": 1.6551876717484666e-05,
      "loss": 4.4983,
      "num_input_tokens_seen": 33054064,
      "step": 24390
    },
    {
      "epoch": 0.7808398950131233,
      "grad_norm": 40.0,
      "learning_rate": 1.6542637424153752e-05,
      "loss": 4.0678,
      "num_input_tokens_seen": 33060992,
      "step": 24395
    },
    {
      "epoch": 0.7809999359836118,
      "grad_norm": 31.875,
      "learning_rate": 1.6533399435054418e-05,
      "loss": 3.9703,
      "num_input_tokens_seen": 33067904,
      "step": 24400
    },
    {
      "epoch": 0.7809999359836118,
      "eval_loss": 4.482758522033691,
      "eval_runtime": 490.5808,
      "eval_samples_per_second": 28.305,
      "eval_steps_per_second": 14.153,
      "num_input_tokens_seen": 33067904,
      "step": 24400
    },
    {
      "epoch": 0.7811599769541002,
      "grad_norm": 21.125,
      "learning_rate": 1.6524162751611304e-05,
      "loss": 5.0017,
      "num_input_tokens_seen": 33074064,
      "step": 24405
    },
    {
      "epoch": 0.7813200179245887,
      "grad_norm": 24.75,
      "learning_rate": 1.6514927375248796e-05,
      "loss": 3.5888,
      "num_input_tokens_seen": 33080640,
      "step": 24410
    },
    {
      "epoch": 0.7814800588950771,
      "grad_norm": 38.25,
      "learning_rate": 1.6505693307391127e-05,
      "loss": 4.1224,
      "num_input_tokens_seen": 33087264,
      "step": 24415
    },
    {
      "epoch": 0.7816400998655656,
      "grad_norm": 27.5,
      "learning_rate": 1.6496460549462288e-05,
      "loss": 4.5002,
      "num_input_tokens_seen": 33093872,
      "step": 24420
    },
    {
      "epoch": 0.781800140836054,
      "grad_norm": 39.0,
      "learning_rate": 1.6487229102886097e-05,
      "loss": 5.0477,
      "num_input_tokens_seen": 33100752,
      "step": 24425
    },
    {
      "epoch": 0.7819601818065425,
      "grad_norm": 27.375,
      "learning_rate": 1.6477998969086155e-05,
      "loss": 4.3953,
      "num_input_tokens_seen": 33107616,
      "step": 24430
    },
    {
      "epoch": 0.782120222777031,
      "grad_norm": 21.375,
      "learning_rate": 1.646877014948587e-05,
      "loss": 3.3113,
      "num_input_tokens_seen": 33114736,
      "step": 24435
    },
    {
      "epoch": 0.7822802637475194,
      "grad_norm": 39.0,
      "learning_rate": 1.6459542645508433e-05,
      "loss": 4.2002,
      "num_input_tokens_seen": 33121504,
      "step": 24440
    },
    {
      "epoch": 0.7824403047180079,
      "grad_norm": 82.0,
      "learning_rate": 1.6450316458576852e-05,
      "loss": 4.5107,
      "num_input_tokens_seen": 33127936,
      "step": 24445
    },
    {
      "epoch": 0.7826003456884962,
      "grad_norm": 39.25,
      "learning_rate": 1.6441091590113912e-05,
      "loss": 3.8839,
      "num_input_tokens_seen": 33134688,
      "step": 24450
    },
    {
      "epoch": 0.7827603866589847,
      "grad_norm": 27.75,
      "learning_rate": 1.6431868041542213e-05,
      "loss": 5.13,
      "num_input_tokens_seen": 33141520,
      "step": 24455
    },
    {
      "epoch": 0.7829204276294731,
      "grad_norm": 17.125,
      "learning_rate": 1.6422645814284123e-05,
      "loss": 4.3084,
      "num_input_tokens_seen": 33148512,
      "step": 24460
    },
    {
      "epoch": 0.7830804685999616,
      "grad_norm": 39.25,
      "learning_rate": 1.6413424909761846e-05,
      "loss": 4.7541,
      "num_input_tokens_seen": 33155232,
      "step": 24465
    },
    {
      "epoch": 0.78324050957045,
      "grad_norm": 34.75,
      "learning_rate": 1.640420532939736e-05,
      "loss": 3.7177,
      "num_input_tokens_seen": 33162032,
      "step": 24470
    },
    {
      "epoch": 0.7834005505409385,
      "grad_norm": 43.25,
      "learning_rate": 1.639498707461242e-05,
      "loss": 4.697,
      "num_input_tokens_seen": 33168800,
      "step": 24475
    },
    {
      "epoch": 0.7835605915114269,
      "grad_norm": 22.125,
      "learning_rate": 1.6385770146828614e-05,
      "loss": 3.406,
      "num_input_tokens_seen": 33175632,
      "step": 24480
    },
    {
      "epoch": 0.7837206324819154,
      "grad_norm": 42.0,
      "learning_rate": 1.637655454746731e-05,
      "loss": 4.5834,
      "num_input_tokens_seen": 33183024,
      "step": 24485
    },
    {
      "epoch": 0.7838806734524039,
      "grad_norm": 40.25,
      "learning_rate": 1.6367340277949658e-05,
      "loss": 3.986,
      "num_input_tokens_seen": 33190032,
      "step": 24490
    },
    {
      "epoch": 0.7840407144228922,
      "grad_norm": 35.5,
      "learning_rate": 1.635812733969663e-05,
      "loss": 4.3349,
      "num_input_tokens_seen": 33196992,
      "step": 24495
    },
    {
      "epoch": 0.7842007553933807,
      "grad_norm": 47.25,
      "learning_rate": 1.634891573412896e-05,
      "loss": 3.9426,
      "num_input_tokens_seen": 33203584,
      "step": 24500
    },
    {
      "epoch": 0.7843607963638691,
      "grad_norm": 25.25,
      "learning_rate": 1.6339705462667196e-05,
      "loss": 3.9274,
      "num_input_tokens_seen": 33210096,
      "step": 24505
    },
    {
      "epoch": 0.7845208373343576,
      "grad_norm": 92.5,
      "learning_rate": 1.633049652673169e-05,
      "loss": 4.2692,
      "num_input_tokens_seen": 33216912,
      "step": 24510
    },
    {
      "epoch": 0.784680878304846,
      "grad_norm": 30.75,
      "learning_rate": 1.632128892774256e-05,
      "loss": 3.3007,
      "num_input_tokens_seen": 33223568,
      "step": 24515
    },
    {
      "epoch": 0.7848409192753345,
      "grad_norm": 17.875,
      "learning_rate": 1.6312082667119737e-05,
      "loss": 3.7943,
      "num_input_tokens_seen": 33230272,
      "step": 24520
    },
    {
      "epoch": 0.7850009602458229,
      "grad_norm": 31.75,
      "learning_rate": 1.630287774628296e-05,
      "loss": 4.4732,
      "num_input_tokens_seen": 33236848,
      "step": 24525
    },
    {
      "epoch": 0.7851610012163114,
      "grad_norm": 21.875,
      "learning_rate": 1.6293674166651718e-05,
      "loss": 3.9203,
      "num_input_tokens_seen": 33243952,
      "step": 24530
    },
    {
      "epoch": 0.7853210421867998,
      "grad_norm": 22.625,
      "learning_rate": 1.6284471929645338e-05,
      "loss": 4.0071,
      "num_input_tokens_seen": 33250864,
      "step": 24535
    },
    {
      "epoch": 0.7854810831572883,
      "grad_norm": 44.75,
      "learning_rate": 1.627527103668291e-05,
      "loss": 5.5295,
      "num_input_tokens_seen": 33257632,
      "step": 24540
    },
    {
      "epoch": 0.7856411241277768,
      "grad_norm": 27.0,
      "learning_rate": 1.6266071489183327e-05,
      "loss": 3.9033,
      "num_input_tokens_seen": 33264000,
      "step": 24545
    },
    {
      "epoch": 0.7858011650982651,
      "grad_norm": 34.25,
      "learning_rate": 1.6256873288565283e-05,
      "loss": 4.3133,
      "num_input_tokens_seen": 33270560,
      "step": 24550
    },
    {
      "epoch": 0.7859612060687536,
      "grad_norm": 25.375,
      "learning_rate": 1.6247676436247245e-05,
      "loss": 3.8611,
      "num_input_tokens_seen": 33277104,
      "step": 24555
    },
    {
      "epoch": 0.786121247039242,
      "grad_norm": 25.875,
      "learning_rate": 1.6238480933647486e-05,
      "loss": 3.8924,
      "num_input_tokens_seen": 33283648,
      "step": 24560
    },
    {
      "epoch": 0.7862812880097305,
      "grad_norm": 33.25,
      "learning_rate": 1.6229286782184083e-05,
      "loss": 3.9951,
      "num_input_tokens_seen": 33290096,
      "step": 24565
    },
    {
      "epoch": 0.7864413289802189,
      "grad_norm": 16.5,
      "learning_rate": 1.622009398327487e-05,
      "loss": 4.0172,
      "num_input_tokens_seen": 33296528,
      "step": 24570
    },
    {
      "epoch": 0.7866013699507074,
      "grad_norm": 21.125,
      "learning_rate": 1.6210902538337502e-05,
      "loss": 3.3439,
      "num_input_tokens_seen": 33303808,
      "step": 24575
    },
    {
      "epoch": 0.7867614109211958,
      "grad_norm": 22.125,
      "learning_rate": 1.6201712448789413e-05,
      "loss": 4.3706,
      "num_input_tokens_seen": 33310224,
      "step": 24580
    },
    {
      "epoch": 0.7869214518916843,
      "grad_norm": 81.0,
      "learning_rate": 1.6192523716047827e-05,
      "loss": 4.7883,
      "num_input_tokens_seen": 33316576,
      "step": 24585
    },
    {
      "epoch": 0.7870814928621728,
      "grad_norm": 30.625,
      "learning_rate": 1.6183336341529776e-05,
      "loss": 3.8556,
      "num_input_tokens_seen": 33323152,
      "step": 24590
    },
    {
      "epoch": 0.7872415338326612,
      "grad_norm": 32.0,
      "learning_rate": 1.6174150326652047e-05,
      "loss": 4.7166,
      "num_input_tokens_seen": 33329920,
      "step": 24595
    },
    {
      "epoch": 0.7874015748031497,
      "grad_norm": 33.25,
      "learning_rate": 1.6164965672831256e-05,
      "loss": 4.0106,
      "num_input_tokens_seen": 33336480,
      "step": 24600
    },
    {
      "epoch": 0.7874015748031497,
      "eval_loss": 4.49838399887085,
      "eval_runtime": 490.4717,
      "eval_samples_per_second": 28.312,
      "eval_steps_per_second": 14.156,
      "num_input_tokens_seen": 33336480,
      "step": 24600
    },
    {
      "epoch": 0.787561615773638,
      "grad_norm": 15.8125,
      "learning_rate": 1.6155782381483784e-05,
      "loss": 3.5924,
      "num_input_tokens_seen": 33343088,
      "step": 24605
    },
    {
      "epoch": 0.7877216567441265,
      "grad_norm": 31.625,
      "learning_rate": 1.6146600454025813e-05,
      "loss": 4.5774,
      "num_input_tokens_seen": 33350272,
      "step": 24610
    },
    {
      "epoch": 0.7878816977146149,
      "grad_norm": 22.75,
      "learning_rate": 1.6137419891873317e-05,
      "loss": 4.0521,
      "num_input_tokens_seen": 33357312,
      "step": 24615
    },
    {
      "epoch": 0.7880417386851034,
      "grad_norm": 15.375,
      "learning_rate": 1.6128240696442038e-05,
      "loss": 3.7919,
      "num_input_tokens_seen": 33363760,
      "step": 24620
    },
    {
      "epoch": 0.7882017796555918,
      "grad_norm": 34.25,
      "learning_rate": 1.611906286914753e-05,
      "loss": 4.4389,
      "num_input_tokens_seen": 33370784,
      "step": 24625
    },
    {
      "epoch": 0.7883618206260803,
      "grad_norm": 28.875,
      "learning_rate": 1.6109886411405144e-05,
      "loss": 3.3113,
      "num_input_tokens_seen": 33377728,
      "step": 24630
    },
    {
      "epoch": 0.7885218615965687,
      "grad_norm": 36.75,
      "learning_rate": 1.6100711324629985e-05,
      "loss": 3.8963,
      "num_input_tokens_seen": 33384560,
      "step": 24635
    },
    {
      "epoch": 0.7886819025670572,
      "grad_norm": 14.3125,
      "learning_rate": 1.609153761023698e-05,
      "loss": 4.2368,
      "num_input_tokens_seen": 33391264,
      "step": 24640
    },
    {
      "epoch": 0.7888419435375457,
      "grad_norm": 29.125,
      "learning_rate": 1.608236526964083e-05,
      "loss": 4.0454,
      "num_input_tokens_seen": 33397856,
      "step": 24645
    },
    {
      "epoch": 0.789001984508034,
      "grad_norm": 17.25,
      "learning_rate": 1.607319430425601e-05,
      "loss": 4.2559,
      "num_input_tokens_seen": 33404864,
      "step": 24650
    },
    {
      "epoch": 0.7891620254785225,
      "grad_norm": 14.875,
      "learning_rate": 1.606402471549682e-05,
      "loss": 3.4629,
      "num_input_tokens_seen": 33411712,
      "step": 24655
    },
    {
      "epoch": 0.7893220664490109,
      "grad_norm": 26.25,
      "learning_rate": 1.6054856504777312e-05,
      "loss": 4.2869,
      "num_input_tokens_seen": 33418320,
      "step": 24660
    },
    {
      "epoch": 0.7894821074194994,
      "grad_norm": 31.375,
      "learning_rate": 1.6045689673511334e-05,
      "loss": 4.1227,
      "num_input_tokens_seen": 33424752,
      "step": 24665
    },
    {
      "epoch": 0.7896421483899878,
      "grad_norm": 37.0,
      "learning_rate": 1.6036524223112548e-05,
      "loss": 5.1567,
      "num_input_tokens_seen": 33431392,
      "step": 24670
    },
    {
      "epoch": 0.7898021893604763,
      "grad_norm": 21.25,
      "learning_rate": 1.602736015499436e-05,
      "loss": 3.5095,
      "num_input_tokens_seen": 33438096,
      "step": 24675
    },
    {
      "epoch": 0.7899622303309647,
      "grad_norm": 32.0,
      "learning_rate": 1.601819747057e-05,
      "loss": 4.499,
      "num_input_tokens_seen": 33444608,
      "step": 24680
    },
    {
      "epoch": 0.7901222713014532,
      "grad_norm": 30.375,
      "learning_rate": 1.6009036171252465e-05,
      "loss": 3.854,
      "num_input_tokens_seen": 33451168,
      "step": 24685
    },
    {
      "epoch": 0.7902823122719416,
      "grad_norm": 29.5,
      "learning_rate": 1.599987625845453e-05,
      "loss": 5.077,
      "num_input_tokens_seen": 33457696,
      "step": 24690
    },
    {
      "epoch": 0.7904423532424301,
      "grad_norm": 21.875,
      "learning_rate": 1.599071773358879e-05,
      "loss": 5.0137,
      "num_input_tokens_seen": 33464768,
      "step": 24695
    },
    {
      "epoch": 0.7906023942129186,
      "grad_norm": 45.25,
      "learning_rate": 1.598156059806758e-05,
      "loss": 4.3313,
      "num_input_tokens_seen": 33471856,
      "step": 24700
    },
    {
      "epoch": 0.7907624351834069,
      "grad_norm": 27.75,
      "learning_rate": 1.5972404853303062e-05,
      "loss": 5.0017,
      "num_input_tokens_seen": 33478512,
      "step": 24705
    },
    {
      "epoch": 0.7909224761538954,
      "grad_norm": 35.75,
      "learning_rate": 1.5963250500707172e-05,
      "loss": 4.9864,
      "num_input_tokens_seen": 33485120,
      "step": 24710
    },
    {
      "epoch": 0.7910825171243838,
      "grad_norm": 23.75,
      "learning_rate": 1.5954097541691612e-05,
      "loss": 4.0263,
      "num_input_tokens_seen": 33491856,
      "step": 24715
    },
    {
      "epoch": 0.7912425580948723,
      "grad_norm": 31.25,
      "learning_rate": 1.5944945977667884e-05,
      "loss": 4.4628,
      "num_input_tokens_seen": 33498416,
      "step": 24720
    },
    {
      "epoch": 0.7914025990653607,
      "grad_norm": 26.375,
      "learning_rate": 1.593579581004729e-05,
      "loss": 4.5278,
      "num_input_tokens_seen": 33505360,
      "step": 24725
    },
    {
      "epoch": 0.7915626400358492,
      "grad_norm": 32.25,
      "learning_rate": 1.592664704024088e-05,
      "loss": 4.4109,
      "num_input_tokens_seen": 33511680,
      "step": 24730
    },
    {
      "epoch": 0.7917226810063376,
      "grad_norm": 42.0,
      "learning_rate": 1.591749966965953e-05,
      "loss": 4.4604,
      "num_input_tokens_seen": 33518624,
      "step": 24735
    },
    {
      "epoch": 0.7918827219768261,
      "grad_norm": 25.875,
      "learning_rate": 1.5908353699713856e-05,
      "loss": 4.4153,
      "num_input_tokens_seen": 33525536,
      "step": 24740
    },
    {
      "epoch": 0.7920427629473145,
      "grad_norm": 35.5,
      "learning_rate": 1.5899209131814298e-05,
      "loss": 5.1627,
      "num_input_tokens_seen": 33532272,
      "step": 24745
    },
    {
      "epoch": 0.792202803917803,
      "grad_norm": 45.25,
      "learning_rate": 1.5890065967371067e-05,
      "loss": 4.1718,
      "num_input_tokens_seen": 33539152,
      "step": 24750
    },
    {
      "epoch": 0.7923628448882915,
      "grad_norm": 31.375,
      "learning_rate": 1.5880924207794144e-05,
      "loss": 4.3589,
      "num_input_tokens_seen": 33546048,
      "step": 24755
    },
    {
      "epoch": 0.7925228858587798,
      "grad_norm": 28.125,
      "learning_rate": 1.5871783854493298e-05,
      "loss": 4.7269,
      "num_input_tokens_seen": 33552624,
      "step": 24760
    },
    {
      "epoch": 0.7926829268292683,
      "grad_norm": 34.25,
      "learning_rate": 1.5862644908878106e-05,
      "loss": 4.1292,
      "num_input_tokens_seen": 33559472,
      "step": 24765
    },
    {
      "epoch": 0.7928429677997567,
      "grad_norm": 43.0,
      "learning_rate": 1.5853507372357885e-05,
      "loss": 5.1858,
      "num_input_tokens_seen": 33566448,
      "step": 24770
    },
    {
      "epoch": 0.7930030087702452,
      "grad_norm": 33.5,
      "learning_rate": 1.5844371246341776e-05,
      "loss": 4.2497,
      "num_input_tokens_seen": 33572816,
      "step": 24775
    },
    {
      "epoch": 0.7931630497407336,
      "grad_norm": 19.125,
      "learning_rate": 1.5835236532238674e-05,
      "loss": 3.9225,
      "num_input_tokens_seen": 33579344,
      "step": 24780
    },
    {
      "epoch": 0.7933230907112221,
      "grad_norm": 39.25,
      "learning_rate": 1.582610323145727e-05,
      "loss": 4.133,
      "num_input_tokens_seen": 33585840,
      "step": 24785
    },
    {
      "epoch": 0.7934831316817105,
      "grad_norm": 31.125,
      "learning_rate": 1.5816971345406035e-05,
      "loss": 4.2858,
      "num_input_tokens_seen": 33592528,
      "step": 24790
    },
    {
      "epoch": 0.793643172652199,
      "grad_norm": 24.125,
      "learning_rate": 1.5807840875493225e-05,
      "loss": 3.9676,
      "num_input_tokens_seen": 33599488,
      "step": 24795
    },
    {
      "epoch": 0.7938032136226874,
      "grad_norm": 35.75,
      "learning_rate": 1.5798711823126854e-05,
      "loss": 4.3804,
      "num_input_tokens_seen": 33606096,
      "step": 24800
    },
    {
      "epoch": 0.7938032136226874,
      "eval_loss": 4.461815357208252,
      "eval_runtime": 489.5318,
      "eval_samples_per_second": 28.366,
      "eval_steps_per_second": 14.183,
      "num_input_tokens_seen": 33606096,
      "step": 24800
    },
    {
      "epoch": 0.7939632545931758,
      "grad_norm": 19.375,
      "learning_rate": 1.578958418971477e-05,
      "loss": 3.9625,
      "num_input_tokens_seen": 33612928,
      "step": 24805
    },
    {
      "epoch": 0.7941232955636643,
      "grad_norm": 20.875,
      "learning_rate": 1.578045797666453e-05,
      "loss": 4.2408,
      "num_input_tokens_seen": 33619296,
      "step": 24810
    },
    {
      "epoch": 0.7942833365341527,
      "grad_norm": 28.75,
      "learning_rate": 1.5771333185383548e-05,
      "loss": 4.0257,
      "num_input_tokens_seen": 33625792,
      "step": 24815
    },
    {
      "epoch": 0.7944433775046412,
      "grad_norm": 19.875,
      "learning_rate": 1.576220981727895e-05,
      "loss": 5.2106,
      "num_input_tokens_seen": 33632784,
      "step": 24820
    },
    {
      "epoch": 0.7946034184751296,
      "grad_norm": 57.25,
      "learning_rate": 1.575308787375769e-05,
      "loss": 4.9004,
      "num_input_tokens_seen": 33639520,
      "step": 24825
    },
    {
      "epoch": 0.7947634594456181,
      "grad_norm": 33.0,
      "learning_rate": 1.5743967356226492e-05,
      "loss": 5.2879,
      "num_input_tokens_seen": 33646640,
      "step": 24830
    },
    {
      "epoch": 0.7949235004161065,
      "grad_norm": 31.5,
      "learning_rate": 1.5734848266091835e-05,
      "loss": 4.9028,
      "num_input_tokens_seen": 33653440,
      "step": 24835
    },
    {
      "epoch": 0.795083541386595,
      "grad_norm": 18.0,
      "learning_rate": 1.572573060476001e-05,
      "loss": 3.8314,
      "num_input_tokens_seen": 33660240,
      "step": 24840
    },
    {
      "epoch": 0.7952435823570834,
      "grad_norm": 25.125,
      "learning_rate": 1.5716614373637085e-05,
      "loss": 4.345,
      "num_input_tokens_seen": 33667168,
      "step": 24845
    },
    {
      "epoch": 0.7954036233275719,
      "grad_norm": 34.5,
      "learning_rate": 1.570749957412887e-05,
      "loss": 4.3242,
      "num_input_tokens_seen": 33674064,
      "step": 24850
    },
    {
      "epoch": 0.7955636642980604,
      "grad_norm": 19.0,
      "learning_rate": 1.5698386207641013e-05,
      "loss": 5.0379,
      "num_input_tokens_seen": 33680544,
      "step": 24855
    },
    {
      "epoch": 0.7957237052685487,
      "grad_norm": 34.75,
      "learning_rate": 1.5689274275578884e-05,
      "loss": 3.9106,
      "num_input_tokens_seen": 33687616,
      "step": 24860
    },
    {
      "epoch": 0.7958837462390372,
      "grad_norm": 20.75,
      "learning_rate": 1.5680163779347667e-05,
      "loss": 4.2009,
      "num_input_tokens_seen": 33694608,
      "step": 24865
    },
    {
      "epoch": 0.7960437872095256,
      "grad_norm": 35.0,
      "learning_rate": 1.5671054720352327e-05,
      "loss": 4.9982,
      "num_input_tokens_seen": 33701536,
      "step": 24870
    },
    {
      "epoch": 0.7962038281800141,
      "grad_norm": 30.25,
      "learning_rate": 1.566194709999757e-05,
      "loss": 4.3812,
      "num_input_tokens_seen": 33708064,
      "step": 24875
    },
    {
      "epoch": 0.7963638691505025,
      "grad_norm": 29.25,
      "learning_rate": 1.5652840919687933e-05,
      "loss": 4.3157,
      "num_input_tokens_seen": 33714736,
      "step": 24880
    },
    {
      "epoch": 0.796523910120991,
      "grad_norm": 38.25,
      "learning_rate": 1.5643736180827676e-05,
      "loss": 4.1637,
      "num_input_tokens_seen": 33721536,
      "step": 24885
    },
    {
      "epoch": 0.7966839510914794,
      "grad_norm": 22.875,
      "learning_rate": 1.5634632884820878e-05,
      "loss": 5.1329,
      "num_input_tokens_seen": 33728576,
      "step": 24890
    },
    {
      "epoch": 0.7968439920619679,
      "grad_norm": 27.625,
      "learning_rate": 1.5625531033071395e-05,
      "loss": 4.1344,
      "num_input_tokens_seen": 33735248,
      "step": 24895
    },
    {
      "epoch": 0.7970040330324563,
      "grad_norm": 42.75,
      "learning_rate": 1.5616430626982828e-05,
      "loss": 5.2317,
      "num_input_tokens_seen": 33741808,
      "step": 24900
    },
    {
      "epoch": 0.7971640740029448,
      "grad_norm": 28.25,
      "learning_rate": 1.5607331667958575e-05,
      "loss": 4.0158,
      "num_input_tokens_seen": 33748880,
      "step": 24905
    },
    {
      "epoch": 0.7973241149734333,
      "grad_norm": 21.125,
      "learning_rate": 1.5598234157401824e-05,
      "loss": 3.7215,
      "num_input_tokens_seen": 33755216,
      "step": 24910
    },
    {
      "epoch": 0.7974841559439216,
      "grad_norm": 30.375,
      "learning_rate": 1.5589138096715503e-05,
      "loss": 4.0043,
      "num_input_tokens_seen": 33762048,
      "step": 24915
    },
    {
      "epoch": 0.7976441969144101,
      "grad_norm": 35.5,
      "learning_rate": 1.5580043487302365e-05,
      "loss": 4.3331,
      "num_input_tokens_seen": 33768896,
      "step": 24920
    },
    {
      "epoch": 0.7978042378848985,
      "grad_norm": 22.25,
      "learning_rate": 1.5570950330564888e-05,
      "loss": 4.707,
      "num_input_tokens_seen": 33775760,
      "step": 24925
    },
    {
      "epoch": 0.797964278855387,
      "grad_norm": 14.0,
      "learning_rate": 1.5561858627905367e-05,
      "loss": 3.6253,
      "num_input_tokens_seen": 33782880,
      "step": 24930
    },
    {
      "epoch": 0.7981243198258754,
      "grad_norm": 20.375,
      "learning_rate": 1.5552768380725857e-05,
      "loss": 4.7628,
      "num_input_tokens_seen": 33789904,
      "step": 24935
    },
    {
      "epoch": 0.7982843607963639,
      "grad_norm": 29.625,
      "learning_rate": 1.5543679590428183e-05,
      "loss": 3.728,
      "num_input_tokens_seen": 33797664,
      "step": 24940
    },
    {
      "epoch": 0.7984444017668523,
      "grad_norm": 23.25,
      "learning_rate": 1.5534592258413943e-05,
      "loss": 3.9182,
      "num_input_tokens_seen": 33804240,
      "step": 24945
    },
    {
      "epoch": 0.7986044427373408,
      "grad_norm": 30.625,
      "learning_rate": 1.5525506386084538e-05,
      "loss": 3.9814,
      "num_input_tokens_seen": 33810976,
      "step": 24950
    },
    {
      "epoch": 0.7987644837078292,
      "grad_norm": 32.0,
      "learning_rate": 1.55164219748411e-05,
      "loss": 4.2836,
      "num_input_tokens_seen": 33817680,
      "step": 24955
    },
    {
      "epoch": 0.7989245246783176,
      "grad_norm": 18.375,
      "learning_rate": 1.550733902608459e-05,
      "loss": 3.8687,
      "num_input_tokens_seen": 33824432,
      "step": 24960
    },
    {
      "epoch": 0.7990845656488061,
      "grad_norm": 21.625,
      "learning_rate": 1.549825754121568e-05,
      "loss": 4.1476,
      "num_input_tokens_seen": 33831296,
      "step": 24965
    },
    {
      "epoch": 0.7992446066192945,
      "grad_norm": 49.5,
      "learning_rate": 1.5489177521634864e-05,
      "loss": 4.0492,
      "num_input_tokens_seen": 33838096,
      "step": 24970
    },
    {
      "epoch": 0.799404647589783,
      "grad_norm": 25.375,
      "learning_rate": 1.5480098968742402e-05,
      "loss": 3.9692,
      "num_input_tokens_seen": 33845056,
      "step": 24975
    },
    {
      "epoch": 0.7995646885602714,
      "grad_norm": 26.875,
      "learning_rate": 1.5471021883938304e-05,
      "loss": 3.815,
      "num_input_tokens_seen": 33852064,
      "step": 24980
    },
    {
      "epoch": 0.7997247295307599,
      "grad_norm": 25.5,
      "learning_rate": 1.546194626862238e-05,
      "loss": 3.7704,
      "num_input_tokens_seen": 33858608,
      "step": 24985
    },
    {
      "epoch": 0.7998847705012483,
      "grad_norm": 30.875,
      "learning_rate": 1.5452872124194216e-05,
      "loss": 5.1345,
      "num_input_tokens_seen": 33865376,
      "step": 24990
    },
    {
      "epoch": 0.8000448114717368,
      "grad_norm": 16.5,
      "learning_rate": 1.5443799452053136e-05,
      "loss": 3.7849,
      "num_input_tokens_seen": 33872064,
      "step": 24995
    },
    {
      "epoch": 0.8002048524422252,
      "grad_norm": 42.75,
      "learning_rate": 1.543472825359828e-05,
      "loss": 4.2001,
      "num_input_tokens_seen": 33878720,
      "step": 25000
    },
    {
      "epoch": 0.8002048524422252,
      "eval_loss": 4.494149684906006,
      "eval_runtime": 489.3953,
      "eval_samples_per_second": 28.374,
      "eval_steps_per_second": 14.187,
      "num_input_tokens_seen": 33878720,
      "step": 25000
    },
    {
      "epoch": 0.8003648934127137,
      "grad_norm": 26.625,
      "learning_rate": 1.5425658530228522e-05,
      "loss": 4.4457,
      "num_input_tokens_seen": 33885824,
      "step": 25005
    },
    {
      "epoch": 0.800524934383202,
      "grad_norm": 35.0,
      "learning_rate": 1.5416590283342546e-05,
      "loss": 4.2035,
      "num_input_tokens_seen": 33892576,
      "step": 25010
    },
    {
      "epoch": 0.8006849753536905,
      "grad_norm": 38.0,
      "learning_rate": 1.5407523514338783e-05,
      "loss": 3.4938,
      "num_input_tokens_seen": 33899344,
      "step": 25015
    },
    {
      "epoch": 0.800845016324179,
      "grad_norm": 26.5,
      "learning_rate": 1.539845822461543e-05,
      "loss": 4.2901,
      "num_input_tokens_seen": 33906144,
      "step": 25020
    },
    {
      "epoch": 0.8010050572946674,
      "grad_norm": 29.75,
      "learning_rate": 1.538939441557048e-05,
      "loss": 5.1025,
      "num_input_tokens_seen": 33912768,
      "step": 25025
    },
    {
      "epoch": 0.8011650982651559,
      "grad_norm": 21.75,
      "learning_rate": 1.5380332088601696e-05,
      "loss": 4.0316,
      "num_input_tokens_seen": 33919936,
      "step": 25030
    },
    {
      "epoch": 0.8013251392356443,
      "grad_norm": 13.1875,
      "learning_rate": 1.537127124510658e-05,
      "loss": 4.0283,
      "num_input_tokens_seen": 33926672,
      "step": 25035
    },
    {
      "epoch": 0.8014851802061328,
      "grad_norm": 54.25,
      "learning_rate": 1.5362211886482457e-05,
      "loss": 4.4331,
      "num_input_tokens_seen": 33933488,
      "step": 25040
    },
    {
      "epoch": 0.8016452211766212,
      "grad_norm": 44.5,
      "learning_rate": 1.5353154014126363e-05,
      "loss": 4.2094,
      "num_input_tokens_seen": 33940256,
      "step": 25045
    },
    {
      "epoch": 0.8018052621471097,
      "grad_norm": 16.25,
      "learning_rate": 1.534409762943515e-05,
      "loss": 4.8775,
      "num_input_tokens_seen": 33946656,
      "step": 25050
    },
    {
      "epoch": 0.8019653031175981,
      "grad_norm": 21.0,
      "learning_rate": 1.5335042733805438e-05,
      "loss": 4.1746,
      "num_input_tokens_seen": 33953120,
      "step": 25055
    },
    {
      "epoch": 0.8021253440880866,
      "grad_norm": 39.5,
      "learning_rate": 1.532598932863358e-05,
      "loss": 4.0907,
      "num_input_tokens_seen": 33959968,
      "step": 25060
    },
    {
      "epoch": 0.8022853850585749,
      "grad_norm": 28.125,
      "learning_rate": 1.531693741531574e-05,
      "loss": 4.0412,
      "num_input_tokens_seen": 33967248,
      "step": 25065
    },
    {
      "epoch": 0.8024454260290634,
      "grad_norm": 35.5,
      "learning_rate": 1.5307886995247844e-05,
      "loss": 4.0855,
      "num_input_tokens_seen": 33973664,
      "step": 25070
    },
    {
      "epoch": 0.8026054669995519,
      "grad_norm": 26.875,
      "learning_rate": 1.529883806982557e-05,
      "loss": 4.5592,
      "num_input_tokens_seen": 33980032,
      "step": 25075
    },
    {
      "epoch": 0.8027655079700403,
      "grad_norm": 46.25,
      "learning_rate": 1.5289790640444376e-05,
      "loss": 4.1145,
      "num_input_tokens_seen": 33986736,
      "step": 25080
    },
    {
      "epoch": 0.8029255489405288,
      "grad_norm": 11.5,
      "learning_rate": 1.5280744708499494e-05,
      "loss": 3.9015,
      "num_input_tokens_seen": 33993584,
      "step": 25085
    },
    {
      "epoch": 0.8030855899110172,
      "grad_norm": 45.75,
      "learning_rate": 1.527170027538591e-05,
      "loss": 3.9453,
      "num_input_tokens_seen": 34000368,
      "step": 25090
    },
    {
      "epoch": 0.8032456308815057,
      "grad_norm": 30.375,
      "learning_rate": 1.5262657342498407e-05,
      "loss": 4.3477,
      "num_input_tokens_seen": 34006752,
      "step": 25095
    },
    {
      "epoch": 0.8034056718519941,
      "grad_norm": 72.0,
      "learning_rate": 1.52536159112315e-05,
      "loss": 3.3949,
      "num_input_tokens_seen": 34013648,
      "step": 25100
    },
    {
      "epoch": 0.8035657128224826,
      "grad_norm": 19.75,
      "learning_rate": 1.5244575982979497e-05,
      "loss": 4.1681,
      "num_input_tokens_seen": 34020080,
      "step": 25105
    },
    {
      "epoch": 0.803725753792971,
      "grad_norm": 22.0,
      "learning_rate": 1.5235537559136487e-05,
      "loss": 4.0478,
      "num_input_tokens_seen": 34026992,
      "step": 25110
    },
    {
      "epoch": 0.8038857947634594,
      "grad_norm": 15.375,
      "learning_rate": 1.5226500641096286e-05,
      "loss": 3.3499,
      "num_input_tokens_seen": 34033552,
      "step": 25115
    },
    {
      "epoch": 0.804045835733948,
      "grad_norm": 26.75,
      "learning_rate": 1.5217465230252509e-05,
      "loss": 4.8615,
      "num_input_tokens_seen": 34040096,
      "step": 25120
    },
    {
      "epoch": 0.8042058767044363,
      "grad_norm": 25.0,
      "learning_rate": 1.5208431327998523e-05,
      "loss": 4.3058,
      "num_input_tokens_seen": 34046896,
      "step": 25125
    },
    {
      "epoch": 0.8043659176749248,
      "grad_norm": 29.875,
      "learning_rate": 1.5199398935727477e-05,
      "loss": 4.6081,
      "num_input_tokens_seen": 34053728,
      "step": 25130
    },
    {
      "epoch": 0.8045259586454132,
      "grad_norm": 24.625,
      "learning_rate": 1.5190368054832282e-05,
      "loss": 3.9066,
      "num_input_tokens_seen": 34060144,
      "step": 25135
    },
    {
      "epoch": 0.8046859996159017,
      "grad_norm": 64.5,
      "learning_rate": 1.5181338686705601e-05,
      "loss": 4.5501,
      "num_input_tokens_seen": 34067120,
      "step": 25140
    },
    {
      "epoch": 0.8048460405863901,
      "grad_norm": 53.5,
      "learning_rate": 1.5172310832739889e-05,
      "loss": 5.1629,
      "num_input_tokens_seen": 34073920,
      "step": 25145
    },
    {
      "epoch": 0.8050060815568786,
      "grad_norm": 31.125,
      "learning_rate": 1.5163284494327346e-05,
      "loss": 4.1833,
      "num_input_tokens_seen": 34080576,
      "step": 25150
    },
    {
      "epoch": 0.805166122527367,
      "grad_norm": 27.625,
      "learning_rate": 1.5154259672859952e-05,
      "loss": 4.7057,
      "num_input_tokens_seen": 34087792,
      "step": 25155
    },
    {
      "epoch": 0.8053261634978555,
      "grad_norm": 28.375,
      "learning_rate": 1.5145236369729452e-05,
      "loss": 4.4067,
      "num_input_tokens_seen": 34094512,
      "step": 25160
    },
    {
      "epoch": 0.8054862044683438,
      "grad_norm": 24.75,
      "learning_rate": 1.5136214586327335e-05,
      "loss": 3.9555,
      "num_input_tokens_seen": 34101344,
      "step": 25165
    },
    {
      "epoch": 0.8056462454388323,
      "grad_norm": 30.5,
      "learning_rate": 1.5127194324044885e-05,
      "loss": 4.2645,
      "num_input_tokens_seen": 34108304,
      "step": 25170
    },
    {
      "epoch": 0.8058062864093208,
      "grad_norm": 35.75,
      "learning_rate": 1.5118175584273148e-05,
      "loss": 4.1035,
      "num_input_tokens_seen": 34115136,
      "step": 25175
    },
    {
      "epoch": 0.8059663273798092,
      "grad_norm": 35.5,
      "learning_rate": 1.5109158368402909e-05,
      "loss": 4.9778,
      "num_input_tokens_seen": 34121744,
      "step": 25180
    },
    {
      "epoch": 0.8061263683502977,
      "grad_norm": 21.75,
      "learning_rate": 1.5100142677824753e-05,
      "loss": 4.9576,
      "num_input_tokens_seen": 34128400,
      "step": 25185
    },
    {
      "epoch": 0.8062864093207861,
      "grad_norm": 24.125,
      "learning_rate": 1.509112851392901e-05,
      "loss": 3.8875,
      "num_input_tokens_seen": 34134880,
      "step": 25190
    },
    {
      "epoch": 0.8064464502912746,
      "grad_norm": 26.125,
      "learning_rate": 1.5082115878105763e-05,
      "loss": 4.1273,
      "num_input_tokens_seen": 34141504,
      "step": 25195
    },
    {
      "epoch": 0.806606491261763,
      "grad_norm": 35.25,
      "learning_rate": 1.5073104771744892e-05,
      "loss": 4.4727,
      "num_input_tokens_seen": 34148496,
      "step": 25200
    },
    {
      "epoch": 0.806606491261763,
      "eval_loss": 4.482816696166992,
      "eval_runtime": 490.1243,
      "eval_samples_per_second": 28.332,
      "eval_steps_per_second": 14.166,
      "num_input_tokens_seen": 34148496,
      "step": 25200
    },
    {
      "epoch": 0.8067665322322515,
      "grad_norm": 17.0,
      "learning_rate": 1.5064095196236006e-05,
      "loss": 3.5504,
      "num_input_tokens_seen": 34155136,
      "step": 25205
    },
    {
      "epoch": 0.8069265732027399,
      "grad_norm": 37.0,
      "learning_rate": 1.50550871529685e-05,
      "loss": 3.413,
      "num_input_tokens_seen": 34162016,
      "step": 25210
    },
    {
      "epoch": 0.8070866141732284,
      "grad_norm": 26.125,
      "learning_rate": 1.5046080643331546e-05,
      "loss": 4.8567,
      "num_input_tokens_seen": 34169072,
      "step": 25215
    },
    {
      "epoch": 0.8072466551437167,
      "grad_norm": 18.5,
      "learning_rate": 1.5037075668714028e-05,
      "loss": 4.9121,
      "num_input_tokens_seen": 34176176,
      "step": 25220
    },
    {
      "epoch": 0.8074066961142052,
      "grad_norm": 22.75,
      "learning_rate": 1.5028072230504656e-05,
      "loss": 4.6347,
      "num_input_tokens_seen": 34182736,
      "step": 25225
    },
    {
      "epoch": 0.8075667370846937,
      "grad_norm": 33.0,
      "learning_rate": 1.5019070330091861e-05,
      "loss": 3.9771,
      "num_input_tokens_seen": 34189536,
      "step": 25230
    },
    {
      "epoch": 0.8077267780551821,
      "grad_norm": 33.0,
      "learning_rate": 1.5010069968863843e-05,
      "loss": 3.9405,
      "num_input_tokens_seen": 34196400,
      "step": 25235
    },
    {
      "epoch": 0.8078868190256706,
      "grad_norm": 27.625,
      "learning_rate": 1.5001071148208584e-05,
      "loss": 4.7093,
      "num_input_tokens_seen": 34203056,
      "step": 25240
    },
    {
      "epoch": 0.808046859996159,
      "grad_norm": 21.25,
      "learning_rate": 1.49920738695138e-05,
      "loss": 3.4865,
      "num_input_tokens_seen": 34210112,
      "step": 25245
    },
    {
      "epoch": 0.8082069009666475,
      "grad_norm": 36.25,
      "learning_rate": 1.4983078134166995e-05,
      "loss": 4.5884,
      "num_input_tokens_seen": 34216704,
      "step": 25250
    },
    {
      "epoch": 0.8083669419371359,
      "grad_norm": 39.0,
      "learning_rate": 1.4974083943555428e-05,
      "loss": 4.4159,
      "num_input_tokens_seen": 34223376,
      "step": 25255
    },
    {
      "epoch": 0.8085269829076244,
      "grad_norm": 21.375,
      "learning_rate": 1.496509129906611e-05,
      "loss": 5.1084,
      "num_input_tokens_seen": 34230416,
      "step": 25260
    },
    {
      "epoch": 0.8086870238781128,
      "grad_norm": 34.0,
      "learning_rate": 1.4956100202085809e-05,
      "loss": 5.1563,
      "num_input_tokens_seen": 34237072,
      "step": 25265
    },
    {
      "epoch": 0.8088470648486012,
      "grad_norm": 45.25,
      "learning_rate": 1.4947110654001093e-05,
      "loss": 4.4006,
      "num_input_tokens_seen": 34243696,
      "step": 25270
    },
    {
      "epoch": 0.8090071058190896,
      "grad_norm": 47.0,
      "learning_rate": 1.4938122656198234e-05,
      "loss": 5.0531,
      "num_input_tokens_seen": 34251296,
      "step": 25275
    },
    {
      "epoch": 0.8091671467895781,
      "grad_norm": 23.375,
      "learning_rate": 1.4929136210063316e-05,
      "loss": 4.3693,
      "num_input_tokens_seen": 34258320,
      "step": 25280
    },
    {
      "epoch": 0.8093271877600666,
      "grad_norm": 16.125,
      "learning_rate": 1.4920151316982146e-05,
      "loss": 3.9026,
      "num_input_tokens_seen": 34265152,
      "step": 25285
    },
    {
      "epoch": 0.809487228730555,
      "grad_norm": 33.25,
      "learning_rate": 1.4911167978340312e-05,
      "loss": 4.6232,
      "num_input_tokens_seen": 34272096,
      "step": 25290
    },
    {
      "epoch": 0.8096472697010435,
      "grad_norm": 42.5,
      "learning_rate": 1.4902186195523166e-05,
      "loss": 4.5706,
      "num_input_tokens_seen": 34278464,
      "step": 25295
    },
    {
      "epoch": 0.8098073106715319,
      "grad_norm": 59.75,
      "learning_rate": 1.4893205969915805e-05,
      "loss": 4.1765,
      "num_input_tokens_seen": 34284976,
      "step": 25300
    },
    {
      "epoch": 0.8099673516420204,
      "grad_norm": 18.0,
      "learning_rate": 1.4884227302903086e-05,
      "loss": 4.7919,
      "num_input_tokens_seen": 34291440,
      "step": 25305
    },
    {
      "epoch": 0.8101273926125088,
      "grad_norm": 32.5,
      "learning_rate": 1.4875250195869653e-05,
      "loss": 4.1159,
      "num_input_tokens_seen": 34298192,
      "step": 25310
    },
    {
      "epoch": 0.8102874335829973,
      "grad_norm": 31.875,
      "learning_rate": 1.4866274650199862e-05,
      "loss": 3.3016,
      "num_input_tokens_seen": 34304800,
      "step": 25315
    },
    {
      "epoch": 0.8104474745534856,
      "grad_norm": 17.5,
      "learning_rate": 1.485730066727788e-05,
      "loss": 4.0088,
      "num_input_tokens_seen": 34311424,
      "step": 25320
    },
    {
      "epoch": 0.8106075155239741,
      "grad_norm": 26.0,
      "learning_rate": 1.4848328248487586e-05,
      "loss": 4.4819,
      "num_input_tokens_seen": 34317984,
      "step": 25325
    },
    {
      "epoch": 0.8107675564944626,
      "grad_norm": 31.375,
      "learning_rate": 1.4839357395212656e-05,
      "loss": 4.8962,
      "num_input_tokens_seen": 34324896,
      "step": 25330
    },
    {
      "epoch": 0.810927597464951,
      "grad_norm": 31.5,
      "learning_rate": 1.4830388108836502e-05,
      "loss": 5.2621,
      "num_input_tokens_seen": 34331888,
      "step": 25335
    },
    {
      "epoch": 0.8110876384354395,
      "grad_norm": 22.625,
      "learning_rate": 1.4821420390742299e-05,
      "loss": 4.7271,
      "num_input_tokens_seen": 34338832,
      "step": 25340
    },
    {
      "epoch": 0.8112476794059279,
      "grad_norm": 24.375,
      "learning_rate": 1.4812454242312979e-05,
      "loss": 3.8804,
      "num_input_tokens_seen": 34345472,
      "step": 25345
    },
    {
      "epoch": 0.8114077203764164,
      "grad_norm": 22.875,
      "learning_rate": 1.4803489664931253e-05,
      "loss": 5.0172,
      "num_input_tokens_seen": 34352192,
      "step": 25350
    },
    {
      "epoch": 0.8115677613469048,
      "grad_norm": 23.25,
      "learning_rate": 1.4794526659979544e-05,
      "loss": 4.4998,
      "num_input_tokens_seen": 34359136,
      "step": 25355
    },
    {
      "epoch": 0.8117278023173933,
      "grad_norm": 41.75,
      "learning_rate": 1.4785565228840086e-05,
      "loss": 3.8199,
      "num_input_tokens_seen": 34366192,
      "step": 25360
    },
    {
      "epoch": 0.8118878432878817,
      "grad_norm": 28.875,
      "learning_rate": 1.4776605372894819e-05,
      "loss": 4.9228,
      "num_input_tokens_seen": 34373280,
      "step": 25365
    },
    {
      "epoch": 0.8120478842583702,
      "grad_norm": 38.5,
      "learning_rate": 1.4767647093525488e-05,
      "loss": 4.8567,
      "num_input_tokens_seen": 34379824,
      "step": 25370
    },
    {
      "epoch": 0.8122079252288585,
      "grad_norm": 35.25,
      "learning_rate": 1.4758690392113566e-05,
      "loss": 3.9377,
      "num_input_tokens_seen": 34387104,
      "step": 25375
    },
    {
      "epoch": 0.812367966199347,
      "grad_norm": 29.375,
      "learning_rate": 1.4749735270040276e-05,
      "loss": 4.9597,
      "num_input_tokens_seen": 34393872,
      "step": 25380
    },
    {
      "epoch": 0.8125280071698355,
      "grad_norm": 30.375,
      "learning_rate": 1.4740781728686623e-05,
      "loss": 3.6765,
      "num_input_tokens_seen": 34400384,
      "step": 25385
    },
    {
      "epoch": 0.8126880481403239,
      "grad_norm": 28.25,
      "learning_rate": 1.4731829769433358e-05,
      "loss": 4.1426,
      "num_input_tokens_seen": 34407392,
      "step": 25390
    },
    {
      "epoch": 0.8128480891108124,
      "grad_norm": 25.375,
      "learning_rate": 1.4722879393660976e-05,
      "loss": 4.2695,
      "num_input_tokens_seen": 34414240,
      "step": 25395
    },
    {
      "epoch": 0.8130081300813008,
      "grad_norm": 42.0,
      "learning_rate": 1.4713930602749748e-05,
      "loss": 4.79,
      "num_input_tokens_seen": 34421392,
      "step": 25400
    },
    {
      "epoch": 0.8130081300813008,
      "eval_loss": 4.5009541511535645,
      "eval_runtime": 489.7712,
      "eval_samples_per_second": 28.352,
      "eval_steps_per_second": 14.176,
      "num_input_tokens_seen": 34421392,
      "step": 25400
    },
    {
      "epoch": 0.8131681710517893,
      "grad_norm": 22.625,
      "learning_rate": 1.470498339807968e-05,
      "loss": 4.256,
      "num_input_tokens_seen": 34428288,
      "step": 25405
    },
    {
      "epoch": 0.8133282120222777,
      "grad_norm": 37.0,
      "learning_rate": 1.4696037781030542e-05,
      "loss": 5.0067,
      "num_input_tokens_seen": 34435104,
      "step": 25410
    },
    {
      "epoch": 0.8134882529927662,
      "grad_norm": 36.5,
      "learning_rate": 1.4687093752981876e-05,
      "loss": 4.1044,
      "num_input_tokens_seen": 34441968,
      "step": 25415
    },
    {
      "epoch": 0.8136482939632546,
      "grad_norm": 20.25,
      "learning_rate": 1.4678151315312943e-05,
      "loss": 4.5726,
      "num_input_tokens_seen": 34448768,
      "step": 25420
    },
    {
      "epoch": 0.813808334933743,
      "grad_norm": 26.875,
      "learning_rate": 1.4669210469402789e-05,
      "loss": 4.7189,
      "num_input_tokens_seen": 34455584,
      "step": 25425
    },
    {
      "epoch": 0.8139683759042314,
      "grad_norm": 22.125,
      "learning_rate": 1.4660271216630218e-05,
      "loss": 3.8836,
      "num_input_tokens_seen": 34462880,
      "step": 25430
    },
    {
      "epoch": 0.8141284168747199,
      "grad_norm": 21.125,
      "learning_rate": 1.4651333558373748e-05,
      "loss": 4.1295,
      "num_input_tokens_seen": 34469744,
      "step": 25435
    },
    {
      "epoch": 0.8142884578452084,
      "grad_norm": 26.5,
      "learning_rate": 1.4642397496011707e-05,
      "loss": 3.6969,
      "num_input_tokens_seen": 34476592,
      "step": 25440
    },
    {
      "epoch": 0.8144484988156968,
      "grad_norm": 31.5,
      "learning_rate": 1.4633463030922129e-05,
      "loss": 4.7248,
      "num_input_tokens_seen": 34483232,
      "step": 25445
    },
    {
      "epoch": 0.8146085397861853,
      "grad_norm": 33.25,
      "learning_rate": 1.462453016448282e-05,
      "loss": 4.4019,
      "num_input_tokens_seen": 34490176,
      "step": 25450
    },
    {
      "epoch": 0.8147685807566737,
      "grad_norm": 43.75,
      "learning_rate": 1.4615598898071354e-05,
      "loss": 4.8365,
      "num_input_tokens_seen": 34496960,
      "step": 25455
    },
    {
      "epoch": 0.8149286217271622,
      "grad_norm": 45.0,
      "learning_rate": 1.4606669233065026e-05,
      "loss": 3.8289,
      "num_input_tokens_seen": 34503824,
      "step": 25460
    },
    {
      "epoch": 0.8150886626976506,
      "grad_norm": 31.125,
      "learning_rate": 1.4597741170840914e-05,
      "loss": 3.8161,
      "num_input_tokens_seen": 34510480,
      "step": 25465
    },
    {
      "epoch": 0.8152487036681391,
      "grad_norm": 18.5,
      "learning_rate": 1.4588814712775853e-05,
      "loss": 4.1862,
      "num_input_tokens_seen": 34517344,
      "step": 25470
    },
    {
      "epoch": 0.8154087446386274,
      "grad_norm": 37.25,
      "learning_rate": 1.4579889860246382e-05,
      "loss": 4.9695,
      "num_input_tokens_seen": 34524112,
      "step": 25475
    },
    {
      "epoch": 0.8155687856091159,
      "grad_norm": 28.625,
      "learning_rate": 1.457096661462885e-05,
      "loss": 4.1742,
      "num_input_tokens_seen": 34530480,
      "step": 25480
    },
    {
      "epoch": 0.8157288265796043,
      "grad_norm": 16.125,
      "learning_rate": 1.4562044977299322e-05,
      "loss": 4.5775,
      "num_input_tokens_seen": 34537152,
      "step": 25485
    },
    {
      "epoch": 0.8158888675500928,
      "grad_norm": 64.5,
      "learning_rate": 1.4553124949633623e-05,
      "loss": 4.5465,
      "num_input_tokens_seen": 34544016,
      "step": 25490
    },
    {
      "epoch": 0.8160489085205813,
      "grad_norm": 27.625,
      "learning_rate": 1.4544206533007354e-05,
      "loss": 4.4044,
      "num_input_tokens_seen": 34550608,
      "step": 25495
    },
    {
      "epoch": 0.8162089494910697,
      "grad_norm": 19.375,
      "learning_rate": 1.4535289728795821e-05,
      "loss": 3.5624,
      "num_input_tokens_seen": 34557760,
      "step": 25500
    },
    {
      "epoch": 0.8163689904615582,
      "grad_norm": 47.0,
      "learning_rate": 1.4526374538374132e-05,
      "loss": 4.9892,
      "num_input_tokens_seen": 34564256,
      "step": 25505
    },
    {
      "epoch": 0.8165290314320466,
      "grad_norm": 28.375,
      "learning_rate": 1.4517460963117097e-05,
      "loss": 4.4842,
      "num_input_tokens_seen": 34571296,
      "step": 25510
    },
    {
      "epoch": 0.8166890724025351,
      "grad_norm": 30.375,
      "learning_rate": 1.4508549004399314e-05,
      "loss": 4.5272,
      "num_input_tokens_seen": 34577968,
      "step": 25515
    },
    {
      "epoch": 0.8168491133730235,
      "grad_norm": 19.625,
      "learning_rate": 1.449963866359513e-05,
      "loss": 4.8789,
      "num_input_tokens_seen": 34584816,
      "step": 25520
    },
    {
      "epoch": 0.817009154343512,
      "grad_norm": 30.375,
      "learning_rate": 1.4490729942078607e-05,
      "loss": 4.317,
      "num_input_tokens_seen": 34591344,
      "step": 25525
    },
    {
      "epoch": 0.8171691953140003,
      "grad_norm": 26.75,
      "learning_rate": 1.4481822841223608e-05,
      "loss": 3.7215,
      "num_input_tokens_seen": 34597776,
      "step": 25530
    },
    {
      "epoch": 0.8173292362844888,
      "grad_norm": 29.0,
      "learning_rate": 1.4472917362403704e-05,
      "loss": 4.13,
      "num_input_tokens_seen": 34604384,
      "step": 25535
    },
    {
      "epoch": 0.8174892772549772,
      "grad_norm": 31.5,
      "learning_rate": 1.4464013506992224e-05,
      "loss": 4.3065,
      "num_input_tokens_seen": 34611328,
      "step": 25540
    },
    {
      "epoch": 0.8176493182254657,
      "grad_norm": 39.75,
      "learning_rate": 1.4455111276362277e-05,
      "loss": 4.5751,
      "num_input_tokens_seen": 34618448,
      "step": 25545
    },
    {
      "epoch": 0.8178093591959542,
      "grad_norm": 15.25,
      "learning_rate": 1.4446210671886676e-05,
      "loss": 3.8976,
      "num_input_tokens_seen": 34625328,
      "step": 25550
    },
    {
      "epoch": 0.8179694001664426,
      "grad_norm": 24.375,
      "learning_rate": 1.4437311694938015e-05,
      "loss": 3.5381,
      "num_input_tokens_seen": 34631920,
      "step": 25555
    },
    {
      "epoch": 0.8181294411369311,
      "grad_norm": 24.125,
      "learning_rate": 1.442841434688864e-05,
      "loss": 4.1434,
      "num_input_tokens_seen": 34638480,
      "step": 25560
    },
    {
      "epoch": 0.8182894821074195,
      "grad_norm": 28.625,
      "learning_rate": 1.4419518629110615e-05,
      "loss": 3.9699,
      "num_input_tokens_seen": 34645072,
      "step": 25565
    },
    {
      "epoch": 0.818449523077908,
      "grad_norm": 41.25,
      "learning_rate": 1.4410624542975778e-05,
      "loss": 3.9419,
      "num_input_tokens_seen": 34652528,
      "step": 25570
    },
    {
      "epoch": 0.8186095640483964,
      "grad_norm": 17.875,
      "learning_rate": 1.4401732089855724e-05,
      "loss": 3.8809,
      "num_input_tokens_seen": 34658976,
      "step": 25575
    },
    {
      "epoch": 0.8187696050188848,
      "grad_norm": 28.5,
      "learning_rate": 1.4392841271121754e-05,
      "loss": 5.3976,
      "num_input_tokens_seen": 34665488,
      "step": 25580
    },
    {
      "epoch": 0.8189296459893732,
      "grad_norm": 72.0,
      "learning_rate": 1.438395208814497e-05,
      "loss": 4.1933,
      "num_input_tokens_seen": 34672624,
      "step": 25585
    },
    {
      "epoch": 0.8190896869598617,
      "grad_norm": 27.875,
      "learning_rate": 1.4375064542296174e-05,
      "loss": 4.4612,
      "num_input_tokens_seen": 34679152,
      "step": 25590
    },
    {
      "epoch": 0.8192497279303502,
      "grad_norm": 27.75,
      "learning_rate": 1.4366178634945946e-05,
      "loss": 4.2138,
      "num_input_tokens_seen": 34685952,
      "step": 25595
    },
    {
      "epoch": 0.8194097689008386,
      "grad_norm": 28.75,
      "learning_rate": 1.4357294367464616e-05,
      "loss": 4.8489,
      "num_input_tokens_seen": 34692880,
      "step": 25600
    },
    {
      "epoch": 0.8194097689008386,
      "eval_loss": 4.487926959991455,
      "eval_runtime": 489.2516,
      "eval_samples_per_second": 28.382,
      "eval_steps_per_second": 14.191,
      "num_input_tokens_seen": 34692880,
      "step": 25600
    },
    {
      "epoch": 0.8195698098713271,
      "grad_norm": 50.25,
      "learning_rate": 1.434841174122224e-05,
      "loss": 4.0463,
      "num_input_tokens_seen": 34699712,
      "step": 25605
    },
    {
      "epoch": 0.8197298508418155,
      "grad_norm": 32.5,
      "learning_rate": 1.4339530757588615e-05,
      "loss": 4.6825,
      "num_input_tokens_seen": 34706976,
      "step": 25610
    },
    {
      "epoch": 0.819889891812304,
      "grad_norm": 39.25,
      "learning_rate": 1.433065141793333e-05,
      "loss": 5.2805,
      "num_input_tokens_seen": 34713904,
      "step": 25615
    },
    {
      "epoch": 0.8200499327827924,
      "grad_norm": 48.0,
      "learning_rate": 1.4321773723625665e-05,
      "loss": 4.6843,
      "num_input_tokens_seen": 34720160,
      "step": 25620
    },
    {
      "epoch": 0.8202099737532809,
      "grad_norm": 31.875,
      "learning_rate": 1.4312897676034693e-05,
      "loss": 4.8113,
      "num_input_tokens_seen": 34726816,
      "step": 25625
    },
    {
      "epoch": 0.8203700147237692,
      "grad_norm": 24.875,
      "learning_rate": 1.4304023276529188e-05,
      "loss": 4.4352,
      "num_input_tokens_seen": 34733584,
      "step": 25630
    },
    {
      "epoch": 0.8205300556942577,
      "grad_norm": 29.75,
      "learning_rate": 1.4295150526477712e-05,
      "loss": 3.4482,
      "num_input_tokens_seen": 34740544,
      "step": 25635
    },
    {
      "epoch": 0.8206900966647461,
      "grad_norm": 27.25,
      "learning_rate": 1.4286279427248562e-05,
      "loss": 3.601,
      "num_input_tokens_seen": 34747504,
      "step": 25640
    },
    {
      "epoch": 0.8208501376352346,
      "grad_norm": 34.75,
      "learning_rate": 1.4277409980209747e-05,
      "loss": 4.3094,
      "num_input_tokens_seen": 34754624,
      "step": 25645
    },
    {
      "epoch": 0.8210101786057231,
      "grad_norm": 30.375,
      "learning_rate": 1.4268542186729061e-05,
      "loss": 4.714,
      "num_input_tokens_seen": 34761376,
      "step": 25650
    },
    {
      "epoch": 0.8211702195762115,
      "grad_norm": 30.875,
      "learning_rate": 1.4259676048174043e-05,
      "loss": 3.8153,
      "num_input_tokens_seen": 34767888,
      "step": 25655
    },
    {
      "epoch": 0.8213302605467,
      "grad_norm": 25.25,
      "learning_rate": 1.4250811565911937e-05,
      "loss": 5.077,
      "num_input_tokens_seen": 34774432,
      "step": 25660
    },
    {
      "epoch": 0.8214903015171884,
      "grad_norm": 32.5,
      "learning_rate": 1.4241948741309782e-05,
      "loss": 4.3457,
      "num_input_tokens_seen": 34781152,
      "step": 25665
    },
    {
      "epoch": 0.8216503424876769,
      "grad_norm": 33.75,
      "learning_rate": 1.4233087575734317e-05,
      "loss": 4.5215,
      "num_input_tokens_seen": 34787584,
      "step": 25670
    },
    {
      "epoch": 0.8218103834581653,
      "grad_norm": 19.375,
      "learning_rate": 1.422422807055206e-05,
      "loss": 3.6159,
      "num_input_tokens_seen": 34794624,
      "step": 25675
    },
    {
      "epoch": 0.8219704244286538,
      "grad_norm": 47.25,
      "learning_rate": 1.4215370227129243e-05,
      "loss": 4.7036,
      "num_input_tokens_seen": 34801392,
      "step": 25680
    },
    {
      "epoch": 0.8221304653991421,
      "grad_norm": 22.75,
      "learning_rate": 1.4206514046831876e-05,
      "loss": 4.5202,
      "num_input_tokens_seen": 34808208,
      "step": 25685
    },
    {
      "epoch": 0.8222905063696306,
      "grad_norm": 24.125,
      "learning_rate": 1.419765953102567e-05,
      "loss": 4.0238,
      "num_input_tokens_seen": 34814864,
      "step": 25690
    },
    {
      "epoch": 0.822450547340119,
      "grad_norm": 30.875,
      "learning_rate": 1.4188806681076125e-05,
      "loss": 4.3143,
      "num_input_tokens_seen": 34821472,
      "step": 25695
    },
    {
      "epoch": 0.8226105883106075,
      "grad_norm": 38.75,
      "learning_rate": 1.4179955498348443e-05,
      "loss": 4.7563,
      "num_input_tokens_seen": 34828064,
      "step": 25700
    },
    {
      "epoch": 0.822770629281096,
      "grad_norm": 32.25,
      "learning_rate": 1.4171105984207605e-05,
      "loss": 5.1614,
      "num_input_tokens_seen": 34834848,
      "step": 25705
    },
    {
      "epoch": 0.8229306702515844,
      "grad_norm": 19.5,
      "learning_rate": 1.4162258140018304e-05,
      "loss": 3.9246,
      "num_input_tokens_seen": 34841712,
      "step": 25710
    },
    {
      "epoch": 0.8230907112220729,
      "grad_norm": 32.75,
      "learning_rate": 1.4153411967144986e-05,
      "loss": 4.5581,
      "num_input_tokens_seen": 34848800,
      "step": 25715
    },
    {
      "epoch": 0.8232507521925613,
      "grad_norm": 36.75,
      "learning_rate": 1.4144567466951864e-05,
      "loss": 4.5274,
      "num_input_tokens_seen": 34855184,
      "step": 25720
    },
    {
      "epoch": 0.8234107931630498,
      "grad_norm": 31.25,
      "learning_rate": 1.4135724640802844e-05,
      "loss": 4.3866,
      "num_input_tokens_seen": 34862160,
      "step": 25725
    },
    {
      "epoch": 0.8235708341335382,
      "grad_norm": 31.625,
      "learning_rate": 1.4126883490061615e-05,
      "loss": 4.7235,
      "num_input_tokens_seen": 34868768,
      "step": 25730
    },
    {
      "epoch": 0.8237308751040266,
      "grad_norm": 23.375,
      "learning_rate": 1.4118044016091603e-05,
      "loss": 4.6099,
      "num_input_tokens_seen": 34875296,
      "step": 25735
    },
    {
      "epoch": 0.823890916074515,
      "grad_norm": 41.75,
      "learning_rate": 1.410920622025594e-05,
      "loss": 5.3223,
      "num_input_tokens_seen": 34882064,
      "step": 25740
    },
    {
      "epoch": 0.8240509570450035,
      "grad_norm": 32.25,
      "learning_rate": 1.4100370103917554e-05,
      "loss": 4.2531,
      "num_input_tokens_seen": 34889008,
      "step": 25745
    },
    {
      "epoch": 0.8242109980154919,
      "grad_norm": 23.75,
      "learning_rate": 1.409153566843907e-05,
      "loss": 4.2786,
      "num_input_tokens_seen": 34895616,
      "step": 25750
    },
    {
      "epoch": 0.8243710389859804,
      "grad_norm": 36.75,
      "learning_rate": 1.408270291518286e-05,
      "loss": 5.4624,
      "num_input_tokens_seen": 34902624,
      "step": 25755
    },
    {
      "epoch": 0.8245310799564689,
      "grad_norm": 41.75,
      "learning_rate": 1.407387184551107e-05,
      "loss": 4.9804,
      "num_input_tokens_seen": 34909312,
      "step": 25760
    },
    {
      "epoch": 0.8246911209269573,
      "grad_norm": 25.75,
      "learning_rate": 1.4065042460785532e-05,
      "loss": 4.0351,
      "num_input_tokens_seen": 34916176,
      "step": 25765
    },
    {
      "epoch": 0.8248511618974458,
      "grad_norm": 22.375,
      "learning_rate": 1.405621476236787e-05,
      "loss": 3.7928,
      "num_input_tokens_seen": 34922992,
      "step": 25770
    },
    {
      "epoch": 0.8250112028679342,
      "grad_norm": 27.5,
      "learning_rate": 1.4047388751619423e-05,
      "loss": 3.7893,
      "num_input_tokens_seen": 34929440,
      "step": 25775
    },
    {
      "epoch": 0.8251712438384227,
      "grad_norm": 25.125,
      "learning_rate": 1.4038564429901264e-05,
      "loss": 3.5308,
      "num_input_tokens_seen": 34936096,
      "step": 25780
    },
    {
      "epoch": 0.825331284808911,
      "grad_norm": 19.0,
      "learning_rate": 1.4029741798574227e-05,
      "loss": 4.0012,
      "num_input_tokens_seen": 34943056,
      "step": 25785
    },
    {
      "epoch": 0.8254913257793995,
      "grad_norm": 28.5,
      "learning_rate": 1.402092085899886e-05,
      "loss": 2.9573,
      "num_input_tokens_seen": 34949904,
      "step": 25790
    },
    {
      "epoch": 0.8256513667498879,
      "grad_norm": 33.0,
      "learning_rate": 1.4012101612535464e-05,
      "loss": 4.8736,
      "num_input_tokens_seen": 34956992,
      "step": 25795
    },
    {
      "epoch": 0.8258114077203764,
      "grad_norm": 25.25,
      "learning_rate": 1.4003284060544092e-05,
      "loss": 4.0376,
      "num_input_tokens_seen": 34964656,
      "step": 25800
    },
    {
      "epoch": 0.8258114077203764,
      "eval_loss": 4.495737552642822,
      "eval_runtime": 489.7129,
      "eval_samples_per_second": 28.355,
      "eval_steps_per_second": 14.178,
      "num_input_tokens_seen": 34964656,
      "step": 25800
    },
    {
      "epoch": 0.8259714486908648,
      "grad_norm": 37.25,
      "learning_rate": 1.3994468204384504e-05,
      "loss": 4.6366,
      "num_input_tokens_seen": 34971184,
      "step": 25805
    },
    {
      "epoch": 0.8261314896613533,
      "grad_norm": 29.375,
      "learning_rate": 1.398565404541622e-05,
      "loss": 3.5837,
      "num_input_tokens_seen": 34977824,
      "step": 25810
    },
    {
      "epoch": 0.8262915306318418,
      "grad_norm": 35.75,
      "learning_rate": 1.3976841584998513e-05,
      "loss": 5.2824,
      "num_input_tokens_seen": 34984336,
      "step": 25815
    },
    {
      "epoch": 0.8264515716023302,
      "grad_norm": 18.5,
      "learning_rate": 1.3968030824490352e-05,
      "loss": 3.6594,
      "num_input_tokens_seen": 34990944,
      "step": 25820
    },
    {
      "epoch": 0.8266116125728187,
      "grad_norm": 30.0,
      "learning_rate": 1.3959221765250469e-05,
      "loss": 4.8415,
      "num_input_tokens_seen": 34997600,
      "step": 25825
    },
    {
      "epoch": 0.8267716535433071,
      "grad_norm": 59.25,
      "learning_rate": 1.3950414408637343e-05,
      "loss": 4.2511,
      "num_input_tokens_seen": 35004464,
      "step": 25830
    },
    {
      "epoch": 0.8269316945137956,
      "grad_norm": 25.375,
      "learning_rate": 1.3941608756009166e-05,
      "loss": 4.8997,
      "num_input_tokens_seen": 35010736,
      "step": 25835
    },
    {
      "epoch": 0.8270917354842839,
      "grad_norm": 23.25,
      "learning_rate": 1.3932804808723898e-05,
      "loss": 4.0108,
      "num_input_tokens_seen": 35017232,
      "step": 25840
    },
    {
      "epoch": 0.8272517764547724,
      "grad_norm": 28.0,
      "learning_rate": 1.3924002568139194e-05,
      "loss": 4.8641,
      "num_input_tokens_seen": 35023472,
      "step": 25845
    },
    {
      "epoch": 0.8274118174252608,
      "grad_norm": 29.375,
      "learning_rate": 1.3915202035612485e-05,
      "loss": 4.4865,
      "num_input_tokens_seen": 35030096,
      "step": 25850
    },
    {
      "epoch": 0.8275718583957493,
      "grad_norm": 28.25,
      "learning_rate": 1.3906403212500935e-05,
      "loss": 4.852,
      "num_input_tokens_seen": 35036784,
      "step": 25855
    },
    {
      "epoch": 0.8277318993662378,
      "grad_norm": 24.75,
      "learning_rate": 1.3897606100161409e-05,
      "loss": 4.6443,
      "num_input_tokens_seen": 35043632,
      "step": 25860
    },
    {
      "epoch": 0.8278919403367262,
      "grad_norm": 37.0,
      "learning_rate": 1.388881069995055e-05,
      "loss": 4.5409,
      "num_input_tokens_seen": 35050544,
      "step": 25865
    },
    {
      "epoch": 0.8280519813072147,
      "grad_norm": 25.875,
      "learning_rate": 1.3880017013224708e-05,
      "loss": 4.4243,
      "num_input_tokens_seen": 35058208,
      "step": 25870
    },
    {
      "epoch": 0.8282120222777031,
      "grad_norm": 50.75,
      "learning_rate": 1.3871225041339984e-05,
      "loss": 3.7964,
      "num_input_tokens_seen": 35065216,
      "step": 25875
    },
    {
      "epoch": 0.8283720632481916,
      "grad_norm": 38.25,
      "learning_rate": 1.386243478565222e-05,
      "loss": 4.8153,
      "num_input_tokens_seen": 35071168,
      "step": 25880
    },
    {
      "epoch": 0.82853210421868,
      "grad_norm": 35.75,
      "learning_rate": 1.3853646247516966e-05,
      "loss": 4.5587,
      "num_input_tokens_seen": 35078224,
      "step": 25885
    },
    {
      "epoch": 0.8286921451891684,
      "grad_norm": 18.875,
      "learning_rate": 1.3844859428289545e-05,
      "loss": 4.4145,
      "num_input_tokens_seen": 35085248,
      "step": 25890
    },
    {
      "epoch": 0.8288521861596568,
      "grad_norm": 30.0,
      "learning_rate": 1.3836074329324984e-05,
      "loss": 4.974,
      "num_input_tokens_seen": 35091520,
      "step": 25895
    },
    {
      "epoch": 0.8290122271301453,
      "grad_norm": 18.625,
      "learning_rate": 1.3827290951978044e-05,
      "loss": 3.3098,
      "num_input_tokens_seen": 35098208,
      "step": 25900
    },
    {
      "epoch": 0.8291722681006337,
      "grad_norm": 20.75,
      "learning_rate": 1.381850929760326e-05,
      "loss": 4.7164,
      "num_input_tokens_seen": 35105360,
      "step": 25905
    },
    {
      "epoch": 0.8293323090711222,
      "grad_norm": 38.5,
      "learning_rate": 1.3809729367554842e-05,
      "loss": 3.7495,
      "num_input_tokens_seen": 35112496,
      "step": 25910
    },
    {
      "epoch": 0.8294923500416107,
      "grad_norm": 30.375,
      "learning_rate": 1.3800951163186784e-05,
      "loss": 4.3284,
      "num_input_tokens_seen": 35119248,
      "step": 25915
    },
    {
      "epoch": 0.8296523910120991,
      "grad_norm": 28.875,
      "learning_rate": 1.3792174685852801e-05,
      "loss": 3.3688,
      "num_input_tokens_seen": 35126000,
      "step": 25920
    },
    {
      "epoch": 0.8298124319825876,
      "grad_norm": 21.75,
      "learning_rate": 1.378339993690632e-05,
      "loss": 4.1122,
      "num_input_tokens_seen": 35132688,
      "step": 25925
    },
    {
      "epoch": 0.829972472953076,
      "grad_norm": 25.0,
      "learning_rate": 1.3774626917700523e-05,
      "loss": 4.8978,
      "num_input_tokens_seen": 35139488,
      "step": 25930
    },
    {
      "epoch": 0.8301325139235645,
      "grad_norm": 20.875,
      "learning_rate": 1.3765855629588334e-05,
      "loss": 4.0604,
      "num_input_tokens_seen": 35146016,
      "step": 25935
    },
    {
      "epoch": 0.8302925548940528,
      "grad_norm": 14.0625,
      "learning_rate": 1.3757086073922374e-05,
      "loss": 4.7497,
      "num_input_tokens_seen": 35152832,
      "step": 25940
    },
    {
      "epoch": 0.8304525958645413,
      "grad_norm": 24.75,
      "learning_rate": 1.3748318252055038e-05,
      "loss": 3.1413,
      "num_input_tokens_seen": 35160272,
      "step": 25945
    },
    {
      "epoch": 0.8306126368350297,
      "grad_norm": 18.0,
      "learning_rate": 1.3739552165338416e-05,
      "loss": 4.2992,
      "num_input_tokens_seen": 35166672,
      "step": 25950
    },
    {
      "epoch": 0.8307726778055182,
      "grad_norm": 42.5,
      "learning_rate": 1.3730787815124354e-05,
      "loss": 4.6545,
      "num_input_tokens_seen": 35173360,
      "step": 25955
    },
    {
      "epoch": 0.8309327187760066,
      "grad_norm": 24.625,
      "learning_rate": 1.3722025202764443e-05,
      "loss": 4.6266,
      "num_input_tokens_seen": 35180640,
      "step": 25960
    },
    {
      "epoch": 0.8310927597464951,
      "grad_norm": 41.5,
      "learning_rate": 1.371326432960997e-05,
      "loss": 6.1899,
      "num_input_tokens_seen": 35187248,
      "step": 25965
    },
    {
      "epoch": 0.8312528007169836,
      "grad_norm": 34.0,
      "learning_rate": 1.3704505197011969e-05,
      "loss": 5.0213,
      "num_input_tokens_seen": 35194192,
      "step": 25970
    },
    {
      "epoch": 0.831412841687472,
      "grad_norm": 22.125,
      "learning_rate": 1.3695747806321224e-05,
      "loss": 3.722,
      "num_input_tokens_seen": 35201008,
      "step": 25975
    },
    {
      "epoch": 0.8315728826579605,
      "grad_norm": 22.125,
      "learning_rate": 1.3686992158888212e-05,
      "loss": 4.0314,
      "num_input_tokens_seen": 35207840,
      "step": 25980
    },
    {
      "epoch": 0.8317329236284489,
      "grad_norm": 29.25,
      "learning_rate": 1.367823825606319e-05,
      "loss": 4.5599,
      "num_input_tokens_seen": 35214544,
      "step": 25985
    },
    {
      "epoch": 0.8318929645989374,
      "grad_norm": 21.0,
      "learning_rate": 1.36694860991961e-05,
      "loss": 4.1223,
      "num_input_tokens_seen": 35221344,
      "step": 25990
    },
    {
      "epoch": 0.8320530055694257,
      "grad_norm": 36.25,
      "learning_rate": 1.3660735689636636e-05,
      "loss": 3.5979,
      "num_input_tokens_seen": 35228000,
      "step": 25995
    },
    {
      "epoch": 0.8322130465399142,
      "grad_norm": 35.5,
      "learning_rate": 1.365198702873424e-05,
      "loss": 5.177,
      "num_input_tokens_seen": 35234256,
      "step": 26000
    },
    {
      "epoch": 0.8322130465399142,
      "eval_loss": 4.486413478851318,
      "eval_runtime": 490.0285,
      "eval_samples_per_second": 28.337,
      "eval_steps_per_second": 14.169,
      "num_input_tokens_seen": 35234256,
      "step": 26000
    },
    {
      "epoch": 0.8323730875104026,
      "grad_norm": 48.25,
      "learning_rate": 1.364324011783804e-05,
      "loss": 4.3454,
      "num_input_tokens_seen": 35240416,
      "step": 26005
    },
    {
      "epoch": 0.8325331284808911,
      "grad_norm": 29.125,
      "learning_rate": 1.3634494958296934e-05,
      "loss": 4.0761,
      "num_input_tokens_seen": 35246864,
      "step": 26010
    },
    {
      "epoch": 0.8326931694513795,
      "grad_norm": 45.25,
      "learning_rate": 1.3625751551459542e-05,
      "loss": 4.4957,
      "num_input_tokens_seen": 35253616,
      "step": 26015
    },
    {
      "epoch": 0.832853210421868,
      "grad_norm": 32.5,
      "learning_rate": 1.3617009898674188e-05,
      "loss": 4.772,
      "num_input_tokens_seen": 35260304,
      "step": 26020
    },
    {
      "epoch": 0.8330132513923565,
      "grad_norm": 29.0,
      "learning_rate": 1.3608270001288967e-05,
      "loss": 4.0718,
      "num_input_tokens_seen": 35267280,
      "step": 26025
    },
    {
      "epoch": 0.8331732923628449,
      "grad_norm": 26.0,
      "learning_rate": 1.359953186065166e-05,
      "loss": 3.9457,
      "num_input_tokens_seen": 35274288,
      "step": 26030
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 27.375,
      "learning_rate": 1.3590795478109814e-05,
      "loss": 4.8029,
      "num_input_tokens_seen": 35280832,
      "step": 26035
    },
    {
      "epoch": 0.8334933743038218,
      "grad_norm": 42.5,
      "learning_rate": 1.3582060855010675e-05,
      "loss": 4.1258,
      "num_input_tokens_seen": 35287808,
      "step": 26040
    },
    {
      "epoch": 0.8336534152743102,
      "grad_norm": 27.875,
      "learning_rate": 1.3573327992701245e-05,
      "loss": 4.0507,
      "num_input_tokens_seen": 35294160,
      "step": 26045
    },
    {
      "epoch": 0.8338134562447986,
      "grad_norm": 29.125,
      "learning_rate": 1.356459689252823e-05,
      "loss": 4.3013,
      "num_input_tokens_seen": 35300944,
      "step": 26050
    },
    {
      "epoch": 0.8339734972152871,
      "grad_norm": 20.25,
      "learning_rate": 1.3555867555838087e-05,
      "loss": 4.4195,
      "num_input_tokens_seen": 35307680,
      "step": 26055
    },
    {
      "epoch": 0.8341335381857755,
      "grad_norm": 14.0,
      "learning_rate": 1.3547139983976975e-05,
      "loss": 4.8362,
      "num_input_tokens_seen": 35314704,
      "step": 26060
    },
    {
      "epoch": 0.834293579156264,
      "grad_norm": 25.625,
      "learning_rate": 1.3538414178290815e-05,
      "loss": 3.5036,
      "num_input_tokens_seen": 35321744,
      "step": 26065
    },
    {
      "epoch": 0.8344536201267524,
      "grad_norm": 30.125,
      "learning_rate": 1.3529690140125209e-05,
      "loss": 4.1427,
      "num_input_tokens_seen": 35328384,
      "step": 26070
    },
    {
      "epoch": 0.8346136610972409,
      "grad_norm": 23.75,
      "learning_rate": 1.352096787082553e-05,
      "loss": 4.5517,
      "num_input_tokens_seen": 35335424,
      "step": 26075
    },
    {
      "epoch": 0.8347737020677294,
      "grad_norm": 14.0625,
      "learning_rate": 1.3512247371736871e-05,
      "loss": 4.722,
      "num_input_tokens_seen": 35342464,
      "step": 26080
    },
    {
      "epoch": 0.8349337430382178,
      "grad_norm": 29.125,
      "learning_rate": 1.3503528644204022e-05,
      "loss": 3.8862,
      "num_input_tokens_seen": 35348976,
      "step": 26085
    },
    {
      "epoch": 0.8350937840087063,
      "grad_norm": 31.25,
      "learning_rate": 1.349481168957153e-05,
      "loss": 3.929,
      "num_input_tokens_seen": 35355952,
      "step": 26090
    },
    {
      "epoch": 0.8352538249791946,
      "grad_norm": 29.25,
      "learning_rate": 1.3486096509183665e-05,
      "loss": 5.2318,
      "num_input_tokens_seen": 35362320,
      "step": 26095
    },
    {
      "epoch": 0.8354138659496831,
      "grad_norm": 27.625,
      "learning_rate": 1.3477383104384406e-05,
      "loss": 4.1087,
      "num_input_tokens_seen": 35369568,
      "step": 26100
    },
    {
      "epoch": 0.8355739069201715,
      "grad_norm": 21.25,
      "learning_rate": 1.3468671476517481e-05,
      "loss": 3.8077,
      "num_input_tokens_seen": 35376576,
      "step": 26105
    },
    {
      "epoch": 0.83573394789066,
      "grad_norm": 29.75,
      "learning_rate": 1.3459961626926326e-05,
      "loss": 4.0275,
      "num_input_tokens_seen": 35383248,
      "step": 26110
    },
    {
      "epoch": 0.8358939888611484,
      "grad_norm": 19.625,
      "learning_rate": 1.3451253556954101e-05,
      "loss": 3.9254,
      "num_input_tokens_seen": 35390640,
      "step": 26115
    },
    {
      "epoch": 0.8360540298316369,
      "grad_norm": 16.0,
      "learning_rate": 1.3442547267943717e-05,
      "loss": 3.7662,
      "num_input_tokens_seen": 35397376,
      "step": 26120
    },
    {
      "epoch": 0.8362140708021254,
      "grad_norm": 25.375,
      "learning_rate": 1.3433842761237774e-05,
      "loss": 4.0187,
      "num_input_tokens_seen": 35404256,
      "step": 26125
    },
    {
      "epoch": 0.8363741117726138,
      "grad_norm": 42.0,
      "learning_rate": 1.3425140038178639e-05,
      "loss": 4.087,
      "num_input_tokens_seen": 35410800,
      "step": 26130
    },
    {
      "epoch": 0.8365341527431023,
      "grad_norm": 44.5,
      "learning_rate": 1.3416439100108358e-05,
      "loss": 4.4892,
      "num_input_tokens_seen": 35417328,
      "step": 26135
    },
    {
      "epoch": 0.8366941937135907,
      "grad_norm": 19.375,
      "learning_rate": 1.3407739948368734e-05,
      "loss": 4.2513,
      "num_input_tokens_seen": 35424144,
      "step": 26140
    },
    {
      "epoch": 0.8368542346840792,
      "grad_norm": 30.75,
      "learning_rate": 1.3399042584301298e-05,
      "loss": 3.7873,
      "num_input_tokens_seen": 35431008,
      "step": 26145
    },
    {
      "epoch": 0.8370142756545675,
      "grad_norm": 38.0,
      "learning_rate": 1.3390347009247272e-05,
      "loss": 4.3018,
      "num_input_tokens_seen": 35437456,
      "step": 26150
    },
    {
      "epoch": 0.837174316625056,
      "grad_norm": 23.625,
      "learning_rate": 1.3381653224547635e-05,
      "loss": 3.6626,
      "num_input_tokens_seen": 35445104,
      "step": 26155
    },
    {
      "epoch": 0.8373343575955444,
      "grad_norm": 35.75,
      "learning_rate": 1.3372961231543086e-05,
      "loss": 3.7459,
      "num_input_tokens_seen": 35451472,
      "step": 26160
    },
    {
      "epoch": 0.8374943985660329,
      "grad_norm": 20.0,
      "learning_rate": 1.3364271031574016e-05,
      "loss": 4.27,
      "num_input_tokens_seen": 35458176,
      "step": 26165
    },
    {
      "epoch": 0.8376544395365213,
      "grad_norm": 25.375,
      "learning_rate": 1.335558262598059e-05,
      "loss": 4.8103,
      "num_input_tokens_seen": 35464896,
      "step": 26170
    },
    {
      "epoch": 0.8378144805070098,
      "grad_norm": 34.0,
      "learning_rate": 1.3346896016102645e-05,
      "loss": 4.1898,
      "num_input_tokens_seen": 35471712,
      "step": 26175
    },
    {
      "epoch": 0.8379745214774983,
      "grad_norm": 13.8125,
      "learning_rate": 1.3338211203279788e-05,
      "loss": 4.3994,
      "num_input_tokens_seen": 35478096,
      "step": 26180
    },
    {
      "epoch": 0.8381345624479867,
      "grad_norm": 41.5,
      "learning_rate": 1.3329528188851303e-05,
      "loss": 3.9289,
      "num_input_tokens_seen": 35484832,
      "step": 26185
    },
    {
      "epoch": 0.8382946034184752,
      "grad_norm": 29.0,
      "learning_rate": 1.3320846974156242e-05,
      "loss": 4.6659,
      "num_input_tokens_seen": 35491344,
      "step": 26190
    },
    {
      "epoch": 0.8384546443889636,
      "grad_norm": 36.25,
      "learning_rate": 1.3312167560533337e-05,
      "loss": 4.852,
      "num_input_tokens_seen": 35498048,
      "step": 26195
    },
    {
      "epoch": 0.838614685359452,
      "grad_norm": 47.75,
      "learning_rate": 1.3303489949321082e-05,
      "loss": 4.6042,
      "num_input_tokens_seen": 35504864,
      "step": 26200
    },
    {
      "epoch": 0.838614685359452,
      "eval_loss": 4.471009254455566,
      "eval_runtime": 489.3195,
      "eval_samples_per_second": 28.378,
      "eval_steps_per_second": 14.189,
      "num_input_tokens_seen": 35504864,
      "step": 26200
    },
    {
      "epoch": 0.8387747263299404,
      "grad_norm": 26.625,
      "learning_rate": 1.3294814141857653e-05,
      "loss": 4.7241,
      "num_input_tokens_seen": 35511760,
      "step": 26205
    },
    {
      "epoch": 0.8389347673004289,
      "grad_norm": 28.875,
      "learning_rate": 1.3286140139480992e-05,
      "loss": 4.0941,
      "num_input_tokens_seen": 35518736,
      "step": 26210
    },
    {
      "epoch": 0.8390948082709173,
      "grad_norm": 38.5,
      "learning_rate": 1.3277467943528719e-05,
      "loss": 4.566,
      "num_input_tokens_seen": 35525360,
      "step": 26215
    },
    {
      "epoch": 0.8392548492414058,
      "grad_norm": 36.25,
      "learning_rate": 1.3268797555338203e-05,
      "loss": 4.864,
      "num_input_tokens_seen": 35532368,
      "step": 26220
    },
    {
      "epoch": 0.8394148902118942,
      "grad_norm": 20.125,
      "learning_rate": 1.3260128976246533e-05,
      "loss": 2.8137,
      "num_input_tokens_seen": 35538976,
      "step": 26225
    },
    {
      "epoch": 0.8395749311823827,
      "grad_norm": 56.25,
      "learning_rate": 1.32514622075905e-05,
      "loss": 4.5056,
      "num_input_tokens_seen": 35546080,
      "step": 26230
    },
    {
      "epoch": 0.8397349721528712,
      "grad_norm": 36.0,
      "learning_rate": 1.3242797250706638e-05,
      "loss": 4.2375,
      "num_input_tokens_seen": 35552832,
      "step": 26235
    },
    {
      "epoch": 0.8398950131233596,
      "grad_norm": 36.0,
      "learning_rate": 1.3234134106931195e-05,
      "loss": 4.5986,
      "num_input_tokens_seen": 35559392,
      "step": 26240
    },
    {
      "epoch": 0.8400550540938481,
      "grad_norm": 18.5,
      "learning_rate": 1.322547277760013e-05,
      "loss": 4.0647,
      "num_input_tokens_seen": 35566512,
      "step": 26245
    },
    {
      "epoch": 0.8402150950643364,
      "grad_norm": 23.625,
      "learning_rate": 1.3216813264049132e-05,
      "loss": 4.4076,
      "num_input_tokens_seen": 35573152,
      "step": 26250
    },
    {
      "epoch": 0.8403751360348249,
      "grad_norm": 56.0,
      "learning_rate": 1.32081555676136e-05,
      "loss": 5.6113,
      "num_input_tokens_seen": 35579952,
      "step": 26255
    },
    {
      "epoch": 0.8405351770053133,
      "grad_norm": 36.75,
      "learning_rate": 1.3199499689628674e-05,
      "loss": 4.6034,
      "num_input_tokens_seen": 35586752,
      "step": 26260
    },
    {
      "epoch": 0.8406952179758018,
      "grad_norm": 40.0,
      "learning_rate": 1.3190845631429192e-05,
      "loss": 4.1097,
      "num_input_tokens_seen": 35594064,
      "step": 26265
    },
    {
      "epoch": 0.8408552589462902,
      "grad_norm": 32.75,
      "learning_rate": 1.3182193394349704e-05,
      "loss": 4.4237,
      "num_input_tokens_seen": 35600624,
      "step": 26270
    },
    {
      "epoch": 0.8410152999167787,
      "grad_norm": 22.75,
      "learning_rate": 1.3173542979724507e-05,
      "loss": 3.863,
      "num_input_tokens_seen": 35607472,
      "step": 26275
    },
    {
      "epoch": 0.8411753408872671,
      "grad_norm": 43.25,
      "learning_rate": 1.3164894388887617e-05,
      "loss": 4.083,
      "num_input_tokens_seen": 35614256,
      "step": 26280
    },
    {
      "epoch": 0.8413353818577556,
      "grad_norm": 27.0,
      "learning_rate": 1.3156247623172727e-05,
      "loss": 4.6549,
      "num_input_tokens_seen": 35620592,
      "step": 26285
    },
    {
      "epoch": 0.8414954228282441,
      "grad_norm": 25.5,
      "learning_rate": 1.3147602683913302e-05,
      "loss": 4.0291,
      "num_input_tokens_seen": 35627088,
      "step": 26290
    },
    {
      "epoch": 0.8416554637987325,
      "grad_norm": 31.125,
      "learning_rate": 1.3138959572442481e-05,
      "loss": 4.5222,
      "num_input_tokens_seen": 35633680,
      "step": 26295
    },
    {
      "epoch": 0.841815504769221,
      "grad_norm": 35.25,
      "learning_rate": 1.3130318290093146e-05,
      "loss": 3.6832,
      "num_input_tokens_seen": 35640464,
      "step": 26300
    },
    {
      "epoch": 0.8419755457397093,
      "grad_norm": 37.0,
      "learning_rate": 1.3121678838197909e-05,
      "loss": 5.6066,
      "num_input_tokens_seen": 35646976,
      "step": 26305
    },
    {
      "epoch": 0.8421355867101978,
      "grad_norm": 16.0,
      "learning_rate": 1.3113041218089056e-05,
      "loss": 3.8262,
      "num_input_tokens_seen": 35653984,
      "step": 26310
    },
    {
      "epoch": 0.8422956276806862,
      "grad_norm": 36.25,
      "learning_rate": 1.3104405431098626e-05,
      "loss": 4.0046,
      "num_input_tokens_seen": 35660688,
      "step": 26315
    },
    {
      "epoch": 0.8424556686511747,
      "grad_norm": 35.25,
      "learning_rate": 1.3095771478558377e-05,
      "loss": 4.621,
      "num_input_tokens_seen": 35667280,
      "step": 26320
    },
    {
      "epoch": 0.8426157096216631,
      "grad_norm": 21.0,
      "learning_rate": 1.3087139361799766e-05,
      "loss": 4.2138,
      "num_input_tokens_seen": 35674064,
      "step": 26325
    },
    {
      "epoch": 0.8427757505921516,
      "grad_norm": 19.875,
      "learning_rate": 1.3078509082153964e-05,
      "loss": 3.8548,
      "num_input_tokens_seen": 35681040,
      "step": 26330
    },
    {
      "epoch": 0.8429357915626401,
      "grad_norm": 58.5,
      "learning_rate": 1.3069880640951885e-05,
      "loss": 4.1815,
      "num_input_tokens_seen": 35687712,
      "step": 26335
    },
    {
      "epoch": 0.8430958325331285,
      "grad_norm": 35.75,
      "learning_rate": 1.3061254039524123e-05,
      "loss": 3.644,
      "num_input_tokens_seen": 35694208,
      "step": 26340
    },
    {
      "epoch": 0.843255873503617,
      "grad_norm": 21.25,
      "learning_rate": 1.3052629279201028e-05,
      "loss": 3.8102,
      "num_input_tokens_seen": 35701184,
      "step": 26345
    },
    {
      "epoch": 0.8434159144741054,
      "grad_norm": 36.0,
      "learning_rate": 1.3044006361312633e-05,
      "loss": 4.3908,
      "num_input_tokens_seen": 35708000,
      "step": 26350
    },
    {
      "epoch": 0.8435759554445939,
      "grad_norm": 44.0,
      "learning_rate": 1.30353852871887e-05,
      "loss": 4.2132,
      "num_input_tokens_seen": 35714816,
      "step": 26355
    },
    {
      "epoch": 0.8437359964150822,
      "grad_norm": 29.625,
      "learning_rate": 1.302676605815873e-05,
      "loss": 4.4274,
      "num_input_tokens_seen": 35721680,
      "step": 26360
    },
    {
      "epoch": 0.8438960373855707,
      "grad_norm": 27.75,
      "learning_rate": 1.3018148675551884e-05,
      "loss": 5.1951,
      "num_input_tokens_seen": 35729344,
      "step": 26365
    },
    {
      "epoch": 0.8440560783560591,
      "grad_norm": 15.75,
      "learning_rate": 1.3009533140697094e-05,
      "loss": 4.3707,
      "num_input_tokens_seen": 35736256,
      "step": 26370
    },
    {
      "epoch": 0.8442161193265476,
      "grad_norm": 37.0,
      "learning_rate": 1.3000919454922966e-05,
      "loss": 3.9898,
      "num_input_tokens_seen": 35743040,
      "step": 26375
    },
    {
      "epoch": 0.844376160297036,
      "grad_norm": 22.75,
      "learning_rate": 1.299230761955785e-05,
      "loss": 4.4421,
      "num_input_tokens_seen": 35750320,
      "step": 26380
    },
    {
      "epoch": 0.8445362012675245,
      "grad_norm": 32.25,
      "learning_rate": 1.2983697635929807e-05,
      "loss": 4.8762,
      "num_input_tokens_seen": 35757200,
      "step": 26385
    },
    {
      "epoch": 0.844696242238013,
      "grad_norm": 41.0,
      "learning_rate": 1.2975089505366584e-05,
      "loss": 4.2528,
      "num_input_tokens_seen": 35764112,
      "step": 26390
    },
    {
      "epoch": 0.8448562832085014,
      "grad_norm": 27.5,
      "learning_rate": 1.2966483229195683e-05,
      "loss": 4.627,
      "num_input_tokens_seen": 35770832,
      "step": 26395
    },
    {
      "epoch": 0.8450163241789899,
      "grad_norm": 38.75,
      "learning_rate": 1.2957878808744283e-05,
      "loss": 4.626,
      "num_input_tokens_seen": 35777296,
      "step": 26400
    },
    {
      "epoch": 0.8450163241789899,
      "eval_loss": 4.507416248321533,
      "eval_runtime": 489.3016,
      "eval_samples_per_second": 28.379,
      "eval_steps_per_second": 14.19,
      "num_input_tokens_seen": 35777296,
      "step": 26400
    },
    {
      "epoch": 0.8451763651494782,
      "grad_norm": 26.875,
      "learning_rate": 1.294927624533931e-05,
      "loss": 4.2858,
      "num_input_tokens_seen": 35783840,
      "step": 26405
    },
    {
      "epoch": 0.8453364061199667,
      "grad_norm": 31.625,
      "learning_rate": 1.2940675540307378e-05,
      "loss": 3.8384,
      "num_input_tokens_seen": 35790400,
      "step": 26410
    },
    {
      "epoch": 0.8454964470904551,
      "grad_norm": 20.625,
      "learning_rate": 1.2932076694974814e-05,
      "loss": 4.4151,
      "num_input_tokens_seen": 35797456,
      "step": 26415
    },
    {
      "epoch": 0.8456564880609436,
      "grad_norm": 42.0,
      "learning_rate": 1.2923479710667682e-05,
      "loss": 4.5352,
      "num_input_tokens_seen": 35804160,
      "step": 26420
    },
    {
      "epoch": 0.845816529031432,
      "grad_norm": 27.625,
      "learning_rate": 1.2914884588711751e-05,
      "loss": 3.5595,
      "num_input_tokens_seen": 35810864,
      "step": 26425
    },
    {
      "epoch": 0.8459765700019205,
      "grad_norm": 43.75,
      "learning_rate": 1.2906291330432475e-05,
      "loss": 5.575,
      "num_input_tokens_seen": 35817088,
      "step": 26430
    },
    {
      "epoch": 0.8461366109724089,
      "grad_norm": 28.625,
      "learning_rate": 1.2897699937155055e-05,
      "loss": 4.349,
      "num_input_tokens_seen": 35823616,
      "step": 26435
    },
    {
      "epoch": 0.8462966519428974,
      "grad_norm": 37.0,
      "learning_rate": 1.2889110410204403e-05,
      "loss": 3.772,
      "num_input_tokens_seen": 35830384,
      "step": 26440
    },
    {
      "epoch": 0.8464566929133859,
      "grad_norm": 44.75,
      "learning_rate": 1.2880522750905111e-05,
      "loss": 4.8548,
      "num_input_tokens_seen": 35837216,
      "step": 26445
    },
    {
      "epoch": 0.8466167338838743,
      "grad_norm": 30.625,
      "learning_rate": 1.2871936960581523e-05,
      "loss": 4.2238,
      "num_input_tokens_seen": 35843760,
      "step": 26450
    },
    {
      "epoch": 0.8467767748543628,
      "grad_norm": 37.75,
      "learning_rate": 1.2863353040557658e-05,
      "loss": 4.0549,
      "num_input_tokens_seen": 35850080,
      "step": 26455
    },
    {
      "epoch": 0.8469368158248511,
      "grad_norm": 34.25,
      "learning_rate": 1.2854770992157273e-05,
      "loss": 5.346,
      "num_input_tokens_seen": 35856720,
      "step": 26460
    },
    {
      "epoch": 0.8470968567953396,
      "grad_norm": 20.5,
      "learning_rate": 1.2846190816703835e-05,
      "loss": 4.9682,
      "num_input_tokens_seen": 35863056,
      "step": 26465
    },
    {
      "epoch": 0.847256897765828,
      "grad_norm": 17.75,
      "learning_rate": 1.2837612515520498e-05,
      "loss": 3.3477,
      "num_input_tokens_seen": 35869632,
      "step": 26470
    },
    {
      "epoch": 0.8474169387363165,
      "grad_norm": 47.5,
      "learning_rate": 1.2829036089930163e-05,
      "loss": 4.8573,
      "num_input_tokens_seen": 35877024,
      "step": 26475
    },
    {
      "epoch": 0.8475769797068049,
      "grad_norm": 29.125,
      "learning_rate": 1.2820461541255412e-05,
      "loss": 4.8629,
      "num_input_tokens_seen": 35883776,
      "step": 26480
    },
    {
      "epoch": 0.8477370206772934,
      "grad_norm": 34.25,
      "learning_rate": 1.2811888870818543e-05,
      "loss": 4.5888,
      "num_input_tokens_seen": 35890496,
      "step": 26485
    },
    {
      "epoch": 0.8478970616477818,
      "grad_norm": 36.0,
      "learning_rate": 1.2803318079941581e-05,
      "loss": 3.7567,
      "num_input_tokens_seen": 35897136,
      "step": 26490
    },
    {
      "epoch": 0.8480571026182703,
      "grad_norm": 24.5,
      "learning_rate": 1.2794749169946235e-05,
      "loss": 4.696,
      "num_input_tokens_seen": 35903776,
      "step": 26495
    },
    {
      "epoch": 0.8482171435887588,
      "grad_norm": 37.5,
      "learning_rate": 1.2786182142153952e-05,
      "loss": 4.4244,
      "num_input_tokens_seen": 35910272,
      "step": 26500
    },
    {
      "epoch": 0.8483771845592472,
      "grad_norm": 29.375,
      "learning_rate": 1.2777616997885878e-05,
      "loss": 4.2308,
      "num_input_tokens_seen": 35917104,
      "step": 26505
    },
    {
      "epoch": 0.8485372255297357,
      "grad_norm": 22.625,
      "learning_rate": 1.2769053738462847e-05,
      "loss": 4.9516,
      "num_input_tokens_seen": 35924160,
      "step": 26510
    },
    {
      "epoch": 0.848697266500224,
      "grad_norm": 27.875,
      "learning_rate": 1.2760492365205434e-05,
      "loss": 3.9563,
      "num_input_tokens_seen": 35930896,
      "step": 26515
    },
    {
      "epoch": 0.8488573074707125,
      "grad_norm": 29.25,
      "learning_rate": 1.2751932879433919e-05,
      "loss": 4.5309,
      "num_input_tokens_seen": 35937584,
      "step": 26520
    },
    {
      "epoch": 0.8490173484412009,
      "grad_norm": 29.375,
      "learning_rate": 1.2743375282468267e-05,
      "loss": 3.4483,
      "num_input_tokens_seen": 35944384,
      "step": 26525
    },
    {
      "epoch": 0.8491773894116894,
      "grad_norm": 47.25,
      "learning_rate": 1.2734819575628182e-05,
      "loss": 4.5533,
      "num_input_tokens_seen": 35950944,
      "step": 26530
    },
    {
      "epoch": 0.8493374303821778,
      "grad_norm": 43.5,
      "learning_rate": 1.2726265760233039e-05,
      "loss": 4.8978,
      "num_input_tokens_seen": 35957616,
      "step": 26535
    },
    {
      "epoch": 0.8494974713526663,
      "grad_norm": 41.25,
      "learning_rate": 1.271771383760197e-05,
      "loss": 3.8082,
      "num_input_tokens_seen": 35964464,
      "step": 26540
    },
    {
      "epoch": 0.8496575123231547,
      "grad_norm": 22.625,
      "learning_rate": 1.2709163809053764e-05,
      "loss": 4.4731,
      "num_input_tokens_seen": 35971216,
      "step": 26545
    },
    {
      "epoch": 0.8498175532936432,
      "grad_norm": 26.5,
      "learning_rate": 1.2700615675906963e-05,
      "loss": 3.823,
      "num_input_tokens_seen": 35978272,
      "step": 26550
    },
    {
      "epoch": 0.8499775942641317,
      "grad_norm": 25.25,
      "learning_rate": 1.269206943947978e-05,
      "loss": 3.9131,
      "num_input_tokens_seen": 35984992,
      "step": 26555
    },
    {
      "epoch": 0.85013763523462,
      "grad_norm": 32.0,
      "learning_rate": 1.2683525101090177e-05,
      "loss": 3.767,
      "num_input_tokens_seen": 35991520,
      "step": 26560
    },
    {
      "epoch": 0.8502976762051085,
      "grad_norm": 52.0,
      "learning_rate": 1.2674982662055765e-05,
      "loss": 4.4106,
      "num_input_tokens_seen": 35998560,
      "step": 26565
    },
    {
      "epoch": 0.8504577171755969,
      "grad_norm": 32.75,
      "learning_rate": 1.2666442123693922e-05,
      "loss": 3.6752,
      "num_input_tokens_seen": 36005200,
      "step": 26570
    },
    {
      "epoch": 0.8506177581460854,
      "grad_norm": 33.5,
      "learning_rate": 1.265790348732169e-05,
      "loss": 3.8784,
      "num_input_tokens_seen": 36012112,
      "step": 26575
    },
    {
      "epoch": 0.8507777991165738,
      "grad_norm": 47.5,
      "learning_rate": 1.264936675425584e-05,
      "loss": 4.2903,
      "num_input_tokens_seen": 36018512,
      "step": 26580
    },
    {
      "epoch": 0.8509378400870623,
      "grad_norm": 50.25,
      "learning_rate": 1.2640831925812852e-05,
      "loss": 5.0387,
      "num_input_tokens_seen": 36025120,
      "step": 26585
    },
    {
      "epoch": 0.8510978810575507,
      "grad_norm": 20.375,
      "learning_rate": 1.263229900330889e-05,
      "loss": 4.4429,
      "num_input_tokens_seen": 36031936,
      "step": 26590
    },
    {
      "epoch": 0.8512579220280392,
      "grad_norm": 40.5,
      "learning_rate": 1.2623767988059843e-05,
      "loss": 4.5109,
      "num_input_tokens_seen": 36038560,
      "step": 26595
    },
    {
      "epoch": 0.8514179629985277,
      "grad_norm": 32.5,
      "learning_rate": 1.2615238881381309e-05,
      "loss": 4.4373,
      "num_input_tokens_seen": 36045376,
      "step": 26600
    },
    {
      "epoch": 0.8514179629985277,
      "eval_loss": 4.470012187957764,
      "eval_runtime": 489.8012,
      "eval_samples_per_second": 28.35,
      "eval_steps_per_second": 14.175,
      "num_input_tokens_seen": 36045376,
      "step": 26600
    },
    {
      "epoch": 0.8515780039690161,
      "grad_norm": 17.375,
      "learning_rate": 1.2606711684588568e-05,
      "loss": 4.7135,
      "num_input_tokens_seen": 36052272,
      "step": 26605
    },
    {
      "epoch": 0.8517380449395046,
      "grad_norm": 28.5,
      "learning_rate": 1.2598186398996636e-05,
      "loss": 3.579,
      "num_input_tokens_seen": 36058848,
      "step": 26610
    },
    {
      "epoch": 0.8518980859099929,
      "grad_norm": 24.625,
      "learning_rate": 1.2589663025920207e-05,
      "loss": 5.1972,
      "num_input_tokens_seen": 36065376,
      "step": 26615
    },
    {
      "epoch": 0.8520581268804814,
      "grad_norm": 22.25,
      "learning_rate": 1.2581141566673705e-05,
      "loss": 4.7821,
      "num_input_tokens_seen": 36072048,
      "step": 26620
    },
    {
      "epoch": 0.8522181678509698,
      "grad_norm": 14.8125,
      "learning_rate": 1.257262202257124e-05,
      "loss": 4.0455,
      "num_input_tokens_seen": 36078816,
      "step": 26625
    },
    {
      "epoch": 0.8523782088214583,
      "grad_norm": 34.25,
      "learning_rate": 1.2564104394926618e-05,
      "loss": 5.8142,
      "num_input_tokens_seen": 36085968,
      "step": 26630
    },
    {
      "epoch": 0.8525382497919467,
      "grad_norm": 35.25,
      "learning_rate": 1.2555588685053383e-05,
      "loss": 4.8623,
      "num_input_tokens_seen": 36092640,
      "step": 26635
    },
    {
      "epoch": 0.8526982907624352,
      "grad_norm": 24.375,
      "learning_rate": 1.2547074894264762e-05,
      "loss": 4.2731,
      "num_input_tokens_seen": 36099920,
      "step": 26640
    },
    {
      "epoch": 0.8528583317329236,
      "grad_norm": 49.25,
      "learning_rate": 1.2538563023873679e-05,
      "loss": 4.679,
      "num_input_tokens_seen": 36106624,
      "step": 26645
    },
    {
      "epoch": 0.8530183727034121,
      "grad_norm": 33.75,
      "learning_rate": 1.2530053075192789e-05,
      "loss": 4.225,
      "num_input_tokens_seen": 36113392,
      "step": 26650
    },
    {
      "epoch": 0.8531784136739006,
      "grad_norm": 33.0,
      "learning_rate": 1.252154504953441e-05,
      "loss": 5.4186,
      "num_input_tokens_seen": 36120176,
      "step": 26655
    },
    {
      "epoch": 0.853338454644389,
      "grad_norm": 19.125,
      "learning_rate": 1.25130389482106e-05,
      "loss": 4.5813,
      "num_input_tokens_seen": 36127008,
      "step": 26660
    },
    {
      "epoch": 0.8534984956148775,
      "grad_norm": 22.875,
      "learning_rate": 1.2504534772533116e-05,
      "loss": 4.0258,
      "num_input_tokens_seen": 36133712,
      "step": 26665
    },
    {
      "epoch": 0.8536585365853658,
      "grad_norm": 40.0,
      "learning_rate": 1.2496032523813387e-05,
      "loss": 4.6133,
      "num_input_tokens_seen": 36140352,
      "step": 26670
    },
    {
      "epoch": 0.8538185775558543,
      "grad_norm": 28.75,
      "learning_rate": 1.2487532203362576e-05,
      "loss": 4.7026,
      "num_input_tokens_seen": 36146992,
      "step": 26675
    },
    {
      "epoch": 0.8539786185263427,
      "grad_norm": 34.75,
      "learning_rate": 1.247903381249155e-05,
      "loss": 4.4546,
      "num_input_tokens_seen": 36153936,
      "step": 26680
    },
    {
      "epoch": 0.8541386594968312,
      "grad_norm": 37.25,
      "learning_rate": 1.2470537352510853e-05,
      "loss": 4.3772,
      "num_input_tokens_seen": 36160992,
      "step": 26685
    },
    {
      "epoch": 0.8542987004673196,
      "grad_norm": 26.0,
      "learning_rate": 1.2462042824730758e-05,
      "loss": 3.7548,
      "num_input_tokens_seen": 36167712,
      "step": 26690
    },
    {
      "epoch": 0.8544587414378081,
      "grad_norm": 26.625,
      "learning_rate": 1.245355023046122e-05,
      "loss": 4.4608,
      "num_input_tokens_seen": 36174528,
      "step": 26695
    },
    {
      "epoch": 0.8546187824082965,
      "grad_norm": 20.5,
      "learning_rate": 1.2445059571011896e-05,
      "loss": 3.454,
      "num_input_tokens_seen": 36181264,
      "step": 26700
    },
    {
      "epoch": 0.854778823378785,
      "grad_norm": 37.25,
      "learning_rate": 1.2436570847692173e-05,
      "loss": 4.2584,
      "num_input_tokens_seen": 36187872,
      "step": 26705
    },
    {
      "epoch": 0.8549388643492735,
      "grad_norm": 37.75,
      "learning_rate": 1.2428084061811096e-05,
      "loss": 4.8078,
      "num_input_tokens_seen": 36194352,
      "step": 26710
    },
    {
      "epoch": 0.8550989053197618,
      "grad_norm": 17.375,
      "learning_rate": 1.2419599214677447e-05,
      "loss": 4.1715,
      "num_input_tokens_seen": 36201168,
      "step": 26715
    },
    {
      "epoch": 0.8552589462902503,
      "grad_norm": 39.25,
      "learning_rate": 1.2411116307599702e-05,
      "loss": 5.4924,
      "num_input_tokens_seen": 36207712,
      "step": 26720
    },
    {
      "epoch": 0.8554189872607387,
      "grad_norm": 34.75,
      "learning_rate": 1.2402635341886016e-05,
      "loss": 3.8876,
      "num_input_tokens_seen": 36214688,
      "step": 26725
    },
    {
      "epoch": 0.8555790282312272,
      "grad_norm": 52.5,
      "learning_rate": 1.2394156318844278e-05,
      "loss": 4.7261,
      "num_input_tokens_seen": 36221424,
      "step": 26730
    },
    {
      "epoch": 0.8557390692017156,
      "grad_norm": 32.25,
      "learning_rate": 1.2385679239782039e-05,
      "loss": 4.5185,
      "num_input_tokens_seen": 36227728,
      "step": 26735
    },
    {
      "epoch": 0.8558991101722041,
      "grad_norm": 31.25,
      "learning_rate": 1.2377204106006585e-05,
      "loss": 4.6533,
      "num_input_tokens_seen": 36234624,
      "step": 26740
    },
    {
      "epoch": 0.8560591511426925,
      "grad_norm": 22.5,
      "learning_rate": 1.2368730918824891e-05,
      "loss": 4.607,
      "num_input_tokens_seen": 36241360,
      "step": 26745
    },
    {
      "epoch": 0.856219192113181,
      "grad_norm": 38.5,
      "learning_rate": 1.236025967954362e-05,
      "loss": 4.9349,
      "num_input_tokens_seen": 36248608,
      "step": 26750
    },
    {
      "epoch": 0.8563792330836694,
      "grad_norm": 39.25,
      "learning_rate": 1.2351790389469153e-05,
      "loss": 4.2721,
      "num_input_tokens_seen": 36255056,
      "step": 26755
    },
    {
      "epoch": 0.8565392740541579,
      "grad_norm": 16.5,
      "learning_rate": 1.234332304990755e-05,
      "loss": 5.0184,
      "num_input_tokens_seen": 36261328,
      "step": 26760
    },
    {
      "epoch": 0.8566993150246464,
      "grad_norm": 20.625,
      "learning_rate": 1.2334857662164593e-05,
      "loss": 4.6178,
      "num_input_tokens_seen": 36268224,
      "step": 26765
    },
    {
      "epoch": 0.8568593559951347,
      "grad_norm": 39.75,
      "learning_rate": 1.2326394227545743e-05,
      "loss": 5.0391,
      "num_input_tokens_seen": 36274880,
      "step": 26770
    },
    {
      "epoch": 0.8570193969656232,
      "grad_norm": 37.5,
      "learning_rate": 1.2317932747356162e-05,
      "loss": 3.9127,
      "num_input_tokens_seen": 36281824,
      "step": 26775
    },
    {
      "epoch": 0.8571794379361116,
      "grad_norm": 32.25,
      "learning_rate": 1.2309473222900726e-05,
      "loss": 4.2777,
      "num_input_tokens_seen": 36288560,
      "step": 26780
    },
    {
      "epoch": 0.8573394789066001,
      "grad_norm": 28.0,
      "learning_rate": 1.2301015655484006e-05,
      "loss": 4.457,
      "num_input_tokens_seen": 36295232,
      "step": 26785
    },
    {
      "epoch": 0.8574995198770885,
      "grad_norm": 26.75,
      "learning_rate": 1.2292560046410245e-05,
      "loss": 3.2473,
      "num_input_tokens_seen": 36302416,
      "step": 26790
    },
    {
      "epoch": 0.857659560847577,
      "grad_norm": 22.875,
      "learning_rate": 1.228410639698343e-05,
      "loss": 3.7468,
      "num_input_tokens_seen": 36309216,
      "step": 26795
    },
    {
      "epoch": 0.8578196018180654,
      "grad_norm": 35.0,
      "learning_rate": 1.2275654708507195e-05,
      "loss": 4.3166,
      "num_input_tokens_seen": 36315872,
      "step": 26800
    },
    {
      "epoch": 0.8578196018180654,
      "eval_loss": 4.487011909484863,
      "eval_runtime": 489.2317,
      "eval_samples_per_second": 28.383,
      "eval_steps_per_second": 14.192,
      "num_input_tokens_seen": 36315872,
      "step": 26800
    },
    {
      "epoch": 0.8579796427885539,
      "grad_norm": 56.75,
      "learning_rate": 1.2267204982284908e-05,
      "loss": 4.5378,
      "num_input_tokens_seen": 36322464,
      "step": 26805
    },
    {
      "epoch": 0.8581396837590423,
      "grad_norm": 34.0,
      "learning_rate": 1.2258757219619635e-05,
      "loss": 4.907,
      "num_input_tokens_seen": 36329168,
      "step": 26810
    },
    {
      "epoch": 0.8582997247295308,
      "grad_norm": 31.5,
      "learning_rate": 1.2250311421814104e-05,
      "loss": 4.3711,
      "num_input_tokens_seen": 36336352,
      "step": 26815
    },
    {
      "epoch": 0.8584597657000193,
      "grad_norm": 27.0,
      "learning_rate": 1.2241867590170772e-05,
      "loss": 4.1594,
      "num_input_tokens_seen": 36343232,
      "step": 26820
    },
    {
      "epoch": 0.8586198066705076,
      "grad_norm": 14.3125,
      "learning_rate": 1.2233425725991799e-05,
      "loss": 4.73,
      "num_input_tokens_seen": 36350240,
      "step": 26825
    },
    {
      "epoch": 0.8587798476409961,
      "grad_norm": 21.25,
      "learning_rate": 1.2224985830579003e-05,
      "loss": 4.4576,
      "num_input_tokens_seen": 36356992,
      "step": 26830
    },
    {
      "epoch": 0.8589398886114845,
      "grad_norm": 29.875,
      "learning_rate": 1.2216547905233944e-05,
      "loss": 5.0134,
      "num_input_tokens_seen": 36363488,
      "step": 26835
    },
    {
      "epoch": 0.859099929581973,
      "grad_norm": 18.75,
      "learning_rate": 1.2208111951257842e-05,
      "loss": 3.9069,
      "num_input_tokens_seen": 36370576,
      "step": 26840
    },
    {
      "epoch": 0.8592599705524614,
      "grad_norm": 29.375,
      "learning_rate": 1.2199677969951622e-05,
      "loss": 4.9632,
      "num_input_tokens_seen": 36377840,
      "step": 26845
    },
    {
      "epoch": 0.8594200115229499,
      "grad_norm": 18.375,
      "learning_rate": 1.2191245962615927e-05,
      "loss": 3.9499,
      "num_input_tokens_seen": 36384624,
      "step": 26850
    },
    {
      "epoch": 0.8595800524934383,
      "grad_norm": 26.75,
      "learning_rate": 1.218281593055106e-05,
      "loss": 4.198,
      "num_input_tokens_seen": 36391504,
      "step": 26855
    },
    {
      "epoch": 0.8597400934639268,
      "grad_norm": 36.5,
      "learning_rate": 1.217438787505705e-05,
      "loss": 4.2037,
      "num_input_tokens_seen": 36398032,
      "step": 26860
    },
    {
      "epoch": 0.8599001344344153,
      "grad_norm": 48.25,
      "learning_rate": 1.2165961797433615e-05,
      "loss": 4.6409,
      "num_input_tokens_seen": 36404880,
      "step": 26865
    },
    {
      "epoch": 0.8600601754049036,
      "grad_norm": 33.75,
      "learning_rate": 1.215753769898014e-05,
      "loss": 4.7904,
      "num_input_tokens_seen": 36411360,
      "step": 26870
    },
    {
      "epoch": 0.8602202163753921,
      "grad_norm": 31.625,
      "learning_rate": 1.2149115580995755e-05,
      "loss": 4.8604,
      "num_input_tokens_seen": 36418000,
      "step": 26875
    },
    {
      "epoch": 0.8603802573458805,
      "grad_norm": 23.625,
      "learning_rate": 1.2140695444779227e-05,
      "loss": 4.0929,
      "num_input_tokens_seen": 36424960,
      "step": 26880
    },
    {
      "epoch": 0.860540298316369,
      "grad_norm": 29.5,
      "learning_rate": 1.2132277291629066e-05,
      "loss": 3.6777,
      "num_input_tokens_seen": 36432096,
      "step": 26885
    },
    {
      "epoch": 0.8607003392868574,
      "grad_norm": 31.125,
      "learning_rate": 1.2123861122843458e-05,
      "loss": 5.2731,
      "num_input_tokens_seen": 36439024,
      "step": 26890
    },
    {
      "epoch": 0.8608603802573459,
      "grad_norm": 54.5,
      "learning_rate": 1.2115446939720271e-05,
      "loss": 4.5826,
      "num_input_tokens_seen": 36446176,
      "step": 26895
    },
    {
      "epoch": 0.8610204212278343,
      "grad_norm": 26.625,
      "learning_rate": 1.210703474355708e-05,
      "loss": 4.4417,
      "num_input_tokens_seen": 36452896,
      "step": 26900
    },
    {
      "epoch": 0.8611804621983228,
      "grad_norm": 13.625,
      "learning_rate": 1.2098624535651164e-05,
      "loss": 3.9375,
      "num_input_tokens_seen": 36459456,
      "step": 26905
    },
    {
      "epoch": 0.8613405031688112,
      "grad_norm": 35.75,
      "learning_rate": 1.2090216317299477e-05,
      "loss": 4.3721,
      "num_input_tokens_seen": 36466240,
      "step": 26910
    },
    {
      "epoch": 0.8615005441392997,
      "grad_norm": 33.25,
      "learning_rate": 1.2081810089798668e-05,
      "loss": 5.977,
      "num_input_tokens_seen": 36472848,
      "step": 26915
    },
    {
      "epoch": 0.8616605851097882,
      "grad_norm": 30.75,
      "learning_rate": 1.2073405854445072e-05,
      "loss": 4.7023,
      "num_input_tokens_seen": 36479792,
      "step": 26920
    },
    {
      "epoch": 0.8618206260802765,
      "grad_norm": 24.375,
      "learning_rate": 1.206500361253474e-05,
      "loss": 2.9584,
      "num_input_tokens_seen": 36486544,
      "step": 26925
    },
    {
      "epoch": 0.861980667050765,
      "grad_norm": 50.0,
      "learning_rate": 1.2056603365363409e-05,
      "loss": 4.4283,
      "num_input_tokens_seen": 36493152,
      "step": 26930
    },
    {
      "epoch": 0.8621407080212534,
      "grad_norm": 28.625,
      "learning_rate": 1.2048205114226487e-05,
      "loss": 4.488,
      "num_input_tokens_seen": 36500256,
      "step": 26935
    },
    {
      "epoch": 0.8623007489917419,
      "grad_norm": 39.0,
      "learning_rate": 1.2039808860419102e-05,
      "loss": 4.1913,
      "num_input_tokens_seen": 36507536,
      "step": 26940
    },
    {
      "epoch": 0.8624607899622303,
      "grad_norm": 37.25,
      "learning_rate": 1.2031414605236066e-05,
      "loss": 3.4921,
      "num_input_tokens_seen": 36514624,
      "step": 26945
    },
    {
      "epoch": 0.8626208309327188,
      "grad_norm": 22.75,
      "learning_rate": 1.2023022349971862e-05,
      "loss": 4.2471,
      "num_input_tokens_seen": 36521504,
      "step": 26950
    },
    {
      "epoch": 0.8627808719032072,
      "grad_norm": 38.5,
      "learning_rate": 1.20146320959207e-05,
      "loss": 5.0222,
      "num_input_tokens_seen": 36528384,
      "step": 26955
    },
    {
      "epoch": 0.8629409128736957,
      "grad_norm": 57.5,
      "learning_rate": 1.2006243844376445e-05,
      "loss": 5.0865,
      "num_input_tokens_seen": 36534864,
      "step": 26960
    },
    {
      "epoch": 0.8631009538441841,
      "grad_norm": 40.0,
      "learning_rate": 1.1997857596632678e-05,
      "loss": 5.2324,
      "num_input_tokens_seen": 36541792,
      "step": 26965
    },
    {
      "epoch": 0.8632609948146726,
      "grad_norm": 34.75,
      "learning_rate": 1.1989473353982672e-05,
      "loss": 3.6047,
      "num_input_tokens_seen": 36548464,
      "step": 26970
    },
    {
      "epoch": 0.863421035785161,
      "grad_norm": 24.125,
      "learning_rate": 1.198109111771937e-05,
      "loss": 3.7495,
      "num_input_tokens_seen": 36555360,
      "step": 26975
    },
    {
      "epoch": 0.8635810767556494,
      "grad_norm": 30.625,
      "learning_rate": 1.197271088913543e-05,
      "loss": 4.6283,
      "num_input_tokens_seen": 36562256,
      "step": 26980
    },
    {
      "epoch": 0.8637411177261379,
      "grad_norm": 44.75,
      "learning_rate": 1.1964332669523182e-05,
      "loss": 4.6663,
      "num_input_tokens_seen": 36569600,
      "step": 26985
    },
    {
      "epoch": 0.8639011586966263,
      "grad_norm": 32.25,
      "learning_rate": 1.1955956460174645e-05,
      "loss": 4.7032,
      "num_input_tokens_seen": 36576400,
      "step": 26990
    },
    {
      "epoch": 0.8640611996671148,
      "grad_norm": 22.125,
      "learning_rate": 1.1947582262381552e-05,
      "loss": 5.0713,
      "num_input_tokens_seen": 36583248,
      "step": 26995
    },
    {
      "epoch": 0.8642212406376032,
      "grad_norm": 21.75,
      "learning_rate": 1.1939210077435293e-05,
      "loss": 4.4594,
      "num_input_tokens_seen": 36590336,
      "step": 27000
    },
    {
      "epoch": 0.8642212406376032,
      "eval_loss": 4.489694118499756,
      "eval_runtime": 489.7883,
      "eval_samples_per_second": 28.351,
      "eval_steps_per_second": 14.176,
      "num_input_tokens_seen": 36590336,
      "step": 27000
    },
    {
      "epoch": 0.8643812816080917,
      "grad_norm": 16.625,
      "learning_rate": 1.193083990662697e-05,
      "loss": 4.5586,
      "num_input_tokens_seen": 36596720,
      "step": 27005
    },
    {
      "epoch": 0.8645413225785801,
      "grad_norm": 39.25,
      "learning_rate": 1.192247175124738e-05,
      "loss": 4.3273,
      "num_input_tokens_seen": 36603088,
      "step": 27010
    },
    {
      "epoch": 0.8647013635490686,
      "grad_norm": 40.0,
      "learning_rate": 1.191410561258698e-05,
      "loss": 4.4288,
      "num_input_tokens_seen": 36609504,
      "step": 27015
    },
    {
      "epoch": 0.864861404519557,
      "grad_norm": 30.0,
      "learning_rate": 1.1905741491935944e-05,
      "loss": 5.5705,
      "num_input_tokens_seen": 36616336,
      "step": 27020
    },
    {
      "epoch": 0.8650214454900454,
      "grad_norm": 27.25,
      "learning_rate": 1.1897379390584129e-05,
      "loss": 3.2762,
      "num_input_tokens_seen": 36623216,
      "step": 27025
    },
    {
      "epoch": 0.8651814864605339,
      "grad_norm": 31.5,
      "learning_rate": 1.1889019309821062e-05,
      "loss": 4.9005,
      "num_input_tokens_seen": 36630032,
      "step": 27030
    },
    {
      "epoch": 0.8653415274310223,
      "grad_norm": 24.375,
      "learning_rate": 1.188066125093599e-05,
      "loss": 4.3282,
      "num_input_tokens_seen": 36636880,
      "step": 27035
    },
    {
      "epoch": 0.8655015684015108,
      "grad_norm": 29.0,
      "learning_rate": 1.1872305215217811e-05,
      "loss": 4.5481,
      "num_input_tokens_seen": 36643312,
      "step": 27040
    },
    {
      "epoch": 0.8656616093719992,
      "grad_norm": 35.5,
      "learning_rate": 1.186395120395514e-05,
      "loss": 4.1643,
      "num_input_tokens_seen": 36649952,
      "step": 27045
    },
    {
      "epoch": 0.8658216503424877,
      "grad_norm": 36.5,
      "learning_rate": 1.1855599218436283e-05,
      "loss": 4.2901,
      "num_input_tokens_seen": 36657024,
      "step": 27050
    },
    {
      "epoch": 0.8659816913129761,
      "grad_norm": 26.375,
      "learning_rate": 1.1847249259949209e-05,
      "loss": 4.136,
      "num_input_tokens_seen": 36663664,
      "step": 27055
    },
    {
      "epoch": 0.8661417322834646,
      "grad_norm": 21.0,
      "learning_rate": 1.1838901329781574e-05,
      "loss": 4.9049,
      "num_input_tokens_seen": 36670384,
      "step": 27060
    },
    {
      "epoch": 0.866301773253953,
      "grad_norm": 39.25,
      "learning_rate": 1.1830555429220758e-05,
      "loss": 5.1056,
      "num_input_tokens_seen": 36676768,
      "step": 27065
    },
    {
      "epoch": 0.8664618142244415,
      "grad_norm": 38.0,
      "learning_rate": 1.1822211559553784e-05,
      "loss": 5.1096,
      "num_input_tokens_seen": 36683328,
      "step": 27070
    },
    {
      "epoch": 0.86662185519493,
      "grad_norm": 31.5,
      "learning_rate": 1.18138697220674e-05,
      "loss": 4.1414,
      "num_input_tokens_seen": 36689984,
      "step": 27075
    },
    {
      "epoch": 0.8667818961654183,
      "grad_norm": 47.25,
      "learning_rate": 1.1805529918048e-05,
      "loss": 5.4621,
      "num_input_tokens_seen": 36696544,
      "step": 27080
    },
    {
      "epoch": 0.8669419371359068,
      "grad_norm": 33.25,
      "learning_rate": 1.1797192148781702e-05,
      "loss": 4.4347,
      "num_input_tokens_seen": 36703072,
      "step": 27085
    },
    {
      "epoch": 0.8671019781063952,
      "grad_norm": 46.75,
      "learning_rate": 1.1788856415554297e-05,
      "loss": 4.5718,
      "num_input_tokens_seen": 36709632,
      "step": 27090
    },
    {
      "epoch": 0.8672620190768837,
      "grad_norm": 36.25,
      "learning_rate": 1.1780522719651249e-05,
      "loss": 4.4216,
      "num_input_tokens_seen": 36716960,
      "step": 27095
    },
    {
      "epoch": 0.8674220600473721,
      "grad_norm": 27.5,
      "learning_rate": 1.1772191062357721e-05,
      "loss": 4.7508,
      "num_input_tokens_seen": 36723584,
      "step": 27100
    },
    {
      "epoch": 0.8675821010178606,
      "grad_norm": 31.0,
      "learning_rate": 1.1763861444958573e-05,
      "loss": 3.9843,
      "num_input_tokens_seen": 36729968,
      "step": 27105
    },
    {
      "epoch": 0.867742141988349,
      "grad_norm": 25.75,
      "learning_rate": 1.1755533868738317e-05,
      "loss": 3.8008,
      "num_input_tokens_seen": 36736944,
      "step": 27110
    },
    {
      "epoch": 0.8679021829588375,
      "grad_norm": 24.875,
      "learning_rate": 1.1747208334981185e-05,
      "loss": 3.7424,
      "num_input_tokens_seen": 36743792,
      "step": 27115
    },
    {
      "epoch": 0.8680622239293259,
      "grad_norm": 32.5,
      "learning_rate": 1.1738884844971067e-05,
      "loss": 4.7051,
      "num_input_tokens_seen": 36750432,
      "step": 27120
    },
    {
      "epoch": 0.8682222648998144,
      "grad_norm": 23.625,
      "learning_rate": 1.1730563399991563e-05,
      "loss": 4.3834,
      "num_input_tokens_seen": 36757216,
      "step": 27125
    },
    {
      "epoch": 0.8683823058703029,
      "grad_norm": 41.5,
      "learning_rate": 1.1722244001325938e-05,
      "loss": 5.0626,
      "num_input_tokens_seen": 36764480,
      "step": 27130
    },
    {
      "epoch": 0.8685423468407912,
      "grad_norm": 22.0,
      "learning_rate": 1.1713926650257137e-05,
      "loss": 3.894,
      "num_input_tokens_seen": 36770688,
      "step": 27135
    },
    {
      "epoch": 0.8687023878112797,
      "grad_norm": 27.875,
      "learning_rate": 1.170561134806781e-05,
      "loss": 4.7348,
      "num_input_tokens_seen": 36777760,
      "step": 27140
    },
    {
      "epoch": 0.8688624287817681,
      "grad_norm": 29.25,
      "learning_rate": 1.1697298096040287e-05,
      "loss": 4.0578,
      "num_input_tokens_seen": 36784464,
      "step": 27145
    },
    {
      "epoch": 0.8690224697522566,
      "grad_norm": 22.125,
      "learning_rate": 1.1688986895456567e-05,
      "loss": 4.7883,
      "num_input_tokens_seen": 36791216,
      "step": 27150
    },
    {
      "epoch": 0.869182510722745,
      "grad_norm": 47.0,
      "learning_rate": 1.1680677747598349e-05,
      "loss": 5.3613,
      "num_input_tokens_seen": 36797824,
      "step": 27155
    },
    {
      "epoch": 0.8693425516932335,
      "grad_norm": 29.625,
      "learning_rate": 1.1672370653746995e-05,
      "loss": 3.9663,
      "num_input_tokens_seen": 36804768,
      "step": 27160
    },
    {
      "epoch": 0.8695025926637219,
      "grad_norm": 36.75,
      "learning_rate": 1.166406561518357e-05,
      "loss": 5.7705,
      "num_input_tokens_seen": 36811120,
      "step": 27165
    },
    {
      "epoch": 0.8696626336342104,
      "grad_norm": 19.25,
      "learning_rate": 1.1655762633188826e-05,
      "loss": 4.1245,
      "num_input_tokens_seen": 36817712,
      "step": 27170
    },
    {
      "epoch": 0.8698226746046988,
      "grad_norm": 15.8125,
      "learning_rate": 1.1647461709043172e-05,
      "loss": 3.9881,
      "num_input_tokens_seen": 36824880,
      "step": 27175
    },
    {
      "epoch": 0.8699827155751872,
      "grad_norm": 25.875,
      "learning_rate": 1.1639162844026722e-05,
      "loss": 3.9623,
      "num_input_tokens_seen": 36831728,
      "step": 27180
    },
    {
      "epoch": 0.8701427565456757,
      "grad_norm": 28.0,
      "learning_rate": 1.163086603941927e-05,
      "loss": 4.7329,
      "num_input_tokens_seen": 36838192,
      "step": 27185
    },
    {
      "epoch": 0.8703027975161641,
      "grad_norm": 28.125,
      "learning_rate": 1.1622571296500273e-05,
      "loss": 4.1214,
      "num_input_tokens_seen": 36844848,
      "step": 27190
    },
    {
      "epoch": 0.8704628384866526,
      "grad_norm": 27.75,
      "learning_rate": 1.1614278616548904e-05,
      "loss": 3.9695,
      "num_input_tokens_seen": 36851392,
      "step": 27195
    },
    {
      "epoch": 0.870622879457141,
      "grad_norm": 25.625,
      "learning_rate": 1.1605988000843986e-05,
      "loss": 4.1458,
      "num_input_tokens_seen": 36858080,
      "step": 27200
    },
    {
      "epoch": 0.870622879457141,
      "eval_loss": 4.480690956115723,
      "eval_runtime": 489.6863,
      "eval_samples_per_second": 28.357,
      "eval_steps_per_second": 14.178,
      "num_input_tokens_seen": 36858080,
      "step": 27200
    },
    {
      "epoch": 0.8707829204276295,
      "grad_norm": 22.625,
      "learning_rate": 1.1597699450664028e-05,
      "loss": 4.5433,
      "num_input_tokens_seen": 36864640,
      "step": 27205
    },
    {
      "epoch": 0.8709429613981179,
      "grad_norm": 30.875,
      "learning_rate": 1.1589412967287252e-05,
      "loss": 5.051,
      "num_input_tokens_seen": 36871392,
      "step": 27210
    },
    {
      "epoch": 0.8711030023686064,
      "grad_norm": 24.25,
      "learning_rate": 1.1581128551991514e-05,
      "loss": 4.186,
      "num_input_tokens_seen": 36878176,
      "step": 27215
    },
    {
      "epoch": 0.8712630433390948,
      "grad_norm": 36.75,
      "learning_rate": 1.1572846206054383e-05,
      "loss": 4.7805,
      "num_input_tokens_seen": 36885152,
      "step": 27220
    },
    {
      "epoch": 0.8714230843095833,
      "grad_norm": 29.875,
      "learning_rate": 1.1564565930753113e-05,
      "loss": 3.8926,
      "num_input_tokens_seen": 36891616,
      "step": 27225
    },
    {
      "epoch": 0.8715831252800716,
      "grad_norm": 30.875,
      "learning_rate": 1.1556287727364606e-05,
      "loss": 4.0097,
      "num_input_tokens_seen": 36898672,
      "step": 27230
    },
    {
      "epoch": 0.8717431662505601,
      "grad_norm": 26.375,
      "learning_rate": 1.1548011597165489e-05,
      "loss": 4.3428,
      "num_input_tokens_seen": 36905840,
      "step": 27235
    },
    {
      "epoch": 0.8719032072210486,
      "grad_norm": 29.75,
      "learning_rate": 1.1539737541432019e-05,
      "loss": 3.9983,
      "num_input_tokens_seen": 36912672,
      "step": 27240
    },
    {
      "epoch": 0.872063248191537,
      "grad_norm": 33.0,
      "learning_rate": 1.1531465561440174e-05,
      "loss": 4.8665,
      "num_input_tokens_seen": 36919264,
      "step": 27245
    },
    {
      "epoch": 0.8722232891620255,
      "grad_norm": 22.5,
      "learning_rate": 1.1523195658465605e-05,
      "loss": 4.6016,
      "num_input_tokens_seen": 36925760,
      "step": 27250
    },
    {
      "epoch": 0.8723833301325139,
      "grad_norm": 42.0,
      "learning_rate": 1.1514927833783618e-05,
      "loss": 5.1981,
      "num_input_tokens_seen": 36932224,
      "step": 27255
    },
    {
      "epoch": 0.8725433711030024,
      "grad_norm": 14.25,
      "learning_rate": 1.150666208866922e-05,
      "loss": 5.0602,
      "num_input_tokens_seen": 36938880,
      "step": 27260
    },
    {
      "epoch": 0.8727034120734908,
      "grad_norm": 21.125,
      "learning_rate": 1.1498398424397106e-05,
      "loss": 3.6424,
      "num_input_tokens_seen": 36945392,
      "step": 27265
    },
    {
      "epoch": 0.8728634530439793,
      "grad_norm": 28.5,
      "learning_rate": 1.1490136842241628e-05,
      "loss": 5.3888,
      "num_input_tokens_seen": 36951936,
      "step": 27270
    },
    {
      "epoch": 0.8730234940144677,
      "grad_norm": 25.625,
      "learning_rate": 1.1481877343476813e-05,
      "loss": 4.9693,
      "num_input_tokens_seen": 36958768,
      "step": 27275
    },
    {
      "epoch": 0.8731835349849562,
      "grad_norm": 24.25,
      "learning_rate": 1.14736199293764e-05,
      "loss": 4.0024,
      "num_input_tokens_seen": 36965552,
      "step": 27280
    },
    {
      "epoch": 0.8733435759554445,
      "grad_norm": 28.625,
      "learning_rate": 1.1465364601213771e-05,
      "loss": 4.241,
      "num_input_tokens_seen": 36972336,
      "step": 27285
    },
    {
      "epoch": 0.873503616925933,
      "grad_norm": 45.0,
      "learning_rate": 1.1457111360262012e-05,
      "loss": 4.7143,
      "num_input_tokens_seen": 36978832,
      "step": 27290
    },
    {
      "epoch": 0.8736636578964215,
      "grad_norm": 29.125,
      "learning_rate": 1.1448860207793869e-05,
      "loss": 4.8345,
      "num_input_tokens_seen": 36985424,
      "step": 27295
    },
    {
      "epoch": 0.8738236988669099,
      "grad_norm": 35.75,
      "learning_rate": 1.144061114508177e-05,
      "loss": 4.2361,
      "num_input_tokens_seen": 36992240,
      "step": 27300
    },
    {
      "epoch": 0.8739837398373984,
      "grad_norm": 37.0,
      "learning_rate": 1.1432364173397842e-05,
      "loss": 4.5064,
      "num_input_tokens_seen": 36998976,
      "step": 27305
    },
    {
      "epoch": 0.8741437808078868,
      "grad_norm": 27.5,
      "learning_rate": 1.1424119294013852e-05,
      "loss": 3.8532,
      "num_input_tokens_seen": 37005584,
      "step": 27310
    },
    {
      "epoch": 0.8743038217783753,
      "grad_norm": 15.3125,
      "learning_rate": 1.1415876508201279e-05,
      "loss": 3.331,
      "num_input_tokens_seen": 37012336,
      "step": 27315
    },
    {
      "epoch": 0.8744638627488637,
      "grad_norm": 25.625,
      "learning_rate": 1.140763581723125e-05,
      "loss": 3.5697,
      "num_input_tokens_seen": 37019184,
      "step": 27320
    },
    {
      "epoch": 0.8746239037193522,
      "grad_norm": 29.25,
      "learning_rate": 1.1399397222374588e-05,
      "loss": 3.4941,
      "num_input_tokens_seen": 37026288,
      "step": 27325
    },
    {
      "epoch": 0.8747839446898406,
      "grad_norm": 27.0,
      "learning_rate": 1.1391160724901804e-05,
      "loss": 3.7269,
      "num_input_tokens_seen": 37032752,
      "step": 27330
    },
    {
      "epoch": 0.874943985660329,
      "grad_norm": 40.5,
      "learning_rate": 1.138292632608304e-05,
      "loss": 3.6859,
      "num_input_tokens_seen": 37039216,
      "step": 27335
    },
    {
      "epoch": 0.8751040266308175,
      "grad_norm": 24.375,
      "learning_rate": 1.1374694027188174e-05,
      "loss": 3.835,
      "num_input_tokens_seen": 37046064,
      "step": 27340
    },
    {
      "epoch": 0.8752640676013059,
      "grad_norm": 31.75,
      "learning_rate": 1.1366463829486711e-05,
      "loss": 4.2165,
      "num_input_tokens_seen": 37052576,
      "step": 27345
    },
    {
      "epoch": 0.8754241085717944,
      "grad_norm": 28.75,
      "learning_rate": 1.1358235734247849e-05,
      "loss": 3.6911,
      "num_input_tokens_seen": 37059200,
      "step": 27350
    },
    {
      "epoch": 0.8755841495422828,
      "grad_norm": 59.25,
      "learning_rate": 1.1350009742740478e-05,
      "loss": 4.1852,
      "num_input_tokens_seen": 37065824,
      "step": 27355
    },
    {
      "epoch": 0.8757441905127713,
      "grad_norm": 17.375,
      "learning_rate": 1.134178585623313e-05,
      "loss": 3.6858,
      "num_input_tokens_seen": 37072336,
      "step": 27360
    },
    {
      "epoch": 0.8759042314832597,
      "grad_norm": 20.25,
      "learning_rate": 1.1333564075994047e-05,
      "loss": 3.7713,
      "num_input_tokens_seen": 37078816,
      "step": 27365
    },
    {
      "epoch": 0.8760642724537482,
      "grad_norm": 30.75,
      "learning_rate": 1.1325344403291133e-05,
      "loss": 4.3289,
      "num_input_tokens_seen": 37085504,
      "step": 27370
    },
    {
      "epoch": 0.8762243134242366,
      "grad_norm": 16.375,
      "learning_rate": 1.1317126839391951e-05,
      "loss": 4.0183,
      "num_input_tokens_seen": 37091824,
      "step": 27375
    },
    {
      "epoch": 0.8763843543947251,
      "grad_norm": 28.875,
      "learning_rate": 1.1308911385563766e-05,
      "loss": 4.5914,
      "num_input_tokens_seen": 37098624,
      "step": 27380
    },
    {
      "epoch": 0.8765443953652134,
      "grad_norm": 30.625,
      "learning_rate": 1.1300698043073494e-05,
      "loss": 4.5059,
      "num_input_tokens_seen": 37105296,
      "step": 27385
    },
    {
      "epoch": 0.8767044363357019,
      "grad_norm": 29.75,
      "learning_rate": 1.1292486813187736e-05,
      "loss": 4.6083,
      "num_input_tokens_seen": 37111888,
      "step": 27390
    },
    {
      "epoch": 0.8768644773061904,
      "grad_norm": 44.0,
      "learning_rate": 1.1284277697172782e-05,
      "loss": 4.4272,
      "num_input_tokens_seen": 37118448,
      "step": 27395
    },
    {
      "epoch": 0.8770245182766788,
      "grad_norm": 26.5,
      "learning_rate": 1.127607069629456e-05,
      "loss": 4.2357,
      "num_input_tokens_seen": 37125216,
      "step": 27400
    },
    {
      "epoch": 0.8770245182766788,
      "eval_loss": 4.500524997711182,
      "eval_runtime": 489.288,
      "eval_samples_per_second": 28.38,
      "eval_steps_per_second": 14.19,
      "num_input_tokens_seen": 37125216,
      "step": 27400
    },
    {
      "epoch": 0.8771845592471673,
      "grad_norm": 32.5,
      "learning_rate": 1.1267865811818701e-05,
      "loss": 3.8048,
      "num_input_tokens_seen": 37132288,
      "step": 27405
    },
    {
      "epoch": 0.8773446002176557,
      "grad_norm": 51.75,
      "learning_rate": 1.1259663045010513e-05,
      "loss": 5.8401,
      "num_input_tokens_seen": 37139056,
      "step": 27410
    },
    {
      "epoch": 0.8775046411881442,
      "grad_norm": 24.0,
      "learning_rate": 1.1251462397134957e-05,
      "loss": 4.3055,
      "num_input_tokens_seen": 37145840,
      "step": 27415
    },
    {
      "epoch": 0.8776646821586326,
      "grad_norm": 43.0,
      "learning_rate": 1.1243263869456664e-05,
      "loss": 5.1264,
      "num_input_tokens_seen": 37152432,
      "step": 27420
    },
    {
      "epoch": 0.8778247231291211,
      "grad_norm": 44.0,
      "learning_rate": 1.1235067463239967e-05,
      "loss": 4.9412,
      "num_input_tokens_seen": 37159280,
      "step": 27425
    },
    {
      "epoch": 0.8779847640996095,
      "grad_norm": 36.25,
      "learning_rate": 1.122687317974884e-05,
      "loss": 5.3019,
      "num_input_tokens_seen": 37166160,
      "step": 27430
    },
    {
      "epoch": 0.878144805070098,
      "grad_norm": 24.375,
      "learning_rate": 1.1218681020246963e-05,
      "loss": 4.026,
      "num_input_tokens_seen": 37173184,
      "step": 27435
    },
    {
      "epoch": 0.8783048460405863,
      "grad_norm": 20.5,
      "learning_rate": 1.1210490985997652e-05,
      "loss": 3.8966,
      "num_input_tokens_seen": 37180272,
      "step": 27440
    },
    {
      "epoch": 0.8784648870110748,
      "grad_norm": 28.125,
      "learning_rate": 1.1202303078263917e-05,
      "loss": 4.9002,
      "num_input_tokens_seen": 37187104,
      "step": 27445
    },
    {
      "epoch": 0.8786249279815633,
      "grad_norm": 26.25,
      "learning_rate": 1.1194117298308451e-05,
      "loss": 4.3795,
      "num_input_tokens_seen": 37193776,
      "step": 27450
    },
    {
      "epoch": 0.8787849689520517,
      "grad_norm": 30.125,
      "learning_rate": 1.1185933647393585e-05,
      "loss": 4.0244,
      "num_input_tokens_seen": 37200672,
      "step": 27455
    },
    {
      "epoch": 0.8789450099225402,
      "grad_norm": 19.125,
      "learning_rate": 1.1177752126781354e-05,
      "loss": 4.5575,
      "num_input_tokens_seen": 37207360,
      "step": 27460
    },
    {
      "epoch": 0.8791050508930286,
      "grad_norm": 38.75,
      "learning_rate": 1.1169572737733441e-05,
      "loss": 4.1553,
      "num_input_tokens_seen": 37214304,
      "step": 27465
    },
    {
      "epoch": 0.8792650918635171,
      "grad_norm": 25.0,
      "learning_rate": 1.1161395481511216e-05,
      "loss": 3.8289,
      "num_input_tokens_seen": 37220928,
      "step": 27470
    },
    {
      "epoch": 0.8794251328340055,
      "grad_norm": 32.0,
      "learning_rate": 1.1153220359375722e-05,
      "loss": 5.3212,
      "num_input_tokens_seen": 37227920,
      "step": 27475
    },
    {
      "epoch": 0.879585173804494,
      "grad_norm": 28.75,
      "learning_rate": 1.114504737258765e-05,
      "loss": 4.3757,
      "num_input_tokens_seen": 37234864,
      "step": 27480
    },
    {
      "epoch": 0.8797452147749824,
      "grad_norm": 16.125,
      "learning_rate": 1.1136876522407393e-05,
      "loss": 4.0287,
      "num_input_tokens_seen": 37241536,
      "step": 27485
    },
    {
      "epoch": 0.8799052557454708,
      "grad_norm": 19.875,
      "learning_rate": 1.1128707810094985e-05,
      "loss": 4.513,
      "num_input_tokens_seen": 37247968,
      "step": 27490
    },
    {
      "epoch": 0.8800652967159592,
      "grad_norm": 21.125,
      "learning_rate": 1.1120541236910157e-05,
      "loss": 4.6679,
      "num_input_tokens_seen": 37254384,
      "step": 27495
    },
    {
      "epoch": 0.8802253376864477,
      "grad_norm": 41.75,
      "learning_rate": 1.111237680411229e-05,
      "loss": 4.2167,
      "num_input_tokens_seen": 37260832,
      "step": 27500
    },
    {
      "epoch": 0.8803853786569362,
      "grad_norm": 30.75,
      "learning_rate": 1.1104214512960433e-05,
      "loss": 4.0688,
      "num_input_tokens_seen": 37267952,
      "step": 27505
    },
    {
      "epoch": 0.8805454196274246,
      "grad_norm": 24.25,
      "learning_rate": 1.1096054364713327e-05,
      "loss": 4.4707,
      "num_input_tokens_seen": 37274672,
      "step": 27510
    },
    {
      "epoch": 0.8807054605979131,
      "grad_norm": 17.625,
      "learning_rate": 1.1087896360629371e-05,
      "loss": 4.3506,
      "num_input_tokens_seen": 37281584,
      "step": 27515
    },
    {
      "epoch": 0.8808655015684015,
      "grad_norm": 24.25,
      "learning_rate": 1.107974050196662e-05,
      "loss": 4.4597,
      "num_input_tokens_seen": 37288208,
      "step": 27520
    },
    {
      "epoch": 0.88102554253889,
      "grad_norm": 40.0,
      "learning_rate": 1.1071586789982816e-05,
      "loss": 4.3025,
      "num_input_tokens_seen": 37294944,
      "step": 27525
    },
    {
      "epoch": 0.8811855835093784,
      "grad_norm": 25.5,
      "learning_rate": 1.1063435225935373e-05,
      "loss": 4.1363,
      "num_input_tokens_seen": 37301808,
      "step": 27530
    },
    {
      "epoch": 0.8813456244798669,
      "grad_norm": 34.5,
      "learning_rate": 1.1055285811081348e-05,
      "loss": 4.7062,
      "num_input_tokens_seen": 37308528,
      "step": 27535
    },
    {
      "epoch": 0.8815056654503552,
      "grad_norm": 30.375,
      "learning_rate": 1.1047138546677499e-05,
      "loss": 3.927,
      "num_input_tokens_seen": 37316336,
      "step": 27540
    },
    {
      "epoch": 0.8816657064208437,
      "grad_norm": 33.25,
      "learning_rate": 1.1038993433980219e-05,
      "loss": 5.0805,
      "num_input_tokens_seen": 37323088,
      "step": 27545
    },
    {
      "epoch": 0.8818257473913321,
      "grad_norm": 19.625,
      "learning_rate": 1.1030850474245597e-05,
      "loss": 4.2992,
      "num_input_tokens_seen": 37329504,
      "step": 27550
    },
    {
      "epoch": 0.8819857883618206,
      "grad_norm": 32.0,
      "learning_rate": 1.102270966872939e-05,
      "loss": 4.5546,
      "num_input_tokens_seen": 37336432,
      "step": 27555
    },
    {
      "epoch": 0.8821458293323091,
      "grad_norm": 34.0,
      "learning_rate": 1.1014571018687e-05,
      "loss": 3.7028,
      "num_input_tokens_seen": 37343184,
      "step": 27560
    },
    {
      "epoch": 0.8823058703027975,
      "grad_norm": 27.625,
      "learning_rate": 1.1006434525373502e-05,
      "loss": 4.4363,
      "num_input_tokens_seen": 37350208,
      "step": 27565
    },
    {
      "epoch": 0.882465911273286,
      "grad_norm": 45.0,
      "learning_rate": 1.0998300190043664e-05,
      "loss": 4.8097,
      "num_input_tokens_seen": 37357040,
      "step": 27570
    },
    {
      "epoch": 0.8826259522437744,
      "grad_norm": 24.25,
      "learning_rate": 1.0990168013951882e-05,
      "loss": 3.8395,
      "num_input_tokens_seen": 37363776,
      "step": 27575
    },
    {
      "epoch": 0.8827859932142629,
      "grad_norm": 20.125,
      "learning_rate": 1.0982037998352263e-05,
      "loss": 4.0301,
      "num_input_tokens_seen": 37370256,
      "step": 27580
    },
    {
      "epoch": 0.8829460341847513,
      "grad_norm": 34.75,
      "learning_rate": 1.0973910144498534e-05,
      "loss": 4.8418,
      "num_input_tokens_seen": 37376944,
      "step": 27585
    },
    {
      "epoch": 0.8831060751552398,
      "grad_norm": 43.0,
      "learning_rate": 1.0965784453644123e-05,
      "loss": 4.6616,
      "num_input_tokens_seen": 37383952,
      "step": 27590
    },
    {
      "epoch": 0.8832661161257281,
      "grad_norm": 43.5,
      "learning_rate": 1.0957660927042127e-05,
      "loss": 4.2185,
      "num_input_tokens_seen": 37390864,
      "step": 27595
    },
    {
      "epoch": 0.8834261570962166,
      "grad_norm": 43.5,
      "learning_rate": 1.094953956594527e-05,
      "loss": 4.6963,
      "num_input_tokens_seen": 37397648,
      "step": 27600
    },
    {
      "epoch": 0.8834261570962166,
      "eval_loss": 4.492526054382324,
      "eval_runtime": 489.2571,
      "eval_samples_per_second": 28.382,
      "eval_steps_per_second": 14.191,
      "num_input_tokens_seen": 37397648,
      "step": 27600
    },
    {
      "epoch": 0.8835861980667051,
      "grad_norm": 42.0,
      "learning_rate": 1.0941420371605981e-05,
      "loss": 3.727,
      "num_input_tokens_seen": 37404528,
      "step": 27605
    },
    {
      "epoch": 0.8837462390371935,
      "grad_norm": 43.5,
      "learning_rate": 1.0933303345276354e-05,
      "loss": 4.6164,
      "num_input_tokens_seen": 37411296,
      "step": 27610
    },
    {
      "epoch": 0.883906280007682,
      "grad_norm": 30.875,
      "learning_rate": 1.0925188488208112e-05,
      "loss": 4.3116,
      "num_input_tokens_seen": 37418064,
      "step": 27615
    },
    {
      "epoch": 0.8840663209781704,
      "grad_norm": 43.0,
      "learning_rate": 1.0917075801652694e-05,
      "loss": 4.1021,
      "num_input_tokens_seen": 37425008,
      "step": 27620
    },
    {
      "epoch": 0.8842263619486589,
      "grad_norm": 21.875,
      "learning_rate": 1.0908965286861151e-05,
      "loss": 4.7467,
      "num_input_tokens_seen": 37431264,
      "step": 27625
    },
    {
      "epoch": 0.8843864029191473,
      "grad_norm": 16.875,
      "learning_rate": 1.090085694508425e-05,
      "loss": 3.8176,
      "num_input_tokens_seen": 37438000,
      "step": 27630
    },
    {
      "epoch": 0.8845464438896358,
      "grad_norm": 18.125,
      "learning_rate": 1.089275077757238e-05,
      "loss": 4.5538,
      "num_input_tokens_seen": 37444144,
      "step": 27635
    },
    {
      "epoch": 0.8847064848601242,
      "grad_norm": 28.875,
      "learning_rate": 1.0884646785575633e-05,
      "loss": 5.0579,
      "num_input_tokens_seen": 37450672,
      "step": 27640
    },
    {
      "epoch": 0.8848665258306126,
      "grad_norm": 34.25,
      "learning_rate": 1.0876544970343728e-05,
      "loss": 3.7007,
      "num_input_tokens_seen": 37457712,
      "step": 27645
    },
    {
      "epoch": 0.885026566801101,
      "grad_norm": 27.0,
      "learning_rate": 1.0868445333126082e-05,
      "loss": 4.1813,
      "num_input_tokens_seen": 37464368,
      "step": 27650
    },
    {
      "epoch": 0.8851866077715895,
      "grad_norm": 28.0,
      "learning_rate": 1.0860347875171745e-05,
      "loss": 3.9688,
      "num_input_tokens_seen": 37471152,
      "step": 27655
    },
    {
      "epoch": 0.885346648742078,
      "grad_norm": 18.125,
      "learning_rate": 1.0852252597729465e-05,
      "loss": 4.4836,
      "num_input_tokens_seen": 37477744,
      "step": 27660
    },
    {
      "epoch": 0.8855066897125664,
      "grad_norm": 27.125,
      "learning_rate": 1.0844159502047615e-05,
      "loss": 4.1327,
      "num_input_tokens_seen": 37484320,
      "step": 27665
    },
    {
      "epoch": 0.8856667306830549,
      "grad_norm": 33.25,
      "learning_rate": 1.0836068589374265e-05,
      "loss": 4.814,
      "num_input_tokens_seen": 37491520,
      "step": 27670
    },
    {
      "epoch": 0.8858267716535433,
      "grad_norm": 27.625,
      "learning_rate": 1.0827979860957144e-05,
      "loss": 4.0325,
      "num_input_tokens_seen": 37498560,
      "step": 27675
    },
    {
      "epoch": 0.8859868126240318,
      "grad_norm": 19.625,
      "learning_rate": 1.0819893318043615e-05,
      "loss": 4.9976,
      "num_input_tokens_seen": 37505392,
      "step": 27680
    },
    {
      "epoch": 0.8861468535945202,
      "grad_norm": 16.5,
      "learning_rate": 1.0811808961880734e-05,
      "loss": 3.7154,
      "num_input_tokens_seen": 37512432,
      "step": 27685
    },
    {
      "epoch": 0.8863068945650087,
      "grad_norm": 54.25,
      "learning_rate": 1.080372679371522e-05,
      "loss": 4.9198,
      "num_input_tokens_seen": 37519248,
      "step": 27690
    },
    {
      "epoch": 0.886466935535497,
      "grad_norm": 28.75,
      "learning_rate": 1.0795646814793428e-05,
      "loss": 4.3647,
      "num_input_tokens_seen": 37525744,
      "step": 27695
    },
    {
      "epoch": 0.8866269765059855,
      "grad_norm": 27.75,
      "learning_rate": 1.078756902636141e-05,
      "loss": 4.1654,
      "num_input_tokens_seen": 37532848,
      "step": 27700
    },
    {
      "epoch": 0.8867870174764739,
      "grad_norm": 34.5,
      "learning_rate": 1.077949342966485e-05,
      "loss": 4.7386,
      "num_input_tokens_seen": 37539728,
      "step": 27705
    },
    {
      "epoch": 0.8869470584469624,
      "grad_norm": 30.75,
      "learning_rate": 1.0771420025949103e-05,
      "loss": 4.7227,
      "num_input_tokens_seen": 37547056,
      "step": 27710
    },
    {
      "epoch": 0.8871070994174509,
      "grad_norm": 26.25,
      "learning_rate": 1.0763348816459204e-05,
      "loss": 5.7447,
      "num_input_tokens_seen": 37553744,
      "step": 27715
    },
    {
      "epoch": 0.8872671403879393,
      "grad_norm": 29.625,
      "learning_rate": 1.0755279802439816e-05,
      "loss": 5.2984,
      "num_input_tokens_seen": 37560320,
      "step": 27720
    },
    {
      "epoch": 0.8874271813584278,
      "grad_norm": 19.5,
      "learning_rate": 1.0747212985135293e-05,
      "loss": 3.9487,
      "num_input_tokens_seen": 37566944,
      "step": 27725
    },
    {
      "epoch": 0.8875872223289162,
      "grad_norm": 21.25,
      "learning_rate": 1.073914836578965e-05,
      "loss": 3.7996,
      "num_input_tokens_seen": 37573712,
      "step": 27730
    },
    {
      "epoch": 0.8877472632994047,
      "grad_norm": 37.0,
      "learning_rate": 1.0731085945646529e-05,
      "loss": 4.9112,
      "num_input_tokens_seen": 37580560,
      "step": 27735
    },
    {
      "epoch": 0.8879073042698931,
      "grad_norm": 59.0,
      "learning_rate": 1.0723025725949285e-05,
      "loss": 3.8013,
      "num_input_tokens_seen": 37587568,
      "step": 27740
    },
    {
      "epoch": 0.8880673452403816,
      "grad_norm": 18.625,
      "learning_rate": 1.0714967707940875e-05,
      "loss": 4.6006,
      "num_input_tokens_seen": 37594608,
      "step": 27745
    },
    {
      "epoch": 0.8882273862108699,
      "grad_norm": 45.5,
      "learning_rate": 1.0706911892863963e-05,
      "loss": 4.0604,
      "num_input_tokens_seen": 37601296,
      "step": 27750
    },
    {
      "epoch": 0.8883874271813584,
      "grad_norm": 73.5,
      "learning_rate": 1.0698858281960866e-05,
      "loss": 4.8353,
      "num_input_tokens_seen": 37608000,
      "step": 27755
    },
    {
      "epoch": 0.8885474681518468,
      "grad_norm": 43.5,
      "learning_rate": 1.069080687647353e-05,
      "loss": 4.3395,
      "num_input_tokens_seen": 37614384,
      "step": 27760
    },
    {
      "epoch": 0.8887075091223353,
      "grad_norm": 36.75,
      "learning_rate": 1.0682757677643596e-05,
      "loss": 4.5464,
      "num_input_tokens_seen": 37620816,
      "step": 27765
    },
    {
      "epoch": 0.8888675500928238,
      "grad_norm": 23.25,
      "learning_rate": 1.0674710686712359e-05,
      "loss": 4.7133,
      "num_input_tokens_seen": 37627664,
      "step": 27770
    },
    {
      "epoch": 0.8890275910633122,
      "grad_norm": 11.4375,
      "learning_rate": 1.0666665904920756e-05,
      "loss": 4.3591,
      "num_input_tokens_seen": 37633920,
      "step": 27775
    },
    {
      "epoch": 0.8891876320338007,
      "grad_norm": 20.375,
      "learning_rate": 1.0658623333509385e-05,
      "loss": 4.2536,
      "num_input_tokens_seen": 37640384,
      "step": 27780
    },
    {
      "epoch": 0.8893476730042891,
      "grad_norm": 31.375,
      "learning_rate": 1.0650582973718532e-05,
      "loss": 3.8796,
      "num_input_tokens_seen": 37646896,
      "step": 27785
    },
    {
      "epoch": 0.8895077139747776,
      "grad_norm": 31.5,
      "learning_rate": 1.0642544826788098e-05,
      "loss": 4.266,
      "num_input_tokens_seen": 37654064,
      "step": 27790
    },
    {
      "epoch": 0.889667754945266,
      "grad_norm": 31.0,
      "learning_rate": 1.063450889395769e-05,
      "loss": 4.145,
      "num_input_tokens_seen": 37660864,
      "step": 27795
    },
    {
      "epoch": 0.8898277959157544,
      "grad_norm": 22.375,
      "learning_rate": 1.062647517646653e-05,
      "loss": 3.6524,
      "num_input_tokens_seen": 37667456,
      "step": 27800
    },
    {
      "epoch": 0.8898277959157544,
      "eval_loss": 4.514245986938477,
      "eval_runtime": 489.206,
      "eval_samples_per_second": 28.385,
      "eval_steps_per_second": 14.192,
      "num_input_tokens_seen": 37667456,
      "step": 27800
    },
    {
      "epoch": 0.8899878368862428,
      "grad_norm": 23.375,
      "learning_rate": 1.0618443675553527e-05,
      "loss": 3.8842,
      "num_input_tokens_seen": 37673872,
      "step": 27805
    },
    {
      "epoch": 0.8901478778567313,
      "grad_norm": 95.5,
      "learning_rate": 1.0610414392457247e-05,
      "loss": 3.9749,
      "num_input_tokens_seen": 37680288,
      "step": 27810
    },
    {
      "epoch": 0.8903079188272197,
      "grad_norm": 31.125,
      "learning_rate": 1.0602387328415888e-05,
      "loss": 4.8834,
      "num_input_tokens_seen": 37687296,
      "step": 27815
    },
    {
      "epoch": 0.8904679597977082,
      "grad_norm": 52.5,
      "learning_rate": 1.0594362484667347e-05,
      "loss": 4.2541,
      "num_input_tokens_seen": 37694016,
      "step": 27820
    },
    {
      "epoch": 0.8906280007681967,
      "grad_norm": 23.0,
      "learning_rate": 1.0586339862449132e-05,
      "loss": 3.6538,
      "num_input_tokens_seen": 37700752,
      "step": 27825
    },
    {
      "epoch": 0.8907880417386851,
      "grad_norm": 31.25,
      "learning_rate": 1.0578319462998445e-05,
      "loss": 4.3083,
      "num_input_tokens_seen": 37707392,
      "step": 27830
    },
    {
      "epoch": 0.8909480827091736,
      "grad_norm": 46.0,
      "learning_rate": 1.057030128755214e-05,
      "loss": 4.2709,
      "num_input_tokens_seen": 37714064,
      "step": 27835
    },
    {
      "epoch": 0.891108123679662,
      "grad_norm": 41.75,
      "learning_rate": 1.0562285337346703e-05,
      "loss": 5.6661,
      "num_input_tokens_seen": 37720624,
      "step": 27840
    },
    {
      "epoch": 0.8912681646501505,
      "grad_norm": 28.875,
      "learning_rate": 1.0554271613618308e-05,
      "loss": 4.5009,
      "num_input_tokens_seen": 37727520,
      "step": 27845
    },
    {
      "epoch": 0.8914282056206388,
      "grad_norm": 20.875,
      "learning_rate": 1.054626011760276e-05,
      "loss": 4.4938,
      "num_input_tokens_seen": 37734784,
      "step": 27850
    },
    {
      "epoch": 0.8915882465911273,
      "grad_norm": 43.5,
      "learning_rate": 1.0538250850535549e-05,
      "loss": 4.5606,
      "num_input_tokens_seen": 37741056,
      "step": 27855
    },
    {
      "epoch": 0.8917482875616157,
      "grad_norm": 26.0,
      "learning_rate": 1.0530243813651794e-05,
      "loss": 4.2859,
      "num_input_tokens_seen": 37747728,
      "step": 27860
    },
    {
      "epoch": 0.8919083285321042,
      "grad_norm": 33.75,
      "learning_rate": 1.0522239008186271e-05,
      "loss": 4.8562,
      "num_input_tokens_seen": 37754496,
      "step": 27865
    },
    {
      "epoch": 0.8920683695025927,
      "grad_norm": 27.625,
      "learning_rate": 1.0514236435373434e-05,
      "loss": 4.5362,
      "num_input_tokens_seen": 37760800,
      "step": 27870
    },
    {
      "epoch": 0.8922284104730811,
      "grad_norm": 26.25,
      "learning_rate": 1.0506236096447386e-05,
      "loss": 4.3689,
      "num_input_tokens_seen": 37767376,
      "step": 27875
    },
    {
      "epoch": 0.8923884514435696,
      "grad_norm": 24.0,
      "learning_rate": 1.049823799264186e-05,
      "loss": 3.2211,
      "num_input_tokens_seen": 37774240,
      "step": 27880
    },
    {
      "epoch": 0.892548492414058,
      "grad_norm": 24.625,
      "learning_rate": 1.049024212519028e-05,
      "loss": 4.4799,
      "num_input_tokens_seen": 37780912,
      "step": 27885
    },
    {
      "epoch": 0.8927085333845465,
      "grad_norm": 18.625,
      "learning_rate": 1.0482248495325713e-05,
      "loss": 4.3038,
      "num_input_tokens_seen": 37787792,
      "step": 27890
    },
    {
      "epoch": 0.8928685743550349,
      "grad_norm": 29.75,
      "learning_rate": 1.047425710428086e-05,
      "loss": 3.7385,
      "num_input_tokens_seen": 37794432,
      "step": 27895
    },
    {
      "epoch": 0.8930286153255234,
      "grad_norm": 44.25,
      "learning_rate": 1.0466267953288114e-05,
      "loss": 4.5765,
      "num_input_tokens_seen": 37801232,
      "step": 27900
    },
    {
      "epoch": 0.8931886562960117,
      "grad_norm": 18.0,
      "learning_rate": 1.0458281043579482e-05,
      "loss": 3.7169,
      "num_input_tokens_seen": 37808144,
      "step": 27905
    },
    {
      "epoch": 0.8933486972665002,
      "grad_norm": 42.0,
      "learning_rate": 1.0450296376386657e-05,
      "loss": 5.1535,
      "num_input_tokens_seen": 37814864,
      "step": 27910
    },
    {
      "epoch": 0.8935087382369886,
      "grad_norm": 18.75,
      "learning_rate": 1.044231395294098e-05,
      "loss": 3.7762,
      "num_input_tokens_seen": 37821616,
      "step": 27915
    },
    {
      "epoch": 0.8936687792074771,
      "grad_norm": 48.0,
      "learning_rate": 1.0434333774473435e-05,
      "loss": 3.8867,
      "num_input_tokens_seen": 37828752,
      "step": 27920
    },
    {
      "epoch": 0.8938288201779656,
      "grad_norm": 30.125,
      "learning_rate": 1.0426355842214657e-05,
      "loss": 3.8446,
      "num_input_tokens_seen": 37835056,
      "step": 27925
    },
    {
      "epoch": 0.893988861148454,
      "grad_norm": 39.75,
      "learning_rate": 1.0418380157394963e-05,
      "loss": 4.7531,
      "num_input_tokens_seen": 37842160,
      "step": 27930
    },
    {
      "epoch": 0.8941489021189425,
      "grad_norm": 34.5,
      "learning_rate": 1.0410406721244281e-05,
      "loss": 5.3867,
      "num_input_tokens_seen": 37848976,
      "step": 27935
    },
    {
      "epoch": 0.8943089430894309,
      "grad_norm": 30.125,
      "learning_rate": 1.0402435534992238e-05,
      "loss": 4.0886,
      "num_input_tokens_seen": 37855824,
      "step": 27940
    },
    {
      "epoch": 0.8944689840599194,
      "grad_norm": 41.5,
      "learning_rate": 1.0394466599868071e-05,
      "loss": 4.7449,
      "num_input_tokens_seen": 37862624,
      "step": 27945
    },
    {
      "epoch": 0.8946290250304078,
      "grad_norm": 26.0,
      "learning_rate": 1.0386499917100697e-05,
      "loss": 4.3786,
      "num_input_tokens_seen": 37869408,
      "step": 27950
    },
    {
      "epoch": 0.8947890660008962,
      "grad_norm": 28.875,
      "learning_rate": 1.0378535487918692e-05,
      "loss": 4.3997,
      "num_input_tokens_seen": 37875952,
      "step": 27955
    },
    {
      "epoch": 0.8949491069713846,
      "grad_norm": 63.75,
      "learning_rate": 1.037057331355025e-05,
      "loss": 3.9746,
      "num_input_tokens_seen": 37882704,
      "step": 27960
    },
    {
      "epoch": 0.8951091479418731,
      "grad_norm": 41.25,
      "learning_rate": 1.0362613395223247e-05,
      "loss": 5.1725,
      "num_input_tokens_seen": 37889360,
      "step": 27965
    },
    {
      "epoch": 0.8952691889123615,
      "grad_norm": 37.75,
      "learning_rate": 1.0354655734165212e-05,
      "loss": 4.9481,
      "num_input_tokens_seen": 37896368,
      "step": 27970
    },
    {
      "epoch": 0.89542922988285,
      "grad_norm": 24.875,
      "learning_rate": 1.03467003316033e-05,
      "loss": 3.96,
      "num_input_tokens_seen": 37902576,
      "step": 27975
    },
    {
      "epoch": 0.8955892708533385,
      "grad_norm": 29.5,
      "learning_rate": 1.033874718876435e-05,
      "loss": 4.7532,
      "num_input_tokens_seen": 37909216,
      "step": 27980
    },
    {
      "epoch": 0.8957493118238269,
      "grad_norm": 27.75,
      "learning_rate": 1.0330796306874818e-05,
      "loss": 4.3703,
      "num_input_tokens_seen": 37915760,
      "step": 27985
    },
    {
      "epoch": 0.8959093527943154,
      "grad_norm": 38.0,
      "learning_rate": 1.032284768716085e-05,
      "loss": 4.6971,
      "num_input_tokens_seen": 37922736,
      "step": 27990
    },
    {
      "epoch": 0.8960693937648038,
      "grad_norm": 28.75,
      "learning_rate": 1.0314901330848206e-05,
      "loss": 3.8832,
      "num_input_tokens_seen": 37929312,
      "step": 27995
    },
    {
      "epoch": 0.8962294347352923,
      "grad_norm": 20.5,
      "learning_rate": 1.030695723916233e-05,
      "loss": 4.0705,
      "num_input_tokens_seen": 37935760,
      "step": 28000
    },
    {
      "epoch": 0.8962294347352923,
      "eval_loss": 4.486621379852295,
      "eval_runtime": 489.2293,
      "eval_samples_per_second": 28.383,
      "eval_steps_per_second": 14.192,
      "num_input_tokens_seen": 37935760,
      "step": 28000
    },
    {
      "epoch": 0.8963894757057806,
      "grad_norm": 41.75,
      "learning_rate": 1.0299015413328289e-05,
      "loss": 4.9591,
      "num_input_tokens_seen": 37942464,
      "step": 28005
    },
    {
      "epoch": 0.8965495166762691,
      "grad_norm": 28.125,
      "learning_rate": 1.0291075854570809e-05,
      "loss": 4.3692,
      "num_input_tokens_seen": 37949264,
      "step": 28010
    },
    {
      "epoch": 0.8967095576467575,
      "grad_norm": 29.25,
      "learning_rate": 1.0283138564114275e-05,
      "loss": 4.2141,
      "num_input_tokens_seen": 37955888,
      "step": 28015
    },
    {
      "epoch": 0.896869598617246,
      "grad_norm": 24.875,
      "learning_rate": 1.027520354318273e-05,
      "loss": 4.0561,
      "num_input_tokens_seen": 37962688,
      "step": 28020
    },
    {
      "epoch": 0.8970296395877344,
      "grad_norm": 29.0,
      "learning_rate": 1.0267270792999828e-05,
      "loss": 4.0807,
      "num_input_tokens_seen": 37969296,
      "step": 28025
    },
    {
      "epoch": 0.8971896805582229,
      "grad_norm": 21.5,
      "learning_rate": 1.0259340314788919e-05,
      "loss": 4.129,
      "num_input_tokens_seen": 37975616,
      "step": 28030
    },
    {
      "epoch": 0.8973497215287114,
      "grad_norm": 23.5,
      "learning_rate": 1.0251412109772979e-05,
      "loss": 4.3724,
      "num_input_tokens_seen": 37982432,
      "step": 28035
    },
    {
      "epoch": 0.8975097624991998,
      "grad_norm": 35.5,
      "learning_rate": 1.0243486179174627e-05,
      "loss": 4.4391,
      "num_input_tokens_seen": 37989008,
      "step": 28040
    },
    {
      "epoch": 0.8976698034696883,
      "grad_norm": 33.25,
      "learning_rate": 1.0235562524216158e-05,
      "loss": 4.8435,
      "num_input_tokens_seen": 37996336,
      "step": 28045
    },
    {
      "epoch": 0.8978298444401767,
      "grad_norm": 46.25,
      "learning_rate": 1.022764114611948e-05,
      "loss": 6.6915,
      "num_input_tokens_seen": 38003104,
      "step": 28050
    },
    {
      "epoch": 0.8979898854106652,
      "grad_norm": 24.625,
      "learning_rate": 1.0219722046106178e-05,
      "loss": 4.63,
      "num_input_tokens_seen": 38009424,
      "step": 28055
    },
    {
      "epoch": 0.8981499263811535,
      "grad_norm": 24.75,
      "learning_rate": 1.0211805225397486e-05,
      "loss": 3.9881,
      "num_input_tokens_seen": 38016208,
      "step": 28060
    },
    {
      "epoch": 0.898309967351642,
      "grad_norm": 39.75,
      "learning_rate": 1.020389068521426e-05,
      "loss": 5.2183,
      "num_input_tokens_seen": 38022544,
      "step": 28065
    },
    {
      "epoch": 0.8984700083221304,
      "grad_norm": 19.375,
      "learning_rate": 1.0195978426777039e-05,
      "loss": 4.1221,
      "num_input_tokens_seen": 38029888,
      "step": 28070
    },
    {
      "epoch": 0.8986300492926189,
      "grad_norm": 39.25,
      "learning_rate": 1.0188068451305982e-05,
      "loss": 4.897,
      "num_input_tokens_seen": 38036608,
      "step": 28075
    },
    {
      "epoch": 0.8987900902631074,
      "grad_norm": 21.0,
      "learning_rate": 1.0180160760020902e-05,
      "loss": 4.1531,
      "num_input_tokens_seen": 38043216,
      "step": 28080
    },
    {
      "epoch": 0.8989501312335958,
      "grad_norm": 18.75,
      "learning_rate": 1.0172255354141278e-05,
      "loss": 4.8436,
      "num_input_tokens_seen": 38049840,
      "step": 28085
    },
    {
      "epoch": 0.8991101722040843,
      "grad_norm": 45.25,
      "learning_rate": 1.0164352234886205e-05,
      "loss": 4.125,
      "num_input_tokens_seen": 38056640,
      "step": 28090
    },
    {
      "epoch": 0.8992702131745727,
      "grad_norm": 36.0,
      "learning_rate": 1.0156451403474454e-05,
      "loss": 4.0476,
      "num_input_tokens_seen": 38063424,
      "step": 28095
    },
    {
      "epoch": 0.8994302541450612,
      "grad_norm": 13.6875,
      "learning_rate": 1.0148552861124443e-05,
      "loss": 4.1166,
      "num_input_tokens_seen": 38070608,
      "step": 28100
    },
    {
      "epoch": 0.8995902951155496,
      "grad_norm": 36.25,
      "learning_rate": 1.0140656609054205e-05,
      "loss": 3.7719,
      "num_input_tokens_seen": 38077616,
      "step": 28105
    },
    {
      "epoch": 0.899750336086038,
      "grad_norm": 15.75,
      "learning_rate": 1.0132762648481455e-05,
      "loss": 3.7416,
      "num_input_tokens_seen": 38084560,
      "step": 28110
    },
    {
      "epoch": 0.8999103770565264,
      "grad_norm": 26.5,
      "learning_rate": 1.0124870980623543e-05,
      "loss": 5.5006,
      "num_input_tokens_seen": 38091408,
      "step": 28115
    },
    {
      "epoch": 0.9000704180270149,
      "grad_norm": 41.5,
      "learning_rate": 1.0116981606697453e-05,
      "loss": 4.9244,
      "num_input_tokens_seen": 38097952,
      "step": 28120
    },
    {
      "epoch": 0.9002304589975033,
      "grad_norm": 17.25,
      "learning_rate": 1.0109094527919838e-05,
      "loss": 4.5368,
      "num_input_tokens_seen": 38104544,
      "step": 28125
    },
    {
      "epoch": 0.9003904999679918,
      "grad_norm": 28.0,
      "learning_rate": 1.010120974550697e-05,
      "loss": 4.4199,
      "num_input_tokens_seen": 38111280,
      "step": 28130
    },
    {
      "epoch": 0.9005505409384803,
      "grad_norm": 31.75,
      "learning_rate": 1.0093327260674795e-05,
      "loss": 4.7513,
      "num_input_tokens_seen": 38118160,
      "step": 28135
    },
    {
      "epoch": 0.9007105819089687,
      "grad_norm": 39.75,
      "learning_rate": 1.0085447074638878e-05,
      "loss": 4.0866,
      "num_input_tokens_seen": 38124640,
      "step": 28140
    },
    {
      "epoch": 0.9008706228794572,
      "grad_norm": 41.75,
      "learning_rate": 1.0077569188614461e-05,
      "loss": 4.361,
      "num_input_tokens_seen": 38131168,
      "step": 28145
    },
    {
      "epoch": 0.9010306638499456,
      "grad_norm": 22.5,
      "learning_rate": 1.0069693603816393e-05,
      "loss": 4.3176,
      "num_input_tokens_seen": 38137728,
      "step": 28150
    },
    {
      "epoch": 0.9011907048204341,
      "grad_norm": 24.125,
      "learning_rate": 1.0061820321459204e-05,
      "loss": 4.2872,
      "num_input_tokens_seen": 38144464,
      "step": 28155
    },
    {
      "epoch": 0.9013507457909224,
      "grad_norm": 27.5,
      "learning_rate": 1.0053949342757038e-05,
      "loss": 3.7725,
      "num_input_tokens_seen": 38150928,
      "step": 28160
    },
    {
      "epoch": 0.9015107867614109,
      "grad_norm": 21.75,
      "learning_rate": 1.0046080668923717e-05,
      "loss": 4.2753,
      "num_input_tokens_seen": 38157872,
      "step": 28165
    },
    {
      "epoch": 0.9016708277318993,
      "grad_norm": 44.0,
      "learning_rate": 1.003821430117267e-05,
      "loss": 4.3826,
      "num_input_tokens_seen": 38164896,
      "step": 28170
    },
    {
      "epoch": 0.9018308687023878,
      "grad_norm": 26.625,
      "learning_rate": 1.0030350240716999e-05,
      "loss": 4.4186,
      "num_input_tokens_seen": 38171520,
      "step": 28175
    },
    {
      "epoch": 0.9019909096728762,
      "grad_norm": 20.5,
      "learning_rate": 1.0022488488769449e-05,
      "loss": 4.7621,
      "num_input_tokens_seen": 38178000,
      "step": 28180
    },
    {
      "epoch": 0.9021509506433647,
      "grad_norm": 32.75,
      "learning_rate": 1.0014629046542387e-05,
      "loss": 4.057,
      "num_input_tokens_seen": 38184912,
      "step": 28185
    },
    {
      "epoch": 0.9023109916138532,
      "grad_norm": 23.625,
      "learning_rate": 1.0006771915247842e-05,
      "loss": 4.3111,
      "num_input_tokens_seen": 38191552,
      "step": 28190
    },
    {
      "epoch": 0.9024710325843416,
      "grad_norm": 36.0,
      "learning_rate": 9.998917096097495e-06,
      "loss": 4.2478,
      "num_input_tokens_seen": 38198224,
      "step": 28195
    },
    {
      "epoch": 0.9026310735548301,
      "grad_norm": 34.25,
      "learning_rate": 9.991064590302638e-06,
      "loss": 4.1514,
      "num_input_tokens_seen": 38204832,
      "step": 28200
    },
    {
      "epoch": 0.9026310735548301,
      "eval_loss": 4.484555244445801,
      "eval_runtime": 489.6073,
      "eval_samples_per_second": 28.362,
      "eval_steps_per_second": 14.181,
      "num_input_tokens_seen": 38204832,
      "step": 28200
    },
    {
      "epoch": 0.9027911145253185,
      "grad_norm": 28.25,
      "learning_rate": 9.983214399074241e-06,
      "loss": 4.4359,
      "num_input_tokens_seen": 38211344,
      "step": 28205
    },
    {
      "epoch": 0.902951155495807,
      "grad_norm": 20.875,
      "learning_rate": 9.975366523622893e-06,
      "loss": 3.5999,
      "num_input_tokens_seen": 38218096,
      "step": 28210
    },
    {
      "epoch": 0.9031111964662953,
      "grad_norm": 44.0,
      "learning_rate": 9.967520965158841e-06,
      "loss": 4.7939,
      "num_input_tokens_seen": 38225648,
      "step": 28215
    },
    {
      "epoch": 0.9032712374367838,
      "grad_norm": 22.375,
      "learning_rate": 9.95967772489197e-06,
      "loss": 4.0535,
      "num_input_tokens_seen": 38232512,
      "step": 28220
    },
    {
      "epoch": 0.9034312784072722,
      "grad_norm": 35.75,
      "learning_rate": 9.951836804031794e-06,
      "loss": 4.3208,
      "num_input_tokens_seen": 38238944,
      "step": 28225
    },
    {
      "epoch": 0.9035913193777607,
      "grad_norm": 24.25,
      "learning_rate": 9.943998203787489e-06,
      "loss": 3.1039,
      "num_input_tokens_seen": 38245712,
      "step": 28230
    },
    {
      "epoch": 0.9037513603482491,
      "grad_norm": 29.125,
      "learning_rate": 9.936161925367874e-06,
      "loss": 4.7431,
      "num_input_tokens_seen": 38252272,
      "step": 28235
    },
    {
      "epoch": 0.9039114013187376,
      "grad_norm": 19.125,
      "learning_rate": 9.928327969981386e-06,
      "loss": 3.8103,
      "num_input_tokens_seen": 38259040,
      "step": 28240
    },
    {
      "epoch": 0.9040714422892261,
      "grad_norm": 33.5,
      "learning_rate": 9.920496338836135e-06,
      "loss": 4.0457,
      "num_input_tokens_seen": 38265664,
      "step": 28245
    },
    {
      "epoch": 0.9042314832597145,
      "grad_norm": 26.75,
      "learning_rate": 9.912667033139844e-06,
      "loss": 3.9957,
      "num_input_tokens_seen": 38272432,
      "step": 28250
    },
    {
      "epoch": 0.904391524230203,
      "grad_norm": 28.125,
      "learning_rate": 9.904840054099893e-06,
      "loss": 4.079,
      "num_input_tokens_seen": 38279984,
      "step": 28255
    },
    {
      "epoch": 0.9045515652006914,
      "grad_norm": 30.0,
      "learning_rate": 9.897015402923312e-06,
      "loss": 3.9224,
      "num_input_tokens_seen": 38286432,
      "step": 28260
    },
    {
      "epoch": 0.9047116061711798,
      "grad_norm": 30.375,
      "learning_rate": 9.889193080816744e-06,
      "loss": 4.6186,
      "num_input_tokens_seen": 38293728,
      "step": 28265
    },
    {
      "epoch": 0.9048716471416682,
      "grad_norm": 36.75,
      "learning_rate": 9.881373088986498e-06,
      "loss": 4.3221,
      "num_input_tokens_seen": 38300736,
      "step": 28270
    },
    {
      "epoch": 0.9050316881121567,
      "grad_norm": 42.0,
      "learning_rate": 9.873555428638523e-06,
      "loss": 4.8302,
      "num_input_tokens_seen": 38307456,
      "step": 28275
    },
    {
      "epoch": 0.9051917290826451,
      "grad_norm": 45.25,
      "learning_rate": 9.865740100978383e-06,
      "loss": 4.3399,
      "num_input_tokens_seen": 38314128,
      "step": 28280
    },
    {
      "epoch": 0.9053517700531336,
      "grad_norm": 51.25,
      "learning_rate": 9.857927107211315e-06,
      "loss": 3.9204,
      "num_input_tokens_seen": 38320832,
      "step": 28285
    },
    {
      "epoch": 0.905511811023622,
      "grad_norm": 38.5,
      "learning_rate": 9.850116448542177e-06,
      "loss": 4.6984,
      "num_input_tokens_seen": 38327296,
      "step": 28290
    },
    {
      "epoch": 0.9056718519941105,
      "grad_norm": 27.625,
      "learning_rate": 9.842308126175457e-06,
      "loss": 4.8058,
      "num_input_tokens_seen": 38334480,
      "step": 28295
    },
    {
      "epoch": 0.905831892964599,
      "grad_norm": 21.375,
      "learning_rate": 9.834502141315315e-06,
      "loss": 3.823,
      "num_input_tokens_seen": 38341504,
      "step": 28300
    },
    {
      "epoch": 0.9059919339350874,
      "grad_norm": 30.75,
      "learning_rate": 9.82669849516552e-06,
      "loss": 4.0961,
      "num_input_tokens_seen": 38348112,
      "step": 28305
    },
    {
      "epoch": 0.9061519749055759,
      "grad_norm": 37.75,
      "learning_rate": 9.818897188929493e-06,
      "loss": 4.4993,
      "num_input_tokens_seen": 38354848,
      "step": 28310
    },
    {
      "epoch": 0.9063120158760642,
      "grad_norm": 50.75,
      "learning_rate": 9.811098223810309e-06,
      "loss": 5.0217,
      "num_input_tokens_seen": 38361872,
      "step": 28315
    },
    {
      "epoch": 0.9064720568465527,
      "grad_norm": 30.5,
      "learning_rate": 9.803301601010641e-06,
      "loss": 4.6988,
      "num_input_tokens_seen": 38368432,
      "step": 28320
    },
    {
      "epoch": 0.9066320978170411,
      "grad_norm": 34.25,
      "learning_rate": 9.795507321732853e-06,
      "loss": 4.065,
      "num_input_tokens_seen": 38375024,
      "step": 28325
    },
    {
      "epoch": 0.9067921387875296,
      "grad_norm": 38.5,
      "learning_rate": 9.787715387178898e-06,
      "loss": 4.5093,
      "num_input_tokens_seen": 38381632,
      "step": 28330
    },
    {
      "epoch": 0.906952179758018,
      "grad_norm": 34.25,
      "learning_rate": 9.779925798550399e-06,
      "loss": 5.1792,
      "num_input_tokens_seen": 38388352,
      "step": 28335
    },
    {
      "epoch": 0.9071122207285065,
      "grad_norm": 40.75,
      "learning_rate": 9.772138557048619e-06,
      "loss": 3.9623,
      "num_input_tokens_seen": 38395024,
      "step": 28340
    },
    {
      "epoch": 0.907272261698995,
      "grad_norm": 32.5,
      "learning_rate": 9.764353663874426e-06,
      "loss": 3.712,
      "num_input_tokens_seen": 38401856,
      "step": 28345
    },
    {
      "epoch": 0.9074323026694834,
      "grad_norm": 60.25,
      "learning_rate": 9.756571120228375e-06,
      "loss": 4.7427,
      "num_input_tokens_seen": 38408848,
      "step": 28350
    },
    {
      "epoch": 0.9075923436399719,
      "grad_norm": 53.25,
      "learning_rate": 9.748790927310605e-06,
      "loss": 4.6674,
      "num_input_tokens_seen": 38415440,
      "step": 28355
    },
    {
      "epoch": 0.9077523846104603,
      "grad_norm": 31.25,
      "learning_rate": 9.741013086320946e-06,
      "loss": 3.7788,
      "num_input_tokens_seen": 38422256,
      "step": 28360
    },
    {
      "epoch": 0.9079124255809488,
      "grad_norm": 33.75,
      "learning_rate": 9.733237598458821e-06,
      "loss": 4.2796,
      "num_input_tokens_seen": 38429200,
      "step": 28365
    },
    {
      "epoch": 0.9080724665514371,
      "grad_norm": 23.125,
      "learning_rate": 9.725464464923308e-06,
      "loss": 4.4789,
      "num_input_tokens_seen": 38435808,
      "step": 28370
    },
    {
      "epoch": 0.9082325075219256,
      "grad_norm": 22.75,
      "learning_rate": 9.717693686913123e-06,
      "loss": 4.4193,
      "num_input_tokens_seen": 38442416,
      "step": 28375
    },
    {
      "epoch": 0.908392548492414,
      "grad_norm": 40.25,
      "learning_rate": 9.709925265626632e-06,
      "loss": 5.7725,
      "num_input_tokens_seen": 38449072,
      "step": 28380
    },
    {
      "epoch": 0.9085525894629025,
      "grad_norm": 31.75,
      "learning_rate": 9.702159202261801e-06,
      "loss": 3.9703,
      "num_input_tokens_seen": 38455568,
      "step": 28385
    },
    {
      "epoch": 0.9087126304333909,
      "grad_norm": 20.875,
      "learning_rate": 9.694395498016268e-06,
      "loss": 4.9803,
      "num_input_tokens_seen": 38462320,
      "step": 28390
    },
    {
      "epoch": 0.9088726714038794,
      "grad_norm": 25.625,
      "learning_rate": 9.686634154087298e-06,
      "loss": 3.9211,
      "num_input_tokens_seen": 38468784,
      "step": 28395
    },
    {
      "epoch": 0.9090327123743679,
      "grad_norm": 39.75,
      "learning_rate": 9.678875171671776e-06,
      "loss": 4.6624,
      "num_input_tokens_seen": 38475552,
      "step": 28400
    },
    {
      "epoch": 0.9090327123743679,
      "eval_loss": 4.510716438293457,
      "eval_runtime": 489.6854,
      "eval_samples_per_second": 28.357,
      "eval_steps_per_second": 14.178,
      "num_input_tokens_seen": 38475552,
      "step": 28400
    },
    {
      "epoch": 0.9091927533448563,
      "grad_norm": 24.75,
      "learning_rate": 9.671118551966246e-06,
      "loss": 5.2081,
      "num_input_tokens_seen": 38482112,
      "step": 28405
    },
    {
      "epoch": 0.9093527943153448,
      "grad_norm": 24.25,
      "learning_rate": 9.66336429616686e-06,
      "loss": 5.0717,
      "num_input_tokens_seen": 38488592,
      "step": 28410
    },
    {
      "epoch": 0.9095128352858332,
      "grad_norm": 23.25,
      "learning_rate": 9.655612405469436e-06,
      "loss": 4.0976,
      "num_input_tokens_seen": 38495456,
      "step": 28415
    },
    {
      "epoch": 0.9096728762563216,
      "grad_norm": 29.875,
      "learning_rate": 9.647862881069413e-06,
      "loss": 4.0526,
      "num_input_tokens_seen": 38502352,
      "step": 28420
    },
    {
      "epoch": 0.90983291722681,
      "grad_norm": 26.25,
      "learning_rate": 9.640115724161855e-06,
      "loss": 3.8312,
      "num_input_tokens_seen": 38509760,
      "step": 28425
    },
    {
      "epoch": 0.9099929581972985,
      "grad_norm": 27.375,
      "learning_rate": 9.632370935941483e-06,
      "loss": 4.098,
      "num_input_tokens_seen": 38516912,
      "step": 28430
    },
    {
      "epoch": 0.9101529991677869,
      "grad_norm": 27.125,
      "learning_rate": 9.624628517602634e-06,
      "loss": 4.255,
      "num_input_tokens_seen": 38523792,
      "step": 28435
    },
    {
      "epoch": 0.9103130401382754,
      "grad_norm": 34.75,
      "learning_rate": 9.61688847033928e-06,
      "loss": 4.8939,
      "num_input_tokens_seen": 38530432,
      "step": 28440
    },
    {
      "epoch": 0.9104730811087638,
      "grad_norm": 31.0,
      "learning_rate": 9.609150795345051e-06,
      "loss": 4.6346,
      "num_input_tokens_seen": 38537280,
      "step": 28445
    },
    {
      "epoch": 0.9106331220792523,
      "grad_norm": 31.75,
      "learning_rate": 9.601415493813171e-06,
      "loss": 4.0058,
      "num_input_tokens_seen": 38544032,
      "step": 28450
    },
    {
      "epoch": 0.9107931630497408,
      "grad_norm": 21.0,
      "learning_rate": 9.593682566936533e-06,
      "loss": 3.5421,
      "num_input_tokens_seen": 38550768,
      "step": 28455
    },
    {
      "epoch": 0.9109532040202292,
      "grad_norm": 25.25,
      "learning_rate": 9.58595201590766e-06,
      "loss": 4.248,
      "num_input_tokens_seen": 38557904,
      "step": 28460
    },
    {
      "epoch": 0.9111132449907177,
      "grad_norm": 22.5,
      "learning_rate": 9.578223841918681e-06,
      "loss": 4.6019,
      "num_input_tokens_seen": 38564192,
      "step": 28465
    },
    {
      "epoch": 0.911273285961206,
      "grad_norm": 25.375,
      "learning_rate": 9.570498046161389e-06,
      "loss": 3.6228,
      "num_input_tokens_seen": 38570592,
      "step": 28470
    },
    {
      "epoch": 0.9114333269316945,
      "grad_norm": 41.25,
      "learning_rate": 9.562774629827206e-06,
      "loss": 3.5009,
      "num_input_tokens_seen": 38577392,
      "step": 28475
    },
    {
      "epoch": 0.9115933679021829,
      "grad_norm": 16.125,
      "learning_rate": 9.555053594107163e-06,
      "loss": 4.1582,
      "num_input_tokens_seen": 38584160,
      "step": 28480
    },
    {
      "epoch": 0.9117534088726714,
      "grad_norm": 22.625,
      "learning_rate": 9.547334940191957e-06,
      "loss": 5.221,
      "num_input_tokens_seen": 38590544,
      "step": 28485
    },
    {
      "epoch": 0.9119134498431598,
      "grad_norm": 23.125,
      "learning_rate": 9.539618669271886e-06,
      "loss": 4.4677,
      "num_input_tokens_seen": 38597584,
      "step": 28490
    },
    {
      "epoch": 0.9120734908136483,
      "grad_norm": 41.5,
      "learning_rate": 9.531904782536904e-06,
      "loss": 6.0995,
      "num_input_tokens_seen": 38604176,
      "step": 28495
    },
    {
      "epoch": 0.9122335317841367,
      "grad_norm": 26.875,
      "learning_rate": 9.524193281176597e-06,
      "loss": 3.3748,
      "num_input_tokens_seen": 38611104,
      "step": 28500
    },
    {
      "epoch": 0.9123935727546252,
      "grad_norm": 25.625,
      "learning_rate": 9.516484166380165e-06,
      "loss": 3.6305,
      "num_input_tokens_seen": 38617728,
      "step": 28505
    },
    {
      "epoch": 0.9125536137251137,
      "grad_norm": 35.0,
      "learning_rate": 9.508777439336447e-06,
      "loss": 4.3139,
      "num_input_tokens_seen": 38624448,
      "step": 28510
    },
    {
      "epoch": 0.9127136546956021,
      "grad_norm": 25.875,
      "learning_rate": 9.50107310123393e-06,
      "loss": 4.0604,
      "num_input_tokens_seen": 38631280,
      "step": 28515
    },
    {
      "epoch": 0.9128736956660906,
      "grad_norm": 22.125,
      "learning_rate": 9.493371153260702e-06,
      "loss": 4.7052,
      "num_input_tokens_seen": 38637696,
      "step": 28520
    },
    {
      "epoch": 0.9130337366365789,
      "grad_norm": 19.0,
      "learning_rate": 9.485671596604523e-06,
      "loss": 4.4218,
      "num_input_tokens_seen": 38644896,
      "step": 28525
    },
    {
      "epoch": 0.9131937776070674,
      "grad_norm": 25.875,
      "learning_rate": 9.477974432452738e-06,
      "loss": 4.5508,
      "num_input_tokens_seen": 38651584,
      "step": 28530
    },
    {
      "epoch": 0.9133538185775558,
      "grad_norm": 18.5,
      "learning_rate": 9.470279661992356e-06,
      "loss": 3.9433,
      "num_input_tokens_seen": 38658592,
      "step": 28535
    },
    {
      "epoch": 0.9135138595480443,
      "grad_norm": 20.5,
      "learning_rate": 9.462587286410021e-06,
      "loss": 4.0128,
      "num_input_tokens_seen": 38665312,
      "step": 28540
    },
    {
      "epoch": 0.9136739005185327,
      "grad_norm": 39.25,
      "learning_rate": 9.454897306891972e-06,
      "loss": 3.6863,
      "num_input_tokens_seen": 38672176,
      "step": 28545
    },
    {
      "epoch": 0.9138339414890212,
      "grad_norm": 25.5,
      "learning_rate": 9.44720972462411e-06,
      "loss": 3.8668,
      "num_input_tokens_seen": 38678624,
      "step": 28550
    },
    {
      "epoch": 0.9139939824595096,
      "grad_norm": 35.75,
      "learning_rate": 9.439524540791964e-06,
      "loss": 4.3453,
      "num_input_tokens_seen": 38685248,
      "step": 28555
    },
    {
      "epoch": 0.9141540234299981,
      "grad_norm": 31.125,
      "learning_rate": 9.431841756580673e-06,
      "loss": 4.4251,
      "num_input_tokens_seen": 38692240,
      "step": 28560
    },
    {
      "epoch": 0.9143140644004866,
      "grad_norm": 42.0,
      "learning_rate": 9.42416137317503e-06,
      "loss": 5.4089,
      "num_input_tokens_seen": 38699664,
      "step": 28565
    },
    {
      "epoch": 0.914474105370975,
      "grad_norm": 19.5,
      "learning_rate": 9.416483391759437e-06,
      "loss": 3.8763,
      "num_input_tokens_seen": 38706496,
      "step": 28570
    },
    {
      "epoch": 0.9146341463414634,
      "grad_norm": 29.625,
      "learning_rate": 9.408807813517945e-06,
      "loss": 4.7288,
      "num_input_tokens_seen": 38713008,
      "step": 28575
    },
    {
      "epoch": 0.9147941873119518,
      "grad_norm": 21.125,
      "learning_rate": 9.401134639634221e-06,
      "loss": 3.9749,
      "num_input_tokens_seen": 38719648,
      "step": 28580
    },
    {
      "epoch": 0.9149542282824403,
      "grad_norm": 23.625,
      "learning_rate": 9.393463871291555e-06,
      "loss": 3.6591,
      "num_input_tokens_seen": 38726512,
      "step": 28585
    },
    {
      "epoch": 0.9151142692529287,
      "grad_norm": 24.75,
      "learning_rate": 9.385795509672881e-06,
      "loss": 3.9984,
      "num_input_tokens_seen": 38733104,
      "step": 28590
    },
    {
      "epoch": 0.9152743102234172,
      "grad_norm": 35.5,
      "learning_rate": 9.378129555960771e-06,
      "loss": 4.3597,
      "num_input_tokens_seen": 38739600,
      "step": 28595
    },
    {
      "epoch": 0.9154343511939056,
      "grad_norm": 14.4375,
      "learning_rate": 9.370466011337392e-06,
      "loss": 3.0155,
      "num_input_tokens_seen": 38746560,
      "step": 28600
    },
    {
      "epoch": 0.9154343511939056,
      "eval_loss": 4.496036529541016,
      "eval_runtime": 489.2185,
      "eval_samples_per_second": 28.384,
      "eval_steps_per_second": 14.192,
      "num_input_tokens_seen": 38746560,
      "step": 28600
    },
    {
      "epoch": 0.9155943921643941,
      "grad_norm": 23.125,
      "learning_rate": 9.362804876984573e-06,
      "loss": 3.5356,
      "num_input_tokens_seen": 38753344,
      "step": 28605
    },
    {
      "epoch": 0.9157544331348826,
      "grad_norm": 23.5,
      "learning_rate": 9.355146154083747e-06,
      "loss": 4.9472,
      "num_input_tokens_seen": 38759728,
      "step": 28610
    },
    {
      "epoch": 0.915914474105371,
      "grad_norm": 23.625,
      "learning_rate": 9.347489843815987e-06,
      "loss": 4.436,
      "num_input_tokens_seen": 38766800,
      "step": 28615
    },
    {
      "epoch": 0.9160745150758595,
      "grad_norm": 26.875,
      "learning_rate": 9.339835947362002e-06,
      "loss": 4.7555,
      "num_input_tokens_seen": 38773632,
      "step": 28620
    },
    {
      "epoch": 0.9162345560463478,
      "grad_norm": 23.375,
      "learning_rate": 9.332184465902105e-06,
      "loss": 3.6746,
      "num_input_tokens_seen": 38780192,
      "step": 28625
    },
    {
      "epoch": 0.9163945970168363,
      "grad_norm": 18.0,
      "learning_rate": 9.324535400616266e-06,
      "loss": 4.5265,
      "num_input_tokens_seen": 38787200,
      "step": 28630
    },
    {
      "epoch": 0.9165546379873247,
      "grad_norm": 36.25,
      "learning_rate": 9.31688875268405e-06,
      "loss": 4.5723,
      "num_input_tokens_seen": 38794176,
      "step": 28635
    },
    {
      "epoch": 0.9167146789578132,
      "grad_norm": 22.5,
      "learning_rate": 9.309244523284674e-06,
      "loss": 4.2284,
      "num_input_tokens_seen": 38801504,
      "step": 28640
    },
    {
      "epoch": 0.9168747199283016,
      "grad_norm": 38.0,
      "learning_rate": 9.301602713596982e-06,
      "loss": 4.6078,
      "num_input_tokens_seen": 38808736,
      "step": 28645
    },
    {
      "epoch": 0.9170347608987901,
      "grad_norm": 25.0,
      "learning_rate": 9.293963324799432e-06,
      "loss": 4.2522,
      "num_input_tokens_seen": 38815376,
      "step": 28650
    },
    {
      "epoch": 0.9171948018692785,
      "grad_norm": 17.5,
      "learning_rate": 9.286326358070104e-06,
      "loss": 5.0885,
      "num_input_tokens_seen": 38821904,
      "step": 28655
    },
    {
      "epoch": 0.917354842839767,
      "grad_norm": 28.25,
      "learning_rate": 9.278691814586729e-06,
      "loss": 3.7627,
      "num_input_tokens_seen": 38828672,
      "step": 28660
    },
    {
      "epoch": 0.9175148838102555,
      "grad_norm": 48.75,
      "learning_rate": 9.271059695526635e-06,
      "loss": 3.8519,
      "num_input_tokens_seen": 38834944,
      "step": 28665
    },
    {
      "epoch": 0.9176749247807439,
      "grad_norm": 31.375,
      "learning_rate": 9.263430002066805e-06,
      "loss": 4.3127,
      "num_input_tokens_seen": 38841440,
      "step": 28670
    },
    {
      "epoch": 0.9178349657512324,
      "grad_norm": 41.25,
      "learning_rate": 9.25580273538382e-06,
      "loss": 3.9681,
      "num_input_tokens_seen": 38848112,
      "step": 28675
    },
    {
      "epoch": 0.9179950067217207,
      "grad_norm": 31.75,
      "learning_rate": 9.248177896653907e-06,
      "loss": 4.9414,
      "num_input_tokens_seen": 38854704,
      "step": 28680
    },
    {
      "epoch": 0.9181550476922092,
      "grad_norm": 57.5,
      "learning_rate": 9.240555487052918e-06,
      "loss": 5.1177,
      "num_input_tokens_seen": 38861856,
      "step": 28685
    },
    {
      "epoch": 0.9183150886626976,
      "grad_norm": 74.0,
      "learning_rate": 9.232935507756313e-06,
      "loss": 4.6412,
      "num_input_tokens_seen": 38868176,
      "step": 28690
    },
    {
      "epoch": 0.9184751296331861,
      "grad_norm": 16.75,
      "learning_rate": 9.225317959939193e-06,
      "loss": 3.8834,
      "num_input_tokens_seen": 38874928,
      "step": 28695
    },
    {
      "epoch": 0.9186351706036745,
      "grad_norm": 40.25,
      "learning_rate": 9.217702844776287e-06,
      "loss": 4.9121,
      "num_input_tokens_seen": 38881456,
      "step": 28700
    },
    {
      "epoch": 0.918795211574163,
      "grad_norm": 39.25,
      "learning_rate": 9.210090163441929e-06,
      "loss": 4.3106,
      "num_input_tokens_seen": 38888048,
      "step": 28705
    },
    {
      "epoch": 0.9189552525446514,
      "grad_norm": 23.625,
      "learning_rate": 9.202479917110105e-06,
      "loss": 4.6738,
      "num_input_tokens_seen": 38894704,
      "step": 28710
    },
    {
      "epoch": 0.9191152935151399,
      "grad_norm": 27.25,
      "learning_rate": 9.194872106954392e-06,
      "loss": 4.5559,
      "num_input_tokens_seen": 38901728,
      "step": 28715
    },
    {
      "epoch": 0.9192753344856284,
      "grad_norm": 44.0,
      "learning_rate": 9.187266734148029e-06,
      "loss": 4.6598,
      "num_input_tokens_seen": 38908528,
      "step": 28720
    },
    {
      "epoch": 0.9194353754561168,
      "grad_norm": 30.75,
      "learning_rate": 9.179663799863849e-06,
      "loss": 4.6491,
      "num_input_tokens_seen": 38915552,
      "step": 28725
    },
    {
      "epoch": 0.9195954164266052,
      "grad_norm": 66.5,
      "learning_rate": 9.172063305274317e-06,
      "loss": 3.6195,
      "num_input_tokens_seen": 38922032,
      "step": 28730
    },
    {
      "epoch": 0.9197554573970936,
      "grad_norm": 26.0,
      "learning_rate": 9.164465251551527e-06,
      "loss": 4.5599,
      "num_input_tokens_seen": 38928336,
      "step": 28735
    },
    {
      "epoch": 0.9199154983675821,
      "grad_norm": 26.125,
      "learning_rate": 9.156869639867205e-06,
      "loss": 4.1801,
      "num_input_tokens_seen": 38934720,
      "step": 28740
    },
    {
      "epoch": 0.9200755393380705,
      "grad_norm": 34.25,
      "learning_rate": 9.149276471392677e-06,
      "loss": 4.6238,
      "num_input_tokens_seen": 38941312,
      "step": 28745
    },
    {
      "epoch": 0.920235580308559,
      "grad_norm": 15.3125,
      "learning_rate": 9.141685747298914e-06,
      "loss": 4.9019,
      "num_input_tokens_seen": 38948064,
      "step": 28750
    },
    {
      "epoch": 0.9203956212790474,
      "grad_norm": 23.75,
      "learning_rate": 9.13409746875649e-06,
      "loss": 4.8485,
      "num_input_tokens_seen": 38954224,
      "step": 28755
    },
    {
      "epoch": 0.9205556622495359,
      "grad_norm": 38.75,
      "learning_rate": 9.12651163693562e-06,
      "loss": 4.395,
      "num_input_tokens_seen": 38961120,
      "step": 28760
    },
    {
      "epoch": 0.9207157032200243,
      "grad_norm": 26.25,
      "learning_rate": 9.11892825300614e-06,
      "loss": 4.7748,
      "num_input_tokens_seen": 38967712,
      "step": 28765
    },
    {
      "epoch": 0.9208757441905128,
      "grad_norm": 34.25,
      "learning_rate": 9.111347318137491e-06,
      "loss": 4.678,
      "num_input_tokens_seen": 38974448,
      "step": 28770
    },
    {
      "epoch": 0.9210357851610013,
      "grad_norm": 31.25,
      "learning_rate": 9.103768833498755e-06,
      "loss": 4.2821,
      "num_input_tokens_seen": 38981488,
      "step": 28775
    },
    {
      "epoch": 0.9211958261314896,
      "grad_norm": 39.25,
      "learning_rate": 9.096192800258639e-06,
      "loss": 4.7958,
      "num_input_tokens_seen": 38988752,
      "step": 28780
    },
    {
      "epoch": 0.9213558671019781,
      "grad_norm": 46.0,
      "learning_rate": 9.088619219585443e-06,
      "loss": 4.2846,
      "num_input_tokens_seen": 38995472,
      "step": 28785
    },
    {
      "epoch": 0.9215159080724665,
      "grad_norm": 16.125,
      "learning_rate": 9.081048092647127e-06,
      "loss": 3.9302,
      "num_input_tokens_seen": 39002528,
      "step": 28790
    },
    {
      "epoch": 0.921675949042955,
      "grad_norm": 31.25,
      "learning_rate": 9.073479420611245e-06,
      "loss": 4.3011,
      "num_input_tokens_seen": 39009600,
      "step": 28795
    },
    {
      "epoch": 0.9218359900134434,
      "grad_norm": 44.0,
      "learning_rate": 9.065913204644974e-06,
      "loss": 5.2439,
      "num_input_tokens_seen": 39016288,
      "step": 28800
    },
    {
      "epoch": 0.9218359900134434,
      "eval_loss": 4.484742164611816,
      "eval_runtime": 489.341,
      "eval_samples_per_second": 28.377,
      "eval_steps_per_second": 14.188,
      "num_input_tokens_seen": 39016288,
      "step": 28800
    },
    {
      "epoch": 0.9219960309839319,
      "grad_norm": 18.25,
      "learning_rate": 9.058349445915135e-06,
      "loss": 4.6472,
      "num_input_tokens_seen": 39023328,
      "step": 28805
    },
    {
      "epoch": 0.9221560719544203,
      "grad_norm": 28.625,
      "learning_rate": 9.050788145588138e-06,
      "loss": 4.4969,
      "num_input_tokens_seen": 39030160,
      "step": 28810
    },
    {
      "epoch": 0.9223161129249088,
      "grad_norm": 31.875,
      "learning_rate": 9.043229304830039e-06,
      "loss": 5.0267,
      "num_input_tokens_seen": 39036992,
      "step": 28815
    },
    {
      "epoch": 0.9224761538953972,
      "grad_norm": 42.25,
      "learning_rate": 9.035672924806515e-06,
      "loss": 3.9902,
      "num_input_tokens_seen": 39043616,
      "step": 28820
    },
    {
      "epoch": 0.9226361948658857,
      "grad_norm": 24.5,
      "learning_rate": 9.028119006682839e-06,
      "loss": 4.4812,
      "num_input_tokens_seen": 39050656,
      "step": 28825
    },
    {
      "epoch": 0.9227962358363742,
      "grad_norm": 27.125,
      "learning_rate": 9.020567551623935e-06,
      "loss": 3.9356,
      "num_input_tokens_seen": 39057456,
      "step": 28830
    },
    {
      "epoch": 0.9229562768068625,
      "grad_norm": 38.75,
      "learning_rate": 9.013018560794318e-06,
      "loss": 4.2925,
      "num_input_tokens_seen": 39064288,
      "step": 28835
    },
    {
      "epoch": 0.923116317777351,
      "grad_norm": 34.75,
      "learning_rate": 9.005472035358139e-06,
      "loss": 4.2681,
      "num_input_tokens_seen": 39071328,
      "step": 28840
    },
    {
      "epoch": 0.9232763587478394,
      "grad_norm": 28.0,
      "learning_rate": 8.997927976479185e-06,
      "loss": 3.9319,
      "num_input_tokens_seen": 39077904,
      "step": 28845
    },
    {
      "epoch": 0.9234363997183279,
      "grad_norm": 20.5,
      "learning_rate": 8.99038638532082e-06,
      "loss": 4.247,
      "num_input_tokens_seen": 39084656,
      "step": 28850
    },
    {
      "epoch": 0.9235964406888163,
      "grad_norm": 47.0,
      "learning_rate": 8.982847263046065e-06,
      "loss": 4.4511,
      "num_input_tokens_seen": 39091408,
      "step": 28855
    },
    {
      "epoch": 0.9237564816593048,
      "grad_norm": 15.375,
      "learning_rate": 8.975310610817555e-06,
      "loss": 4.2614,
      "num_input_tokens_seen": 39097872,
      "step": 28860
    },
    {
      "epoch": 0.9239165226297932,
      "grad_norm": 34.5,
      "learning_rate": 8.967776429797528e-06,
      "loss": 4.5316,
      "num_input_tokens_seen": 39105232,
      "step": 28865
    },
    {
      "epoch": 0.9240765636002817,
      "grad_norm": 27.75,
      "learning_rate": 8.960244721147842e-06,
      "loss": 3.4277,
      "num_input_tokens_seen": 39111792,
      "step": 28870
    },
    {
      "epoch": 0.9242366045707702,
      "grad_norm": 28.75,
      "learning_rate": 8.952715486029995e-06,
      "loss": 4.365,
      "num_input_tokens_seen": 39118736,
      "step": 28875
    },
    {
      "epoch": 0.9243966455412586,
      "grad_norm": 40.0,
      "learning_rate": 8.945188725605075e-06,
      "loss": 4.8887,
      "num_input_tokens_seen": 39125216,
      "step": 28880
    },
    {
      "epoch": 0.924556686511747,
      "grad_norm": 28.375,
      "learning_rate": 8.937664441033817e-06,
      "loss": 5.0542,
      "num_input_tokens_seen": 39131952,
      "step": 28885
    },
    {
      "epoch": 0.9247167274822354,
      "grad_norm": 31.5,
      "learning_rate": 8.930142633476549e-06,
      "loss": 4.0623,
      "num_input_tokens_seen": 39138832,
      "step": 28890
    },
    {
      "epoch": 0.9248767684527239,
      "grad_norm": 47.5,
      "learning_rate": 8.92262330409323e-06,
      "loss": 4.2702,
      "num_input_tokens_seen": 39145792,
      "step": 28895
    },
    {
      "epoch": 0.9250368094232123,
      "grad_norm": 43.25,
      "learning_rate": 8.915106454043448e-06,
      "loss": 4.0328,
      "num_input_tokens_seen": 39152880,
      "step": 28900
    },
    {
      "epoch": 0.9251968503937008,
      "grad_norm": 19.875,
      "learning_rate": 8.90759208448638e-06,
      "loss": 3.8828,
      "num_input_tokens_seen": 39159760,
      "step": 28905
    },
    {
      "epoch": 0.9253568913641892,
      "grad_norm": 53.0,
      "learning_rate": 8.900080196580848e-06,
      "loss": 4.112,
      "num_input_tokens_seen": 39166752,
      "step": 28910
    },
    {
      "epoch": 0.9255169323346777,
      "grad_norm": 43.0,
      "learning_rate": 8.892570791485267e-06,
      "loss": 4.9824,
      "num_input_tokens_seen": 39173760,
      "step": 28915
    },
    {
      "epoch": 0.9256769733051661,
      "grad_norm": 36.75,
      "learning_rate": 8.885063870357688e-06,
      "loss": 4.1299,
      "num_input_tokens_seen": 39180448,
      "step": 28920
    },
    {
      "epoch": 0.9258370142756546,
      "grad_norm": 24.125,
      "learning_rate": 8.87755943435578e-06,
      "loss": 3.5602,
      "num_input_tokens_seen": 39187392,
      "step": 28925
    },
    {
      "epoch": 0.9259970552461431,
      "grad_norm": 32.75,
      "learning_rate": 8.87005748463681e-06,
      "loss": 4.3912,
      "num_input_tokens_seen": 39194208,
      "step": 28930
    },
    {
      "epoch": 0.9261570962166314,
      "grad_norm": 36.0,
      "learning_rate": 8.862558022357681e-06,
      "loss": 4.9496,
      "num_input_tokens_seen": 39200720,
      "step": 28935
    },
    {
      "epoch": 0.9263171371871199,
      "grad_norm": 40.25,
      "learning_rate": 8.855061048674903e-06,
      "loss": 5.0461,
      "num_input_tokens_seen": 39207552,
      "step": 28940
    },
    {
      "epoch": 0.9264771781576083,
      "grad_norm": 32.5,
      "learning_rate": 8.847566564744595e-06,
      "loss": 4.5986,
      "num_input_tokens_seen": 39214272,
      "step": 28945
    },
    {
      "epoch": 0.9266372191280968,
      "grad_norm": 26.25,
      "learning_rate": 8.840074571722512e-06,
      "loss": 5.4304,
      "num_input_tokens_seen": 39221056,
      "step": 28950
    },
    {
      "epoch": 0.9267972600985852,
      "grad_norm": 29.5,
      "learning_rate": 8.832585070764002e-06,
      "loss": 4.2658,
      "num_input_tokens_seen": 39227568,
      "step": 28955
    },
    {
      "epoch": 0.9269573010690737,
      "grad_norm": 26.125,
      "learning_rate": 8.825098063024045e-06,
      "loss": 3.6949,
      "num_input_tokens_seen": 39234256,
      "step": 28960
    },
    {
      "epoch": 0.9271173420395621,
      "grad_norm": 45.5,
      "learning_rate": 8.817613549657244e-06,
      "loss": 3.8258,
      "num_input_tokens_seen": 39240864,
      "step": 28965
    },
    {
      "epoch": 0.9272773830100506,
      "grad_norm": 31.125,
      "learning_rate": 8.810131531817783e-06,
      "loss": 5.2827,
      "num_input_tokens_seen": 39247696,
      "step": 28970
    },
    {
      "epoch": 0.927437423980539,
      "grad_norm": 43.5,
      "learning_rate": 8.802652010659496e-06,
      "loss": 4.3622,
      "num_input_tokens_seen": 39254032,
      "step": 28975
    },
    {
      "epoch": 0.9275974649510275,
      "grad_norm": 26.375,
      "learning_rate": 8.795174987335827e-06,
      "loss": 4.1461,
      "num_input_tokens_seen": 39260656,
      "step": 28980
    },
    {
      "epoch": 0.927757505921516,
      "grad_norm": 27.5,
      "learning_rate": 8.787700462999807e-06,
      "loss": 4.0571,
      "num_input_tokens_seen": 39267296,
      "step": 28985
    },
    {
      "epoch": 0.9279175468920043,
      "grad_norm": 35.0,
      "learning_rate": 8.780228438804122e-06,
      "loss": 4.7174,
      "num_input_tokens_seen": 39273840,
      "step": 28990
    },
    {
      "epoch": 0.9280775878624928,
      "grad_norm": 32.5,
      "learning_rate": 8.772758915901032e-06,
      "loss": 4.5565,
      "num_input_tokens_seen": 39281136,
      "step": 28995
    },
    {
      "epoch": 0.9282376288329812,
      "grad_norm": 17.125,
      "learning_rate": 8.765291895442443e-06,
      "loss": 5.1229,
      "num_input_tokens_seen": 39287360,
      "step": 29000
    },
    {
      "epoch": 0.9282376288329812,
      "eval_loss": 4.491867542266846,
      "eval_runtime": 489.1078,
      "eval_samples_per_second": 28.39,
      "eval_steps_per_second": 14.195,
      "num_input_tokens_seen": 39287360,
      "step": 29000
    },
    {
      "epoch": 0.9283976698034697,
      "grad_norm": 21.875,
      "learning_rate": 8.75782737857987e-06,
      "loss": 4.6098,
      "num_input_tokens_seen": 39293776,
      "step": 29005
    },
    {
      "epoch": 0.9285577107739581,
      "grad_norm": 25.0,
      "learning_rate": 8.750365366464425e-06,
      "loss": 4.4074,
      "num_input_tokens_seen": 39300688,
      "step": 29010
    },
    {
      "epoch": 0.9287177517444466,
      "grad_norm": 38.5,
      "learning_rate": 8.742905860246838e-06,
      "loss": 3.866,
      "num_input_tokens_seen": 39307216,
      "step": 29015
    },
    {
      "epoch": 0.928877792714935,
      "grad_norm": 16.75,
      "learning_rate": 8.735448861077478e-06,
      "loss": 4.3414,
      "num_input_tokens_seen": 39314064,
      "step": 29020
    },
    {
      "epoch": 0.9290378336854235,
      "grad_norm": 34.25,
      "learning_rate": 8.727994370106288e-06,
      "loss": 4.4555,
      "num_input_tokens_seen": 39321072,
      "step": 29025
    },
    {
      "epoch": 0.9291978746559119,
      "grad_norm": 40.5,
      "learning_rate": 8.720542388482861e-06,
      "loss": 5.5758,
      "num_input_tokens_seen": 39327696,
      "step": 29030
    },
    {
      "epoch": 0.9293579156264004,
      "grad_norm": 20.875,
      "learning_rate": 8.71309291735637e-06,
      "loss": 4.0919,
      "num_input_tokens_seen": 39335008,
      "step": 29035
    },
    {
      "epoch": 0.9295179565968889,
      "grad_norm": 38.25,
      "learning_rate": 8.705645957875621e-06,
      "loss": 4.7857,
      "num_input_tokens_seen": 39341776,
      "step": 29040
    },
    {
      "epoch": 0.9296779975673772,
      "grad_norm": 32.0,
      "learning_rate": 8.698201511189048e-06,
      "loss": 4.5878,
      "num_input_tokens_seen": 39348320,
      "step": 29045
    },
    {
      "epoch": 0.9298380385378657,
      "grad_norm": 31.625,
      "learning_rate": 8.690759578444649e-06,
      "loss": 4.509,
      "num_input_tokens_seen": 39355232,
      "step": 29050
    },
    {
      "epoch": 0.9299980795083541,
      "grad_norm": 29.125,
      "learning_rate": 8.68332016079008e-06,
      "loss": 4.4129,
      "num_input_tokens_seen": 39361680,
      "step": 29055
    },
    {
      "epoch": 0.9301581204788426,
      "grad_norm": 23.875,
      "learning_rate": 8.6758832593726e-06,
      "loss": 5.5321,
      "num_input_tokens_seen": 39368048,
      "step": 29060
    },
    {
      "epoch": 0.930318161449331,
      "grad_norm": 30.5,
      "learning_rate": 8.668448875339053e-06,
      "loss": 3.9134,
      "num_input_tokens_seen": 39374576,
      "step": 29065
    },
    {
      "epoch": 0.9304782024198195,
      "grad_norm": 27.125,
      "learning_rate": 8.661017009835933e-06,
      "loss": 3.6188,
      "num_input_tokens_seen": 39381264,
      "step": 29070
    },
    {
      "epoch": 0.9306382433903079,
      "grad_norm": 37.0,
      "learning_rate": 8.653587664009311e-06,
      "loss": 4.831,
      "num_input_tokens_seen": 39388368,
      "step": 29075
    },
    {
      "epoch": 0.9307982843607964,
      "grad_norm": 18.25,
      "learning_rate": 8.646160839004902e-06,
      "loss": 3.8167,
      "num_input_tokens_seen": 39395088,
      "step": 29080
    },
    {
      "epoch": 0.9309583253312849,
      "grad_norm": 37.25,
      "learning_rate": 8.638736535967998e-06,
      "loss": 4.6595,
      "num_input_tokens_seen": 39401792,
      "step": 29085
    },
    {
      "epoch": 0.9311183663017732,
      "grad_norm": 19.625,
      "learning_rate": 8.631314756043535e-06,
      "loss": 4.3501,
      "num_input_tokens_seen": 39408240,
      "step": 29090
    },
    {
      "epoch": 0.9312784072722617,
      "grad_norm": 38.75,
      "learning_rate": 8.62389550037603e-06,
      "loss": 4.0815,
      "num_input_tokens_seen": 39415376,
      "step": 29095
    },
    {
      "epoch": 0.9314384482427501,
      "grad_norm": 31.875,
      "learning_rate": 8.616478770109646e-06,
      "loss": 4.0695,
      "num_input_tokens_seen": 39421744,
      "step": 29100
    },
    {
      "epoch": 0.9315984892132386,
      "grad_norm": 31.625,
      "learning_rate": 8.609064566388111e-06,
      "loss": 4.2889,
      "num_input_tokens_seen": 39428304,
      "step": 29105
    },
    {
      "epoch": 0.931758530183727,
      "grad_norm": 35.25,
      "learning_rate": 8.601652890354815e-06,
      "loss": 3.7856,
      "num_input_tokens_seen": 39434896,
      "step": 29110
    },
    {
      "epoch": 0.9319185711542155,
      "grad_norm": 42.25,
      "learning_rate": 8.594243743152705e-06,
      "loss": 4.6898,
      "num_input_tokens_seen": 39441552,
      "step": 29115
    },
    {
      "epoch": 0.9320786121247039,
      "grad_norm": 35.0,
      "learning_rate": 8.58683712592438e-06,
      "loss": 4.8414,
      "num_input_tokens_seen": 39448496,
      "step": 29120
    },
    {
      "epoch": 0.9322386530951924,
      "grad_norm": 25.75,
      "learning_rate": 8.579433039812037e-06,
      "loss": 3.7418,
      "num_input_tokens_seen": 39455472,
      "step": 29125
    },
    {
      "epoch": 0.9323986940656808,
      "grad_norm": 29.875,
      "learning_rate": 8.572031485957466e-06,
      "loss": 4.6475,
      "num_input_tokens_seen": 39461920,
      "step": 29130
    },
    {
      "epoch": 0.9325587350361693,
      "grad_norm": 30.75,
      "learning_rate": 8.564632465502084e-06,
      "loss": 5.0485,
      "num_input_tokens_seen": 39468656,
      "step": 29135
    },
    {
      "epoch": 0.9327187760066578,
      "grad_norm": 23.375,
      "learning_rate": 8.557235979586928e-06,
      "loss": 4.3439,
      "num_input_tokens_seen": 39475456,
      "step": 29140
    },
    {
      "epoch": 0.9328788169771461,
      "grad_norm": 24.875,
      "learning_rate": 8.549842029352606e-06,
      "loss": 4.7572,
      "num_input_tokens_seen": 39482160,
      "step": 29145
    },
    {
      "epoch": 0.9330388579476346,
      "grad_norm": 15.375,
      "learning_rate": 8.542450615939376e-06,
      "loss": 4.0069,
      "num_input_tokens_seen": 39489296,
      "step": 29150
    },
    {
      "epoch": 0.933198898918123,
      "grad_norm": 32.5,
      "learning_rate": 8.535061740487082e-06,
      "loss": 4.0018,
      "num_input_tokens_seen": 39495856,
      "step": 29155
    },
    {
      "epoch": 0.9333589398886115,
      "grad_norm": 36.75,
      "learning_rate": 8.527675404135168e-06,
      "loss": 4.1271,
      "num_input_tokens_seen": 39502608,
      "step": 29160
    },
    {
      "epoch": 0.9335189808590999,
      "grad_norm": 35.75,
      "learning_rate": 8.520291608022724e-06,
      "loss": 5.4539,
      "num_input_tokens_seen": 39509680,
      "step": 29165
    },
    {
      "epoch": 0.9336790218295884,
      "grad_norm": 35.5,
      "learning_rate": 8.512910353288398e-06,
      "loss": 4.8796,
      "num_input_tokens_seen": 39516624,
      "step": 29170
    },
    {
      "epoch": 0.9338390628000768,
      "grad_norm": 25.25,
      "learning_rate": 8.505531641070486e-06,
      "loss": 4.1573,
      "num_input_tokens_seen": 39523472,
      "step": 29175
    },
    {
      "epoch": 0.9339991037705653,
      "grad_norm": 29.25,
      "learning_rate": 8.498155472506885e-06,
      "loss": 3.349,
      "num_input_tokens_seen": 39530832,
      "step": 29180
    },
    {
      "epoch": 0.9341591447410537,
      "grad_norm": 34.5,
      "learning_rate": 8.49078184873508e-06,
      "loss": 5.109,
      "num_input_tokens_seen": 39537376,
      "step": 29185
    },
    {
      "epoch": 0.9343191857115422,
      "grad_norm": 36.75,
      "learning_rate": 8.483410770892188e-06,
      "loss": 4.005,
      "num_input_tokens_seen": 39544176,
      "step": 29190
    },
    {
      "epoch": 0.9344792266820307,
      "grad_norm": 19.125,
      "learning_rate": 8.476042240114909e-06,
      "loss": 4.3072,
      "num_input_tokens_seen": 39550960,
      "step": 29195
    },
    {
      "epoch": 0.934639267652519,
      "grad_norm": 34.75,
      "learning_rate": 8.468676257539568e-06,
      "loss": 4.7314,
      "num_input_tokens_seen": 39557440,
      "step": 29200
    },
    {
      "epoch": 0.934639267652519,
      "eval_loss": 4.48408842086792,
      "eval_runtime": 489.6003,
      "eval_samples_per_second": 28.362,
      "eval_steps_per_second": 14.181,
      "num_input_tokens_seen": 39557440,
      "step": 29200
    },
    {
      "epoch": 0.9347993086230075,
      "grad_norm": 29.125,
      "learning_rate": 8.4613128243021e-06,
      "loss": 3.5855,
      "num_input_tokens_seen": 39564512,
      "step": 29205
    },
    {
      "epoch": 0.9349593495934959,
      "grad_norm": 27.75,
      "learning_rate": 8.453951941538028e-06,
      "loss": 4.4897,
      "num_input_tokens_seen": 39571456,
      "step": 29210
    },
    {
      "epoch": 0.9351193905639844,
      "grad_norm": 27.75,
      "learning_rate": 8.446593610382495e-06,
      "loss": 4.6963,
      "num_input_tokens_seen": 39578624,
      "step": 29215
    },
    {
      "epoch": 0.9352794315344728,
      "grad_norm": 64.0,
      "learning_rate": 8.439237831970259e-06,
      "loss": 4.3305,
      "num_input_tokens_seen": 39585712,
      "step": 29220
    },
    {
      "epoch": 0.9354394725049613,
      "grad_norm": 54.5,
      "learning_rate": 8.431884607435667e-06,
      "loss": 4.1238,
      "num_input_tokens_seen": 39591904,
      "step": 29225
    },
    {
      "epoch": 0.9355995134754497,
      "grad_norm": 33.25,
      "learning_rate": 8.424533937912665e-06,
      "loss": 4.1805,
      "num_input_tokens_seen": 39598800,
      "step": 29230
    },
    {
      "epoch": 0.9357595544459382,
      "grad_norm": 29.375,
      "learning_rate": 8.41718582453484e-06,
      "loss": 4.9268,
      "num_input_tokens_seen": 39605680,
      "step": 29235
    },
    {
      "epoch": 0.9359195954164266,
      "grad_norm": 35.5,
      "learning_rate": 8.409840268435346e-06,
      "loss": 4.5311,
      "num_input_tokens_seen": 39612528,
      "step": 29240
    },
    {
      "epoch": 0.936079636386915,
      "grad_norm": 42.5,
      "learning_rate": 8.402497270746976e-06,
      "loss": 3.9728,
      "num_input_tokens_seen": 39619680,
      "step": 29245
    },
    {
      "epoch": 0.9362396773574035,
      "grad_norm": 45.5,
      "learning_rate": 8.395156832602095e-06,
      "loss": 4.3797,
      "num_input_tokens_seen": 39626096,
      "step": 29250
    },
    {
      "epoch": 0.9363997183278919,
      "grad_norm": 27.5,
      "learning_rate": 8.387818955132707e-06,
      "loss": 4.0432,
      "num_input_tokens_seen": 39632656,
      "step": 29255
    },
    {
      "epoch": 0.9365597592983804,
      "grad_norm": 38.0,
      "learning_rate": 8.38048363947039e-06,
      "loss": 4.3947,
      "num_input_tokens_seen": 39639600,
      "step": 29260
    },
    {
      "epoch": 0.9367198002688688,
      "grad_norm": 33.0,
      "learning_rate": 8.373150886746351e-06,
      "loss": 5.3862,
      "num_input_tokens_seen": 39646320,
      "step": 29265
    },
    {
      "epoch": 0.9368798412393573,
      "grad_norm": 28.875,
      "learning_rate": 8.365820698091397e-06,
      "loss": 4.498,
      "num_input_tokens_seen": 39653504,
      "step": 29270
    },
    {
      "epoch": 0.9370398822098457,
      "grad_norm": 38.5,
      "learning_rate": 8.358493074635922e-06,
      "loss": 4.806,
      "num_input_tokens_seen": 39660032,
      "step": 29275
    },
    {
      "epoch": 0.9371999231803342,
      "grad_norm": 35.75,
      "learning_rate": 8.351168017509948e-06,
      "loss": 4.9199,
      "num_input_tokens_seen": 39666752,
      "step": 29280
    },
    {
      "epoch": 0.9373599641508226,
      "grad_norm": 47.25,
      "learning_rate": 8.343845527843094e-06,
      "loss": 3.9959,
      "num_input_tokens_seen": 39673344,
      "step": 29285
    },
    {
      "epoch": 0.9375200051213111,
      "grad_norm": 34.25,
      "learning_rate": 8.336525606764566e-06,
      "loss": 4.3903,
      "num_input_tokens_seen": 39679808,
      "step": 29290
    },
    {
      "epoch": 0.9376800460917994,
      "grad_norm": 39.25,
      "learning_rate": 8.329208255403204e-06,
      "loss": 4.2615,
      "num_input_tokens_seen": 39686416,
      "step": 29295
    },
    {
      "epoch": 0.9378400870622879,
      "grad_norm": 30.875,
      "learning_rate": 8.321893474887426e-06,
      "loss": 5.0511,
      "num_input_tokens_seen": 39693264,
      "step": 29300
    },
    {
      "epoch": 0.9380001280327764,
      "grad_norm": 24.125,
      "learning_rate": 8.31458126634526e-06,
      "loss": 4.1604,
      "num_input_tokens_seen": 39700112,
      "step": 29305
    },
    {
      "epoch": 0.9381601690032648,
      "grad_norm": 22.625,
      "learning_rate": 8.30727163090435e-06,
      "loss": 4.7484,
      "num_input_tokens_seen": 39706848,
      "step": 29310
    },
    {
      "epoch": 0.9383202099737533,
      "grad_norm": 22.375,
      "learning_rate": 8.29996456969192e-06,
      "loss": 4.3932,
      "num_input_tokens_seen": 39713472,
      "step": 29315
    },
    {
      "epoch": 0.9384802509442417,
      "grad_norm": 20.375,
      "learning_rate": 8.292660083834818e-06,
      "loss": 4.3314,
      "num_input_tokens_seen": 39720208,
      "step": 29320
    },
    {
      "epoch": 0.9386402919147302,
      "grad_norm": 26.25,
      "learning_rate": 8.2853581744595e-06,
      "loss": 4.1299,
      "num_input_tokens_seen": 39727200,
      "step": 29325
    },
    {
      "epoch": 0.9388003328852186,
      "grad_norm": 34.75,
      "learning_rate": 8.278058842691991e-06,
      "loss": 3.6433,
      "num_input_tokens_seen": 39734112,
      "step": 29330
    },
    {
      "epoch": 0.9389603738557071,
      "grad_norm": 27.875,
      "learning_rate": 8.27076208965796e-06,
      "loss": 4.8712,
      "num_input_tokens_seen": 39741328,
      "step": 29335
    },
    {
      "epoch": 0.9391204148261955,
      "grad_norm": 25.375,
      "learning_rate": 8.263467916482637e-06,
      "loss": 3.4883,
      "num_input_tokens_seen": 39748144,
      "step": 29340
    },
    {
      "epoch": 0.939280455796684,
      "grad_norm": 34.0,
      "learning_rate": 8.256176324290885e-06,
      "loss": 4.5481,
      "num_input_tokens_seen": 39754544,
      "step": 29345
    },
    {
      "epoch": 0.9394404967671725,
      "grad_norm": 23.75,
      "learning_rate": 8.248887314207168e-06,
      "loss": 3.8102,
      "num_input_tokens_seen": 39762496,
      "step": 29350
    },
    {
      "epoch": 0.9396005377376608,
      "grad_norm": 20.25,
      "learning_rate": 8.24160088735553e-06,
      "loss": 4.342,
      "num_input_tokens_seen": 39769536,
      "step": 29355
    },
    {
      "epoch": 0.9397605787081493,
      "grad_norm": 31.125,
      "learning_rate": 8.234317044859629e-06,
      "loss": 4.2684,
      "num_input_tokens_seen": 39776848,
      "step": 29360
    },
    {
      "epoch": 0.9399206196786377,
      "grad_norm": 40.5,
      "learning_rate": 8.227035787842744e-06,
      "loss": 4.8273,
      "num_input_tokens_seen": 39783440,
      "step": 29365
    },
    {
      "epoch": 0.9400806606491262,
      "grad_norm": 28.125,
      "learning_rate": 8.219757117427721e-06,
      "loss": 5.4211,
      "num_input_tokens_seen": 39790512,
      "step": 29370
    },
    {
      "epoch": 0.9402407016196146,
      "grad_norm": 20.125,
      "learning_rate": 8.212481034737014e-06,
      "loss": 3.4989,
      "num_input_tokens_seen": 39797488,
      "step": 29375
    },
    {
      "epoch": 0.9404007425901031,
      "grad_norm": 32.75,
      "learning_rate": 8.205207540892707e-06,
      "loss": 4.0167,
      "num_input_tokens_seen": 39804064,
      "step": 29380
    },
    {
      "epoch": 0.9405607835605915,
      "grad_norm": 34.0,
      "learning_rate": 8.197936637016442e-06,
      "loss": 4.352,
      "num_input_tokens_seen": 39810576,
      "step": 29385
    },
    {
      "epoch": 0.94072082453108,
      "grad_norm": 24.375,
      "learning_rate": 8.190668324229508e-06,
      "loss": 3.8818,
      "num_input_tokens_seen": 39817152,
      "step": 29390
    },
    {
      "epoch": 0.9408808655015684,
      "grad_norm": 30.25,
      "learning_rate": 8.183402603652749e-06,
      "loss": 3.7434,
      "num_input_tokens_seen": 39823792,
      "step": 29395
    },
    {
      "epoch": 0.9410409064720568,
      "grad_norm": 45.75,
      "learning_rate": 8.176139476406635e-06,
      "loss": 3.8531,
      "num_input_tokens_seen": 39830256,
      "step": 29400
    },
    {
      "epoch": 0.9410409064720568,
      "eval_loss": 4.476461887359619,
      "eval_runtime": 489.3226,
      "eval_samples_per_second": 28.378,
      "eval_steps_per_second": 14.189,
      "num_input_tokens_seen": 39830256,
      "step": 29400
    },
    {
      "epoch": 0.9412009474425453,
      "grad_norm": 32.5,
      "learning_rate": 8.16887894361125e-06,
      "loss": 4.3376,
      "num_input_tokens_seen": 39837200,
      "step": 29405
    },
    {
      "epoch": 0.9413609884130337,
      "grad_norm": 19.625,
      "learning_rate": 8.161621006386233e-06,
      "loss": 4.2038,
      "num_input_tokens_seen": 39844304,
      "step": 29410
    },
    {
      "epoch": 0.9415210293835222,
      "grad_norm": 35.75,
      "learning_rate": 8.154365665850869e-06,
      "loss": 3.649,
      "num_input_tokens_seen": 39851024,
      "step": 29415
    },
    {
      "epoch": 0.9416810703540106,
      "grad_norm": 22.5,
      "learning_rate": 8.147112923124005e-06,
      "loss": 4.654,
      "num_input_tokens_seen": 39857648,
      "step": 29420
    },
    {
      "epoch": 0.9418411113244991,
      "grad_norm": 21.375,
      "learning_rate": 8.13986277932412e-06,
      "loss": 4.9636,
      "num_input_tokens_seen": 39864352,
      "step": 29425
    },
    {
      "epoch": 0.9420011522949875,
      "grad_norm": 37.0,
      "learning_rate": 8.132615235569277e-06,
      "loss": 3.8716,
      "num_input_tokens_seen": 39871072,
      "step": 29430
    },
    {
      "epoch": 0.942161193265476,
      "grad_norm": 25.5,
      "learning_rate": 8.125370292977124e-06,
      "loss": 4.2138,
      "num_input_tokens_seen": 39877968,
      "step": 29435
    },
    {
      "epoch": 0.9423212342359644,
      "grad_norm": 28.25,
      "learning_rate": 8.118127952664944e-06,
      "loss": 3.8523,
      "num_input_tokens_seen": 39884592,
      "step": 29440
    },
    {
      "epoch": 0.9424812752064529,
      "grad_norm": 11.5625,
      "learning_rate": 8.110888215749574e-06,
      "loss": 4.3865,
      "num_input_tokens_seen": 39891408,
      "step": 29445
    },
    {
      "epoch": 0.9426413161769412,
      "grad_norm": 45.75,
      "learning_rate": 8.10365108334749e-06,
      "loss": 3.8474,
      "num_input_tokens_seen": 39898224,
      "step": 29450
    },
    {
      "epoch": 0.9428013571474297,
      "grad_norm": 33.0,
      "learning_rate": 8.096416556574743e-06,
      "loss": 4.2019,
      "num_input_tokens_seen": 39905120,
      "step": 29455
    },
    {
      "epoch": 0.9429613981179182,
      "grad_norm": 33.5,
      "learning_rate": 8.08918463654698e-06,
      "loss": 4.6457,
      "num_input_tokens_seen": 39911712,
      "step": 29460
    },
    {
      "epoch": 0.9431214390884066,
      "grad_norm": 24.25,
      "learning_rate": 8.081955324379458e-06,
      "loss": 4.3451,
      "num_input_tokens_seen": 39918544,
      "step": 29465
    },
    {
      "epoch": 0.9432814800588951,
      "grad_norm": 16.625,
      "learning_rate": 8.074728621187039e-06,
      "loss": 3.4997,
      "num_input_tokens_seen": 39925312,
      "step": 29470
    },
    {
      "epoch": 0.9434415210293835,
      "grad_norm": 29.375,
      "learning_rate": 8.067504528084158e-06,
      "loss": 5.817,
      "num_input_tokens_seen": 39932256,
      "step": 29475
    },
    {
      "epoch": 0.943601561999872,
      "grad_norm": 30.5,
      "learning_rate": 8.060283046184861e-06,
      "loss": 4.3887,
      "num_input_tokens_seen": 39939552,
      "step": 29480
    },
    {
      "epoch": 0.9437616029703604,
      "grad_norm": 30.875,
      "learning_rate": 8.053064176602806e-06,
      "loss": 3.8015,
      "num_input_tokens_seen": 39946352,
      "step": 29485
    },
    {
      "epoch": 0.9439216439408489,
      "grad_norm": 23.0,
      "learning_rate": 8.045847920451216e-06,
      "loss": 4.1422,
      "num_input_tokens_seen": 39953344,
      "step": 29490
    },
    {
      "epoch": 0.9440816849113373,
      "grad_norm": 33.75,
      "learning_rate": 8.038634278842944e-06,
      "loss": 4.9648,
      "num_input_tokens_seen": 39960384,
      "step": 29495
    },
    {
      "epoch": 0.9442417258818258,
      "grad_norm": 27.625,
      "learning_rate": 8.031423252890408e-06,
      "loss": 3.8877,
      "num_input_tokens_seen": 39967632,
      "step": 29500
    },
    {
      "epoch": 0.9444017668523141,
      "grad_norm": 42.5,
      "learning_rate": 8.024214843705646e-06,
      "loss": 5.7155,
      "num_input_tokens_seen": 39974064,
      "step": 29505
    },
    {
      "epoch": 0.9445618078228026,
      "grad_norm": 55.5,
      "learning_rate": 8.017009052400295e-06,
      "loss": 5.2066,
      "num_input_tokens_seen": 39981104,
      "step": 29510
    },
    {
      "epoch": 0.9447218487932911,
      "grad_norm": 22.125,
      "learning_rate": 8.00980588008557e-06,
      "loss": 3.9089,
      "num_input_tokens_seen": 39987616,
      "step": 29515
    },
    {
      "epoch": 0.9448818897637795,
      "grad_norm": 37.5,
      "learning_rate": 8.002605327872282e-06,
      "loss": 4.0104,
      "num_input_tokens_seen": 39994416,
      "step": 29520
    },
    {
      "epoch": 0.945041930734268,
      "grad_norm": 22.25,
      "learning_rate": 7.995407396870862e-06,
      "loss": 4.6049,
      "num_input_tokens_seen": 40001136,
      "step": 29525
    },
    {
      "epoch": 0.9452019717047564,
      "grad_norm": 23.625,
      "learning_rate": 7.988212088191307e-06,
      "loss": 4.345,
      "num_input_tokens_seen": 40008496,
      "step": 29530
    },
    {
      "epoch": 0.9453620126752449,
      "grad_norm": 38.5,
      "learning_rate": 7.98101940294324e-06,
      "loss": 3.673,
      "num_input_tokens_seen": 40015136,
      "step": 29535
    },
    {
      "epoch": 0.9455220536457333,
      "grad_norm": 12.8125,
      "learning_rate": 7.973829342235847e-06,
      "loss": 4.4753,
      "num_input_tokens_seen": 40021760,
      "step": 29540
    },
    {
      "epoch": 0.9456820946162218,
      "grad_norm": 31.875,
      "learning_rate": 7.966641907177936e-06,
      "loss": 4.4082,
      "num_input_tokens_seen": 40028272,
      "step": 29545
    },
    {
      "epoch": 0.9458421355867102,
      "grad_norm": 38.0,
      "learning_rate": 7.959457098877901e-06,
      "loss": 5.0301,
      "num_input_tokens_seen": 40035152,
      "step": 29550
    },
    {
      "epoch": 0.9460021765571986,
      "grad_norm": 54.5,
      "learning_rate": 7.952274918443719e-06,
      "loss": 4.3503,
      "num_input_tokens_seen": 40041936,
      "step": 29555
    },
    {
      "epoch": 0.946162217527687,
      "grad_norm": 25.0,
      "learning_rate": 7.945095366982983e-06,
      "loss": 4.3817,
      "num_input_tokens_seen": 40049328,
      "step": 29560
    },
    {
      "epoch": 0.9463222584981755,
      "grad_norm": 22.5,
      "learning_rate": 7.937918445602871e-06,
      "loss": 3.5309,
      "num_input_tokens_seen": 40056096,
      "step": 29565
    },
    {
      "epoch": 0.946482299468664,
      "grad_norm": 45.25,
      "learning_rate": 7.930744155410145e-06,
      "loss": 4.6383,
      "num_input_tokens_seen": 40062880,
      "step": 29570
    },
    {
      "epoch": 0.9466423404391524,
      "grad_norm": 29.75,
      "learning_rate": 7.923572497511181e-06,
      "loss": 4.4344,
      "num_input_tokens_seen": 40069696,
      "step": 29575
    },
    {
      "epoch": 0.9468023814096409,
      "grad_norm": 27.625,
      "learning_rate": 7.916403473011927e-06,
      "loss": 4.0955,
      "num_input_tokens_seen": 40076112,
      "step": 29580
    },
    {
      "epoch": 0.9469624223801293,
      "grad_norm": 36.5,
      "learning_rate": 7.909237083017953e-06,
      "loss": 3.8368,
      "num_input_tokens_seen": 40082544,
      "step": 29585
    },
    {
      "epoch": 0.9471224633506178,
      "grad_norm": 27.5,
      "learning_rate": 7.902073328634389e-06,
      "loss": 4.5329,
      "num_input_tokens_seen": 40089328,
      "step": 29590
    },
    {
      "epoch": 0.9472825043211062,
      "grad_norm": 26.0,
      "learning_rate": 7.894912210965987e-06,
      "loss": 4.3501,
      "num_input_tokens_seen": 40095936,
      "step": 29595
    },
    {
      "epoch": 0.9474425452915947,
      "grad_norm": 19.0,
      "learning_rate": 7.887753731117075e-06,
      "loss": 4.4226,
      "num_input_tokens_seen": 40102464,
      "step": 29600
    },
    {
      "epoch": 0.9474425452915947,
      "eval_loss": 4.481018543243408,
      "eval_runtime": 489.6631,
      "eval_samples_per_second": 28.358,
      "eval_steps_per_second": 14.179,
      "num_input_tokens_seen": 40102464,
      "step": 29600
    },
    {
      "epoch": 0.947602586262083,
      "grad_norm": 39.25,
      "learning_rate": 7.880597890191587e-06,
      "loss": 4.3299,
      "num_input_tokens_seen": 40108848,
      "step": 29605
    },
    {
      "epoch": 0.9477626272325715,
      "grad_norm": 32.25,
      "learning_rate": 7.873444689293036e-06,
      "loss": 4.879,
      "num_input_tokens_seen": 40115600,
      "step": 29610
    },
    {
      "epoch": 0.94792266820306,
      "grad_norm": 16.875,
      "learning_rate": 7.866294129524548e-06,
      "loss": 3.5029,
      "num_input_tokens_seen": 40122784,
      "step": 29615
    },
    {
      "epoch": 0.9480827091735484,
      "grad_norm": 36.5,
      "learning_rate": 7.859146211988811e-06,
      "loss": 3.9519,
      "num_input_tokens_seen": 40129376,
      "step": 29620
    },
    {
      "epoch": 0.9482427501440369,
      "grad_norm": 27.375,
      "learning_rate": 7.852000937788134e-06,
      "loss": 3.714,
      "num_input_tokens_seen": 40135984,
      "step": 29625
    },
    {
      "epoch": 0.9484027911145253,
      "grad_norm": 49.25,
      "learning_rate": 7.844858308024416e-06,
      "loss": 4.8847,
      "num_input_tokens_seen": 40142880,
      "step": 29630
    },
    {
      "epoch": 0.9485628320850138,
      "grad_norm": 25.875,
      "learning_rate": 7.837718323799122e-06,
      "loss": 4.6442,
      "num_input_tokens_seen": 40149424,
      "step": 29635
    },
    {
      "epoch": 0.9487228730555022,
      "grad_norm": 25.75,
      "learning_rate": 7.83058098621334e-06,
      "loss": 4.425,
      "num_input_tokens_seen": 40155888,
      "step": 29640
    },
    {
      "epoch": 0.9488829140259907,
      "grad_norm": 38.25,
      "learning_rate": 7.823446296367739e-06,
      "loss": 4.6593,
      "num_input_tokens_seen": 40162544,
      "step": 29645
    },
    {
      "epoch": 0.9490429549964791,
      "grad_norm": 33.25,
      "learning_rate": 7.81631425536257e-06,
      "loss": 4.151,
      "num_input_tokens_seen": 40169328,
      "step": 29650
    },
    {
      "epoch": 0.9492029959669676,
      "grad_norm": 33.5,
      "learning_rate": 7.809184864297689e-06,
      "loss": 4.0052,
      "num_input_tokens_seen": 40175808,
      "step": 29655
    },
    {
      "epoch": 0.9493630369374559,
      "grad_norm": 33.5,
      "learning_rate": 7.802058124272532e-06,
      "loss": 4.1668,
      "num_input_tokens_seen": 40182640,
      "step": 29660
    },
    {
      "epoch": 0.9495230779079444,
      "grad_norm": 27.875,
      "learning_rate": 7.79493403638614e-06,
      "loss": 3.6026,
      "num_input_tokens_seen": 40189264,
      "step": 29665
    },
    {
      "epoch": 0.9496831188784329,
      "grad_norm": 49.5,
      "learning_rate": 7.787812601737132e-06,
      "loss": 3.4589,
      "num_input_tokens_seen": 40196224,
      "step": 29670
    },
    {
      "epoch": 0.9498431598489213,
      "grad_norm": 29.5,
      "learning_rate": 7.780693821423715e-06,
      "loss": 4.9865,
      "num_input_tokens_seen": 40203088,
      "step": 29675
    },
    {
      "epoch": 0.9500032008194098,
      "grad_norm": 26.25,
      "learning_rate": 7.773577696543705e-06,
      "loss": 3.7259,
      "num_input_tokens_seen": 40210096,
      "step": 29680
    },
    {
      "epoch": 0.9501632417898982,
      "grad_norm": 23.25,
      "learning_rate": 7.7664642281945e-06,
      "loss": 4.1325,
      "num_input_tokens_seen": 40217072,
      "step": 29685
    },
    {
      "epoch": 0.9503232827603867,
      "grad_norm": 37.5,
      "learning_rate": 7.759353417473072e-06,
      "loss": 4.361,
      "num_input_tokens_seen": 40223760,
      "step": 29690
    },
    {
      "epoch": 0.9504833237308751,
      "grad_norm": 20.375,
      "learning_rate": 7.752245265476016e-06,
      "loss": 3.6131,
      "num_input_tokens_seen": 40230288,
      "step": 29695
    },
    {
      "epoch": 0.9506433647013636,
      "grad_norm": 24.125,
      "learning_rate": 7.745139773299481e-06,
      "loss": 3.8956,
      "num_input_tokens_seen": 40237152,
      "step": 29700
    },
    {
      "epoch": 0.950803405671852,
      "grad_norm": 30.75,
      "learning_rate": 7.738036942039232e-06,
      "loss": 5.3194,
      "num_input_tokens_seen": 40243712,
      "step": 29705
    },
    {
      "epoch": 0.9509634466423404,
      "grad_norm": 25.375,
      "learning_rate": 7.73093677279062e-06,
      "loss": 4.0729,
      "num_input_tokens_seen": 40250528,
      "step": 29710
    },
    {
      "epoch": 0.9511234876128288,
      "grad_norm": 23.125,
      "learning_rate": 7.72383926664857e-06,
      "loss": 4.7906,
      "num_input_tokens_seen": 40257168,
      "step": 29715
    },
    {
      "epoch": 0.9512835285833173,
      "grad_norm": 20.625,
      "learning_rate": 7.716744424707606e-06,
      "loss": 4.1295,
      "num_input_tokens_seen": 40264064,
      "step": 29720
    },
    {
      "epoch": 0.9514435695538058,
      "grad_norm": 34.0,
      "learning_rate": 7.709652248061858e-06,
      "loss": 4.6387,
      "num_input_tokens_seen": 40270720,
      "step": 29725
    },
    {
      "epoch": 0.9516036105242942,
      "grad_norm": 27.5,
      "learning_rate": 7.702562737805017e-06,
      "loss": 5.0007,
      "num_input_tokens_seen": 40277312,
      "step": 29730
    },
    {
      "epoch": 0.9517636514947827,
      "grad_norm": 31.5,
      "learning_rate": 7.695475895030365e-06,
      "loss": 5.1939,
      "num_input_tokens_seen": 40283904,
      "step": 29735
    },
    {
      "epoch": 0.9519236924652711,
      "grad_norm": 29.125,
      "learning_rate": 7.6883917208308e-06,
      "loss": 5.3261,
      "num_input_tokens_seen": 40290656,
      "step": 29740
    },
    {
      "epoch": 0.9520837334357596,
      "grad_norm": 24.375,
      "learning_rate": 7.681310216298778e-06,
      "loss": 3.9111,
      "num_input_tokens_seen": 40297216,
      "step": 29745
    },
    {
      "epoch": 0.952243774406248,
      "grad_norm": 15.875,
      "learning_rate": 7.674231382526367e-06,
      "loss": 3.8801,
      "num_input_tokens_seen": 40303632,
      "step": 29750
    },
    {
      "epoch": 0.9524038153767365,
      "grad_norm": 34.25,
      "learning_rate": 7.667155220605198e-06,
      "loss": 3.8799,
      "num_input_tokens_seen": 40310416,
      "step": 29755
    },
    {
      "epoch": 0.9525638563472248,
      "grad_norm": 32.75,
      "learning_rate": 7.660081731626515e-06,
      "loss": 4.8822,
      "num_input_tokens_seen": 40316896,
      "step": 29760
    },
    {
      "epoch": 0.9527238973177133,
      "grad_norm": 16.75,
      "learning_rate": 7.653010916681141e-06,
      "loss": 4.6627,
      "num_input_tokens_seen": 40323344,
      "step": 29765
    },
    {
      "epoch": 0.9528839382882017,
      "grad_norm": 21.0,
      "learning_rate": 7.645942776859472e-06,
      "loss": 3.8542,
      "num_input_tokens_seen": 40330176,
      "step": 29770
    },
    {
      "epoch": 0.9530439792586902,
      "grad_norm": 20.625,
      "learning_rate": 7.63887731325152e-06,
      "loss": 4.7054,
      "num_input_tokens_seen": 40337344,
      "step": 29775
    },
    {
      "epoch": 0.9532040202291787,
      "grad_norm": 43.25,
      "learning_rate": 7.63181452694685e-06,
      "loss": 4.12,
      "num_input_tokens_seen": 40344064,
      "step": 29780
    },
    {
      "epoch": 0.9533640611996671,
      "grad_norm": 27.125,
      "learning_rate": 7.624754419034644e-06,
      "loss": 3.1071,
      "num_input_tokens_seen": 40351136,
      "step": 29785
    },
    {
      "epoch": 0.9535241021701556,
      "grad_norm": 13.5625,
      "learning_rate": 7.6176969906036645e-06,
      "loss": 3.6596,
      "num_input_tokens_seen": 40358272,
      "step": 29790
    },
    {
      "epoch": 0.953684143140644,
      "grad_norm": 14.3125,
      "learning_rate": 7.610642242742242e-06,
      "loss": 3.5088,
      "num_input_tokens_seen": 40365216,
      "step": 29795
    },
    {
      "epoch": 0.9538441841111325,
      "grad_norm": 32.75,
      "learning_rate": 7.603590176538322e-06,
      "loss": 3.617,
      "num_input_tokens_seen": 40371968,
      "step": 29800
    },
    {
      "epoch": 0.9538441841111325,
      "eval_loss": 4.489213466644287,
      "eval_runtime": 490.5075,
      "eval_samples_per_second": 28.309,
      "eval_steps_per_second": 14.155,
      "num_input_tokens_seen": 40371968,
      "step": 29800
    },
    {
      "epoch": 0.9540042250816209,
      "grad_norm": 27.375,
      "learning_rate": 7.596540793079404e-06,
      "loss": 4.0266,
      "num_input_tokens_seen": 40379104,
      "step": 29805
    },
    {
      "epoch": 0.9541642660521094,
      "grad_norm": 25.625,
      "learning_rate": 7.5894940934526125e-06,
      "loss": 3.5983,
      "num_input_tokens_seen": 40385904,
      "step": 29810
    },
    {
      "epoch": 0.9543243070225977,
      "grad_norm": 23.75,
      "learning_rate": 7.582450078744621e-06,
      "loss": 4.3593,
      "num_input_tokens_seen": 40392480,
      "step": 29815
    },
    {
      "epoch": 0.9544843479930862,
      "grad_norm": 32.5,
      "learning_rate": 7.575408750041707e-06,
      "loss": 5.0582,
      "num_input_tokens_seen": 40398976,
      "step": 29820
    },
    {
      "epoch": 0.9546443889635747,
      "grad_norm": 26.25,
      "learning_rate": 7.568370108429732e-06,
      "loss": 5.4929,
      "num_input_tokens_seen": 40405504,
      "step": 29825
    },
    {
      "epoch": 0.9548044299340631,
      "grad_norm": 36.25,
      "learning_rate": 7.561334154994154e-06,
      "loss": 5.2392,
      "num_input_tokens_seen": 40412192,
      "step": 29830
    },
    {
      "epoch": 0.9549644709045516,
      "grad_norm": 14.3125,
      "learning_rate": 7.55430089081999e-06,
      "loss": 3.906,
      "num_input_tokens_seen": 40419040,
      "step": 29835
    },
    {
      "epoch": 0.95512451187504,
      "grad_norm": 19.5,
      "learning_rate": 7.547270316991864e-06,
      "loss": 4.5544,
      "num_input_tokens_seen": 40425968,
      "step": 29840
    },
    {
      "epoch": 0.9552845528455285,
      "grad_norm": 25.125,
      "learning_rate": 7.5402424345939884e-06,
      "loss": 4.025,
      "num_input_tokens_seen": 40432688,
      "step": 29845
    },
    {
      "epoch": 0.9554445938160169,
      "grad_norm": 22.5,
      "learning_rate": 7.533217244710133e-06,
      "loss": 3.6797,
      "num_input_tokens_seen": 40439152,
      "step": 29850
    },
    {
      "epoch": 0.9556046347865054,
      "grad_norm": 27.875,
      "learning_rate": 7.52619474842369e-06,
      "loss": 4.468,
      "num_input_tokens_seen": 40445776,
      "step": 29855
    },
    {
      "epoch": 0.9557646757569938,
      "grad_norm": 39.75,
      "learning_rate": 7.519174946817597e-06,
      "loss": 4.6518,
      "num_input_tokens_seen": 40452624,
      "step": 29860
    },
    {
      "epoch": 0.9559247167274822,
      "grad_norm": 26.25,
      "learning_rate": 7.512157840974407e-06,
      "loss": 4.0131,
      "num_input_tokens_seen": 40459184,
      "step": 29865
    },
    {
      "epoch": 0.9560847576979706,
      "grad_norm": 28.25,
      "learning_rate": 7.5051434319762496e-06,
      "loss": 4.5122,
      "num_input_tokens_seen": 40465904,
      "step": 29870
    },
    {
      "epoch": 0.9562447986684591,
      "grad_norm": 22.125,
      "learning_rate": 7.498131720904822e-06,
      "loss": 3.707,
      "num_input_tokens_seen": 40472432,
      "step": 29875
    },
    {
      "epoch": 0.9564048396389476,
      "grad_norm": 22.75,
      "learning_rate": 7.491122708841433e-06,
      "loss": 4.4789,
      "num_input_tokens_seen": 40479824,
      "step": 29880
    },
    {
      "epoch": 0.956564880609436,
      "grad_norm": 32.75,
      "learning_rate": 7.4841163968669524e-06,
      "loss": 4.4249,
      "num_input_tokens_seen": 40486448,
      "step": 29885
    },
    {
      "epoch": 0.9567249215799245,
      "grad_norm": 23.75,
      "learning_rate": 7.4771127860618355e-06,
      "loss": 3.8755,
      "num_input_tokens_seen": 40493088,
      "step": 29890
    },
    {
      "epoch": 0.9568849625504129,
      "grad_norm": 35.75,
      "learning_rate": 7.470111877506139e-06,
      "loss": 4.252,
      "num_input_tokens_seen": 40500176,
      "step": 29895
    },
    {
      "epoch": 0.9570450035209014,
      "grad_norm": 34.25,
      "learning_rate": 7.463113672279479e-06,
      "loss": 4.1648,
      "num_input_tokens_seen": 40507104,
      "step": 29900
    },
    {
      "epoch": 0.9572050444913898,
      "grad_norm": 38.0,
      "learning_rate": 7.456118171461071e-06,
      "loss": 5.3788,
      "num_input_tokens_seen": 40514112,
      "step": 29905
    },
    {
      "epoch": 0.9573650854618783,
      "grad_norm": 29.875,
      "learning_rate": 7.449125376129721e-06,
      "loss": 4.7065,
      "num_input_tokens_seen": 40521088,
      "step": 29910
    },
    {
      "epoch": 0.9575251264323666,
      "grad_norm": 38.5,
      "learning_rate": 7.442135287363788e-06,
      "loss": 4.2488,
      "num_input_tokens_seen": 40528048,
      "step": 29915
    },
    {
      "epoch": 0.9576851674028551,
      "grad_norm": 30.25,
      "learning_rate": 7.435147906241247e-06,
      "loss": 3.7245,
      "num_input_tokens_seen": 40534736,
      "step": 29920
    },
    {
      "epoch": 0.9578452083733435,
      "grad_norm": 20.5,
      "learning_rate": 7.428163233839624e-06,
      "loss": 4.4216,
      "num_input_tokens_seen": 40541232,
      "step": 29925
    },
    {
      "epoch": 0.958005249343832,
      "grad_norm": 23.875,
      "learning_rate": 7.4211812712360525e-06,
      "loss": 3.6125,
      "num_input_tokens_seen": 40548480,
      "step": 29930
    },
    {
      "epoch": 0.9581652903143205,
      "grad_norm": 27.875,
      "learning_rate": 7.4142020195072464e-06,
      "loss": 5.0085,
      "num_input_tokens_seen": 40555200,
      "step": 29935
    },
    {
      "epoch": 0.9583253312848089,
      "grad_norm": 38.25,
      "learning_rate": 7.407225479729479e-06,
      "loss": 4.2115,
      "num_input_tokens_seen": 40562192,
      "step": 29940
    },
    {
      "epoch": 0.9584853722552974,
      "grad_norm": 15.5625,
      "learning_rate": 7.400251652978632e-06,
      "loss": 3.9231,
      "num_input_tokens_seen": 40569408,
      "step": 29945
    },
    {
      "epoch": 0.9586454132257858,
      "grad_norm": 20.625,
      "learning_rate": 7.393280540330147e-06,
      "loss": 4.6375,
      "num_input_tokens_seen": 40576272,
      "step": 29950
    },
    {
      "epoch": 0.9588054541962743,
      "grad_norm": 21.5,
      "learning_rate": 7.386312142859069e-06,
      "loss": 4.2438,
      "num_input_tokens_seen": 40582976,
      "step": 29955
    },
    {
      "epoch": 0.9589654951667627,
      "grad_norm": 42.0,
      "learning_rate": 7.379346461640008e-06,
      "loss": 4.9349,
      "num_input_tokens_seen": 40589696,
      "step": 29960
    },
    {
      "epoch": 0.9591255361372512,
      "grad_norm": 19.375,
      "learning_rate": 7.372383497747149e-06,
      "loss": 4.2442,
      "num_input_tokens_seen": 40596560,
      "step": 29965
    },
    {
      "epoch": 0.9592855771077395,
      "grad_norm": 37.25,
      "learning_rate": 7.3654232522542775e-06,
      "loss": 4.7084,
      "num_input_tokens_seen": 40603504,
      "step": 29970
    },
    {
      "epoch": 0.959445618078228,
      "grad_norm": 29.375,
      "learning_rate": 7.358465726234756e-06,
      "loss": 3.8158,
      "num_input_tokens_seen": 40610176,
      "step": 29975
    },
    {
      "epoch": 0.9596056590487164,
      "grad_norm": 19.375,
      "learning_rate": 7.351510920761512e-06,
      "loss": 3.2318,
      "num_input_tokens_seen": 40617040,
      "step": 29980
    },
    {
      "epoch": 0.9597657000192049,
      "grad_norm": 14.0,
      "learning_rate": 7.344558836907067e-06,
      "loss": 4.1414,
      "num_input_tokens_seen": 40623888,
      "step": 29985
    },
    {
      "epoch": 0.9599257409896934,
      "grad_norm": 29.25,
      "learning_rate": 7.3376094757435285e-06,
      "loss": 3.2729,
      "num_input_tokens_seen": 40630352,
      "step": 29990
    },
    {
      "epoch": 0.9600857819601818,
      "grad_norm": 35.5,
      "learning_rate": 7.330662838342561e-06,
      "loss": 4.8723,
      "num_input_tokens_seen": 40637120,
      "step": 29995
    },
    {
      "epoch": 0.9602458229306703,
      "grad_norm": 13.8125,
      "learning_rate": 7.323718925775438e-06,
      "loss": 3.6916,
      "num_input_tokens_seen": 40643632,
      "step": 30000
    },
    {
      "epoch": 0.9602458229306703,
      "eval_loss": 4.487293243408203,
      "eval_runtime": 490.5018,
      "eval_samples_per_second": 28.31,
      "eval_steps_per_second": 14.155,
      "num_input_tokens_seen": 40643632,
      "step": 30000
    },
    {
      "epoch": 0.9604058639011587,
      "grad_norm": 38.75,
      "learning_rate": 7.316777739112985e-06,
      "loss": 4.1552,
      "num_input_tokens_seen": 40650960,
      "step": 30005
    },
    {
      "epoch": 0.9605659048716472,
      "grad_norm": 14.3125,
      "learning_rate": 7.309839279425626e-06,
      "loss": 3.6181,
      "num_input_tokens_seen": 40657856,
      "step": 30010
    },
    {
      "epoch": 0.9607259458421356,
      "grad_norm": 27.25,
      "learning_rate": 7.302903547783366e-06,
      "loss": 4.3624,
      "num_input_tokens_seen": 40664800,
      "step": 30015
    },
    {
      "epoch": 0.960885986812624,
      "grad_norm": 29.375,
      "learning_rate": 7.2959705452557644e-06,
      "loss": 3.9608,
      "num_input_tokens_seen": 40671408,
      "step": 30020
    },
    {
      "epoch": 0.9610460277831124,
      "grad_norm": 20.75,
      "learning_rate": 7.289040272911996e-06,
      "loss": 4.7129,
      "num_input_tokens_seen": 40678048,
      "step": 30025
    },
    {
      "epoch": 0.9612060687536009,
      "grad_norm": 24.375,
      "learning_rate": 7.282112731820789e-06,
      "loss": 3.357,
      "num_input_tokens_seen": 40684672,
      "step": 30030
    },
    {
      "epoch": 0.9613661097240893,
      "grad_norm": 21.5,
      "learning_rate": 7.275187923050447e-06,
      "loss": 4.7131,
      "num_input_tokens_seen": 40691264,
      "step": 30035
    },
    {
      "epoch": 0.9615261506945778,
      "grad_norm": 18.75,
      "learning_rate": 7.268265847668879e-06,
      "loss": 3.4241,
      "num_input_tokens_seen": 40698160,
      "step": 30040
    },
    {
      "epoch": 0.9616861916650663,
      "grad_norm": 34.0,
      "learning_rate": 7.261346506743538e-06,
      "loss": 3.9448,
      "num_input_tokens_seen": 40704816,
      "step": 30045
    },
    {
      "epoch": 0.9618462326355547,
      "grad_norm": 34.25,
      "learning_rate": 7.254429901341486e-06,
      "loss": 5.6957,
      "num_input_tokens_seen": 40711552,
      "step": 30050
    },
    {
      "epoch": 0.9620062736060432,
      "grad_norm": 30.375,
      "learning_rate": 7.247516032529356e-06,
      "loss": 4.4931,
      "num_input_tokens_seen": 40717952,
      "step": 30055
    },
    {
      "epoch": 0.9621663145765316,
      "grad_norm": 19.625,
      "learning_rate": 7.240604901373338e-06,
      "loss": 3.4118,
      "num_input_tokens_seen": 40725024,
      "step": 30060
    },
    {
      "epoch": 0.9623263555470201,
      "grad_norm": 16.875,
      "learning_rate": 7.233696508939223e-06,
      "loss": 4.1202,
      "num_input_tokens_seen": 40731744,
      "step": 30065
    },
    {
      "epoch": 0.9624863965175084,
      "grad_norm": 49.0,
      "learning_rate": 7.226790856292376e-06,
      "loss": 4.1467,
      "num_input_tokens_seen": 40738208,
      "step": 30070
    },
    {
      "epoch": 0.9626464374879969,
      "grad_norm": 32.5,
      "learning_rate": 7.219887944497727e-06,
      "loss": 3.4289,
      "num_input_tokens_seen": 40745168,
      "step": 30075
    },
    {
      "epoch": 0.9628064784584853,
      "grad_norm": 46.0,
      "learning_rate": 7.2129877746198e-06,
      "loss": 3.7114,
      "num_input_tokens_seen": 40752064,
      "step": 30080
    },
    {
      "epoch": 0.9629665194289738,
      "grad_norm": 37.5,
      "learning_rate": 7.20609034772268e-06,
      "loss": 4.4057,
      "num_input_tokens_seen": 40758736,
      "step": 30085
    },
    {
      "epoch": 0.9631265603994623,
      "grad_norm": 21.625,
      "learning_rate": 7.19919566487004e-06,
      "loss": 4.5238,
      "num_input_tokens_seen": 40765392,
      "step": 30090
    },
    {
      "epoch": 0.9632866013699507,
      "grad_norm": 53.5,
      "learning_rate": 7.192303727125132e-06,
      "loss": 4.3895,
      "num_input_tokens_seen": 40772688,
      "step": 30095
    },
    {
      "epoch": 0.9634466423404392,
      "grad_norm": 42.75,
      "learning_rate": 7.185414535550777e-06,
      "loss": 4.2008,
      "num_input_tokens_seen": 40779568,
      "step": 30100
    },
    {
      "epoch": 0.9636066833109276,
      "grad_norm": 31.875,
      "learning_rate": 7.178528091209363e-06,
      "loss": 4.8157,
      "num_input_tokens_seen": 40786240,
      "step": 30105
    },
    {
      "epoch": 0.9637667242814161,
      "grad_norm": 34.25,
      "learning_rate": 7.171644395162888e-06,
      "loss": 4.2137,
      "num_input_tokens_seen": 40793648,
      "step": 30110
    },
    {
      "epoch": 0.9639267652519045,
      "grad_norm": 33.5,
      "learning_rate": 7.164763448472881e-06,
      "loss": 4.5797,
      "num_input_tokens_seen": 40800304,
      "step": 30115
    },
    {
      "epoch": 0.964086806222393,
      "grad_norm": 34.25,
      "learning_rate": 7.157885252200491e-06,
      "loss": 4.9396,
      "num_input_tokens_seen": 40806544,
      "step": 30120
    },
    {
      "epoch": 0.9642468471928813,
      "grad_norm": 37.25,
      "learning_rate": 7.151009807406403e-06,
      "loss": 4.3256,
      "num_input_tokens_seen": 40813152,
      "step": 30125
    },
    {
      "epoch": 0.9644068881633698,
      "grad_norm": 34.25,
      "learning_rate": 7.144137115150909e-06,
      "loss": 4.2525,
      "num_input_tokens_seen": 40820032,
      "step": 30130
    },
    {
      "epoch": 0.9645669291338582,
      "grad_norm": 27.125,
      "learning_rate": 7.1372671764938725e-06,
      "loss": 4.4656,
      "num_input_tokens_seen": 40827168,
      "step": 30135
    },
    {
      "epoch": 0.9647269701043467,
      "grad_norm": 27.875,
      "learning_rate": 7.130399992494705e-06,
      "loss": 4.175,
      "num_input_tokens_seen": 40833360,
      "step": 30140
    },
    {
      "epoch": 0.9648870110748352,
      "grad_norm": 31.25,
      "learning_rate": 7.123535564212419e-06,
      "loss": 3.7633,
      "num_input_tokens_seen": 40840032,
      "step": 30145
    },
    {
      "epoch": 0.9650470520453236,
      "grad_norm": 63.25,
      "learning_rate": 7.116673892705611e-06,
      "loss": 5.1046,
      "num_input_tokens_seen": 40846368,
      "step": 30150
    },
    {
      "epoch": 0.9652070930158121,
      "grad_norm": 31.25,
      "learning_rate": 7.109814979032415e-06,
      "loss": 5.4702,
      "num_input_tokens_seen": 40853296,
      "step": 30155
    },
    {
      "epoch": 0.9653671339863005,
      "grad_norm": 16.5,
      "learning_rate": 7.102958824250577e-06,
      "loss": 3.5074,
      "num_input_tokens_seen": 40859936,
      "step": 30160
    },
    {
      "epoch": 0.965527174956789,
      "grad_norm": 27.75,
      "learning_rate": 7.096105429417393e-06,
      "loss": 3.1923,
      "num_input_tokens_seen": 40866672,
      "step": 30165
    },
    {
      "epoch": 0.9656872159272774,
      "grad_norm": 50.25,
      "learning_rate": 7.0892547955897506e-06,
      "loss": 3.7865,
      "num_input_tokens_seen": 40873584,
      "step": 30170
    },
    {
      "epoch": 0.9658472568977658,
      "grad_norm": 39.5,
      "learning_rate": 7.0824069238241e-06,
      "loss": 4.2845,
      "num_input_tokens_seen": 40879808,
      "step": 30175
    },
    {
      "epoch": 0.9660072978682542,
      "grad_norm": 24.625,
      "learning_rate": 7.075561815176462e-06,
      "loss": 3.8005,
      "num_input_tokens_seen": 40886656,
      "step": 30180
    },
    {
      "epoch": 0.9661673388387427,
      "grad_norm": 19.875,
      "learning_rate": 7.068719470702445e-06,
      "loss": 4.2233,
      "num_input_tokens_seen": 40893648,
      "step": 30185
    },
    {
      "epoch": 0.9663273798092311,
      "grad_norm": 14.875,
      "learning_rate": 7.061879891457229e-06,
      "loss": 3.3303,
      "num_input_tokens_seen": 40900400,
      "step": 30190
    },
    {
      "epoch": 0.9664874207797196,
      "grad_norm": 35.75,
      "learning_rate": 7.0550430784955515e-06,
      "loss": 5.5124,
      "num_input_tokens_seen": 40907328,
      "step": 30195
    },
    {
      "epoch": 0.9666474617502081,
      "grad_norm": 42.0,
      "learning_rate": 7.048209032871752e-06,
      "loss": 3.6975,
      "num_input_tokens_seen": 40914064,
      "step": 30200
    },
    {
      "epoch": 0.9666474617502081,
      "eval_loss": 4.48817253112793,
      "eval_runtime": 490.7039,
      "eval_samples_per_second": 28.298,
      "eval_steps_per_second": 14.149,
      "num_input_tokens_seen": 40914064,
      "step": 30200
    },
    {
      "epoch": 0.9668075027206965,
      "grad_norm": 21.125,
      "learning_rate": 7.0413777556397055e-06,
      "loss": 4.3793,
      "num_input_tokens_seen": 40920576,
      "step": 30205
    },
    {
      "epoch": 0.966967543691185,
      "grad_norm": 44.0,
      "learning_rate": 7.0345492478528925e-06,
      "loss": 4.3975,
      "num_input_tokens_seen": 40927200,
      "step": 30210
    },
    {
      "epoch": 0.9671275846616734,
      "grad_norm": 32.75,
      "learning_rate": 7.02772351056436e-06,
      "loss": 5.1556,
      "num_input_tokens_seen": 40933456,
      "step": 30215
    },
    {
      "epoch": 0.9672876256321619,
      "grad_norm": 25.75,
      "learning_rate": 7.020900544826709e-06,
      "loss": 4.9334,
      "num_input_tokens_seen": 40940096,
      "step": 30220
    },
    {
      "epoch": 0.9674476666026502,
      "grad_norm": 39.0,
      "learning_rate": 7.014080351692134e-06,
      "loss": 4.5888,
      "num_input_tokens_seen": 40946720,
      "step": 30225
    },
    {
      "epoch": 0.9676077075731387,
      "grad_norm": 41.5,
      "learning_rate": 7.0072629322124024e-06,
      "loss": 4.0393,
      "num_input_tokens_seen": 40953600,
      "step": 30230
    },
    {
      "epoch": 0.9677677485436271,
      "grad_norm": 23.875,
      "learning_rate": 7.000448287438827e-06,
      "loss": 4.0344,
      "num_input_tokens_seen": 40960208,
      "step": 30235
    },
    {
      "epoch": 0.9679277895141156,
      "grad_norm": 19.5,
      "learning_rate": 6.993636418422331e-06,
      "loss": 5.343,
      "num_input_tokens_seen": 40966848,
      "step": 30240
    },
    {
      "epoch": 0.968087830484604,
      "grad_norm": 27.5,
      "learning_rate": 6.986827326213383e-06,
      "loss": 5.269,
      "num_input_tokens_seen": 40973600,
      "step": 30245
    },
    {
      "epoch": 0.9682478714550925,
      "grad_norm": 37.5,
      "learning_rate": 6.9800210118620205e-06,
      "loss": 3.4432,
      "num_input_tokens_seen": 40981120,
      "step": 30250
    },
    {
      "epoch": 0.968407912425581,
      "grad_norm": 24.625,
      "learning_rate": 6.973217476417876e-06,
      "loss": 3.8994,
      "num_input_tokens_seen": 40988096,
      "step": 30255
    },
    {
      "epoch": 0.9685679533960694,
      "grad_norm": 27.125,
      "learning_rate": 6.96641672093013e-06,
      "loss": 4.633,
      "num_input_tokens_seen": 40994704,
      "step": 30260
    },
    {
      "epoch": 0.9687279943665579,
      "grad_norm": 19.125,
      "learning_rate": 6.95961874644755e-06,
      "loss": 4.4607,
      "num_input_tokens_seen": 41001024,
      "step": 30265
    },
    {
      "epoch": 0.9688880353370463,
      "grad_norm": 35.0,
      "learning_rate": 6.952823554018476e-06,
      "loss": 4.3703,
      "num_input_tokens_seen": 41007712,
      "step": 30270
    },
    {
      "epoch": 0.9690480763075348,
      "grad_norm": 27.75,
      "learning_rate": 6.946031144690798e-06,
      "loss": 4.2936,
      "num_input_tokens_seen": 41014496,
      "step": 30275
    },
    {
      "epoch": 0.9692081172780231,
      "grad_norm": 26.5,
      "learning_rate": 6.939241519512005e-06,
      "loss": 4.7164,
      "num_input_tokens_seen": 41021184,
      "step": 30280
    },
    {
      "epoch": 0.9693681582485116,
      "grad_norm": 40.5,
      "learning_rate": 6.932454679529129e-06,
      "loss": 4.4201,
      "num_input_tokens_seen": 41027648,
      "step": 30285
    },
    {
      "epoch": 0.969528199219,
      "grad_norm": 43.25,
      "learning_rate": 6.925670625788791e-06,
      "loss": 3.6882,
      "num_input_tokens_seen": 41034512,
      "step": 30290
    },
    {
      "epoch": 0.9696882401894885,
      "grad_norm": 28.5,
      "learning_rate": 6.918889359337186e-06,
      "loss": 3.79,
      "num_input_tokens_seen": 41041440,
      "step": 30295
    },
    {
      "epoch": 0.9698482811599769,
      "grad_norm": 24.75,
      "learning_rate": 6.912110881220058e-06,
      "loss": 3.9505,
      "num_input_tokens_seen": 41048016,
      "step": 30300
    },
    {
      "epoch": 0.9700083221304654,
      "grad_norm": 23.625,
      "learning_rate": 6.905335192482735e-06,
      "loss": 4.4015,
      "num_input_tokens_seen": 41054912,
      "step": 30305
    },
    {
      "epoch": 0.9701683631009539,
      "grad_norm": 30.625,
      "learning_rate": 6.8985622941701275e-06,
      "loss": 4.6524,
      "num_input_tokens_seen": 41061792,
      "step": 30310
    },
    {
      "epoch": 0.9703284040714423,
      "grad_norm": 23.75,
      "learning_rate": 6.89179218732669e-06,
      "loss": 3.909,
      "num_input_tokens_seen": 41068144,
      "step": 30315
    },
    {
      "epoch": 0.9704884450419308,
      "grad_norm": 27.875,
      "learning_rate": 6.8850248729964595e-06,
      "loss": 4.1592,
      "num_input_tokens_seen": 41074784,
      "step": 30320
    },
    {
      "epoch": 0.9706484860124192,
      "grad_norm": 30.625,
      "learning_rate": 6.8782603522230314e-06,
      "loss": 3.821,
      "num_input_tokens_seen": 41081280,
      "step": 30325
    },
    {
      "epoch": 0.9708085269829076,
      "grad_norm": 34.0,
      "learning_rate": 6.871498626049591e-06,
      "loss": 4.5631,
      "num_input_tokens_seen": 41088000,
      "step": 30330
    },
    {
      "epoch": 0.970968567953396,
      "grad_norm": 27.125,
      "learning_rate": 6.8647396955188875e-06,
      "loss": 4.6337,
      "num_input_tokens_seen": 41094256,
      "step": 30335
    },
    {
      "epoch": 0.9711286089238845,
      "grad_norm": 29.5,
      "learning_rate": 6.857983561673218e-06,
      "loss": 4.2481,
      "num_input_tokens_seen": 41100848,
      "step": 30340
    },
    {
      "epoch": 0.9712886498943729,
      "grad_norm": 18.25,
      "learning_rate": 6.851230225554467e-06,
      "loss": 4.6861,
      "num_input_tokens_seen": 41107568,
      "step": 30345
    },
    {
      "epoch": 0.9714486908648614,
      "grad_norm": 26.75,
      "learning_rate": 6.8444796882040946e-06,
      "loss": 5.5928,
      "num_input_tokens_seen": 41114096,
      "step": 30350
    },
    {
      "epoch": 0.9716087318353499,
      "grad_norm": 53.0,
      "learning_rate": 6.837731950663106e-06,
      "loss": 4.2478,
      "num_input_tokens_seen": 41120736,
      "step": 30355
    },
    {
      "epoch": 0.9717687728058383,
      "grad_norm": 26.375,
      "learning_rate": 6.830987013972098e-06,
      "loss": 3.7548,
      "num_input_tokens_seen": 41127488,
      "step": 30360
    },
    {
      "epoch": 0.9719288137763268,
      "grad_norm": 21.25,
      "learning_rate": 6.82424487917121e-06,
      "loss": 4.8961,
      "num_input_tokens_seen": 41133696,
      "step": 30365
    },
    {
      "epoch": 0.9720888547468152,
      "grad_norm": 39.0,
      "learning_rate": 6.8175055473001735e-06,
      "loss": 5.2386,
      "num_input_tokens_seen": 41140832,
      "step": 30370
    },
    {
      "epoch": 0.9722488957173037,
      "grad_norm": 19.0,
      "learning_rate": 6.8107690193982855e-06,
      "loss": 3.9532,
      "num_input_tokens_seen": 41147664,
      "step": 30375
    },
    {
      "epoch": 0.972408936687792,
      "grad_norm": 22.5,
      "learning_rate": 6.804035296504385e-06,
      "loss": 4.2696,
      "num_input_tokens_seen": 41154464,
      "step": 30380
    },
    {
      "epoch": 0.9725689776582805,
      "grad_norm": 15.375,
      "learning_rate": 6.797304379656916e-06,
      "loss": 3.6192,
      "num_input_tokens_seen": 41161024,
      "step": 30385
    },
    {
      "epoch": 0.9727290186287689,
      "grad_norm": 57.0,
      "learning_rate": 6.790576269893861e-06,
      "loss": 3.9877,
      "num_input_tokens_seen": 41168000,
      "step": 30390
    },
    {
      "epoch": 0.9728890595992574,
      "grad_norm": 33.5,
      "learning_rate": 6.783850968252772e-06,
      "loss": 3.9177,
      "num_input_tokens_seen": 41174976,
      "step": 30395
    },
    {
      "epoch": 0.9730491005697458,
      "grad_norm": 31.875,
      "learning_rate": 6.777128475770789e-06,
      "loss": 4.0617,
      "num_input_tokens_seen": 41182128,
      "step": 30400
    },
    {
      "epoch": 0.9730491005697458,
      "eval_loss": 4.490996837615967,
      "eval_runtime": 489.9857,
      "eval_samples_per_second": 28.34,
      "eval_steps_per_second": 14.17,
      "num_input_tokens_seen": 41182128,
      "step": 30400
    },
    {
      "epoch": 0.9732091415402343,
      "grad_norm": 24.875,
      "learning_rate": 6.77040879348459e-06,
      "loss": 4.1095,
      "num_input_tokens_seen": 41188800,
      "step": 30405
    },
    {
      "epoch": 0.9733691825107228,
      "grad_norm": 29.625,
      "learning_rate": 6.763691922430443e-06,
      "loss": 3.8498,
      "num_input_tokens_seen": 41195904,
      "step": 30410
    },
    {
      "epoch": 0.9735292234812112,
      "grad_norm": 36.25,
      "learning_rate": 6.756977863644178e-06,
      "loss": 4.0537,
      "num_input_tokens_seen": 41202912,
      "step": 30415
    },
    {
      "epoch": 0.9736892644516997,
      "grad_norm": 24.0,
      "learning_rate": 6.7502666181611804e-06,
      "loss": 4.3529,
      "num_input_tokens_seen": 41209424,
      "step": 30420
    },
    {
      "epoch": 0.9738493054221881,
      "grad_norm": 50.75,
      "learning_rate": 6.743558187016405e-06,
      "loss": 4.3507,
      "num_input_tokens_seen": 41216720,
      "step": 30425
    },
    {
      "epoch": 0.9740093463926766,
      "grad_norm": 23.75,
      "learning_rate": 6.7368525712443925e-06,
      "loss": 3.9887,
      "num_input_tokens_seen": 41223280,
      "step": 30430
    },
    {
      "epoch": 0.9741693873631649,
      "grad_norm": 20.25,
      "learning_rate": 6.7301497718792155e-06,
      "loss": 4.142,
      "num_input_tokens_seen": 41230016,
      "step": 30435
    },
    {
      "epoch": 0.9743294283336534,
      "grad_norm": 27.0,
      "learning_rate": 6.723449789954544e-06,
      "loss": 4.89,
      "num_input_tokens_seen": 41237120,
      "step": 30440
    },
    {
      "epoch": 0.9744894693041418,
      "grad_norm": 37.5,
      "learning_rate": 6.716752626503586e-06,
      "loss": 5.2731,
      "num_input_tokens_seen": 41243776,
      "step": 30445
    },
    {
      "epoch": 0.9746495102746303,
      "grad_norm": 25.625,
      "learning_rate": 6.710058282559131e-06,
      "loss": 4.7785,
      "num_input_tokens_seen": 41250592,
      "step": 30450
    },
    {
      "epoch": 0.9748095512451187,
      "grad_norm": 34.5,
      "learning_rate": 6.703366759153545e-06,
      "loss": 4.4103,
      "num_input_tokens_seen": 41257168,
      "step": 30455
    },
    {
      "epoch": 0.9749695922156072,
      "grad_norm": 26.75,
      "learning_rate": 6.6966780573187335e-06,
      "loss": 3.7686,
      "num_input_tokens_seen": 41264128,
      "step": 30460
    },
    {
      "epoch": 0.9751296331860957,
      "grad_norm": 32.0,
      "learning_rate": 6.689992178086174e-06,
      "loss": 4.7031,
      "num_input_tokens_seen": 41270752,
      "step": 30465
    },
    {
      "epoch": 0.9752896741565841,
      "grad_norm": 22.875,
      "learning_rate": 6.683309122486925e-06,
      "loss": 3.7856,
      "num_input_tokens_seen": 41277392,
      "step": 30470
    },
    {
      "epoch": 0.9754497151270726,
      "grad_norm": 42.0,
      "learning_rate": 6.676628891551584e-06,
      "loss": 4.3352,
      "num_input_tokens_seen": 41284000,
      "step": 30475
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 27.625,
      "learning_rate": 6.6699514863103385e-06,
      "loss": 4.3629,
      "num_input_tokens_seen": 41290816,
      "step": 30480
    },
    {
      "epoch": 0.9757697970680494,
      "grad_norm": 31.5,
      "learning_rate": 6.663276907792921e-06,
      "loss": 4.5899,
      "num_input_tokens_seen": 41297504,
      "step": 30485
    },
    {
      "epoch": 0.9759298380385378,
      "grad_norm": 40.5,
      "learning_rate": 6.656605157028634e-06,
      "loss": 4.4315,
      "num_input_tokens_seen": 41303984,
      "step": 30490
    },
    {
      "epoch": 0.9760898790090263,
      "grad_norm": 14.0625,
      "learning_rate": 6.649936235046358e-06,
      "loss": 3.9916,
      "num_input_tokens_seen": 41310432,
      "step": 30495
    },
    {
      "epoch": 0.9762499199795147,
      "grad_norm": 33.75,
      "learning_rate": 6.643270142874508e-06,
      "loss": 5.0173,
      "num_input_tokens_seen": 41317120,
      "step": 30500
    },
    {
      "epoch": 0.9764099609500032,
      "grad_norm": 18.375,
      "learning_rate": 6.636606881541094e-06,
      "loss": 5.725,
      "num_input_tokens_seen": 41323600,
      "step": 30505
    },
    {
      "epoch": 0.9765700019204916,
      "grad_norm": 41.75,
      "learning_rate": 6.629946452073662e-06,
      "loss": 6.1276,
      "num_input_tokens_seen": 41330256,
      "step": 30510
    },
    {
      "epoch": 0.9767300428909801,
      "grad_norm": 18.125,
      "learning_rate": 6.6232888554993375e-06,
      "loss": 3.5961,
      "num_input_tokens_seen": 41336688,
      "step": 30515
    },
    {
      "epoch": 0.9768900838614686,
      "grad_norm": 61.5,
      "learning_rate": 6.616634092844817e-06,
      "loss": 4.3697,
      "num_input_tokens_seen": 41343504,
      "step": 30520
    },
    {
      "epoch": 0.977050124831957,
      "grad_norm": 17.0,
      "learning_rate": 6.609982165136331e-06,
      "loss": 4.1089,
      "num_input_tokens_seen": 41349920,
      "step": 30525
    },
    {
      "epoch": 0.9772101658024455,
      "grad_norm": 51.75,
      "learning_rate": 6.603333073399706e-06,
      "loss": 4.9238,
      "num_input_tokens_seen": 41356816,
      "step": 30530
    },
    {
      "epoch": 0.9773702067729338,
      "grad_norm": 24.25,
      "learning_rate": 6.596686818660308e-06,
      "loss": 3.825,
      "num_input_tokens_seen": 41363808,
      "step": 30535
    },
    {
      "epoch": 0.9775302477434223,
      "grad_norm": 15.9375,
      "learning_rate": 6.590043401943066e-06,
      "loss": 3.6073,
      "num_input_tokens_seen": 41370928,
      "step": 30540
    },
    {
      "epoch": 0.9776902887139107,
      "grad_norm": 31.25,
      "learning_rate": 6.583402824272494e-06,
      "loss": 4.8328,
      "num_input_tokens_seen": 41377936,
      "step": 30545
    },
    {
      "epoch": 0.9778503296843992,
      "grad_norm": 19.0,
      "learning_rate": 6.576765086672634e-06,
      "loss": 4.1753,
      "num_input_tokens_seen": 41384704,
      "step": 30550
    },
    {
      "epoch": 0.9780103706548876,
      "grad_norm": 37.5,
      "learning_rate": 6.57013019016712e-06,
      "loss": 4.8479,
      "num_input_tokens_seen": 41391440,
      "step": 30555
    },
    {
      "epoch": 0.9781704116253761,
      "grad_norm": 20.5,
      "learning_rate": 6.563498135779142e-06,
      "loss": 3.7463,
      "num_input_tokens_seen": 41399136,
      "step": 30560
    },
    {
      "epoch": 0.9783304525958645,
      "grad_norm": 30.5,
      "learning_rate": 6.556868924531431e-06,
      "loss": 4.5935,
      "num_input_tokens_seen": 41405760,
      "step": 30565
    },
    {
      "epoch": 0.978490493566353,
      "grad_norm": 23.375,
      "learning_rate": 6.550242557446304e-06,
      "loss": 4.4563,
      "num_input_tokens_seen": 41412560,
      "step": 30570
    },
    {
      "epoch": 0.9786505345368415,
      "grad_norm": 24.125,
      "learning_rate": 6.543619035545634e-06,
      "loss": 4.3092,
      "num_input_tokens_seen": 41419040,
      "step": 30575
    },
    {
      "epoch": 0.9788105755073299,
      "grad_norm": 26.625,
      "learning_rate": 6.53699835985084e-06,
      "loss": 4.8658,
      "num_input_tokens_seen": 41425760,
      "step": 30580
    },
    {
      "epoch": 0.9789706164778184,
      "grad_norm": 30.375,
      "learning_rate": 6.530380531382927e-06,
      "loss": 4.6706,
      "num_input_tokens_seen": 41432432,
      "step": 30585
    },
    {
      "epoch": 0.9791306574483067,
      "grad_norm": 22.625,
      "learning_rate": 6.523765551162433e-06,
      "loss": 4.3958,
      "num_input_tokens_seen": 41439104,
      "step": 30590
    },
    {
      "epoch": 0.9792906984187952,
      "grad_norm": 39.5,
      "learning_rate": 6.517153420209476e-06,
      "loss": 4.5495,
      "num_input_tokens_seen": 41445952,
      "step": 30595
    },
    {
      "epoch": 0.9794507393892836,
      "grad_norm": 33.75,
      "learning_rate": 6.510544139543739e-06,
      "loss": 4.0031,
      "num_input_tokens_seen": 41452688,
      "step": 30600
    },
    {
      "epoch": 0.9794507393892836,
      "eval_loss": 4.490927219390869,
      "eval_runtime": 490.089,
      "eval_samples_per_second": 28.334,
      "eval_steps_per_second": 14.167,
      "num_input_tokens_seen": 41452688,
      "step": 30600
    },
    {
      "epoch": 0.9796107803597721,
      "grad_norm": 38.75,
      "learning_rate": 6.503937710184452e-06,
      "loss": 4.7633,
      "num_input_tokens_seen": 41459408,
      "step": 30605
    },
    {
      "epoch": 0.9797708213302605,
      "grad_norm": 30.625,
      "learning_rate": 6.4973341331503954e-06,
      "loss": 4.5314,
      "num_input_tokens_seen": 41466032,
      "step": 30610
    },
    {
      "epoch": 0.979930862300749,
      "grad_norm": 46.0,
      "learning_rate": 6.490733409459942e-06,
      "loss": 4.5742,
      "num_input_tokens_seen": 41472816,
      "step": 30615
    },
    {
      "epoch": 0.9800909032712375,
      "grad_norm": 24.875,
      "learning_rate": 6.484135540130995e-06,
      "loss": 4.2962,
      "num_input_tokens_seen": 41479888,
      "step": 30620
    },
    {
      "epoch": 0.9802509442417259,
      "grad_norm": 24.5,
      "learning_rate": 6.4775405261810364e-06,
      "loss": 4.9703,
      "num_input_tokens_seen": 41486800,
      "step": 30625
    },
    {
      "epoch": 0.9804109852122144,
      "grad_norm": 18.5,
      "learning_rate": 6.470948368627092e-06,
      "loss": 3.6766,
      "num_input_tokens_seen": 41493648,
      "step": 30630
    },
    {
      "epoch": 0.9805710261827028,
      "grad_norm": 44.5,
      "learning_rate": 6.464359068485756e-06,
      "loss": 5.1048,
      "num_input_tokens_seen": 41500080,
      "step": 30635
    },
    {
      "epoch": 0.9807310671531912,
      "grad_norm": 36.25,
      "learning_rate": 6.457772626773195e-06,
      "loss": 5.2212,
      "num_input_tokens_seen": 41506912,
      "step": 30640
    },
    {
      "epoch": 0.9808911081236796,
      "grad_norm": 44.75,
      "learning_rate": 6.451189044505104e-06,
      "loss": 4.0161,
      "num_input_tokens_seen": 41513392,
      "step": 30645
    },
    {
      "epoch": 0.9810511490941681,
      "grad_norm": 25.875,
      "learning_rate": 6.44460832269676e-06,
      "loss": 4.678,
      "num_input_tokens_seen": 41520192,
      "step": 30650
    },
    {
      "epoch": 0.9812111900646565,
      "grad_norm": 16.625,
      "learning_rate": 6.438030462363001e-06,
      "loss": 4.9167,
      "num_input_tokens_seen": 41526592,
      "step": 30655
    },
    {
      "epoch": 0.981371231035145,
      "grad_norm": 19.25,
      "learning_rate": 6.431455464518205e-06,
      "loss": 4.4445,
      "num_input_tokens_seen": 41533408,
      "step": 30660
    },
    {
      "epoch": 0.9815312720056334,
      "grad_norm": 37.0,
      "learning_rate": 6.424883330176326e-06,
      "loss": 2.9862,
      "num_input_tokens_seen": 41540128,
      "step": 30665
    },
    {
      "epoch": 0.9816913129761219,
      "grad_norm": 25.875,
      "learning_rate": 6.418314060350864e-06,
      "loss": 4.1259,
      "num_input_tokens_seen": 41546784,
      "step": 30670
    },
    {
      "epoch": 0.9818513539466104,
      "grad_norm": 45.0,
      "learning_rate": 6.4117476560548895e-06,
      "loss": 4.9611,
      "num_input_tokens_seen": 41553808,
      "step": 30675
    },
    {
      "epoch": 0.9820113949170988,
      "grad_norm": 27.375,
      "learning_rate": 6.405184118301016e-06,
      "loss": 4.1782,
      "num_input_tokens_seen": 41560432,
      "step": 30680
    },
    {
      "epoch": 0.9821714358875873,
      "grad_norm": 26.625,
      "learning_rate": 6.398623448101434e-06,
      "loss": 4.4679,
      "num_input_tokens_seen": 41567232,
      "step": 30685
    },
    {
      "epoch": 0.9823314768580756,
      "grad_norm": 21.0,
      "learning_rate": 6.392065646467871e-06,
      "loss": 3.7084,
      "num_input_tokens_seen": 41573568,
      "step": 30690
    },
    {
      "epoch": 0.9824915178285641,
      "grad_norm": 35.5,
      "learning_rate": 6.385510714411632e-06,
      "loss": 3.8751,
      "num_input_tokens_seen": 41580192,
      "step": 30695
    },
    {
      "epoch": 0.9826515587990525,
      "grad_norm": 32.5,
      "learning_rate": 6.378958652943559e-06,
      "loss": 4.6444,
      "num_input_tokens_seen": 41586768,
      "step": 30700
    },
    {
      "epoch": 0.982811599769541,
      "grad_norm": 17.625,
      "learning_rate": 6.3724094630740776e-06,
      "loss": 4.104,
      "num_input_tokens_seen": 41593648,
      "step": 30705
    },
    {
      "epoch": 0.9829716407400294,
      "grad_norm": 21.25,
      "learning_rate": 6.365863145813136e-06,
      "loss": 4.1005,
      "num_input_tokens_seen": 41600368,
      "step": 30710
    },
    {
      "epoch": 0.9831316817105179,
      "grad_norm": 19.75,
      "learning_rate": 6.359319702170269e-06,
      "loss": 3.8816,
      "num_input_tokens_seen": 41607056,
      "step": 30715
    },
    {
      "epoch": 0.9832917226810063,
      "grad_norm": 13.0625,
      "learning_rate": 6.352779133154566e-06,
      "loss": 5.0974,
      "num_input_tokens_seen": 41613776,
      "step": 30720
    },
    {
      "epoch": 0.9834517636514948,
      "grad_norm": 18.375,
      "learning_rate": 6.346241439774648e-06,
      "loss": 4.312,
      "num_input_tokens_seen": 41620288,
      "step": 30725
    },
    {
      "epoch": 0.9836118046219833,
      "grad_norm": 36.25,
      "learning_rate": 6.339706623038716e-06,
      "loss": 3.0973,
      "num_input_tokens_seen": 41627088,
      "step": 30730
    },
    {
      "epoch": 0.9837718455924717,
      "grad_norm": 19.875,
      "learning_rate": 6.333174683954532e-06,
      "loss": 3.9819,
      "num_input_tokens_seen": 41633776,
      "step": 30735
    },
    {
      "epoch": 0.9839318865629602,
      "grad_norm": 26.625,
      "learning_rate": 6.326645623529387e-06,
      "loss": 4.4802,
      "num_input_tokens_seen": 41640688,
      "step": 30740
    },
    {
      "epoch": 0.9840919275334485,
      "grad_norm": 25.25,
      "learning_rate": 6.320119442770156e-06,
      "loss": 4.1052,
      "num_input_tokens_seen": 41647232,
      "step": 30745
    },
    {
      "epoch": 0.984251968503937,
      "grad_norm": 28.0,
      "learning_rate": 6.313596142683254e-06,
      "loss": 3.5685,
      "num_input_tokens_seen": 41654064,
      "step": 30750
    },
    {
      "epoch": 0.9844120094744254,
      "grad_norm": 34.25,
      "learning_rate": 6.307075724274647e-06,
      "loss": 3.5819,
      "num_input_tokens_seen": 41660800,
      "step": 30755
    },
    {
      "epoch": 0.9845720504449139,
      "grad_norm": 34.0,
      "learning_rate": 6.300558188549882e-06,
      "loss": 5.9252,
      "num_input_tokens_seen": 41667456,
      "step": 30760
    },
    {
      "epoch": 0.9847320914154023,
      "grad_norm": 46.75,
      "learning_rate": 6.29404353651403e-06,
      "loss": 3.9845,
      "num_input_tokens_seen": 41674048,
      "step": 30765
    },
    {
      "epoch": 0.9848921323858908,
      "grad_norm": 32.25,
      "learning_rate": 6.287531769171737e-06,
      "loss": 3.9793,
      "num_input_tokens_seen": 41680608,
      "step": 30770
    },
    {
      "epoch": 0.9850521733563792,
      "grad_norm": 26.625,
      "learning_rate": 6.2810228875272045e-06,
      "loss": 4.6643,
      "num_input_tokens_seen": 41687056,
      "step": 30775
    },
    {
      "epoch": 0.9852122143268677,
      "grad_norm": 23.875,
      "learning_rate": 6.274516892584179e-06,
      "loss": 3.4119,
      "num_input_tokens_seen": 41693760,
      "step": 30780
    },
    {
      "epoch": 0.9853722552973562,
      "grad_norm": 40.25,
      "learning_rate": 6.268013785345969e-06,
      "loss": 4.3505,
      "num_input_tokens_seen": 41700672,
      "step": 30785
    },
    {
      "epoch": 0.9855322962678446,
      "grad_norm": 19.125,
      "learning_rate": 6.26151356681543e-06,
      "loss": 3.784,
      "num_input_tokens_seen": 41707312,
      "step": 30790
    },
    {
      "epoch": 0.985692337238333,
      "grad_norm": 28.25,
      "learning_rate": 6.255016237994981e-06,
      "loss": 5.2918,
      "num_input_tokens_seen": 41714064,
      "step": 30795
    },
    {
      "epoch": 0.9858523782088214,
      "grad_norm": 38.25,
      "learning_rate": 6.248521799886603e-06,
      "loss": 4.8552,
      "num_input_tokens_seen": 41721056,
      "step": 30800
    },
    {
      "epoch": 0.9858523782088214,
      "eval_loss": 4.488006114959717,
      "eval_runtime": 490.6933,
      "eval_samples_per_second": 28.299,
      "eval_steps_per_second": 14.149,
      "num_input_tokens_seen": 41721056,
      "step": 30800
    },
    {
      "epoch": 0.9860124191793099,
      "grad_norm": 38.75,
      "learning_rate": 6.242030253491798e-06,
      "loss": 5.1203,
      "num_input_tokens_seen": 41728128,
      "step": 30805
    },
    {
      "epoch": 0.9861724601497983,
      "grad_norm": 22.125,
      "learning_rate": 6.235541599811656e-06,
      "loss": 4.9068,
      "num_input_tokens_seen": 41734864,
      "step": 30810
    },
    {
      "epoch": 0.9863325011202868,
      "grad_norm": 34.25,
      "learning_rate": 6.229055839846814e-06,
      "loss": 5.2181,
      "num_input_tokens_seen": 41741472,
      "step": 30815
    },
    {
      "epoch": 0.9864925420907752,
      "grad_norm": 59.75,
      "learning_rate": 6.222572974597455e-06,
      "loss": 4.4479,
      "num_input_tokens_seen": 41747936,
      "step": 30820
    },
    {
      "epoch": 0.9866525830612637,
      "grad_norm": 19.625,
      "learning_rate": 6.216093005063306e-06,
      "loss": 3.8824,
      "num_input_tokens_seen": 41754800,
      "step": 30825
    },
    {
      "epoch": 0.9868126240317522,
      "grad_norm": 26.875,
      "learning_rate": 6.209615932243678e-06,
      "loss": 3.8647,
      "num_input_tokens_seen": 41761600,
      "step": 30830
    },
    {
      "epoch": 0.9869726650022406,
      "grad_norm": 27.375,
      "learning_rate": 6.203141757137399e-06,
      "loss": 4.28,
      "num_input_tokens_seen": 41768496,
      "step": 30835
    },
    {
      "epoch": 0.9871327059727291,
      "grad_norm": 27.875,
      "learning_rate": 6.196670480742886e-06,
      "loss": 3.7478,
      "num_input_tokens_seen": 41775264,
      "step": 30840
    },
    {
      "epoch": 0.9872927469432174,
      "grad_norm": 31.75,
      "learning_rate": 6.190202104058074e-06,
      "loss": 4.3294,
      "num_input_tokens_seen": 41782048,
      "step": 30845
    },
    {
      "epoch": 0.9874527879137059,
      "grad_norm": 47.0,
      "learning_rate": 6.183736628080475e-06,
      "loss": 3.9228,
      "num_input_tokens_seen": 41789168,
      "step": 30850
    },
    {
      "epoch": 0.9876128288841943,
      "grad_norm": 43.0,
      "learning_rate": 6.177274053807155e-06,
      "loss": 4.3962,
      "num_input_tokens_seen": 41795680,
      "step": 30855
    },
    {
      "epoch": 0.9877728698546828,
      "grad_norm": 20.375,
      "learning_rate": 6.170814382234713e-06,
      "loss": 4.3785,
      "num_input_tokens_seen": 41802432,
      "step": 30860
    },
    {
      "epoch": 0.9879329108251712,
      "grad_norm": 36.75,
      "learning_rate": 6.16435761435932e-06,
      "loss": 5.1576,
      "num_input_tokens_seen": 41809280,
      "step": 30865
    },
    {
      "epoch": 0.9880929517956597,
      "grad_norm": 24.875,
      "learning_rate": 6.157903751176681e-06,
      "loss": 4.4609,
      "num_input_tokens_seen": 41815968,
      "step": 30870
    },
    {
      "epoch": 0.9882529927661481,
      "grad_norm": 20.125,
      "learning_rate": 6.151452793682066e-06,
      "loss": 3.7862,
      "num_input_tokens_seen": 41822672,
      "step": 30875
    },
    {
      "epoch": 0.9884130337366366,
      "grad_norm": 22.25,
      "learning_rate": 6.145004742870305e-06,
      "loss": 4.3126,
      "num_input_tokens_seen": 41829888,
      "step": 30880
    },
    {
      "epoch": 0.9885730747071251,
      "grad_norm": 18.875,
      "learning_rate": 6.138559599735752e-06,
      "loss": 4.9249,
      "num_input_tokens_seen": 41836464,
      "step": 30885
    },
    {
      "epoch": 0.9887331156776135,
      "grad_norm": 31.125,
      "learning_rate": 6.132117365272344e-06,
      "loss": 4.1374,
      "num_input_tokens_seen": 41843312,
      "step": 30890
    },
    {
      "epoch": 0.988893156648102,
      "grad_norm": 32.5,
      "learning_rate": 6.125678040473545e-06,
      "loss": 4.0083,
      "num_input_tokens_seen": 41850320,
      "step": 30895
    },
    {
      "epoch": 0.9890531976185903,
      "grad_norm": 43.75,
      "learning_rate": 6.1192416263323755e-06,
      "loss": 3.7731,
      "num_input_tokens_seen": 41857344,
      "step": 30900
    },
    {
      "epoch": 0.9892132385890788,
      "grad_norm": 23.75,
      "learning_rate": 6.112808123841424e-06,
      "loss": 3.8485,
      "num_input_tokens_seen": 41864096,
      "step": 30905
    },
    {
      "epoch": 0.9893732795595672,
      "grad_norm": 36.25,
      "learning_rate": 6.106377533992805e-06,
      "loss": 3.9707,
      "num_input_tokens_seen": 41870848,
      "step": 30910
    },
    {
      "epoch": 0.9895333205300557,
      "grad_norm": 32.25,
      "learning_rate": 6.099949857778204e-06,
      "loss": 3.7508,
      "num_input_tokens_seen": 41877696,
      "step": 30915
    },
    {
      "epoch": 0.9896933615005441,
      "grad_norm": 27.875,
      "learning_rate": 6.093525096188852e-06,
      "loss": 4.5832,
      "num_input_tokens_seen": 41884384,
      "step": 30920
    },
    {
      "epoch": 0.9898534024710326,
      "grad_norm": 26.75,
      "learning_rate": 6.087103250215518e-06,
      "loss": 4.7252,
      "num_input_tokens_seen": 41890976,
      "step": 30925
    },
    {
      "epoch": 0.990013443441521,
      "grad_norm": 36.25,
      "learning_rate": 6.080684320848537e-06,
      "loss": 4.1808,
      "num_input_tokens_seen": 41897760,
      "step": 30930
    },
    {
      "epoch": 0.9901734844120095,
      "grad_norm": 22.875,
      "learning_rate": 6.074268309077794e-06,
      "loss": 3.7741,
      "num_input_tokens_seen": 41904752,
      "step": 30935
    },
    {
      "epoch": 0.990333525382498,
      "grad_norm": 27.75,
      "learning_rate": 6.067855215892709e-06,
      "loss": 4.4771,
      "num_input_tokens_seen": 41911552,
      "step": 30940
    },
    {
      "epoch": 0.9904935663529864,
      "grad_norm": 27.5,
      "learning_rate": 6.061445042282271e-06,
      "loss": 4.9569,
      "num_input_tokens_seen": 41918384,
      "step": 30945
    },
    {
      "epoch": 0.9906536073234748,
      "grad_norm": 36.0,
      "learning_rate": 6.055037789234999e-06,
      "loss": 5.4271,
      "num_input_tokens_seen": 41925568,
      "step": 30950
    },
    {
      "epoch": 0.9908136482939632,
      "grad_norm": 35.5,
      "learning_rate": 6.048633457738975e-06,
      "loss": 5.8296,
      "num_input_tokens_seen": 41932592,
      "step": 30955
    },
    {
      "epoch": 0.9909736892644517,
      "grad_norm": 38.25,
      "learning_rate": 6.042232048781837e-06,
      "loss": 4.6143,
      "num_input_tokens_seen": 41939264,
      "step": 30960
    },
    {
      "epoch": 0.9911337302349401,
      "grad_norm": 25.75,
      "learning_rate": 6.035833563350757e-06,
      "loss": 5.0128,
      "num_input_tokens_seen": 41946080,
      "step": 30965
    },
    {
      "epoch": 0.9912937712054286,
      "grad_norm": 18.125,
      "learning_rate": 6.0294380024324525e-06,
      "loss": 3.5159,
      "num_input_tokens_seen": 41952944,
      "step": 30970
    },
    {
      "epoch": 0.991453812175917,
      "grad_norm": 35.25,
      "learning_rate": 6.023045367013213e-06,
      "loss": 3.607,
      "num_input_tokens_seen": 41959504,
      "step": 30975
    },
    {
      "epoch": 0.9916138531464055,
      "grad_norm": 37.75,
      "learning_rate": 6.016655658078851e-06,
      "loss": 4.6487,
      "num_input_tokens_seen": 41966000,
      "step": 30980
    },
    {
      "epoch": 0.9917738941168939,
      "grad_norm": 44.0,
      "learning_rate": 6.010268876614753e-06,
      "loss": 4.3035,
      "num_input_tokens_seen": 41972768,
      "step": 30985
    },
    {
      "epoch": 0.9919339350873824,
      "grad_norm": 47.25,
      "learning_rate": 6.0038850236058266e-06,
      "loss": 4.898,
      "num_input_tokens_seen": 41979584,
      "step": 30990
    },
    {
      "epoch": 0.9920939760578709,
      "grad_norm": 34.0,
      "learning_rate": 5.997504100036549e-06,
      "loss": 5.0576,
      "num_input_tokens_seen": 41986784,
      "step": 30995
    },
    {
      "epoch": 0.9922540170283592,
      "grad_norm": 10.4375,
      "learning_rate": 5.991126106890949e-06,
      "loss": 3.6645,
      "num_input_tokens_seen": 41993584,
      "step": 31000
    },
    {
      "epoch": 0.9922540170283592,
      "eval_loss": 4.484321117401123,
      "eval_runtime": 490.5672,
      "eval_samples_per_second": 28.306,
      "eval_steps_per_second": 14.153,
      "num_input_tokens_seen": 41993584,
      "step": 31000
    },
    {
      "epoch": 0.9924140579988477,
      "grad_norm": 25.5,
      "learning_rate": 5.984751045152576e-06,
      "loss": 4.3995,
      "num_input_tokens_seen": 42000864,
      "step": 31005
    },
    {
      "epoch": 0.9925740989693361,
      "grad_norm": 18.75,
      "learning_rate": 5.978378915804553e-06,
      "loss": 3.8902,
      "num_input_tokens_seen": 42008128,
      "step": 31010
    },
    {
      "epoch": 0.9927341399398246,
      "grad_norm": 20.75,
      "learning_rate": 5.972009719829547e-06,
      "loss": 3.7722,
      "num_input_tokens_seen": 42014944,
      "step": 31015
    },
    {
      "epoch": 0.992894180910313,
      "grad_norm": 20.625,
      "learning_rate": 5.965643458209755e-06,
      "loss": 4.0109,
      "num_input_tokens_seen": 42021936,
      "step": 31020
    },
    {
      "epoch": 0.9930542218808015,
      "grad_norm": 39.5,
      "learning_rate": 5.95928013192695e-06,
      "loss": 4.5746,
      "num_input_tokens_seen": 42028592,
      "step": 31025
    },
    {
      "epoch": 0.9932142628512899,
      "grad_norm": 14.8125,
      "learning_rate": 5.952919741962423e-06,
      "loss": 3.8138,
      "num_input_tokens_seen": 42035216,
      "step": 31030
    },
    {
      "epoch": 0.9933743038217784,
      "grad_norm": 26.375,
      "learning_rate": 5.946562289297042e-06,
      "loss": 4.1118,
      "num_input_tokens_seen": 42041936,
      "step": 31035
    },
    {
      "epoch": 0.9935343447922668,
      "grad_norm": 21.125,
      "learning_rate": 5.9402077749111855e-06,
      "loss": 4.7771,
      "num_input_tokens_seen": 42048832,
      "step": 31040
    },
    {
      "epoch": 0.9936943857627553,
      "grad_norm": 21.25,
      "learning_rate": 5.933856199784821e-06,
      "loss": 3.9418,
      "num_input_tokens_seen": 42055840,
      "step": 31045
    },
    {
      "epoch": 0.9938544267332438,
      "grad_norm": 22.625,
      "learning_rate": 5.927507564897419e-06,
      "loss": 3.5823,
      "num_input_tokens_seen": 42062960,
      "step": 31050
    },
    {
      "epoch": 0.9940144677037321,
      "grad_norm": 17.5,
      "learning_rate": 5.9211618712280395e-06,
      "loss": 3.5495,
      "num_input_tokens_seen": 42069456,
      "step": 31055
    },
    {
      "epoch": 0.9941745086742206,
      "grad_norm": 42.25,
      "learning_rate": 5.914819119755255e-06,
      "loss": 3.6632,
      "num_input_tokens_seen": 42075952,
      "step": 31060
    },
    {
      "epoch": 0.994334549644709,
      "grad_norm": 28.125,
      "learning_rate": 5.908479311457205e-06,
      "loss": 3.9407,
      "num_input_tokens_seen": 42083136,
      "step": 31065
    },
    {
      "epoch": 0.9944945906151975,
      "grad_norm": 28.5,
      "learning_rate": 5.902142447311559e-06,
      "loss": 4.8063,
      "num_input_tokens_seen": 42090048,
      "step": 31070
    },
    {
      "epoch": 0.9946546315856859,
      "grad_norm": 49.75,
      "learning_rate": 5.895808528295546e-06,
      "loss": 4.5887,
      "num_input_tokens_seen": 42096624,
      "step": 31075
    },
    {
      "epoch": 0.9948146725561744,
      "grad_norm": 31.875,
      "learning_rate": 5.889477555385941e-06,
      "loss": 4.5216,
      "num_input_tokens_seen": 42103184,
      "step": 31080
    },
    {
      "epoch": 0.9949747135266628,
      "grad_norm": 27.5,
      "learning_rate": 5.883149529559051e-06,
      "loss": 4.3178,
      "num_input_tokens_seen": 42110272,
      "step": 31085
    },
    {
      "epoch": 0.9951347544971513,
      "grad_norm": 34.25,
      "learning_rate": 5.876824451790738e-06,
      "loss": 4.2239,
      "num_input_tokens_seen": 42117216,
      "step": 31090
    },
    {
      "epoch": 0.9952947954676398,
      "grad_norm": 19.75,
      "learning_rate": 5.87050232305642e-06,
      "loss": 3.951,
      "num_input_tokens_seen": 42123856,
      "step": 31095
    },
    {
      "epoch": 0.9954548364381282,
      "grad_norm": 47.5,
      "learning_rate": 5.864183144331034e-06,
      "loss": 3.9069,
      "num_input_tokens_seen": 42130800,
      "step": 31100
    },
    {
      "epoch": 0.9956148774086166,
      "grad_norm": 26.5,
      "learning_rate": 5.857866916589089e-06,
      "loss": 4.6685,
      "num_input_tokens_seen": 42138240,
      "step": 31105
    },
    {
      "epoch": 0.995774918379105,
      "grad_norm": 31.0,
      "learning_rate": 5.8515536408046216e-06,
      "loss": 5.5523,
      "num_input_tokens_seen": 42144576,
      "step": 31110
    },
    {
      "epoch": 0.9959349593495935,
      "grad_norm": 34.25,
      "learning_rate": 5.845243317951208e-06,
      "loss": 4.4144,
      "num_input_tokens_seen": 42151728,
      "step": 31115
    },
    {
      "epoch": 0.9960950003200819,
      "grad_norm": 40.5,
      "learning_rate": 5.838935949001997e-06,
      "loss": 3.8192,
      "num_input_tokens_seen": 42158864,
      "step": 31120
    },
    {
      "epoch": 0.9962550412905704,
      "grad_norm": 25.625,
      "learning_rate": 5.8326315349296476e-06,
      "loss": 4.2363,
      "num_input_tokens_seen": 42165648,
      "step": 31125
    },
    {
      "epoch": 0.9964150822610588,
      "grad_norm": 17.375,
      "learning_rate": 5.826330076706396e-06,
      "loss": 4.5115,
      "num_input_tokens_seen": 42172384,
      "step": 31130
    },
    {
      "epoch": 0.9965751232315473,
      "grad_norm": 26.625,
      "learning_rate": 5.820031575303988e-06,
      "loss": 4.5158,
      "num_input_tokens_seen": 42179200,
      "step": 31135
    },
    {
      "epoch": 0.9967351642020357,
      "grad_norm": 20.0,
      "learning_rate": 5.813736031693745e-06,
      "loss": 4.3851,
      "num_input_tokens_seen": 42185904,
      "step": 31140
    },
    {
      "epoch": 0.9968952051725242,
      "grad_norm": 59.5,
      "learning_rate": 5.807443446846522e-06,
      "loss": 5.0777,
      "num_input_tokens_seen": 42192592,
      "step": 31145
    },
    {
      "epoch": 0.9970552461430127,
      "grad_norm": 22.375,
      "learning_rate": 5.801153821732699e-06,
      "loss": 3.7877,
      "num_input_tokens_seen": 42199024,
      "step": 31150
    },
    {
      "epoch": 0.997215287113501,
      "grad_norm": 26.25,
      "learning_rate": 5.794867157322229e-06,
      "loss": 4.438,
      "num_input_tokens_seen": 42206496,
      "step": 31155
    },
    {
      "epoch": 0.9973753280839895,
      "grad_norm": 19.75,
      "learning_rate": 5.788583454584593e-06,
      "loss": 4.1463,
      "num_input_tokens_seen": 42213152,
      "step": 31160
    },
    {
      "epoch": 0.9975353690544779,
      "grad_norm": 40.75,
      "learning_rate": 5.7823027144888075e-06,
      "loss": 4.0303,
      "num_input_tokens_seen": 42219744,
      "step": 31165
    },
    {
      "epoch": 0.9976954100249664,
      "grad_norm": 46.0,
      "learning_rate": 5.776024938003455e-06,
      "loss": 4.2212,
      "num_input_tokens_seen": 42226304,
      "step": 31170
    },
    {
      "epoch": 0.9978554509954548,
      "grad_norm": 31.625,
      "learning_rate": 5.7697501260966345e-06,
      "loss": 4.0372,
      "num_input_tokens_seen": 42232960,
      "step": 31175
    },
    {
      "epoch": 0.9980154919659433,
      "grad_norm": 28.625,
      "learning_rate": 5.7634782797360145e-06,
      "loss": 4.0809,
      "num_input_tokens_seen": 42239568,
      "step": 31180
    },
    {
      "epoch": 0.9981755329364317,
      "grad_norm": 29.625,
      "learning_rate": 5.757209399888777e-06,
      "loss": 4.1816,
      "num_input_tokens_seen": 42246624,
      "step": 31185
    },
    {
      "epoch": 0.9983355739069202,
      "grad_norm": 20.625,
      "learning_rate": 5.750943487521679e-06,
      "loss": 4.6441,
      "num_input_tokens_seen": 42253376,
      "step": 31190
    },
    {
      "epoch": 0.9984956148774086,
      "grad_norm": 24.875,
      "learning_rate": 5.744680543600986e-06,
      "loss": 4.1237,
      "num_input_tokens_seen": 42259968,
      "step": 31195
    },
    {
      "epoch": 0.9986556558478971,
      "grad_norm": 24.0,
      "learning_rate": 5.738420569092537e-06,
      "loss": 3.9784,
      "num_input_tokens_seen": 42266304,
      "step": 31200
    },
    {
      "epoch": 0.9986556558478971,
      "eval_loss": 4.476981163024902,
      "eval_runtime": 490.579,
      "eval_samples_per_second": 28.305,
      "eval_steps_per_second": 14.153,
      "num_input_tokens_seen": 42266304,
      "step": 31200
    },
    {
      "epoch": 0.9988156968183856,
      "grad_norm": 24.25,
      "learning_rate": 5.732163564961684e-06,
      "loss": 4.6412,
      "num_input_tokens_seen": 42273360,
      "step": 31205
    },
    {
      "epoch": 0.9989757377888739,
      "grad_norm": 27.5,
      "learning_rate": 5.725909532173354e-06,
      "loss": 4.5179,
      "num_input_tokens_seen": 42279920,
      "step": 31210
    },
    {
      "epoch": 0.9991357787593624,
      "grad_norm": 44.0,
      "learning_rate": 5.719658471691977e-06,
      "loss": 4.7844,
      "num_input_tokens_seen": 42287072,
      "step": 31215
    },
    {
      "epoch": 0.9992958197298508,
      "grad_norm": 28.0,
      "learning_rate": 5.71341038448156e-06,
      "loss": 4.4176,
      "num_input_tokens_seen": 42294064,
      "step": 31220
    },
    {
      "epoch": 0.9994558607003393,
      "grad_norm": 18.0,
      "learning_rate": 5.707165271505635e-06,
      "loss": 4.1926,
      "num_input_tokens_seen": 42300832,
      "step": 31225
    },
    {
      "epoch": 0.9996159016708277,
      "grad_norm": 18.375,
      "learning_rate": 5.700923133727271e-06,
      "loss": 4.8138,
      "num_input_tokens_seen": 42308272,
      "step": 31230
    },
    {
      "epoch": 0.9997759426413162,
      "grad_norm": 25.625,
      "learning_rate": 5.694683972109083e-06,
      "loss": 4.0568,
      "num_input_tokens_seen": 42314800,
      "step": 31235
    },
    {
      "epoch": 0.9999359836118046,
      "grad_norm": 39.75,
      "learning_rate": 5.688447787613241e-06,
      "loss": 4.2695,
      "num_input_tokens_seen": 42321728,
      "step": 31240
    },
    {
      "epoch": 1.000096024582293,
      "grad_norm": 36.75,
      "learning_rate": 5.6822145812014285e-06,
      "loss": 4.3975,
      "num_input_tokens_seen": 42328064,
      "step": 31245
    },
    {
      "epoch": 1.0002560655527815,
      "grad_norm": 24.625,
      "learning_rate": 5.675984353834896e-06,
      "loss": 4.1932,
      "num_input_tokens_seen": 42334896,
      "step": 31250
    },
    {
      "epoch": 1.00041610652327,
      "grad_norm": 26.375,
      "learning_rate": 5.66975710647441e-06,
      "loss": 4.1835,
      "num_input_tokens_seen": 42341760,
      "step": 31255
    },
    {
      "epoch": 1.0005761474937584,
      "grad_norm": 19.5,
      "learning_rate": 5.663532840080304e-06,
      "loss": 4.0147,
      "num_input_tokens_seen": 42348448,
      "step": 31260
    },
    {
      "epoch": 1.000736188464247,
      "grad_norm": 33.5,
      "learning_rate": 5.6573115556124325e-06,
      "loss": 3.6232,
      "num_input_tokens_seen": 42355344,
      "step": 31265
    },
    {
      "epoch": 1.0008962294347352,
      "grad_norm": 23.375,
      "learning_rate": 5.651093254030185e-06,
      "loss": 4.0474,
      "num_input_tokens_seen": 42361968,
      "step": 31270
    },
    {
      "epoch": 1.0010562704052237,
      "grad_norm": 29.625,
      "learning_rate": 5.644877936292514e-06,
      "loss": 4.3961,
      "num_input_tokens_seen": 42368944,
      "step": 31275
    },
    {
      "epoch": 1.0012163113757122,
      "grad_norm": 28.25,
      "learning_rate": 5.638665603357901e-06,
      "loss": 4.1672,
      "num_input_tokens_seen": 42375600,
      "step": 31280
    },
    {
      "epoch": 1.0013763523462007,
      "grad_norm": 31.0,
      "learning_rate": 5.632456256184357e-06,
      "loss": 4.3291,
      "num_input_tokens_seen": 42382784,
      "step": 31285
    },
    {
      "epoch": 1.001536393316689,
      "grad_norm": 27.625,
      "learning_rate": 5.626249895729452e-06,
      "loss": 3.9637,
      "num_input_tokens_seen": 42389856,
      "step": 31290
    },
    {
      "epoch": 1.0016964342871775,
      "grad_norm": 22.25,
      "learning_rate": 5.620046522950273e-06,
      "loss": 4.4754,
      "num_input_tokens_seen": 42396384,
      "step": 31295
    },
    {
      "epoch": 1.001856475257666,
      "grad_norm": 42.25,
      "learning_rate": 5.613846138803464e-06,
      "loss": 5.5947,
      "num_input_tokens_seen": 42403072,
      "step": 31300
    },
    {
      "epoch": 1.0020165162281545,
      "grad_norm": 17.625,
      "learning_rate": 5.607648744245206e-06,
      "loss": 3.2184,
      "num_input_tokens_seen": 42409648,
      "step": 31305
    },
    {
      "epoch": 1.002176557198643,
      "grad_norm": 25.5,
      "learning_rate": 5.601454340231207e-06,
      "loss": 3.6858,
      "num_input_tokens_seen": 42416176,
      "step": 31310
    },
    {
      "epoch": 1.0023365981691312,
      "grad_norm": 21.125,
      "learning_rate": 5.595262927716724e-06,
      "loss": 4.4712,
      "num_input_tokens_seen": 42422976,
      "step": 31315
    },
    {
      "epoch": 1.0024966391396197,
      "grad_norm": 28.125,
      "learning_rate": 5.589074507656561e-06,
      "loss": 4.9155,
      "num_input_tokens_seen": 42429744,
      "step": 31320
    },
    {
      "epoch": 1.0026566801101082,
      "grad_norm": 15.625,
      "learning_rate": 5.582889081005044e-06,
      "loss": 4.3577,
      "num_input_tokens_seen": 42436432,
      "step": 31325
    },
    {
      "epoch": 1.0028167210805967,
      "grad_norm": 29.875,
      "learning_rate": 5.5767066487160316e-06,
      "loss": 4.2923,
      "num_input_tokens_seen": 42442816,
      "step": 31330
    },
    {
      "epoch": 1.002976762051085,
      "grad_norm": 30.875,
      "learning_rate": 5.570527211742949e-06,
      "loss": 4.347,
      "num_input_tokens_seen": 42449376,
      "step": 31335
    },
    {
      "epoch": 1.0031368030215735,
      "grad_norm": 26.5,
      "learning_rate": 5.564350771038731e-06,
      "loss": 4.7394,
      "num_input_tokens_seen": 42456240,
      "step": 31340
    },
    {
      "epoch": 1.003296843992062,
      "grad_norm": 30.25,
      "learning_rate": 5.558177327555875e-06,
      "loss": 4.6168,
      "num_input_tokens_seen": 42462656,
      "step": 31345
    },
    {
      "epoch": 1.0034568849625505,
      "grad_norm": 27.25,
      "learning_rate": 5.552006882246388e-06,
      "loss": 4.6688,
      "num_input_tokens_seen": 42469152,
      "step": 31350
    },
    {
      "epoch": 1.0036169259330388,
      "grad_norm": 27.125,
      "learning_rate": 5.545839436061839e-06,
      "loss": 4.9279,
      "num_input_tokens_seen": 42476048,
      "step": 31355
    },
    {
      "epoch": 1.0037769669035272,
      "grad_norm": 32.25,
      "learning_rate": 5.539674989953331e-06,
      "loss": 4.3942,
      "num_input_tokens_seen": 42482848,
      "step": 31360
    },
    {
      "epoch": 1.0039370078740157,
      "grad_norm": 24.75,
      "learning_rate": 5.533513544871488e-06,
      "loss": 3.5042,
      "num_input_tokens_seen": 42489888,
      "step": 31365
    },
    {
      "epoch": 1.0040970488445042,
      "grad_norm": 25.125,
      "learning_rate": 5.527355101766493e-06,
      "loss": 3.6478,
      "num_input_tokens_seen": 42496464,
      "step": 31370
    },
    {
      "epoch": 1.0042570898149927,
      "grad_norm": 27.375,
      "learning_rate": 5.521199661588044e-06,
      "loss": 5.293,
      "num_input_tokens_seen": 42502912,
      "step": 31375
    },
    {
      "epoch": 1.004417130785481,
      "grad_norm": 36.75,
      "learning_rate": 5.5150472252853944e-06,
      "loss": 4.0672,
      "num_input_tokens_seen": 42509520,
      "step": 31380
    },
    {
      "epoch": 1.0045771717559695,
      "grad_norm": 35.75,
      "learning_rate": 5.50889779380733e-06,
      "loss": 4.0153,
      "num_input_tokens_seen": 42515872,
      "step": 31385
    },
    {
      "epoch": 1.004737212726458,
      "grad_norm": 27.875,
      "learning_rate": 5.5027513681021605e-06,
      "loss": 4.0188,
      "num_input_tokens_seen": 42522352,
      "step": 31390
    },
    {
      "epoch": 1.0048972536969465,
      "grad_norm": 38.75,
      "learning_rate": 5.4966079491177545e-06,
      "loss": 3.4568,
      "num_input_tokens_seen": 42529104,
      "step": 31395
    },
    {
      "epoch": 1.0050572946674348,
      "grad_norm": 19.5,
      "learning_rate": 5.490467537801491e-06,
      "loss": 3.6653,
      "num_input_tokens_seen": 42536720,
      "step": 31400
    },
    {
      "epoch": 1.0050572946674348,
      "eval_loss": 4.477981090545654,
      "eval_runtime": 490.4641,
      "eval_samples_per_second": 28.312,
      "eval_steps_per_second": 14.156,
      "num_input_tokens_seen": 42536720,
      "step": 31400
    },
    {
      "epoch": 1.0052173356379233,
      "grad_norm": 30.75,
      "learning_rate": 5.484330135100313e-06,
      "loss": 4.9726,
      "num_input_tokens_seen": 42544096,
      "step": 31405
    },
    {
      "epoch": 1.0053773766084118,
      "grad_norm": 28.5,
      "learning_rate": 5.4781957419606785e-06,
      "loss": 3.4016,
      "num_input_tokens_seen": 42550832,
      "step": 31410
    },
    {
      "epoch": 1.0055374175789002,
      "grad_norm": 22.125,
      "learning_rate": 5.472064359328577e-06,
      "loss": 4.3055,
      "num_input_tokens_seen": 42558336,
      "step": 31415
    },
    {
      "epoch": 1.0056974585493887,
      "grad_norm": 40.0,
      "learning_rate": 5.4659359881495565e-06,
      "loss": 4.8319,
      "num_input_tokens_seen": 42565536,
      "step": 31420
    },
    {
      "epoch": 1.005857499519877,
      "grad_norm": 39.5,
      "learning_rate": 5.4598106293686916e-06,
      "loss": 4.4008,
      "num_input_tokens_seen": 42572208,
      "step": 31425
    },
    {
      "epoch": 1.0060175404903655,
      "grad_norm": 40.75,
      "learning_rate": 5.45368828393058e-06,
      "loss": 4.7579,
      "num_input_tokens_seen": 42578816,
      "step": 31430
    },
    {
      "epoch": 1.006177581460854,
      "grad_norm": 19.25,
      "learning_rate": 5.44756895277937e-06,
      "loss": 3.7968,
      "num_input_tokens_seen": 42585520,
      "step": 31435
    },
    {
      "epoch": 1.0063376224313425,
      "grad_norm": 31.875,
      "learning_rate": 5.441452636858746e-06,
      "loss": 4.1853,
      "num_input_tokens_seen": 42592192,
      "step": 31440
    },
    {
      "epoch": 1.0064976634018308,
      "grad_norm": 27.375,
      "learning_rate": 5.435339337111905e-06,
      "loss": 5.0703,
      "num_input_tokens_seen": 42599040,
      "step": 31445
    },
    {
      "epoch": 1.0066577043723193,
      "grad_norm": 33.0,
      "learning_rate": 5.42922905448161e-06,
      "loss": 4.4278,
      "num_input_tokens_seen": 42605856,
      "step": 31450
    },
    {
      "epoch": 1.0068177453428078,
      "grad_norm": 37.0,
      "learning_rate": 5.423121789910129e-06,
      "loss": 4.1326,
      "num_input_tokens_seen": 42612752,
      "step": 31455
    },
    {
      "epoch": 1.0069777863132963,
      "grad_norm": 31.0,
      "learning_rate": 5.417017544339287e-06,
      "loss": 4.595,
      "num_input_tokens_seen": 42620032,
      "step": 31460
    },
    {
      "epoch": 1.0071378272837848,
      "grad_norm": 33.5,
      "learning_rate": 5.410916318710443e-06,
      "loss": 5.0747,
      "num_input_tokens_seen": 42626992,
      "step": 31465
    },
    {
      "epoch": 1.007297868254273,
      "grad_norm": 33.5,
      "learning_rate": 5.404818113964466e-06,
      "loss": 3.6337,
      "num_input_tokens_seen": 42634016,
      "step": 31470
    },
    {
      "epoch": 1.0074579092247615,
      "grad_norm": 25.625,
      "learning_rate": 5.398722931041792e-06,
      "loss": 4.3093,
      "num_input_tokens_seen": 42640784,
      "step": 31475
    },
    {
      "epoch": 1.00761795019525,
      "grad_norm": 27.5,
      "learning_rate": 5.392630770882367e-06,
      "loss": 3.8345,
      "num_input_tokens_seen": 42648080,
      "step": 31480
    },
    {
      "epoch": 1.0077779911657385,
      "grad_norm": 22.0,
      "learning_rate": 5.3865416344256705e-06,
      "loss": 4.3131,
      "num_input_tokens_seen": 42655072,
      "step": 31485
    },
    {
      "epoch": 1.0079380321362268,
      "grad_norm": 24.625,
      "learning_rate": 5.380455522610742e-06,
      "loss": 4.3277,
      "num_input_tokens_seen": 42661728,
      "step": 31490
    },
    {
      "epoch": 1.0080980731067153,
      "grad_norm": 26.0,
      "learning_rate": 5.374372436376116e-06,
      "loss": 3.6086,
      "num_input_tokens_seen": 42668160,
      "step": 31495
    },
    {
      "epoch": 1.0082581140772038,
      "grad_norm": 33.0,
      "learning_rate": 5.368292376659895e-06,
      "loss": 3.9544,
      "num_input_tokens_seen": 42674736,
      "step": 31500
    },
    {
      "epoch": 1.0084181550476923,
      "grad_norm": 20.375,
      "learning_rate": 5.362215344399701e-06,
      "loss": 3.6292,
      "num_input_tokens_seen": 42681648,
      "step": 31505
    },
    {
      "epoch": 1.0085781960181806,
      "grad_norm": 22.25,
      "learning_rate": 5.356141340532678e-06,
      "loss": 4.1726,
      "num_input_tokens_seen": 42688176,
      "step": 31510
    },
    {
      "epoch": 1.008738236988669,
      "grad_norm": 28.5,
      "learning_rate": 5.350070365995522e-06,
      "loss": 3.5737,
      "num_input_tokens_seen": 42695360,
      "step": 31515
    },
    {
      "epoch": 1.0088982779591575,
      "grad_norm": 25.625,
      "learning_rate": 5.344002421724459e-06,
      "loss": 4.636,
      "num_input_tokens_seen": 42701952,
      "step": 31520
    },
    {
      "epoch": 1.009058318929646,
      "grad_norm": 24.5,
      "learning_rate": 5.337937508655228e-06,
      "loss": 4.5778,
      "num_input_tokens_seen": 42708752,
      "step": 31525
    },
    {
      "epoch": 1.0092183599001345,
      "grad_norm": 22.625,
      "learning_rate": 5.331875627723126e-06,
      "loss": 4.7448,
      "num_input_tokens_seen": 42715488,
      "step": 31530
    },
    {
      "epoch": 1.0093784008706228,
      "grad_norm": 22.5,
      "learning_rate": 5.325816779862963e-06,
      "loss": 4.1619,
      "num_input_tokens_seen": 42722144,
      "step": 31535
    },
    {
      "epoch": 1.0095384418411113,
      "grad_norm": 27.75,
      "learning_rate": 5.319760966009102e-06,
      "loss": 3.6096,
      "num_input_tokens_seen": 42729024,
      "step": 31540
    },
    {
      "epoch": 1.0096984828115998,
      "grad_norm": 24.75,
      "learning_rate": 5.3137081870954096e-06,
      "loss": 3.3211,
      "num_input_tokens_seen": 42735712,
      "step": 31545
    },
    {
      "epoch": 1.0098585237820883,
      "grad_norm": 20.125,
      "learning_rate": 5.307658444055313e-06,
      "loss": 4.1579,
      "num_input_tokens_seen": 42742448,
      "step": 31550
    },
    {
      "epoch": 1.0100185647525766,
      "grad_norm": 20.125,
      "learning_rate": 5.301611737821749e-06,
      "loss": 4.4395,
      "num_input_tokens_seen": 42749456,
      "step": 31555
    },
    {
      "epoch": 1.010178605723065,
      "grad_norm": 65.5,
      "learning_rate": 5.295568069327206e-06,
      "loss": 2.973,
      "num_input_tokens_seen": 42756544,
      "step": 31560
    },
    {
      "epoch": 1.0103386466935536,
      "grad_norm": 25.75,
      "learning_rate": 5.289527439503683e-06,
      "loss": 3.768,
      "num_input_tokens_seen": 42763216,
      "step": 31565
    },
    {
      "epoch": 1.010498687664042,
      "grad_norm": 19.25,
      "learning_rate": 5.28348984928273e-06,
      "loss": 4.3913,
      "num_input_tokens_seen": 42769968,
      "step": 31570
    },
    {
      "epoch": 1.0106587286345305,
      "grad_norm": 26.125,
      "learning_rate": 5.27745529959541e-06,
      "loss": 4.1934,
      "num_input_tokens_seen": 42776416,
      "step": 31575
    },
    {
      "epoch": 1.0108187696050188,
      "grad_norm": 27.125,
      "learning_rate": 5.271423791372335e-06,
      "loss": 4.3999,
      "num_input_tokens_seen": 42783392,
      "step": 31580
    },
    {
      "epoch": 1.0109788105755073,
      "grad_norm": 26.125,
      "learning_rate": 5.26539532554364e-06,
      "loss": 3.5814,
      "num_input_tokens_seen": 42790128,
      "step": 31585
    },
    {
      "epoch": 1.0111388515459958,
      "grad_norm": 37.75,
      "learning_rate": 5.25936990303898e-06,
      "loss": 5.7195,
      "num_input_tokens_seen": 42797344,
      "step": 31590
    },
    {
      "epoch": 1.0112988925164843,
      "grad_norm": 29.375,
      "learning_rate": 5.253347524787555e-06,
      "loss": 4.5492,
      "num_input_tokens_seen": 42804080,
      "step": 31595
    },
    {
      "epoch": 1.0114589334869726,
      "grad_norm": 36.25,
      "learning_rate": 5.2473281917181035e-06,
      "loss": 5.1739,
      "num_input_tokens_seen": 42810528,
      "step": 31600
    },
    {
      "epoch": 1.0114589334869726,
      "eval_loss": 4.490750789642334,
      "eval_runtime": 491.0932,
      "eval_samples_per_second": 28.276,
      "eval_steps_per_second": 14.138,
      "num_input_tokens_seen": 42810528,
      "step": 31600
    },
    {
      "epoch": 1.011618974457461,
      "grad_norm": 22.0,
      "learning_rate": 5.241311904758864e-06,
      "loss": 4.0924,
      "num_input_tokens_seen": 42817248,
      "step": 31605
    },
    {
      "epoch": 1.0117790154279496,
      "grad_norm": 33.75,
      "learning_rate": 5.23529866483764e-06,
      "loss": 4.7294,
      "num_input_tokens_seen": 42823936,
      "step": 31610
    },
    {
      "epoch": 1.011939056398438,
      "grad_norm": 15.0,
      "learning_rate": 5.229288472881732e-06,
      "loss": 3.4805,
      "num_input_tokens_seen": 42831072,
      "step": 31615
    },
    {
      "epoch": 1.0120990973689263,
      "grad_norm": 29.875,
      "learning_rate": 5.2232813298180025e-06,
      "loss": 4.4133,
      "num_input_tokens_seen": 42837840,
      "step": 31620
    },
    {
      "epoch": 1.0122591383394148,
      "grad_norm": 57.0,
      "learning_rate": 5.217277236572824e-06,
      "loss": 5.6134,
      "num_input_tokens_seen": 42844736,
      "step": 31625
    },
    {
      "epoch": 1.0124191793099033,
      "grad_norm": 35.25,
      "learning_rate": 5.211276194072093e-06,
      "loss": 4.5219,
      "num_input_tokens_seen": 42851536,
      "step": 31630
    },
    {
      "epoch": 1.0125792202803918,
      "grad_norm": 18.75,
      "learning_rate": 5.205278203241254e-06,
      "loss": 3.707,
      "num_input_tokens_seen": 42858144,
      "step": 31635
    },
    {
      "epoch": 1.0127392612508803,
      "grad_norm": 33.0,
      "learning_rate": 5.199283265005278e-06,
      "loss": 4.537,
      "num_input_tokens_seen": 42864576,
      "step": 31640
    },
    {
      "epoch": 1.0128993022213686,
      "grad_norm": 36.75,
      "learning_rate": 5.193291380288648e-06,
      "loss": 4.356,
      "num_input_tokens_seen": 42871280,
      "step": 31645
    },
    {
      "epoch": 1.013059343191857,
      "grad_norm": 14.4375,
      "learning_rate": 5.1873025500153995e-06,
      "loss": 4.3636,
      "num_input_tokens_seen": 42878272,
      "step": 31650
    },
    {
      "epoch": 1.0132193841623456,
      "grad_norm": 18.75,
      "learning_rate": 5.181316775109071e-06,
      "loss": 3.0514,
      "num_input_tokens_seen": 42884976,
      "step": 31655
    },
    {
      "epoch": 1.013379425132834,
      "grad_norm": 24.0,
      "learning_rate": 5.1753340564927564e-06,
      "loss": 4.5982,
      "num_input_tokens_seen": 42891888,
      "step": 31660
    },
    {
      "epoch": 1.0135394661033224,
      "grad_norm": 39.25,
      "learning_rate": 5.169354395089068e-06,
      "loss": 4.2933,
      "num_input_tokens_seen": 42898352,
      "step": 31665
    },
    {
      "epoch": 1.0136995070738108,
      "grad_norm": 23.375,
      "learning_rate": 5.1633777918201346e-06,
      "loss": 3.7384,
      "num_input_tokens_seen": 42905296,
      "step": 31670
    },
    {
      "epoch": 1.0138595480442993,
      "grad_norm": 40.5,
      "learning_rate": 5.157404247607625e-06,
      "loss": 3.9827,
      "num_input_tokens_seen": 42911648,
      "step": 31675
    },
    {
      "epoch": 1.0140195890147878,
      "grad_norm": 25.375,
      "learning_rate": 5.1514337633727454e-06,
      "loss": 4.5061,
      "num_input_tokens_seen": 42918256,
      "step": 31680
    },
    {
      "epoch": 1.0141796299852763,
      "grad_norm": 24.75,
      "learning_rate": 5.145466340036206e-06,
      "loss": 4.6643,
      "num_input_tokens_seen": 42924880,
      "step": 31685
    },
    {
      "epoch": 1.0143396709557646,
      "grad_norm": 22.75,
      "learning_rate": 5.139501978518274e-06,
      "loss": 4.8727,
      "num_input_tokens_seen": 42931328,
      "step": 31690
    },
    {
      "epoch": 1.014499711926253,
      "grad_norm": 39.0,
      "learning_rate": 5.133540679738716e-06,
      "loss": 3.9452,
      "num_input_tokens_seen": 42938368,
      "step": 31695
    },
    {
      "epoch": 1.0146597528967416,
      "grad_norm": 33.25,
      "learning_rate": 5.127582444616838e-06,
      "loss": 4.7383,
      "num_input_tokens_seen": 42944848,
      "step": 31700
    },
    {
      "epoch": 1.01481979386723,
      "grad_norm": 39.25,
      "learning_rate": 5.121627274071486e-06,
      "loss": 4.9857,
      "num_input_tokens_seen": 42951440,
      "step": 31705
    },
    {
      "epoch": 1.0149798348377184,
      "grad_norm": 33.5,
      "learning_rate": 5.115675169021009e-06,
      "loss": 4.9369,
      "num_input_tokens_seen": 42958336,
      "step": 31710
    },
    {
      "epoch": 1.0151398758082069,
      "grad_norm": 24.5,
      "learning_rate": 5.1097261303832994e-06,
      "loss": 4.7977,
      "num_input_tokens_seen": 42965312,
      "step": 31715
    },
    {
      "epoch": 1.0152999167786954,
      "grad_norm": 18.375,
      "learning_rate": 5.103780159075788e-06,
      "loss": 3.7639,
      "num_input_tokens_seen": 42971968,
      "step": 31720
    },
    {
      "epoch": 1.0154599577491839,
      "grad_norm": 30.5,
      "learning_rate": 5.0978372560154e-06,
      "loss": 4.7013,
      "num_input_tokens_seen": 42979040,
      "step": 31725
    },
    {
      "epoch": 1.0156199987196723,
      "grad_norm": 42.25,
      "learning_rate": 5.091897422118619e-06,
      "loss": 4.4421,
      "num_input_tokens_seen": 42986048,
      "step": 31730
    },
    {
      "epoch": 1.0157800396901606,
      "grad_norm": 20.0,
      "learning_rate": 5.0859606583014305e-06,
      "loss": 2.8098,
      "num_input_tokens_seen": 42992880,
      "step": 31735
    },
    {
      "epoch": 1.0159400806606491,
      "grad_norm": 35.75,
      "learning_rate": 5.080026965479365e-06,
      "loss": 4.434,
      "num_input_tokens_seen": 42999392,
      "step": 31740
    },
    {
      "epoch": 1.0161001216311376,
      "grad_norm": 29.5,
      "learning_rate": 5.074096344567475e-06,
      "loss": 4.2798,
      "num_input_tokens_seen": 43006272,
      "step": 31745
    },
    {
      "epoch": 1.016260162601626,
      "grad_norm": 22.25,
      "learning_rate": 5.0681687964803294e-06,
      "loss": 3.731,
      "num_input_tokens_seen": 43013328,
      "step": 31750
    },
    {
      "epoch": 1.0164202035721144,
      "grad_norm": 19.375,
      "learning_rate": 5.06224432213204e-06,
      "loss": 3.2928,
      "num_input_tokens_seen": 43019856,
      "step": 31755
    },
    {
      "epoch": 1.0165802445426029,
      "grad_norm": 24.875,
      "learning_rate": 5.056322922436224e-06,
      "loss": 4.6412,
      "num_input_tokens_seen": 43026512,
      "step": 31760
    },
    {
      "epoch": 1.0167402855130914,
      "grad_norm": 35.0,
      "learning_rate": 5.0504045983060465e-06,
      "loss": 5.0985,
      "num_input_tokens_seen": 43033248,
      "step": 31765
    },
    {
      "epoch": 1.0169003264835799,
      "grad_norm": 26.75,
      "learning_rate": 5.044489350654183e-06,
      "loss": 4.4965,
      "num_input_tokens_seen": 43039952,
      "step": 31770
    },
    {
      "epoch": 1.0170603674540681,
      "grad_norm": 41.5,
      "learning_rate": 5.038577180392831e-06,
      "loss": 4.7601,
      "num_input_tokens_seen": 43047024,
      "step": 31775
    },
    {
      "epoch": 1.0172204084245566,
      "grad_norm": 20.25,
      "learning_rate": 5.032668088433729e-06,
      "loss": 5.0415,
      "num_input_tokens_seen": 43053456,
      "step": 31780
    },
    {
      "epoch": 1.0173804493950451,
      "grad_norm": 41.25,
      "learning_rate": 5.02676207568814e-06,
      "loss": 6.662,
      "num_input_tokens_seen": 43060416,
      "step": 31785
    },
    {
      "epoch": 1.0175404903655336,
      "grad_norm": 18.5,
      "learning_rate": 5.02085914306683e-06,
      "loss": 4.2281,
      "num_input_tokens_seen": 43066944,
      "step": 31790
    },
    {
      "epoch": 1.0177005313360221,
      "grad_norm": 32.5,
      "learning_rate": 5.014959291480123e-06,
      "loss": 4.566,
      "num_input_tokens_seen": 43074432,
      "step": 31795
    },
    {
      "epoch": 1.0178605723065104,
      "grad_norm": 22.375,
      "learning_rate": 5.009062521837835e-06,
      "loss": 4.4113,
      "num_input_tokens_seen": 43081488,
      "step": 31800
    },
    {
      "epoch": 1.0178605723065104,
      "eval_loss": 4.480593204498291,
      "eval_runtime": 490.4937,
      "eval_samples_per_second": 28.31,
      "eval_steps_per_second": 14.155,
      "num_input_tokens_seen": 43081488,
      "step": 31800
    },
    {
      "epoch": 1.0180206132769989,
      "grad_norm": 42.25,
      "learning_rate": 5.003168835049324e-06,
      "loss": 4.1295,
      "num_input_tokens_seen": 43088224,
      "step": 31805
    },
    {
      "epoch": 1.0181806542474874,
      "grad_norm": 14.0,
      "learning_rate": 4.997278232023483e-06,
      "loss": 4.0621,
      "num_input_tokens_seen": 43094912,
      "step": 31810
    },
    {
      "epoch": 1.0183406952179759,
      "grad_norm": 32.0,
      "learning_rate": 4.9913907136687036e-06,
      "loss": 4.3705,
      "num_input_tokens_seen": 43102064,
      "step": 31815
    },
    {
      "epoch": 1.0185007361884642,
      "grad_norm": 39.0,
      "learning_rate": 4.985506280892918e-06,
      "loss": 4.9619,
      "num_input_tokens_seen": 43109120,
      "step": 31820
    },
    {
      "epoch": 1.0186607771589526,
      "grad_norm": 48.0,
      "learning_rate": 4.979624934603589e-06,
      "loss": 5.1883,
      "num_input_tokens_seen": 43115696,
      "step": 31825
    },
    {
      "epoch": 1.0188208181294411,
      "grad_norm": 25.75,
      "learning_rate": 4.97374667570768e-06,
      "loss": 4.7347,
      "num_input_tokens_seen": 43122272,
      "step": 31830
    },
    {
      "epoch": 1.0189808590999296,
      "grad_norm": 15.875,
      "learning_rate": 4.967871505111704e-06,
      "loss": 4.2992,
      "num_input_tokens_seen": 43129168,
      "step": 31835
    },
    {
      "epoch": 1.0191409000704181,
      "grad_norm": 24.25,
      "learning_rate": 4.961999423721686e-06,
      "loss": 3.8823,
      "num_input_tokens_seen": 43135888,
      "step": 31840
    },
    {
      "epoch": 1.0193009410409064,
      "grad_norm": 35.25,
      "learning_rate": 4.956130432443159e-06,
      "loss": 3.8615,
      "num_input_tokens_seen": 43142352,
      "step": 31845
    },
    {
      "epoch": 1.019460982011395,
      "grad_norm": 21.625,
      "learning_rate": 4.950264532181215e-06,
      "loss": 4.1541,
      "num_input_tokens_seen": 43148912,
      "step": 31850
    },
    {
      "epoch": 1.0196210229818834,
      "grad_norm": 17.25,
      "learning_rate": 4.944401723840433e-06,
      "loss": 4.1885,
      "num_input_tokens_seen": 43155392,
      "step": 31855
    },
    {
      "epoch": 1.019781063952372,
      "grad_norm": 49.25,
      "learning_rate": 4.938542008324942e-06,
      "loss": 5.5632,
      "num_input_tokens_seen": 43162288,
      "step": 31860
    },
    {
      "epoch": 1.0199411049228602,
      "grad_norm": 41.75,
      "learning_rate": 4.9326853865383855e-06,
      "loss": 5.236,
      "num_input_tokens_seen": 43169280,
      "step": 31865
    },
    {
      "epoch": 1.0201011458933487,
      "grad_norm": 50.0,
      "learning_rate": 4.926831859383918e-06,
      "loss": 4.5373,
      "num_input_tokens_seen": 43175872,
      "step": 31870
    },
    {
      "epoch": 1.0202611868638372,
      "grad_norm": 26.0,
      "learning_rate": 4.92098142776424e-06,
      "loss": 4.4486,
      "num_input_tokens_seen": 43182544,
      "step": 31875
    },
    {
      "epoch": 1.0204212278343257,
      "grad_norm": 33.75,
      "learning_rate": 4.91513409258155e-06,
      "loss": 4.3941,
      "num_input_tokens_seen": 43189200,
      "step": 31880
    },
    {
      "epoch": 1.020581268804814,
      "grad_norm": 40.5,
      "learning_rate": 4.909289854737581e-06,
      "loss": 3.4955,
      "num_input_tokens_seen": 43196064,
      "step": 31885
    },
    {
      "epoch": 1.0207413097753024,
      "grad_norm": 26.75,
      "learning_rate": 4.903448715133602e-06,
      "loss": 3.7862,
      "num_input_tokens_seen": 43203024,
      "step": 31890
    },
    {
      "epoch": 1.020901350745791,
      "grad_norm": 21.125,
      "learning_rate": 4.897610674670372e-06,
      "loss": 4.5641,
      "num_input_tokens_seen": 43209600,
      "step": 31895
    },
    {
      "epoch": 1.0210613917162794,
      "grad_norm": 35.5,
      "learning_rate": 4.8917757342482e-06,
      "loss": 4.9797,
      "num_input_tokens_seen": 43216416,
      "step": 31900
    },
    {
      "epoch": 1.021221432686768,
      "grad_norm": 16.625,
      "learning_rate": 4.885943894766909e-06,
      "loss": 4.2097,
      "num_input_tokens_seen": 43223936,
      "step": 31905
    },
    {
      "epoch": 1.0213814736572562,
      "grad_norm": 18.375,
      "learning_rate": 4.880115157125842e-06,
      "loss": 3.5941,
      "num_input_tokens_seen": 43230592,
      "step": 31910
    },
    {
      "epoch": 1.0215415146277447,
      "grad_norm": 22.375,
      "learning_rate": 4.874289522223857e-06,
      "loss": 3.7871,
      "num_input_tokens_seen": 43237568,
      "step": 31915
    },
    {
      "epoch": 1.0217015555982332,
      "grad_norm": 46.75,
      "learning_rate": 4.868466990959339e-06,
      "loss": 4.8483,
      "num_input_tokens_seen": 43245024,
      "step": 31920
    },
    {
      "epoch": 1.0218615965687217,
      "grad_norm": 28.125,
      "learning_rate": 4.8626475642301964e-06,
      "loss": 4.5614,
      "num_input_tokens_seen": 43251328,
      "step": 31925
    },
    {
      "epoch": 1.02202163753921,
      "grad_norm": 32.0,
      "learning_rate": 4.856831242933871e-06,
      "loss": 4.7818,
      "num_input_tokens_seen": 43258288,
      "step": 31930
    },
    {
      "epoch": 1.0221816785096984,
      "grad_norm": 53.75,
      "learning_rate": 4.851018027967294e-06,
      "loss": 5.0961,
      "num_input_tokens_seen": 43265040,
      "step": 31935
    },
    {
      "epoch": 1.022341719480187,
      "grad_norm": 26.875,
      "learning_rate": 4.845207920226946e-06,
      "loss": 4.4177,
      "num_input_tokens_seen": 43271520,
      "step": 31940
    },
    {
      "epoch": 1.0225017604506754,
      "grad_norm": 23.625,
      "learning_rate": 4.839400920608825e-06,
      "loss": 4.329,
      "num_input_tokens_seen": 43278384,
      "step": 31945
    },
    {
      "epoch": 1.022661801421164,
      "grad_norm": 53.5,
      "learning_rate": 4.83359703000843e-06,
      "loss": 3.7809,
      "num_input_tokens_seen": 43285024,
      "step": 31950
    },
    {
      "epoch": 1.0228218423916522,
      "grad_norm": 18.25,
      "learning_rate": 4.827796249320804e-06,
      "loss": 4.0931,
      "num_input_tokens_seen": 43291184,
      "step": 31955
    },
    {
      "epoch": 1.0229818833621407,
      "grad_norm": 30.625,
      "learning_rate": 4.82199857944049e-06,
      "loss": 4.2055,
      "num_input_tokens_seen": 43297616,
      "step": 31960
    },
    {
      "epoch": 1.0231419243326292,
      "grad_norm": 27.375,
      "learning_rate": 4.8162040212615695e-06,
      "loss": 4.3851,
      "num_input_tokens_seen": 43304592,
      "step": 31965
    },
    {
      "epoch": 1.0233019653031177,
      "grad_norm": 21.375,
      "learning_rate": 4.810412575677639e-06,
      "loss": 4.1668,
      "num_input_tokens_seen": 43311664,
      "step": 31970
    },
    {
      "epoch": 1.023462006273606,
      "grad_norm": 31.75,
      "learning_rate": 4.804624243581801e-06,
      "loss": 4.0885,
      "num_input_tokens_seen": 43318320,
      "step": 31975
    },
    {
      "epoch": 1.0236220472440944,
      "grad_norm": 38.75,
      "learning_rate": 4.798839025866703e-06,
      "loss": 3.9262,
      "num_input_tokens_seen": 43325552,
      "step": 31980
    },
    {
      "epoch": 1.023782088214583,
      "grad_norm": 29.5,
      "learning_rate": 4.793056923424491e-06,
      "loss": 3.7884,
      "num_input_tokens_seen": 43331920,
      "step": 31985
    },
    {
      "epoch": 1.0239421291850714,
      "grad_norm": 29.25,
      "learning_rate": 4.78727793714683e-06,
      "loss": 4.5647,
      "num_input_tokens_seen": 43338400,
      "step": 31990
    },
    {
      "epoch": 1.02410217015556,
      "grad_norm": 33.25,
      "learning_rate": 4.7815020679249285e-06,
      "loss": 3.169,
      "num_input_tokens_seen": 43345296,
      "step": 31995
    },
    {
      "epoch": 1.0242622111260482,
      "grad_norm": 45.75,
      "learning_rate": 4.775729316649483e-06,
      "loss": 4.0525,
      "num_input_tokens_seen": 43351904,
      "step": 32000
    },
    {
      "epoch": 1.0242622111260482,
      "eval_loss": 4.4965596199035645,
      "eval_runtime": 491.1386,
      "eval_samples_per_second": 28.273,
      "eval_steps_per_second": 14.137,
      "num_input_tokens_seen": 43351904,
      "step": 32000
    },
    {
      "epoch": 1.0244222520965367,
      "grad_norm": 32.5,
      "learning_rate": 4.769959684210728e-06,
      "loss": 4.6228,
      "num_input_tokens_seen": 43358480,
      "step": 32005
    },
    {
      "epoch": 1.0245822930670252,
      "grad_norm": 26.25,
      "learning_rate": 4.764193171498426e-06,
      "loss": 3.7897,
      "num_input_tokens_seen": 43365072,
      "step": 32010
    },
    {
      "epoch": 1.0247423340375137,
      "grad_norm": 23.25,
      "learning_rate": 4.75842977940183e-06,
      "loss": 4.1388,
      "num_input_tokens_seen": 43372032,
      "step": 32015
    },
    {
      "epoch": 1.024902375008002,
      "grad_norm": 32.5,
      "learning_rate": 4.752669508809729e-06,
      "loss": 4.5174,
      "num_input_tokens_seen": 43378816,
      "step": 32020
    },
    {
      "epoch": 1.0250624159784905,
      "grad_norm": 34.5,
      "learning_rate": 4.746912360610445e-06,
      "loss": 4.1892,
      "num_input_tokens_seen": 43385456,
      "step": 32025
    },
    {
      "epoch": 1.025222456948979,
      "grad_norm": 29.875,
      "learning_rate": 4.741158335691781e-06,
      "loss": 4.6604,
      "num_input_tokens_seen": 43392096,
      "step": 32030
    },
    {
      "epoch": 1.0253824979194675,
      "grad_norm": 19.75,
      "learning_rate": 4.7354074349410994e-06,
      "loss": 3.9235,
      "num_input_tokens_seen": 43398848,
      "step": 32035
    },
    {
      "epoch": 1.0255425388899557,
      "grad_norm": 21.25,
      "learning_rate": 4.729659659245245e-06,
      "loss": 3.5004,
      "num_input_tokens_seen": 43405376,
      "step": 32040
    },
    {
      "epoch": 1.0257025798604442,
      "grad_norm": 39.0,
      "learning_rate": 4.723915009490601e-06,
      "loss": 3.7726,
      "num_input_tokens_seen": 43412176,
      "step": 32045
    },
    {
      "epoch": 1.0258626208309327,
      "grad_norm": 31.25,
      "learning_rate": 4.718173486563077e-06,
      "loss": 4.2546,
      "num_input_tokens_seen": 43418992,
      "step": 32050
    },
    {
      "epoch": 1.0260226618014212,
      "grad_norm": 30.25,
      "learning_rate": 4.71243509134808e-06,
      "loss": 4.0854,
      "num_input_tokens_seen": 43425328,
      "step": 32055
    },
    {
      "epoch": 1.0261827027719097,
      "grad_norm": 49.0,
      "learning_rate": 4.706699824730532e-06,
      "loss": 4.894,
      "num_input_tokens_seen": 43432368,
      "step": 32060
    },
    {
      "epoch": 1.026342743742398,
      "grad_norm": 36.5,
      "learning_rate": 4.700967687594901e-06,
      "loss": 4.8092,
      "num_input_tokens_seen": 43439248,
      "step": 32065
    },
    {
      "epoch": 1.0265027847128865,
      "grad_norm": 26.125,
      "learning_rate": 4.69523868082514e-06,
      "loss": 5.4213,
      "num_input_tokens_seen": 43445824,
      "step": 32070
    },
    {
      "epoch": 1.026662825683375,
      "grad_norm": 24.625,
      "learning_rate": 4.689512805304747e-06,
      "loss": 5.0054,
      "num_input_tokens_seen": 43452448,
      "step": 32075
    },
    {
      "epoch": 1.0268228666538635,
      "grad_norm": 19.375,
      "learning_rate": 4.683790061916707e-06,
      "loss": 3.4287,
      "num_input_tokens_seen": 43459600,
      "step": 32080
    },
    {
      "epoch": 1.0269829076243517,
      "grad_norm": 32.75,
      "learning_rate": 4.678070451543551e-06,
      "loss": 4.8935,
      "num_input_tokens_seen": 43466144,
      "step": 32085
    },
    {
      "epoch": 1.0271429485948402,
      "grad_norm": 28.375,
      "learning_rate": 4.6723539750673204e-06,
      "loss": 4.2109,
      "num_input_tokens_seen": 43472688,
      "step": 32090
    },
    {
      "epoch": 1.0273029895653287,
      "grad_norm": 41.0,
      "learning_rate": 4.666640633369551e-06,
      "loss": 4.8371,
      "num_input_tokens_seen": 43479248,
      "step": 32095
    },
    {
      "epoch": 1.0274630305358172,
      "grad_norm": 29.5,
      "learning_rate": 4.660930427331323e-06,
      "loss": 4.1583,
      "num_input_tokens_seen": 43485984,
      "step": 32100
    },
    {
      "epoch": 1.0276230715063057,
      "grad_norm": 34.5,
      "learning_rate": 4.6552233578332244e-06,
      "loss": 4.3486,
      "num_input_tokens_seen": 43492848,
      "step": 32105
    },
    {
      "epoch": 1.027783112476794,
      "grad_norm": 36.25,
      "learning_rate": 4.649519425755347e-06,
      "loss": 4.719,
      "num_input_tokens_seen": 43499312,
      "step": 32110
    },
    {
      "epoch": 1.0279431534472825,
      "grad_norm": 23.25,
      "learning_rate": 4.64381863197732e-06,
      "loss": 4.0477,
      "num_input_tokens_seen": 43506320,
      "step": 32115
    },
    {
      "epoch": 1.028103194417771,
      "grad_norm": 28.25,
      "learning_rate": 4.638120977378269e-06,
      "loss": 5.1573,
      "num_input_tokens_seen": 43512864,
      "step": 32120
    },
    {
      "epoch": 1.0282632353882595,
      "grad_norm": 30.125,
      "learning_rate": 4.632426462836848e-06,
      "loss": 4.9763,
      "num_input_tokens_seen": 43520064,
      "step": 32125
    },
    {
      "epoch": 1.0284232763587478,
      "grad_norm": 34.0,
      "learning_rate": 4.626735089231224e-06,
      "loss": 5.0516,
      "num_input_tokens_seen": 43526528,
      "step": 32130
    },
    {
      "epoch": 1.0285833173292362,
      "grad_norm": 38.0,
      "learning_rate": 4.621046857439068e-06,
      "loss": 3.9741,
      "num_input_tokens_seen": 43533216,
      "step": 32135
    },
    {
      "epoch": 1.0287433582997247,
      "grad_norm": 25.625,
      "learning_rate": 4.615361768337587e-06,
      "loss": 5.1905,
      "num_input_tokens_seen": 43540976,
      "step": 32140
    },
    {
      "epoch": 1.0289033992702132,
      "grad_norm": 26.75,
      "learning_rate": 4.6096798228034946e-06,
      "loss": 4.5715,
      "num_input_tokens_seen": 43547680,
      "step": 32145
    },
    {
      "epoch": 1.0290634402407015,
      "grad_norm": 36.25,
      "learning_rate": 4.604001021713008e-06,
      "loss": 5.1968,
      "num_input_tokens_seen": 43554960,
      "step": 32150
    },
    {
      "epoch": 1.02922348121119,
      "grad_norm": 35.25,
      "learning_rate": 4.598325365941883e-06,
      "loss": 5.4157,
      "num_input_tokens_seen": 43561664,
      "step": 32155
    },
    {
      "epoch": 1.0293835221816785,
      "grad_norm": 24.875,
      "learning_rate": 4.5926528563653645e-06,
      "loss": 3.0095,
      "num_input_tokens_seen": 43568320,
      "step": 32160
    },
    {
      "epoch": 1.029543563152167,
      "grad_norm": 24.25,
      "learning_rate": 4.5869834938582295e-06,
      "loss": 4.5354,
      "num_input_tokens_seen": 43575024,
      "step": 32165
    },
    {
      "epoch": 1.0297036041226555,
      "grad_norm": 33.75,
      "learning_rate": 4.581317279294772e-06,
      "loss": 4.9662,
      "num_input_tokens_seen": 43582032,
      "step": 32170
    },
    {
      "epoch": 1.0298636450931438,
      "grad_norm": 43.0,
      "learning_rate": 4.57565421354878e-06,
      "loss": 5.1014,
      "num_input_tokens_seen": 43588752,
      "step": 32175
    },
    {
      "epoch": 1.0300236860636323,
      "grad_norm": 21.0,
      "learning_rate": 4.569994297493579e-06,
      "loss": 4.8472,
      "num_input_tokens_seen": 43595680,
      "step": 32180
    },
    {
      "epoch": 1.0301837270341208,
      "grad_norm": 34.25,
      "learning_rate": 4.564337532002002e-06,
      "loss": 4.336,
      "num_input_tokens_seen": 43602416,
      "step": 32185
    },
    {
      "epoch": 1.0303437680046093,
      "grad_norm": 24.75,
      "learning_rate": 4.55868391794638e-06,
      "loss": 5.1542,
      "num_input_tokens_seen": 43609440,
      "step": 32190
    },
    {
      "epoch": 1.0305038089750975,
      "grad_norm": 32.0,
      "learning_rate": 4.553033456198588e-06,
      "loss": 4.4567,
      "num_input_tokens_seen": 43615920,
      "step": 32195
    },
    {
      "epoch": 1.030663849945586,
      "grad_norm": 75.0,
      "learning_rate": 4.54738614762999e-06,
      "loss": 4.5779,
      "num_input_tokens_seen": 43622640,
      "step": 32200
    },
    {
      "epoch": 1.030663849945586,
      "eval_loss": 4.499609470367432,
      "eval_runtime": 490.9778,
      "eval_samples_per_second": 28.282,
      "eval_steps_per_second": 14.141,
      "num_input_tokens_seen": 43622640,
      "step": 32200
    },
    {
      "epoch": 1.0308238909160745,
      "grad_norm": 18.875,
      "learning_rate": 4.541741993111465e-06,
      "loss": 3.4267,
      "num_input_tokens_seen": 43629408,
      "step": 32205
    },
    {
      "epoch": 1.030983931886563,
      "grad_norm": 24.125,
      "learning_rate": 4.536100993513423e-06,
      "loss": 4.4337,
      "num_input_tokens_seen": 43636064,
      "step": 32210
    },
    {
      "epoch": 1.0311439728570515,
      "grad_norm": 14.25,
      "learning_rate": 4.530463149705768e-06,
      "loss": 4.8362,
      "num_input_tokens_seen": 43642672,
      "step": 32215
    },
    {
      "epoch": 1.0313040138275398,
      "grad_norm": 28.25,
      "learning_rate": 4.524828462557934e-06,
      "loss": 4.2503,
      "num_input_tokens_seen": 43649344,
      "step": 32220
    },
    {
      "epoch": 1.0314640547980283,
      "grad_norm": 29.5,
      "learning_rate": 4.5191969329388625e-06,
      "loss": 5.5422,
      "num_input_tokens_seen": 43655648,
      "step": 32225
    },
    {
      "epoch": 1.0316240957685168,
      "grad_norm": 33.75,
      "learning_rate": 4.5135685617169965e-06,
      "loss": 4.6841,
      "num_input_tokens_seen": 43662720,
      "step": 32230
    },
    {
      "epoch": 1.0317841367390053,
      "grad_norm": 24.375,
      "learning_rate": 4.507943349760313e-06,
      "loss": 4.1481,
      "num_input_tokens_seen": 43669136,
      "step": 32235
    },
    {
      "epoch": 1.0319441777094935,
      "grad_norm": 28.875,
      "learning_rate": 4.502321297936277e-06,
      "loss": 4.7747,
      "num_input_tokens_seen": 43676224,
      "step": 32240
    },
    {
      "epoch": 1.032104218679982,
      "grad_norm": 26.25,
      "learning_rate": 4.496702407111888e-06,
      "loss": 3.5252,
      "num_input_tokens_seen": 43683728,
      "step": 32245
    },
    {
      "epoch": 1.0322642596504705,
      "grad_norm": 35.5,
      "learning_rate": 4.491086678153653e-06,
      "loss": 3.7087,
      "num_input_tokens_seen": 43690416,
      "step": 32250
    },
    {
      "epoch": 1.032424300620959,
      "grad_norm": 19.75,
      "learning_rate": 4.485474111927579e-06,
      "loss": 3.6517,
      "num_input_tokens_seen": 43696880,
      "step": 32255
    },
    {
      "epoch": 1.0325843415914475,
      "grad_norm": 24.0,
      "learning_rate": 4.479864709299197e-06,
      "loss": 3.2627,
      "num_input_tokens_seen": 43703520,
      "step": 32260
    },
    {
      "epoch": 1.0327443825619358,
      "grad_norm": 27.25,
      "learning_rate": 4.474258471133555e-06,
      "loss": 5.1757,
      "num_input_tokens_seen": 43710800,
      "step": 32265
    },
    {
      "epoch": 1.0329044235324243,
      "grad_norm": 24.625,
      "learning_rate": 4.4686553982952014e-06,
      "loss": 4.5261,
      "num_input_tokens_seen": 43717296,
      "step": 32270
    },
    {
      "epoch": 1.0330644645029128,
      "grad_norm": 32.0,
      "learning_rate": 4.463055491648191e-06,
      "loss": 4.3525,
      "num_input_tokens_seen": 43724032,
      "step": 32275
    },
    {
      "epoch": 1.0332245054734013,
      "grad_norm": 47.75,
      "learning_rate": 4.457458752056112e-06,
      "loss": 4.4758,
      "num_input_tokens_seen": 43730976,
      "step": 32280
    },
    {
      "epoch": 1.0333845464438896,
      "grad_norm": 29.625,
      "learning_rate": 4.451865180382042e-06,
      "loss": 4.1577,
      "num_input_tokens_seen": 43737744,
      "step": 32285
    },
    {
      "epoch": 1.033544587414378,
      "grad_norm": 49.75,
      "learning_rate": 4.4462747774885936e-06,
      "loss": 5.4271,
      "num_input_tokens_seen": 43744528,
      "step": 32290
    },
    {
      "epoch": 1.0337046283848665,
      "grad_norm": 16.125,
      "learning_rate": 4.440687544237859e-06,
      "loss": 4.398,
      "num_input_tokens_seen": 43751376,
      "step": 32295
    },
    {
      "epoch": 1.033864669355355,
      "grad_norm": 15.25,
      "learning_rate": 4.435103481491471e-06,
      "loss": 3.9159,
      "num_input_tokens_seen": 43758192,
      "step": 32300
    },
    {
      "epoch": 1.0340247103258433,
      "grad_norm": 31.75,
      "learning_rate": 4.429522590110569e-06,
      "loss": 4.5774,
      "num_input_tokens_seen": 43765184,
      "step": 32305
    },
    {
      "epoch": 1.0341847512963318,
      "grad_norm": 42.75,
      "learning_rate": 4.423944870955779e-06,
      "loss": 5.2746,
      "num_input_tokens_seen": 43772176,
      "step": 32310
    },
    {
      "epoch": 1.0343447922668203,
      "grad_norm": 16.625,
      "learning_rate": 4.418370324887272e-06,
      "loss": 4.5314,
      "num_input_tokens_seen": 43778864,
      "step": 32315
    },
    {
      "epoch": 1.0345048332373088,
      "grad_norm": 19.375,
      "learning_rate": 4.412798952764699e-06,
      "loss": 4.1884,
      "num_input_tokens_seen": 43785408,
      "step": 32320
    },
    {
      "epoch": 1.0346648742077973,
      "grad_norm": 24.75,
      "learning_rate": 4.407230755447245e-06,
      "loss": 4.6191,
      "num_input_tokens_seen": 43792544,
      "step": 32325
    },
    {
      "epoch": 1.0348249151782856,
      "grad_norm": 38.0,
      "learning_rate": 4.401665733793598e-06,
      "loss": 3.5609,
      "num_input_tokens_seen": 43799104,
      "step": 32330
    },
    {
      "epoch": 1.034984956148774,
      "grad_norm": 19.375,
      "learning_rate": 4.3961038886619425e-06,
      "loss": 4.5993,
      "num_input_tokens_seen": 43806336,
      "step": 32335
    },
    {
      "epoch": 1.0351449971192626,
      "grad_norm": 30.5,
      "learning_rate": 4.39054522091e-06,
      "loss": 4.8689,
      "num_input_tokens_seen": 43812912,
      "step": 32340
    },
    {
      "epoch": 1.035305038089751,
      "grad_norm": 46.25,
      "learning_rate": 4.384989731394979e-06,
      "loss": 4.6705,
      "num_input_tokens_seen": 43819760,
      "step": 32345
    },
    {
      "epoch": 1.0354650790602393,
      "grad_norm": 32.0,
      "learning_rate": 4.379437420973598e-06,
      "loss": 4.9613,
      "num_input_tokens_seen": 43826736,
      "step": 32350
    },
    {
      "epoch": 1.0356251200307278,
      "grad_norm": 39.75,
      "learning_rate": 4.373888290502107e-06,
      "loss": 4.3237,
      "num_input_tokens_seen": 43833600,
      "step": 32355
    },
    {
      "epoch": 1.0357851610012163,
      "grad_norm": 34.0,
      "learning_rate": 4.36834234083624e-06,
      "loss": 4.588,
      "num_input_tokens_seen": 43839920,
      "step": 32360
    },
    {
      "epoch": 1.0359452019717048,
      "grad_norm": 22.0,
      "learning_rate": 4.362799572831258e-06,
      "loss": 4.4995,
      "num_input_tokens_seen": 43846112,
      "step": 32365
    },
    {
      "epoch": 1.0361052429421933,
      "grad_norm": 38.25,
      "learning_rate": 4.35725998734193e-06,
      "loss": 3.5462,
      "num_input_tokens_seen": 43852896,
      "step": 32370
    },
    {
      "epoch": 1.0362652839126816,
      "grad_norm": 31.25,
      "learning_rate": 4.3517235852225195e-06,
      "loss": 3.8972,
      "num_input_tokens_seen": 43859648,
      "step": 32375
    },
    {
      "epoch": 1.03642532488317,
      "grad_norm": 27.75,
      "learning_rate": 4.346190367326822e-06,
      "loss": 4.3558,
      "num_input_tokens_seen": 43866576,
      "step": 32380
    },
    {
      "epoch": 1.0365853658536586,
      "grad_norm": 23.875,
      "learning_rate": 4.340660334508115e-06,
      "loss": 5.5985,
      "num_input_tokens_seen": 43873232,
      "step": 32385
    },
    {
      "epoch": 1.036745406824147,
      "grad_norm": 13.4375,
      "learning_rate": 4.335133487619206e-06,
      "loss": 4.1319,
      "num_input_tokens_seen": 43879872,
      "step": 32390
    },
    {
      "epoch": 1.0369054477946353,
      "grad_norm": 25.0,
      "learning_rate": 4.329609827512409e-06,
      "loss": 4.5793,
      "num_input_tokens_seen": 43887472,
      "step": 32395
    },
    {
      "epoch": 1.0370654887651238,
      "grad_norm": 22.75,
      "learning_rate": 4.324089355039531e-06,
      "loss": 4.4427,
      "num_input_tokens_seen": 43893856,
      "step": 32400
    },
    {
      "epoch": 1.0370654887651238,
      "eval_loss": 4.490220546722412,
      "eval_runtime": 491.1284,
      "eval_samples_per_second": 28.274,
      "eval_steps_per_second": 14.137,
      "num_input_tokens_seen": 43893856,
      "step": 32400
    },
    {
      "epoch": 1.0372255297356123,
      "grad_norm": 25.0,
      "learning_rate": 4.3185720710519075e-06,
      "loss": 3.9701,
      "num_input_tokens_seen": 43900704,
      "step": 32405
    },
    {
      "epoch": 1.0373855707061008,
      "grad_norm": 29.125,
      "learning_rate": 4.3130579764003724e-06,
      "loss": 4.4636,
      "num_input_tokens_seen": 43907376,
      "step": 32410
    },
    {
      "epoch": 1.037545611676589,
      "grad_norm": 39.25,
      "learning_rate": 4.307547071935267e-06,
      "loss": 4.3718,
      "num_input_tokens_seen": 43913760,
      "step": 32415
    },
    {
      "epoch": 1.0377056526470776,
      "grad_norm": 32.5,
      "learning_rate": 4.302039358506435e-06,
      "loss": 4.8945,
      "num_input_tokens_seen": 43920496,
      "step": 32420
    },
    {
      "epoch": 1.037865693617566,
      "grad_norm": 23.125,
      "learning_rate": 4.296534836963245e-06,
      "loss": 2.9068,
      "num_input_tokens_seen": 43927360,
      "step": 32425
    },
    {
      "epoch": 1.0380257345880546,
      "grad_norm": 43.5,
      "learning_rate": 4.291033508154555e-06,
      "loss": 4.7123,
      "num_input_tokens_seen": 43934496,
      "step": 32430
    },
    {
      "epoch": 1.038185775558543,
      "grad_norm": 20.125,
      "learning_rate": 4.285535372928748e-06,
      "loss": 3.6466,
      "num_input_tokens_seen": 43941280,
      "step": 32435
    },
    {
      "epoch": 1.0383458165290314,
      "grad_norm": 20.5,
      "learning_rate": 4.280040432133695e-06,
      "loss": 4.7679,
      "num_input_tokens_seen": 43947776,
      "step": 32440
    },
    {
      "epoch": 1.0385058574995198,
      "grad_norm": 24.25,
      "learning_rate": 4.274548686616789e-06,
      "loss": 2.889,
      "num_input_tokens_seen": 43954800,
      "step": 32445
    },
    {
      "epoch": 1.0386658984700083,
      "grad_norm": 28.875,
      "learning_rate": 4.2690601372249364e-06,
      "loss": 4.1789,
      "num_input_tokens_seen": 43962128,
      "step": 32450
    },
    {
      "epoch": 1.0388259394404968,
      "grad_norm": 22.375,
      "learning_rate": 4.263574784804525e-06,
      "loss": 3.7181,
      "num_input_tokens_seen": 43968864,
      "step": 32455
    },
    {
      "epoch": 1.038985980410985,
      "grad_norm": 29.625,
      "learning_rate": 4.258092630201479e-06,
      "loss": 3.5187,
      "num_input_tokens_seen": 43975456,
      "step": 32460
    },
    {
      "epoch": 1.0391460213814736,
      "grad_norm": 46.5,
      "learning_rate": 4.252613674261202e-06,
      "loss": 4.2477,
      "num_input_tokens_seen": 43982592,
      "step": 32465
    },
    {
      "epoch": 1.039306062351962,
      "grad_norm": 17.75,
      "learning_rate": 4.2471379178286224e-06,
      "loss": 4.1965,
      "num_input_tokens_seen": 43989440,
      "step": 32470
    },
    {
      "epoch": 1.0394661033224506,
      "grad_norm": 32.0,
      "learning_rate": 4.241665361748181e-06,
      "loss": 3.8303,
      "num_input_tokens_seen": 43996432,
      "step": 32475
    },
    {
      "epoch": 1.039626144292939,
      "grad_norm": 24.25,
      "learning_rate": 4.2361960068637994e-06,
      "loss": 4.1344,
      "num_input_tokens_seen": 44003264,
      "step": 32480
    },
    {
      "epoch": 1.0397861852634274,
      "grad_norm": 28.75,
      "learning_rate": 4.230729854018933e-06,
      "loss": 3.7225,
      "num_input_tokens_seen": 44010080,
      "step": 32485
    },
    {
      "epoch": 1.0399462262339159,
      "grad_norm": 34.25,
      "learning_rate": 4.225266904056521e-06,
      "loss": 3.8001,
      "num_input_tokens_seen": 44016896,
      "step": 32490
    },
    {
      "epoch": 1.0401062672044044,
      "grad_norm": 38.25,
      "learning_rate": 4.21980715781903e-06,
      "loss": 4.745,
      "num_input_tokens_seen": 44023488,
      "step": 32495
    },
    {
      "epoch": 1.0402663081748929,
      "grad_norm": 22.25,
      "learning_rate": 4.214350616148416e-06,
      "loss": 3.5117,
      "num_input_tokens_seen": 44029824,
      "step": 32500
    },
    {
      "epoch": 1.0404263491453811,
      "grad_norm": 47.0,
      "learning_rate": 4.20889727988614e-06,
      "loss": 4.8475,
      "num_input_tokens_seen": 44036800,
      "step": 32505
    },
    {
      "epoch": 1.0405863901158696,
      "grad_norm": 33.25,
      "learning_rate": 4.20344714987318e-06,
      "loss": 5.4454,
      "num_input_tokens_seen": 44043360,
      "step": 32510
    },
    {
      "epoch": 1.0407464310863581,
      "grad_norm": 18.0,
      "learning_rate": 4.198000226950022e-06,
      "loss": 4.2207,
      "num_input_tokens_seen": 44049856,
      "step": 32515
    },
    {
      "epoch": 1.0409064720568466,
      "grad_norm": 37.75,
      "learning_rate": 4.192556511956635e-06,
      "loss": 4.8838,
      "num_input_tokens_seen": 44056432,
      "step": 32520
    },
    {
      "epoch": 1.041066513027335,
      "grad_norm": 31.75,
      "learning_rate": 4.18711600573252e-06,
      "loss": 5.0132,
      "num_input_tokens_seen": 44063296,
      "step": 32525
    },
    {
      "epoch": 1.0412265539978234,
      "grad_norm": 33.75,
      "learning_rate": 4.181678709116671e-06,
      "loss": 4.4106,
      "num_input_tokens_seen": 44070688,
      "step": 32530
    },
    {
      "epoch": 1.0413865949683119,
      "grad_norm": 26.875,
      "learning_rate": 4.1762446229475785e-06,
      "loss": 4.2004,
      "num_input_tokens_seen": 44077472,
      "step": 32535
    },
    {
      "epoch": 1.0415466359388004,
      "grad_norm": 25.125,
      "learning_rate": 4.17081374806326e-06,
      "loss": 4.5531,
      "num_input_tokens_seen": 44083904,
      "step": 32540
    },
    {
      "epoch": 1.0417066769092889,
      "grad_norm": 43.75,
      "learning_rate": 4.165386085301212e-06,
      "loss": 4.4403,
      "num_input_tokens_seen": 44090560,
      "step": 32545
    },
    {
      "epoch": 1.0418667178797771,
      "grad_norm": 31.5,
      "learning_rate": 4.1599616354984525e-06,
      "loss": 4.8635,
      "num_input_tokens_seen": 44097056,
      "step": 32550
    },
    {
      "epoch": 1.0420267588502656,
      "grad_norm": 27.0,
      "learning_rate": 4.154540399491508e-06,
      "loss": 4.1574,
      "num_input_tokens_seen": 44103376,
      "step": 32555
    },
    {
      "epoch": 1.0421867998207541,
      "grad_norm": 35.25,
      "learning_rate": 4.149122378116394e-06,
      "loss": 4.3093,
      "num_input_tokens_seen": 44110336,
      "step": 32560
    },
    {
      "epoch": 1.0423468407912426,
      "grad_norm": 32.0,
      "learning_rate": 4.14370757220863e-06,
      "loss": 5.555,
      "num_input_tokens_seen": 44117056,
      "step": 32565
    },
    {
      "epoch": 1.042506881761731,
      "grad_norm": 21.25,
      "learning_rate": 4.138295982603263e-06,
      "loss": 3.8743,
      "num_input_tokens_seen": 44123920,
      "step": 32570
    },
    {
      "epoch": 1.0426669227322194,
      "grad_norm": 38.0,
      "learning_rate": 4.132887610134814e-06,
      "loss": 4.4364,
      "num_input_tokens_seen": 44130528,
      "step": 32575
    },
    {
      "epoch": 1.0428269637027079,
      "grad_norm": 27.5,
      "learning_rate": 4.127482455637335e-06,
      "loss": 4.2391,
      "num_input_tokens_seen": 44137360,
      "step": 32580
    },
    {
      "epoch": 1.0429870046731964,
      "grad_norm": 25.5,
      "learning_rate": 4.1220805199443545e-06,
      "loss": 5.1847,
      "num_input_tokens_seen": 44143792,
      "step": 32585
    },
    {
      "epoch": 1.0431470456436849,
      "grad_norm": 34.0,
      "learning_rate": 4.116681803888925e-06,
      "loss": 4.7475,
      "num_input_tokens_seen": 44150800,
      "step": 32590
    },
    {
      "epoch": 1.0433070866141732,
      "grad_norm": 31.5,
      "learning_rate": 4.111286308303605e-06,
      "loss": 3.8321,
      "num_input_tokens_seen": 44157728,
      "step": 32595
    },
    {
      "epoch": 1.0434671275846616,
      "grad_norm": 27.625,
      "learning_rate": 4.105894034020433e-06,
      "loss": 3.1415,
      "num_input_tokens_seen": 44164592,
      "step": 32600
    },
    {
      "epoch": 1.0434671275846616,
      "eval_loss": 4.477707862854004,
      "eval_runtime": 491.1868,
      "eval_samples_per_second": 28.27,
      "eval_steps_per_second": 14.135,
      "num_input_tokens_seen": 44164592,
      "step": 32600
    },
    {
      "epoch": 1.0436271685551501,
      "grad_norm": 24.0,
      "learning_rate": 4.100504981870975e-06,
      "loss": 4.3866,
      "num_input_tokens_seen": 44171600,
      "step": 32605
    },
    {
      "epoch": 1.0437872095256386,
      "grad_norm": 34.25,
      "learning_rate": 4.0951191526862915e-06,
      "loss": 3.8612,
      "num_input_tokens_seen": 44178016,
      "step": 32610
    },
    {
      "epoch": 1.043947250496127,
      "grad_norm": 27.125,
      "learning_rate": 4.089736547296938e-06,
      "loss": 4.6662,
      "num_input_tokens_seen": 44184752,
      "step": 32615
    },
    {
      "epoch": 1.0441072914666154,
      "grad_norm": 26.625,
      "learning_rate": 4.08435716653299e-06,
      "loss": 4.9148,
      "num_input_tokens_seen": 44191104,
      "step": 32620
    },
    {
      "epoch": 1.044267332437104,
      "grad_norm": 21.75,
      "learning_rate": 4.0789810112240005e-06,
      "loss": 4.4711,
      "num_input_tokens_seen": 44197520,
      "step": 32625
    },
    {
      "epoch": 1.0444273734075924,
      "grad_norm": 31.75,
      "learning_rate": 4.073608082199057e-06,
      "loss": 3.9266,
      "num_input_tokens_seen": 44204080,
      "step": 32630
    },
    {
      "epoch": 1.044587414378081,
      "grad_norm": 26.375,
      "learning_rate": 4.068238380286718e-06,
      "loss": 4.2026,
      "num_input_tokens_seen": 44210560,
      "step": 32635
    },
    {
      "epoch": 1.0447474553485692,
      "grad_norm": 26.0,
      "learning_rate": 4.062871906315072e-06,
      "loss": 5.7516,
      "num_input_tokens_seen": 44217024,
      "step": 32640
    },
    {
      "epoch": 1.0449074963190577,
      "grad_norm": 17.5,
      "learning_rate": 4.057508661111686e-06,
      "loss": 2.918,
      "num_input_tokens_seen": 44223936,
      "step": 32645
    },
    {
      "epoch": 1.0450675372895462,
      "grad_norm": 32.25,
      "learning_rate": 4.052148645503648e-06,
      "loss": 3.976,
      "num_input_tokens_seen": 44231264,
      "step": 32650
    },
    {
      "epoch": 1.0452275782600347,
      "grad_norm": 51.75,
      "learning_rate": 4.046791860317531e-06,
      "loss": 4.5542,
      "num_input_tokens_seen": 44238464,
      "step": 32655
    },
    {
      "epoch": 1.045387619230523,
      "grad_norm": 36.0,
      "learning_rate": 4.041438306379431e-06,
      "loss": 4.8655,
      "num_input_tokens_seen": 44245488,
      "step": 32660
    },
    {
      "epoch": 1.0455476602010114,
      "grad_norm": 35.25,
      "learning_rate": 4.036087984514916e-06,
      "loss": 4.9584,
      "num_input_tokens_seen": 44252608,
      "step": 32665
    },
    {
      "epoch": 1.0457077011715,
      "grad_norm": 35.5,
      "learning_rate": 4.030740895549084e-06,
      "loss": 4.1195,
      "num_input_tokens_seen": 44259344,
      "step": 32670
    },
    {
      "epoch": 1.0458677421419884,
      "grad_norm": 21.5,
      "learning_rate": 4.025397040306531e-06,
      "loss": 4.9629,
      "num_input_tokens_seen": 44266000,
      "step": 32675
    },
    {
      "epoch": 1.0460277831124767,
      "grad_norm": 26.875,
      "learning_rate": 4.0200564196113285e-06,
      "loss": 5.1278,
      "num_input_tokens_seen": 44272912,
      "step": 32680
    },
    {
      "epoch": 1.0461878240829652,
      "grad_norm": 14.125,
      "learning_rate": 4.014719034287079e-06,
      "loss": 4.4856,
      "num_input_tokens_seen": 44280144,
      "step": 32685
    },
    {
      "epoch": 1.0463478650534537,
      "grad_norm": 35.5,
      "learning_rate": 4.0093848851568775e-06,
      "loss": 3.8049,
      "num_input_tokens_seen": 44287072,
      "step": 32690
    },
    {
      "epoch": 1.0465079060239422,
      "grad_norm": 25.75,
      "learning_rate": 4.004053973043304e-06,
      "loss": 3.8077,
      "num_input_tokens_seen": 44293696,
      "step": 32695
    },
    {
      "epoch": 1.0466679469944307,
      "grad_norm": 23.125,
      "learning_rate": 3.998726298768465e-06,
      "loss": 4.844,
      "num_input_tokens_seen": 44300464,
      "step": 32700
    },
    {
      "epoch": 1.046827987964919,
      "grad_norm": 17.375,
      "learning_rate": 3.99340186315395e-06,
      "loss": 4.1114,
      "num_input_tokens_seen": 44306944,
      "step": 32705
    },
    {
      "epoch": 1.0469880289354074,
      "grad_norm": 28.125,
      "learning_rate": 3.988080667020849e-06,
      "loss": 4.0475,
      "num_input_tokens_seen": 44313984,
      "step": 32710
    },
    {
      "epoch": 1.047148069905896,
      "grad_norm": 30.0,
      "learning_rate": 3.982762711189766e-06,
      "loss": 4.6984,
      "num_input_tokens_seen": 44320912,
      "step": 32715
    },
    {
      "epoch": 1.0473081108763844,
      "grad_norm": 29.875,
      "learning_rate": 3.977447996480785e-06,
      "loss": 4.5606,
      "num_input_tokens_seen": 44327968,
      "step": 32720
    },
    {
      "epoch": 1.0474681518468727,
      "grad_norm": 29.375,
      "learning_rate": 3.97213652371351e-06,
      "loss": 4.8193,
      "num_input_tokens_seen": 44335248,
      "step": 32725
    },
    {
      "epoch": 1.0476281928173612,
      "grad_norm": 34.0,
      "learning_rate": 3.966828293707042e-06,
      "loss": 4.0896,
      "num_input_tokens_seen": 44342304,
      "step": 32730
    },
    {
      "epoch": 1.0477882337878497,
      "grad_norm": 35.5,
      "learning_rate": 3.961523307279963e-06,
      "loss": 4.1069,
      "num_input_tokens_seen": 44349088,
      "step": 32735
    },
    {
      "epoch": 1.0479482747583382,
      "grad_norm": 30.625,
      "learning_rate": 3.956221565250382e-06,
      "loss": 4.9049,
      "num_input_tokens_seen": 44356272,
      "step": 32740
    },
    {
      "epoch": 1.0481083157288267,
      "grad_norm": 36.0,
      "learning_rate": 3.950923068435883e-06,
      "loss": 4.3,
      "num_input_tokens_seen": 44363632,
      "step": 32745
    },
    {
      "epoch": 1.048268356699315,
      "grad_norm": 42.5,
      "learning_rate": 3.945627817653566e-06,
      "loss": 4.5237,
      "num_input_tokens_seen": 44370288,
      "step": 32750
    },
    {
      "epoch": 1.0484283976698034,
      "grad_norm": 28.5,
      "learning_rate": 3.9403358137200335e-06,
      "loss": 3.3608,
      "num_input_tokens_seen": 44377520,
      "step": 32755
    },
    {
      "epoch": 1.048588438640292,
      "grad_norm": 47.25,
      "learning_rate": 3.9350470574513605e-06,
      "loss": 4.5407,
      "num_input_tokens_seen": 44384544,
      "step": 32760
    },
    {
      "epoch": 1.0487484796107804,
      "grad_norm": 41.75,
      "learning_rate": 3.9297615496631525e-06,
      "loss": 3.9203,
      "num_input_tokens_seen": 44391536,
      "step": 32765
    },
    {
      "epoch": 1.0489085205812687,
      "grad_norm": 27.75,
      "learning_rate": 3.924479291170505e-06,
      "loss": 3.8764,
      "num_input_tokens_seen": 44398368,
      "step": 32770
    },
    {
      "epoch": 1.0490685615517572,
      "grad_norm": 19.25,
      "learning_rate": 3.919200282788002e-06,
      "loss": 3.8676,
      "num_input_tokens_seen": 44405536,
      "step": 32775
    },
    {
      "epoch": 1.0492286025222457,
      "grad_norm": 39.25,
      "learning_rate": 3.913924525329726e-06,
      "loss": 4.4118,
      "num_input_tokens_seen": 44412176,
      "step": 32780
    },
    {
      "epoch": 1.0493886434927342,
      "grad_norm": 26.75,
      "learning_rate": 3.908652019609279e-06,
      "loss": 5.0581,
      "num_input_tokens_seen": 44418880,
      "step": 32785
    },
    {
      "epoch": 1.0495486844632227,
      "grad_norm": 39.0,
      "learning_rate": 3.9033827664397364e-06,
      "loss": 5.0122,
      "num_input_tokens_seen": 44425648,
      "step": 32790
    },
    {
      "epoch": 1.049708725433711,
      "grad_norm": 33.25,
      "learning_rate": 3.898116766633694e-06,
      "loss": 4.1384,
      "num_input_tokens_seen": 44431952,
      "step": 32795
    },
    {
      "epoch": 1.0498687664041995,
      "grad_norm": 24.75,
      "learning_rate": 3.8928540210032225e-06,
      "loss": 3.9178,
      "num_input_tokens_seen": 44438640,
      "step": 32800
    },
    {
      "epoch": 1.0498687664041995,
      "eval_loss": 4.481185436248779,
      "eval_runtime": 490.5951,
      "eval_samples_per_second": 28.304,
      "eval_steps_per_second": 14.152,
      "num_input_tokens_seen": 44438640,
      "step": 32800
    },
    {
      "epoch": 1.050028807374688,
      "grad_norm": 28.5,
      "learning_rate": 3.887594530359909e-06,
      "loss": 3.7765,
      "num_input_tokens_seen": 44445600,
      "step": 32805
    },
    {
      "epoch": 1.0501888483451765,
      "grad_norm": 23.75,
      "learning_rate": 3.88233829551484e-06,
      "loss": 4.1731,
      "num_input_tokens_seen": 44452560,
      "step": 32810
    },
    {
      "epoch": 1.0503488893156647,
      "grad_norm": 29.0,
      "learning_rate": 3.877085317278581e-06,
      "loss": 4.2319,
      "num_input_tokens_seen": 44459424,
      "step": 32815
    },
    {
      "epoch": 1.0505089302861532,
      "grad_norm": 35.5,
      "learning_rate": 3.87183559646122e-06,
      "loss": 4.4168,
      "num_input_tokens_seen": 44466192,
      "step": 32820
    },
    {
      "epoch": 1.0506689712566417,
      "grad_norm": 31.5,
      "learning_rate": 3.866589133872317e-06,
      "loss": 4.7873,
      "num_input_tokens_seen": 44473360,
      "step": 32825
    },
    {
      "epoch": 1.0508290122271302,
      "grad_norm": 17.125,
      "learning_rate": 3.861345930320948e-06,
      "loss": 4.3147,
      "num_input_tokens_seen": 44479872,
      "step": 32830
    },
    {
      "epoch": 1.0509890531976185,
      "grad_norm": 50.0,
      "learning_rate": 3.856105986615688e-06,
      "loss": 5.8453,
      "num_input_tokens_seen": 44486912,
      "step": 32835
    },
    {
      "epoch": 1.051149094168107,
      "grad_norm": 54.0,
      "learning_rate": 3.850869303564589e-06,
      "loss": 3.8148,
      "num_input_tokens_seen": 44493984,
      "step": 32840
    },
    {
      "epoch": 1.0513091351385955,
      "grad_norm": 30.375,
      "learning_rate": 3.845635881975226e-06,
      "loss": 5.0869,
      "num_input_tokens_seen": 44501280,
      "step": 32845
    },
    {
      "epoch": 1.051469176109084,
      "grad_norm": 28.125,
      "learning_rate": 3.840405722654647e-06,
      "loss": 4.1772,
      "num_input_tokens_seen": 44508544,
      "step": 32850
    },
    {
      "epoch": 1.0516292170795725,
      "grad_norm": 22.875,
      "learning_rate": 3.835178826409419e-06,
      "loss": 4.2213,
      "num_input_tokens_seen": 44515248,
      "step": 32855
    },
    {
      "epoch": 1.0517892580500607,
      "grad_norm": 30.625,
      "learning_rate": 3.8299551940455895e-06,
      "loss": 4.3065,
      "num_input_tokens_seen": 44522016,
      "step": 32860
    },
    {
      "epoch": 1.0519492990205492,
      "grad_norm": 31.125,
      "learning_rate": 3.824734826368703e-06,
      "loss": 3.9148,
      "num_input_tokens_seen": 44528928,
      "step": 32865
    },
    {
      "epoch": 1.0521093399910377,
      "grad_norm": 23.125,
      "learning_rate": 3.819517724183813e-06,
      "loss": 3.0472,
      "num_input_tokens_seen": 44535936,
      "step": 32870
    },
    {
      "epoch": 1.0522693809615262,
      "grad_norm": 27.5,
      "learning_rate": 3.8143038882954648e-06,
      "loss": 4.7059,
      "num_input_tokens_seen": 44542544,
      "step": 32875
    },
    {
      "epoch": 1.0524294219320145,
      "grad_norm": 29.125,
      "learning_rate": 3.8090933195076867e-06,
      "loss": 4.5638,
      "num_input_tokens_seen": 44549472,
      "step": 32880
    },
    {
      "epoch": 1.052589462902503,
      "grad_norm": 24.75,
      "learning_rate": 3.8038860186240198e-06,
      "loss": 3.7923,
      "num_input_tokens_seen": 44556768,
      "step": 32885
    },
    {
      "epoch": 1.0527495038729915,
      "grad_norm": 20.875,
      "learning_rate": 3.7986819864475026e-06,
      "loss": 4.6258,
      "num_input_tokens_seen": 44562928,
      "step": 32890
    },
    {
      "epoch": 1.05290954484348,
      "grad_norm": 40.75,
      "learning_rate": 3.793481223780651e-06,
      "loss": 5.1487,
      "num_input_tokens_seen": 44569440,
      "step": 32895
    },
    {
      "epoch": 1.0530695858139685,
      "grad_norm": 35.25,
      "learning_rate": 3.788283731425496e-06,
      "loss": 4.5995,
      "num_input_tokens_seen": 44576304,
      "step": 32900
    },
    {
      "epoch": 1.0532296267844568,
      "grad_norm": 32.5,
      "learning_rate": 3.7830895101835488e-06,
      "loss": 3.5218,
      "num_input_tokens_seen": 44583296,
      "step": 32905
    },
    {
      "epoch": 1.0533896677549452,
      "grad_norm": 45.0,
      "learning_rate": 3.7778985608558274e-06,
      "loss": 4.4517,
      "num_input_tokens_seen": 44590128,
      "step": 32910
    },
    {
      "epoch": 1.0535497087254337,
      "grad_norm": 60.25,
      "learning_rate": 3.7727108842428443e-06,
      "loss": 4.7705,
      "num_input_tokens_seen": 44596624,
      "step": 32915
    },
    {
      "epoch": 1.0537097496959222,
      "grad_norm": 23.875,
      "learning_rate": 3.7675264811446065e-06,
      "loss": 4.4876,
      "num_input_tokens_seen": 44603456,
      "step": 32920
    },
    {
      "epoch": 1.0538697906664105,
      "grad_norm": 25.125,
      "learning_rate": 3.7623453523605994e-06,
      "loss": 4.4014,
      "num_input_tokens_seen": 44610064,
      "step": 32925
    },
    {
      "epoch": 1.054029831636899,
      "grad_norm": 25.875,
      "learning_rate": 3.757167498689834e-06,
      "loss": 4.1924,
      "num_input_tokens_seen": 44616832,
      "step": 32930
    },
    {
      "epoch": 1.0541898726073875,
      "grad_norm": 23.125,
      "learning_rate": 3.7519929209307914e-06,
      "loss": 4.8838,
      "num_input_tokens_seen": 44623360,
      "step": 32935
    },
    {
      "epoch": 1.054349913577876,
      "grad_norm": 17.25,
      "learning_rate": 3.746821619881463e-06,
      "loss": 4.2449,
      "num_input_tokens_seen": 44630480,
      "step": 32940
    },
    {
      "epoch": 1.0545099545483643,
      "grad_norm": 45.0,
      "learning_rate": 3.74165359633932e-06,
      "loss": 3.9767,
      "num_input_tokens_seen": 44637184,
      "step": 32945
    },
    {
      "epoch": 1.0546699955188528,
      "grad_norm": 22.625,
      "learning_rate": 3.736488851101341e-06,
      "loss": 5.2824,
      "num_input_tokens_seen": 44644096,
      "step": 32950
    },
    {
      "epoch": 1.0548300364893413,
      "grad_norm": 22.0,
      "learning_rate": 3.7313273849640035e-06,
      "loss": 4.3639,
      "num_input_tokens_seen": 44651120,
      "step": 32955
    },
    {
      "epoch": 1.0549900774598298,
      "grad_norm": 28.5,
      "learning_rate": 3.7261691987232533e-06,
      "loss": 4.5867,
      "num_input_tokens_seen": 44658208,
      "step": 32960
    },
    {
      "epoch": 1.0551501184303183,
      "grad_norm": 33.5,
      "learning_rate": 3.7210142931745575e-06,
      "loss": 4.4816,
      "num_input_tokens_seen": 44664448,
      "step": 32965
    },
    {
      "epoch": 1.0553101594008065,
      "grad_norm": 40.0,
      "learning_rate": 3.7158626691128712e-06,
      "loss": 4.6016,
      "num_input_tokens_seen": 44671584,
      "step": 32970
    },
    {
      "epoch": 1.055470200371295,
      "grad_norm": 36.75,
      "learning_rate": 3.710714327332629e-06,
      "loss": 4.019,
      "num_input_tokens_seen": 44678080,
      "step": 32975
    },
    {
      "epoch": 1.0556302413417835,
      "grad_norm": 25.75,
      "learning_rate": 3.7055692686277815e-06,
      "loss": 3.4388,
      "num_input_tokens_seen": 44684736,
      "step": 32980
    },
    {
      "epoch": 1.055790282312272,
      "grad_norm": 29.75,
      "learning_rate": 3.70042749379175e-06,
      "loss": 5.0061,
      "num_input_tokens_seen": 44692656,
      "step": 32985
    },
    {
      "epoch": 1.0559503232827603,
      "grad_norm": 46.5,
      "learning_rate": 3.6952890036174693e-06,
      "loss": 3.7345,
      "num_input_tokens_seen": 44699328,
      "step": 32990
    },
    {
      "epoch": 1.0561103642532488,
      "grad_norm": 25.625,
      "learning_rate": 3.690153798897353e-06,
      "loss": 4.095,
      "num_input_tokens_seen": 44706000,
      "step": 32995
    },
    {
      "epoch": 1.0562704052237373,
      "grad_norm": 38.25,
      "learning_rate": 3.6850218804233225e-06,
      "loss": 4.1746,
      "num_input_tokens_seen": 44712640,
      "step": 33000
    },
    {
      "epoch": 1.0562704052237373,
      "eval_loss": 4.479618072509766,
      "eval_runtime": 490.5385,
      "eval_samples_per_second": 28.308,
      "eval_steps_per_second": 14.154,
      "num_input_tokens_seen": 44712640,
      "step": 33000
    },
    {
      "epoch": 1.0564304461942258,
      "grad_norm": 22.75,
      "learning_rate": 3.679893248986779e-06,
      "loss": 4.0516,
      "num_input_tokens_seen": 44719136,
      "step": 33005
    },
    {
      "epoch": 1.0565904871647143,
      "grad_norm": 26.75,
      "learning_rate": 3.6747679053786147e-06,
      "loss": 4.2359,
      "num_input_tokens_seen": 44725824,
      "step": 33010
    },
    {
      "epoch": 1.0567505281352025,
      "grad_norm": 54.25,
      "learning_rate": 3.669645850389228e-06,
      "loss": 5.3294,
      "num_input_tokens_seen": 44732816,
      "step": 33015
    },
    {
      "epoch": 1.056910569105691,
      "grad_norm": 26.25,
      "learning_rate": 3.664527084808514e-06,
      "loss": 5.2656,
      "num_input_tokens_seen": 44739520,
      "step": 33020
    },
    {
      "epoch": 1.0570706100761795,
      "grad_norm": 40.5,
      "learning_rate": 3.6594116094258337e-06,
      "loss": 4.3631,
      "num_input_tokens_seen": 44746768,
      "step": 33025
    },
    {
      "epoch": 1.057230651046668,
      "grad_norm": 20.25,
      "learning_rate": 3.6542994250300665e-06,
      "loss": 5.1673,
      "num_input_tokens_seen": 44753184,
      "step": 33030
    },
    {
      "epoch": 1.0573906920171563,
      "grad_norm": 36.0,
      "learning_rate": 3.6491905324095825e-06,
      "loss": 4.0134,
      "num_input_tokens_seen": 44759776,
      "step": 33035
    },
    {
      "epoch": 1.0575507329876448,
      "grad_norm": 25.5,
      "learning_rate": 3.644084932352221e-06,
      "loss": 4.4837,
      "num_input_tokens_seen": 44766400,
      "step": 33040
    },
    {
      "epoch": 1.0577107739581333,
      "grad_norm": 21.125,
      "learning_rate": 3.6389826256453457e-06,
      "loss": 4.4899,
      "num_input_tokens_seen": 44772896,
      "step": 33045
    },
    {
      "epoch": 1.0578708149286218,
      "grad_norm": 24.625,
      "learning_rate": 3.633883613075781e-06,
      "loss": 4.0623,
      "num_input_tokens_seen": 44779456,
      "step": 33050
    },
    {
      "epoch": 1.0580308558991103,
      "grad_norm": 43.5,
      "learning_rate": 3.6287878954298693e-06,
      "loss": 3.9497,
      "num_input_tokens_seen": 44786064,
      "step": 33055
    },
    {
      "epoch": 1.0581908968695986,
      "grad_norm": 20.875,
      "learning_rate": 3.6236954734934354e-06,
      "loss": 5.0472,
      "num_input_tokens_seen": 44792976,
      "step": 33060
    },
    {
      "epoch": 1.058350937840087,
      "grad_norm": 28.875,
      "learning_rate": 3.618606348051784e-06,
      "loss": 4.0699,
      "num_input_tokens_seen": 44799648,
      "step": 33065
    },
    {
      "epoch": 1.0585109788105755,
      "grad_norm": 34.75,
      "learning_rate": 3.6135205198897376e-06,
      "loss": 4.5375,
      "num_input_tokens_seen": 44806080,
      "step": 33070
    },
    {
      "epoch": 1.058671019781064,
      "grad_norm": 33.0,
      "learning_rate": 3.6084379897915854e-06,
      "loss": 5.6123,
      "num_input_tokens_seen": 44812592,
      "step": 33075
    },
    {
      "epoch": 1.0588310607515523,
      "grad_norm": 15.625,
      "learning_rate": 3.6033587585411115e-06,
      "loss": 4.623,
      "num_input_tokens_seen": 44819232,
      "step": 33080
    },
    {
      "epoch": 1.0589911017220408,
      "grad_norm": 21.625,
      "learning_rate": 3.5982828269216117e-06,
      "loss": 3.6352,
      "num_input_tokens_seen": 44825696,
      "step": 33085
    },
    {
      "epoch": 1.0591511426925293,
      "grad_norm": 44.0,
      "learning_rate": 3.593210195715843e-06,
      "loss": 4.6439,
      "num_input_tokens_seen": 44832400,
      "step": 33090
    },
    {
      "epoch": 1.0593111836630178,
      "grad_norm": 21.375,
      "learning_rate": 3.5881408657060773e-06,
      "loss": 3.9297,
      "num_input_tokens_seen": 44839152,
      "step": 33095
    },
    {
      "epoch": 1.059471224633506,
      "grad_norm": 29.875,
      "learning_rate": 3.583074837674075e-06,
      "loss": 5.1296,
      "num_input_tokens_seen": 44845952,
      "step": 33100
    },
    {
      "epoch": 1.0596312656039946,
      "grad_norm": 41.75,
      "learning_rate": 3.578012112401069e-06,
      "loss": 4.7923,
      "num_input_tokens_seen": 44852384,
      "step": 33105
    },
    {
      "epoch": 1.059791306574483,
      "grad_norm": 33.5,
      "learning_rate": 3.5729526906677996e-06,
      "loss": 4.9377,
      "num_input_tokens_seen": 44859392,
      "step": 33110
    },
    {
      "epoch": 1.0599513475449716,
      "grad_norm": 40.75,
      "learning_rate": 3.5678965732545007e-06,
      "loss": 3.9418,
      "num_input_tokens_seen": 44866752,
      "step": 33115
    },
    {
      "epoch": 1.06011138851546,
      "grad_norm": 20.0,
      "learning_rate": 3.562843760940876e-06,
      "loss": 3.4917,
      "num_input_tokens_seen": 44873456,
      "step": 33120
    },
    {
      "epoch": 1.0602714294859483,
      "grad_norm": 39.75,
      "learning_rate": 3.5577942545061473e-06,
      "loss": 4.47,
      "num_input_tokens_seen": 44880112,
      "step": 33125
    },
    {
      "epoch": 1.0604314704564368,
      "grad_norm": 21.625,
      "learning_rate": 3.5527480547289967e-06,
      "loss": 4.3534,
      "num_input_tokens_seen": 44886512,
      "step": 33130
    },
    {
      "epoch": 1.0605915114269253,
      "grad_norm": 44.75,
      "learning_rate": 3.547705162387624e-06,
      "loss": 4.5391,
      "num_input_tokens_seen": 44893136,
      "step": 33135
    },
    {
      "epoch": 1.0607515523974138,
      "grad_norm": 46.5,
      "learning_rate": 3.542665578259699e-06,
      "loss": 5.797,
      "num_input_tokens_seen": 44900944,
      "step": 33140
    },
    {
      "epoch": 1.060911593367902,
      "grad_norm": 65.0,
      "learning_rate": 3.5376293031223945e-06,
      "loss": 5.303,
      "num_input_tokens_seen": 44907552,
      "step": 33145
    },
    {
      "epoch": 1.0610716343383906,
      "grad_norm": 34.5,
      "learning_rate": 3.5325963377523614e-06,
      "loss": 4.5936,
      "num_input_tokens_seen": 44913888,
      "step": 33150
    },
    {
      "epoch": 1.061231675308879,
      "grad_norm": 13.0625,
      "learning_rate": 3.5275666829257536e-06,
      "loss": 4.5285,
      "num_input_tokens_seen": 44920848,
      "step": 33155
    },
    {
      "epoch": 1.0613917162793676,
      "grad_norm": 31.875,
      "learning_rate": 3.5225403394181955e-06,
      "loss": 4.3325,
      "num_input_tokens_seen": 44927664,
      "step": 33160
    },
    {
      "epoch": 1.061551757249856,
      "grad_norm": 21.25,
      "learning_rate": 3.517517308004828e-06,
      "loss": 4.1346,
      "num_input_tokens_seen": 44934400,
      "step": 33165
    },
    {
      "epoch": 1.0617117982203443,
      "grad_norm": 18.375,
      "learning_rate": 3.512497589460251e-06,
      "loss": 3.7896,
      "num_input_tokens_seen": 44941360,
      "step": 33170
    },
    {
      "epoch": 1.0618718391908328,
      "grad_norm": 24.5,
      "learning_rate": 3.5074811845585727e-06,
      "loss": 3.8064,
      "num_input_tokens_seen": 44948160,
      "step": 33175
    },
    {
      "epoch": 1.0620318801613213,
      "grad_norm": 32.5,
      "learning_rate": 3.5024680940733937e-06,
      "loss": 4.6874,
      "num_input_tokens_seen": 44954912,
      "step": 33180
    },
    {
      "epoch": 1.0621919211318098,
      "grad_norm": 33.5,
      "learning_rate": 3.4974583187777852e-06,
      "loss": 3.5811,
      "num_input_tokens_seen": 44961488,
      "step": 33185
    },
    {
      "epoch": 1.062351962102298,
      "grad_norm": 45.25,
      "learning_rate": 3.4924518594443204e-06,
      "loss": 5.4092,
      "num_input_tokens_seen": 44967984,
      "step": 33190
    },
    {
      "epoch": 1.0625120030727866,
      "grad_norm": 23.25,
      "learning_rate": 3.4874487168450682e-06,
      "loss": 4.5044,
      "num_input_tokens_seen": 44974624,
      "step": 33195
    },
    {
      "epoch": 1.062672044043275,
      "grad_norm": 27.75,
      "learning_rate": 3.482448891751558e-06,
      "loss": 4.973,
      "num_input_tokens_seen": 44980912,
      "step": 33200
    },
    {
      "epoch": 1.062672044043275,
      "eval_loss": 4.489758491516113,
      "eval_runtime": 489.1806,
      "eval_samples_per_second": 28.386,
      "eval_steps_per_second": 14.193,
      "num_input_tokens_seen": 44980912,
      "step": 33200
    },
    {
      "epoch": 1.0628320850137636,
      "grad_norm": 24.875,
      "learning_rate": 3.477452384934843e-06,
      "loss": 5.3476,
      "num_input_tokens_seen": 44987664,
      "step": 33205
    },
    {
      "epoch": 1.0629921259842519,
      "grad_norm": 30.25,
      "learning_rate": 3.472459197165434e-06,
      "loss": 5.0406,
      "num_input_tokens_seen": 44994384,
      "step": 33210
    },
    {
      "epoch": 1.0631521669547404,
      "grad_norm": 13.8125,
      "learning_rate": 3.4674693292133518e-06,
      "loss": 4.388,
      "num_input_tokens_seen": 45000768,
      "step": 33215
    },
    {
      "epoch": 1.0633122079252288,
      "grad_norm": 35.5,
      "learning_rate": 3.4624827818480977e-06,
      "loss": 3.4964,
      "num_input_tokens_seen": 45007568,
      "step": 33220
    },
    {
      "epoch": 1.0634722488957173,
      "grad_norm": 23.375,
      "learning_rate": 3.4574995558386474e-06,
      "loss": 3.805,
      "num_input_tokens_seen": 45014336,
      "step": 33225
    },
    {
      "epoch": 1.0636322898662058,
      "grad_norm": 29.0,
      "learning_rate": 3.452519651953487e-06,
      "loss": 3.39,
      "num_input_tokens_seen": 45021168,
      "step": 33230
    },
    {
      "epoch": 1.0637923308366941,
      "grad_norm": 45.5,
      "learning_rate": 3.447543070960585e-06,
      "loss": 4.5333,
      "num_input_tokens_seen": 45027776,
      "step": 33235
    },
    {
      "epoch": 1.0639523718071826,
      "grad_norm": 19.625,
      "learning_rate": 3.4425698136273778e-06,
      "loss": 3.2985,
      "num_input_tokens_seen": 45034176,
      "step": 33240
    },
    {
      "epoch": 1.064112412777671,
      "grad_norm": 34.25,
      "learning_rate": 3.437599880720821e-06,
      "loss": 4.4127,
      "num_input_tokens_seen": 45040992,
      "step": 33245
    },
    {
      "epoch": 1.0642724537481596,
      "grad_norm": 26.125,
      "learning_rate": 3.4326332730073267e-06,
      "loss": 4.1088,
      "num_input_tokens_seen": 45047808,
      "step": 33250
    },
    {
      "epoch": 1.0644324947186479,
      "grad_norm": 25.0,
      "learning_rate": 3.427669991252813e-06,
      "loss": 3.7451,
      "num_input_tokens_seen": 45054384,
      "step": 33255
    },
    {
      "epoch": 1.0645925356891364,
      "grad_norm": 25.5,
      "learning_rate": 3.42271003622269e-06,
      "loss": 5.6944,
      "num_input_tokens_seen": 45061456,
      "step": 33260
    },
    {
      "epoch": 1.0647525766596249,
      "grad_norm": 17.75,
      "learning_rate": 3.4177534086818286e-06,
      "loss": 4.384,
      "num_input_tokens_seen": 45068144,
      "step": 33265
    },
    {
      "epoch": 1.0649126176301134,
      "grad_norm": 41.5,
      "learning_rate": 3.412800109394612e-06,
      "loss": 3.7585,
      "num_input_tokens_seen": 45074944,
      "step": 33270
    },
    {
      "epoch": 1.0650726586006019,
      "grad_norm": 31.625,
      "learning_rate": 3.4078501391249044e-06,
      "loss": 4.8546,
      "num_input_tokens_seen": 45081888,
      "step": 33275
    },
    {
      "epoch": 1.0652326995710901,
      "grad_norm": 24.125,
      "learning_rate": 3.4029034986360453e-06,
      "loss": 4.2737,
      "num_input_tokens_seen": 45088912,
      "step": 33280
    },
    {
      "epoch": 1.0653927405415786,
      "grad_norm": 21.875,
      "learning_rate": 3.397960188690877e-06,
      "loss": 4.6922,
      "num_input_tokens_seen": 45095728,
      "step": 33285
    },
    {
      "epoch": 1.0655527815120671,
      "grad_norm": 36.5,
      "learning_rate": 3.393020210051717e-06,
      "loss": 4.4483,
      "num_input_tokens_seen": 45102448,
      "step": 33290
    },
    {
      "epoch": 1.0657128224825556,
      "grad_norm": 46.0,
      "learning_rate": 3.3880835634803655e-06,
      "loss": 4.7066,
      "num_input_tokens_seen": 45109600,
      "step": 33295
    },
    {
      "epoch": 1.0658728634530439,
      "grad_norm": 30.25,
      "learning_rate": 3.383150249738126e-06,
      "loss": 4.1551,
      "num_input_tokens_seen": 45116240,
      "step": 33300
    },
    {
      "epoch": 1.0660329044235324,
      "grad_norm": 18.625,
      "learning_rate": 3.3782202695857663e-06,
      "loss": 4.0096,
      "num_input_tokens_seen": 45122784,
      "step": 33305
    },
    {
      "epoch": 1.0661929453940209,
      "grad_norm": 33.5,
      "learning_rate": 3.373293623783558e-06,
      "loss": 4.1172,
      "num_input_tokens_seen": 45129136,
      "step": 33310
    },
    {
      "epoch": 1.0663529863645094,
      "grad_norm": 31.125,
      "learning_rate": 3.368370313091257e-06,
      "loss": 3.631,
      "num_input_tokens_seen": 45135856,
      "step": 33315
    },
    {
      "epoch": 1.0665130273349979,
      "grad_norm": 31.5,
      "learning_rate": 3.363450338268087e-06,
      "loss": 4.1693,
      "num_input_tokens_seen": 45142992,
      "step": 33320
    },
    {
      "epoch": 1.0666730683054861,
      "grad_norm": 32.0,
      "learning_rate": 3.358533700072783e-06,
      "loss": 4.6825,
      "num_input_tokens_seen": 45149872,
      "step": 33325
    },
    {
      "epoch": 1.0668331092759746,
      "grad_norm": 21.5,
      "learning_rate": 3.3536203992635377e-06,
      "loss": 3.7485,
      "num_input_tokens_seen": 45156512,
      "step": 33330
    },
    {
      "epoch": 1.0669931502464631,
      "grad_norm": 23.25,
      "learning_rate": 3.348710436598057e-06,
      "loss": 4.9921,
      "num_input_tokens_seen": 45163440,
      "step": 33335
    },
    {
      "epoch": 1.0671531912169516,
      "grad_norm": 25.5,
      "learning_rate": 3.3438038128335155e-06,
      "loss": 3.8356,
      "num_input_tokens_seen": 45170384,
      "step": 33340
    },
    {
      "epoch": 1.06731323218744,
      "grad_norm": 30.125,
      "learning_rate": 3.338900528726571e-06,
      "loss": 3.6285,
      "num_input_tokens_seen": 45177072,
      "step": 33345
    },
    {
      "epoch": 1.0674732731579284,
      "grad_norm": 23.5,
      "learning_rate": 3.3340005850333812e-06,
      "loss": 3.7487,
      "num_input_tokens_seen": 45183936,
      "step": 33350
    },
    {
      "epoch": 1.0676333141284169,
      "grad_norm": 21.625,
      "learning_rate": 3.329103982509568e-06,
      "loss": 4.2312,
      "num_input_tokens_seen": 45190736,
      "step": 33355
    },
    {
      "epoch": 1.0677933550989054,
      "grad_norm": 24.25,
      "learning_rate": 3.324210721910259e-06,
      "loss": 3.7976,
      "num_input_tokens_seen": 45197616,
      "step": 33360
    },
    {
      "epoch": 1.0679533960693939,
      "grad_norm": 19.125,
      "learning_rate": 3.319320803990053e-06,
      "loss": 4.7653,
      "num_input_tokens_seen": 45204576,
      "step": 33365
    },
    {
      "epoch": 1.0681134370398822,
      "grad_norm": 34.25,
      "learning_rate": 3.3144342295030274e-06,
      "loss": 4.1016,
      "num_input_tokens_seen": 45210960,
      "step": 33370
    },
    {
      "epoch": 1.0682734780103706,
      "grad_norm": 25.0,
      "learning_rate": 3.309550999202765e-06,
      "loss": 3.7528,
      "num_input_tokens_seen": 45217824,
      "step": 33375
    },
    {
      "epoch": 1.0684335189808591,
      "grad_norm": 14.1875,
      "learning_rate": 3.3046711138423197e-06,
      "loss": 3.2399,
      "num_input_tokens_seen": 45224528,
      "step": 33380
    },
    {
      "epoch": 1.0685935599513476,
      "grad_norm": 18.875,
      "learning_rate": 3.2997945741742255e-06,
      "loss": 3.9262,
      "num_input_tokens_seen": 45231168,
      "step": 33385
    },
    {
      "epoch": 1.068753600921836,
      "grad_norm": 26.875,
      "learning_rate": 3.2949213809505082e-06,
      "loss": 4.5265,
      "num_input_tokens_seen": 45237600,
      "step": 33390
    },
    {
      "epoch": 1.0689136418923244,
      "grad_norm": 19.375,
      "learning_rate": 3.2900515349226834e-06,
      "loss": 4.3031,
      "num_input_tokens_seen": 45244512,
      "step": 33395
    },
    {
      "epoch": 1.069073682862813,
      "grad_norm": 18.875,
      "learning_rate": 3.285185036841731e-06,
      "loss": 4.4304,
      "num_input_tokens_seen": 45251328,
      "step": 33400
    },
    {
      "epoch": 1.069073682862813,
      "eval_loss": 4.477181911468506,
      "eval_runtime": 489.8134,
      "eval_samples_per_second": 28.35,
      "eval_steps_per_second": 14.175,
      "num_input_tokens_seen": 45251328,
      "step": 33400
    },
    {
      "epoch": 1.0692337238333014,
      "grad_norm": 28.375,
      "learning_rate": 3.2803218874581377e-06,
      "loss": 4.9433,
      "num_input_tokens_seen": 45257632,
      "step": 33405
    },
    {
      "epoch": 1.0693937648037897,
      "grad_norm": 23.875,
      "learning_rate": 3.2754620875218494e-06,
      "loss": 3.497,
      "num_input_tokens_seen": 45264496,
      "step": 33410
    },
    {
      "epoch": 1.0695538057742782,
      "grad_norm": 26.125,
      "learning_rate": 3.2706056377823146e-06,
      "loss": 4.9754,
      "num_input_tokens_seen": 45271440,
      "step": 33415
    },
    {
      "epoch": 1.0697138467447667,
      "grad_norm": 40.5,
      "learning_rate": 3.2657525389884647e-06,
      "loss": 4.3479,
      "num_input_tokens_seen": 45278336,
      "step": 33420
    },
    {
      "epoch": 1.0698738877152552,
      "grad_norm": 21.75,
      "learning_rate": 3.260902791888698e-06,
      "loss": 3.1352,
      "num_input_tokens_seen": 45285168,
      "step": 33425
    },
    {
      "epoch": 1.0700339286857437,
      "grad_norm": 21.375,
      "learning_rate": 3.2560563972309166e-06,
      "loss": 4.171,
      "num_input_tokens_seen": 45292736,
      "step": 33430
    },
    {
      "epoch": 1.070193969656232,
      "grad_norm": 44.5,
      "learning_rate": 3.251213355762489e-06,
      "loss": 3.6282,
      "num_input_tokens_seen": 45299552,
      "step": 33435
    },
    {
      "epoch": 1.0703540106267204,
      "grad_norm": 33.25,
      "learning_rate": 3.2463736682302707e-06,
      "loss": 4.6835,
      "num_input_tokens_seen": 45306704,
      "step": 33440
    },
    {
      "epoch": 1.070514051597209,
      "grad_norm": 28.0,
      "learning_rate": 3.2415373353806124e-06,
      "loss": 4.3633,
      "num_input_tokens_seen": 45313808,
      "step": 33445
    },
    {
      "epoch": 1.0706740925676974,
      "grad_norm": 38.25,
      "learning_rate": 3.236704357959322e-06,
      "loss": 3.6589,
      "num_input_tokens_seen": 45320544,
      "step": 33450
    },
    {
      "epoch": 1.0708341335381857,
      "grad_norm": 54.75,
      "learning_rate": 3.2318747367117154e-06,
      "loss": 4.1027,
      "num_input_tokens_seen": 45327248,
      "step": 33455
    },
    {
      "epoch": 1.0709941745086742,
      "grad_norm": 45.25,
      "learning_rate": 3.227048472382585e-06,
      "loss": 4.9385,
      "num_input_tokens_seen": 45334336,
      "step": 33460
    },
    {
      "epoch": 1.0711542154791627,
      "grad_norm": 23.125,
      "learning_rate": 3.2222255657161915e-06,
      "loss": 4.6382,
      "num_input_tokens_seen": 45340688,
      "step": 33465
    },
    {
      "epoch": 1.0713142564496512,
      "grad_norm": 31.25,
      "learning_rate": 3.2174060174562924e-06,
      "loss": 4.2857,
      "num_input_tokens_seen": 45347952,
      "step": 33470
    },
    {
      "epoch": 1.0714742974201394,
      "grad_norm": 22.625,
      "learning_rate": 3.2125898283461298e-06,
      "loss": 4.3363,
      "num_input_tokens_seen": 45354576,
      "step": 33475
    },
    {
      "epoch": 1.071634338390628,
      "grad_norm": 38.75,
      "learning_rate": 3.207776999128406e-06,
      "loss": 5.1304,
      "num_input_tokens_seen": 45361472,
      "step": 33480
    },
    {
      "epoch": 1.0717943793611164,
      "grad_norm": 29.875,
      "learning_rate": 3.202967530545331e-06,
      "loss": 4.6275,
      "num_input_tokens_seen": 45368464,
      "step": 33485
    },
    {
      "epoch": 1.071954420331605,
      "grad_norm": 27.125,
      "learning_rate": 3.1981614233385778e-06,
      "loss": 4.6668,
      "num_input_tokens_seen": 45375872,
      "step": 33490
    },
    {
      "epoch": 1.0721144613020934,
      "grad_norm": 23.0,
      "learning_rate": 3.1933586782493115e-06,
      "loss": 3.6403,
      "num_input_tokens_seen": 45382640,
      "step": 33495
    },
    {
      "epoch": 1.0722745022725817,
      "grad_norm": 32.0,
      "learning_rate": 3.188559296018184e-06,
      "loss": 4.1756,
      "num_input_tokens_seen": 45389344,
      "step": 33500
    },
    {
      "epoch": 1.0724345432430702,
      "grad_norm": 44.25,
      "learning_rate": 3.1837632773853098e-06,
      "loss": 4.5241,
      "num_input_tokens_seen": 45395680,
      "step": 33505
    },
    {
      "epoch": 1.0725945842135587,
      "grad_norm": 25.875,
      "learning_rate": 3.178970623090294e-06,
      "loss": 3.9064,
      "num_input_tokens_seen": 45402384,
      "step": 33510
    },
    {
      "epoch": 1.0727546251840472,
      "grad_norm": 27.5,
      "learning_rate": 3.174181333872234e-06,
      "loss": 4.9379,
      "num_input_tokens_seen": 45408752,
      "step": 33515
    },
    {
      "epoch": 1.0729146661545355,
      "grad_norm": 34.25,
      "learning_rate": 3.169395410469686e-06,
      "loss": 4.5804,
      "num_input_tokens_seen": 45415328,
      "step": 33520
    },
    {
      "epoch": 1.073074707125024,
      "grad_norm": 33.25,
      "learning_rate": 3.164612853620713e-06,
      "loss": 3.942,
      "num_input_tokens_seen": 45422000,
      "step": 33525
    },
    {
      "epoch": 1.0732347480955124,
      "grad_norm": 35.0,
      "learning_rate": 3.1598336640628333e-06,
      "loss": 5.3993,
      "num_input_tokens_seen": 45429040,
      "step": 33530
    },
    {
      "epoch": 1.073394789066001,
      "grad_norm": 18.375,
      "learning_rate": 3.155057842533063e-06,
      "loss": 4.2267,
      "num_input_tokens_seen": 45435360,
      "step": 33535
    },
    {
      "epoch": 1.0735548300364894,
      "grad_norm": 43.0,
      "learning_rate": 3.1502853897678984e-06,
      "loss": 5.7472,
      "num_input_tokens_seen": 45441920,
      "step": 33540
    },
    {
      "epoch": 1.0737148710069777,
      "grad_norm": 30.625,
      "learning_rate": 3.1455163065033017e-06,
      "loss": 4.7578,
      "num_input_tokens_seen": 45448864,
      "step": 33545
    },
    {
      "epoch": 1.0738749119774662,
      "grad_norm": 33.25,
      "learning_rate": 3.140750593474734e-06,
      "loss": 3.9493,
      "num_input_tokens_seen": 45456176,
      "step": 33550
    },
    {
      "epoch": 1.0740349529479547,
      "grad_norm": 26.5,
      "learning_rate": 3.1359882514171294e-06,
      "loss": 4.6528,
      "num_input_tokens_seen": 45462992,
      "step": 33555
    },
    {
      "epoch": 1.0741949939184432,
      "grad_norm": 27.375,
      "learning_rate": 3.1312292810648903e-06,
      "loss": 4.4628,
      "num_input_tokens_seen": 45469728,
      "step": 33560
    },
    {
      "epoch": 1.0743550348889315,
      "grad_norm": 20.25,
      "learning_rate": 3.1264736831519204e-06,
      "loss": 4.4995,
      "num_input_tokens_seen": 45476576,
      "step": 33565
    },
    {
      "epoch": 1.07451507585942,
      "grad_norm": 15.4375,
      "learning_rate": 3.1217214584115863e-06,
      "loss": 3.4992,
      "num_input_tokens_seen": 45483552,
      "step": 33570
    },
    {
      "epoch": 1.0746751168299085,
      "grad_norm": 42.0,
      "learning_rate": 3.116972607576746e-06,
      "loss": 4.7112,
      "num_input_tokens_seen": 45490192,
      "step": 33575
    },
    {
      "epoch": 1.074835157800397,
      "grad_norm": 35.0,
      "learning_rate": 3.1122271313797303e-06,
      "loss": 4.0159,
      "num_input_tokens_seen": 45496800,
      "step": 33580
    },
    {
      "epoch": 1.0749951987708855,
      "grad_norm": 24.375,
      "learning_rate": 3.107485030552343e-06,
      "loss": 3.5316,
      "num_input_tokens_seen": 45503328,
      "step": 33585
    },
    {
      "epoch": 1.0751552397413737,
      "grad_norm": 29.25,
      "learning_rate": 3.1027463058258848e-06,
      "loss": 3.9872,
      "num_input_tokens_seen": 45510112,
      "step": 33590
    },
    {
      "epoch": 1.0753152807118622,
      "grad_norm": 45.75,
      "learning_rate": 3.0980109579311273e-06,
      "loss": 4.6658,
      "num_input_tokens_seen": 45516752,
      "step": 33595
    },
    {
      "epoch": 1.0754753216823507,
      "grad_norm": 23.125,
      "learning_rate": 3.093278987598314e-06,
      "loss": 4.1773,
      "num_input_tokens_seen": 45523792,
      "step": 33600
    },
    {
      "epoch": 1.0754753216823507,
      "eval_loss": 4.491732120513916,
      "eval_runtime": 488.1344,
      "eval_samples_per_second": 28.447,
      "eval_steps_per_second": 14.224,
      "num_input_tokens_seen": 45523792,
      "step": 33600
    },
    {
      "epoch": 1.0756353626528392,
      "grad_norm": 32.75,
      "learning_rate": 3.0885503955571826e-06,
      "loss": 4.0541,
      "num_input_tokens_seen": 45530608,
      "step": 33605
    },
    {
      "epoch": 1.0757954036233275,
      "grad_norm": 25.75,
      "learning_rate": 3.0838251825369313e-06,
      "loss": 4.3954,
      "num_input_tokens_seen": 45537568,
      "step": 33610
    },
    {
      "epoch": 1.075955444593816,
      "grad_norm": 29.25,
      "learning_rate": 3.0791033492662517e-06,
      "loss": 4.1968,
      "num_input_tokens_seen": 45544592,
      "step": 33615
    },
    {
      "epoch": 1.0761154855643045,
      "grad_norm": 22.75,
      "learning_rate": 3.0743848964733203e-06,
      "loss": 4.4406,
      "num_input_tokens_seen": 45551440,
      "step": 33620
    },
    {
      "epoch": 1.076275526534793,
      "grad_norm": 34.5,
      "learning_rate": 3.0696698248857625e-06,
      "loss": 4.497,
      "num_input_tokens_seen": 45558096,
      "step": 33625
    },
    {
      "epoch": 1.0764355675052815,
      "grad_norm": 31.0,
      "learning_rate": 3.0649581352307192e-06,
      "loss": 4.321,
      "num_input_tokens_seen": 45564736,
      "step": 33630
    },
    {
      "epoch": 1.0765956084757697,
      "grad_norm": 33.25,
      "learning_rate": 3.060249828234776e-06,
      "loss": 4.1493,
      "num_input_tokens_seen": 45571504,
      "step": 33635
    },
    {
      "epoch": 1.0767556494462582,
      "grad_norm": 25.125,
      "learning_rate": 3.055544904624025e-06,
      "loss": 4.3225,
      "num_input_tokens_seen": 45578848,
      "step": 33640
    },
    {
      "epoch": 1.0769156904167467,
      "grad_norm": 19.625,
      "learning_rate": 3.050843365124026e-06,
      "loss": 4.3789,
      "num_input_tokens_seen": 45585472,
      "step": 33645
    },
    {
      "epoch": 1.0770757313872352,
      "grad_norm": 19.375,
      "learning_rate": 3.0461452104598083e-06,
      "loss": 4.6993,
      "num_input_tokens_seen": 45592560,
      "step": 33650
    },
    {
      "epoch": 1.0772357723577235,
      "grad_norm": 43.0,
      "learning_rate": 3.0414504413558836e-06,
      "loss": 4.3266,
      "num_input_tokens_seen": 45599792,
      "step": 33655
    },
    {
      "epoch": 1.077395813328212,
      "grad_norm": 29.25,
      "learning_rate": 3.0367590585362564e-06,
      "loss": 3.9356,
      "num_input_tokens_seen": 45606608,
      "step": 33660
    },
    {
      "epoch": 1.0775558542987005,
      "grad_norm": 22.25,
      "learning_rate": 3.0320710627243813e-06,
      "loss": 3.7897,
      "num_input_tokens_seen": 45614192,
      "step": 33665
    },
    {
      "epoch": 1.077715895269189,
      "grad_norm": 24.5,
      "learning_rate": 3.027386454643222e-06,
      "loss": 4.4315,
      "num_input_tokens_seen": 45620720,
      "step": 33670
    },
    {
      "epoch": 1.0778759362396773,
      "grad_norm": 36.5,
      "learning_rate": 3.0227052350151914e-06,
      "loss": 4.2861,
      "num_input_tokens_seen": 45627392,
      "step": 33675
    },
    {
      "epoch": 1.0780359772101658,
      "grad_norm": 20.625,
      "learning_rate": 3.0180274045621957e-06,
      "loss": 3.822,
      "num_input_tokens_seen": 45633776,
      "step": 33680
    },
    {
      "epoch": 1.0781960181806542,
      "grad_norm": 41.25,
      "learning_rate": 3.013352964005625e-06,
      "loss": 4.617,
      "num_input_tokens_seen": 45640912,
      "step": 33685
    },
    {
      "epoch": 1.0783560591511427,
      "grad_norm": 30.625,
      "learning_rate": 3.0086819140663218e-06,
      "loss": 3.8636,
      "num_input_tokens_seen": 45647456,
      "step": 33690
    },
    {
      "epoch": 1.0785161001216312,
      "grad_norm": 35.25,
      "learning_rate": 3.0040142554646265e-06,
      "loss": 4.9908,
      "num_input_tokens_seen": 45654048,
      "step": 33695
    },
    {
      "epoch": 1.0786761410921195,
      "grad_norm": 22.875,
      "learning_rate": 2.999349988920361e-06,
      "loss": 3.9946,
      "num_input_tokens_seen": 45661216,
      "step": 33700
    },
    {
      "epoch": 1.078836182062608,
      "grad_norm": 27.5,
      "learning_rate": 2.994689115152796e-06,
      "loss": 3.63,
      "num_input_tokens_seen": 45667968,
      "step": 33705
    },
    {
      "epoch": 1.0789962230330965,
      "grad_norm": 26.25,
      "learning_rate": 2.9900316348807105e-06,
      "loss": 3.8352,
      "num_input_tokens_seen": 45674592,
      "step": 33710
    },
    {
      "epoch": 1.079156264003585,
      "grad_norm": 17.25,
      "learning_rate": 2.985377548822338e-06,
      "loss": 3.9705,
      "num_input_tokens_seen": 45681632,
      "step": 33715
    },
    {
      "epoch": 1.0793163049740733,
      "grad_norm": 29.125,
      "learning_rate": 2.980726857695404e-06,
      "loss": 5.125,
      "num_input_tokens_seen": 45687984,
      "step": 33720
    },
    {
      "epoch": 1.0794763459445618,
      "grad_norm": 35.5,
      "learning_rate": 2.9760795622171017e-06,
      "loss": 3.9536,
      "num_input_tokens_seen": 45694224,
      "step": 33725
    },
    {
      "epoch": 1.0796363869150503,
      "grad_norm": 18.75,
      "learning_rate": 2.971435663104094e-06,
      "loss": 4.2683,
      "num_input_tokens_seen": 45701056,
      "step": 33730
    },
    {
      "epoch": 1.0797964278855388,
      "grad_norm": 35.0,
      "learning_rate": 2.9667951610725385e-06,
      "loss": 4.2601,
      "num_input_tokens_seen": 45707456,
      "step": 33735
    },
    {
      "epoch": 1.079956468856027,
      "grad_norm": 24.125,
      "learning_rate": 2.9621580568380575e-06,
      "loss": 4.1038,
      "num_input_tokens_seen": 45714752,
      "step": 33740
    },
    {
      "epoch": 1.0801165098265155,
      "grad_norm": 18.625,
      "learning_rate": 2.9575243511157453e-06,
      "loss": 4.2582,
      "num_input_tokens_seen": 45721392,
      "step": 33745
    },
    {
      "epoch": 1.080276550797004,
      "grad_norm": 24.875,
      "learning_rate": 2.952894044620186e-06,
      "loss": 4.0735,
      "num_input_tokens_seen": 45727664,
      "step": 33750
    },
    {
      "epoch": 1.0804365917674925,
      "grad_norm": 36.0,
      "learning_rate": 2.948267138065419e-06,
      "loss": 5.3501,
      "num_input_tokens_seen": 45734976,
      "step": 33755
    },
    {
      "epoch": 1.080596632737981,
      "grad_norm": 22.375,
      "learning_rate": 2.943643632164983e-06,
      "loss": 4.2838,
      "num_input_tokens_seen": 45741936,
      "step": 33760
    },
    {
      "epoch": 1.0807566737084693,
      "grad_norm": 39.25,
      "learning_rate": 2.939023527631879e-06,
      "loss": 3.4266,
      "num_input_tokens_seen": 45748784,
      "step": 33765
    },
    {
      "epoch": 1.0809167146789578,
      "grad_norm": 29.125,
      "learning_rate": 2.934406825178576e-06,
      "loss": 4.3293,
      "num_input_tokens_seen": 45755616,
      "step": 33770
    },
    {
      "epoch": 1.0810767556494463,
      "grad_norm": 28.5,
      "learning_rate": 2.9297935255170357e-06,
      "loss": 4.3746,
      "num_input_tokens_seen": 45762352,
      "step": 33775
    },
    {
      "epoch": 1.0812367966199348,
      "grad_norm": 30.0,
      "learning_rate": 2.925183629358691e-06,
      "loss": 4.6169,
      "num_input_tokens_seen": 45768816,
      "step": 33780
    },
    {
      "epoch": 1.081396837590423,
      "grad_norm": 34.25,
      "learning_rate": 2.9205771374144346e-06,
      "loss": 3.7287,
      "num_input_tokens_seen": 45775776,
      "step": 33785
    },
    {
      "epoch": 1.0815568785609115,
      "grad_norm": 33.0,
      "learning_rate": 2.915974050394657e-06,
      "loss": 4.3589,
      "num_input_tokens_seen": 45782736,
      "step": 33790
    },
    {
      "epoch": 1.0817169195314,
      "grad_norm": 52.0,
      "learning_rate": 2.9113743690092067e-06,
      "loss": 4.1766,
      "num_input_tokens_seen": 45789968,
      "step": 33795
    },
    {
      "epoch": 1.0818769605018885,
      "grad_norm": 21.0,
      "learning_rate": 2.906778093967402e-06,
      "loss": 4.3373,
      "num_input_tokens_seen": 45796960,
      "step": 33800
    },
    {
      "epoch": 1.0818769605018885,
      "eval_loss": 4.4945597648620605,
      "eval_runtime": 488.5656,
      "eval_samples_per_second": 28.422,
      "eval_steps_per_second": 14.211,
      "num_input_tokens_seen": 45796960,
      "step": 33800
    },
    {
      "epoch": 1.082037001472377,
      "grad_norm": 32.5,
      "learning_rate": 2.9021852259780656e-06,
      "loss": 5.3061,
      "num_input_tokens_seen": 45803328,
      "step": 33805
    },
    {
      "epoch": 1.0821970424428653,
      "grad_norm": 46.25,
      "learning_rate": 2.8975957657494583e-06,
      "loss": 4.7259,
      "num_input_tokens_seen": 45810352,
      "step": 33810
    },
    {
      "epoch": 1.0823570834133538,
      "grad_norm": 20.75,
      "learning_rate": 2.8930097139893417e-06,
      "loss": 4.1605,
      "num_input_tokens_seen": 45817344,
      "step": 33815
    },
    {
      "epoch": 1.0825171243838423,
      "grad_norm": 37.0,
      "learning_rate": 2.888427071404945e-06,
      "loss": 5.1734,
      "num_input_tokens_seen": 45824032,
      "step": 33820
    },
    {
      "epoch": 1.0826771653543308,
      "grad_norm": 23.5,
      "learning_rate": 2.8838478387029606e-06,
      "loss": 4.56,
      "num_input_tokens_seen": 45830944,
      "step": 33825
    },
    {
      "epoch": 1.082837206324819,
      "grad_norm": 22.0,
      "learning_rate": 2.8792720165895737e-06,
      "loss": 4.5625,
      "num_input_tokens_seen": 45837632,
      "step": 33830
    },
    {
      "epoch": 1.0829972472953076,
      "grad_norm": 40.75,
      "learning_rate": 2.874699605770423e-06,
      "loss": 3.6176,
      "num_input_tokens_seen": 45844800,
      "step": 33835
    },
    {
      "epoch": 1.083157288265796,
      "grad_norm": 23.375,
      "learning_rate": 2.8701306069506383e-06,
      "loss": 3.7917,
      "num_input_tokens_seen": 45851456,
      "step": 33840
    },
    {
      "epoch": 1.0833173292362845,
      "grad_norm": 20.625,
      "learning_rate": 2.8655650208348178e-06,
      "loss": 3.6793,
      "num_input_tokens_seen": 45858176,
      "step": 33845
    },
    {
      "epoch": 1.083477370206773,
      "grad_norm": 24.875,
      "learning_rate": 2.8610028481270257e-06,
      "loss": 4.4526,
      "num_input_tokens_seen": 45864864,
      "step": 33850
    },
    {
      "epoch": 1.0836374111772613,
      "grad_norm": 28.375,
      "learning_rate": 2.856444089530813e-06,
      "loss": 4.759,
      "num_input_tokens_seen": 45871856,
      "step": 33855
    },
    {
      "epoch": 1.0837974521477498,
      "grad_norm": 41.5,
      "learning_rate": 2.8518887457491955e-06,
      "loss": 4.1402,
      "num_input_tokens_seen": 45878368,
      "step": 33860
    },
    {
      "epoch": 1.0839574931182383,
      "grad_norm": 32.75,
      "learning_rate": 2.8473368174846666e-06,
      "loss": 4.5861,
      "num_input_tokens_seen": 45885328,
      "step": 33865
    },
    {
      "epoch": 1.0841175340887268,
      "grad_norm": 31.75,
      "learning_rate": 2.842788305439184e-06,
      "loss": 4.067,
      "num_input_tokens_seen": 45892304,
      "step": 33870
    },
    {
      "epoch": 1.084277575059215,
      "grad_norm": 36.0,
      "learning_rate": 2.8382432103141925e-06,
      "loss": 3.4341,
      "num_input_tokens_seen": 45899120,
      "step": 33875
    },
    {
      "epoch": 1.0844376160297036,
      "grad_norm": 27.0,
      "learning_rate": 2.833701532810598e-06,
      "loss": 3.6873,
      "num_input_tokens_seen": 45905696,
      "step": 33880
    },
    {
      "epoch": 1.084597657000192,
      "grad_norm": 19.375,
      "learning_rate": 2.8291632736287877e-06,
      "loss": 4.3476,
      "num_input_tokens_seen": 45912496,
      "step": 33885
    },
    {
      "epoch": 1.0847576979706806,
      "grad_norm": 18.0,
      "learning_rate": 2.824628433468615e-06,
      "loss": 3.5132,
      "num_input_tokens_seen": 45919600,
      "step": 33890
    },
    {
      "epoch": 1.084917738941169,
      "grad_norm": 33.75,
      "learning_rate": 2.8200970130294073e-06,
      "loss": 4.3101,
      "num_input_tokens_seen": 45926864,
      "step": 33895
    },
    {
      "epoch": 1.0850777799116573,
      "grad_norm": 27.625,
      "learning_rate": 2.8155690130099775e-06,
      "loss": 4.7013,
      "num_input_tokens_seen": 45933536,
      "step": 33900
    },
    {
      "epoch": 1.0852378208821458,
      "grad_norm": 41.0,
      "learning_rate": 2.8110444341085895e-06,
      "loss": 4.2022,
      "num_input_tokens_seen": 45940304,
      "step": 33905
    },
    {
      "epoch": 1.0853978618526343,
      "grad_norm": 22.125,
      "learning_rate": 2.806523277022996e-06,
      "loss": 4.3563,
      "num_input_tokens_seen": 45946768,
      "step": 33910
    },
    {
      "epoch": 1.0855579028231228,
      "grad_norm": 34.75,
      "learning_rate": 2.802005542450409e-06,
      "loss": 5.6157,
      "num_input_tokens_seen": 45953360,
      "step": 33915
    },
    {
      "epoch": 1.085717943793611,
      "grad_norm": 32.0,
      "learning_rate": 2.797491231087526e-06,
      "loss": 4.3701,
      "num_input_tokens_seen": 45959904,
      "step": 33920
    },
    {
      "epoch": 1.0858779847640996,
      "grad_norm": 15.625,
      "learning_rate": 2.7929803436305137e-06,
      "loss": 4.253,
      "num_input_tokens_seen": 45966512,
      "step": 33925
    },
    {
      "epoch": 1.086038025734588,
      "grad_norm": 40.75,
      "learning_rate": 2.788472880774998e-06,
      "loss": 4.45,
      "num_input_tokens_seen": 45973264,
      "step": 33930
    },
    {
      "epoch": 1.0861980667050766,
      "grad_norm": 19.375,
      "learning_rate": 2.7839688432160977e-06,
      "loss": 3.9369,
      "num_input_tokens_seen": 45980416,
      "step": 33935
    },
    {
      "epoch": 1.0863581076755648,
      "grad_norm": 25.5,
      "learning_rate": 2.779468231648383e-06,
      "loss": 4.5478,
      "num_input_tokens_seen": 45987008,
      "step": 33940
    },
    {
      "epoch": 1.0865181486460533,
      "grad_norm": 20.0,
      "learning_rate": 2.774971046765906e-06,
      "loss": 4.1415,
      "num_input_tokens_seen": 45993984,
      "step": 33945
    },
    {
      "epoch": 1.0866781896165418,
      "grad_norm": 26.0,
      "learning_rate": 2.770477289262194e-06,
      "loss": 4.6512,
      "num_input_tokens_seen": 46000464,
      "step": 33950
    },
    {
      "epoch": 1.0868382305870303,
      "grad_norm": 27.25,
      "learning_rate": 2.765986959830233e-06,
      "loss": 4.6969,
      "num_input_tokens_seen": 46007584,
      "step": 33955
    },
    {
      "epoch": 1.0869982715575188,
      "grad_norm": 31.875,
      "learning_rate": 2.761500059162492e-06,
      "loss": 3.9052,
      "num_input_tokens_seen": 46014160,
      "step": 33960
    },
    {
      "epoch": 1.087158312528007,
      "grad_norm": 36.0,
      "learning_rate": 2.757016587950914e-06,
      "loss": 4.7686,
      "num_input_tokens_seen": 46020912,
      "step": 33965
    },
    {
      "epoch": 1.0873183534984956,
      "grad_norm": 30.5,
      "learning_rate": 2.752536546886897e-06,
      "loss": 3.4299,
      "num_input_tokens_seen": 46027616,
      "step": 33970
    },
    {
      "epoch": 1.087478394468984,
      "grad_norm": 50.25,
      "learning_rate": 2.7480599366613234e-06,
      "loss": 5.0844,
      "num_input_tokens_seen": 46034208,
      "step": 33975
    },
    {
      "epoch": 1.0876384354394726,
      "grad_norm": 20.125,
      "learning_rate": 2.7435867579645473e-06,
      "loss": 3.5416,
      "num_input_tokens_seen": 46041296,
      "step": 33980
    },
    {
      "epoch": 1.0877984764099609,
      "grad_norm": 38.75,
      "learning_rate": 2.739117011486378e-06,
      "loss": 3.9119,
      "num_input_tokens_seen": 46048032,
      "step": 33985
    },
    {
      "epoch": 1.0879585173804494,
      "grad_norm": 27.625,
      "learning_rate": 2.7346506979161216e-06,
      "loss": 3.551,
      "num_input_tokens_seen": 46055056,
      "step": 33990
    },
    {
      "epoch": 1.0881185583509378,
      "grad_norm": 36.25,
      "learning_rate": 2.7301878179425227e-06,
      "loss": 4.3357,
      "num_input_tokens_seen": 46061632,
      "step": 33995
    },
    {
      "epoch": 1.0882785993214263,
      "grad_norm": 39.5,
      "learning_rate": 2.7257283722538244e-06,
      "loss": 5.0644,
      "num_input_tokens_seen": 46067712,
      "step": 34000
    },
    {
      "epoch": 1.0882785993214263,
      "eval_loss": 4.498800277709961,
      "eval_runtime": 488.0224,
      "eval_samples_per_second": 28.454,
      "eval_steps_per_second": 14.227,
      "num_input_tokens_seen": 46067712,
      "step": 34000
    },
    {
      "epoch": 1.0884386402919146,
      "grad_norm": 17.0,
      "learning_rate": 2.7212723615377326e-06,
      "loss": 4.4503,
      "num_input_tokens_seen": 46074592,
      "step": 34005
    },
    {
      "epoch": 1.0885986812624031,
      "grad_norm": 28.875,
      "learning_rate": 2.7168197864814145e-06,
      "loss": 4.4307,
      "num_input_tokens_seen": 46081232,
      "step": 34010
    },
    {
      "epoch": 1.0887587222328916,
      "grad_norm": 53.5,
      "learning_rate": 2.712370647771509e-06,
      "loss": 4.9148,
      "num_input_tokens_seen": 46087744,
      "step": 34015
    },
    {
      "epoch": 1.08891876320338,
      "grad_norm": 18.875,
      "learning_rate": 2.707924946094137e-06,
      "loss": 3.5779,
      "num_input_tokens_seen": 46094208,
      "step": 34020
    },
    {
      "epoch": 1.0890788041738686,
      "grad_norm": 22.625,
      "learning_rate": 2.7034826821348723e-06,
      "loss": 3.413,
      "num_input_tokens_seen": 46101104,
      "step": 34025
    },
    {
      "epoch": 1.0892388451443569,
      "grad_norm": 34.5,
      "learning_rate": 2.6990438565787786e-06,
      "loss": 4.1685,
      "num_input_tokens_seen": 46107920,
      "step": 34030
    },
    {
      "epoch": 1.0893988861148454,
      "grad_norm": 15.3125,
      "learning_rate": 2.6946084701103714e-06,
      "loss": 3.6185,
      "num_input_tokens_seen": 46114752,
      "step": 34035
    },
    {
      "epoch": 1.0895589270853339,
      "grad_norm": 26.625,
      "learning_rate": 2.6901765234136428e-06,
      "loss": 4.8077,
      "num_input_tokens_seen": 46122320,
      "step": 34040
    },
    {
      "epoch": 1.0897189680558224,
      "grad_norm": 24.25,
      "learning_rate": 2.685748017172063e-06,
      "loss": 4.819,
      "num_input_tokens_seen": 46128752,
      "step": 34045
    },
    {
      "epoch": 1.0898790090263106,
      "grad_norm": 29.875,
      "learning_rate": 2.681322952068549e-06,
      "loss": 4.417,
      "num_input_tokens_seen": 46135792,
      "step": 34050
    },
    {
      "epoch": 1.0900390499967991,
      "grad_norm": 40.75,
      "learning_rate": 2.6769013287855137e-06,
      "loss": 4.4413,
      "num_input_tokens_seen": 46142240,
      "step": 34055
    },
    {
      "epoch": 1.0901990909672876,
      "grad_norm": 28.625,
      "learning_rate": 2.6724831480048286e-06,
      "loss": 3.8173,
      "num_input_tokens_seen": 46149120,
      "step": 34060
    },
    {
      "epoch": 1.0903591319377761,
      "grad_norm": 23.625,
      "learning_rate": 2.66806841040782e-06,
      "loss": 3.8965,
      "num_input_tokens_seen": 46155904,
      "step": 34065
    },
    {
      "epoch": 1.0905191729082646,
      "grad_norm": 34.25,
      "learning_rate": 2.6636571166753083e-06,
      "loss": 3.6587,
      "num_input_tokens_seen": 46162784,
      "step": 34070
    },
    {
      "epoch": 1.0906792138787529,
      "grad_norm": 33.5,
      "learning_rate": 2.6592492674875598e-06,
      "loss": 3.5179,
      "num_input_tokens_seen": 46169424,
      "step": 34075
    },
    {
      "epoch": 1.0908392548492414,
      "grad_norm": 29.125,
      "learning_rate": 2.6548448635243305e-06,
      "loss": 4.2232,
      "num_input_tokens_seen": 46176144,
      "step": 34080
    },
    {
      "epoch": 1.0909992958197299,
      "grad_norm": 16.875,
      "learning_rate": 2.650443905464828e-06,
      "loss": 5.0093,
      "num_input_tokens_seen": 46182976,
      "step": 34085
    },
    {
      "epoch": 1.0911593367902184,
      "grad_norm": 34.25,
      "learning_rate": 2.646046393987739e-06,
      "loss": 5.6558,
      "num_input_tokens_seen": 46189264,
      "step": 34090
    },
    {
      "epoch": 1.0913193777607066,
      "grad_norm": 40.0,
      "learning_rate": 2.64165232977121e-06,
      "loss": 4.0515,
      "num_input_tokens_seen": 46195856,
      "step": 34095
    },
    {
      "epoch": 1.0914794187311951,
      "grad_norm": 32.5,
      "learning_rate": 2.6372617134928695e-06,
      "loss": 3.683,
      "num_input_tokens_seen": 46202208,
      "step": 34100
    },
    {
      "epoch": 1.0916394597016836,
      "grad_norm": 31.0,
      "learning_rate": 2.6328745458297943e-06,
      "loss": 4.6455,
      "num_input_tokens_seen": 46208800,
      "step": 34105
    },
    {
      "epoch": 1.0917995006721721,
      "grad_norm": 24.375,
      "learning_rate": 2.6284908274585546e-06,
      "loss": 3.78,
      "num_input_tokens_seen": 46215520,
      "step": 34110
    },
    {
      "epoch": 1.0919595416426606,
      "grad_norm": 27.75,
      "learning_rate": 2.6241105590551595e-06,
      "loss": 4.4836,
      "num_input_tokens_seen": 46222208,
      "step": 34115
    },
    {
      "epoch": 1.092119582613149,
      "grad_norm": 21.75,
      "learning_rate": 2.6197337412951105e-06,
      "loss": 4.8104,
      "num_input_tokens_seen": 46229008,
      "step": 34120
    },
    {
      "epoch": 1.0922796235836374,
      "grad_norm": 20.375,
      "learning_rate": 2.6153603748533705e-06,
      "loss": 4.4141,
      "num_input_tokens_seen": 46235680,
      "step": 34125
    },
    {
      "epoch": 1.0924396645541259,
      "grad_norm": 23.25,
      "learning_rate": 2.6109904604043585e-06,
      "loss": 4.7745,
      "num_input_tokens_seen": 46242416,
      "step": 34130
    },
    {
      "epoch": 1.0925997055246144,
      "grad_norm": 31.0,
      "learning_rate": 2.6066239986219765e-06,
      "loss": 4.192,
      "num_input_tokens_seen": 46248928,
      "step": 34135
    },
    {
      "epoch": 1.0927597464951027,
      "grad_norm": 36.75,
      "learning_rate": 2.602260990179592e-06,
      "loss": 5.2854,
      "num_input_tokens_seen": 46255984,
      "step": 34140
    },
    {
      "epoch": 1.0929197874655912,
      "grad_norm": 22.5,
      "learning_rate": 2.5979014357500248e-06,
      "loss": 5.2881,
      "num_input_tokens_seen": 46262480,
      "step": 34145
    },
    {
      "epoch": 1.0930798284360796,
      "grad_norm": 31.5,
      "learning_rate": 2.5935453360055844e-06,
      "loss": 4.1875,
      "num_input_tokens_seen": 46269168,
      "step": 34150
    },
    {
      "epoch": 1.0932398694065681,
      "grad_norm": 21.875,
      "learning_rate": 2.5891926916180283e-06,
      "loss": 4.9966,
      "num_input_tokens_seen": 46275536,
      "step": 34155
    },
    {
      "epoch": 1.0933999103770566,
      "grad_norm": 24.5,
      "learning_rate": 2.5848435032585883e-06,
      "loss": 4.0937,
      "num_input_tokens_seen": 46282656,
      "step": 34160
    },
    {
      "epoch": 1.093559951347545,
      "grad_norm": 32.25,
      "learning_rate": 2.58049777159797e-06,
      "loss": 4.2968,
      "num_input_tokens_seen": 46289200,
      "step": 34165
    },
    {
      "epoch": 1.0937199923180334,
      "grad_norm": 19.25,
      "learning_rate": 2.576155497306332e-06,
      "loss": 4.7503,
      "num_input_tokens_seen": 46296576,
      "step": 34170
    },
    {
      "epoch": 1.093880033288522,
      "grad_norm": 34.25,
      "learning_rate": 2.57181668105331e-06,
      "loss": 3.6871,
      "num_input_tokens_seen": 46303728,
      "step": 34175
    },
    {
      "epoch": 1.0940400742590104,
      "grad_norm": 31.0,
      "learning_rate": 2.567481323508014e-06,
      "loss": 4.8234,
      "num_input_tokens_seen": 46310352,
      "step": 34180
    },
    {
      "epoch": 1.0942001152294987,
      "grad_norm": 20.625,
      "learning_rate": 2.5631494253389954e-06,
      "loss": 4.0851,
      "num_input_tokens_seen": 46316960,
      "step": 34185
    },
    {
      "epoch": 1.0943601561999872,
      "grad_norm": 45.25,
      "learning_rate": 2.5588209872142997e-06,
      "loss": 4.8124,
      "num_input_tokens_seen": 46324288,
      "step": 34190
    },
    {
      "epoch": 1.0945201971704757,
      "grad_norm": 36.75,
      "learning_rate": 2.5544960098014186e-06,
      "loss": 4.0923,
      "num_input_tokens_seen": 46330880,
      "step": 34195
    },
    {
      "epoch": 1.0946802381409642,
      "grad_norm": 32.5,
      "learning_rate": 2.550174493767318e-06,
      "loss": 3.8478,
      "num_input_tokens_seen": 46337408,
      "step": 34200
    },
    {
      "epoch": 1.0946802381409642,
      "eval_loss": 4.474913120269775,
      "eval_runtime": 487.9103,
      "eval_samples_per_second": 28.46,
      "eval_steps_per_second": 14.23,
      "num_input_tokens_seen": 46337408,
      "step": 34200
    },
    {
      "epoch": 1.0948402791114524,
      "grad_norm": 29.375,
      "learning_rate": 2.545856439778438e-06,
      "loss": 4.8782,
      "num_input_tokens_seen": 46344112,
      "step": 34205
    },
    {
      "epoch": 1.095000320081941,
      "grad_norm": 25.0,
      "learning_rate": 2.541541848500667e-06,
      "loss": 4.3285,
      "num_input_tokens_seen": 46350784,
      "step": 34210
    },
    {
      "epoch": 1.0951603610524294,
      "grad_norm": 30.5,
      "learning_rate": 2.5372307205993733e-06,
      "loss": 3.9452,
      "num_input_tokens_seen": 46357904,
      "step": 34215
    },
    {
      "epoch": 1.095320402022918,
      "grad_norm": 30.375,
      "learning_rate": 2.5329230567393917e-06,
      "loss": 4.6155,
      "num_input_tokens_seen": 46364912,
      "step": 34220
    },
    {
      "epoch": 1.0954804429934064,
      "grad_norm": 9.0,
      "learning_rate": 2.5286188575850164e-06,
      "loss": 3.4358,
      "num_input_tokens_seen": 46371824,
      "step": 34225
    },
    {
      "epoch": 1.0956404839638947,
      "grad_norm": 38.0,
      "learning_rate": 2.5243181237999984e-06,
      "loss": 3.7552,
      "num_input_tokens_seen": 46378800,
      "step": 34230
    },
    {
      "epoch": 1.0958005249343832,
      "grad_norm": 31.375,
      "learning_rate": 2.520020856047578e-06,
      "loss": 4.5976,
      "num_input_tokens_seen": 46385872,
      "step": 34235
    },
    {
      "epoch": 1.0959605659048717,
      "grad_norm": 18.5,
      "learning_rate": 2.515727054990438e-06,
      "loss": 4.9378,
      "num_input_tokens_seen": 46392704,
      "step": 34240
    },
    {
      "epoch": 1.0961206068753602,
      "grad_norm": 22.75,
      "learning_rate": 2.511436721290747e-06,
      "loss": 3.3016,
      "num_input_tokens_seen": 46400144,
      "step": 34245
    },
    {
      "epoch": 1.0962806478458484,
      "grad_norm": 22.0,
      "learning_rate": 2.5071498556101164e-06,
      "loss": 4.2349,
      "num_input_tokens_seen": 46406880,
      "step": 34250
    },
    {
      "epoch": 1.096440688816337,
      "grad_norm": 35.25,
      "learning_rate": 2.5028664586096485e-06,
      "loss": 4.3097,
      "num_input_tokens_seen": 46413488,
      "step": 34255
    },
    {
      "epoch": 1.0966007297868254,
      "grad_norm": 31.25,
      "learning_rate": 2.498586530949881e-06,
      "loss": 3.8382,
      "num_input_tokens_seen": 46420320,
      "step": 34260
    },
    {
      "epoch": 1.096760770757314,
      "grad_norm": 29.375,
      "learning_rate": 2.4943100732908427e-06,
      "loss": 5.0095,
      "num_input_tokens_seen": 46427120,
      "step": 34265
    },
    {
      "epoch": 1.0969208117278022,
      "grad_norm": 26.875,
      "learning_rate": 2.4900370862920188e-06,
      "loss": 3.6602,
      "num_input_tokens_seen": 46434272,
      "step": 34270
    },
    {
      "epoch": 1.0970808526982907,
      "grad_norm": 38.75,
      "learning_rate": 2.4857675706123518e-06,
      "loss": 5.3835,
      "num_input_tokens_seen": 46441104,
      "step": 34275
    },
    {
      "epoch": 1.0972408936687792,
      "grad_norm": 18.625,
      "learning_rate": 2.4815015269102543e-06,
      "loss": 3.9998,
      "num_input_tokens_seen": 46448048,
      "step": 34280
    },
    {
      "epoch": 1.0974009346392677,
      "grad_norm": 31.125,
      "learning_rate": 2.477238955843611e-06,
      "loss": 4.0525,
      "num_input_tokens_seen": 46455136,
      "step": 34285
    },
    {
      "epoch": 1.0975609756097562,
      "grad_norm": 26.125,
      "learning_rate": 2.4729798580697573e-06,
      "loss": 4.2893,
      "num_input_tokens_seen": 46461872,
      "step": 34290
    },
    {
      "epoch": 1.0977210165802445,
      "grad_norm": 23.75,
      "learning_rate": 2.4687242342455034e-06,
      "loss": 4.1331,
      "num_input_tokens_seen": 46468752,
      "step": 34295
    },
    {
      "epoch": 1.097881057550733,
      "grad_norm": 28.625,
      "learning_rate": 2.4644720850271196e-06,
      "loss": 3.8453,
      "num_input_tokens_seen": 46475536,
      "step": 34300
    },
    {
      "epoch": 1.0980410985212214,
      "grad_norm": 26.625,
      "learning_rate": 2.4602234110703364e-06,
      "loss": 4.2954,
      "num_input_tokens_seen": 46482208,
      "step": 34305
    },
    {
      "epoch": 1.09820113949171,
      "grad_norm": 21.75,
      "learning_rate": 2.4559782130303576e-06,
      "loss": 4.0001,
      "num_input_tokens_seen": 46489280,
      "step": 34310
    },
    {
      "epoch": 1.0983611804621982,
      "grad_norm": 35.0,
      "learning_rate": 2.451736491561843e-06,
      "loss": 4.123,
      "num_input_tokens_seen": 46495984,
      "step": 34315
    },
    {
      "epoch": 1.0985212214326867,
      "grad_norm": 26.0,
      "learning_rate": 2.4474982473189163e-06,
      "loss": 3.9483,
      "num_input_tokens_seen": 46502400,
      "step": 34320
    },
    {
      "epoch": 1.0986812624031752,
      "grad_norm": 30.75,
      "learning_rate": 2.4432634809551796e-06,
      "loss": 4.7483,
      "num_input_tokens_seen": 46509216,
      "step": 34325
    },
    {
      "epoch": 1.0988413033736637,
      "grad_norm": 21.125,
      "learning_rate": 2.439032193123675e-06,
      "loss": 4.4518,
      "num_input_tokens_seen": 46516144,
      "step": 34330
    },
    {
      "epoch": 1.0990013443441522,
      "grad_norm": 29.125,
      "learning_rate": 2.4348043844769297e-06,
      "loss": 4.7004,
      "num_input_tokens_seen": 46523024,
      "step": 34335
    },
    {
      "epoch": 1.0991613853146405,
      "grad_norm": 28.0,
      "learning_rate": 2.4305800556669146e-06,
      "loss": 4.0925,
      "num_input_tokens_seen": 46529728,
      "step": 34340
    },
    {
      "epoch": 1.099321426285129,
      "grad_norm": 42.0,
      "learning_rate": 2.426359207345083e-06,
      "loss": 4.2532,
      "num_input_tokens_seen": 46536560,
      "step": 34345
    },
    {
      "epoch": 1.0994814672556175,
      "grad_norm": 28.375,
      "learning_rate": 2.4221418401623396e-06,
      "loss": 4.4559,
      "num_input_tokens_seen": 46544128,
      "step": 34350
    },
    {
      "epoch": 1.099641508226106,
      "grad_norm": 31.125,
      "learning_rate": 2.4179279547690557e-06,
      "loss": 4.0376,
      "num_input_tokens_seen": 46550880,
      "step": 34355
    },
    {
      "epoch": 1.0998015491965942,
      "grad_norm": 27.125,
      "learning_rate": 2.413717551815062e-06,
      "loss": 3.4967,
      "num_input_tokens_seen": 46557408,
      "step": 34360
    },
    {
      "epoch": 1.0999615901670827,
      "grad_norm": 42.0,
      "learning_rate": 2.409510631949666e-06,
      "loss": 4.0051,
      "num_input_tokens_seen": 46563840,
      "step": 34365
    },
    {
      "epoch": 1.1001216311375712,
      "grad_norm": 24.625,
      "learning_rate": 2.405307195821618e-06,
      "loss": 4.7191,
      "num_input_tokens_seen": 46570384,
      "step": 34370
    },
    {
      "epoch": 1.1002816721080597,
      "grad_norm": 31.5,
      "learning_rate": 2.4011072440791372e-06,
      "loss": 3.703,
      "num_input_tokens_seen": 46577168,
      "step": 34375
    },
    {
      "epoch": 1.1004417130785482,
      "grad_norm": 27.375,
      "learning_rate": 2.3969107773699233e-06,
      "loss": 3.8482,
      "num_input_tokens_seen": 46583936,
      "step": 34380
    },
    {
      "epoch": 1.1006017540490365,
      "grad_norm": 13.25,
      "learning_rate": 2.3927177963411096e-06,
      "loss": 4.3488,
      "num_input_tokens_seen": 46590736,
      "step": 34385
    },
    {
      "epoch": 1.100761795019525,
      "grad_norm": 30.25,
      "learning_rate": 2.3885283016393144e-06,
      "loss": 3.749,
      "num_input_tokens_seen": 46597824,
      "step": 34390
    },
    {
      "epoch": 1.1009218359900135,
      "grad_norm": 37.5,
      "learning_rate": 2.3843422939106076e-06,
      "loss": 4.7951,
      "num_input_tokens_seen": 46604656,
      "step": 34395
    },
    {
      "epoch": 1.101081876960502,
      "grad_norm": 51.25,
      "learning_rate": 2.380159773800525e-06,
      "loss": 4.7971,
      "num_input_tokens_seen": 46611232,
      "step": 34400
    },
    {
      "epoch": 1.101081876960502,
      "eval_loss": 4.494030475616455,
      "eval_runtime": 488.0331,
      "eval_samples_per_second": 28.453,
      "eval_steps_per_second": 14.226,
      "num_input_tokens_seen": 46611232,
      "step": 34400
    },
    {
      "epoch": 1.1012419179309902,
      "grad_norm": 24.375,
      "learning_rate": 2.3759807419540675e-06,
      "loss": 3.34,
      "num_input_tokens_seen": 46618000,
      "step": 34405
    },
    {
      "epoch": 1.1014019589014787,
      "grad_norm": 28.125,
      "learning_rate": 2.3718051990156835e-06,
      "loss": 4.1614,
      "num_input_tokens_seen": 46624464,
      "step": 34410
    },
    {
      "epoch": 1.1015619998719672,
      "grad_norm": 49.5,
      "learning_rate": 2.367633145629311e-06,
      "loss": 4.1314,
      "num_input_tokens_seen": 46631088,
      "step": 34415
    },
    {
      "epoch": 1.1017220408424557,
      "grad_norm": 32.5,
      "learning_rate": 2.363464582438316e-06,
      "loss": 4.1332,
      "num_input_tokens_seen": 46637664,
      "step": 34420
    },
    {
      "epoch": 1.1018820818129442,
      "grad_norm": 37.25,
      "learning_rate": 2.3592995100855526e-06,
      "loss": 4.1822,
      "num_input_tokens_seen": 46644608,
      "step": 34425
    },
    {
      "epoch": 1.1020421227834325,
      "grad_norm": 33.25,
      "learning_rate": 2.3551379292133273e-06,
      "loss": 4.2371,
      "num_input_tokens_seen": 46651088,
      "step": 34430
    },
    {
      "epoch": 1.102202163753921,
      "grad_norm": 33.0,
      "learning_rate": 2.3509798404634047e-06,
      "loss": 4.2613,
      "num_input_tokens_seen": 46657472,
      "step": 34435
    },
    {
      "epoch": 1.1023622047244095,
      "grad_norm": 32.75,
      "learning_rate": 2.346825244477019e-06,
      "loss": 4.8734,
      "num_input_tokens_seen": 46664016,
      "step": 34440
    },
    {
      "epoch": 1.102522245694898,
      "grad_norm": 41.5,
      "learning_rate": 2.3426741418948545e-06,
      "loss": 3.5844,
      "num_input_tokens_seen": 46671008,
      "step": 34445
    },
    {
      "epoch": 1.1026822866653863,
      "grad_norm": 39.0,
      "learning_rate": 2.3385265333570715e-06,
      "loss": 4.4027,
      "num_input_tokens_seen": 46677632,
      "step": 34450
    },
    {
      "epoch": 1.1028423276358748,
      "grad_norm": 29.125,
      "learning_rate": 2.334382419503278e-06,
      "loss": 4.6724,
      "num_input_tokens_seen": 46684288,
      "step": 34455
    },
    {
      "epoch": 1.1030023686063632,
      "grad_norm": 35.25,
      "learning_rate": 2.3302418009725465e-06,
      "loss": 4.4603,
      "num_input_tokens_seen": 46691264,
      "step": 34460
    },
    {
      "epoch": 1.1031624095768517,
      "grad_norm": 25.625,
      "learning_rate": 2.326104678403415e-06,
      "loss": 4.4254,
      "num_input_tokens_seen": 46698464,
      "step": 34465
    },
    {
      "epoch": 1.10332245054734,
      "grad_norm": 36.25,
      "learning_rate": 2.321971052433883e-06,
      "loss": 4.2513,
      "num_input_tokens_seen": 46705408,
      "step": 34470
    },
    {
      "epoch": 1.1034824915178285,
      "grad_norm": 16.0,
      "learning_rate": 2.3178409237014004e-06,
      "loss": 5.052,
      "num_input_tokens_seen": 46711968,
      "step": 34475
    },
    {
      "epoch": 1.103642532488317,
      "grad_norm": 29.5,
      "learning_rate": 2.313714292842889e-06,
      "loss": 4.4391,
      "num_input_tokens_seen": 46718928,
      "step": 34480
    },
    {
      "epoch": 1.1038025734588055,
      "grad_norm": 38.25,
      "learning_rate": 2.309591160494734e-06,
      "loss": 4.8038,
      "num_input_tokens_seen": 46725264,
      "step": 34485
    },
    {
      "epoch": 1.103962614429294,
      "grad_norm": 31.625,
      "learning_rate": 2.305471527292763e-06,
      "loss": 3.2942,
      "num_input_tokens_seen": 46732256,
      "step": 34490
    },
    {
      "epoch": 1.1041226553997823,
      "grad_norm": 25.875,
      "learning_rate": 2.3013553938722817e-06,
      "loss": 4.8757,
      "num_input_tokens_seen": 46738560,
      "step": 34495
    },
    {
      "epoch": 1.1042826963702708,
      "grad_norm": 30.625,
      "learning_rate": 2.297242760868043e-06,
      "loss": 4.4224,
      "num_input_tokens_seen": 46745504,
      "step": 34500
    },
    {
      "epoch": 1.1044427373407593,
      "grad_norm": 32.75,
      "learning_rate": 2.2931336289142735e-06,
      "loss": 3.6119,
      "num_input_tokens_seen": 46752128,
      "step": 34505
    },
    {
      "epoch": 1.1046027783112478,
      "grad_norm": 28.125,
      "learning_rate": 2.289027998644655e-06,
      "loss": 5.0392,
      "num_input_tokens_seen": 46759136,
      "step": 34510
    },
    {
      "epoch": 1.104762819281736,
      "grad_norm": 29.875,
      "learning_rate": 2.2849258706923228e-06,
      "loss": 3.9209,
      "num_input_tokens_seen": 46766272,
      "step": 34515
    },
    {
      "epoch": 1.1049228602522245,
      "grad_norm": 22.25,
      "learning_rate": 2.2808272456898705e-06,
      "loss": 3.3684,
      "num_input_tokens_seen": 46773200,
      "step": 34520
    },
    {
      "epoch": 1.105082901222713,
      "grad_norm": 35.25,
      "learning_rate": 2.2767321242693707e-06,
      "loss": 4.2123,
      "num_input_tokens_seen": 46779440,
      "step": 34525
    },
    {
      "epoch": 1.1052429421932015,
      "grad_norm": 39.5,
      "learning_rate": 2.272640507062329e-06,
      "loss": 4.6146,
      "num_input_tokens_seen": 46786448,
      "step": 34530
    },
    {
      "epoch": 1.1054029831636898,
      "grad_norm": 50.5,
      "learning_rate": 2.2685523946997382e-06,
      "loss": 4.6109,
      "num_input_tokens_seen": 46793024,
      "step": 34535
    },
    {
      "epoch": 1.1055630241341783,
      "grad_norm": 31.75,
      "learning_rate": 2.2644677878120245e-06,
      "loss": 5.0328,
      "num_input_tokens_seen": 46799888,
      "step": 34540
    },
    {
      "epoch": 1.1057230651046668,
      "grad_norm": 19.5,
      "learning_rate": 2.2603866870290897e-06,
      "loss": 4.3515,
      "num_input_tokens_seen": 46806384,
      "step": 34545
    },
    {
      "epoch": 1.1058831060751553,
      "grad_norm": 64.5,
      "learning_rate": 2.256309092980294e-06,
      "loss": 4.3654,
      "num_input_tokens_seen": 46813216,
      "step": 34550
    },
    {
      "epoch": 1.1060431470456438,
      "grad_norm": 20.375,
      "learning_rate": 2.252235006294448e-06,
      "loss": 4.4028,
      "num_input_tokens_seen": 46819424,
      "step": 34555
    },
    {
      "epoch": 1.106203188016132,
      "grad_norm": 30.875,
      "learning_rate": 2.2481644275998333e-06,
      "loss": 4.2491,
      "num_input_tokens_seen": 46826576,
      "step": 34560
    },
    {
      "epoch": 1.1063632289866205,
      "grad_norm": 32.25,
      "learning_rate": 2.2440973575241832e-06,
      "loss": 3.7543,
      "num_input_tokens_seen": 46833248,
      "step": 34565
    },
    {
      "epoch": 1.106523269957109,
      "grad_norm": 36.0,
      "learning_rate": 2.240033796694685e-06,
      "loss": 4.1757,
      "num_input_tokens_seen": 46839680,
      "step": 34570
    },
    {
      "epoch": 1.1066833109275975,
      "grad_norm": 27.125,
      "learning_rate": 2.235973745737999e-06,
      "loss": 5.2091,
      "num_input_tokens_seen": 46846304,
      "step": 34575
    },
    {
      "epoch": 1.1068433518980858,
      "grad_norm": 44.5,
      "learning_rate": 2.2319172052802263e-06,
      "loss": 4.7648,
      "num_input_tokens_seen": 46852976,
      "step": 34580
    },
    {
      "epoch": 1.1070033928685743,
      "grad_norm": 28.5,
      "learning_rate": 2.2278641759469477e-06,
      "loss": 4.0913,
      "num_input_tokens_seen": 46859808,
      "step": 34585
    },
    {
      "epoch": 1.1071634338390628,
      "grad_norm": 18.875,
      "learning_rate": 2.2238146583631825e-06,
      "loss": 3.1329,
      "num_input_tokens_seen": 46866512,
      "step": 34590
    },
    {
      "epoch": 1.1073234748095513,
      "grad_norm": 27.875,
      "learning_rate": 2.2197686531534256e-06,
      "loss": 5.5602,
      "num_input_tokens_seen": 46873232,
      "step": 34595
    },
    {
      "epoch": 1.1074835157800398,
      "grad_norm": 33.25,
      "learning_rate": 2.2157261609416087e-06,
      "loss": 4.4257,
      "num_input_tokens_seen": 46879824,
      "step": 34600
    },
    {
      "epoch": 1.1074835157800398,
      "eval_loss": 4.47395133972168,
      "eval_runtime": 487.8552,
      "eval_samples_per_second": 28.463,
      "eval_steps_per_second": 14.232,
      "num_input_tokens_seen": 46879824,
      "step": 34600
    },
    {
      "epoch": 1.107643556750528,
      "grad_norm": 25.0,
      "learning_rate": 2.211687182351149e-06,
      "loss": 3.5576,
      "num_input_tokens_seen": 46886368,
      "step": 34605
    },
    {
      "epoch": 1.1078035977210166,
      "grad_norm": 23.5,
      "learning_rate": 2.2076517180048993e-06,
      "loss": 4.0855,
      "num_input_tokens_seen": 46893424,
      "step": 34610
    },
    {
      "epoch": 1.107963638691505,
      "grad_norm": 23.875,
      "learning_rate": 2.2036197685251834e-06,
      "loss": 4.4485,
      "num_input_tokens_seen": 46899936,
      "step": 34615
    },
    {
      "epoch": 1.1081236796619935,
      "grad_norm": 49.75,
      "learning_rate": 2.199591334533771e-06,
      "loss": 4.5709,
      "num_input_tokens_seen": 46906512,
      "step": 34620
    },
    {
      "epoch": 1.1082837206324818,
      "grad_norm": 38.25,
      "learning_rate": 2.1955664166519036e-06,
      "loss": 4.4198,
      "num_input_tokens_seen": 46913456,
      "step": 34625
    },
    {
      "epoch": 1.1084437616029703,
      "grad_norm": 45.75,
      "learning_rate": 2.1915450155002793e-06,
      "loss": 4.8822,
      "num_input_tokens_seen": 46920016,
      "step": 34630
    },
    {
      "epoch": 1.1086038025734588,
      "grad_norm": 20.875,
      "learning_rate": 2.187527131699038e-06,
      "loss": 3.8104,
      "num_input_tokens_seen": 46926960,
      "step": 34635
    },
    {
      "epoch": 1.1087638435439473,
      "grad_norm": 26.625,
      "learning_rate": 2.18351276586779e-06,
      "loss": 4.271,
      "num_input_tokens_seen": 46934048,
      "step": 34640
    },
    {
      "epoch": 1.1089238845144358,
      "grad_norm": 33.75,
      "learning_rate": 2.1795019186256092e-06,
      "loss": 4.5528,
      "num_input_tokens_seen": 46940928,
      "step": 34645
    },
    {
      "epoch": 1.109083925484924,
      "grad_norm": 32.25,
      "learning_rate": 2.1754945905910094e-06,
      "loss": 4.027,
      "num_input_tokens_seen": 46948032,
      "step": 34650
    },
    {
      "epoch": 1.1092439664554126,
      "grad_norm": 29.5,
      "learning_rate": 2.171490782381977e-06,
      "loss": 3.3969,
      "num_input_tokens_seen": 46955120,
      "step": 34655
    },
    {
      "epoch": 1.109404007425901,
      "grad_norm": 28.625,
      "learning_rate": 2.1674904946159425e-06,
      "loss": 3.8859,
      "num_input_tokens_seen": 46962208,
      "step": 34660
    },
    {
      "epoch": 1.1095640483963896,
      "grad_norm": 24.75,
      "learning_rate": 2.16349372790981e-06,
      "loss": 3.6519,
      "num_input_tokens_seen": 46969168,
      "step": 34665
    },
    {
      "epoch": 1.1097240893668778,
      "grad_norm": 36.75,
      "learning_rate": 2.159500482879928e-06,
      "loss": 4.5301,
      "num_input_tokens_seen": 46976592,
      "step": 34670
    },
    {
      "epoch": 1.1098841303373663,
      "grad_norm": 33.5,
      "learning_rate": 2.155510760142096e-06,
      "loss": 4.5557,
      "num_input_tokens_seen": 46983536,
      "step": 34675
    },
    {
      "epoch": 1.1100441713078548,
      "grad_norm": 35.75,
      "learning_rate": 2.151524560311588e-06,
      "loss": 4.9627,
      "num_input_tokens_seen": 46990704,
      "step": 34680
    },
    {
      "epoch": 1.1102042122783433,
      "grad_norm": 27.25,
      "learning_rate": 2.147541884003129e-06,
      "loss": 4.96,
      "num_input_tokens_seen": 46997616,
      "step": 34685
    },
    {
      "epoch": 1.1103642532488318,
      "grad_norm": 27.75,
      "learning_rate": 2.1435627318308895e-06,
      "loss": 4.6573,
      "num_input_tokens_seen": 47004544,
      "step": 34690
    },
    {
      "epoch": 1.11052429421932,
      "grad_norm": 28.75,
      "learning_rate": 2.139587104408511e-06,
      "loss": 4.2293,
      "num_input_tokens_seen": 47011520,
      "step": 34695
    },
    {
      "epoch": 1.1106843351898086,
      "grad_norm": 25.625,
      "learning_rate": 2.1356150023490783e-06,
      "loss": 4.1462,
      "num_input_tokens_seen": 47018704,
      "step": 34700
    },
    {
      "epoch": 1.110844376160297,
      "grad_norm": 31.125,
      "learning_rate": 2.1316464262651464e-06,
      "loss": 3.5204,
      "num_input_tokens_seen": 47025376,
      "step": 34705
    },
    {
      "epoch": 1.1110044171307856,
      "grad_norm": 28.75,
      "learning_rate": 2.1276813767687224e-06,
      "loss": 4.4583,
      "num_input_tokens_seen": 47031872,
      "step": 34710
    },
    {
      "epoch": 1.1111644581012738,
      "grad_norm": 19.625,
      "learning_rate": 2.123719854471254e-06,
      "loss": 3.3644,
      "num_input_tokens_seen": 47038464,
      "step": 34715
    },
    {
      "epoch": 1.1113244990717623,
      "grad_norm": 37.0,
      "learning_rate": 2.119761859983668e-06,
      "loss": 4.1376,
      "num_input_tokens_seen": 47045296,
      "step": 34720
    },
    {
      "epoch": 1.1114845400422508,
      "grad_norm": 49.25,
      "learning_rate": 2.1158073939163386e-06,
      "loss": 4.7876,
      "num_input_tokens_seen": 47052368,
      "step": 34725
    },
    {
      "epoch": 1.1116445810127393,
      "grad_norm": 24.625,
      "learning_rate": 2.111856456879088e-06,
      "loss": 3.6694,
      "num_input_tokens_seen": 47059104,
      "step": 34730
    },
    {
      "epoch": 1.1118046219832276,
      "grad_norm": 24.625,
      "learning_rate": 2.1079090494811993e-06,
      "loss": 4.4916,
      "num_input_tokens_seen": 47066144,
      "step": 34735
    },
    {
      "epoch": 1.111964662953716,
      "grad_norm": 23.5,
      "learning_rate": 2.103965172331418e-06,
      "loss": 4.9292,
      "num_input_tokens_seen": 47072848,
      "step": 34740
    },
    {
      "epoch": 1.1121247039242046,
      "grad_norm": 22.0,
      "learning_rate": 2.100024826037933e-06,
      "loss": 4.9187,
      "num_input_tokens_seen": 47079376,
      "step": 34745
    },
    {
      "epoch": 1.112284744894693,
      "grad_norm": 40.75,
      "learning_rate": 2.0960880112084027e-06,
      "loss": 4.5836,
      "num_input_tokens_seen": 47086160,
      "step": 34750
    },
    {
      "epoch": 1.1124447858651816,
      "grad_norm": 27.0,
      "learning_rate": 2.092154728449927e-06,
      "loss": 4.42,
      "num_input_tokens_seen": 47092848,
      "step": 34755
    },
    {
      "epoch": 1.1126048268356699,
      "grad_norm": 22.5,
      "learning_rate": 2.0882249783690687e-06,
      "loss": 4.06,
      "num_input_tokens_seen": 47100064,
      "step": 34760
    },
    {
      "epoch": 1.1127648678061584,
      "grad_norm": 31.25,
      "learning_rate": 2.084298761571851e-06,
      "loss": 4.1869,
      "num_input_tokens_seen": 47106928,
      "step": 34765
    },
    {
      "epoch": 1.1129249087766468,
      "grad_norm": 23.125,
      "learning_rate": 2.080376078663737e-06,
      "loss": 4.2129,
      "num_input_tokens_seen": 47113776,
      "step": 34770
    },
    {
      "epoch": 1.1130849497471353,
      "grad_norm": 28.0,
      "learning_rate": 2.0764569302496593e-06,
      "loss": 4.8417,
      "num_input_tokens_seen": 47120352,
      "step": 34775
    },
    {
      "epoch": 1.1132449907176236,
      "grad_norm": 29.0,
      "learning_rate": 2.0725413169339957e-06,
      "loss": 3.9071,
      "num_input_tokens_seen": 47127280,
      "step": 34780
    },
    {
      "epoch": 1.1134050316881121,
      "grad_norm": 26.75,
      "learning_rate": 2.068629239320588e-06,
      "loss": 3.3046,
      "num_input_tokens_seen": 47134160,
      "step": 34785
    },
    {
      "epoch": 1.1135650726586006,
      "grad_norm": 20.75,
      "learning_rate": 2.064720698012726e-06,
      "loss": 4.2697,
      "num_input_tokens_seen": 47141184,
      "step": 34790
    },
    {
      "epoch": 1.113725113629089,
      "grad_norm": 24.625,
      "learning_rate": 2.0608156936131522e-06,
      "loss": 4.9303,
      "num_input_tokens_seen": 47148000,
      "step": 34795
    },
    {
      "epoch": 1.1138851545995774,
      "grad_norm": 17.25,
      "learning_rate": 2.056914226724074e-06,
      "loss": 3.9397,
      "num_input_tokens_seen": 47155008,
      "step": 34800
    },
    {
      "epoch": 1.1138851545995774,
      "eval_loss": 4.492156028747559,
      "eval_runtime": 487.8625,
      "eval_samples_per_second": 28.463,
      "eval_steps_per_second": 14.231,
      "num_input_tokens_seen": 47155008,
      "step": 34800
    },
    {
      "epoch": 1.1140451955700659,
      "grad_norm": 37.5,
      "learning_rate": 2.0530162979471385e-06,
      "loss": 4.7058,
      "num_input_tokens_seen": 47161728,
      "step": 34805
    },
    {
      "epoch": 1.1142052365405544,
      "grad_norm": 29.5,
      "learning_rate": 2.0491219078834667e-06,
      "loss": 3.6824,
      "num_input_tokens_seen": 47168816,
      "step": 34810
    },
    {
      "epoch": 1.1143652775110429,
      "grad_norm": 18.0,
      "learning_rate": 2.045231057133612e-06,
      "loss": 4.1179,
      "num_input_tokens_seen": 47175744,
      "step": 34815
    },
    {
      "epoch": 1.1145253184815314,
      "grad_norm": 46.25,
      "learning_rate": 2.0413437462975944e-06,
      "loss": 5.5704,
      "num_input_tokens_seen": 47183120,
      "step": 34820
    },
    {
      "epoch": 1.1146853594520196,
      "grad_norm": 25.25,
      "learning_rate": 2.0374599759748843e-06,
      "loss": 3.8596,
      "num_input_tokens_seen": 47189568,
      "step": 34825
    },
    {
      "epoch": 1.1148454004225081,
      "grad_norm": 29.875,
      "learning_rate": 2.033579746764419e-06,
      "loss": 3.9505,
      "num_input_tokens_seen": 47196480,
      "step": 34830
    },
    {
      "epoch": 1.1150054413929966,
      "grad_norm": 38.5,
      "learning_rate": 2.029703059264565e-06,
      "loss": 5.9737,
      "num_input_tokens_seen": 47203568,
      "step": 34835
    },
    {
      "epoch": 1.1151654823634851,
      "grad_norm": 31.5,
      "learning_rate": 2.02582991407316e-06,
      "loss": 4.6601,
      "num_input_tokens_seen": 47210208,
      "step": 34840
    },
    {
      "epoch": 1.1153255233339734,
      "grad_norm": 28.375,
      "learning_rate": 2.0219603117874992e-06,
      "loss": 5.4915,
      "num_input_tokens_seen": 47217008,
      "step": 34845
    },
    {
      "epoch": 1.1154855643044619,
      "grad_norm": 18.375,
      "learning_rate": 2.0180942530043156e-06,
      "loss": 3.9759,
      "num_input_tokens_seen": 47224320,
      "step": 34850
    },
    {
      "epoch": 1.1156456052749504,
      "grad_norm": 27.625,
      "learning_rate": 2.0142317383198107e-06,
      "loss": 4.8334,
      "num_input_tokens_seen": 47231168,
      "step": 34855
    },
    {
      "epoch": 1.1158056462454389,
      "grad_norm": 42.75,
      "learning_rate": 2.0103727683296243e-06,
      "loss": 4.5915,
      "num_input_tokens_seen": 47238336,
      "step": 34860
    },
    {
      "epoch": 1.1159656872159274,
      "grad_norm": 27.375,
      "learning_rate": 2.0065173436288636e-06,
      "loss": 4.2046,
      "num_input_tokens_seen": 47245056,
      "step": 34865
    },
    {
      "epoch": 1.1161257281864156,
      "grad_norm": 53.5,
      "learning_rate": 2.002665464812087e-06,
      "loss": 4.9634,
      "num_input_tokens_seen": 47251920,
      "step": 34870
    },
    {
      "epoch": 1.1162857691569041,
      "grad_norm": 28.75,
      "learning_rate": 1.998817132473291e-06,
      "loss": 4.5787,
      "num_input_tokens_seen": 47258864,
      "step": 34875
    },
    {
      "epoch": 1.1164458101273926,
      "grad_norm": 35.5,
      "learning_rate": 1.9949723472059507e-06,
      "loss": 3.4015,
      "num_input_tokens_seen": 47265488,
      "step": 34880
    },
    {
      "epoch": 1.1166058510978811,
      "grad_norm": 44.5,
      "learning_rate": 1.9911311096029726e-06,
      "loss": 4.8721,
      "num_input_tokens_seen": 47272736,
      "step": 34885
    },
    {
      "epoch": 1.1167658920683694,
      "grad_norm": 28.75,
      "learning_rate": 1.9872934202567224e-06,
      "loss": 5.3843,
      "num_input_tokens_seen": 47279408,
      "step": 34890
    },
    {
      "epoch": 1.116925933038858,
      "grad_norm": 35.25,
      "learning_rate": 1.9834592797590257e-06,
      "loss": 4.4552,
      "num_input_tokens_seen": 47285984,
      "step": 34895
    },
    {
      "epoch": 1.1170859740093464,
      "grad_norm": 35.75,
      "learning_rate": 1.979628688701149e-06,
      "loss": 4.0122,
      "num_input_tokens_seen": 47293056,
      "step": 34900
    },
    {
      "epoch": 1.117246014979835,
      "grad_norm": 36.25,
      "learning_rate": 1.9758016476738193e-06,
      "loss": 4.0618,
      "num_input_tokens_seen": 47299392,
      "step": 34905
    },
    {
      "epoch": 1.1174060559503234,
      "grad_norm": 44.0,
      "learning_rate": 1.971978157267221e-06,
      "loss": 4.651,
      "num_input_tokens_seen": 47305904,
      "step": 34910
    },
    {
      "epoch": 1.1175660969208117,
      "grad_norm": 26.375,
      "learning_rate": 1.968158218070973e-06,
      "loss": 4.5632,
      "num_input_tokens_seen": 47312896,
      "step": 34915
    },
    {
      "epoch": 1.1177261378913002,
      "grad_norm": 19.25,
      "learning_rate": 1.9643418306741682e-06,
      "loss": 4.182,
      "num_input_tokens_seen": 47319456,
      "step": 34920
    },
    {
      "epoch": 1.1178861788617886,
      "grad_norm": 27.125,
      "learning_rate": 1.9605289956653337e-06,
      "loss": 4.3667,
      "num_input_tokens_seen": 47326848,
      "step": 34925
    },
    {
      "epoch": 1.1180462198322771,
      "grad_norm": 13.0,
      "learning_rate": 1.9567197136324626e-06,
      "loss": 3.2926,
      "num_input_tokens_seen": 47333520,
      "step": 34930
    },
    {
      "epoch": 1.1182062608027654,
      "grad_norm": 26.75,
      "learning_rate": 1.9529139851629935e-06,
      "loss": 4.8599,
      "num_input_tokens_seen": 47340176,
      "step": 34935
    },
    {
      "epoch": 1.118366301773254,
      "grad_norm": 30.125,
      "learning_rate": 1.949111810843812e-06,
      "loss": 4.5797,
      "num_input_tokens_seen": 47346448,
      "step": 34940
    },
    {
      "epoch": 1.1185263427437424,
      "grad_norm": 19.0,
      "learning_rate": 1.9453131912612694e-06,
      "loss": 4.2379,
      "num_input_tokens_seen": 47353120,
      "step": 34945
    },
    {
      "epoch": 1.118686383714231,
      "grad_norm": 22.5,
      "learning_rate": 1.941518127001149e-06,
      "loss": 4.4763,
      "num_input_tokens_seen": 47359664,
      "step": 34950
    },
    {
      "epoch": 1.1188464246847194,
      "grad_norm": 38.0,
      "learning_rate": 1.9377266186487107e-06,
      "loss": 4.9386,
      "num_input_tokens_seen": 47366464,
      "step": 34955
    },
    {
      "epoch": 1.1190064656552077,
      "grad_norm": 52.0,
      "learning_rate": 1.9339386667886483e-06,
      "loss": 4.8487,
      "num_input_tokens_seen": 47373120,
      "step": 34960
    },
    {
      "epoch": 1.1191665066256962,
      "grad_norm": 21.875,
      "learning_rate": 1.9301542720051024e-06,
      "loss": 4.4582,
      "num_input_tokens_seen": 47379568,
      "step": 34965
    },
    {
      "epoch": 1.1193265475961847,
      "grad_norm": 30.75,
      "learning_rate": 1.926373434881684e-06,
      "loss": 4.2274,
      "num_input_tokens_seen": 47386176,
      "step": 34970
    },
    {
      "epoch": 1.1194865885666732,
      "grad_norm": 37.25,
      "learning_rate": 1.9225961560014468e-06,
      "loss": 4.5695,
      "num_input_tokens_seen": 47393168,
      "step": 34975
    },
    {
      "epoch": 1.1196466295371614,
      "grad_norm": 25.125,
      "learning_rate": 1.918822435946885e-06,
      "loss": 3.956,
      "num_input_tokens_seen": 47399696,
      "step": 34980
    },
    {
      "epoch": 1.11980667050765,
      "grad_norm": 57.5,
      "learning_rate": 1.915052275299961e-06,
      "loss": 5.1516,
      "num_input_tokens_seen": 47406576,
      "step": 34985
    },
    {
      "epoch": 1.1199667114781384,
      "grad_norm": 34.25,
      "learning_rate": 1.9112856746420854e-06,
      "loss": 4.1045,
      "num_input_tokens_seen": 47413744,
      "step": 34990
    },
    {
      "epoch": 1.120126752448627,
      "grad_norm": 26.5,
      "learning_rate": 1.907522634554104e-06,
      "loss": 4.4713,
      "num_input_tokens_seen": 47420480,
      "step": 34995
    },
    {
      "epoch": 1.1202867934191152,
      "grad_norm": 67.5,
      "learning_rate": 1.9037631556163337e-06,
      "loss": 4.4246,
      "num_input_tokens_seen": 47426864,
      "step": 35000
    },
    {
      "epoch": 1.1202867934191152,
      "eval_loss": 4.505105972290039,
      "eval_runtime": 488.6886,
      "eval_samples_per_second": 28.415,
      "eval_steps_per_second": 14.207,
      "num_input_tokens_seen": 47426864,
      "step": 35000
    },
    {
      "epoch": 1.1204468343896037,
      "grad_norm": 26.75,
      "learning_rate": 1.9000072384085272e-06,
      "loss": 4.2814,
      "num_input_tokens_seen": 47434352,
      "step": 35005
    },
    {
      "epoch": 1.1206068753600922,
      "grad_norm": 28.875,
      "learning_rate": 1.8962548835098987e-06,
      "loss": 4.5333,
      "num_input_tokens_seen": 47440976,
      "step": 35010
    },
    {
      "epoch": 1.1207669163305807,
      "grad_norm": 29.875,
      "learning_rate": 1.8925060914991077e-06,
      "loss": 4.2291,
      "num_input_tokens_seen": 47447696,
      "step": 35015
    },
    {
      "epoch": 1.1209269573010692,
      "grad_norm": 29.0,
      "learning_rate": 1.888760862954264e-06,
      "loss": 5.266,
      "num_input_tokens_seen": 47454256,
      "step": 35020
    },
    {
      "epoch": 1.1210869982715574,
      "grad_norm": 21.875,
      "learning_rate": 1.8850191984529309e-06,
      "loss": 4.1171,
      "num_input_tokens_seen": 47461136,
      "step": 35025
    },
    {
      "epoch": 1.121247039242046,
      "grad_norm": 33.0,
      "learning_rate": 1.8812810985721186e-06,
      "loss": 4.5687,
      "num_input_tokens_seen": 47468256,
      "step": 35030
    },
    {
      "epoch": 1.1214070802125344,
      "grad_norm": 27.25,
      "learning_rate": 1.8775465638882856e-06,
      "loss": 3.4143,
      "num_input_tokens_seen": 47474912,
      "step": 35035
    },
    {
      "epoch": 1.121567121183023,
      "grad_norm": 33.5,
      "learning_rate": 1.8738155949773517e-06,
      "loss": 4.5877,
      "num_input_tokens_seen": 47481696,
      "step": 35040
    },
    {
      "epoch": 1.1217271621535112,
      "grad_norm": 23.75,
      "learning_rate": 1.8700881924146707e-06,
      "loss": 4.9101,
      "num_input_tokens_seen": 47488448,
      "step": 35045
    },
    {
      "epoch": 1.1218872031239997,
      "grad_norm": 24.625,
      "learning_rate": 1.8663643567750577e-06,
      "loss": 4.2895,
      "num_input_tokens_seen": 47495360,
      "step": 35050
    },
    {
      "epoch": 1.1220472440944882,
      "grad_norm": 40.0,
      "learning_rate": 1.8626440886327813e-06,
      "loss": 4.9172,
      "num_input_tokens_seen": 47502176,
      "step": 35055
    },
    {
      "epoch": 1.1222072850649767,
      "grad_norm": 21.125,
      "learning_rate": 1.8589273885615432e-06,
      "loss": 4.8748,
      "num_input_tokens_seen": 47509280,
      "step": 35060
    },
    {
      "epoch": 1.122367326035465,
      "grad_norm": 26.25,
      "learning_rate": 1.8552142571345133e-06,
      "loss": 3.9917,
      "num_input_tokens_seen": 47515536,
      "step": 35065
    },
    {
      "epoch": 1.1225273670059535,
      "grad_norm": 36.0,
      "learning_rate": 1.8515046949243025e-06,
      "loss": 4.2694,
      "num_input_tokens_seen": 47522192,
      "step": 35070
    },
    {
      "epoch": 1.122687407976442,
      "grad_norm": 22.25,
      "learning_rate": 1.8477987025029674e-06,
      "loss": 4.617,
      "num_input_tokens_seen": 47528752,
      "step": 35075
    },
    {
      "epoch": 1.1228474489469304,
      "grad_norm": 12.5,
      "learning_rate": 1.8440962804420232e-06,
      "loss": 4.5448,
      "num_input_tokens_seen": 47535904,
      "step": 35080
    },
    {
      "epoch": 1.123007489917419,
      "grad_norm": 48.25,
      "learning_rate": 1.8403974293124265e-06,
      "loss": 4.1324,
      "num_input_tokens_seen": 47543120,
      "step": 35085
    },
    {
      "epoch": 1.1231675308879072,
      "grad_norm": 37.0,
      "learning_rate": 1.8367021496845854e-06,
      "loss": 5.1584,
      "num_input_tokens_seen": 47549984,
      "step": 35090
    },
    {
      "epoch": 1.1233275718583957,
      "grad_norm": 24.5,
      "learning_rate": 1.8330104421283662e-06,
      "loss": 5.3236,
      "num_input_tokens_seen": 47556960,
      "step": 35095
    },
    {
      "epoch": 1.1234876128288842,
      "grad_norm": 76.5,
      "learning_rate": 1.8293223072130717e-06,
      "loss": 6.0109,
      "num_input_tokens_seen": 47563616,
      "step": 35100
    },
    {
      "epoch": 1.1236476537993727,
      "grad_norm": 22.625,
      "learning_rate": 1.8256377455074525e-06,
      "loss": 4.2862,
      "num_input_tokens_seen": 47570144,
      "step": 35105
    },
    {
      "epoch": 1.123807694769861,
      "grad_norm": 27.875,
      "learning_rate": 1.8219567575797263e-06,
      "loss": 4.2156,
      "num_input_tokens_seen": 47576576,
      "step": 35110
    },
    {
      "epoch": 1.1239677357403495,
      "grad_norm": 33.25,
      "learning_rate": 1.8182793439975365e-06,
      "loss": 3.5468,
      "num_input_tokens_seen": 47583584,
      "step": 35115
    },
    {
      "epoch": 1.124127776710838,
      "grad_norm": 29.25,
      "learning_rate": 1.8146055053279958e-06,
      "loss": 3.6874,
      "num_input_tokens_seen": 47590304,
      "step": 35120
    },
    {
      "epoch": 1.1242878176813265,
      "grad_norm": 34.25,
      "learning_rate": 1.8109352421376486e-06,
      "loss": 4.482,
      "num_input_tokens_seen": 47596560,
      "step": 35125
    },
    {
      "epoch": 1.124447858651815,
      "grad_norm": 22.5,
      "learning_rate": 1.8072685549924972e-06,
      "loss": 5.152,
      "num_input_tokens_seen": 47603424,
      "step": 35130
    },
    {
      "epoch": 1.1246078996223032,
      "grad_norm": 25.25,
      "learning_rate": 1.8036054444579982e-06,
      "loss": 4.3969,
      "num_input_tokens_seen": 47610160,
      "step": 35135
    },
    {
      "epoch": 1.1247679405927917,
      "grad_norm": 37.0,
      "learning_rate": 1.7999459110990407e-06,
      "loss": 4.4444,
      "num_input_tokens_seen": 47617136,
      "step": 35140
    },
    {
      "epoch": 1.1249279815632802,
      "grad_norm": 38.0,
      "learning_rate": 1.7962899554799712e-06,
      "loss": 4.0082,
      "num_input_tokens_seen": 47623536,
      "step": 35145
    },
    {
      "epoch": 1.1250880225337687,
      "grad_norm": 17.375,
      "learning_rate": 1.7926375781645937e-06,
      "loss": 3.9765,
      "num_input_tokens_seen": 47630400,
      "step": 35150
    },
    {
      "epoch": 1.125248063504257,
      "grad_norm": 38.0,
      "learning_rate": 1.7889887797161359e-06,
      "loss": 5.1346,
      "num_input_tokens_seen": 47637072,
      "step": 35155
    },
    {
      "epoch": 1.1254081044747455,
      "grad_norm": 38.25,
      "learning_rate": 1.7853435606973028e-06,
      "loss": 5.2647,
      "num_input_tokens_seen": 47643760,
      "step": 35160
    },
    {
      "epoch": 1.125568145445234,
      "grad_norm": 31.25,
      "learning_rate": 1.781701921670223e-06,
      "loss": 4.4695,
      "num_input_tokens_seen": 47650768,
      "step": 35165
    },
    {
      "epoch": 1.1257281864157225,
      "grad_norm": 26.0,
      "learning_rate": 1.7780638631964886e-06,
      "loss": 3.8111,
      "num_input_tokens_seen": 47657600,
      "step": 35170
    },
    {
      "epoch": 1.125888227386211,
      "grad_norm": 22.5,
      "learning_rate": 1.7744293858371314e-06,
      "loss": 4.0419,
      "num_input_tokens_seen": 47664304,
      "step": 35175
    },
    {
      "epoch": 1.1260482683566992,
      "grad_norm": 20.25,
      "learning_rate": 1.770798490152631e-06,
      "loss": 4.4942,
      "num_input_tokens_seen": 47670992,
      "step": 35180
    },
    {
      "epoch": 1.1262083093271877,
      "grad_norm": 15.25,
      "learning_rate": 1.767171176702917e-06,
      "loss": 3.7051,
      "num_input_tokens_seen": 47677856,
      "step": 35185
    },
    {
      "epoch": 1.1263683502976762,
      "grad_norm": 25.5,
      "learning_rate": 1.7635474460473755e-06,
      "loss": 4.185,
      "num_input_tokens_seen": 47684640,
      "step": 35190
    },
    {
      "epoch": 1.1265283912681647,
      "grad_norm": 19.375,
      "learning_rate": 1.7599272987448206e-06,
      "loss": 4.3818,
      "num_input_tokens_seen": 47691424,
      "step": 35195
    },
    {
      "epoch": 1.126688432238653,
      "grad_norm": 32.25,
      "learning_rate": 1.7563107353535362e-06,
      "loss": 4.6721,
      "num_input_tokens_seen": 47698224,
      "step": 35200
    },
    {
      "epoch": 1.126688432238653,
      "eval_loss": 4.490024566650391,
      "eval_runtime": 488.5672,
      "eval_samples_per_second": 28.422,
      "eval_steps_per_second": 14.211,
      "num_input_tokens_seen": 47698224,
      "step": 35200
    },
    {
      "epoch": 1.1268484732091415,
      "grad_norm": 14.1875,
      "learning_rate": 1.7526977564312263e-06,
      "loss": 4.0999,
      "num_input_tokens_seen": 47705296,
      "step": 35205
    },
    {
      "epoch": 1.12700851417963,
      "grad_norm": 19.5,
      "learning_rate": 1.7490883625350701e-06,
      "loss": 4.8385,
      "num_input_tokens_seen": 47711824,
      "step": 35210
    },
    {
      "epoch": 1.1271685551501185,
      "grad_norm": 19.125,
      "learning_rate": 1.7454825542216807e-06,
      "loss": 4.9883,
      "num_input_tokens_seen": 47718032,
      "step": 35215
    },
    {
      "epoch": 1.127328596120607,
      "grad_norm": 28.875,
      "learning_rate": 1.7418803320471105e-06,
      "loss": 5.1861,
      "num_input_tokens_seen": 47724544,
      "step": 35220
    },
    {
      "epoch": 1.1274886370910953,
      "grad_norm": 35.5,
      "learning_rate": 1.7382816965668737e-06,
      "loss": 4.7036,
      "num_input_tokens_seen": 47730992,
      "step": 35225
    },
    {
      "epoch": 1.1276486780615838,
      "grad_norm": 20.375,
      "learning_rate": 1.7346866483359285e-06,
      "loss": 4.2778,
      "num_input_tokens_seen": 47737712,
      "step": 35230
    },
    {
      "epoch": 1.1278087190320722,
      "grad_norm": 45.0,
      "learning_rate": 1.7310951879086657e-06,
      "loss": 5.0302,
      "num_input_tokens_seen": 47744128,
      "step": 35235
    },
    {
      "epoch": 1.1279687600025607,
      "grad_norm": 38.25,
      "learning_rate": 1.7275073158389471e-06,
      "loss": 4.0254,
      "num_input_tokens_seen": 47750768,
      "step": 35240
    },
    {
      "epoch": 1.128128800973049,
      "grad_norm": 20.625,
      "learning_rate": 1.723923032680061e-06,
      "loss": 4.6425,
      "num_input_tokens_seen": 47757232,
      "step": 35245
    },
    {
      "epoch": 1.1282888419435375,
      "grad_norm": 22.875,
      "learning_rate": 1.7203423389847428e-06,
      "loss": 4.4138,
      "num_input_tokens_seen": 47763696,
      "step": 35250
    },
    {
      "epoch": 1.128448882914026,
      "grad_norm": 30.875,
      "learning_rate": 1.7167652353051928e-06,
      "loss": 3.8835,
      "num_input_tokens_seen": 47770944,
      "step": 35255
    },
    {
      "epoch": 1.1286089238845145,
      "grad_norm": 26.375,
      "learning_rate": 1.7131917221930333e-06,
      "loss": 3.923,
      "num_input_tokens_seen": 47777536,
      "step": 35260
    },
    {
      "epoch": 1.128768964855003,
      "grad_norm": 25.875,
      "learning_rate": 1.7096218001993513e-06,
      "loss": 5.1948,
      "num_input_tokens_seen": 47784320,
      "step": 35265
    },
    {
      "epoch": 1.1289290058254913,
      "grad_norm": 40.25,
      "learning_rate": 1.706055469874676e-06,
      "loss": 3.7263,
      "num_input_tokens_seen": 47791488,
      "step": 35270
    },
    {
      "epoch": 1.1290890467959798,
      "grad_norm": 31.375,
      "learning_rate": 1.702492731768976e-06,
      "loss": 3.4256,
      "num_input_tokens_seen": 47798400,
      "step": 35275
    },
    {
      "epoch": 1.1292490877664683,
      "grad_norm": 33.75,
      "learning_rate": 1.6989335864316724e-06,
      "loss": 4.8323,
      "num_input_tokens_seen": 47804848,
      "step": 35280
    },
    {
      "epoch": 1.1294091287369568,
      "grad_norm": 24.25,
      "learning_rate": 1.6953780344116265e-06,
      "loss": 4.5677,
      "num_input_tokens_seen": 47811808,
      "step": 35285
    },
    {
      "epoch": 1.129569169707445,
      "grad_norm": 24.75,
      "learning_rate": 1.6918260762571497e-06,
      "loss": 4.76,
      "num_input_tokens_seen": 47818416,
      "step": 35290
    },
    {
      "epoch": 1.1297292106779335,
      "grad_norm": 32.75,
      "learning_rate": 1.6882777125160093e-06,
      "loss": 3.294,
      "num_input_tokens_seen": 47825296,
      "step": 35295
    },
    {
      "epoch": 1.129889251648422,
      "grad_norm": 45.5,
      "learning_rate": 1.6847329437353899e-06,
      "loss": 4.1277,
      "num_input_tokens_seen": 47832512,
      "step": 35300
    },
    {
      "epoch": 1.1300492926189105,
      "grad_norm": 23.25,
      "learning_rate": 1.6811917704619511e-06,
      "loss": 4.464,
      "num_input_tokens_seen": 47838816,
      "step": 35305
    },
    {
      "epoch": 1.1302093335893988,
      "grad_norm": 29.0,
      "learning_rate": 1.67765419324179e-06,
      "loss": 3.7369,
      "num_input_tokens_seen": 47845440,
      "step": 35310
    },
    {
      "epoch": 1.1303693745598873,
      "grad_norm": 23.0,
      "learning_rate": 1.6741202126204364e-06,
      "loss": 4.6929,
      "num_input_tokens_seen": 47852144,
      "step": 35315
    },
    {
      "epoch": 1.1305294155303758,
      "grad_norm": 20.75,
      "learning_rate": 1.6705898291428767e-06,
      "loss": 4.7539,
      "num_input_tokens_seen": 47859008,
      "step": 35320
    },
    {
      "epoch": 1.1306894565008643,
      "grad_norm": 28.0,
      "learning_rate": 1.6670630433535395e-06,
      "loss": 4.9205,
      "num_input_tokens_seen": 47865904,
      "step": 35325
    },
    {
      "epoch": 1.1308494974713525,
      "grad_norm": 26.125,
      "learning_rate": 1.6635398557962979e-06,
      "loss": 4.2388,
      "num_input_tokens_seen": 47872736,
      "step": 35330
    },
    {
      "epoch": 1.131009538441841,
      "grad_norm": 21.25,
      "learning_rate": 1.660020267014481e-06,
      "loss": 4.3437,
      "num_input_tokens_seen": 47879744,
      "step": 35335
    },
    {
      "epoch": 1.1311695794123295,
      "grad_norm": 30.0,
      "learning_rate": 1.6565042775508438e-06,
      "loss": 5.1484,
      "num_input_tokens_seen": 47886256,
      "step": 35340
    },
    {
      "epoch": 1.131329620382818,
      "grad_norm": 24.625,
      "learning_rate": 1.6529918879475997e-06,
      "loss": 3.4167,
      "num_input_tokens_seen": 47893344,
      "step": 35345
    },
    {
      "epoch": 1.1314896613533065,
      "grad_norm": 22.625,
      "learning_rate": 1.6494830987464043e-06,
      "loss": 4.3117,
      "num_input_tokens_seen": 47900752,
      "step": 35350
    },
    {
      "epoch": 1.1316497023237948,
      "grad_norm": 26.0,
      "learning_rate": 1.6459779104883555e-06,
      "loss": 4.0474,
      "num_input_tokens_seen": 47907408,
      "step": 35355
    },
    {
      "epoch": 1.1318097432942833,
      "grad_norm": 38.5,
      "learning_rate": 1.6424763237140013e-06,
      "loss": 3.3225,
      "num_input_tokens_seen": 47914432,
      "step": 35360
    },
    {
      "epoch": 1.1319697842647718,
      "grad_norm": 50.0,
      "learning_rate": 1.6389783389633207e-06,
      "loss": 3.8075,
      "num_input_tokens_seen": 47921328,
      "step": 35365
    },
    {
      "epoch": 1.1321298252352603,
      "grad_norm": 33.75,
      "learning_rate": 1.6354839567757546e-06,
      "loss": 4.1326,
      "num_input_tokens_seen": 47928096,
      "step": 35370
    },
    {
      "epoch": 1.1322898662057486,
      "grad_norm": 34.0,
      "learning_rate": 1.6319931776901831e-06,
      "loss": 3.793,
      "num_input_tokens_seen": 47934608,
      "step": 35375
    },
    {
      "epoch": 1.132449907176237,
      "grad_norm": 27.5,
      "learning_rate": 1.6285060022449229e-06,
      "loss": 4.397,
      "num_input_tokens_seen": 47941216,
      "step": 35380
    },
    {
      "epoch": 1.1326099481467256,
      "grad_norm": 23.25,
      "learning_rate": 1.6250224309777434e-06,
      "loss": 4.6442,
      "num_input_tokens_seen": 47947584,
      "step": 35385
    },
    {
      "epoch": 1.132769989117214,
      "grad_norm": 17.625,
      "learning_rate": 1.6215424644258515e-06,
      "loss": 4.1369,
      "num_input_tokens_seen": 47954240,
      "step": 35390
    },
    {
      "epoch": 1.1329300300877025,
      "grad_norm": 18.875,
      "learning_rate": 1.6180661031259036e-06,
      "loss": 5.3598,
      "num_input_tokens_seen": 47961168,
      "step": 35395
    },
    {
      "epoch": 1.1330900710581908,
      "grad_norm": 23.0,
      "learning_rate": 1.614593347613999e-06,
      "loss": 3.6621,
      "num_input_tokens_seen": 47967840,
      "step": 35400
    },
    {
      "epoch": 1.1330900710581908,
      "eval_loss": 4.487605094909668,
      "eval_runtime": 487.8599,
      "eval_samples_per_second": 28.463,
      "eval_steps_per_second": 14.232,
      "num_input_tokens_seen": 47967840,
      "step": 35400
    },
    {
      "epoch": 1.1332501120286793,
      "grad_norm": 51.0,
      "learning_rate": 1.6111241984256758e-06,
      "loss": 4.0425,
      "num_input_tokens_seen": 47974240,
      "step": 35405
    },
    {
      "epoch": 1.1334101529991678,
      "grad_norm": 29.5,
      "learning_rate": 1.6076586560959257e-06,
      "loss": 4.644,
      "num_input_tokens_seen": 47980848,
      "step": 35410
    },
    {
      "epoch": 1.1335701939696563,
      "grad_norm": 25.25,
      "learning_rate": 1.604196721159182e-06,
      "loss": 4.1265,
      "num_input_tokens_seen": 47987168,
      "step": 35415
    },
    {
      "epoch": 1.1337302349401446,
      "grad_norm": 28.625,
      "learning_rate": 1.6007383941493092e-06,
      "loss": 6.0291,
      "num_input_tokens_seen": 47994256,
      "step": 35420
    },
    {
      "epoch": 1.133890275910633,
      "grad_norm": 29.25,
      "learning_rate": 1.5972836755996285e-06,
      "loss": 4.8568,
      "num_input_tokens_seen": 48001232,
      "step": 35425
    },
    {
      "epoch": 1.1340503168811216,
      "grad_norm": 21.875,
      "learning_rate": 1.5938325660429076e-06,
      "loss": 3.4926,
      "num_input_tokens_seen": 48008080,
      "step": 35430
    },
    {
      "epoch": 1.13421035785161,
      "grad_norm": 26.625,
      "learning_rate": 1.5903850660113378e-06,
      "loss": 4.6879,
      "num_input_tokens_seen": 48014608,
      "step": 35435
    },
    {
      "epoch": 1.1343703988220986,
      "grad_norm": 23.25,
      "learning_rate": 1.5869411760365826e-06,
      "loss": 4.1681,
      "num_input_tokens_seen": 48021584,
      "step": 35440
    },
    {
      "epoch": 1.1345304397925868,
      "grad_norm": 38.5,
      "learning_rate": 1.58350089664972e-06,
      "loss": 4.9917,
      "num_input_tokens_seen": 48028192,
      "step": 35445
    },
    {
      "epoch": 1.1346904807630753,
      "grad_norm": 27.375,
      "learning_rate": 1.5800642283812865e-06,
      "loss": 4.8018,
      "num_input_tokens_seen": 48034848,
      "step": 35450
    },
    {
      "epoch": 1.1348505217335638,
      "grad_norm": 34.25,
      "learning_rate": 1.5766311717612698e-06,
      "loss": 4.9516,
      "num_input_tokens_seen": 48041632,
      "step": 35455
    },
    {
      "epoch": 1.1350105627040523,
      "grad_norm": 20.5,
      "learning_rate": 1.5732017273190818e-06,
      "loss": 4.272,
      "num_input_tokens_seen": 48048160,
      "step": 35460
    },
    {
      "epoch": 1.1351706036745406,
      "grad_norm": 26.5,
      "learning_rate": 1.5697758955835806e-06,
      "loss": 3.831,
      "num_input_tokens_seen": 48054784,
      "step": 35465
    },
    {
      "epoch": 1.135330644645029,
      "grad_norm": 18.625,
      "learning_rate": 1.566353677083085e-06,
      "loss": 4.43,
      "num_input_tokens_seen": 48062144,
      "step": 35470
    },
    {
      "epoch": 1.1354906856155176,
      "grad_norm": 25.125,
      "learning_rate": 1.562935072345334e-06,
      "loss": 4.8943,
      "num_input_tokens_seen": 48069712,
      "step": 35475
    },
    {
      "epoch": 1.135650726586006,
      "grad_norm": 28.0,
      "learning_rate": 1.5595200818975281e-06,
      "loss": 4.1436,
      "num_input_tokens_seen": 48076336,
      "step": 35480
    },
    {
      "epoch": 1.1358107675564946,
      "grad_norm": 36.25,
      "learning_rate": 1.5561087062662905e-06,
      "loss": 3.2942,
      "num_input_tokens_seen": 48083056,
      "step": 35485
    },
    {
      "epoch": 1.1359708085269828,
      "grad_norm": 42.0,
      "learning_rate": 1.5527009459777087e-06,
      "loss": 4.8974,
      "num_input_tokens_seen": 48089440,
      "step": 35490
    },
    {
      "epoch": 1.1361308494974713,
      "grad_norm": 34.25,
      "learning_rate": 1.5492968015572984e-06,
      "loss": 4.8475,
      "num_input_tokens_seen": 48096336,
      "step": 35495
    },
    {
      "epoch": 1.1362908904679598,
      "grad_norm": 54.0,
      "learning_rate": 1.5458962735300203e-06,
      "loss": 4.2642,
      "num_input_tokens_seen": 48103312,
      "step": 35500
    },
    {
      "epoch": 1.1364509314384483,
      "grad_norm": 30.75,
      "learning_rate": 1.54249936242028e-06,
      "loss": 5.4328,
      "num_input_tokens_seen": 48110304,
      "step": 35505
    },
    {
      "epoch": 1.1366109724089366,
      "grad_norm": 20.375,
      "learning_rate": 1.5391060687519222e-06,
      "loss": 3.411,
      "num_input_tokens_seen": 48116864,
      "step": 35510
    },
    {
      "epoch": 1.136771013379425,
      "grad_norm": 27.5,
      "learning_rate": 1.5357163930482367e-06,
      "loss": 4.2048,
      "num_input_tokens_seen": 48123632,
      "step": 35515
    },
    {
      "epoch": 1.1369310543499136,
      "grad_norm": 32.75,
      "learning_rate": 1.532330335831955e-06,
      "loss": 3.7082,
      "num_input_tokens_seen": 48130848,
      "step": 35520
    },
    {
      "epoch": 1.137091095320402,
      "grad_norm": 35.25,
      "learning_rate": 1.5289478976252491e-06,
      "loss": 3.6215,
      "num_input_tokens_seen": 48137296,
      "step": 35525
    },
    {
      "epoch": 1.1372511362908906,
      "grad_norm": 29.875,
      "learning_rate": 1.5255690789497345e-06,
      "loss": 4.5689,
      "num_input_tokens_seen": 48143840,
      "step": 35530
    },
    {
      "epoch": 1.1374111772613789,
      "grad_norm": 37.75,
      "learning_rate": 1.5221938803264641e-06,
      "loss": 4.8454,
      "num_input_tokens_seen": 48150272,
      "step": 35535
    },
    {
      "epoch": 1.1375712182318674,
      "grad_norm": 26.0,
      "learning_rate": 1.518822302275938e-06,
      "loss": 3.2767,
      "num_input_tokens_seen": 48157024,
      "step": 35540
    },
    {
      "epoch": 1.1377312592023558,
      "grad_norm": 17.375,
      "learning_rate": 1.5154543453180958e-06,
      "loss": 3.8102,
      "num_input_tokens_seen": 48163792,
      "step": 35545
    },
    {
      "epoch": 1.1378913001728443,
      "grad_norm": 31.25,
      "learning_rate": 1.5120900099723167e-06,
      "loss": 4.3983,
      "num_input_tokens_seen": 48170496,
      "step": 35550
    },
    {
      "epoch": 1.1380513411433326,
      "grad_norm": 28.375,
      "learning_rate": 1.5087292967574273e-06,
      "loss": 4.4006,
      "num_input_tokens_seen": 48177488,
      "step": 35555
    },
    {
      "epoch": 1.1382113821138211,
      "grad_norm": 40.5,
      "learning_rate": 1.5053722061916908e-06,
      "loss": 4.1316,
      "num_input_tokens_seen": 48183968,
      "step": 35560
    },
    {
      "epoch": 1.1383714230843096,
      "grad_norm": 30.875,
      "learning_rate": 1.5020187387928124e-06,
      "loss": 4.0438,
      "num_input_tokens_seen": 48190608,
      "step": 35565
    },
    {
      "epoch": 1.138531464054798,
      "grad_norm": 41.0,
      "learning_rate": 1.4986688950779343e-06,
      "loss": 4.7761,
      "num_input_tokens_seen": 48197456,
      "step": 35570
    },
    {
      "epoch": 1.1386915050252864,
      "grad_norm": 43.75,
      "learning_rate": 1.495322675563654e-06,
      "loss": 3.8504,
      "num_input_tokens_seen": 48204736,
      "step": 35575
    },
    {
      "epoch": 1.1388515459957749,
      "grad_norm": 33.5,
      "learning_rate": 1.4919800807659922e-06,
      "loss": 4.2114,
      "num_input_tokens_seen": 48212144,
      "step": 35580
    },
    {
      "epoch": 1.1390115869662634,
      "grad_norm": 29.125,
      "learning_rate": 1.4886411112004255e-06,
      "loss": 5.6383,
      "num_input_tokens_seen": 48219072,
      "step": 35585
    },
    {
      "epoch": 1.1391716279367519,
      "grad_norm": 28.75,
      "learning_rate": 1.4853057673818588e-06,
      "loss": 4.1196,
      "num_input_tokens_seen": 48225696,
      "step": 35590
    },
    {
      "epoch": 1.1393316689072401,
      "grad_norm": 18.875,
      "learning_rate": 1.481974049824647e-06,
      "loss": 3.9784,
      "num_input_tokens_seen": 48232496,
      "step": 35595
    },
    {
      "epoch": 1.1394917098777286,
      "grad_norm": 39.25,
      "learning_rate": 1.4786459590425849e-06,
      "loss": 4.291,
      "num_input_tokens_seen": 48239792,
      "step": 35600
    },
    {
      "epoch": 1.1394917098777286,
      "eval_loss": 4.474913120269775,
      "eval_runtime": 488.3451,
      "eval_samples_per_second": 28.435,
      "eval_steps_per_second": 14.217,
      "num_input_tokens_seen": 48239792,
      "step": 35600
    },
    {
      "epoch": 1.1396517508482171,
      "grad_norm": 33.25,
      "learning_rate": 1.4753214955489036e-06,
      "loss": 4.3788,
      "num_input_tokens_seen": 48246432,
      "step": 35605
    },
    {
      "epoch": 1.1398117918187056,
      "grad_norm": 38.0,
      "learning_rate": 1.4720006598562737e-06,
      "loss": 4.1739,
      "num_input_tokens_seen": 48253616,
      "step": 35610
    },
    {
      "epoch": 1.1399718327891941,
      "grad_norm": 20.125,
      "learning_rate": 1.4686834524768185e-06,
      "loss": 4.284,
      "num_input_tokens_seen": 48260080,
      "step": 35615
    },
    {
      "epoch": 1.1401318737596824,
      "grad_norm": 29.625,
      "learning_rate": 1.4653698739220844e-06,
      "loss": 4.8698,
      "num_input_tokens_seen": 48266800,
      "step": 35620
    },
    {
      "epoch": 1.1402919147301709,
      "grad_norm": 25.375,
      "learning_rate": 1.4620599247030715e-06,
      "loss": 3.9347,
      "num_input_tokens_seen": 48273760,
      "step": 35625
    },
    {
      "epoch": 1.1404519557006594,
      "grad_norm": 20.75,
      "learning_rate": 1.4587536053302125e-06,
      "loss": 4.4317,
      "num_input_tokens_seen": 48280784,
      "step": 35630
    },
    {
      "epoch": 1.1406119966711479,
      "grad_norm": 44.5,
      "learning_rate": 1.4554509163133862e-06,
      "loss": 4.6592,
      "num_input_tokens_seen": 48287552,
      "step": 35635
    },
    {
      "epoch": 1.1407720376416361,
      "grad_norm": 37.25,
      "learning_rate": 1.4521518581619098e-06,
      "loss": 4.2376,
      "num_input_tokens_seen": 48295008,
      "step": 35640
    },
    {
      "epoch": 1.1409320786121246,
      "grad_norm": 20.125,
      "learning_rate": 1.4488564313845348e-06,
      "loss": 3.9789,
      "num_input_tokens_seen": 48301888,
      "step": 35645
    },
    {
      "epoch": 1.1410921195826131,
      "grad_norm": 30.25,
      "learning_rate": 1.4455646364894603e-06,
      "loss": 5.8819,
      "num_input_tokens_seen": 48309472,
      "step": 35650
    },
    {
      "epoch": 1.1412521605531016,
      "grad_norm": 30.25,
      "learning_rate": 1.4422764739843247e-06,
      "loss": 4.9651,
      "num_input_tokens_seen": 48316592,
      "step": 35655
    },
    {
      "epoch": 1.1414122015235901,
      "grad_norm": 23.0,
      "learning_rate": 1.4389919443762e-06,
      "loss": 4.8589,
      "num_input_tokens_seen": 48323456,
      "step": 35660
    },
    {
      "epoch": 1.1415722424940784,
      "grad_norm": 39.75,
      "learning_rate": 1.4357110481716063e-06,
      "loss": 4.6649,
      "num_input_tokens_seen": 48330688,
      "step": 35665
    },
    {
      "epoch": 1.141732283464567,
      "grad_norm": 26.5,
      "learning_rate": 1.4324337858764941e-06,
      "loss": 3.5452,
      "num_input_tokens_seen": 48337536,
      "step": 35670
    },
    {
      "epoch": 1.1418923244350554,
      "grad_norm": 31.625,
      "learning_rate": 1.4291601579962622e-06,
      "loss": 3.4995,
      "num_input_tokens_seen": 48344336,
      "step": 35675
    },
    {
      "epoch": 1.142052365405544,
      "grad_norm": 49.5,
      "learning_rate": 1.42589016503574e-06,
      "loss": 5.1434,
      "num_input_tokens_seen": 48351104,
      "step": 35680
    },
    {
      "epoch": 1.1422124063760322,
      "grad_norm": 48.0,
      "learning_rate": 1.4226238074992099e-06,
      "loss": 5.3437,
      "num_input_tokens_seen": 48357904,
      "step": 35685
    },
    {
      "epoch": 1.1423724473465207,
      "grad_norm": 47.25,
      "learning_rate": 1.4193610858903778e-06,
      "loss": 4.1607,
      "num_input_tokens_seen": 48364752,
      "step": 35690
    },
    {
      "epoch": 1.1425324883170092,
      "grad_norm": 24.125,
      "learning_rate": 1.416102000712402e-06,
      "loss": 4.4321,
      "num_input_tokens_seen": 48371568,
      "step": 35695
    },
    {
      "epoch": 1.1426925292874976,
      "grad_norm": 34.0,
      "learning_rate": 1.4128465524678668e-06,
      "loss": 4.456,
      "num_input_tokens_seen": 48378576,
      "step": 35700
    },
    {
      "epoch": 1.1428525702579861,
      "grad_norm": 21.125,
      "learning_rate": 1.4095947416588124e-06,
      "loss": 3.5089,
      "num_input_tokens_seen": 48384928,
      "step": 35705
    },
    {
      "epoch": 1.1430126112284744,
      "grad_norm": 28.125,
      "learning_rate": 1.4063465687866983e-06,
      "loss": 5.272,
      "num_input_tokens_seen": 48392224,
      "step": 35710
    },
    {
      "epoch": 1.143172652198963,
      "grad_norm": 28.25,
      "learning_rate": 1.4031020343524438e-06,
      "loss": 5.1614,
      "num_input_tokens_seen": 48398928,
      "step": 35715
    },
    {
      "epoch": 1.1433326931694514,
      "grad_norm": 51.5,
      "learning_rate": 1.3998611388563926e-06,
      "loss": 4.0608,
      "num_input_tokens_seen": 48405824,
      "step": 35720
    },
    {
      "epoch": 1.14349273413994,
      "grad_norm": 26.25,
      "learning_rate": 1.3966238827983314e-06,
      "loss": 4.0844,
      "num_input_tokens_seen": 48412912,
      "step": 35725
    },
    {
      "epoch": 1.1436527751104282,
      "grad_norm": 39.25,
      "learning_rate": 1.393390266677483e-06,
      "loss": 4.5816,
      "num_input_tokens_seen": 48419472,
      "step": 35730
    },
    {
      "epoch": 1.1438128160809167,
      "grad_norm": 37.25,
      "learning_rate": 1.3901602909925204e-06,
      "loss": 4.2316,
      "num_input_tokens_seen": 48426112,
      "step": 35735
    },
    {
      "epoch": 1.1439728570514052,
      "grad_norm": 20.0,
      "learning_rate": 1.3869339562415373e-06,
      "loss": 3.5145,
      "num_input_tokens_seen": 48432736,
      "step": 35740
    },
    {
      "epoch": 1.1441328980218937,
      "grad_norm": 25.625,
      "learning_rate": 1.38371126292208e-06,
      "loss": 4.3169,
      "num_input_tokens_seen": 48439616,
      "step": 35745
    },
    {
      "epoch": 1.1442929389923822,
      "grad_norm": 20.75,
      "learning_rate": 1.3804922115311286e-06,
      "loss": 4.7657,
      "num_input_tokens_seen": 48446240,
      "step": 35750
    },
    {
      "epoch": 1.1444529799628704,
      "grad_norm": 34.5,
      "learning_rate": 1.3772768025650945e-06,
      "loss": 4.2253,
      "num_input_tokens_seen": 48453648,
      "step": 35755
    },
    {
      "epoch": 1.144613020933359,
      "grad_norm": 40.25,
      "learning_rate": 1.3740650365198448e-06,
      "loss": 3.3049,
      "num_input_tokens_seen": 48460336,
      "step": 35760
    },
    {
      "epoch": 1.1447730619038474,
      "grad_norm": 22.125,
      "learning_rate": 1.3708569138906612e-06,
      "loss": 4.5175,
      "num_input_tokens_seen": 48467008,
      "step": 35765
    },
    {
      "epoch": 1.144933102874336,
      "grad_norm": 50.5,
      "learning_rate": 1.367652435172287e-06,
      "loss": 4.2563,
      "num_input_tokens_seen": 48473664,
      "step": 35770
    },
    {
      "epoch": 1.1450931438448242,
      "grad_norm": 30.875,
      "learning_rate": 1.364451600858893e-06,
      "loss": 4.826,
      "num_input_tokens_seen": 48480656,
      "step": 35775
    },
    {
      "epoch": 1.1452531848153127,
      "grad_norm": 50.75,
      "learning_rate": 1.3612544114440823e-06,
      "loss": 3.997,
      "num_input_tokens_seen": 48487536,
      "step": 35780
    },
    {
      "epoch": 1.1454132257858012,
      "grad_norm": 43.25,
      "learning_rate": 1.3580608674209072e-06,
      "loss": 4.0375,
      "num_input_tokens_seen": 48494416,
      "step": 35785
    },
    {
      "epoch": 1.1455732667562897,
      "grad_norm": 18.0,
      "learning_rate": 1.3548709692818434e-06,
      "loss": 4.563,
      "num_input_tokens_seen": 48501520,
      "step": 35790
    },
    {
      "epoch": 1.1457333077267782,
      "grad_norm": 44.0,
      "learning_rate": 1.3516847175188223e-06,
      "loss": 4.6852,
      "num_input_tokens_seen": 48508432,
      "step": 35795
    },
    {
      "epoch": 1.1458933486972664,
      "grad_norm": 35.25,
      "learning_rate": 1.348502112623204e-06,
      "loss": 4.2103,
      "num_input_tokens_seen": 48514752,
      "step": 35800
    },
    {
      "epoch": 1.1458933486972664,
      "eval_loss": 4.5065388679504395,
      "eval_runtime": 487.9541,
      "eval_samples_per_second": 28.458,
      "eval_steps_per_second": 14.229,
      "num_input_tokens_seen": 48514752,
      "step": 35800
    },
    {
      "epoch": 1.146053389667755,
      "grad_norm": 37.0,
      "learning_rate": 1.3453231550857787e-06,
      "loss": 4.0265,
      "num_input_tokens_seen": 48521664,
      "step": 35805
    },
    {
      "epoch": 1.1462134306382434,
      "grad_norm": 28.5,
      "learning_rate": 1.3421478453967878e-06,
      "loss": 4.4236,
      "num_input_tokens_seen": 48528480,
      "step": 35810
    },
    {
      "epoch": 1.146373471608732,
      "grad_norm": 27.0,
      "learning_rate": 1.3389761840459065e-06,
      "loss": 4.3317,
      "num_input_tokens_seen": 48534784,
      "step": 35815
    },
    {
      "epoch": 1.1465335125792202,
      "grad_norm": 26.5,
      "learning_rate": 1.3358081715222376e-06,
      "loss": 3.7781,
      "num_input_tokens_seen": 48541248,
      "step": 35820
    },
    {
      "epoch": 1.1466935535497087,
      "grad_norm": 71.0,
      "learning_rate": 1.3326438083143295e-06,
      "loss": 5.0402,
      "num_input_tokens_seen": 48547936,
      "step": 35825
    },
    {
      "epoch": 1.1468535945201972,
      "grad_norm": 26.875,
      "learning_rate": 1.3294830949101723e-06,
      "loss": 4.4972,
      "num_input_tokens_seen": 48554576,
      "step": 35830
    },
    {
      "epoch": 1.1470136354906857,
      "grad_norm": 35.25,
      "learning_rate": 1.3263260317971815e-06,
      "loss": 4.5911,
      "num_input_tokens_seen": 48561360,
      "step": 35835
    },
    {
      "epoch": 1.1471736764611742,
      "grad_norm": 31.25,
      "learning_rate": 1.3231726194622208e-06,
      "loss": 4.8396,
      "num_input_tokens_seen": 48568016,
      "step": 35840
    },
    {
      "epoch": 1.1473337174316625,
      "grad_norm": 14.75,
      "learning_rate": 1.3200228583915814e-06,
      "loss": 4.532,
      "num_input_tokens_seen": 48574672,
      "step": 35845
    },
    {
      "epoch": 1.147493758402151,
      "grad_norm": 52.75,
      "learning_rate": 1.3168767490709971e-06,
      "loss": 5.6417,
      "num_input_tokens_seen": 48581216,
      "step": 35850
    },
    {
      "epoch": 1.1476537993726394,
      "grad_norm": 23.25,
      "learning_rate": 1.3137342919856437e-06,
      "loss": 5.2771,
      "num_input_tokens_seen": 48587792,
      "step": 35855
    },
    {
      "epoch": 1.1478138403431277,
      "grad_norm": 21.5,
      "learning_rate": 1.310595487620117e-06,
      "loss": 3.5818,
      "num_input_tokens_seen": 48594656,
      "step": 35860
    },
    {
      "epoch": 1.1479738813136162,
      "grad_norm": 31.5,
      "learning_rate": 1.3074603364584715e-06,
      "loss": 3.8736,
      "num_input_tokens_seen": 48601696,
      "step": 35865
    },
    {
      "epoch": 1.1481339222841047,
      "grad_norm": 28.75,
      "learning_rate": 1.3043288389841758e-06,
      "loss": 3.8699,
      "num_input_tokens_seen": 48608320,
      "step": 35870
    },
    {
      "epoch": 1.1482939632545932,
      "grad_norm": 30.5,
      "learning_rate": 1.3012009956801546e-06,
      "loss": 3.4683,
      "num_input_tokens_seen": 48615184,
      "step": 35875
    },
    {
      "epoch": 1.1484540042250817,
      "grad_norm": 17.75,
      "learning_rate": 1.2980768070287586e-06,
      "loss": 5.3304,
      "num_input_tokens_seen": 48622144,
      "step": 35880
    },
    {
      "epoch": 1.14861404519557,
      "grad_norm": 29.375,
      "learning_rate": 1.2949562735117716e-06,
      "loss": 4.5857,
      "num_input_tokens_seen": 48628576,
      "step": 35885
    },
    {
      "epoch": 1.1487740861660585,
      "grad_norm": 36.25,
      "learning_rate": 1.291839395610428e-06,
      "loss": 5.5055,
      "num_input_tokens_seen": 48635136,
      "step": 35890
    },
    {
      "epoch": 1.148934127136547,
      "grad_norm": 24.875,
      "learning_rate": 1.2887261738053852e-06,
      "loss": 4.1182,
      "num_input_tokens_seen": 48641744,
      "step": 35895
    },
    {
      "epoch": 1.1490941681070355,
      "grad_norm": 22.625,
      "learning_rate": 1.2856166085767396e-06,
      "loss": 4.3671,
      "num_input_tokens_seen": 48648560,
      "step": 35900
    },
    {
      "epoch": 1.1492542090775237,
      "grad_norm": 25.625,
      "learning_rate": 1.2825107004040272e-06,
      "loss": 3.7549,
      "num_input_tokens_seen": 48655504,
      "step": 35905
    },
    {
      "epoch": 1.1494142500480122,
      "grad_norm": 20.125,
      "learning_rate": 1.2794084497662146e-06,
      "loss": 4.4923,
      "num_input_tokens_seen": 48662400,
      "step": 35910
    },
    {
      "epoch": 1.1495742910185007,
      "grad_norm": 23.625,
      "learning_rate": 1.276309857141711e-06,
      "loss": 3.4424,
      "num_input_tokens_seen": 48669264,
      "step": 35915
    },
    {
      "epoch": 1.1497343319889892,
      "grad_norm": 33.75,
      "learning_rate": 1.273214923008359e-06,
      "loss": 4.7565,
      "num_input_tokens_seen": 48675920,
      "step": 35920
    },
    {
      "epoch": 1.1498943729594777,
      "grad_norm": 29.625,
      "learning_rate": 1.2701236478434352e-06,
      "loss": 4.8068,
      "num_input_tokens_seen": 48682576,
      "step": 35925
    },
    {
      "epoch": 1.150054413929966,
      "grad_norm": 40.75,
      "learning_rate": 1.2670360321236502e-06,
      "loss": 4.7583,
      "num_input_tokens_seen": 48689024,
      "step": 35930
    },
    {
      "epoch": 1.1502144549004545,
      "grad_norm": 23.25,
      "learning_rate": 1.2639520763251617e-06,
      "loss": 3.8465,
      "num_input_tokens_seen": 48695920,
      "step": 35935
    },
    {
      "epoch": 1.150374495870943,
      "grad_norm": 26.5,
      "learning_rate": 1.2608717809235448e-06,
      "loss": 3.9231,
      "num_input_tokens_seen": 48702464,
      "step": 35940
    },
    {
      "epoch": 1.1505345368414315,
      "grad_norm": 30.25,
      "learning_rate": 1.2577951463938282e-06,
      "loss": 4.5021,
      "num_input_tokens_seen": 48708848,
      "step": 35945
    },
    {
      "epoch": 1.1506945778119197,
      "grad_norm": 23.875,
      "learning_rate": 1.2547221732104569e-06,
      "loss": 4.1103,
      "num_input_tokens_seen": 48715392,
      "step": 35950
    },
    {
      "epoch": 1.1508546187824082,
      "grad_norm": 24.0,
      "learning_rate": 1.25165286184733e-06,
      "loss": 5.7494,
      "num_input_tokens_seen": 48722176,
      "step": 35955
    },
    {
      "epoch": 1.1510146597528967,
      "grad_norm": 21.75,
      "learning_rate": 1.248587212777777e-06,
      "loss": 4.8193,
      "num_input_tokens_seen": 48728832,
      "step": 35960
    },
    {
      "epoch": 1.1511747007233852,
      "grad_norm": 27.875,
      "learning_rate": 1.2455252264745532e-06,
      "loss": 4.6498,
      "num_input_tokens_seen": 48735712,
      "step": 35965
    },
    {
      "epoch": 1.1513347416938737,
      "grad_norm": 29.875,
      "learning_rate": 1.2424669034098528e-06,
      "loss": 4.3781,
      "num_input_tokens_seen": 48742464,
      "step": 35970
    },
    {
      "epoch": 1.151494782664362,
      "grad_norm": 35.0,
      "learning_rate": 1.2394122440553185e-06,
      "loss": 4.6647,
      "num_input_tokens_seen": 48749136,
      "step": 35975
    },
    {
      "epoch": 1.1516548236348505,
      "grad_norm": 27.875,
      "learning_rate": 1.2363612488820037e-06,
      "loss": 4.5726,
      "num_input_tokens_seen": 48755840,
      "step": 35980
    },
    {
      "epoch": 1.151814864605339,
      "grad_norm": 30.875,
      "learning_rate": 1.2333139183604208e-06,
      "loss": 4.427,
      "num_input_tokens_seen": 48762592,
      "step": 35985
    },
    {
      "epoch": 1.1519749055758275,
      "grad_norm": 53.25,
      "learning_rate": 1.2302702529604998e-06,
      "loss": 4.2281,
      "num_input_tokens_seen": 48769584,
      "step": 35990
    },
    {
      "epoch": 1.1521349465463158,
      "grad_norm": 31.0,
      "learning_rate": 1.227230253151615e-06,
      "loss": 3.488,
      "num_input_tokens_seen": 48776592,
      "step": 35995
    },
    {
      "epoch": 1.1522949875168043,
      "grad_norm": 21.875,
      "learning_rate": 1.2241939194025748e-06,
      "loss": 4.6992,
      "num_input_tokens_seen": 48783136,
      "step": 36000
    },
    {
      "epoch": 1.1522949875168043,
      "eval_loss": 4.504053592681885,
      "eval_runtime": 488.0699,
      "eval_samples_per_second": 28.451,
      "eval_steps_per_second": 14.225,
      "num_input_tokens_seen": 48783136,
      "step": 36000
    },
    {
      "epoch": 1.1524550284872928,
      "grad_norm": 30.625,
      "learning_rate": 1.2211612521816156e-06,
      "loss": 4.6172,
      "num_input_tokens_seen": 48789776,
      "step": 36005
    },
    {
      "epoch": 1.1526150694577812,
      "grad_norm": 15.125,
      "learning_rate": 1.2181322519564137e-06,
      "loss": 5.122,
      "num_input_tokens_seen": 48796720,
      "step": 36010
    },
    {
      "epoch": 1.1527751104282697,
      "grad_norm": 21.5,
      "learning_rate": 1.2151069191940839e-06,
      "loss": 4.9618,
      "num_input_tokens_seen": 48804208,
      "step": 36015
    },
    {
      "epoch": 1.152935151398758,
      "grad_norm": 33.0,
      "learning_rate": 1.2120852543611644e-06,
      "loss": 4.7222,
      "num_input_tokens_seen": 48810848,
      "step": 36020
    },
    {
      "epoch": 1.1530951923692465,
      "grad_norm": 52.75,
      "learning_rate": 1.2090672579236379e-06,
      "loss": 4.8157,
      "num_input_tokens_seen": 48817440,
      "step": 36025
    },
    {
      "epoch": 1.153255233339735,
      "grad_norm": 54.25,
      "learning_rate": 1.2060529303469126e-06,
      "loss": 4.5668,
      "num_input_tokens_seen": 48824064,
      "step": 36030
    },
    {
      "epoch": 1.1534152743102235,
      "grad_norm": 34.75,
      "learning_rate": 1.2030422720958445e-06,
      "loss": 4.067,
      "num_input_tokens_seen": 48830976,
      "step": 36035
    },
    {
      "epoch": 1.1535753152807118,
      "grad_norm": 16.25,
      "learning_rate": 1.200035283634704e-06,
      "loss": 5.2367,
      "num_input_tokens_seen": 48837824,
      "step": 36040
    },
    {
      "epoch": 1.1537353562512003,
      "grad_norm": 19.25,
      "learning_rate": 1.1970319654272144e-06,
      "loss": 5.0594,
      "num_input_tokens_seen": 48844336,
      "step": 36045
    },
    {
      "epoch": 1.1538953972216888,
      "grad_norm": 19.375,
      "learning_rate": 1.1940323179365192e-06,
      "loss": 3.9815,
      "num_input_tokens_seen": 48851696,
      "step": 36050
    },
    {
      "epoch": 1.1540554381921773,
      "grad_norm": 29.75,
      "learning_rate": 1.1910363416252095e-06,
      "loss": 3.92,
      "num_input_tokens_seen": 48858128,
      "step": 36055
    },
    {
      "epoch": 1.1542154791626658,
      "grad_norm": 18.125,
      "learning_rate": 1.1880440369552964e-06,
      "loss": 4.5472,
      "num_input_tokens_seen": 48864928,
      "step": 36060
    },
    {
      "epoch": 1.154375520133154,
      "grad_norm": 19.75,
      "learning_rate": 1.1850554043882328e-06,
      "loss": 4.7674,
      "num_input_tokens_seen": 48871808,
      "step": 36065
    },
    {
      "epoch": 1.1545355611036425,
      "grad_norm": 28.875,
      "learning_rate": 1.1820704443849028e-06,
      "loss": 4.1498,
      "num_input_tokens_seen": 48878768,
      "step": 36070
    },
    {
      "epoch": 1.154695602074131,
      "grad_norm": 35.25,
      "learning_rate": 1.1790891574056219e-06,
      "loss": 4.6765,
      "num_input_tokens_seen": 48885232,
      "step": 36075
    },
    {
      "epoch": 1.1548556430446195,
      "grad_norm": 32.75,
      "learning_rate": 1.1761115439101523e-06,
      "loss": 5.5887,
      "num_input_tokens_seen": 48891728,
      "step": 36080
    },
    {
      "epoch": 1.1550156840151078,
      "grad_norm": 30.5,
      "learning_rate": 1.1731376043576659e-06,
      "loss": 5.1996,
      "num_input_tokens_seen": 48898176,
      "step": 36085
    },
    {
      "epoch": 1.1551757249855963,
      "grad_norm": 42.5,
      "learning_rate": 1.1701673392067875e-06,
      "loss": 4.741,
      "num_input_tokens_seen": 48905008,
      "step": 36090
    },
    {
      "epoch": 1.1553357659560848,
      "grad_norm": 34.75,
      "learning_rate": 1.1672007489155757e-06,
      "loss": 4.1154,
      "num_input_tokens_seen": 48911808,
      "step": 36095
    },
    {
      "epoch": 1.1554958069265733,
      "grad_norm": 25.125,
      "learning_rate": 1.164237833941506e-06,
      "loss": 4.0609,
      "num_input_tokens_seen": 48918432,
      "step": 36100
    },
    {
      "epoch": 1.1556558478970618,
      "grad_norm": 21.875,
      "learning_rate": 1.1612785947415022e-06,
      "loss": 4.1566,
      "num_input_tokens_seen": 48924912,
      "step": 36105
    },
    {
      "epoch": 1.15581588886755,
      "grad_norm": 30.75,
      "learning_rate": 1.1583230317719185e-06,
      "loss": 5.2229,
      "num_input_tokens_seen": 48931488,
      "step": 36110
    },
    {
      "epoch": 1.1559759298380385,
      "grad_norm": 19.375,
      "learning_rate": 1.1553711454885318e-06,
      "loss": 4.0229,
      "num_input_tokens_seen": 48938240,
      "step": 36115
    },
    {
      "epoch": 1.156135970808527,
      "grad_norm": 35.0,
      "learning_rate": 1.152422936346567e-06,
      "loss": 3.8091,
      "num_input_tokens_seen": 48944880,
      "step": 36120
    },
    {
      "epoch": 1.1562960117790153,
      "grad_norm": 27.0,
      "learning_rate": 1.1494784048006718e-06,
      "loss": 3.8434,
      "num_input_tokens_seen": 48951392,
      "step": 36125
    },
    {
      "epoch": 1.1564560527495038,
      "grad_norm": 32.5,
      "learning_rate": 1.1465375513049326e-06,
      "loss": 4.6268,
      "num_input_tokens_seen": 48958016,
      "step": 36130
    },
    {
      "epoch": 1.1566160937199923,
      "grad_norm": 39.0,
      "learning_rate": 1.1436003763128616e-06,
      "loss": 4.6912,
      "num_input_tokens_seen": 48964560,
      "step": 36135
    },
    {
      "epoch": 1.1567761346904808,
      "grad_norm": 18.875,
      "learning_rate": 1.1406668802774106e-06,
      "loss": 4.1557,
      "num_input_tokens_seen": 48971136,
      "step": 36140
    },
    {
      "epoch": 1.1569361756609693,
      "grad_norm": 22.375,
      "learning_rate": 1.137737063650965e-06,
      "loss": 4.6614,
      "num_input_tokens_seen": 48978336,
      "step": 36145
    },
    {
      "epoch": 1.1570962166314576,
      "grad_norm": 19.125,
      "learning_rate": 1.1348109268853323e-06,
      "loss": 5.0476,
      "num_input_tokens_seen": 48985280,
      "step": 36150
    },
    {
      "epoch": 1.157256257601946,
      "grad_norm": 43.75,
      "learning_rate": 1.1318884704317634e-06,
      "loss": 4.9517,
      "num_input_tokens_seen": 48991936,
      "step": 36155
    },
    {
      "epoch": 1.1574162985724346,
      "grad_norm": 39.5,
      "learning_rate": 1.1289696947409417e-06,
      "loss": 4.2583,
      "num_input_tokens_seen": 48998736,
      "step": 36160
    },
    {
      "epoch": 1.157576339542923,
      "grad_norm": 21.0,
      "learning_rate": 1.126054600262974e-06,
      "loss": 3.5469,
      "num_input_tokens_seen": 49005568,
      "step": 36165
    },
    {
      "epoch": 1.1577363805134113,
      "grad_norm": 18.875,
      "learning_rate": 1.1231431874474064e-06,
      "loss": 4.7262,
      "num_input_tokens_seen": 49012128,
      "step": 36170
    },
    {
      "epoch": 1.1578964214838998,
      "grad_norm": 48.25,
      "learning_rate": 1.12023545674321e-06,
      "loss": 4.6586,
      "num_input_tokens_seen": 49018960,
      "step": 36175
    },
    {
      "epoch": 1.1580564624543883,
      "grad_norm": 38.0,
      "learning_rate": 1.117331408598804e-06,
      "loss": 4.1211,
      "num_input_tokens_seen": 49025520,
      "step": 36180
    },
    {
      "epoch": 1.1582165034248768,
      "grad_norm": 46.0,
      "learning_rate": 1.1144310434620191e-06,
      "loss": 4.6733,
      "num_input_tokens_seen": 49032224,
      "step": 36185
    },
    {
      "epoch": 1.1583765443953653,
      "grad_norm": 45.0,
      "learning_rate": 1.1115343617801365e-06,
      "loss": 4.79,
      "num_input_tokens_seen": 49039184,
      "step": 36190
    },
    {
      "epoch": 1.1585365853658536,
      "grad_norm": 34.75,
      "learning_rate": 1.1086413639998515e-06,
      "loss": 4.7887,
      "num_input_tokens_seen": 49045920,
      "step": 36195
    },
    {
      "epoch": 1.158696626336342,
      "grad_norm": 26.125,
      "learning_rate": 1.1057520505673103e-06,
      "loss": 4.4254,
      "num_input_tokens_seen": 49052640,
      "step": 36200
    },
    {
      "epoch": 1.158696626336342,
      "eval_loss": 4.4957380294799805,
      "eval_runtime": 487.6787,
      "eval_samples_per_second": 28.474,
      "eval_steps_per_second": 14.237,
      "num_input_tokens_seen": 49052640,
      "step": 36200
    },
    {
      "epoch": 1.1588566673068306,
      "grad_norm": 22.0,
      "learning_rate": 1.1028664219280727e-06,
      "loss": 4.8018,
      "num_input_tokens_seen": 49059936,
      "step": 36205
    },
    {
      "epoch": 1.159016708277319,
      "grad_norm": 23.0,
      "learning_rate": 1.0999844785271468e-06,
      "loss": 4.8196,
      "num_input_tokens_seen": 49066624,
      "step": 36210
    },
    {
      "epoch": 1.1591767492478073,
      "grad_norm": 32.75,
      "learning_rate": 1.097106220808955e-06,
      "loss": 3.8354,
      "num_input_tokens_seen": 49073024,
      "step": 36215
    },
    {
      "epoch": 1.1593367902182958,
      "grad_norm": 28.75,
      "learning_rate": 1.0942316492173698e-06,
      "loss": 5.4567,
      "num_input_tokens_seen": 49079824,
      "step": 36220
    },
    {
      "epoch": 1.1594968311887843,
      "grad_norm": 23.875,
      "learning_rate": 1.0913607641956841e-06,
      "loss": 3.9753,
      "num_input_tokens_seen": 49086736,
      "step": 36225
    },
    {
      "epoch": 1.1596568721592728,
      "grad_norm": 27.125,
      "learning_rate": 1.0884935661866213e-06,
      "loss": 3.8289,
      "num_input_tokens_seen": 49093248,
      "step": 36230
    },
    {
      "epoch": 1.1598169131297613,
      "grad_norm": 25.875,
      "learning_rate": 1.0856300556323418e-06,
      "loss": 4.53,
      "num_input_tokens_seen": 49099696,
      "step": 36235
    },
    {
      "epoch": 1.1599769541002496,
      "grad_norm": 19.25,
      "learning_rate": 1.0827702329744365e-06,
      "loss": 3.7349,
      "num_input_tokens_seen": 49106320,
      "step": 36240
    },
    {
      "epoch": 1.160136995070738,
      "grad_norm": 33.0,
      "learning_rate": 1.0799140986539197e-06,
      "loss": 4.4786,
      "num_input_tokens_seen": 49113344,
      "step": 36245
    },
    {
      "epoch": 1.1602970360412266,
      "grad_norm": 24.0,
      "learning_rate": 1.0770616531112526e-06,
      "loss": 3.6079,
      "num_input_tokens_seen": 49119952,
      "step": 36250
    },
    {
      "epoch": 1.160457077011715,
      "grad_norm": 22.625,
      "learning_rate": 1.0742128967863085e-06,
      "loss": 4.1053,
      "num_input_tokens_seen": 49126848,
      "step": 36255
    },
    {
      "epoch": 1.1606171179822034,
      "grad_norm": 18.125,
      "learning_rate": 1.071367830118411e-06,
      "loss": 3.2677,
      "num_input_tokens_seen": 49134112,
      "step": 36260
    },
    {
      "epoch": 1.1607771589526918,
      "grad_norm": 28.625,
      "learning_rate": 1.068526453546298e-06,
      "loss": 4.307,
      "num_input_tokens_seen": 49140864,
      "step": 36265
    },
    {
      "epoch": 1.1609371999231803,
      "grad_norm": 31.375,
      "learning_rate": 1.0656887675081467e-06,
      "loss": 3.8064,
      "num_input_tokens_seen": 49148032,
      "step": 36270
    },
    {
      "epoch": 1.1610972408936688,
      "grad_norm": 28.25,
      "learning_rate": 1.0628547724415628e-06,
      "loss": 4.4803,
      "num_input_tokens_seen": 49154592,
      "step": 36275
    },
    {
      "epoch": 1.1612572818641573,
      "grad_norm": 33.25,
      "learning_rate": 1.0600244687835881e-06,
      "loss": 5.1837,
      "num_input_tokens_seen": 49161264,
      "step": 36280
    },
    {
      "epoch": 1.1614173228346456,
      "grad_norm": 40.0,
      "learning_rate": 1.0571978569706876e-06,
      "loss": 5.1386,
      "num_input_tokens_seen": 49168416,
      "step": 36285
    },
    {
      "epoch": 1.161577363805134,
      "grad_norm": 21.5,
      "learning_rate": 1.0543749374387652e-06,
      "loss": 3.756,
      "num_input_tokens_seen": 49175120,
      "step": 36290
    },
    {
      "epoch": 1.1617374047756226,
      "grad_norm": 32.75,
      "learning_rate": 1.051555710623142e-06,
      "loss": 4.1246,
      "num_input_tokens_seen": 49182240,
      "step": 36295
    },
    {
      "epoch": 1.161897445746111,
      "grad_norm": 29.375,
      "learning_rate": 1.0487401769585847e-06,
      "loss": 4.4187,
      "num_input_tokens_seen": 49189168,
      "step": 36300
    },
    {
      "epoch": 1.1620574867165994,
      "grad_norm": 20.25,
      "learning_rate": 1.0459283368792845e-06,
      "loss": 4.3707,
      "num_input_tokens_seen": 49195744,
      "step": 36305
    },
    {
      "epoch": 1.1622175276870879,
      "grad_norm": 21.5,
      "learning_rate": 1.043120190818858e-06,
      "loss": 4.0213,
      "num_input_tokens_seen": 49202432,
      "step": 36310
    },
    {
      "epoch": 1.1623775686575764,
      "grad_norm": 18.0,
      "learning_rate": 1.0403157392103596e-06,
      "loss": 4.1195,
      "num_input_tokens_seen": 49209104,
      "step": 36315
    },
    {
      "epoch": 1.1625376096280648,
      "grad_norm": 16.875,
      "learning_rate": 1.0375149824862735e-06,
      "loss": 4.5532,
      "num_input_tokens_seen": 49215952,
      "step": 36320
    },
    {
      "epoch": 1.1626976505985533,
      "grad_norm": 34.75,
      "learning_rate": 1.034717921078507e-06,
      "loss": 4.6528,
      "num_input_tokens_seen": 49222784,
      "step": 36325
    },
    {
      "epoch": 1.1628576915690416,
      "grad_norm": 16.125,
      "learning_rate": 1.0319245554184009e-06,
      "loss": 4.0502,
      "num_input_tokens_seen": 49229472,
      "step": 36330
    },
    {
      "epoch": 1.1630177325395301,
      "grad_norm": 32.75,
      "learning_rate": 1.0291348859367361e-06,
      "loss": 4.7719,
      "num_input_tokens_seen": 49235904,
      "step": 36335
    },
    {
      "epoch": 1.1631777735100186,
      "grad_norm": 67.5,
      "learning_rate": 1.0263489130637016e-06,
      "loss": 4.5833,
      "num_input_tokens_seen": 49242464,
      "step": 36340
    },
    {
      "epoch": 1.163337814480507,
      "grad_norm": 28.625,
      "learning_rate": 1.0235666372289427e-06,
      "loss": 4.7051,
      "num_input_tokens_seen": 49249104,
      "step": 36345
    },
    {
      "epoch": 1.1634978554509954,
      "grad_norm": 22.625,
      "learning_rate": 1.0207880588615076e-06,
      "loss": 3.9982,
      "num_input_tokens_seen": 49255856,
      "step": 36350
    },
    {
      "epoch": 1.1636578964214839,
      "grad_norm": 11.8125,
      "learning_rate": 1.0180131783898984e-06,
      "loss": 3.6727,
      "num_input_tokens_seen": 49262800,
      "step": 36355
    },
    {
      "epoch": 1.1638179373919724,
      "grad_norm": 42.0,
      "learning_rate": 1.0152419962420362e-06,
      "loss": 4.3051,
      "num_input_tokens_seen": 49269216,
      "step": 36360
    },
    {
      "epoch": 1.1639779783624609,
      "grad_norm": 25.625,
      "learning_rate": 1.0124745128452685e-06,
      "loss": 4.3032,
      "num_input_tokens_seen": 49275840,
      "step": 36365
    },
    {
      "epoch": 1.1641380193329494,
      "grad_norm": 33.0,
      "learning_rate": 1.0097107286263758e-06,
      "loss": 4.7636,
      "num_input_tokens_seen": 49282048,
      "step": 36370
    },
    {
      "epoch": 1.1642980603034376,
      "grad_norm": 16.875,
      "learning_rate": 1.00695064401157e-06,
      "loss": 3.8633,
      "num_input_tokens_seen": 49288544,
      "step": 36375
    },
    {
      "epoch": 1.1644581012739261,
      "grad_norm": 31.25,
      "learning_rate": 1.0041942594264886e-06,
      "loss": 5.1077,
      "num_input_tokens_seen": 49295408,
      "step": 36380
    },
    {
      "epoch": 1.1646181422444146,
      "grad_norm": 45.25,
      "learning_rate": 1.001441575296208e-06,
      "loss": 4.3172,
      "num_input_tokens_seen": 49302160,
      "step": 36385
    },
    {
      "epoch": 1.164778183214903,
      "grad_norm": 27.75,
      "learning_rate": 9.986925920452139e-07,
      "loss": 3.769,
      "num_input_tokens_seen": 49308864,
      "step": 36390
    },
    {
      "epoch": 1.1649382241853914,
      "grad_norm": 28.0,
      "learning_rate": 9.959473100974475e-07,
      "loss": 4.2014,
      "num_input_tokens_seen": 49315248,
      "step": 36395
    },
    {
      "epoch": 1.1650982651558799,
      "grad_norm": 60.0,
      "learning_rate": 9.932057298762564e-07,
      "loss": 5.1153,
      "num_input_tokens_seen": 49321648,
      "step": 36400
    },
    {
      "epoch": 1.1650982651558799,
      "eval_loss": 4.4957380294799805,
      "eval_runtime": 487.9958,
      "eval_samples_per_second": 28.455,
      "eval_steps_per_second": 14.228,
      "num_input_tokens_seen": 49321648,
      "step": 36400
    },
    {
      "epoch": 1.1652583061263684,
      "grad_norm": 31.375,
      "learning_rate": 9.90467851804433e-07,
      "loss": 5.0595,
      "num_input_tokens_seen": 49328544,
      "step": 36405
    },
    {
      "epoch": 1.1654183470968569,
      "grad_norm": 36.0,
      "learning_rate": 9.877336763041895e-07,
      "loss": 3.5012,
      "num_input_tokens_seen": 49335264,
      "step": 36410
    },
    {
      "epoch": 1.1655783880673452,
      "grad_norm": 38.0,
      "learning_rate": 9.850032037971662e-07,
      "loss": 4.8373,
      "num_input_tokens_seen": 49341888,
      "step": 36415
    },
    {
      "epoch": 1.1657384290378336,
      "grad_norm": 19.5,
      "learning_rate": 9.822764347044406e-07,
      "loss": 4.2715,
      "num_input_tokens_seen": 49348512,
      "step": 36420
    },
    {
      "epoch": 1.1658984700083221,
      "grad_norm": 14.9375,
      "learning_rate": 9.795533694465175e-07,
      "loss": 4.2883,
      "num_input_tokens_seen": 49355648,
      "step": 36425
    },
    {
      "epoch": 1.1660585109788106,
      "grad_norm": 42.75,
      "learning_rate": 9.768340084433197e-07,
      "loss": 4.9679,
      "num_input_tokens_seen": 49362320,
      "step": 36430
    },
    {
      "epoch": 1.166218551949299,
      "grad_norm": 23.875,
      "learning_rate": 9.741183521142143e-07,
      "loss": 3.8098,
      "num_input_tokens_seen": 49369088,
      "step": 36435
    },
    {
      "epoch": 1.1663785929197874,
      "grad_norm": 27.5,
      "learning_rate": 9.714064008779889e-07,
      "loss": 4.0174,
      "num_input_tokens_seen": 49376192,
      "step": 36440
    },
    {
      "epoch": 1.166538633890276,
      "grad_norm": 30.625,
      "learning_rate": 9.686981551528584e-07,
      "loss": 4.5659,
      "num_input_tokens_seen": 49382848,
      "step": 36445
    },
    {
      "epoch": 1.1666986748607644,
      "grad_norm": 13.0625,
      "learning_rate": 9.65993615356467e-07,
      "loss": 3.5461,
      "num_input_tokens_seen": 49389840,
      "step": 36450
    },
    {
      "epoch": 1.166858715831253,
      "grad_norm": 26.625,
      "learning_rate": 9.632927819058917e-07,
      "loss": 4.0522,
      "num_input_tokens_seen": 49396592,
      "step": 36455
    },
    {
      "epoch": 1.1670187568017412,
      "grad_norm": 42.0,
      "learning_rate": 9.605956552176305e-07,
      "loss": 5.7125,
      "num_input_tokens_seen": 49403120,
      "step": 36460
    },
    {
      "epoch": 1.1671787977722297,
      "grad_norm": 37.5,
      "learning_rate": 9.579022357076223e-07,
      "loss": 4.2027,
      "num_input_tokens_seen": 49409872,
      "step": 36465
    },
    {
      "epoch": 1.1673388387427182,
      "grad_norm": 32.75,
      "learning_rate": 9.552125237912158e-07,
      "loss": 3.9796,
      "num_input_tokens_seen": 49416768,
      "step": 36470
    },
    {
      "epoch": 1.1674988797132066,
      "grad_norm": 28.625,
      "learning_rate": 9.525265198832096e-07,
      "loss": 5.3688,
      "num_input_tokens_seen": 49423616,
      "step": 36475
    },
    {
      "epoch": 1.167658920683695,
      "grad_norm": 30.125,
      "learning_rate": 9.498442243978112e-07,
      "loss": 3.4954,
      "num_input_tokens_seen": 49430384,
      "step": 36480
    },
    {
      "epoch": 1.1678189616541834,
      "grad_norm": 18.75,
      "learning_rate": 9.471656377486649e-07,
      "loss": 3.6741,
      "num_input_tokens_seen": 49436752,
      "step": 36485
    },
    {
      "epoch": 1.167979002624672,
      "grad_norm": 22.125,
      "learning_rate": 9.444907603488456e-07,
      "loss": 4.3123,
      "num_input_tokens_seen": 49443392,
      "step": 36490
    },
    {
      "epoch": 1.1681390435951604,
      "grad_norm": 24.0,
      "learning_rate": 9.418195926108514e-07,
      "loss": 4.5137,
      "num_input_tokens_seen": 49449792,
      "step": 36495
    },
    {
      "epoch": 1.168299084565649,
      "grad_norm": 66.0,
      "learning_rate": 9.391521349466053e-07,
      "loss": 4.2778,
      "num_input_tokens_seen": 49456800,
      "step": 36500
    },
    {
      "epoch": 1.1684591255361372,
      "grad_norm": 23.625,
      "learning_rate": 9.364883877674758e-07,
      "loss": 3.6372,
      "num_input_tokens_seen": 49463840,
      "step": 36505
    },
    {
      "epoch": 1.1686191665066257,
      "grad_norm": 19.375,
      "learning_rate": 9.33828351484231e-07,
      "loss": 4.4891,
      "num_input_tokens_seen": 49470944,
      "step": 36510
    },
    {
      "epoch": 1.1687792074771142,
      "grad_norm": 24.5,
      "learning_rate": 9.311720265070906e-07,
      "loss": 5.0068,
      "num_input_tokens_seen": 49477904,
      "step": 36515
    },
    {
      "epoch": 1.1689392484476027,
      "grad_norm": 28.5,
      "learning_rate": 9.285194132456931e-07,
      "loss": 3.972,
      "num_input_tokens_seen": 49484352,
      "step": 36520
    },
    {
      "epoch": 1.169099289418091,
      "grad_norm": 36.25,
      "learning_rate": 9.258705121091032e-07,
      "loss": 3.5469,
      "num_input_tokens_seen": 49490912,
      "step": 36525
    },
    {
      "epoch": 1.1692593303885794,
      "grad_norm": 16.75,
      "learning_rate": 9.232253235058136e-07,
      "loss": 4.8619,
      "num_input_tokens_seen": 49497456,
      "step": 36530
    },
    {
      "epoch": 1.169419371359068,
      "grad_norm": 20.375,
      "learning_rate": 9.205838478437478e-07,
      "loss": 4.3973,
      "num_input_tokens_seen": 49504400,
      "step": 36535
    },
    {
      "epoch": 1.1695794123295564,
      "grad_norm": 23.375,
      "learning_rate": 9.179460855302524e-07,
      "loss": 3.9188,
      "num_input_tokens_seen": 49511104,
      "step": 36540
    },
    {
      "epoch": 1.169739453300045,
      "grad_norm": 45.75,
      "learning_rate": 9.153120369721046e-07,
      "loss": 4.0967,
      "num_input_tokens_seen": 49518016,
      "step": 36545
    },
    {
      "epoch": 1.1698994942705332,
      "grad_norm": 27.0,
      "learning_rate": 9.126817025755103e-07,
      "loss": 4.6573,
      "num_input_tokens_seen": 49524448,
      "step": 36550
    },
    {
      "epoch": 1.1700595352410217,
      "grad_norm": 45.0,
      "learning_rate": 9.100550827460947e-07,
      "loss": 4.4844,
      "num_input_tokens_seen": 49531088,
      "step": 36555
    },
    {
      "epoch": 1.1702195762115102,
      "grad_norm": 56.5,
      "learning_rate": 9.0743217788892e-07,
      "loss": 4.4183,
      "num_input_tokens_seen": 49537920,
      "step": 36560
    },
    {
      "epoch": 1.1703796171819987,
      "grad_norm": 26.75,
      "learning_rate": 9.048129884084683e-07,
      "loss": 3.4812,
      "num_input_tokens_seen": 49544560,
      "step": 36565
    },
    {
      "epoch": 1.170539658152487,
      "grad_norm": 15.5,
      "learning_rate": 9.021975147086553e-07,
      "loss": 3.8045,
      "num_input_tokens_seen": 49551424,
      "step": 36570
    },
    {
      "epoch": 1.1706996991229754,
      "grad_norm": 28.125,
      "learning_rate": 8.995857571928141e-07,
      "loss": 4.5636,
      "num_input_tokens_seen": 49558448,
      "step": 36575
    },
    {
      "epoch": 1.170859740093464,
      "grad_norm": 36.0,
      "learning_rate": 8.969777162637139e-07,
      "loss": 4.7949,
      "num_input_tokens_seen": 49564912,
      "step": 36580
    },
    {
      "epoch": 1.1710197810639524,
      "grad_norm": 26.375,
      "learning_rate": 8.943733923235525e-07,
      "loss": 3.6967,
      "num_input_tokens_seen": 49571984,
      "step": 36585
    },
    {
      "epoch": 1.171179822034441,
      "grad_norm": 32.75,
      "learning_rate": 8.917727857739394e-07,
      "loss": 4.8706,
      "num_input_tokens_seen": 49578688,
      "step": 36590
    },
    {
      "epoch": 1.1713398630049292,
      "grad_norm": 34.0,
      "learning_rate": 8.891758970159258e-07,
      "loss": 3.8826,
      "num_input_tokens_seen": 49585344,
      "step": 36595
    },
    {
      "epoch": 1.1714999039754177,
      "grad_norm": 13.4375,
      "learning_rate": 8.86582726449986e-07,
      "loss": 3.9983,
      "num_input_tokens_seen": 49592352,
      "step": 36600
    },
    {
      "epoch": 1.1714999039754177,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 487.9176,
      "eval_samples_per_second": 28.46,
      "eval_steps_per_second": 14.23,
      "num_input_tokens_seen": 49592352,
      "step": 36600
    },
    {
      "epoch": 1.1716599449459062,
      "grad_norm": 35.0,
      "learning_rate": 8.839932744760165e-07,
      "loss": 4.2794,
      "num_input_tokens_seen": 49599200,
      "step": 36605
    },
    {
      "epoch": 1.1718199859163947,
      "grad_norm": 22.5,
      "learning_rate": 8.814075414933482e-07,
      "loss": 4.8887,
      "num_input_tokens_seen": 49605952,
      "step": 36610
    },
    {
      "epoch": 1.171980026886883,
      "grad_norm": 47.75,
      "learning_rate": 8.788255279007257e-07,
      "loss": 4.0379,
      "num_input_tokens_seen": 49612608,
      "step": 36615
    },
    {
      "epoch": 1.1721400678573715,
      "grad_norm": 21.5,
      "learning_rate": 8.762472340963362e-07,
      "loss": 3.7254,
      "num_input_tokens_seen": 49619152,
      "step": 36620
    },
    {
      "epoch": 1.17230010882786,
      "grad_norm": 44.5,
      "learning_rate": 8.736726604777811e-07,
      "loss": 5.1067,
      "num_input_tokens_seen": 49625568,
      "step": 36625
    },
    {
      "epoch": 1.1724601497983484,
      "grad_norm": 24.625,
      "learning_rate": 8.711018074420901e-07,
      "loss": 5.1765,
      "num_input_tokens_seen": 49632128,
      "step": 36630
    },
    {
      "epoch": 1.172620190768837,
      "grad_norm": 30.875,
      "learning_rate": 8.685346753857209e-07,
      "loss": 3.8873,
      "num_input_tokens_seen": 49639040,
      "step": 36635
    },
    {
      "epoch": 1.1727802317393252,
      "grad_norm": 31.375,
      "learning_rate": 8.659712647045654e-07,
      "loss": 3.4861,
      "num_input_tokens_seen": 49645632,
      "step": 36640
    },
    {
      "epoch": 1.1729402727098137,
      "grad_norm": 49.75,
      "learning_rate": 8.634115757939209e-07,
      "loss": 4.6647,
      "num_input_tokens_seen": 49652448,
      "step": 36645
    },
    {
      "epoch": 1.1731003136803022,
      "grad_norm": 33.5,
      "learning_rate": 8.608556090485387e-07,
      "loss": 5.0001,
      "num_input_tokens_seen": 49659136,
      "step": 36650
    },
    {
      "epoch": 1.1732603546507905,
      "grad_norm": 33.5,
      "learning_rate": 8.583033648625671e-07,
      "loss": 4.384,
      "num_input_tokens_seen": 49666048,
      "step": 36655
    },
    {
      "epoch": 1.173420395621279,
      "grad_norm": 18.875,
      "learning_rate": 8.557548436295998e-07,
      "loss": 3.3087,
      "num_input_tokens_seen": 49672544,
      "step": 36660
    },
    {
      "epoch": 1.1735804365917675,
      "grad_norm": 18.75,
      "learning_rate": 8.532100457426556e-07,
      "loss": 3.8425,
      "num_input_tokens_seen": 49679488,
      "step": 36665
    },
    {
      "epoch": 1.173740477562256,
      "grad_norm": 20.125,
      "learning_rate": 8.506689715941679e-07,
      "loss": 4.6583,
      "num_input_tokens_seen": 49686224,
      "step": 36670
    },
    {
      "epoch": 1.1739005185327445,
      "grad_norm": 27.125,
      "learning_rate": 8.481316215760011e-07,
      "loss": 5.109,
      "num_input_tokens_seen": 49692736,
      "step": 36675
    },
    {
      "epoch": 1.1740605595032327,
      "grad_norm": 27.25,
      "learning_rate": 8.455979960794558e-07,
      "loss": 4.867,
      "num_input_tokens_seen": 49699392,
      "step": 36680
    },
    {
      "epoch": 1.1742206004737212,
      "grad_norm": 30.625,
      "learning_rate": 8.430680954952364e-07,
      "loss": 4.1743,
      "num_input_tokens_seen": 49705920,
      "step": 36685
    },
    {
      "epoch": 1.1743806414442097,
      "grad_norm": 19.5,
      "learning_rate": 8.405419202134974e-07,
      "loss": 4.0713,
      "num_input_tokens_seen": 49712960,
      "step": 36690
    },
    {
      "epoch": 1.1745406824146982,
      "grad_norm": 21.125,
      "learning_rate": 8.380194706237993e-07,
      "loss": 4.4084,
      "num_input_tokens_seen": 49719440,
      "step": 36695
    },
    {
      "epoch": 1.1747007233851865,
      "grad_norm": 46.75,
      "learning_rate": 8.355007471151366e-07,
      "loss": 5.5865,
      "num_input_tokens_seen": 49726032,
      "step": 36700
    },
    {
      "epoch": 1.174860764355675,
      "grad_norm": 26.875,
      "learning_rate": 8.329857500759292e-07,
      "loss": 4.3569,
      "num_input_tokens_seen": 49732432,
      "step": 36705
    },
    {
      "epoch": 1.1750208053261635,
      "grad_norm": 30.625,
      "learning_rate": 8.304744798940194e-07,
      "loss": 4.6073,
      "num_input_tokens_seen": 49739536,
      "step": 36710
    },
    {
      "epoch": 1.175180846296652,
      "grad_norm": 22.25,
      "learning_rate": 8.279669369566756e-07,
      "loss": 3.4748,
      "num_input_tokens_seen": 49746224,
      "step": 36715
    },
    {
      "epoch": 1.1753408872671405,
      "grad_norm": 34.5,
      "learning_rate": 8.254631216505993e-07,
      "loss": 4.9728,
      "num_input_tokens_seen": 49753344,
      "step": 36720
    },
    {
      "epoch": 1.1755009282376288,
      "grad_norm": 21.0,
      "learning_rate": 8.229630343619038e-07,
      "loss": 4.4667,
      "num_input_tokens_seen": 49760096,
      "step": 36725
    },
    {
      "epoch": 1.1756609692081172,
      "grad_norm": 16.875,
      "learning_rate": 8.204666754761392e-07,
      "loss": 5.0887,
      "num_input_tokens_seen": 49767216,
      "step": 36730
    },
    {
      "epoch": 1.1758210101786057,
      "grad_norm": 36.0,
      "learning_rate": 8.179740453782669e-07,
      "loss": 4.1349,
      "num_input_tokens_seen": 49774112,
      "step": 36735
    },
    {
      "epoch": 1.1759810511490942,
      "grad_norm": 55.5,
      "learning_rate": 8.154851444526907e-07,
      "loss": 4.3211,
      "num_input_tokens_seen": 49780928,
      "step": 36740
    },
    {
      "epoch": 1.1761410921195825,
      "grad_norm": 20.25,
      "learning_rate": 8.129999730832283e-07,
      "loss": 3.7147,
      "num_input_tokens_seen": 49788128,
      "step": 36745
    },
    {
      "epoch": 1.176301133090071,
      "grad_norm": 25.375,
      "learning_rate": 8.105185316531178e-07,
      "loss": 4.9528,
      "num_input_tokens_seen": 49794864,
      "step": 36750
    },
    {
      "epoch": 1.1764611740605595,
      "grad_norm": 25.625,
      "learning_rate": 8.08040820545039e-07,
      "loss": 4.4495,
      "num_input_tokens_seen": 49801648,
      "step": 36755
    },
    {
      "epoch": 1.176621215031048,
      "grad_norm": 43.0,
      "learning_rate": 8.055668401410782e-07,
      "loss": 4.6951,
      "num_input_tokens_seen": 49808288,
      "step": 36760
    },
    {
      "epoch": 1.1767812560015365,
      "grad_norm": 28.625,
      "learning_rate": 8.030965908227578e-07,
      "loss": 4.742,
      "num_input_tokens_seen": 49815296,
      "step": 36765
    },
    {
      "epoch": 1.1769412969720248,
      "grad_norm": 19.75,
      "learning_rate": 8.006300729710203e-07,
      "loss": 3.3187,
      "num_input_tokens_seen": 49822544,
      "step": 36770
    },
    {
      "epoch": 1.1771013379425133,
      "grad_norm": 32.25,
      "learning_rate": 7.981672869662337e-07,
      "loss": 4.359,
      "num_input_tokens_seen": 49829504,
      "step": 36775
    },
    {
      "epoch": 1.1772613789130018,
      "grad_norm": 24.375,
      "learning_rate": 7.957082331881888e-07,
      "loss": 4.2699,
      "num_input_tokens_seen": 49836288,
      "step": 36780
    },
    {
      "epoch": 1.1774214198834903,
      "grad_norm": 29.125,
      "learning_rate": 7.932529120161069e-07,
      "loss": 4.3485,
      "num_input_tokens_seen": 49842800,
      "step": 36785
    },
    {
      "epoch": 1.1775814608539785,
      "grad_norm": 27.5,
      "learning_rate": 7.908013238286243e-07,
      "loss": 4.6002,
      "num_input_tokens_seen": 49849312,
      "step": 36790
    },
    {
      "epoch": 1.177741501824467,
      "grad_norm": 34.75,
      "learning_rate": 7.883534690038136e-07,
      "loss": 4.2554,
      "num_input_tokens_seen": 49856240,
      "step": 36795
    },
    {
      "epoch": 1.1779015427949555,
      "grad_norm": 30.75,
      "learning_rate": 7.859093479191559e-07,
      "loss": 3.6453,
      "num_input_tokens_seen": 49863184,
      "step": 36800
    },
    {
      "epoch": 1.1779015427949555,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 487.9201,
      "eval_samples_per_second": 28.46,
      "eval_steps_per_second": 14.23,
      "num_input_tokens_seen": 49863184,
      "step": 36800
    },
    {
      "epoch": 1.178061583765444,
      "grad_norm": 30.375,
      "learning_rate": 7.834689609515722e-07,
      "loss": 4.2611,
      "num_input_tokens_seen": 49869568,
      "step": 36805
    },
    {
      "epoch": 1.1782216247359325,
      "grad_norm": 35.5,
      "learning_rate": 7.810323084774002e-07,
      "loss": 4.4236,
      "num_input_tokens_seen": 49876288,
      "step": 36810
    },
    {
      "epoch": 1.1783816657064208,
      "grad_norm": 28.75,
      "learning_rate": 7.785993908723976e-07,
      "loss": 3.397,
      "num_input_tokens_seen": 49883168,
      "step": 36815
    },
    {
      "epoch": 1.1785417066769093,
      "grad_norm": 22.0,
      "learning_rate": 7.761702085117534e-07,
      "loss": 4.1169,
      "num_input_tokens_seen": 49890416,
      "step": 36820
    },
    {
      "epoch": 1.1787017476473978,
      "grad_norm": 37.5,
      "learning_rate": 7.737447617700844e-07,
      "loss": 4.4413,
      "num_input_tokens_seen": 49896544,
      "step": 36825
    },
    {
      "epoch": 1.1788617886178863,
      "grad_norm": 14.625,
      "learning_rate": 7.713230510214136e-07,
      "loss": 3.7655,
      "num_input_tokens_seen": 49903232,
      "step": 36830
    },
    {
      "epoch": 1.1790218295883745,
      "grad_norm": 21.125,
      "learning_rate": 7.689050766392092e-07,
      "loss": 4.3113,
      "num_input_tokens_seen": 49909744,
      "step": 36835
    },
    {
      "epoch": 1.179181870558863,
      "grad_norm": 21.875,
      "learning_rate": 7.664908389963477e-07,
      "loss": 4.2219,
      "num_input_tokens_seen": 49916288,
      "step": 36840
    },
    {
      "epoch": 1.1793419115293515,
      "grad_norm": 30.75,
      "learning_rate": 7.64080338465134e-07,
      "loss": 4.1115,
      "num_input_tokens_seen": 49923584,
      "step": 36845
    },
    {
      "epoch": 1.17950195249984,
      "grad_norm": 46.25,
      "learning_rate": 7.616735754173043e-07,
      "loss": 4.0413,
      "num_input_tokens_seen": 49930800,
      "step": 36850
    },
    {
      "epoch": 1.1796619934703285,
      "grad_norm": 26.5,
      "learning_rate": 7.592705502240005e-07,
      "loss": 3.3313,
      "num_input_tokens_seen": 49937744,
      "step": 36855
    },
    {
      "epoch": 1.1798220344408168,
      "grad_norm": 27.625,
      "learning_rate": 7.568712632558095e-07,
      "loss": 4.1718,
      "num_input_tokens_seen": 49944752,
      "step": 36860
    },
    {
      "epoch": 1.1799820754113053,
      "grad_norm": 27.375,
      "learning_rate": 7.544757148827297e-07,
      "loss": 4.5544,
      "num_input_tokens_seen": 49951680,
      "step": 36865
    },
    {
      "epoch": 1.1801421163817938,
      "grad_norm": 41.5,
      "learning_rate": 7.520839054741797e-07,
      "loss": 4.6274,
      "num_input_tokens_seen": 49958608,
      "step": 36870
    },
    {
      "epoch": 1.1803021573522823,
      "grad_norm": 20.875,
      "learning_rate": 7.496958353990113e-07,
      "loss": 3.7456,
      "num_input_tokens_seen": 49965184,
      "step": 36875
    },
    {
      "epoch": 1.1804621983227706,
      "grad_norm": 19.375,
      "learning_rate": 7.473115050254941e-07,
      "loss": 4.3628,
      "num_input_tokens_seen": 49971936,
      "step": 36880
    },
    {
      "epoch": 1.180622239293259,
      "grad_norm": 19.5,
      "learning_rate": 7.449309147213173e-07,
      "loss": 4.5243,
      "num_input_tokens_seen": 49978880,
      "step": 36885
    },
    {
      "epoch": 1.1807822802637475,
      "grad_norm": 43.75,
      "learning_rate": 7.425540648536067e-07,
      "loss": 5.0475,
      "num_input_tokens_seen": 49985504,
      "step": 36890
    },
    {
      "epoch": 1.180942321234236,
      "grad_norm": 33.25,
      "learning_rate": 7.40180955788894e-07,
      "loss": 4.3881,
      "num_input_tokens_seen": 49992112,
      "step": 36895
    },
    {
      "epoch": 1.1811023622047245,
      "grad_norm": 30.5,
      "learning_rate": 7.378115878931474e-07,
      "loss": 3.9943,
      "num_input_tokens_seen": 49998768,
      "step": 36900
    },
    {
      "epoch": 1.1812624031752128,
      "grad_norm": 44.25,
      "learning_rate": 7.354459615317527e-07,
      "loss": 3.4572,
      "num_input_tokens_seen": 50005072,
      "step": 36905
    },
    {
      "epoch": 1.1814224441457013,
      "grad_norm": 32.0,
      "learning_rate": 7.33084077069518e-07,
      "loss": 4.1617,
      "num_input_tokens_seen": 50011984,
      "step": 36910
    },
    {
      "epoch": 1.1815824851161898,
      "grad_norm": 22.25,
      "learning_rate": 7.307259348706768e-07,
      "loss": 4.7329,
      "num_input_tokens_seen": 50019232,
      "step": 36915
    },
    {
      "epoch": 1.181742526086678,
      "grad_norm": 41.5,
      "learning_rate": 7.283715352988801e-07,
      "loss": 4.3376,
      "num_input_tokens_seen": 50026736,
      "step": 36920
    },
    {
      "epoch": 1.1819025670571666,
      "grad_norm": 28.875,
      "learning_rate": 7.260208787172068e-07,
      "loss": 3.9421,
      "num_input_tokens_seen": 50033552,
      "step": 36925
    },
    {
      "epoch": 1.182062608027655,
      "grad_norm": 23.5,
      "learning_rate": 7.23673965488167e-07,
      "loss": 4.7566,
      "num_input_tokens_seen": 50040448,
      "step": 36930
    },
    {
      "epoch": 1.1822226489981436,
      "grad_norm": 34.25,
      "learning_rate": 7.213307959736709e-07,
      "loss": 4.9704,
      "num_input_tokens_seen": 50046768,
      "step": 36935
    },
    {
      "epoch": 1.182382689968632,
      "grad_norm": 23.625,
      "learning_rate": 7.189913705350715e-07,
      "loss": 3.6945,
      "num_input_tokens_seen": 50053568,
      "step": 36940
    },
    {
      "epoch": 1.1825427309391203,
      "grad_norm": 31.25,
      "learning_rate": 7.166556895331411e-07,
      "loss": 4.3077,
      "num_input_tokens_seen": 50060448,
      "step": 36945
    },
    {
      "epoch": 1.1827027719096088,
      "grad_norm": 38.0,
      "learning_rate": 7.143237533280639e-07,
      "loss": 3.7765,
      "num_input_tokens_seen": 50067184,
      "step": 36950
    },
    {
      "epoch": 1.1828628128800973,
      "grad_norm": 18.5,
      "learning_rate": 7.119955622794578e-07,
      "loss": 3.6287,
      "num_input_tokens_seen": 50073648,
      "step": 36955
    },
    {
      "epoch": 1.1830228538505858,
      "grad_norm": 16.875,
      "learning_rate": 7.096711167463577e-07,
      "loss": 4.0135,
      "num_input_tokens_seen": 50080480,
      "step": 36960
    },
    {
      "epoch": 1.183182894821074,
      "grad_norm": 29.625,
      "learning_rate": 7.073504170872213e-07,
      "loss": 3.9038,
      "num_input_tokens_seen": 50087392,
      "step": 36965
    },
    {
      "epoch": 1.1833429357915626,
      "grad_norm": 20.125,
      "learning_rate": 7.05033463659932e-07,
      "loss": 4.1708,
      "num_input_tokens_seen": 50094048,
      "step": 36970
    },
    {
      "epoch": 1.183502976762051,
      "grad_norm": 37.5,
      "learning_rate": 7.027202568217928e-07,
      "loss": 3.619,
      "num_input_tokens_seen": 50100800,
      "step": 36975
    },
    {
      "epoch": 1.1836630177325396,
      "grad_norm": 18.875,
      "learning_rate": 7.004107969295293e-07,
      "loss": 3.3879,
      "num_input_tokens_seen": 50107648,
      "step": 36980
    },
    {
      "epoch": 1.183823058703028,
      "grad_norm": 27.875,
      "learning_rate": 6.9810508433929e-07,
      "loss": 4.1978,
      "num_input_tokens_seen": 50114800,
      "step": 36985
    },
    {
      "epoch": 1.1839830996735163,
      "grad_norm": 27.75,
      "learning_rate": 6.958031194066406e-07,
      "loss": 4.3703,
      "num_input_tokens_seen": 50121776,
      "step": 36990
    },
    {
      "epoch": 1.1841431406440048,
      "grad_norm": 59.75,
      "learning_rate": 6.935049024865776e-07,
      "loss": 4.259,
      "num_input_tokens_seen": 50128400,
      "step": 36995
    },
    {
      "epoch": 1.1843031816144933,
      "grad_norm": 27.125,
      "learning_rate": 6.912104339335118e-07,
      "loss": 4.6426,
      "num_input_tokens_seen": 50135184,
      "step": 37000
    },
    {
      "epoch": 1.1843031816144933,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 487.1901,
      "eval_samples_per_second": 28.502,
      "eval_steps_per_second": 14.251,
      "num_input_tokens_seen": 50135184,
      "step": 37000
    },
    {
      "epoch": 1.1844632225849818,
      "grad_norm": 43.25,
      "learning_rate": 6.889197141012799e-07,
      "loss": 3.6079,
      "num_input_tokens_seen": 50142496,
      "step": 37005
    },
    {
      "epoch": 1.18462326355547,
      "grad_norm": 18.625,
      "learning_rate": 6.866327433431435e-07,
      "loss": 4.7466,
      "num_input_tokens_seen": 50149744,
      "step": 37010
    },
    {
      "epoch": 1.1847833045259586,
      "grad_norm": 20.25,
      "learning_rate": 6.843495220117735e-07,
      "loss": 4.4091,
      "num_input_tokens_seen": 50156592,
      "step": 37015
    },
    {
      "epoch": 1.184943345496447,
      "grad_norm": 26.5,
      "learning_rate": 6.820700504592798e-07,
      "loss": 4.5189,
      "num_input_tokens_seen": 50163232,
      "step": 37020
    },
    {
      "epoch": 1.1851033864669356,
      "grad_norm": 44.0,
      "learning_rate": 6.797943290371839e-07,
      "loss": 4.0568,
      "num_input_tokens_seen": 50169872,
      "step": 37025
    },
    {
      "epoch": 1.185263427437424,
      "grad_norm": 43.25,
      "learning_rate": 6.775223580964274e-07,
      "loss": 5.1505,
      "num_input_tokens_seen": 50176880,
      "step": 37030
    },
    {
      "epoch": 1.1854234684079124,
      "grad_norm": 20.875,
      "learning_rate": 6.7525413798738e-07,
      "loss": 4.2342,
      "num_input_tokens_seen": 50183840,
      "step": 37035
    },
    {
      "epoch": 1.1855835093784008,
      "grad_norm": 17.0,
      "learning_rate": 6.729896690598259e-07,
      "loss": 4.6971,
      "num_input_tokens_seen": 50190528,
      "step": 37040
    },
    {
      "epoch": 1.1857435503488893,
      "grad_norm": 23.25,
      "learning_rate": 6.707289516629772e-07,
      "loss": 4.7063,
      "num_input_tokens_seen": 50196992,
      "step": 37045
    },
    {
      "epoch": 1.1859035913193778,
      "grad_norm": 27.875,
      "learning_rate": 6.684719861454692e-07,
      "loss": 4.5781,
      "num_input_tokens_seen": 50203520,
      "step": 37050
    },
    {
      "epoch": 1.186063632289866,
      "grad_norm": 35.0,
      "learning_rate": 6.662187728553481e-07,
      "loss": 4.7868,
      "num_input_tokens_seen": 50210240,
      "step": 37055
    },
    {
      "epoch": 1.1862236732603546,
      "grad_norm": 24.875,
      "learning_rate": 6.639693121400892e-07,
      "loss": 4.4009,
      "num_input_tokens_seen": 50216816,
      "step": 37060
    },
    {
      "epoch": 1.186383714230843,
      "grad_norm": 24.625,
      "learning_rate": 6.617236043465868e-07,
      "loss": 3.335,
      "num_input_tokens_seen": 50223648,
      "step": 37065
    },
    {
      "epoch": 1.1865437552013316,
      "grad_norm": 37.25,
      "learning_rate": 6.594816498211587e-07,
      "loss": 5.1151,
      "num_input_tokens_seen": 50230160,
      "step": 37070
    },
    {
      "epoch": 1.18670379617182,
      "grad_norm": 35.75,
      "learning_rate": 6.572434489095447e-07,
      "loss": 4.5728,
      "num_input_tokens_seen": 50236832,
      "step": 37075
    },
    {
      "epoch": 1.1868638371423084,
      "grad_norm": 32.25,
      "learning_rate": 6.550090019568994e-07,
      "loss": 4.1322,
      "num_input_tokens_seen": 50243616,
      "step": 37080
    },
    {
      "epoch": 1.1870238781127969,
      "grad_norm": 26.0,
      "learning_rate": 6.527783093078027e-07,
      "loss": 3.8861,
      "num_input_tokens_seen": 50251216,
      "step": 37085
    },
    {
      "epoch": 1.1871839190832854,
      "grad_norm": 24.875,
      "learning_rate": 6.5055137130626e-07,
      "loss": 3.9392,
      "num_input_tokens_seen": 50257952,
      "step": 37090
    },
    {
      "epoch": 1.1873439600537739,
      "grad_norm": 21.0,
      "learning_rate": 6.483281882956854e-07,
      "loss": 4.9571,
      "num_input_tokens_seen": 50264544,
      "step": 37095
    },
    {
      "epoch": 1.1875040010242621,
      "grad_norm": 22.25,
      "learning_rate": 6.461087606189298e-07,
      "loss": 4.0422,
      "num_input_tokens_seen": 50271696,
      "step": 37100
    },
    {
      "epoch": 1.1876640419947506,
      "grad_norm": 14.375,
      "learning_rate": 6.438930886182554e-07,
      "loss": 4.3807,
      "num_input_tokens_seen": 50278544,
      "step": 37105
    },
    {
      "epoch": 1.1878240829652391,
      "grad_norm": 39.75,
      "learning_rate": 6.416811726353417e-07,
      "loss": 3.5836,
      "num_input_tokens_seen": 50285056,
      "step": 37110
    },
    {
      "epoch": 1.1879841239357276,
      "grad_norm": 17.75,
      "learning_rate": 6.394730130112991e-07,
      "loss": 4.9689,
      "num_input_tokens_seen": 50291840,
      "step": 37115
    },
    {
      "epoch": 1.188144164906216,
      "grad_norm": 48.75,
      "learning_rate": 6.372686100866471e-07,
      "loss": 3.5992,
      "num_input_tokens_seen": 50298384,
      "step": 37120
    },
    {
      "epoch": 1.1883042058767044,
      "grad_norm": 55.5,
      "learning_rate": 6.350679642013413e-07,
      "loss": 4.9736,
      "num_input_tokens_seen": 50305008,
      "step": 37125
    },
    {
      "epoch": 1.1884642468471929,
      "grad_norm": 27.75,
      "learning_rate": 6.328710756947437e-07,
      "loss": 4.48,
      "num_input_tokens_seen": 50311680,
      "step": 37130
    },
    {
      "epoch": 1.1886242878176814,
      "grad_norm": 22.625,
      "learning_rate": 6.306779449056416e-07,
      "loss": 4.6363,
      "num_input_tokens_seen": 50319072,
      "step": 37135
    },
    {
      "epoch": 1.1887843287881699,
      "grad_norm": 30.125,
      "learning_rate": 6.284885721722422e-07,
      "loss": 3.5002,
      "num_input_tokens_seen": 50325968,
      "step": 37140
    },
    {
      "epoch": 1.1889443697586581,
      "grad_norm": 21.125,
      "learning_rate": 6.26302957832181e-07,
      "loss": 4.3105,
      "num_input_tokens_seen": 50332576,
      "step": 37145
    },
    {
      "epoch": 1.1891044107291466,
      "grad_norm": 19.125,
      "learning_rate": 6.241211022224997e-07,
      "loss": 4.7545,
      "num_input_tokens_seen": 50339024,
      "step": 37150
    },
    {
      "epoch": 1.1892644516996351,
      "grad_norm": 21.125,
      "learning_rate": 6.219430056796732e-07,
      "loss": 4.9334,
      "num_input_tokens_seen": 50345472,
      "step": 37155
    },
    {
      "epoch": 1.1894244926701236,
      "grad_norm": 29.75,
      "learning_rate": 6.19768668539586e-07,
      "loss": 3.668,
      "num_input_tokens_seen": 50352320,
      "step": 37160
    },
    {
      "epoch": 1.1895845336406121,
      "grad_norm": 21.25,
      "learning_rate": 6.175980911375528e-07,
      "loss": 3.6185,
      "num_input_tokens_seen": 50359152,
      "step": 37165
    },
    {
      "epoch": 1.1897445746111004,
      "grad_norm": 28.75,
      "learning_rate": 6.154312738083034e-07,
      "loss": 4.4449,
      "num_input_tokens_seen": 50365856,
      "step": 37170
    },
    {
      "epoch": 1.1899046155815889,
      "grad_norm": 37.0,
      "learning_rate": 6.132682168859843e-07,
      "loss": 4.3873,
      "num_input_tokens_seen": 50372768,
      "step": 37175
    },
    {
      "epoch": 1.1900646565520774,
      "grad_norm": 38.0,
      "learning_rate": 6.111089207041704e-07,
      "loss": 3.7211,
      "num_input_tokens_seen": 50379712,
      "step": 37180
    },
    {
      "epoch": 1.1902246975225657,
      "grad_norm": 31.0,
      "learning_rate": 6.089533855958507e-07,
      "loss": 5.0946,
      "num_input_tokens_seen": 50386768,
      "step": 37185
    },
    {
      "epoch": 1.1903847384930542,
      "grad_norm": 40.25,
      "learning_rate": 6.068016118934372e-07,
      "loss": 4.3698,
      "num_input_tokens_seen": 50393680,
      "step": 37190
    },
    {
      "epoch": 1.1905447794635426,
      "grad_norm": 38.0,
      "learning_rate": 6.04653599928759e-07,
      "loss": 4.0989,
      "num_input_tokens_seen": 50400672,
      "step": 37195
    },
    {
      "epoch": 1.1907048204340311,
      "grad_norm": 32.0,
      "learning_rate": 6.025093500330675e-07,
      "loss": 4.5039,
      "num_input_tokens_seen": 50407568,
      "step": 37200
    },
    {
      "epoch": 1.1907048204340311,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 486.069,
      "eval_samples_per_second": 28.568,
      "eval_steps_per_second": 14.284,
      "num_input_tokens_seen": 50407568,
      "step": 37200
    },
    {
      "epoch": 1.1908648614045196,
      "grad_norm": 22.0,
      "learning_rate": 6.003688625370291e-07,
      "loss": 3.7086,
      "num_input_tokens_seen": 50414576,
      "step": 37205
    },
    {
      "epoch": 1.191024902375008,
      "grad_norm": 19.25,
      "learning_rate": 5.982321377707406e-07,
      "loss": 3.6205,
      "num_input_tokens_seen": 50420864,
      "step": 37210
    },
    {
      "epoch": 1.1911849433454964,
      "grad_norm": 26.375,
      "learning_rate": 5.96099176063708e-07,
      "loss": 3.7508,
      "num_input_tokens_seen": 50427632,
      "step": 37215
    },
    {
      "epoch": 1.191344984315985,
      "grad_norm": 23.25,
      "learning_rate": 5.93969977744857e-07,
      "loss": 4.9343,
      "num_input_tokens_seen": 50434208,
      "step": 37220
    },
    {
      "epoch": 1.1915050252864734,
      "grad_norm": 41.75,
      "learning_rate": 5.918445431425445e-07,
      "loss": 4.7839,
      "num_input_tokens_seen": 50441264,
      "step": 37225
    },
    {
      "epoch": 1.1916650662569617,
      "grad_norm": 54.5,
      "learning_rate": 5.897228725845333e-07,
      "loss": 5.0391,
      "num_input_tokens_seen": 50447792,
      "step": 37230
    },
    {
      "epoch": 1.1918251072274502,
      "grad_norm": 28.875,
      "learning_rate": 5.876049663980171e-07,
      "loss": 3.532,
      "num_input_tokens_seen": 50454560,
      "step": 37235
    },
    {
      "epoch": 1.1919851481979387,
      "grad_norm": 39.0,
      "learning_rate": 5.854908249095959e-07,
      "loss": 4.1193,
      "num_input_tokens_seen": 50461088,
      "step": 37240
    },
    {
      "epoch": 1.1921451891684272,
      "grad_norm": 14.875,
      "learning_rate": 5.833804484453031e-07,
      "loss": 4.2712,
      "num_input_tokens_seen": 50468064,
      "step": 37245
    },
    {
      "epoch": 1.1923052301389157,
      "grad_norm": 39.0,
      "learning_rate": 5.81273837330587e-07,
      "loss": 4.5628,
      "num_input_tokens_seen": 50474864,
      "step": 37250
    },
    {
      "epoch": 1.192465271109404,
      "grad_norm": 34.5,
      "learning_rate": 5.791709918903071e-07,
      "loss": 3.8754,
      "num_input_tokens_seen": 50481808,
      "step": 37255
    },
    {
      "epoch": 1.1926253120798924,
      "grad_norm": 46.75,
      "learning_rate": 5.770719124487483e-07,
      "loss": 5.1967,
      "num_input_tokens_seen": 50488400,
      "step": 37260
    },
    {
      "epoch": 1.192785353050381,
      "grad_norm": 34.75,
      "learning_rate": 5.749765993296241e-07,
      "loss": 4.1561,
      "num_input_tokens_seen": 50494896,
      "step": 37265
    },
    {
      "epoch": 1.1929453940208694,
      "grad_norm": 30.875,
      "learning_rate": 5.728850528560509e-07,
      "loss": 4.3963,
      "num_input_tokens_seen": 50501264,
      "step": 37270
    },
    {
      "epoch": 1.1931054349913577,
      "grad_norm": 32.75,
      "learning_rate": 5.707972733505707e-07,
      "loss": 3.2171,
      "num_input_tokens_seen": 50507792,
      "step": 37275
    },
    {
      "epoch": 1.1932654759618462,
      "grad_norm": 14.625,
      "learning_rate": 5.687132611351509e-07,
      "loss": 3.7103,
      "num_input_tokens_seen": 50514880,
      "step": 37280
    },
    {
      "epoch": 1.1934255169323347,
      "grad_norm": 39.0,
      "learning_rate": 5.666330165311651e-07,
      "loss": 5.5812,
      "num_input_tokens_seen": 50521472,
      "step": 37285
    },
    {
      "epoch": 1.1935855579028232,
      "grad_norm": 30.0,
      "learning_rate": 5.645565398594204e-07,
      "loss": 3.7773,
      "num_input_tokens_seen": 50528384,
      "step": 37290
    },
    {
      "epoch": 1.1937455988733117,
      "grad_norm": 42.25,
      "learning_rate": 5.624838314401304e-07,
      "loss": 4.4751,
      "num_input_tokens_seen": 50535184,
      "step": 37295
    },
    {
      "epoch": 1.1939056398438,
      "grad_norm": 29.375,
      "learning_rate": 5.604148915929336e-07,
      "loss": 4.3987,
      "num_input_tokens_seen": 50542448,
      "step": 37300
    },
    {
      "epoch": 1.1940656808142884,
      "grad_norm": 21.75,
      "learning_rate": 5.583497206368887e-07,
      "loss": 3.7749,
      "num_input_tokens_seen": 50549024,
      "step": 37305
    },
    {
      "epoch": 1.194225721784777,
      "grad_norm": 40.0,
      "learning_rate": 5.562883188904688e-07,
      "loss": 4.2088,
      "num_input_tokens_seen": 50555744,
      "step": 37310
    },
    {
      "epoch": 1.1943857627552654,
      "grad_norm": 27.875,
      "learning_rate": 5.542306866715724e-07,
      "loss": 4.1289,
      "num_input_tokens_seen": 50562512,
      "step": 37315
    },
    {
      "epoch": 1.1945458037257537,
      "grad_norm": 18.375,
      "learning_rate": 5.52176824297504e-07,
      "loss": 4.0964,
      "num_input_tokens_seen": 50569648,
      "step": 37320
    },
    {
      "epoch": 1.1947058446962422,
      "grad_norm": 48.5,
      "learning_rate": 5.501267320850018e-07,
      "loss": 5.4765,
      "num_input_tokens_seen": 50576768,
      "step": 37325
    },
    {
      "epoch": 1.1948658856667307,
      "grad_norm": 24.375,
      "learning_rate": 5.480804103502157e-07,
      "loss": 3.645,
      "num_input_tokens_seen": 50583344,
      "step": 37330
    },
    {
      "epoch": 1.1950259266372192,
      "grad_norm": 13.8125,
      "learning_rate": 5.460378594087101e-07,
      "loss": 3.2934,
      "num_input_tokens_seen": 50590208,
      "step": 37335
    },
    {
      "epoch": 1.1951859676077077,
      "grad_norm": 25.125,
      "learning_rate": 5.439990795754773e-07,
      "loss": 4.3736,
      "num_input_tokens_seen": 50597024,
      "step": 37340
    },
    {
      "epoch": 1.195346008578196,
      "grad_norm": 23.75,
      "learning_rate": 5.419640711649188e-07,
      "loss": 5.1506,
      "num_input_tokens_seen": 50604064,
      "step": 37345
    },
    {
      "epoch": 1.1955060495486844,
      "grad_norm": 48.5,
      "learning_rate": 5.399328344908583e-07,
      "loss": 4.4019,
      "num_input_tokens_seen": 50611104,
      "step": 37350
    },
    {
      "epoch": 1.195666090519173,
      "grad_norm": 22.75,
      "learning_rate": 5.379053698665399e-07,
      "loss": 4.0385,
      "num_input_tokens_seen": 50617664,
      "step": 37355
    },
    {
      "epoch": 1.1958261314896614,
      "grad_norm": 27.125,
      "learning_rate": 5.358816776046216e-07,
      "loss": 4.5006,
      "num_input_tokens_seen": 50624240,
      "step": 37360
    },
    {
      "epoch": 1.1959861724601497,
      "grad_norm": 34.5,
      "learning_rate": 5.338617580171817e-07,
      "loss": 4.1548,
      "num_input_tokens_seen": 50631360,
      "step": 37365
    },
    {
      "epoch": 1.1961462134306382,
      "grad_norm": 20.0,
      "learning_rate": 5.318456114157239e-07,
      "loss": 3.7054,
      "num_input_tokens_seen": 50638384,
      "step": 37370
    },
    {
      "epoch": 1.1963062544011267,
      "grad_norm": 24.125,
      "learning_rate": 5.298332381111576e-07,
      "loss": 4.1464,
      "num_input_tokens_seen": 50645040,
      "step": 37375
    },
    {
      "epoch": 1.1964662953716152,
      "grad_norm": 22.5,
      "learning_rate": 5.27824638413818e-07,
      "loss": 4.0502,
      "num_input_tokens_seen": 50651904,
      "step": 37380
    },
    {
      "epoch": 1.1966263363421037,
      "grad_norm": 29.5,
      "learning_rate": 5.258198126334546e-07,
      "loss": 2.9258,
      "num_input_tokens_seen": 50658576,
      "step": 37385
    },
    {
      "epoch": 1.196786377312592,
      "grad_norm": 27.375,
      "learning_rate": 5.238187610792367e-07,
      "loss": 4.2026,
      "num_input_tokens_seen": 50665200,
      "step": 37390
    },
    {
      "epoch": 1.1969464182830805,
      "grad_norm": 37.5,
      "learning_rate": 5.218214840597563e-07,
      "loss": 4.1635,
      "num_input_tokens_seen": 50671872,
      "step": 37395
    },
    {
      "epoch": 1.197106459253569,
      "grad_norm": 50.0,
      "learning_rate": 5.198279818830115e-07,
      "loss": 6.2586,
      "num_input_tokens_seen": 50678192,
      "step": 37400
    },
    {
      "epoch": 1.197106459253569,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.8257,
      "eval_samples_per_second": 28.582,
      "eval_steps_per_second": 14.291,
      "num_input_tokens_seen": 50678192,
      "step": 37400
    },
    {
      "epoch": 1.1972665002240575,
      "grad_norm": 37.75,
      "learning_rate": 5.178382548564287e-07,
      "loss": 4.0653,
      "num_input_tokens_seen": 50684880,
      "step": 37405
    },
    {
      "epoch": 1.1974265411945457,
      "grad_norm": 28.375,
      "learning_rate": 5.15852303286854e-07,
      "loss": 4.1508,
      "num_input_tokens_seen": 50691520,
      "step": 37410
    },
    {
      "epoch": 1.1975865821650342,
      "grad_norm": 22.75,
      "learning_rate": 5.138701274805396e-07,
      "loss": 4.1245,
      "num_input_tokens_seen": 50698272,
      "step": 37415
    },
    {
      "epoch": 1.1977466231355227,
      "grad_norm": 39.5,
      "learning_rate": 5.118917277431606e-07,
      "loss": 4.2009,
      "num_input_tokens_seen": 50704976,
      "step": 37420
    },
    {
      "epoch": 1.1979066641060112,
      "grad_norm": 37.0,
      "learning_rate": 5.099171043798145e-07,
      "loss": 4.6336,
      "num_input_tokens_seen": 50711536,
      "step": 37425
    },
    {
      "epoch": 1.1980667050764997,
      "grad_norm": 27.375,
      "learning_rate": 5.079462576950133e-07,
      "loss": 5.0437,
      "num_input_tokens_seen": 50718928,
      "step": 37430
    },
    {
      "epoch": 1.198226746046988,
      "grad_norm": 34.5,
      "learning_rate": 5.059791879926862e-07,
      "loss": 4.069,
      "num_input_tokens_seen": 50725264,
      "step": 37435
    },
    {
      "epoch": 1.1983867870174765,
      "grad_norm": 20.5,
      "learning_rate": 5.040158955761793e-07,
      "loss": 3.9985,
      "num_input_tokens_seen": 50731664,
      "step": 37440
    },
    {
      "epoch": 1.198546827987965,
      "grad_norm": 24.25,
      "learning_rate": 5.020563807482559e-07,
      "loss": 3.7195,
      "num_input_tokens_seen": 50738704,
      "step": 37445
    },
    {
      "epoch": 1.1987068689584532,
      "grad_norm": 72.0,
      "learning_rate": 5.001006438110995e-07,
      "loss": 3.973,
      "num_input_tokens_seen": 50745136,
      "step": 37450
    },
    {
      "epoch": 1.1988669099289417,
      "grad_norm": 32.5,
      "learning_rate": 4.981486850663075e-07,
      "loss": 3.7893,
      "num_input_tokens_seen": 50752016,
      "step": 37455
    },
    {
      "epoch": 1.1990269508994302,
      "grad_norm": 20.875,
      "learning_rate": 4.962005048149005e-07,
      "loss": 4.5722,
      "num_input_tokens_seen": 50758800,
      "step": 37460
    },
    {
      "epoch": 1.1991869918699187,
      "grad_norm": 30.5,
      "learning_rate": 4.942561033573073e-07,
      "loss": 3.7741,
      "num_input_tokens_seen": 50765552,
      "step": 37465
    },
    {
      "epoch": 1.1993470328404072,
      "grad_norm": 28.75,
      "learning_rate": 4.923154809933827e-07,
      "loss": 3.752,
      "num_input_tokens_seen": 50772464,
      "step": 37470
    },
    {
      "epoch": 1.1995070738108955,
      "grad_norm": 28.5,
      "learning_rate": 4.903786380223957e-07,
      "loss": 3.9406,
      "num_input_tokens_seen": 50779392,
      "step": 37475
    },
    {
      "epoch": 1.199667114781384,
      "grad_norm": 33.5,
      "learning_rate": 4.884455747430266e-07,
      "loss": 5.9822,
      "num_input_tokens_seen": 50786448,
      "step": 37480
    },
    {
      "epoch": 1.1998271557518725,
      "grad_norm": 18.5,
      "learning_rate": 4.865162914533816e-07,
      "loss": 3.6091,
      "num_input_tokens_seen": 50793008,
      "step": 37485
    },
    {
      "epoch": 1.199987196722361,
      "grad_norm": 27.25,
      "learning_rate": 4.845907884509809e-07,
      "loss": 4.4473,
      "num_input_tokens_seen": 50799984,
      "step": 37490
    },
    {
      "epoch": 1.2001472376928493,
      "grad_norm": 17.375,
      "learning_rate": 4.82669066032762e-07,
      "loss": 4.2413,
      "num_input_tokens_seen": 50806400,
      "step": 37495
    },
    {
      "epoch": 1.2003072786633378,
      "grad_norm": 40.25,
      "learning_rate": 4.807511244950768e-07,
      "loss": 4.8354,
      "num_input_tokens_seen": 50813152,
      "step": 37500
    },
    {
      "epoch": 1.2004673196338262,
      "grad_norm": 17.125,
      "learning_rate": 4.788369641336943e-07,
      "loss": 3.3747,
      "num_input_tokens_seen": 50820288,
      "step": 37505
    },
    {
      "epoch": 1.2006273606043147,
      "grad_norm": 13.4375,
      "learning_rate": 4.769265852438032e-07,
      "loss": 4.2198,
      "num_input_tokens_seen": 50826928,
      "step": 37510
    },
    {
      "epoch": 1.2007874015748032,
      "grad_norm": 33.5,
      "learning_rate": 4.750199881200124e-07,
      "loss": 4.5294,
      "num_input_tokens_seen": 50833600,
      "step": 37515
    },
    {
      "epoch": 1.2009474425452915,
      "grad_norm": 20.5,
      "learning_rate": 4.7311717305633664e-07,
      "loss": 4.681,
      "num_input_tokens_seen": 50840224,
      "step": 37520
    },
    {
      "epoch": 1.20110748351578,
      "grad_norm": 28.875,
      "learning_rate": 4.7121814034621623e-07,
      "loss": 4.6892,
      "num_input_tokens_seen": 50847136,
      "step": 37525
    },
    {
      "epoch": 1.2012675244862685,
      "grad_norm": 19.625,
      "learning_rate": 4.693228902825114e-07,
      "loss": 4.7233,
      "num_input_tokens_seen": 50853648,
      "step": 37530
    },
    {
      "epoch": 1.201427565456757,
      "grad_norm": 23.0,
      "learning_rate": 4.6743142315748277e-07,
      "loss": 4.4949,
      "num_input_tokens_seen": 50860224,
      "step": 37535
    },
    {
      "epoch": 1.2015876064272453,
      "grad_norm": 27.5,
      "learning_rate": 4.655437392628276e-07,
      "loss": 4.6437,
      "num_input_tokens_seen": 50866752,
      "step": 37540
    },
    {
      "epoch": 1.2017476473977338,
      "grad_norm": 20.25,
      "learning_rate": 4.636598388896463e-07,
      "loss": 4.7672,
      "num_input_tokens_seen": 50873536,
      "step": 37545
    },
    {
      "epoch": 1.2019076883682223,
      "grad_norm": 29.375,
      "learning_rate": 4.6177972232845925e-07,
      "loss": 4.3946,
      "num_input_tokens_seen": 50880256,
      "step": 37550
    },
    {
      "epoch": 1.2020677293387108,
      "grad_norm": 56.75,
      "learning_rate": 4.5990338986920953e-07,
      "loss": 4.7367,
      "num_input_tokens_seen": 50887744,
      "step": 37555
    },
    {
      "epoch": 1.2022277703091993,
      "grad_norm": 33.0,
      "learning_rate": 4.5803084180124633e-07,
      "loss": 4.7907,
      "num_input_tokens_seen": 50894784,
      "step": 37560
    },
    {
      "epoch": 1.2023878112796875,
      "grad_norm": 43.5,
      "learning_rate": 4.561620784133386e-07,
      "loss": 4.4578,
      "num_input_tokens_seen": 50901472,
      "step": 37565
    },
    {
      "epoch": 1.202547852250176,
      "grad_norm": 41.25,
      "learning_rate": 4.5429709999367796e-07,
      "loss": 3.8765,
      "num_input_tokens_seen": 50908112,
      "step": 37570
    },
    {
      "epoch": 1.2027078932206645,
      "grad_norm": 24.25,
      "learning_rate": 4.5243590682986223e-07,
      "loss": 3.0515,
      "num_input_tokens_seen": 50914960,
      "step": 37575
    },
    {
      "epoch": 1.202867934191153,
      "grad_norm": 43.25,
      "learning_rate": 4.5057849920891735e-07,
      "loss": 4.978,
      "num_input_tokens_seen": 50921920,
      "step": 37580
    },
    {
      "epoch": 1.2030279751616413,
      "grad_norm": 19.5,
      "learning_rate": 4.487248774172698e-07,
      "loss": 3.7118,
      "num_input_tokens_seen": 50928880,
      "step": 37585
    },
    {
      "epoch": 1.2031880161321298,
      "grad_norm": 28.5,
      "learning_rate": 4.4687504174077965e-07,
      "loss": 5.0278,
      "num_input_tokens_seen": 50939504,
      "step": 37590
    },
    {
      "epoch": 1.2033480571026183,
      "grad_norm": 44.75,
      "learning_rate": 4.450289924647133e-07,
      "loss": 3.9681,
      "num_input_tokens_seen": 50946352,
      "step": 37595
    },
    {
      "epoch": 1.2035080980731068,
      "grad_norm": 53.75,
      "learning_rate": 4.431867298737513e-07,
      "loss": 4.6008,
      "num_input_tokens_seen": 50953312,
      "step": 37600
    },
    {
      "epoch": 1.2035080980731068,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.8772,
      "eval_samples_per_second": 28.579,
      "eval_steps_per_second": 14.29,
      "num_input_tokens_seen": 50953312,
      "step": 37600
    },
    {
      "epoch": 1.2036681390435953,
      "grad_norm": 56.0,
      "learning_rate": 4.41348254251997e-07,
      "loss": 4.6512,
      "num_input_tokens_seen": 50959648,
      "step": 37605
    },
    {
      "epoch": 1.2038281800140835,
      "grad_norm": 41.25,
      "learning_rate": 4.395135658829652e-07,
      "loss": 4.0786,
      "num_input_tokens_seen": 50966240,
      "step": 37610
    },
    {
      "epoch": 1.203988220984572,
      "grad_norm": 39.5,
      "learning_rate": 4.376826650495852e-07,
      "loss": 5.7336,
      "num_input_tokens_seen": 50972944,
      "step": 37615
    },
    {
      "epoch": 1.2041482619550605,
      "grad_norm": 33.0,
      "learning_rate": 4.358555520342117e-07,
      "loss": 4.8504,
      "num_input_tokens_seen": 50979328,
      "step": 37620
    },
    {
      "epoch": 1.204308302925549,
      "grad_norm": 37.0,
      "learning_rate": 4.3403222711860257e-07,
      "loss": 4.6413,
      "num_input_tokens_seen": 50985792,
      "step": 37625
    },
    {
      "epoch": 1.2044683438960373,
      "grad_norm": 37.25,
      "learning_rate": 4.3221269058394133e-07,
      "loss": 4.3637,
      "num_input_tokens_seen": 50993024,
      "step": 37630
    },
    {
      "epoch": 1.2046283848665258,
      "grad_norm": 25.75,
      "learning_rate": 4.303969427108173e-07,
      "loss": 3.6031,
      "num_input_tokens_seen": 51000256,
      "step": 37635
    },
    {
      "epoch": 1.2047884258370143,
      "grad_norm": 22.5,
      "learning_rate": 4.2858498377924825e-07,
      "loss": 4.842,
      "num_input_tokens_seen": 51007008,
      "step": 37640
    },
    {
      "epoch": 1.2049484668075028,
      "grad_norm": 29.125,
      "learning_rate": 4.267768140686579e-07,
      "loss": 4.3039,
      "num_input_tokens_seen": 51013760,
      "step": 37645
    },
    {
      "epoch": 1.2051085077779913,
      "grad_norm": 26.75,
      "learning_rate": 4.2497243385788975e-07,
      "loss": 3.8446,
      "num_input_tokens_seen": 51021104,
      "step": 37650
    },
    {
      "epoch": 1.2052685487484796,
      "grad_norm": 31.75,
      "learning_rate": 4.231718434251991e-07,
      "loss": 4.1574,
      "num_input_tokens_seen": 51027712,
      "step": 37655
    },
    {
      "epoch": 1.205428589718968,
      "grad_norm": 33.25,
      "learning_rate": 4.213750430482666e-07,
      "loss": 5.73,
      "num_input_tokens_seen": 51034416,
      "step": 37660
    },
    {
      "epoch": 1.2055886306894565,
      "grad_norm": 34.5,
      "learning_rate": 4.1958203300417054e-07,
      "loss": 4.6582,
      "num_input_tokens_seen": 51040912,
      "step": 37665
    },
    {
      "epoch": 1.205748671659945,
      "grad_norm": 36.5,
      "learning_rate": 4.177928135694259e-07,
      "loss": 5.3854,
      "num_input_tokens_seen": 51047616,
      "step": 37670
    },
    {
      "epoch": 1.2059087126304333,
      "grad_norm": 34.5,
      "learning_rate": 4.1600738501994807e-07,
      "loss": 4.2722,
      "num_input_tokens_seen": 51054400,
      "step": 37675
    },
    {
      "epoch": 1.2060687536009218,
      "grad_norm": 16.125,
      "learning_rate": 4.1422574763107237e-07,
      "loss": 4.8867,
      "num_input_tokens_seen": 51061952,
      "step": 37680
    },
    {
      "epoch": 1.2062287945714103,
      "grad_norm": 21.625,
      "learning_rate": 4.124479016775512e-07,
      "loss": 4.3845,
      "num_input_tokens_seen": 51068560,
      "step": 37685
    },
    {
      "epoch": 1.2063888355418988,
      "grad_norm": 20.625,
      "learning_rate": 4.106738474335514e-07,
      "loss": 5.0289,
      "num_input_tokens_seen": 51075344,
      "step": 37690
    },
    {
      "epoch": 1.2065488765123873,
      "grad_norm": 47.5,
      "learning_rate": 4.089035851726486e-07,
      "loss": 4.6027,
      "num_input_tokens_seen": 51082416,
      "step": 37695
    },
    {
      "epoch": 1.2067089174828756,
      "grad_norm": 25.25,
      "learning_rate": 4.0713711516784937e-07,
      "loss": 4.372,
      "num_input_tokens_seen": 51089200,
      "step": 37700
    },
    {
      "epoch": 1.206868958453364,
      "grad_norm": 25.875,
      "learning_rate": 4.05374437691558e-07,
      "loss": 3.8812,
      "num_input_tokens_seen": 51095840,
      "step": 37705
    },
    {
      "epoch": 1.2070289994238526,
      "grad_norm": 26.125,
      "learning_rate": 4.036155530156044e-07,
      "loss": 4.1361,
      "num_input_tokens_seen": 51102608,
      "step": 37710
    },
    {
      "epoch": 1.2071890403943408,
      "grad_norm": 24.75,
      "learning_rate": 4.018604614112298e-07,
      "loss": 4.558,
      "num_input_tokens_seen": 51108928,
      "step": 37715
    },
    {
      "epoch": 1.2073490813648293,
      "grad_norm": 15.875,
      "learning_rate": 4.0010916314908996e-07,
      "loss": 3.7521,
      "num_input_tokens_seen": 51116048,
      "step": 37720
    },
    {
      "epoch": 1.2075091223353178,
      "grad_norm": 30.0,
      "learning_rate": 3.983616584992578e-07,
      "loss": 3.7311,
      "num_input_tokens_seen": 51122816,
      "step": 37725
    },
    {
      "epoch": 1.2076691633058063,
      "grad_norm": 15.6875,
      "learning_rate": 3.9661794773122595e-07,
      "loss": 4.7073,
      "num_input_tokens_seen": 51129280,
      "step": 37730
    },
    {
      "epoch": 1.2078292042762948,
      "grad_norm": 30.75,
      "learning_rate": 3.9487803111388777e-07,
      "loss": 4.2508,
      "num_input_tokens_seen": 51136208,
      "step": 37735
    },
    {
      "epoch": 1.207989245246783,
      "grad_norm": 22.5,
      "learning_rate": 3.9314190891556747e-07,
      "loss": 3.9618,
      "num_input_tokens_seen": 51142928,
      "step": 37740
    },
    {
      "epoch": 1.2081492862172716,
      "grad_norm": 25.375,
      "learning_rate": 3.914095814039925e-07,
      "loss": 3.2917,
      "num_input_tokens_seen": 51149584,
      "step": 37745
    },
    {
      "epoch": 1.20830932718776,
      "grad_norm": 23.875,
      "learning_rate": 3.896810488463104e-07,
      "loss": 3.9719,
      "num_input_tokens_seen": 51155984,
      "step": 37750
    },
    {
      "epoch": 1.2084693681582486,
      "grad_norm": 20.875,
      "learning_rate": 3.8795631150908565e-07,
      "loss": 3.8917,
      "num_input_tokens_seen": 51162864,
      "step": 37755
    },
    {
      "epoch": 1.2086294091287368,
      "grad_norm": 31.5,
      "learning_rate": 3.862353696582888e-07,
      "loss": 4.4276,
      "num_input_tokens_seen": 51169984,
      "step": 37760
    },
    {
      "epoch": 1.2087894500992253,
      "grad_norm": 22.125,
      "learning_rate": 3.8451822355931313e-07,
      "loss": 4.4941,
      "num_input_tokens_seen": 51176592,
      "step": 37765
    },
    {
      "epoch": 1.2089494910697138,
      "grad_norm": 29.0,
      "learning_rate": 3.82804873476969e-07,
      "loss": 4.3226,
      "num_input_tokens_seen": 51182992,
      "step": 37770
    },
    {
      "epoch": 1.2091095320402023,
      "grad_norm": 30.0,
      "learning_rate": 3.810953196754702e-07,
      "loss": 4.5315,
      "num_input_tokens_seen": 51189760,
      "step": 37775
    },
    {
      "epoch": 1.2092695730106908,
      "grad_norm": 34.75,
      "learning_rate": 3.793895624184529e-07,
      "loss": 4.6547,
      "num_input_tokens_seen": 51196304,
      "step": 37780
    },
    {
      "epoch": 1.209429613981179,
      "grad_norm": 25.625,
      "learning_rate": 3.776876019689679e-07,
      "loss": 4.7372,
      "num_input_tokens_seen": 51202656,
      "step": 37785
    },
    {
      "epoch": 1.2095896549516676,
      "grad_norm": 31.875,
      "learning_rate": 3.7598943858947743e-07,
      "loss": 4.3038,
      "num_input_tokens_seen": 51209440,
      "step": 37790
    },
    {
      "epoch": 1.209749695922156,
      "grad_norm": 31.0,
      "learning_rate": 3.742950725418637e-07,
      "loss": 4.2928,
      "num_input_tokens_seen": 51216320,
      "step": 37795
    },
    {
      "epoch": 1.2099097368926446,
      "grad_norm": 38.75,
      "learning_rate": 3.726045040874093e-07,
      "loss": 4.0899,
      "num_input_tokens_seen": 51223392,
      "step": 37800
    },
    {
      "epoch": 1.2099097368926446,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.8393,
      "eval_samples_per_second": 28.581,
      "eval_steps_per_second": 14.291,
      "num_input_tokens_seen": 51223392,
      "step": 37800
    },
    {
      "epoch": 1.2100697778631329,
      "grad_norm": 23.125,
      "learning_rate": 3.709177334868308e-07,
      "loss": 4.491,
      "num_input_tokens_seen": 51230128,
      "step": 37805
    },
    {
      "epoch": 1.2102298188336214,
      "grad_norm": 27.5,
      "learning_rate": 3.692347610002478e-07,
      "loss": 4.7843,
      "num_input_tokens_seen": 51236544,
      "step": 37810
    },
    {
      "epoch": 1.2103898598041098,
      "grad_norm": 43.75,
      "learning_rate": 3.675555868871916e-07,
      "loss": 4.7065,
      "num_input_tokens_seen": 51243264,
      "step": 37815
    },
    {
      "epoch": 1.2105499007745983,
      "grad_norm": 19.375,
      "learning_rate": 3.658802114066162e-07,
      "loss": 5.3883,
      "num_input_tokens_seen": 51250048,
      "step": 37820
    },
    {
      "epoch": 1.2107099417450868,
      "grad_norm": 32.5,
      "learning_rate": 3.6420863481688437e-07,
      "loss": 4.4478,
      "num_input_tokens_seen": 51256576,
      "step": 37825
    },
    {
      "epoch": 1.210869982715575,
      "grad_norm": 30.625,
      "learning_rate": 3.625408573757705e-07,
      "loss": 4.5415,
      "num_input_tokens_seen": 51262960,
      "step": 37830
    },
    {
      "epoch": 1.2110300236860636,
      "grad_norm": 27.25,
      "learning_rate": 3.608768793404743e-07,
      "loss": 4.7021,
      "num_input_tokens_seen": 51270176,
      "step": 37835
    },
    {
      "epoch": 1.211190064656552,
      "grad_norm": 22.25,
      "learning_rate": 3.592167009675934e-07,
      "loss": 3.4635,
      "num_input_tokens_seen": 51276752,
      "step": 37840
    },
    {
      "epoch": 1.2113501056270406,
      "grad_norm": 22.5,
      "learning_rate": 3.575603225131563e-07,
      "loss": 4.0145,
      "num_input_tokens_seen": 51283712,
      "step": 37845
    },
    {
      "epoch": 1.2115101465975289,
      "grad_norm": 34.0,
      "learning_rate": 3.55907744232592e-07,
      "loss": 3.5923,
      "num_input_tokens_seen": 51290112,
      "step": 37850
    },
    {
      "epoch": 1.2116701875680174,
      "grad_norm": 22.875,
      "learning_rate": 3.5425896638075217e-07,
      "loss": 3.5661,
      "num_input_tokens_seen": 51296672,
      "step": 37855
    },
    {
      "epoch": 1.2118302285385059,
      "grad_norm": 33.0,
      "learning_rate": 3.5261398921189736e-07,
      "loss": 3.6048,
      "num_input_tokens_seen": 51303168,
      "step": 37860
    },
    {
      "epoch": 1.2119902695089944,
      "grad_norm": 20.875,
      "learning_rate": 3.509728129797024e-07,
      "loss": 5.0545,
      "num_input_tokens_seen": 51309552,
      "step": 37865
    },
    {
      "epoch": 1.2121503104794829,
      "grad_norm": 30.375,
      "learning_rate": 3.4933543793725656e-07,
      "loss": 4.1733,
      "num_input_tokens_seen": 51316112,
      "step": 37870
    },
    {
      "epoch": 1.2123103514499711,
      "grad_norm": 32.0,
      "learning_rate": 3.4770186433707163e-07,
      "loss": 3.9594,
      "num_input_tokens_seen": 51322704,
      "step": 37875
    },
    {
      "epoch": 1.2124703924204596,
      "grad_norm": 48.75,
      "learning_rate": 3.4607209243105453e-07,
      "loss": 4.3739,
      "num_input_tokens_seen": 51329696,
      "step": 37880
    },
    {
      "epoch": 1.2126304333909481,
      "grad_norm": 27.25,
      "learning_rate": 3.444461224705431e-07,
      "loss": 4.562,
      "num_input_tokens_seen": 51336400,
      "step": 37885
    },
    {
      "epoch": 1.2127904743614366,
      "grad_norm": 32.0,
      "learning_rate": 3.4282395470628116e-07,
      "loss": 4.2227,
      "num_input_tokens_seen": 51343184,
      "step": 37890
    },
    {
      "epoch": 1.2129505153319249,
      "grad_norm": 19.25,
      "learning_rate": 3.4120558938842417e-07,
      "loss": 3.7039,
      "num_input_tokens_seen": 51350096,
      "step": 37895
    },
    {
      "epoch": 1.2131105563024134,
      "grad_norm": 27.375,
      "learning_rate": 3.395910267665503e-07,
      "loss": 3.9455,
      "num_input_tokens_seen": 51356960,
      "step": 37900
    },
    {
      "epoch": 1.2132705972729019,
      "grad_norm": 36.0,
      "learning_rate": 3.3798026708964094e-07,
      "loss": 4.5479,
      "num_input_tokens_seen": 51363360,
      "step": 37905
    },
    {
      "epoch": 1.2134306382433904,
      "grad_norm": 27.375,
      "learning_rate": 3.3637331060609456e-07,
      "loss": 4.1849,
      "num_input_tokens_seen": 51370144,
      "step": 37910
    },
    {
      "epoch": 1.2135906792138789,
      "grad_norm": 40.0,
      "learning_rate": 3.3477015756372966e-07,
      "loss": 4.3185,
      "num_input_tokens_seen": 51377216,
      "step": 37915
    },
    {
      "epoch": 1.2137507201843671,
      "grad_norm": 27.5,
      "learning_rate": 3.3317080820976785e-07,
      "loss": 4.3497,
      "num_input_tokens_seen": 51383712,
      "step": 37920
    },
    {
      "epoch": 1.2139107611548556,
      "grad_norm": 37.25,
      "learning_rate": 3.315752627908508e-07,
      "loss": 5.3036,
      "num_input_tokens_seen": 51390016,
      "step": 37925
    },
    {
      "epoch": 1.2140708021253441,
      "grad_norm": 27.25,
      "learning_rate": 3.299835215530317e-07,
      "loss": 5.506,
      "num_input_tokens_seen": 51396464,
      "step": 37930
    },
    {
      "epoch": 1.2142308430958326,
      "grad_norm": 13.8125,
      "learning_rate": 3.2839558474177245e-07,
      "loss": 3.3992,
      "num_input_tokens_seen": 51403200,
      "step": 37935
    },
    {
      "epoch": 1.214390884066321,
      "grad_norm": 32.75,
      "learning_rate": 3.2681145260196056e-07,
      "loss": 3.7672,
      "num_input_tokens_seen": 51409904,
      "step": 37940
    },
    {
      "epoch": 1.2145509250368094,
      "grad_norm": 37.25,
      "learning_rate": 3.252311253778839e-07,
      "loss": 3.9291,
      "num_input_tokens_seen": 51417088,
      "step": 37945
    },
    {
      "epoch": 1.2147109660072979,
      "grad_norm": 45.5,
      "learning_rate": 3.2365460331325034e-07,
      "loss": 5.9173,
      "num_input_tokens_seen": 51423776,
      "step": 37950
    },
    {
      "epoch": 1.2148710069777864,
      "grad_norm": 44.0,
      "learning_rate": 3.2208188665117934e-07,
      "loss": 4.206,
      "num_input_tokens_seen": 51430928,
      "step": 37955
    },
    {
      "epoch": 1.2150310479482749,
      "grad_norm": 22.875,
      "learning_rate": 3.205129756342018e-07,
      "loss": 4.0198,
      "num_input_tokens_seen": 51437744,
      "step": 37960
    },
    {
      "epoch": 1.2151910889187632,
      "grad_norm": 38.75,
      "learning_rate": 3.189478705042659e-07,
      "loss": 4.172,
      "num_input_tokens_seen": 51444496,
      "step": 37965
    },
    {
      "epoch": 1.2153511298892516,
      "grad_norm": 34.25,
      "learning_rate": 3.173865715027341e-07,
      "loss": 4.8177,
      "num_input_tokens_seen": 51451136,
      "step": 37970
    },
    {
      "epoch": 1.2155111708597401,
      "grad_norm": 23.125,
      "learning_rate": 3.158290788703694e-07,
      "loss": 4.5468,
      "num_input_tokens_seen": 51457680,
      "step": 37975
    },
    {
      "epoch": 1.2156712118302284,
      "grad_norm": 35.0,
      "learning_rate": 3.1427539284736297e-07,
      "loss": 4.1536,
      "num_input_tokens_seen": 51464368,
      "step": 37980
    },
    {
      "epoch": 1.215831252800717,
      "grad_norm": 38.0,
      "learning_rate": 3.127255136733093e-07,
      "loss": 4.9291,
      "num_input_tokens_seen": 51470880,
      "step": 37985
    },
    {
      "epoch": 1.2159912937712054,
      "grad_norm": 33.5,
      "learning_rate": 3.1117944158722544e-07,
      "loss": 4.7529,
      "num_input_tokens_seen": 51477952,
      "step": 37990
    },
    {
      "epoch": 1.216151334741694,
      "grad_norm": 47.75,
      "learning_rate": 3.0963717682752635e-07,
      "loss": 4.8884,
      "num_input_tokens_seen": 51485184,
      "step": 37995
    },
    {
      "epoch": 1.2163113757121824,
      "grad_norm": 20.0,
      "learning_rate": 3.080987196320578e-07,
      "loss": 4.6102,
      "num_input_tokens_seen": 51491824,
      "step": 38000
    },
    {
      "epoch": 1.2163113757121824,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.8613,
      "eval_samples_per_second": 28.58,
      "eval_steps_per_second": 14.29,
      "num_input_tokens_seen": 51491824,
      "step": 38000
    },
    {
      "epoch": 1.2164714166826707,
      "grad_norm": 32.0,
      "learning_rate": 3.065640702380607e-07,
      "loss": 4.5875,
      "num_input_tokens_seen": 51498448,
      "step": 38005
    },
    {
      "epoch": 1.2166314576531592,
      "grad_norm": 16.75,
      "learning_rate": 3.050332288822011e-07,
      "loss": 4.2605,
      "num_input_tokens_seen": 51505056,
      "step": 38010
    },
    {
      "epoch": 1.2167914986236477,
      "grad_norm": 37.5,
      "learning_rate": 3.035061958005542e-07,
      "loss": 3.5232,
      "num_input_tokens_seen": 51512032,
      "step": 38015
    },
    {
      "epoch": 1.2169515395941362,
      "grad_norm": 55.0,
      "learning_rate": 3.019829712286093e-07,
      "loss": 4.3355,
      "num_input_tokens_seen": 51518816,
      "step": 38020
    },
    {
      "epoch": 1.2171115805646244,
      "grad_norm": 39.25,
      "learning_rate": 3.004635554012647e-07,
      "loss": 5.2768,
      "num_input_tokens_seen": 51525248,
      "step": 38025
    },
    {
      "epoch": 1.217271621535113,
      "grad_norm": 42.5,
      "learning_rate": 2.9894794855283017e-07,
      "loss": 4.1378,
      "num_input_tokens_seen": 51532144,
      "step": 38030
    },
    {
      "epoch": 1.2174316625056014,
      "grad_norm": 25.75,
      "learning_rate": 2.9743615091703816e-07,
      "loss": 4.1499,
      "num_input_tokens_seen": 51538912,
      "step": 38035
    },
    {
      "epoch": 1.21759170347609,
      "grad_norm": 20.5,
      "learning_rate": 2.959281627270216e-07,
      "loss": 3.6592,
      "num_input_tokens_seen": 51545824,
      "step": 38040
    },
    {
      "epoch": 1.2177517444465784,
      "grad_norm": 21.25,
      "learning_rate": 2.944239842153362e-07,
      "loss": 3.7435,
      "num_input_tokens_seen": 51552304,
      "step": 38045
    },
    {
      "epoch": 1.2179117854170667,
      "grad_norm": 36.75,
      "learning_rate": 2.929236156139381e-07,
      "loss": 4.8325,
      "num_input_tokens_seen": 51559056,
      "step": 38050
    },
    {
      "epoch": 1.2180718263875552,
      "grad_norm": 36.0,
      "learning_rate": 2.9142705715420883e-07,
      "loss": 4.9056,
      "num_input_tokens_seen": 51565808,
      "step": 38055
    },
    {
      "epoch": 1.2182318673580437,
      "grad_norm": 25.625,
      "learning_rate": 2.8993430906693595e-07,
      "loss": 4.8516,
      "num_input_tokens_seen": 51572448,
      "step": 38060
    },
    {
      "epoch": 1.2183919083285322,
      "grad_norm": 30.625,
      "learning_rate": 2.88445371582316e-07,
      "loss": 4.576,
      "num_input_tokens_seen": 51579232,
      "step": 38065
    },
    {
      "epoch": 1.2185519492990204,
      "grad_norm": 48.25,
      "learning_rate": 2.8696024492996796e-07,
      "loss": 4.6203,
      "num_input_tokens_seen": 51585664,
      "step": 38070
    },
    {
      "epoch": 1.218711990269509,
      "grad_norm": 15.1875,
      "learning_rate": 2.854789293389115e-07,
      "loss": 4.3441,
      "num_input_tokens_seen": 51592464,
      "step": 38075
    },
    {
      "epoch": 1.2188720312399974,
      "grad_norm": 38.5,
      "learning_rate": 2.8400142503758606e-07,
      "loss": 4.6733,
      "num_input_tokens_seen": 51599184,
      "step": 38080
    },
    {
      "epoch": 1.219032072210486,
      "grad_norm": 55.0,
      "learning_rate": 2.8252773225384276e-07,
      "loss": 3.9492,
      "num_input_tokens_seen": 51605904,
      "step": 38085
    },
    {
      "epoch": 1.2191921131809744,
      "grad_norm": 30.375,
      "learning_rate": 2.8105785121494143e-07,
      "loss": 3.6819,
      "num_input_tokens_seen": 51612800,
      "step": 38090
    },
    {
      "epoch": 1.2193521541514627,
      "grad_norm": 16.25,
      "learning_rate": 2.795917821475563e-07,
      "loss": 4.7769,
      "num_input_tokens_seen": 51619456,
      "step": 38095
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 20.125,
      "learning_rate": 2.78129525277776e-07,
      "loss": 4.4059,
      "num_input_tokens_seen": 51626464,
      "step": 38100
    },
    {
      "epoch": 1.2196722360924397,
      "grad_norm": 21.5,
      "learning_rate": 2.766710808310952e-07,
      "loss": 4.9338,
      "num_input_tokens_seen": 51633616,
      "step": 38105
    },
    {
      "epoch": 1.2198322770629282,
      "grad_norm": 27.25,
      "learning_rate": 2.7521644903242827e-07,
      "loss": 3.3562,
      "num_input_tokens_seen": 51640160,
      "step": 38110
    },
    {
      "epoch": 1.2199923180334165,
      "grad_norm": 21.0,
      "learning_rate": 2.7376563010609593e-07,
      "loss": 4.5258,
      "num_input_tokens_seen": 51646864,
      "step": 38115
    },
    {
      "epoch": 1.220152359003905,
      "grad_norm": 19.0,
      "learning_rate": 2.72318624275833e-07,
      "loss": 4.3531,
      "num_input_tokens_seen": 51653520,
      "step": 38120
    },
    {
      "epoch": 1.2203123999743934,
      "grad_norm": 29.25,
      "learning_rate": 2.7087543176478324e-07,
      "loss": 4.0297,
      "num_input_tokens_seen": 51660560,
      "step": 38125
    },
    {
      "epoch": 1.220472440944882,
      "grad_norm": 21.375,
      "learning_rate": 2.694360527955103e-07,
      "loss": 3.4705,
      "num_input_tokens_seen": 51667392,
      "step": 38130
    },
    {
      "epoch": 1.2206324819153704,
      "grad_norm": 25.375,
      "learning_rate": 2.680004875899811e-07,
      "loss": 3.7024,
      "num_input_tokens_seen": 51674640,
      "step": 38135
    },
    {
      "epoch": 1.2207925228858587,
      "grad_norm": 32.5,
      "learning_rate": 2.665687363695768e-07,
      "loss": 4.6589,
      "num_input_tokens_seen": 51681232,
      "step": 38140
    },
    {
      "epoch": 1.2209525638563472,
      "grad_norm": 34.5,
      "learning_rate": 2.6514079935509584e-07,
      "loss": 4.6641,
      "num_input_tokens_seen": 51689152,
      "step": 38145
    },
    {
      "epoch": 1.2211126048268357,
      "grad_norm": 28.375,
      "learning_rate": 2.6371667676673983e-07,
      "loss": 4.5545,
      "num_input_tokens_seen": 51696208,
      "step": 38150
    },
    {
      "epoch": 1.2212726457973242,
      "grad_norm": 20.0,
      "learning_rate": 2.6229636882412755e-07,
      "loss": 3.8131,
      "num_input_tokens_seen": 51702928,
      "step": 38155
    },
    {
      "epoch": 1.2214326867678125,
      "grad_norm": 25.0,
      "learning_rate": 2.6087987574628935e-07,
      "loss": 3.8803,
      "num_input_tokens_seen": 51709808,
      "step": 38160
    },
    {
      "epoch": 1.221592727738301,
      "grad_norm": 31.375,
      "learning_rate": 2.5946719775166437e-07,
      "loss": 3.4975,
      "num_input_tokens_seen": 51716256,
      "step": 38165
    },
    {
      "epoch": 1.2217527687087895,
      "grad_norm": 16.625,
      "learning_rate": 2.5805833505810616e-07,
      "loss": 4.2116,
      "num_input_tokens_seen": 51723136,
      "step": 38170
    },
    {
      "epoch": 1.221912809679278,
      "grad_norm": 53.25,
      "learning_rate": 2.566532878828798e-07,
      "loss": 4.6984,
      "num_input_tokens_seen": 51729776,
      "step": 38175
    },
    {
      "epoch": 1.2220728506497665,
      "grad_norm": 33.25,
      "learning_rate": 2.552520564426619e-07,
      "loss": 4.3023,
      "num_input_tokens_seen": 51736480,
      "step": 38180
    },
    {
      "epoch": 1.2222328916202547,
      "grad_norm": 36.0,
      "learning_rate": 2.5385464095353803e-07,
      "loss": 3.6266,
      "num_input_tokens_seen": 51743008,
      "step": 38185
    },
    {
      "epoch": 1.2223929325907432,
      "grad_norm": 21.75,
      "learning_rate": 2.5246104163100804e-07,
      "loss": 4.1133,
      "num_input_tokens_seen": 51749376,
      "step": 38190
    },
    {
      "epoch": 1.2225529735612317,
      "grad_norm": 33.0,
      "learning_rate": 2.510712586899833e-07,
      "loss": 3.3113,
      "num_input_tokens_seen": 51755936,
      "step": 38195
    },
    {
      "epoch": 1.2227130145317202,
      "grad_norm": 54.25,
      "learning_rate": 2.4968529234478124e-07,
      "loss": 5.2984,
      "num_input_tokens_seen": 51763040,
      "step": 38200
    },
    {
      "epoch": 1.2227130145317202,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.7535,
      "eval_samples_per_second": 28.587,
      "eval_steps_per_second": 14.293,
      "num_input_tokens_seen": 51763040,
      "step": 38200
    },
    {
      "epoch": 1.2228730555022085,
      "grad_norm": 31.625,
      "learning_rate": 2.483031428091448e-07,
      "loss": 4.1643,
      "num_input_tokens_seen": 51769456,
      "step": 38205
    },
    {
      "epoch": 1.223033096472697,
      "grad_norm": 30.375,
      "learning_rate": 2.469248102962091e-07,
      "loss": 3.3596,
      "num_input_tokens_seen": 51776864,
      "step": 38210
    },
    {
      "epoch": 1.2231931374431855,
      "grad_norm": 28.75,
      "learning_rate": 2.4555029501853455e-07,
      "loss": 4.0616,
      "num_input_tokens_seen": 51784096,
      "step": 38215
    },
    {
      "epoch": 1.223353178413674,
      "grad_norm": 21.875,
      "learning_rate": 2.441795971880906e-07,
      "loss": 4.5514,
      "num_input_tokens_seen": 51790624,
      "step": 38220
    },
    {
      "epoch": 1.2235132193841625,
      "grad_norm": 43.5,
      "learning_rate": 2.4281271701625255e-07,
      "loss": 3.4676,
      "num_input_tokens_seen": 51797808,
      "step": 38225
    },
    {
      "epoch": 1.2236732603546507,
      "grad_norm": 26.5,
      "learning_rate": 2.4144965471381007e-07,
      "loss": 3.8205,
      "num_input_tokens_seen": 51804592,
      "step": 38230
    },
    {
      "epoch": 1.2238333013251392,
      "grad_norm": 38.25,
      "learning_rate": 2.400904104909674e-07,
      "loss": 4.6088,
      "num_input_tokens_seen": 51811488,
      "step": 38235
    },
    {
      "epoch": 1.2239933422956277,
      "grad_norm": 32.5,
      "learning_rate": 2.3873498455733725e-07,
      "loss": 4.7497,
      "num_input_tokens_seen": 51818336,
      "step": 38240
    },
    {
      "epoch": 1.224153383266116,
      "grad_norm": 63.0,
      "learning_rate": 2.3738337712194137e-07,
      "loss": 4.4251,
      "num_input_tokens_seen": 51825120,
      "step": 38245
    },
    {
      "epoch": 1.2243134242366045,
      "grad_norm": 40.5,
      "learning_rate": 2.3603558839321305e-07,
      "loss": 4.4446,
      "num_input_tokens_seen": 51832128,
      "step": 38250
    },
    {
      "epoch": 1.224473465207093,
      "grad_norm": 25.625,
      "learning_rate": 2.3469161857900267e-07,
      "loss": 5.0602,
      "num_input_tokens_seen": 51838816,
      "step": 38255
    },
    {
      "epoch": 1.2246335061775815,
      "grad_norm": 28.75,
      "learning_rate": 2.3335146788656393e-07,
      "loss": 3.7664,
      "num_input_tokens_seen": 51845248,
      "step": 38260
    },
    {
      "epoch": 1.22479354714807,
      "grad_norm": 43.0,
      "learning_rate": 2.3201513652256757e-07,
      "loss": 4.7319,
      "num_input_tokens_seen": 51851744,
      "step": 38265
    },
    {
      "epoch": 1.2249535881185583,
      "grad_norm": 29.25,
      "learning_rate": 2.3068262469308766e-07,
      "loss": 3.3559,
      "num_input_tokens_seen": 51858640,
      "step": 38270
    },
    {
      "epoch": 1.2251136290890468,
      "grad_norm": 41.0,
      "learning_rate": 2.2935393260362093e-07,
      "loss": 4.8554,
      "num_input_tokens_seen": 51865072,
      "step": 38275
    },
    {
      "epoch": 1.2252736700595352,
      "grad_norm": 29.5,
      "learning_rate": 2.2802906045906458e-07,
      "loss": 4.0615,
      "num_input_tokens_seen": 51872080,
      "step": 38280
    },
    {
      "epoch": 1.2254337110300237,
      "grad_norm": 30.5,
      "learning_rate": 2.2670800846373018e-07,
      "loss": 4.454,
      "num_input_tokens_seen": 51878784,
      "step": 38285
    },
    {
      "epoch": 1.225593752000512,
      "grad_norm": 25.5,
      "learning_rate": 2.2539077682134367e-07,
      "loss": 5.0736,
      "num_input_tokens_seen": 51885360,
      "step": 38290
    },
    {
      "epoch": 1.2257537929710005,
      "grad_norm": 20.625,
      "learning_rate": 2.2407736573503423e-07,
      "loss": 3.2214,
      "num_input_tokens_seen": 51892848,
      "step": 38295
    },
    {
      "epoch": 1.225913833941489,
      "grad_norm": 28.75,
      "learning_rate": 2.2276777540735093e-07,
      "loss": 3.6971,
      "num_input_tokens_seen": 51899408,
      "step": 38300
    },
    {
      "epoch": 1.2260738749119775,
      "grad_norm": 33.0,
      "learning_rate": 2.2146200604024613e-07,
      "loss": 4.7438,
      "num_input_tokens_seen": 51905920,
      "step": 38305
    },
    {
      "epoch": 1.226233915882466,
      "grad_norm": 44.0,
      "learning_rate": 2.2016005783508375e-07,
      "loss": 4.6395,
      "num_input_tokens_seen": 51912560,
      "step": 38310
    },
    {
      "epoch": 1.2263939568529543,
      "grad_norm": 29.0,
      "learning_rate": 2.1886193099264763e-07,
      "loss": 4.781,
      "num_input_tokens_seen": 51919376,
      "step": 38315
    },
    {
      "epoch": 1.2265539978234428,
      "grad_norm": 21.625,
      "learning_rate": 2.175676257131165e-07,
      "loss": 4.9748,
      "num_input_tokens_seen": 51926432,
      "step": 38320
    },
    {
      "epoch": 1.2267140387939313,
      "grad_norm": 29.375,
      "learning_rate": 2.162771421960974e-07,
      "loss": 4.5781,
      "num_input_tokens_seen": 51933232,
      "step": 38325
    },
    {
      "epoch": 1.2268740797644198,
      "grad_norm": 38.75,
      "learning_rate": 2.1499048064059224e-07,
      "loss": 4.5999,
      "num_input_tokens_seen": 51940208,
      "step": 38330
    },
    {
      "epoch": 1.227034120734908,
      "grad_norm": 29.75,
      "learning_rate": 2.1370764124502285e-07,
      "loss": 5.3523,
      "num_input_tokens_seen": 51947184,
      "step": 38335
    },
    {
      "epoch": 1.2271941617053965,
      "grad_norm": 19.0,
      "learning_rate": 2.1242862420721988e-07,
      "loss": 4.2213,
      "num_input_tokens_seen": 51953760,
      "step": 38340
    },
    {
      "epoch": 1.227354202675885,
      "grad_norm": 51.0,
      "learning_rate": 2.1115342972442276e-07,
      "loss": 4.3625,
      "num_input_tokens_seen": 51960400,
      "step": 38345
    },
    {
      "epoch": 1.2275142436463735,
      "grad_norm": 34.5,
      "learning_rate": 2.0988205799328252e-07,
      "loss": 3.8823,
      "num_input_tokens_seen": 51967024,
      "step": 38350
    },
    {
      "epoch": 1.227674284616862,
      "grad_norm": 26.375,
      "learning_rate": 2.0861450920986182e-07,
      "loss": 3.4012,
      "num_input_tokens_seen": 51974144,
      "step": 38355
    },
    {
      "epoch": 1.2278343255873503,
      "grad_norm": 24.5,
      "learning_rate": 2.07350783569632e-07,
      "loss": 3.8632,
      "num_input_tokens_seen": 51980896,
      "step": 38360
    },
    {
      "epoch": 1.2279943665578388,
      "grad_norm": 29.125,
      "learning_rate": 2.060908812674761e-07,
      "loss": 4.7667,
      "num_input_tokens_seen": 51987568,
      "step": 38365
    },
    {
      "epoch": 1.2281544075283273,
      "grad_norm": 23.125,
      "learning_rate": 2.0483480249768317e-07,
      "loss": 3.8882,
      "num_input_tokens_seen": 51994288,
      "step": 38370
    },
    {
      "epoch": 1.2283144484988158,
      "grad_norm": 34.5,
      "learning_rate": 2.035825474539621e-07,
      "loss": 5.4296,
      "num_input_tokens_seen": 52001184,
      "step": 38375
    },
    {
      "epoch": 1.228474489469304,
      "grad_norm": 25.375,
      "learning_rate": 2.0233411632942235e-07,
      "loss": 3.8652,
      "num_input_tokens_seen": 52007584,
      "step": 38380
    },
    {
      "epoch": 1.2286345304397925,
      "grad_norm": 25.375,
      "learning_rate": 2.0108950931658764e-07,
      "loss": 6.1087,
      "num_input_tokens_seen": 52013888,
      "step": 38385
    },
    {
      "epoch": 1.228794571410281,
      "grad_norm": 22.5,
      "learning_rate": 1.998487266073934e-07,
      "loss": 4.0945,
      "num_input_tokens_seen": 52020128,
      "step": 38390
    },
    {
      "epoch": 1.2289546123807695,
      "grad_norm": 28.75,
      "learning_rate": 1.986117683931865e-07,
      "loss": 3.9315,
      "num_input_tokens_seen": 52026592,
      "step": 38395
    },
    {
      "epoch": 1.229114653351258,
      "grad_norm": 36.75,
      "learning_rate": 1.9737863486471442e-07,
      "loss": 3.8424,
      "num_input_tokens_seen": 52033392,
      "step": 38400
    },
    {
      "epoch": 1.229114653351258,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.994,
      "eval_samples_per_second": 28.572,
      "eval_steps_per_second": 14.286,
      "num_input_tokens_seen": 52033392,
      "step": 38400
    },
    {
      "epoch": 1.2292746943217463,
      "grad_norm": 27.75,
      "learning_rate": 1.9614932621215e-07,
      "loss": 4.8051,
      "num_input_tokens_seen": 52040176,
      "step": 38405
    },
    {
      "epoch": 1.2294347352922348,
      "grad_norm": 22.625,
      "learning_rate": 1.9492384262506102e-07,
      "loss": 4.4187,
      "num_input_tokens_seen": 52046832,
      "step": 38410
    },
    {
      "epoch": 1.2295947762627233,
      "grad_norm": 23.125,
      "learning_rate": 1.9370218429243524e-07,
      "loss": 4.5019,
      "num_input_tokens_seen": 52053520,
      "step": 38415
    },
    {
      "epoch": 1.2297548172332118,
      "grad_norm": 27.875,
      "learning_rate": 1.9248435140267197e-07,
      "loss": 4.5229,
      "num_input_tokens_seen": 52060336,
      "step": 38420
    },
    {
      "epoch": 1.2299148582037,
      "grad_norm": 20.0,
      "learning_rate": 1.9127034414356814e-07,
      "loss": 4.6181,
      "num_input_tokens_seen": 52066880,
      "step": 38425
    },
    {
      "epoch": 1.2300748991741886,
      "grad_norm": 30.5,
      "learning_rate": 1.9006016270234627e-07,
      "loss": 4.8169,
      "num_input_tokens_seen": 52073344,
      "step": 38430
    },
    {
      "epoch": 1.230234940144677,
      "grad_norm": 16.375,
      "learning_rate": 1.888538072656293e-07,
      "loss": 3.9977,
      "num_input_tokens_seen": 52080112,
      "step": 38435
    },
    {
      "epoch": 1.2303949811151655,
      "grad_norm": 47.0,
      "learning_rate": 1.8765127801944893e-07,
      "loss": 4.0626,
      "num_input_tokens_seen": 52086896,
      "step": 38440
    },
    {
      "epoch": 1.230555022085654,
      "grad_norm": 18.875,
      "learning_rate": 1.8645257514925406e-07,
      "loss": 4.6847,
      "num_input_tokens_seen": 52093568,
      "step": 38445
    },
    {
      "epoch": 1.2307150630561423,
      "grad_norm": 20.125,
      "learning_rate": 1.8525769883989685e-07,
      "loss": 4.2068,
      "num_input_tokens_seen": 52100496,
      "step": 38450
    },
    {
      "epoch": 1.2308751040266308,
      "grad_norm": 54.75,
      "learning_rate": 1.8406664927564654e-07,
      "loss": 4.008,
      "num_input_tokens_seen": 52107392,
      "step": 38455
    },
    {
      "epoch": 1.2310351449971193,
      "grad_norm": 44.25,
      "learning_rate": 1.8287942664017566e-07,
      "loss": 4.6542,
      "num_input_tokens_seen": 52114064,
      "step": 38460
    },
    {
      "epoch": 1.2311951859676078,
      "grad_norm": 14.5625,
      "learning_rate": 1.8169603111656552e-07,
      "loss": 4.8812,
      "num_input_tokens_seen": 52120992,
      "step": 38465
    },
    {
      "epoch": 1.231355226938096,
      "grad_norm": 23.125,
      "learning_rate": 1.805164628873146e-07,
      "loss": 5.6227,
      "num_input_tokens_seen": 52128400,
      "step": 38470
    },
    {
      "epoch": 1.2315152679085846,
      "grad_norm": 51.0,
      "learning_rate": 1.793407221343274e-07,
      "loss": 5.1458,
      "num_input_tokens_seen": 52135360,
      "step": 38475
    },
    {
      "epoch": 1.231675308879073,
      "grad_norm": 34.25,
      "learning_rate": 1.781688090389172e-07,
      "loss": 3.9285,
      "num_input_tokens_seen": 52142272,
      "step": 38480
    },
    {
      "epoch": 1.2318353498495616,
      "grad_norm": 38.25,
      "learning_rate": 1.770007237818061e-07,
      "loss": 4.1257,
      "num_input_tokens_seen": 52149056,
      "step": 38485
    },
    {
      "epoch": 1.23199539082005,
      "grad_norm": 38.75,
      "learning_rate": 1.7583646654313059e-07,
      "loss": 5.3309,
      "num_input_tokens_seen": 52155472,
      "step": 38490
    },
    {
      "epoch": 1.2321554317905383,
      "grad_norm": 40.5,
      "learning_rate": 1.7467603750242757e-07,
      "loss": 4.3646,
      "num_input_tokens_seen": 52162176,
      "step": 38495
    },
    {
      "epoch": 1.2323154727610268,
      "grad_norm": 41.0,
      "learning_rate": 1.7351943683865944e-07,
      "loss": 4.726,
      "num_input_tokens_seen": 52168352,
      "step": 38500
    },
    {
      "epoch": 1.2324755137315153,
      "grad_norm": 25.375,
      "learning_rate": 1.723666647301808e-07,
      "loss": 4.5014,
      "num_input_tokens_seen": 52174944,
      "step": 38505
    },
    {
      "epoch": 1.2326355547020036,
      "grad_norm": 25.0,
      "learning_rate": 1.712177213547661e-07,
      "loss": 3.754,
      "num_input_tokens_seen": 52181648,
      "step": 38510
    },
    {
      "epoch": 1.232795595672492,
      "grad_norm": 29.25,
      "learning_rate": 1.7007260688959581e-07,
      "loss": 4.5077,
      "num_input_tokens_seen": 52188496,
      "step": 38515
    },
    {
      "epoch": 1.2329556366429806,
      "grad_norm": 32.0,
      "learning_rate": 1.68931321511262e-07,
      "loss": 5.3479,
      "num_input_tokens_seen": 52195184,
      "step": 38520
    },
    {
      "epoch": 1.233115677613469,
      "grad_norm": 34.5,
      "learning_rate": 1.6779386539576835e-07,
      "loss": 3.7521,
      "num_input_tokens_seen": 52202368,
      "step": 38525
    },
    {
      "epoch": 1.2332757185839576,
      "grad_norm": 33.25,
      "learning_rate": 1.666602387185162e-07,
      "loss": 4.4477,
      "num_input_tokens_seen": 52209024,
      "step": 38530
    },
    {
      "epoch": 1.2334357595544458,
      "grad_norm": 16.625,
      "learning_rate": 1.655304416543352e-07,
      "loss": 4.5712,
      "num_input_tokens_seen": 52215904,
      "step": 38535
    },
    {
      "epoch": 1.2335958005249343,
      "grad_norm": 37.5,
      "learning_rate": 1.6440447437744698e-07,
      "loss": 4.23,
      "num_input_tokens_seen": 52222544,
      "step": 38540
    },
    {
      "epoch": 1.2337558414954228,
      "grad_norm": 16.25,
      "learning_rate": 1.6328233706149332e-07,
      "loss": 5.0862,
      "num_input_tokens_seen": 52229440,
      "step": 38545
    },
    {
      "epoch": 1.2339158824659113,
      "grad_norm": 46.5,
      "learning_rate": 1.6216402987951906e-07,
      "loss": 4.7653,
      "num_input_tokens_seen": 52235680,
      "step": 38550
    },
    {
      "epoch": 1.2340759234363996,
      "grad_norm": 19.5,
      "learning_rate": 1.6104955300398627e-07,
      "loss": 3.861,
      "num_input_tokens_seen": 52242464,
      "step": 38555
    },
    {
      "epoch": 1.234235964406888,
      "grad_norm": 41.25,
      "learning_rate": 1.5993890660675748e-07,
      "loss": 4.7979,
      "num_input_tokens_seen": 52249296,
      "step": 38560
    },
    {
      "epoch": 1.2343960053773766,
      "grad_norm": 20.875,
      "learning_rate": 1.5883209085910678e-07,
      "loss": 4.1123,
      "num_input_tokens_seen": 52256480,
      "step": 38565
    },
    {
      "epoch": 1.234556046347865,
      "grad_norm": 34.0,
      "learning_rate": 1.5772910593172264e-07,
      "loss": 5.1481,
      "num_input_tokens_seen": 52263424,
      "step": 38570
    },
    {
      "epoch": 1.2347160873183536,
      "grad_norm": 38.75,
      "learning_rate": 1.5662995199469954e-07,
      "loss": 4.4644,
      "num_input_tokens_seen": 52270512,
      "step": 38575
    },
    {
      "epoch": 1.2348761282888419,
      "grad_norm": 27.625,
      "learning_rate": 1.5553462921753802e-07,
      "loss": 4.5257,
      "num_input_tokens_seen": 52277552,
      "step": 38580
    },
    {
      "epoch": 1.2350361692593304,
      "grad_norm": 29.5,
      "learning_rate": 1.544431377691502e-07,
      "loss": 4.2569,
      "num_input_tokens_seen": 52284256,
      "step": 38585
    },
    {
      "epoch": 1.2351962102298188,
      "grad_norm": 24.75,
      "learning_rate": 1.5335547781785975e-07,
      "loss": 3.7849,
      "num_input_tokens_seen": 52290960,
      "step": 38590
    },
    {
      "epoch": 1.2353562512003073,
      "grad_norm": 39.0,
      "learning_rate": 1.5227164953139917e-07,
      "loss": 4.3092,
      "num_input_tokens_seen": 52297792,
      "step": 38595
    },
    {
      "epoch": 1.2355162921707956,
      "grad_norm": 24.375,
      "learning_rate": 1.511916530769042e-07,
      "loss": 4.7151,
      "num_input_tokens_seen": 52304608,
      "step": 38600
    },
    {
      "epoch": 1.2355162921707956,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.933,
      "eval_samples_per_second": 28.576,
      "eval_steps_per_second": 14.288,
      "num_input_tokens_seen": 52304608,
      "step": 38600
    },
    {
      "epoch": 1.2356763331412841,
      "grad_norm": 8.375,
      "learning_rate": 1.5011548862092773e-07,
      "loss": 4.2747,
      "num_input_tokens_seen": 52311568,
      "step": 38605
    },
    {
      "epoch": 1.2358363741117726,
      "grad_norm": 50.5,
      "learning_rate": 1.490431563294231e-07,
      "loss": 4.3628,
      "num_input_tokens_seen": 52318512,
      "step": 38610
    },
    {
      "epoch": 1.235996415082261,
      "grad_norm": 16.75,
      "learning_rate": 1.4797465636776365e-07,
      "loss": 4.0041,
      "num_input_tokens_seen": 52325792,
      "step": 38615
    },
    {
      "epoch": 1.2361564560527496,
      "grad_norm": 50.75,
      "learning_rate": 1.4690998890072027e-07,
      "loss": 4.2741,
      "num_input_tokens_seen": 52332016,
      "step": 38620
    },
    {
      "epoch": 1.2363164970232379,
      "grad_norm": 36.5,
      "learning_rate": 1.4584915409248112e-07,
      "loss": 3.1077,
      "num_input_tokens_seen": 52338656,
      "step": 38625
    },
    {
      "epoch": 1.2364765379937264,
      "grad_norm": 17.75,
      "learning_rate": 1.4479215210663754e-07,
      "loss": 5.2377,
      "num_input_tokens_seen": 52345712,
      "step": 38630
    },
    {
      "epoch": 1.2366365789642149,
      "grad_norm": 25.875,
      "learning_rate": 1.4373898310619528e-07,
      "loss": 4.575,
      "num_input_tokens_seen": 52352416,
      "step": 38635
    },
    {
      "epoch": 1.2367966199347034,
      "grad_norm": 23.375,
      "learning_rate": 1.4268964725356604e-07,
      "loss": 4.0985,
      "num_input_tokens_seen": 52359024,
      "step": 38640
    },
    {
      "epoch": 1.2369566609051916,
      "grad_norm": 14.3125,
      "learning_rate": 1.4164414471056764e-07,
      "loss": 4.7246,
      "num_input_tokens_seen": 52366144,
      "step": 38645
    },
    {
      "epoch": 1.2371167018756801,
      "grad_norm": 37.25,
      "learning_rate": 1.4060247563843497e-07,
      "loss": 5.2229,
      "num_input_tokens_seen": 52372992,
      "step": 38650
    },
    {
      "epoch": 1.2372767428461686,
      "grad_norm": 29.875,
      "learning_rate": 1.3956464019780068e-07,
      "loss": 3.8551,
      "num_input_tokens_seen": 52379632,
      "step": 38655
    },
    {
      "epoch": 1.2374367838166571,
      "grad_norm": 20.75,
      "learning_rate": 1.385306385487145e-07,
      "loss": 4.0969,
      "num_input_tokens_seen": 52386432,
      "step": 38660
    },
    {
      "epoch": 1.2375968247871456,
      "grad_norm": 31.875,
      "learning_rate": 1.3750047085063222e-07,
      "loss": 3.6994,
      "num_input_tokens_seen": 52393024,
      "step": 38665
    },
    {
      "epoch": 1.2377568657576339,
      "grad_norm": 41.75,
      "learning_rate": 1.3647413726242119e-07,
      "loss": 4.9494,
      "num_input_tokens_seen": 52399616,
      "step": 38670
    },
    {
      "epoch": 1.2379169067281224,
      "grad_norm": 18.625,
      "learning_rate": 1.3545163794235205e-07,
      "loss": 4.5529,
      "num_input_tokens_seen": 52406368,
      "step": 38675
    },
    {
      "epoch": 1.2380769476986109,
      "grad_norm": 29.375,
      "learning_rate": 1.3443297304810698e-07,
      "loss": 4.7243,
      "num_input_tokens_seen": 52412752,
      "step": 38680
    },
    {
      "epoch": 1.2382369886690994,
      "grad_norm": 28.25,
      "learning_rate": 1.3341814273677977e-07,
      "loss": 4.1692,
      "num_input_tokens_seen": 52419248,
      "step": 38685
    },
    {
      "epoch": 1.2383970296395876,
      "grad_norm": 37.5,
      "learning_rate": 1.324071471648647e-07,
      "loss": 6.1174,
      "num_input_tokens_seen": 52426048,
      "step": 38690
    },
    {
      "epoch": 1.2385570706100761,
      "grad_norm": 35.5,
      "learning_rate": 1.3139998648827312e-07,
      "loss": 4.3609,
      "num_input_tokens_seen": 52432752,
      "step": 38695
    },
    {
      "epoch": 1.2387171115805646,
      "grad_norm": 15.0,
      "learning_rate": 1.3039666086232526e-07,
      "loss": 3.9051,
      "num_input_tokens_seen": 52439376,
      "step": 38700
    },
    {
      "epoch": 1.2388771525510531,
      "grad_norm": 53.75,
      "learning_rate": 1.2939717044174183e-07,
      "loss": 4.1436,
      "num_input_tokens_seen": 52446192,
      "step": 38705
    },
    {
      "epoch": 1.2390371935215416,
      "grad_norm": 28.875,
      "learning_rate": 1.284015153806578e-07,
      "loss": 3.8553,
      "num_input_tokens_seen": 52452736,
      "step": 38710
    },
    {
      "epoch": 1.23919723449203,
      "grad_norm": 18.75,
      "learning_rate": 1.274096958326171e-07,
      "loss": 4.6064,
      "num_input_tokens_seen": 52459968,
      "step": 38715
    },
    {
      "epoch": 1.2393572754625184,
      "grad_norm": 42.25,
      "learning_rate": 1.2642171195056952e-07,
      "loss": 4.7882,
      "num_input_tokens_seen": 52466976,
      "step": 38720
    },
    {
      "epoch": 1.2395173164330069,
      "grad_norm": 24.25,
      "learning_rate": 1.2543756388687377e-07,
      "loss": 3.7343,
      "num_input_tokens_seen": 52473920,
      "step": 38725
    },
    {
      "epoch": 1.2396773574034954,
      "grad_norm": 30.5,
      "learning_rate": 1.2445725179330014e-07,
      "loss": 4.9702,
      "num_input_tokens_seen": 52480640,
      "step": 38730
    },
    {
      "epoch": 1.2398373983739837,
      "grad_norm": 20.375,
      "learning_rate": 1.2348077582102212e-07,
      "loss": 3.7485,
      "num_input_tokens_seen": 52487136,
      "step": 38735
    },
    {
      "epoch": 1.2399974393444722,
      "grad_norm": 35.0,
      "learning_rate": 1.2250813612062762e-07,
      "loss": 4.6604,
      "num_input_tokens_seen": 52493824,
      "step": 38740
    },
    {
      "epoch": 1.2401574803149606,
      "grad_norm": 40.25,
      "learning_rate": 1.215393328421105e-07,
      "loss": 3.9772,
      "num_input_tokens_seen": 52500192,
      "step": 38745
    },
    {
      "epoch": 1.2403175212854491,
      "grad_norm": 48.0,
      "learning_rate": 1.2057436613486796e-07,
      "loss": 4.4802,
      "num_input_tokens_seen": 52506928,
      "step": 38750
    },
    {
      "epoch": 1.2404775622559376,
      "grad_norm": 19.375,
      "learning_rate": 1.1961323614771424e-07,
      "loss": 4.5022,
      "num_input_tokens_seen": 52513680,
      "step": 38755
    },
    {
      "epoch": 1.240637603226426,
      "grad_norm": 29.375,
      "learning_rate": 1.1865594302886418e-07,
      "loss": 4.0424,
      "num_input_tokens_seen": 52520400,
      "step": 38760
    },
    {
      "epoch": 1.2407976441969144,
      "grad_norm": 24.875,
      "learning_rate": 1.1770248692594687e-07,
      "loss": 5.3184,
      "num_input_tokens_seen": 52527168,
      "step": 38765
    },
    {
      "epoch": 1.240957685167403,
      "grad_norm": 26.25,
      "learning_rate": 1.167528679859975e-07,
      "loss": 5.9621,
      "num_input_tokens_seen": 52534112,
      "step": 38770
    },
    {
      "epoch": 1.2411177261378914,
      "grad_norm": 20.125,
      "learning_rate": 1.1580708635545446e-07,
      "loss": 4.2441,
      "num_input_tokens_seen": 52541104,
      "step": 38775
    },
    {
      "epoch": 1.2412777671083797,
      "grad_norm": 25.5,
      "learning_rate": 1.1486514218017885e-07,
      "loss": 4.3666,
      "num_input_tokens_seen": 52547696,
      "step": 38780
    },
    {
      "epoch": 1.2414378080788682,
      "grad_norm": 31.625,
      "learning_rate": 1.1392703560542117e-07,
      "loss": 4.3083,
      "num_input_tokens_seen": 52554464,
      "step": 38785
    },
    {
      "epoch": 1.2415978490493567,
      "grad_norm": 28.875,
      "learning_rate": 1.129927667758518e-07,
      "loss": 3.8738,
      "num_input_tokens_seen": 52560752,
      "step": 38790
    },
    {
      "epoch": 1.2417578900198452,
      "grad_norm": 36.0,
      "learning_rate": 1.1206233583554992e-07,
      "loss": 4.7761,
      "num_input_tokens_seen": 52567584,
      "step": 38795
    },
    {
      "epoch": 1.2419179309903334,
      "grad_norm": 36.25,
      "learning_rate": 1.1113574292799523e-07,
      "loss": 4.5647,
      "num_input_tokens_seen": 52574352,
      "step": 38800
    },
    {
      "epoch": 1.2419179309903334,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.9449,
      "eval_samples_per_second": 28.575,
      "eval_steps_per_second": 14.288,
      "num_input_tokens_seen": 52574352,
      "step": 38800
    },
    {
      "epoch": 1.242077971960822,
      "grad_norm": 25.5,
      "learning_rate": 1.1021298819608449e-07,
      "loss": 4.6376,
      "num_input_tokens_seen": 52581264,
      "step": 38805
    },
    {
      "epoch": 1.2422380129313104,
      "grad_norm": 46.5,
      "learning_rate": 1.0929407178211226e-07,
      "loss": 4.302,
      "num_input_tokens_seen": 52587696,
      "step": 38810
    },
    {
      "epoch": 1.242398053901799,
      "grad_norm": 25.125,
      "learning_rate": 1.0837899382779293e-07,
      "loss": 5.1923,
      "num_input_tokens_seen": 52594320,
      "step": 38815
    },
    {
      "epoch": 1.2425580948722872,
      "grad_norm": 16.375,
      "learning_rate": 1.0746775447423862e-07,
      "loss": 4.4447,
      "num_input_tokens_seen": 52601040,
      "step": 38820
    },
    {
      "epoch": 1.2427181358427757,
      "grad_norm": 28.75,
      "learning_rate": 1.0656035386197583e-07,
      "loss": 3.7244,
      "num_input_tokens_seen": 52607840,
      "step": 38825
    },
    {
      "epoch": 1.2428781768132642,
      "grad_norm": 24.875,
      "learning_rate": 1.0565679213093982e-07,
      "loss": 4.4214,
      "num_input_tokens_seen": 52614704,
      "step": 38830
    },
    {
      "epoch": 1.2430382177837527,
      "grad_norm": 27.875,
      "learning_rate": 1.0475706942046638e-07,
      "loss": 4.3279,
      "num_input_tokens_seen": 52621120,
      "step": 38835
    },
    {
      "epoch": 1.2431982587542412,
      "grad_norm": 42.75,
      "learning_rate": 1.0386118586930282e-07,
      "loss": 5.0715,
      "num_input_tokens_seen": 52627840,
      "step": 38840
    },
    {
      "epoch": 1.2433582997247294,
      "grad_norm": 36.25,
      "learning_rate": 1.0296914161561367e-07,
      "loss": 4.6805,
      "num_input_tokens_seen": 52634336,
      "step": 38845
    },
    {
      "epoch": 1.243518340695218,
      "grad_norm": 24.5,
      "learning_rate": 1.0208093679695552e-07,
      "loss": 4.1433,
      "num_input_tokens_seen": 52641008,
      "step": 38850
    },
    {
      "epoch": 1.2436783816657064,
      "grad_norm": 26.75,
      "learning_rate": 1.0119657155030493e-07,
      "loss": 4.4316,
      "num_input_tokens_seen": 52647440,
      "step": 38855
    },
    {
      "epoch": 1.243838422636195,
      "grad_norm": 42.5,
      "learning_rate": 1.003160460120417e-07,
      "loss": 4.4571,
      "num_input_tokens_seen": 52654160,
      "step": 38860
    },
    {
      "epoch": 1.2439984636066832,
      "grad_norm": 28.0,
      "learning_rate": 9.943936031795165e-08,
      "loss": 4.9426,
      "num_input_tokens_seen": 52661024,
      "step": 38865
    },
    {
      "epoch": 1.2441585045771717,
      "grad_norm": 12.125,
      "learning_rate": 9.856651460323219e-08,
      "loss": 3.9107,
      "num_input_tokens_seen": 52667808,
      "step": 38870
    },
    {
      "epoch": 1.2443185455476602,
      "grad_norm": 17.5,
      "learning_rate": 9.769750900248953e-08,
      "loss": 4.1208,
      "num_input_tokens_seen": 52674640,
      "step": 38875
    },
    {
      "epoch": 1.2444785865181487,
      "grad_norm": 39.25,
      "learning_rate": 9.683234364973038e-08,
      "loss": 4.3767,
      "num_input_tokens_seen": 52681568,
      "step": 38880
    },
    {
      "epoch": 1.2446386274886372,
      "grad_norm": 31.125,
      "learning_rate": 9.597101867837854e-08,
      "loss": 4.5996,
      "num_input_tokens_seen": 52688032,
      "step": 38885
    },
    {
      "epoch": 1.2447986684591255,
      "grad_norm": 18.625,
      "learning_rate": 9.511353422125835e-08,
      "loss": 3.7108,
      "num_input_tokens_seen": 52695136,
      "step": 38890
    },
    {
      "epoch": 1.244958709429614,
      "grad_norm": 40.0,
      "learning_rate": 9.42598904106029e-08,
      "loss": 4.4566,
      "num_input_tokens_seen": 52702464,
      "step": 38895
    },
    {
      "epoch": 1.2451187504001024,
      "grad_norm": 24.125,
      "learning_rate": 9.341008737806245e-08,
      "loss": 3.961,
      "num_input_tokens_seen": 52709232,
      "step": 38900
    },
    {
      "epoch": 1.245278791370591,
      "grad_norm": 25.125,
      "learning_rate": 9.256412525467661e-08,
      "loss": 5.1734,
      "num_input_tokens_seen": 52715904,
      "step": 38905
    },
    {
      "epoch": 1.2454388323410792,
      "grad_norm": 17.875,
      "learning_rate": 9.172200417091326e-08,
      "loss": 4.6847,
      "num_input_tokens_seen": 52723024,
      "step": 38910
    },
    {
      "epoch": 1.2455988733115677,
      "grad_norm": 21.875,
      "learning_rate": 9.088372425663239e-08,
      "loss": 3.6955,
      "num_input_tokens_seen": 52729616,
      "step": 38915
    },
    {
      "epoch": 1.2457589142820562,
      "grad_norm": 22.5,
      "learning_rate": 9.004928564110837e-08,
      "loss": 3.6031,
      "num_input_tokens_seen": 52736400,
      "step": 38920
    },
    {
      "epoch": 1.2459189552525447,
      "grad_norm": 30.375,
      "learning_rate": 8.92186884530244e-08,
      "loss": 3.4737,
      "num_input_tokens_seen": 52743184,
      "step": 38925
    },
    {
      "epoch": 1.2460789962230332,
      "grad_norm": 19.25,
      "learning_rate": 8.83919328204641e-08,
      "loss": 4.6499,
      "num_input_tokens_seen": 52750112,
      "step": 38930
    },
    {
      "epoch": 1.2462390371935215,
      "grad_norm": 39.75,
      "learning_rate": 8.756901887093105e-08,
      "loss": 5.4508,
      "num_input_tokens_seen": 52756992,
      "step": 38935
    },
    {
      "epoch": 1.24639907816401,
      "grad_norm": 21.625,
      "learning_rate": 8.674994673132098e-08,
      "loss": 4.3683,
      "num_input_tokens_seen": 52763600,
      "step": 38940
    },
    {
      "epoch": 1.2465591191344985,
      "grad_norm": 30.125,
      "learning_rate": 8.593471652794949e-08,
      "loss": 4.2646,
      "num_input_tokens_seen": 52770336,
      "step": 38945
    },
    {
      "epoch": 1.246719160104987,
      "grad_norm": 22.25,
      "learning_rate": 8.512332838653548e-08,
      "loss": 4.311,
      "num_input_tokens_seen": 52777552,
      "step": 38950
    },
    {
      "epoch": 1.2468792010754752,
      "grad_norm": 41.0,
      "learning_rate": 8.431578243220106e-08,
      "loss": 5.0567,
      "num_input_tokens_seen": 52784064,
      "step": 38955
    },
    {
      "epoch": 1.2470392420459637,
      "grad_norm": 20.375,
      "learning_rate": 8.351207878948552e-08,
      "loss": 4.776,
      "num_input_tokens_seen": 52791008,
      "step": 38960
    },
    {
      "epoch": 1.2471992830164522,
      "grad_norm": 23.25,
      "learning_rate": 8.271221758232583e-08,
      "loss": 4.3092,
      "num_input_tokens_seen": 52798048,
      "step": 38965
    },
    {
      "epoch": 1.2473593239869407,
      "grad_norm": 31.5,
      "learning_rate": 8.191619893407332e-08,
      "loss": 4.8299,
      "num_input_tokens_seen": 52805008,
      "step": 38970
    },
    {
      "epoch": 1.2475193649574292,
      "grad_norm": 25.25,
      "learning_rate": 8.112402296748534e-08,
      "loss": 4.6812,
      "num_input_tokens_seen": 52811840,
      "step": 38975
    },
    {
      "epoch": 1.2476794059279175,
      "grad_norm": 22.875,
      "learning_rate": 8.033568980471973e-08,
      "loss": 4.0659,
      "num_input_tokens_seen": 52818800,
      "step": 38980
    },
    {
      "epoch": 1.247839446898406,
      "grad_norm": 21.125,
      "learning_rate": 7.955119956735146e-08,
      "loss": 3.7201,
      "num_input_tokens_seen": 52825584,
      "step": 38985
    },
    {
      "epoch": 1.2479994878688945,
      "grad_norm": 21.25,
      "learning_rate": 7.877055237636155e-08,
      "loss": 4.1058,
      "num_input_tokens_seen": 52832656,
      "step": 38990
    },
    {
      "epoch": 1.248159528839383,
      "grad_norm": 24.125,
      "learning_rate": 7.79937483521287e-08,
      "loss": 4.7921,
      "num_input_tokens_seen": 52839104,
      "step": 38995
    },
    {
      "epoch": 1.2483195698098712,
      "grad_norm": 17.0,
      "learning_rate": 7.722078761444873e-08,
      "loss": 4.2167,
      "num_input_tokens_seen": 52846048,
      "step": 39000
    },
    {
      "epoch": 1.2483195698098712,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.937,
      "eval_samples_per_second": 28.576,
      "eval_steps_per_second": 14.288,
      "num_input_tokens_seen": 52846048,
      "step": 39000
    },
    {
      "epoch": 1.2484796107803597,
      "grad_norm": 37.0,
      "learning_rate": 7.645167028252631e-08,
      "loss": 4.039,
      "num_input_tokens_seen": 52853200,
      "step": 39005
    },
    {
      "epoch": 1.2486396517508482,
      "grad_norm": 27.0,
      "learning_rate": 7.568639647496379e-08,
      "loss": 3.917,
      "num_input_tokens_seen": 52860000,
      "step": 39010
    },
    {
      "epoch": 1.2487996927213367,
      "grad_norm": 25.25,
      "learning_rate": 7.492496630977508e-08,
      "loss": 4.2875,
      "num_input_tokens_seen": 52866976,
      "step": 39015
    },
    {
      "epoch": 1.2489597336918252,
      "grad_norm": 25.5,
      "learning_rate": 7.416737990438571e-08,
      "loss": 4.1906,
      "num_input_tokens_seen": 52873968,
      "step": 39020
    },
    {
      "epoch": 1.2491197746623135,
      "grad_norm": 40.0,
      "learning_rate": 7.341363737562445e-08,
      "loss": 4.4296,
      "num_input_tokens_seen": 52881184,
      "step": 39025
    },
    {
      "epoch": 1.249279815632802,
      "grad_norm": 19.125,
      "learning_rate": 7.266373883972887e-08,
      "loss": 5.3886,
      "num_input_tokens_seen": 52888304,
      "step": 39030
    },
    {
      "epoch": 1.2494398566032905,
      "grad_norm": 21.25,
      "learning_rate": 7.191768441233981e-08,
      "loss": 3.8641,
      "num_input_tokens_seen": 52895504,
      "step": 39035
    },
    {
      "epoch": 1.249599897573779,
      "grad_norm": 35.75,
      "learning_rate": 7.11754742085069e-08,
      "loss": 4.5124,
      "num_input_tokens_seen": 52902192,
      "step": 39040
    },
    {
      "epoch": 1.2497599385442673,
      "grad_norm": 38.75,
      "learning_rate": 7.043710834269413e-08,
      "loss": 5.2931,
      "num_input_tokens_seen": 52908720,
      "step": 39045
    },
    {
      "epoch": 1.2499199795147558,
      "grad_norm": 24.25,
      "learning_rate": 6.970258692876319e-08,
      "loss": 4.6903,
      "num_input_tokens_seen": 52915744,
      "step": 39050
    },
    {
      "epoch": 1.2500800204852442,
      "grad_norm": 16.0,
      "learning_rate": 6.897191007998738e-08,
      "loss": 4.202,
      "num_input_tokens_seen": 52922432,
      "step": 39055
    },
    {
      "epoch": 1.2502400614557327,
      "grad_norm": 23.25,
      "learning_rate": 6.824507790904599e-08,
      "loss": 3.8934,
      "num_input_tokens_seen": 52929584,
      "step": 39060
    },
    {
      "epoch": 1.2504001024262212,
      "grad_norm": 31.75,
      "learning_rate": 6.752209052802439e-08,
      "loss": 3.8644,
      "num_input_tokens_seen": 52936480,
      "step": 39065
    },
    {
      "epoch": 1.2505601433967095,
      "grad_norm": 20.5,
      "learning_rate": 6.680294804841946e-08,
      "loss": 4.9768,
      "num_input_tokens_seen": 52943600,
      "step": 39070
    },
    {
      "epoch": 1.250720184367198,
      "grad_norm": 27.375,
      "learning_rate": 6.608765058112865e-08,
      "loss": 4.3787,
      "num_input_tokens_seen": 52950208,
      "step": 39075
    },
    {
      "epoch": 1.2508802253376865,
      "grad_norm": 35.0,
      "learning_rate": 6.537619823646368e-08,
      "loss": 4.4566,
      "num_input_tokens_seen": 52957376,
      "step": 39080
    },
    {
      "epoch": 1.2510402663081748,
      "grad_norm": 25.5,
      "learning_rate": 6.466859112413404e-08,
      "loss": 3.979,
      "num_input_tokens_seen": 52964384,
      "step": 39085
    },
    {
      "epoch": 1.2512003072786633,
      "grad_norm": 27.0,
      "learning_rate": 6.39648293532663e-08,
      "loss": 4.2295,
      "num_input_tokens_seen": 52970736,
      "step": 39090
    },
    {
      "epoch": 1.2513603482491518,
      "grad_norm": 19.375,
      "learning_rate": 6.32649130323848e-08,
      "loss": 3.7725,
      "num_input_tokens_seen": 52977696,
      "step": 39095
    },
    {
      "epoch": 1.2515203892196403,
      "grad_norm": 38.5,
      "learning_rate": 6.256884226943094e-08,
      "loss": 4.789,
      "num_input_tokens_seen": 52984080,
      "step": 39100
    },
    {
      "epoch": 1.2516804301901288,
      "grad_norm": 31.75,
      "learning_rate": 6.187661717174386e-08,
      "loss": 4.6272,
      "num_input_tokens_seen": 52990864,
      "step": 39105
    },
    {
      "epoch": 1.2518404711606173,
      "grad_norm": 22.25,
      "learning_rate": 6.118823784607708e-08,
      "loss": 3.61,
      "num_input_tokens_seen": 52997616,
      "step": 39110
    },
    {
      "epoch": 1.2520005121311055,
      "grad_norm": 25.875,
      "learning_rate": 6.050370439858178e-08,
      "loss": 3.8933,
      "num_input_tokens_seen": 53004672,
      "step": 39115
    },
    {
      "epoch": 1.252160553101594,
      "grad_norm": 35.0,
      "learning_rate": 5.98230169348235e-08,
      "loss": 4.5864,
      "num_input_tokens_seen": 53011328,
      "step": 39120
    },
    {
      "epoch": 1.2523205940720825,
      "grad_norm": 33.0,
      "learning_rate": 5.914617555977664e-08,
      "loss": 3.2907,
      "num_input_tokens_seen": 53018096,
      "step": 39125
    },
    {
      "epoch": 1.2524806350425708,
      "grad_norm": 29.125,
      "learning_rate": 5.8473180377816017e-08,
      "loss": 4.3071,
      "num_input_tokens_seen": 53024816,
      "step": 39130
    },
    {
      "epoch": 1.2526406760130593,
      "grad_norm": 33.5,
      "learning_rate": 5.780403149272251e-08,
      "loss": 5.0735,
      "num_input_tokens_seen": 53031488,
      "step": 39135
    },
    {
      "epoch": 1.2528007169835478,
      "grad_norm": 16.25,
      "learning_rate": 5.7138729007694126e-08,
      "loss": 4.1689,
      "num_input_tokens_seen": 53038448,
      "step": 39140
    },
    {
      "epoch": 1.2529607579540363,
      "grad_norm": 30.75,
      "learning_rate": 5.64772730253238e-08,
      "loss": 4.8078,
      "num_input_tokens_seen": 53045424,
      "step": 39145
    },
    {
      "epoch": 1.2531207989245248,
      "grad_norm": 48.25,
      "learning_rate": 5.5819663647618814e-08,
      "loss": 4.3825,
      "num_input_tokens_seen": 53052160,
      "step": 39150
    },
    {
      "epoch": 1.253280839895013,
      "grad_norm": 31.125,
      "learning_rate": 5.5165900975989723e-08,
      "loss": 4.3317,
      "num_input_tokens_seen": 53059312,
      "step": 39155
    },
    {
      "epoch": 1.2534408808655015,
      "grad_norm": 20.75,
      "learning_rate": 5.451598511125311e-08,
      "loss": 4.5228,
      "num_input_tokens_seen": 53065872,
      "step": 39160
    },
    {
      "epoch": 1.25360092183599,
      "grad_norm": 47.75,
      "learning_rate": 5.3869916153637124e-08,
      "loss": 4.3067,
      "num_input_tokens_seen": 53072672,
      "step": 39165
    },
    {
      "epoch": 1.2537609628064785,
      "grad_norm": 35.75,
      "learning_rate": 5.322769420277318e-08,
      "loss": 5.031,
      "num_input_tokens_seen": 53079632,
      "step": 39170
    },
    {
      "epoch": 1.2539210037769668,
      "grad_norm": 34.75,
      "learning_rate": 5.258931935769873e-08,
      "loss": 5.7991,
      "num_input_tokens_seen": 53086112,
      "step": 39175
    },
    {
      "epoch": 1.2540810447474553,
      "grad_norm": 32.75,
      "learning_rate": 5.19547917168628e-08,
      "loss": 5.2999,
      "num_input_tokens_seen": 53092544,
      "step": 39180
    },
    {
      "epoch": 1.2542410857179438,
      "grad_norm": 15.1875,
      "learning_rate": 5.13241113781121e-08,
      "loss": 4.7889,
      "num_input_tokens_seen": 53098768,
      "step": 39185
    },
    {
      "epoch": 1.2544011266884323,
      "grad_norm": 21.125,
      "learning_rate": 5.0697278438707755e-08,
      "loss": 4.5244,
      "num_input_tokens_seen": 53105520,
      "step": 39190
    },
    {
      "epoch": 1.2545611676589208,
      "grad_norm": 19.875,
      "learning_rate": 5.0074292995316854e-08,
      "loss": 4.8165,
      "num_input_tokens_seen": 53112032,
      "step": 39195
    },
    {
      "epoch": 1.254721208629409,
      "grad_norm": 30.875,
      "learning_rate": 4.945515514400978e-08,
      "loss": 4.1054,
      "num_input_tokens_seen": 53118576,
      "step": 39200
    },
    {
      "epoch": 1.254721208629409,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.8471,
      "eval_samples_per_second": 28.581,
      "eval_steps_per_second": 14.291,
      "num_input_tokens_seen": 53118576,
      "step": 39200
    },
    {
      "epoch": 1.2548812495998976,
      "grad_norm": 56.0,
      "learning_rate": 4.883986498026571e-08,
      "loss": 5.1117,
      "num_input_tokens_seen": 53125136,
      "step": 39205
    },
    {
      "epoch": 1.255041290570386,
      "grad_norm": 31.125,
      "learning_rate": 4.822842259896987e-08,
      "loss": 4.9617,
      "num_input_tokens_seen": 53131696,
      "step": 39210
    },
    {
      "epoch": 1.2552013315408745,
      "grad_norm": 50.0,
      "learning_rate": 4.762082809441626e-08,
      "loss": 4.2788,
      "num_input_tokens_seen": 53138496,
      "step": 39215
    },
    {
      "epoch": 1.2553613725113628,
      "grad_norm": 22.0,
      "learning_rate": 4.7017081560302156e-08,
      "loss": 4.6173,
      "num_input_tokens_seen": 53145024,
      "step": 39220
    },
    {
      "epoch": 1.2555214134818513,
      "grad_norm": 29.5,
      "learning_rate": 4.6417183089730866e-08,
      "loss": 4.2835,
      "num_input_tokens_seen": 53151632,
      "step": 39225
    },
    {
      "epoch": 1.2556814544523398,
      "grad_norm": 36.5,
      "learning_rate": 4.5821132775217265e-08,
      "loss": 4.3373,
      "num_input_tokens_seen": 53158320,
      "step": 39230
    },
    {
      "epoch": 1.2558414954228283,
      "grad_norm": 26.25,
      "learning_rate": 4.5228930708679504e-08,
      "loss": 4.5941,
      "num_input_tokens_seen": 53165088,
      "step": 39235
    },
    {
      "epoch": 1.2560015363933168,
      "grad_norm": 29.25,
      "learning_rate": 4.464057698144175e-08,
      "loss": 4.5332,
      "num_input_tokens_seen": 53171360,
      "step": 39240
    },
    {
      "epoch": 1.256161577363805,
      "grad_norm": 20.5,
      "learning_rate": 4.4056071684236974e-08,
      "loss": 3.7002,
      "num_input_tokens_seen": 53178304,
      "step": 39245
    },
    {
      "epoch": 1.2563216183342936,
      "grad_norm": 31.125,
      "learning_rate": 4.347541490719864e-08,
      "loss": 4.2645,
      "num_input_tokens_seen": 53184912,
      "step": 39250
    },
    {
      "epoch": 1.256481659304782,
      "grad_norm": 23.125,
      "learning_rate": 4.2898606739877336e-08,
      "loss": 4.0853,
      "num_input_tokens_seen": 53191952,
      "step": 39255
    },
    {
      "epoch": 1.2566417002752703,
      "grad_norm": 35.5,
      "learning_rate": 4.232564727122135e-08,
      "loss": 4.5213,
      "num_input_tokens_seen": 53198848,
      "step": 39260
    },
    {
      "epoch": 1.2568017412457588,
      "grad_norm": 32.5,
      "learning_rate": 4.1756536589585004e-08,
      "loss": 4.0932,
      "num_input_tokens_seen": 53205728,
      "step": 39265
    },
    {
      "epoch": 1.2569617822162473,
      "grad_norm": 22.25,
      "learning_rate": 4.119127478273976e-08,
      "loss": 4.3377,
      "num_input_tokens_seen": 53212544,
      "step": 39270
    },
    {
      "epoch": 1.2571218231867358,
      "grad_norm": 29.125,
      "learning_rate": 4.062986193784923e-08,
      "loss": 4.6051,
      "num_input_tokens_seen": 53219296,
      "step": 39275
    },
    {
      "epoch": 1.2572818641572243,
      "grad_norm": 30.875,
      "learning_rate": 4.007229814149416e-08,
      "loss": 5.2153,
      "num_input_tokens_seen": 53226016,
      "step": 39280
    },
    {
      "epoch": 1.2574419051277128,
      "grad_norm": 32.25,
      "learning_rate": 3.951858347965576e-08,
      "loss": 4.4457,
      "num_input_tokens_seen": 53232848,
      "step": 39285
    },
    {
      "epoch": 1.257601946098201,
      "grad_norm": 24.5,
      "learning_rate": 3.896871803772684e-08,
      "loss": 3.7601,
      "num_input_tokens_seen": 53239472,
      "step": 39290
    },
    {
      "epoch": 1.2577619870686896,
      "grad_norm": 19.5,
      "learning_rate": 3.842270190050068e-08,
      "loss": 3.6477,
      "num_input_tokens_seen": 53246592,
      "step": 39295
    },
    {
      "epoch": 1.257922028039178,
      "grad_norm": 37.0,
      "learning_rate": 3.7880535152179376e-08,
      "loss": 4.6784,
      "num_input_tokens_seen": 53253376,
      "step": 39300
    },
    {
      "epoch": 1.2580820690096663,
      "grad_norm": 38.5,
      "learning_rate": 3.734221787637382e-08,
      "loss": 3.824,
      "num_input_tokens_seen": 53260000,
      "step": 39305
    },
    {
      "epoch": 1.2582421099801548,
      "grad_norm": 22.875,
      "learning_rate": 3.680775015609817e-08,
      "loss": 4.0576,
      "num_input_tokens_seen": 53266752,
      "step": 39310
    },
    {
      "epoch": 1.2584021509506433,
      "grad_norm": 36.75,
      "learning_rate": 3.627713207377537e-08,
      "loss": 4.2325,
      "num_input_tokens_seen": 53273712,
      "step": 39315
    },
    {
      "epoch": 1.2585621919211318,
      "grad_norm": 27.875,
      "learning_rate": 3.575036371123164e-08,
      "loss": 4.6856,
      "num_input_tokens_seen": 53280208,
      "step": 39320
    },
    {
      "epoch": 1.2587222328916203,
      "grad_norm": 39.5,
      "learning_rate": 3.5227445149704776e-08,
      "loss": 4.3343,
      "num_input_tokens_seen": 53287520,
      "step": 39325
    },
    {
      "epoch": 1.2588822738621088,
      "grad_norm": 28.0,
      "learning_rate": 3.470837646983027e-08,
      "loss": 5.091,
      "num_input_tokens_seen": 53294160,
      "step": 39330
    },
    {
      "epoch": 1.259042314832597,
      "grad_norm": 25.125,
      "learning_rate": 3.419315775165799e-08,
      "loss": 4.0073,
      "num_input_tokens_seen": 53301008,
      "step": 39335
    },
    {
      "epoch": 1.2592023558030856,
      "grad_norm": 32.0,
      "learning_rate": 3.368178907464103e-08,
      "loss": 4.3363,
      "num_input_tokens_seen": 53307664,
      "step": 39340
    },
    {
      "epoch": 1.259362396773574,
      "grad_norm": 28.75,
      "learning_rate": 3.317427051763855e-08,
      "loss": 4.4137,
      "num_input_tokens_seen": 53314512,
      "step": 39345
    },
    {
      "epoch": 1.2595224377440624,
      "grad_norm": 43.5,
      "learning_rate": 3.267060215891571e-08,
      "loss": 3.4823,
      "num_input_tokens_seen": 53321008,
      "step": 39350
    },
    {
      "epoch": 1.2596824787145509,
      "grad_norm": 21.5,
      "learning_rate": 3.217078407614649e-08,
      "loss": 3.0929,
      "num_input_tokens_seen": 53327520,
      "step": 39355
    },
    {
      "epoch": 1.2598425196850394,
      "grad_norm": 23.125,
      "learning_rate": 3.1674816346405345e-08,
      "loss": 5.1883,
      "num_input_tokens_seen": 53334272,
      "step": 39360
    },
    {
      "epoch": 1.2600025606555278,
      "grad_norm": 28.375,
      "learning_rate": 3.11826990461811e-08,
      "loss": 3.9269,
      "num_input_tokens_seen": 53340800,
      "step": 39365
    },
    {
      "epoch": 1.2601626016260163,
      "grad_norm": 32.25,
      "learning_rate": 3.069443225136304e-08,
      "loss": 3.9297,
      "num_input_tokens_seen": 53347328,
      "step": 39370
    },
    {
      "epoch": 1.2603226425965048,
      "grad_norm": 32.5,
      "learning_rate": 3.021001603724372e-08,
      "loss": 5.1067,
      "num_input_tokens_seen": 53353920,
      "step": 39375
    },
    {
      "epoch": 1.2604826835669931,
      "grad_norm": 13.5,
      "learning_rate": 2.9729450478532818e-08,
      "loss": 4.4417,
      "num_input_tokens_seen": 53360848,
      "step": 39380
    },
    {
      "epoch": 1.2606427245374816,
      "grad_norm": 134.0,
      "learning_rate": 2.9252735649337726e-08,
      "loss": 4.3825,
      "num_input_tokens_seen": 53367328,
      "step": 39385
    },
    {
      "epoch": 1.26080276550797,
      "grad_norm": 21.625,
      "learning_rate": 2.8779871623171863e-08,
      "loss": 4.8634,
      "num_input_tokens_seen": 53374256,
      "step": 39390
    },
    {
      "epoch": 1.2609628064784584,
      "grad_norm": 37.0,
      "learning_rate": 2.8310858472957448e-08,
      "loss": 4.4711,
      "num_input_tokens_seen": 53381072,
      "step": 39395
    },
    {
      "epoch": 1.2611228474489469,
      "grad_norm": 21.75,
      "learning_rate": 2.784569627101996e-08,
      "loss": 3.7602,
      "num_input_tokens_seen": 53387872,
      "step": 39400
    },
    {
      "epoch": 1.2611228474489469,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.9221,
      "eval_samples_per_second": 28.577,
      "eval_steps_per_second": 14.288,
      "num_input_tokens_seen": 53387872,
      "step": 39400
    },
    {
      "epoch": 1.2612828884194354,
      "grad_norm": 22.125,
      "learning_rate": 2.738438508909924e-08,
      "loss": 4.9825,
      "num_input_tokens_seen": 53394400,
      "step": 39405
    },
    {
      "epoch": 1.2614429293899239,
      "grad_norm": 30.875,
      "learning_rate": 2.692692499833005e-08,
      "loss": 5.1585,
      "num_input_tokens_seen": 53401888,
      "step": 39410
    },
    {
      "epoch": 1.2616029703604124,
      "grad_norm": 34.5,
      "learning_rate": 2.647331606926151e-08,
      "loss": 3.9911,
      "num_input_tokens_seen": 53408848,
      "step": 39415
    },
    {
      "epoch": 1.2617630113309006,
      "grad_norm": 26.375,
      "learning_rate": 2.6023558371843225e-08,
      "loss": 4.2258,
      "num_input_tokens_seen": 53415632,
      "step": 39420
    },
    {
      "epoch": 1.2619230523013891,
      "grad_norm": 21.25,
      "learning_rate": 2.557765197543638e-08,
      "loss": 4.6292,
      "num_input_tokens_seen": 53422368,
      "step": 39425
    },
    {
      "epoch": 1.2620830932718776,
      "grad_norm": 26.625,
      "learning_rate": 2.513559694880263e-08,
      "loss": 3.7945,
      "num_input_tokens_seen": 53428848,
      "step": 39430
    },
    {
      "epoch": 1.2622431342423661,
      "grad_norm": 26.25,
      "learning_rate": 2.469739336011523e-08,
      "loss": 3.583,
      "num_input_tokens_seen": 53435552,
      "step": 39435
    },
    {
      "epoch": 1.2624031752128544,
      "grad_norm": 38.25,
      "learning_rate": 2.4263041276947894e-08,
      "loss": 4.1208,
      "num_input_tokens_seen": 53442032,
      "step": 39440
    },
    {
      "epoch": 1.2625632161833429,
      "grad_norm": 33.0,
      "learning_rate": 2.3832540766283164e-08,
      "loss": 4.5726,
      "num_input_tokens_seen": 53448720,
      "step": 39445
    },
    {
      "epoch": 1.2627232571538314,
      "grad_norm": 25.5,
      "learning_rate": 2.3405891894512366e-08,
      "loss": 3.8476,
      "num_input_tokens_seen": 53455408,
      "step": 39450
    },
    {
      "epoch": 1.2628832981243199,
      "grad_norm": 33.25,
      "learning_rate": 2.29830947274301e-08,
      "loss": 4.4394,
      "num_input_tokens_seen": 53462528,
      "step": 39455
    },
    {
      "epoch": 1.2630433390948084,
      "grad_norm": 19.875,
      "learning_rate": 2.2564149330231432e-08,
      "loss": 4.3878,
      "num_input_tokens_seen": 53469184,
      "step": 39460
    },
    {
      "epoch": 1.2632033800652966,
      "grad_norm": 38.25,
      "learning_rate": 2.2149055767528572e-08,
      "loss": 4.2975,
      "num_input_tokens_seen": 53475568,
      "step": 39465
    },
    {
      "epoch": 1.2633634210357851,
      "grad_norm": 29.125,
      "learning_rate": 2.1737814103334197e-08,
      "loss": 5.0666,
      "num_input_tokens_seen": 53482416,
      "step": 39470
    },
    {
      "epoch": 1.2635234620062736,
      "grad_norm": 24.5,
      "learning_rate": 2.1330424401064253e-08,
      "loss": 3.9342,
      "num_input_tokens_seen": 53489088,
      "step": 39475
    },
    {
      "epoch": 1.2636835029767621,
      "grad_norm": 50.0,
      "learning_rate": 2.092688672354348e-08,
      "loss": 5.3275,
      "num_input_tokens_seen": 53495760,
      "step": 39480
    },
    {
      "epoch": 1.2638435439472504,
      "grad_norm": 49.5,
      "learning_rate": 2.0527201133005435e-08,
      "loss": 4.5419,
      "num_input_tokens_seen": 53502752,
      "step": 39485
    },
    {
      "epoch": 1.264003584917739,
      "grad_norm": 35.25,
      "learning_rate": 2.0131367691084148e-08,
      "loss": 5.0827,
      "num_input_tokens_seen": 53509584,
      "step": 39490
    },
    {
      "epoch": 1.2641636258882274,
      "grad_norm": 35.5,
      "learning_rate": 1.9739386458819675e-08,
      "loss": 4.1054,
      "num_input_tokens_seen": 53516544,
      "step": 39495
    },
    {
      "epoch": 1.2643236668587159,
      "grad_norm": 33.25,
      "learning_rate": 1.9351257496666442e-08,
      "loss": 3.9213,
      "num_input_tokens_seen": 53524000,
      "step": 39500
    },
    {
      "epoch": 1.2644837078292044,
      "grad_norm": 36.5,
      "learning_rate": 1.896698086447657e-08,
      "loss": 3.8321,
      "num_input_tokens_seen": 53530976,
      "step": 39505
    },
    {
      "epoch": 1.2646437487996927,
      "grad_norm": 42.5,
      "learning_rate": 1.8586556621505436e-08,
      "loss": 4.489,
      "num_input_tokens_seen": 53537600,
      "step": 39510
    },
    {
      "epoch": 1.2648037897701812,
      "grad_norm": 18.0,
      "learning_rate": 1.820998482642833e-08,
      "loss": 4.0119,
      "num_input_tokens_seen": 53544032,
      "step": 39515
    },
    {
      "epoch": 1.2649638307406696,
      "grad_norm": 38.75,
      "learning_rate": 1.7837265537309912e-08,
      "loss": 4.3097,
      "num_input_tokens_seen": 53551104,
      "step": 39520
    },
    {
      "epoch": 1.265123871711158,
      "grad_norm": 29.0,
      "learning_rate": 1.7468398811629206e-08,
      "loss": 4.581,
      "num_input_tokens_seen": 53557840,
      "step": 39525
    },
    {
      "epoch": 1.2652839126816464,
      "grad_norm": 17.625,
      "learning_rate": 1.710338470627404e-08,
      "loss": 4.2163,
      "num_input_tokens_seen": 53564448,
      "step": 39530
    },
    {
      "epoch": 1.265443953652135,
      "grad_norm": 34.75,
      "learning_rate": 1.6742223277529945e-08,
      "loss": 5.315,
      "num_input_tokens_seen": 53570736,
      "step": 39535
    },
    {
      "epoch": 1.2656039946226234,
      "grad_norm": 24.375,
      "learning_rate": 1.6384914581094036e-08,
      "loss": 3.9707,
      "num_input_tokens_seen": 53577792,
      "step": 39540
    },
    {
      "epoch": 1.265764035593112,
      "grad_norm": 61.5,
      "learning_rate": 1.6031458672069455e-08,
      "loss": 5.0353,
      "num_input_tokens_seen": 53584688,
      "step": 39545
    },
    {
      "epoch": 1.2659240765636004,
      "grad_norm": 38.0,
      "learning_rate": 1.5681855604962602e-08,
      "loss": 4.2603,
      "num_input_tokens_seen": 53591264,
      "step": 39550
    },
    {
      "epoch": 1.2660841175340887,
      "grad_norm": 28.625,
      "learning_rate": 1.5336105433683135e-08,
      "loss": 4.4,
      "num_input_tokens_seen": 53598288,
      "step": 39555
    },
    {
      "epoch": 1.2662441585045772,
      "grad_norm": 27.125,
      "learning_rate": 1.499420821155506e-08,
      "loss": 4.535,
      "num_input_tokens_seen": 53605312,
      "step": 39560
    },
    {
      "epoch": 1.2664041994750657,
      "grad_norm": 20.0,
      "learning_rate": 1.4656163991302874e-08,
      "loss": 3.3366,
      "num_input_tokens_seen": 53612032,
      "step": 39565
    },
    {
      "epoch": 1.266564240445554,
      "grad_norm": 24.75,
      "learning_rate": 1.4321972825051544e-08,
      "loss": 3.8122,
      "num_input_tokens_seen": 53619216,
      "step": 39570
    },
    {
      "epoch": 1.2667242814160424,
      "grad_norm": 26.0,
      "learning_rate": 1.3991634764345951e-08,
      "loss": 3.9841,
      "num_input_tokens_seen": 53625888,
      "step": 39575
    },
    {
      "epoch": 1.266884322386531,
      "grad_norm": 46.25,
      "learning_rate": 1.3665149860120352e-08,
      "loss": 4.898,
      "num_input_tokens_seen": 53633136,
      "step": 39580
    },
    {
      "epoch": 1.2670443633570194,
      "grad_norm": 40.5,
      "learning_rate": 1.3342518162728912e-08,
      "loss": 3.5526,
      "num_input_tokens_seen": 53639776,
      "step": 39585
    },
    {
      "epoch": 1.267204404327508,
      "grad_norm": 33.5,
      "learning_rate": 1.30237397219235e-08,
      "loss": 3.3673,
      "num_input_tokens_seen": 53646528,
      "step": 39590
    },
    {
      "epoch": 1.2673644452979964,
      "grad_norm": 26.125,
      "learning_rate": 1.2708814586862016e-08,
      "loss": 4.415,
      "num_input_tokens_seen": 53653008,
      "step": 39595
    },
    {
      "epoch": 1.2675244862684847,
      "grad_norm": 18.625,
      "learning_rate": 1.2397742806111168e-08,
      "loss": 3.4438,
      "num_input_tokens_seen": 53659856,
      "step": 39600
    },
    {
      "epoch": 1.2675244862684847,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.9745,
      "eval_samples_per_second": 28.574,
      "eval_steps_per_second": 14.287,
      "num_input_tokens_seen": 53659856,
      "step": 39600
    },
    {
      "epoch": 1.2676845272389732,
      "grad_norm": 23.0,
      "learning_rate": 1.209052442764369e-08,
      "loss": 3.9122,
      "num_input_tokens_seen": 53666864,
      "step": 39605
    },
    {
      "epoch": 1.2678445682094617,
      "grad_norm": 16.5,
      "learning_rate": 1.17871594988328e-08,
      "loss": 4.1206,
      "num_input_tokens_seen": 53673408,
      "step": 39610
    },
    {
      "epoch": 1.26800460917995,
      "grad_norm": 21.875,
      "learning_rate": 1.1487648066466072e-08,
      "loss": 4.4076,
      "num_input_tokens_seen": 53679936,
      "step": 39615
    },
    {
      "epoch": 1.2681646501504384,
      "grad_norm": 27.75,
      "learning_rate": 1.1191990176728784e-08,
      "loss": 4.9926,
      "num_input_tokens_seen": 53686512,
      "step": 39620
    },
    {
      "epoch": 1.268324691120927,
      "grad_norm": 20.625,
      "learning_rate": 1.0900185875215018e-08,
      "loss": 4.108,
      "num_input_tokens_seen": 53693072,
      "step": 39625
    },
    {
      "epoch": 1.2684847320914154,
      "grad_norm": 18.375,
      "learning_rate": 1.0612235206924891e-08,
      "loss": 3.5686,
      "num_input_tokens_seen": 53699664,
      "step": 39630
    },
    {
      "epoch": 1.268644773061904,
      "grad_norm": 30.75,
      "learning_rate": 1.0328138216264549e-08,
      "loss": 3.9492,
      "num_input_tokens_seen": 53706320,
      "step": 39635
    },
    {
      "epoch": 1.2688048140323924,
      "grad_norm": 26.75,
      "learning_rate": 1.004789494704339e-08,
      "loss": 4.9192,
      "num_input_tokens_seen": 53713072,
      "step": 39640
    },
    {
      "epoch": 1.2689648550028807,
      "grad_norm": 33.0,
      "learning_rate": 9.771505442482397e-09,
      "loss": 5.0073,
      "num_input_tokens_seen": 53719504,
      "step": 39645
    },
    {
      "epoch": 1.2691248959733692,
      "grad_norm": 24.25,
      "learning_rate": 9.498969745200259e-09,
      "loss": 5.024,
      "num_input_tokens_seen": 53726752,
      "step": 39650
    },
    {
      "epoch": 1.2692849369438577,
      "grad_norm": 25.125,
      "learning_rate": 9.230287897230017e-09,
      "loss": 4.4155,
      "num_input_tokens_seen": 53733392,
      "step": 39655
    },
    {
      "epoch": 1.269444977914346,
      "grad_norm": 33.0,
      "learning_rate": 8.965459940002419e-09,
      "loss": 4.7688,
      "num_input_tokens_seen": 53739984,
      "step": 39660
    },
    {
      "epoch": 1.2696050188848345,
      "grad_norm": 42.25,
      "learning_rate": 8.704485914357019e-09,
      "loss": 4.7639,
      "num_input_tokens_seen": 53746656,
      "step": 39665
    },
    {
      "epoch": 1.269765059855323,
      "grad_norm": 20.375,
      "learning_rate": 8.447365860539402e-09,
      "loss": 3.9814,
      "num_input_tokens_seen": 53753648,
      "step": 39670
    },
    {
      "epoch": 1.2699251008258114,
      "grad_norm": 47.75,
      "learning_rate": 8.194099818201184e-09,
      "loss": 4.4671,
      "num_input_tokens_seen": 53760592,
      "step": 39675
    },
    {
      "epoch": 1.2700851417963,
      "grad_norm": 57.5,
      "learning_rate": 7.944687826400011e-09,
      "loss": 4.9851,
      "num_input_tokens_seen": 53767264,
      "step": 39680
    },
    {
      "epoch": 1.2702451827667882,
      "grad_norm": 38.25,
      "learning_rate": 7.699129923599557e-09,
      "loss": 5.935,
      "num_input_tokens_seen": 53774224,
      "step": 39685
    },
    {
      "epoch": 1.2704052237372767,
      "grad_norm": 26.125,
      "learning_rate": 7.457426147663982e-09,
      "loss": 4.3385,
      "num_input_tokens_seen": 53781088,
      "step": 39690
    },
    {
      "epoch": 1.2705652647077652,
      "grad_norm": 22.75,
      "learning_rate": 7.219576535871797e-09,
      "loss": 4.7791,
      "num_input_tokens_seen": 53787520,
      "step": 39695
    },
    {
      "epoch": 1.2707253056782537,
      "grad_norm": 36.25,
      "learning_rate": 6.985581124896445e-09,
      "loss": 4.4866,
      "num_input_tokens_seen": 53793856,
      "step": 39700
    },
    {
      "epoch": 1.270885346648742,
      "grad_norm": 44.75,
      "learning_rate": 6.755439950828501e-09,
      "loss": 4.4506,
      "num_input_tokens_seen": 53800896,
      "step": 39705
    },
    {
      "epoch": 1.2710453876192305,
      "grad_norm": 30.75,
      "learning_rate": 6.5291530491562444e-09,
      "loss": 5.1017,
      "num_input_tokens_seen": 53807840,
      "step": 39710
    },
    {
      "epoch": 1.271205428589719,
      "grad_norm": 27.0,
      "learning_rate": 6.3067204547739845e-09,
      "loss": 4.474,
      "num_input_tokens_seen": 53814128,
      "step": 39715
    },
    {
      "epoch": 1.2713654695602075,
      "grad_norm": 18.125,
      "learning_rate": 6.088142201987612e-09,
      "loss": 4.4599,
      "num_input_tokens_seen": 53821072,
      "step": 39720
    },
    {
      "epoch": 1.271525510530696,
      "grad_norm": 25.625,
      "learning_rate": 5.873418324503499e-09,
      "loss": 4.2431,
      "num_input_tokens_seen": 53827408,
      "step": 39725
    },
    {
      "epoch": 1.2716855515011842,
      "grad_norm": 22.125,
      "learning_rate": 5.6625488554340465e-09,
      "loss": 4.3636,
      "num_input_tokens_seen": 53833664,
      "step": 39730
    },
    {
      "epoch": 1.2718455924716727,
      "grad_norm": 56.5,
      "learning_rate": 5.455533827297688e-09,
      "loss": 4.6568,
      "num_input_tokens_seen": 53840672,
      "step": 39735
    },
    {
      "epoch": 1.2720056334421612,
      "grad_norm": 32.75,
      "learning_rate": 5.252373272018885e-09,
      "loss": 5.6014,
      "num_input_tokens_seen": 53847472,
      "step": 39740
    },
    {
      "epoch": 1.2721656744126497,
      "grad_norm": 33.75,
      "learning_rate": 5.053067220925356e-09,
      "loss": 3.5757,
      "num_input_tokens_seen": 53854208,
      "step": 39745
    },
    {
      "epoch": 1.272325715383138,
      "grad_norm": 23.875,
      "learning_rate": 4.857615704759177e-09,
      "loss": 4.8894,
      "num_input_tokens_seen": 53860848,
      "step": 39750
    },
    {
      "epoch": 1.2724857563536265,
      "grad_norm": 31.875,
      "learning_rate": 4.666018753654577e-09,
      "loss": 3.3481,
      "num_input_tokens_seen": 53867712,
      "step": 39755
    },
    {
      "epoch": 1.272645797324115,
      "grad_norm": 27.75,
      "learning_rate": 4.478276397162917e-09,
      "loss": 4.5211,
      "num_input_tokens_seen": 53874368,
      "step": 39760
    },
    {
      "epoch": 1.2728058382946035,
      "grad_norm": 33.75,
      "learning_rate": 4.294388664233262e-09,
      "loss": 5.2062,
      "num_input_tokens_seen": 53881024,
      "step": 39765
    },
    {
      "epoch": 1.272965879265092,
      "grad_norm": 31.75,
      "learning_rate": 4.114355583223484e-09,
      "loss": 4.9259,
      "num_input_tokens_seen": 53887184,
      "step": 39770
    },
    {
      "epoch": 1.2731259202355802,
      "grad_norm": 25.125,
      "learning_rate": 3.9381771818974845e-09,
      "loss": 4.1545,
      "num_input_tokens_seen": 53893648,
      "step": 39775
    },
    {
      "epoch": 1.2732859612060687,
      "grad_norm": 36.25,
      "learning_rate": 3.765853487427973e-09,
      "loss": 3.68,
      "num_input_tokens_seen": 53900624,
      "step": 39780
    },
    {
      "epoch": 1.2734460021765572,
      "grad_norm": 27.875,
      "learning_rate": 3.5973845263825857e-09,
      "loss": 3.4487,
      "num_input_tokens_seen": 53907744,
      "step": 39785
    },
    {
      "epoch": 1.2736060431470455,
      "grad_norm": 26.125,
      "learning_rate": 3.4327703247488684e-09,
      "loss": 3.972,
      "num_input_tokens_seen": 53914800,
      "step": 39790
    },
    {
      "epoch": 1.273766084117534,
      "grad_norm": 36.0,
      "learning_rate": 3.2720109079037443e-09,
      "loss": 4.952,
      "num_input_tokens_seen": 53921712,
      "step": 39795
    },
    {
      "epoch": 1.2739261250880225,
      "grad_norm": 33.25,
      "learning_rate": 3.1151063006468193e-09,
      "loss": 4.9077,
      "num_input_tokens_seen": 53928784,
      "step": 39800
    },
    {
      "epoch": 1.2739261250880225,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.9227,
      "eval_samples_per_second": 28.577,
      "eval_steps_per_second": 14.288,
      "num_input_tokens_seen": 53928784,
      "step": 39800
    },
    {
      "epoch": 1.274086166058511,
      "grad_norm": 23.75,
      "learning_rate": 2.962056527169854e-09,
      "loss": 5.485,
      "num_input_tokens_seen": 53935552,
      "step": 39805
    },
    {
      "epoch": 1.2742462070289995,
      "grad_norm": 29.0,
      "learning_rate": 2.8128616110761898e-09,
      "loss": 3.8043,
      "num_input_tokens_seen": 53942256,
      "step": 39810
    },
    {
      "epoch": 1.274406247999488,
      "grad_norm": 22.375,
      "learning_rate": 2.6675215753724223e-09,
      "loss": 4.4255,
      "num_input_tokens_seen": 53948512,
      "step": 39815
    },
    {
      "epoch": 1.2745662889699763,
      "grad_norm": 39.75,
      "learning_rate": 2.5260364424739557e-09,
      "loss": 4.6643,
      "num_input_tokens_seen": 53955520,
      "step": 39820
    },
    {
      "epoch": 1.2747263299404648,
      "grad_norm": 38.75,
      "learning_rate": 2.3884062341994475e-09,
      "loss": 4.8305,
      "num_input_tokens_seen": 53962352,
      "step": 39825
    },
    {
      "epoch": 1.2748863709109532,
      "grad_norm": 24.125,
      "learning_rate": 2.25463097177081e-09,
      "loss": 4.1801,
      "num_input_tokens_seen": 53969264,
      "step": 39830
    },
    {
      "epoch": 1.2750464118814415,
      "grad_norm": 25.5,
      "learning_rate": 2.1247106758215397e-09,
      "loss": 3.4962,
      "num_input_tokens_seen": 53976032,
      "step": 39835
    },
    {
      "epoch": 1.27520645285193,
      "grad_norm": 24.625,
      "learning_rate": 1.998645366382834e-09,
      "loss": 3.9358,
      "num_input_tokens_seen": 53982976,
      "step": 39840
    },
    {
      "epoch": 1.2753664938224185,
      "grad_norm": 26.5,
      "learning_rate": 1.876435062897475e-09,
      "loss": 4.8239,
      "num_input_tokens_seen": 53990112,
      "step": 39845
    },
    {
      "epoch": 1.275526534792907,
      "grad_norm": 25.625,
      "learning_rate": 1.758079784211497e-09,
      "loss": 4.3512,
      "num_input_tokens_seen": 53996960,
      "step": 39850
    },
    {
      "epoch": 1.2756865757633955,
      "grad_norm": 23.5,
      "learning_rate": 1.6435795485797434e-09,
      "loss": 3.6973,
      "num_input_tokens_seen": 54003616,
      "step": 39855
    },
    {
      "epoch": 1.275846616733884,
      "grad_norm": 29.875,
      "learning_rate": 1.5329343736547596e-09,
      "loss": 4.4603,
      "num_input_tokens_seen": 54010320,
      "step": 39860
    },
    {
      "epoch": 1.2760066577043723,
      "grad_norm": 18.25,
      "learning_rate": 1.4261442765006739e-09,
      "loss": 3.6412,
      "num_input_tokens_seen": 54016752,
      "step": 39865
    },
    {
      "epoch": 1.2761666986748608,
      "grad_norm": 50.0,
      "learning_rate": 1.3232092735876445e-09,
      "loss": 5.2725,
      "num_input_tokens_seen": 54023792,
      "step": 39870
    },
    {
      "epoch": 1.2763267396453493,
      "grad_norm": 26.375,
      "learning_rate": 1.2241293807918607e-09,
      "loss": 3.5467,
      "num_input_tokens_seen": 54030800,
      "step": 39875
    },
    {
      "epoch": 1.2764867806158375,
      "grad_norm": 29.875,
      "learning_rate": 1.128904613387216e-09,
      "loss": 4.8853,
      "num_input_tokens_seen": 54038272,
      "step": 39880
    },
    {
      "epoch": 1.276646821586326,
      "grad_norm": 23.25,
      "learning_rate": 1.0375349860591853e-09,
      "loss": 4.7676,
      "num_input_tokens_seen": 54044944,
      "step": 39885
    },
    {
      "epoch": 1.2768068625568145,
      "grad_norm": 41.25,
      "learning_rate": 9.5002051290205e-10,
      "loss": 3.9065,
      "num_input_tokens_seen": 54051744,
      "step": 39890
    },
    {
      "epoch": 1.276966903527303,
      "grad_norm": 24.375,
      "learning_rate": 8.663612074077954e-10,
      "loss": 4.2099,
      "num_input_tokens_seen": 54058176,
      "step": 39895
    },
    {
      "epoch": 1.2771269444977915,
      "grad_norm": 19.5,
      "learning_rate": 7.865570824799884e-10,
      "loss": 4.1818,
      "num_input_tokens_seen": 54065248,
      "step": 39900
    },
    {
      "epoch": 1.27728698546828,
      "grad_norm": 28.5,
      "learning_rate": 7.106081504254514e-10,
      "loss": 4.4327,
      "num_input_tokens_seen": 54071648,
      "step": 39905
    },
    {
      "epoch": 1.2774470264387683,
      "grad_norm": 34.5,
      "learning_rate": 6.385144229570372e-10,
      "loss": 4.1104,
      "num_input_tokens_seen": 54078176,
      "step": 39910
    },
    {
      "epoch": 1.2776070674092568,
      "grad_norm": 24.875,
      "learning_rate": 5.70275911190854e-10,
      "loss": 4.3922,
      "num_input_tokens_seen": 54084912,
      "step": 39915
    },
    {
      "epoch": 1.2777671083797453,
      "grad_norm": 29.25,
      "learning_rate": 5.058926256490403e-10,
      "loss": 4.0688,
      "num_input_tokens_seen": 54091424,
      "step": 39920
    },
    {
      "epoch": 1.2779271493502335,
      "grad_norm": 10.375,
      "learning_rate": 4.4536457626254134e-10,
      "loss": 3.7378,
      "num_input_tokens_seen": 54098000,
      "step": 39925
    },
    {
      "epoch": 1.278087190320722,
      "grad_norm": 23.625,
      "learning_rate": 3.88691772365557e-10,
      "loss": 5.0885,
      "num_input_tokens_seen": 54104816,
      "step": 39930
    },
    {
      "epoch": 1.2782472312912105,
      "grad_norm": 23.875,
      "learning_rate": 3.358742226955425e-10,
      "loss": 4.0524,
      "num_input_tokens_seen": 54111488,
      "step": 39935
    },
    {
      "epoch": 1.278407272261699,
      "grad_norm": 23.125,
      "learning_rate": 2.8691193539875925e-10,
      "loss": 5.281,
      "num_input_tokens_seen": 54118352,
      "step": 39940
    },
    {
      "epoch": 1.2785673132321875,
      "grad_norm": 37.0,
      "learning_rate": 2.418049180274995e-10,
      "loss": 4.4591,
      "num_input_tokens_seen": 54125232,
      "step": 39945
    },
    {
      "epoch": 1.2787273542026758,
      "grad_norm": 24.75,
      "learning_rate": 2.005531775373104e-10,
      "loss": 4.9959,
      "num_input_tokens_seen": 54131792,
      "step": 39950
    },
    {
      "epoch": 1.2788873951731643,
      "grad_norm": 17.375,
      "learning_rate": 1.6315672028699435e-10,
      "loss": 4.05,
      "num_input_tokens_seen": 54138736,
      "step": 39955
    },
    {
      "epoch": 1.2790474361436528,
      "grad_norm": 43.75,
      "learning_rate": 1.2961555204693555e-10,
      "loss": 4.7774,
      "num_input_tokens_seen": 54145120,
      "step": 39960
    },
    {
      "epoch": 1.2792074771141413,
      "grad_norm": 34.75,
      "learning_rate": 9.992967798799768e-11,
      "loss": 3.6987,
      "num_input_tokens_seen": 54151872,
      "step": 39965
    },
    {
      "epoch": 1.2793675180846296,
      "grad_norm": 35.5,
      "learning_rate": 7.409910268707521e-11,
      "loss": 4.404,
      "num_input_tokens_seen": 54158800,
      "step": 39970
    },
    {
      "epoch": 1.279527559055118,
      "grad_norm": 27.125,
      "learning_rate": 5.212383012986877e-11,
      "loss": 3.6288,
      "num_input_tokens_seen": 54165648,
      "step": 39975
    },
    {
      "epoch": 1.2796876000256066,
      "grad_norm": 32.75,
      "learning_rate": 3.400386370533415e-11,
      "loss": 3.8687,
      "num_input_tokens_seen": 54172192,
      "step": 39980
    },
    {
      "epoch": 1.279847640996095,
      "grad_norm": 26.125,
      "learning_rate": 1.9739206205682258e-11,
      "loss": 3.6627,
      "num_input_tokens_seen": 54178880,
      "step": 39985
    },
    {
      "epoch": 1.2800076819665835,
      "grad_norm": 25.625,
      "learning_rate": 9.329859829154685e-12,
      "loss": 4.8128,
      "num_input_tokens_seen": 54185392,
      "step": 39990
    },
    {
      "epoch": 1.2801677229370718,
      "grad_norm": 18.75,
      "learning_rate": 2.7758261855748148e-12,
      "loss": 3.763,
      "num_input_tokens_seen": 54192000,
      "step": 39995
    },
    {
      "epoch": 1.2803277639075603,
      "grad_norm": 44.25,
      "learning_rate": 7.710628524559838e-14,
      "loss": 4.6876,
      "num_input_tokens_seen": 54198768,
      "step": 40000
    },
    {
      "epoch": 1.2803277639075603,
      "eval_loss": 4.496033191680908,
      "eval_runtime": 485.9695,
      "eval_samples_per_second": 28.574,
      "eval_steps_per_second": 14.287,
      "num_input_tokens_seen": 54198768,
      "step": 40000
    },
    {
      "epoch": 1.2803277639075603,
      "num_input_tokens_seen": 54198768,
      "step": 40000,
      "total_flos": 2.4406326062111785e+18,
      "train_loss": 4.3420969699859615,
      "train_runtime": 115401.9951,
      "train_samples_per_second": 1.386,
      "train_steps_per_second": 0.347
    }
  ],
  "logging_steps": 5,
  "max_steps": 40000,
  "num_input_tokens_seen": 54198768,
  "num_train_epochs": 2,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4406326062111785e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}