{
  "best_global_step": 31400,
  "best_metric": 0.19886288046836853,
  "best_model_checkpoint": "saves/ia3/gemma-3-1b-it/train_boolq_1745950271/checkpoint-31400",
  "epoch": 18.85902876001886,
  "eval_steps": 200,
  "global_step": 40000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0023573785950023575,
      "grad_norm": 15.485270500183105,
      "learning_rate": 4.999999876629946e-05,
      "loss": 3.2685,
      "num_input_tokens_seen": 4304,
      "step": 5
    },
    {
      "epoch": 0.004714757190004715,
      "grad_norm": 16.38265037536621,
      "learning_rate": 4.999999375439123e-05,
      "loss": 3.3775,
      "num_input_tokens_seen": 8432,
      "step": 10
    },
    {
      "epoch": 0.007072135785007072,
      "grad_norm": 17.36084747314453,
      "learning_rate": 4.9999984887169785e-05,
      "loss": 3.7333,
      "num_input_tokens_seen": 12688,
      "step": 15
    },
    {
      "epoch": 0.00942951438000943,
      "grad_norm": 14.106888771057129,
      "learning_rate": 4.9999972164636506e-05,
      "loss": 2.6515,
      "num_input_tokens_seen": 17360,
      "step": 20
    },
    {
      "epoch": 0.011786892975011787,
      "grad_norm": 13.816413879394531,
      "learning_rate": 4.999995558679334e-05,
      "loss": 3.65,
      "num_input_tokens_seen": 21808,
      "step": 25
    },
    {
      "epoch": 0.014144271570014143,
      "grad_norm": 17.012653350830078,
      "learning_rate": 4.999993515364287e-05,
      "loss": 2.9843,
      "num_input_tokens_seen": 25952,
      "step": 30
    },
    {
      "epoch": 0.0165016501650165,
      "grad_norm": 14.573676109313965,
      "learning_rate": 4.999991086518822e-05,
      "loss": 3.2827,
      "num_input_tokens_seen": 29568,
      "step": 35
    },
    {
      "epoch": 0.01885902876001886,
      "grad_norm": 15.425125122070312,
      "learning_rate": 4.999988272143315e-05,
      "loss": 3.7168,
      "num_input_tokens_seen": 33712,
      "step": 40
    },
    {
      "epoch": 0.021216407355021217,
      "grad_norm": 19.60480499267578,
      "learning_rate": 4.999985072238199e-05,
      "loss": 3.8686,
      "num_input_tokens_seen": 37856,
      "step": 45
    },
    {
      "epoch": 0.023573785950023574,
      "grad_norm": 11.83559513092041,
      "learning_rate": 4.999981486803969e-05,
      "loss": 2.2628,
      "num_input_tokens_seen": 41936,
      "step": 50
    },
    {
      "epoch": 0.02593116454502593,
      "grad_norm": 11.358976364135742,
      "learning_rate": 4.999977515841176e-05,
      "loss": 2.2008,
      "num_input_tokens_seen": 46448,
      "step": 55
    },
    {
      "epoch": 0.028288543140028287,
      "grad_norm": 5.608299255371094,
      "learning_rate": 4.9999731593504344e-05,
      "loss": 1.7321,
      "num_input_tokens_seen": 51120,
      "step": 60
    },
    {
      "epoch": 0.030645921735030647,
      "grad_norm": 6.975717067718506,
      "learning_rate": 4.999968417332415e-05,
      "loss": 1.2905,
      "num_input_tokens_seen": 55520,
      "step": 65
    },
    {
      "epoch": 0.033003300330033,
      "grad_norm": 3.0788180828094482,
      "learning_rate": 4.999963289787848e-05,
      "loss": 0.6561,
      "num_input_tokens_seen": 60496,
      "step": 70
    },
    {
      "epoch": 0.03536067892503536,
      "grad_norm": 14.735339164733887,
      "learning_rate": 4.999957776717526e-05,
      "loss": 0.8052,
      "num_input_tokens_seen": 64368,
      "step": 75
    },
    {
      "epoch": 0.03771805752003772,
      "grad_norm": 2.976858615875244,
      "learning_rate": 4.9999518781222984e-05,
      "loss": 0.8036,
      "num_input_tokens_seen": 68720,
      "step": 80
    },
    {
      "epoch": 0.040075436115040074,
      "grad_norm": 6.195841312408447,
      "learning_rate": 4.9999455940030746e-05,
      "loss": 0.4994,
      "num_input_tokens_seen": 72592,
      "step": 85
    },
    {
      "epoch": 0.042432814710042434,
      "grad_norm": 4.408599853515625,
      "learning_rate": 4.999938924360824e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 76384,
      "step": 90
    },
    {
      "epoch": 0.04479019330504479,
      "grad_norm": 3.0907809734344482,
      "learning_rate": 4.999931869196575e-05,
      "loss": 1.0307,
      "num_input_tokens_seen": 80224,
      "step": 95
    },
    {
      "epoch": 0.04714757190004715,
      "grad_norm": 2.002164363861084,
      "learning_rate": 4.999924428511416e-05,
      "loss": 0.6227,
      "num_input_tokens_seen": 84704,
      "step": 100
    },
    {
      "epoch": 0.04950495049504951,
      "grad_norm": 5.804714679718018,
      "learning_rate": 4.999916602306494e-05,
      "loss": 0.3729,
      "num_input_tokens_seen": 88688,
      "step": 105
    },
    {
      "epoch": 0.05186232909005186,
      "grad_norm": 3.962733030319214,
      "learning_rate": 4.999908390583016e-05,
      "loss": 0.2935,
      "num_input_tokens_seen": 93184,
      "step": 110
    },
    {
      "epoch": 0.05421970768505422,
      "grad_norm": 4.103965759277344,
      "learning_rate": 4.999899793342247e-05,
      "loss": 0.4128,
      "num_input_tokens_seen": 98192,
      "step": 115
    },
    {
      "epoch": 0.056577086280056574,
      "grad_norm": 3.6429519653320312,
      "learning_rate": 4.999890810585516e-05,
      "loss": 0.5566,
      "num_input_tokens_seen": 102912,
      "step": 120
    },
    {
      "epoch": 0.058934464875058934,
      "grad_norm": 6.972048759460449,
      "learning_rate": 4.999881442314206e-05,
      "loss": 0.3367,
      "num_input_tokens_seen": 108384,
      "step": 125
    },
    {
      "epoch": 0.061291843470061294,
      "grad_norm": 5.4712700843811035,
      "learning_rate": 4.9998716885297617e-05,
      "loss": 0.7951,
      "num_input_tokens_seen": 113248,
      "step": 130
    },
    {
      "epoch": 0.06364922206506365,
      "grad_norm": 3.1452548503875732,
      "learning_rate": 4.999861549233688e-05,
      "loss": 0.9523,
      "num_input_tokens_seen": 117872,
      "step": 135
    },
    {
      "epoch": 0.066006600660066,
      "grad_norm": 3.023514986038208,
      "learning_rate": 4.999851024427548e-05,
      "loss": 0.8009,
      "num_input_tokens_seen": 122176,
      "step": 140
    },
    {
      "epoch": 0.06836397925506836,
      "grad_norm": 2.1132586002349854,
      "learning_rate": 4.999840114112965e-05,
      "loss": 0.2684,
      "num_input_tokens_seen": 126432,
      "step": 145
    },
    {
      "epoch": 0.07072135785007072,
      "grad_norm": 4.9118218421936035,
      "learning_rate": 4.999828818291621e-05,
      "loss": 0.2397,
      "num_input_tokens_seen": 130576,
      "step": 150
    },
    {
      "epoch": 0.07307873644507308,
      "grad_norm": 1.9960168600082397,
      "learning_rate": 4.999817136965259e-05,
      "loss": 0.6852,
      "num_input_tokens_seen": 135184,
      "step": 155
    },
    {
      "epoch": 0.07543611504007544,
      "grad_norm": 0.6213404536247253,
      "learning_rate": 4.9998050701356794e-05,
      "loss": 0.2484,
      "num_input_tokens_seen": 139248,
      "step": 160
    },
    {
      "epoch": 0.07779349363507779,
      "grad_norm": 0.2970295250415802,
      "learning_rate": 4.999792617804744e-05,
      "loss": 0.4397,
      "num_input_tokens_seen": 143984,
      "step": 165
    },
    {
      "epoch": 0.08015087223008015,
      "grad_norm": 1.5859484672546387,
      "learning_rate": 4.9997797799743724e-05,
      "loss": 0.6151,
      "num_input_tokens_seen": 148704,
      "step": 170
    },
    {
      "epoch": 0.08250825082508251,
      "grad_norm": 4.297430038452148,
      "learning_rate": 4.999766556646545e-05,
      "loss": 0.5068,
      "num_input_tokens_seen": 153440,
      "step": 175
    },
    {
      "epoch": 0.08486562942008487,
      "grad_norm": 0.30845218896865845,
      "learning_rate": 4.9997529478232996e-05,
      "loss": 0.3963,
      "num_input_tokens_seen": 157264,
      "step": 180
    },
    {
      "epoch": 0.08722300801508723,
      "grad_norm": 1.8172276020050049,
      "learning_rate": 4.9997389535067365e-05,
      "loss": 0.8911,
      "num_input_tokens_seen": 161840,
      "step": 185
    },
    {
      "epoch": 0.08958038661008957,
      "grad_norm": 5.605410575866699,
      "learning_rate": 4.999724573699012e-05,
      "loss": 0.5995,
      "num_input_tokens_seen": 165616,
      "step": 190
    },
    {
      "epoch": 0.09193776520509193,
      "grad_norm": 4.350292205810547,
      "learning_rate": 4.9997098084023457e-05,
      "loss": 0.4971,
      "num_input_tokens_seen": 169760,
      "step": 195
    },
    {
      "epoch": 0.0942951438000943,
      "grad_norm": 3.403632879257202,
      "learning_rate": 4.999694657619013e-05,
      "loss": 0.4743,
      "num_input_tokens_seen": 174096,
      "step": 200
    },
    {
      "epoch": 0.0942951438000943,
      "eval_loss": 0.5494534969329834,
      "eval_runtime": 22.157,
      "eval_samples_per_second": 42.56,
      "eval_steps_per_second": 21.303,
      "num_input_tokens_seen": 174096,
      "step": 200
    },
    {
      "epoch": 0.09665252239509665,
      "grad_norm": 3.810647964477539,
      "learning_rate": 4.999679121351352e-05,
      "loss": 0.6237,
      "num_input_tokens_seen": 178704,
      "step": 205
    },
    {
      "epoch": 0.09900990099009901,
      "grad_norm": 3.8721418380737305,
      "learning_rate": 4.9996631996017565e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 182912,
      "step": 210
    },
    {
      "epoch": 0.10136727958510136,
      "grad_norm": 1.310941457748413,
      "learning_rate": 4.9996468923726835e-05,
      "loss": 0.4077,
      "num_input_tokens_seen": 186880,
      "step": 215
    },
    {
      "epoch": 0.10372465818010372,
      "grad_norm": 2.5896449089050293,
      "learning_rate": 4.999630199666647e-05,
      "loss": 0.4879,
      "num_input_tokens_seen": 190976,
      "step": 220
    },
    {
      "epoch": 0.10608203677510608,
      "grad_norm": 4.082836151123047,
      "learning_rate": 4.999613121486222e-05,
      "loss": 0.3663,
      "num_input_tokens_seen": 195136,
      "step": 225
    },
    {
      "epoch": 0.10843941537010844,
      "grad_norm": 1.714237928390503,
      "learning_rate": 4.999595657834041e-05,
      "loss": 0.3463,
      "num_input_tokens_seen": 199376,
      "step": 230
    },
    {
      "epoch": 0.1107967939651108,
      "grad_norm": 2.753406286239624,
      "learning_rate": 4.999577808712798e-05,
      "loss": 0.4577,
      "num_input_tokens_seen": 203104,
      "step": 235
    },
    {
      "epoch": 0.11315417256011315,
      "grad_norm": 1.9329020977020264,
      "learning_rate": 4.999559574125244e-05,
      "loss": 0.1846,
      "num_input_tokens_seen": 207440,
      "step": 240
    },
    {
      "epoch": 0.11551155115511551,
      "grad_norm": 2.888082265853882,
      "learning_rate": 4.9995409540741934e-05,
      "loss": 0.7608,
      "num_input_tokens_seen": 211520,
      "step": 245
    },
    {
      "epoch": 0.11786892975011787,
      "grad_norm": 2.217763662338257,
      "learning_rate": 4.999521948562516e-05,
      "loss": 0.1919,
      "num_input_tokens_seen": 216224,
      "step": 250
    },
    {
      "epoch": 0.12022630834512023,
      "grad_norm": 2.4813849925994873,
      "learning_rate": 4.999502557593143e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 220352,
      "step": 255
    },
    {
      "epoch": 0.12258368694012259,
      "grad_norm": 7.869880199432373,
      "learning_rate": 4.999482781169066e-05,
      "loss": 0.4345,
      "num_input_tokens_seen": 224352,
      "step": 260
    },
    {
      "epoch": 0.12494106553512493,
      "grad_norm": 3.345714569091797,
      "learning_rate": 4.9994626192933324e-05,
      "loss": 0.7938,
      "num_input_tokens_seen": 228240,
      "step": 265
    },
    {
      "epoch": 0.1272984441301273,
      "grad_norm": 4.560533046722412,
      "learning_rate": 4.999442071969054e-05,
      "loss": 0.4596,
      "num_input_tokens_seen": 232304,
      "step": 270
    },
    {
      "epoch": 0.12965582272512965,
      "grad_norm": 4.127234935760498,
      "learning_rate": 4.999421139199397e-05,
      "loss": 0.2265,
      "num_input_tokens_seen": 236464,
      "step": 275
    },
    {
      "epoch": 0.132013201320132,
      "grad_norm": 1.1112598180770874,
      "learning_rate": 4.999399820987592e-05,
      "loss": 0.4097,
      "num_input_tokens_seen": 241440,
      "step": 280
    },
    {
      "epoch": 0.13437057991513437,
      "grad_norm": 2.470679759979248,
      "learning_rate": 4.999378117336924e-05,
      "loss": 0.2632,
      "num_input_tokens_seen": 245104,
      "step": 285
    },
    {
      "epoch": 0.13672795851013672,
      "grad_norm": 3.0626373291015625,
      "learning_rate": 4.9993560282507415e-05,
      "loss": 0.5101,
      "num_input_tokens_seen": 249552,
      "step": 290
    },
    {
      "epoch": 0.1390853371051391,
      "grad_norm": 1.756316900253296,
      "learning_rate": 4.9993335537324495e-05,
      "loss": 0.2586,
      "num_input_tokens_seen": 253232,
      "step": 295
    },
    {
      "epoch": 0.14144271570014144,
      "grad_norm": 1.8645142316818237,
      "learning_rate": 4.999310693785516e-05,
      "loss": 0.3428,
      "num_input_tokens_seen": 257456,
      "step": 300
    },
    {
      "epoch": 0.1438000942951438,
      "grad_norm": 3.156534433364868,
      "learning_rate": 4.9992874484134653e-05,
      "loss": 0.7119,
      "num_input_tokens_seen": 262112,
      "step": 305
    },
    {
      "epoch": 0.14615747289014616,
      "grad_norm": 1.1955878734588623,
      "learning_rate": 4.999263817619882e-05,
      "loss": 0.3523,
      "num_input_tokens_seen": 266208,
      "step": 310
    },
    {
      "epoch": 0.1485148514851485,
      "grad_norm": 0.7811552286148071,
      "learning_rate": 4.9992398014084105e-05,
      "loss": 0.4033,
      "num_input_tokens_seen": 269792,
      "step": 315
    },
    {
      "epoch": 0.15087223008015088,
      "grad_norm": 3.125236749649048,
      "learning_rate": 4.999215399782754e-05,
      "loss": 0.5893,
      "num_input_tokens_seen": 274000,
      "step": 320
    },
    {
      "epoch": 0.15322960867515323,
      "grad_norm": 7.45491886138916,
      "learning_rate": 4.999190612746675e-05,
      "loss": 0.3085,
      "num_input_tokens_seen": 278304,
      "step": 325
    },
    {
      "epoch": 0.15558698727015557,
      "grad_norm": 3.9699885845184326,
      "learning_rate": 4.999165440303998e-05,
      "loss": 0.4413,
      "num_input_tokens_seen": 282976,
      "step": 330
    },
    {
      "epoch": 0.15794436586515795,
      "grad_norm": 0.3015575408935547,
      "learning_rate": 4.999139882458603e-05,
      "loss": 0.2762,
      "num_input_tokens_seen": 287584,
      "step": 335
    },
    {
      "epoch": 0.1603017444601603,
      "grad_norm": 2.034245014190674,
      "learning_rate": 4.9991139392144314e-05,
      "loss": 0.3905,
      "num_input_tokens_seen": 293632,
      "step": 340
    },
    {
      "epoch": 0.16265912305516267,
      "grad_norm": 1.6614937782287598,
      "learning_rate": 4.999087610575485e-05,
      "loss": 0.2129,
      "num_input_tokens_seen": 297712,
      "step": 345
    },
    {
      "epoch": 0.16501650165016502,
      "grad_norm": 3.0024757385253906,
      "learning_rate": 4.999060896545824e-05,
      "loss": 0.3115,
      "num_input_tokens_seen": 302240,
      "step": 350
    },
    {
      "epoch": 0.16737388024516736,
      "grad_norm": 6.979857921600342,
      "learning_rate": 4.999033797129568e-05,
      "loss": 0.2782,
      "num_input_tokens_seen": 306336,
      "step": 355
    },
    {
      "epoch": 0.16973125884016974,
      "grad_norm": 2.6858768463134766,
      "learning_rate": 4.999006312330894e-05,
      "loss": 0.3571,
      "num_input_tokens_seen": 310864,
      "step": 360
    },
    {
      "epoch": 0.17208863743517208,
      "grad_norm": 8.033185958862305,
      "learning_rate": 4.998978442154043e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 315072,
      "step": 365
    },
    {
      "epoch": 0.17444601603017446,
      "grad_norm": 1.980529546737671,
      "learning_rate": 4.9989501866033125e-05,
      "loss": 0.411,
      "num_input_tokens_seen": 319552,
      "step": 370
    },
    {
      "epoch": 0.1768033946251768,
      "grad_norm": 3.291781187057495,
      "learning_rate": 4.998921545683059e-05,
      "loss": 0.3859,
      "num_input_tokens_seen": 323952,
      "step": 375
    },
    {
      "epoch": 0.17916077322017915,
      "grad_norm": 0.6430925726890564,
      "learning_rate": 4.9988925193976996e-05,
      "loss": 0.2949,
      "num_input_tokens_seen": 327856,
      "step": 380
    },
    {
      "epoch": 0.18151815181518152,
      "grad_norm": 2.6830475330352783,
      "learning_rate": 4.998863107751711e-05,
      "loss": 0.1764,
      "num_input_tokens_seen": 332384,
      "step": 385
    },
    {
      "epoch": 0.18387553041018387,
      "grad_norm": 3.4273626804351807,
      "learning_rate": 4.998833310749629e-05,
      "loss": 0.2755,
      "num_input_tokens_seen": 336992,
      "step": 390
    },
    {
      "epoch": 0.18623290900518624,
      "grad_norm": 3.934072732925415,
      "learning_rate": 4.998803128396047e-05,
      "loss": 0.2294,
      "num_input_tokens_seen": 340592,
      "step": 395
    },
    {
      "epoch": 0.1885902876001886,
      "grad_norm": 4.120837688446045,
      "learning_rate": 4.9987725606956215e-05,
      "loss": 0.4901,
      "num_input_tokens_seen": 344560,
      "step": 400
    },
    {
      "epoch": 0.1885902876001886,
      "eval_loss": 0.3760131597518921,
      "eval_runtime": 22.1521,
      "eval_samples_per_second": 42.569,
      "eval_steps_per_second": 21.307,
      "num_input_tokens_seen": 344560,
      "step": 400
    },
    {
      "epoch": 0.19094766619519093,
      "grad_norm": 3.286654472351074,
      "learning_rate": 4.998741607653066e-05,
      "loss": 0.4333,
      "num_input_tokens_seen": 349088,
      "step": 405
    },
    {
      "epoch": 0.1933050447901933,
      "grad_norm": 1.840010166168213,
      "learning_rate": 4.9987102692731523e-05,
      "loss": 0.1458,
      "num_input_tokens_seen": 353472,
      "step": 410
    },
    {
      "epoch": 0.19566242338519566,
      "grad_norm": 2.575014591217041,
      "learning_rate": 4.9986785455607157e-05,
      "loss": 0.3014,
      "num_input_tokens_seen": 357904,
      "step": 415
    },
    {
      "epoch": 0.19801980198019803,
      "grad_norm": 2.5661110877990723,
      "learning_rate": 4.9986464365206456e-05,
      "loss": 0.3221,
      "num_input_tokens_seen": 362448,
      "step": 420
    },
    {
      "epoch": 0.20037718057520038,
      "grad_norm": 0.9874239563941956,
      "learning_rate": 4.9986139421578956e-05,
      "loss": 0.2325,
      "num_input_tokens_seen": 367744,
      "step": 425
    },
    {
      "epoch": 0.20273455917020272,
      "grad_norm": 1.544314980506897,
      "learning_rate": 4.998581062477477e-05,
      "loss": 0.1713,
      "num_input_tokens_seen": 371920,
      "step": 430
    },
    {
      "epoch": 0.2050919377652051,
      "grad_norm": 1.5154350996017456,
      "learning_rate": 4.998547797484458e-05,
      "loss": 0.3238,
      "num_input_tokens_seen": 376832,
      "step": 435
    },
    {
      "epoch": 0.20744931636020744,
      "grad_norm": 1.0819036960601807,
      "learning_rate": 4.9985141471839706e-05,
      "loss": 0.4176,
      "num_input_tokens_seen": 380800,
      "step": 440
    },
    {
      "epoch": 0.20980669495520982,
      "grad_norm": 2.0719358921051025,
      "learning_rate": 4.998480111581203e-05,
      "loss": 0.2675,
      "num_input_tokens_seen": 384736,
      "step": 445
    },
    {
      "epoch": 0.21216407355021216,
      "grad_norm": 3.10402512550354,
      "learning_rate": 4.998445690681405e-05,
      "loss": 0.3555,
      "num_input_tokens_seen": 389840,
      "step": 450
    },
    {
      "epoch": 0.2145214521452145,
      "grad_norm": 1.9251759052276611,
      "learning_rate": 4.9984108844898834e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 393680,
      "step": 455
    },
    {
      "epoch": 0.21687883074021688,
      "grad_norm": 1.5773593187332153,
      "learning_rate": 4.9983756930120076e-05,
      "loss": 0.4691,
      "num_input_tokens_seen": 397328,
      "step": 460
    },
    {
      "epoch": 0.21923620933521923,
      "grad_norm": 2.6630897521972656,
      "learning_rate": 4.9983401162532025e-05,
      "loss": 0.3377,
      "num_input_tokens_seen": 401648,
      "step": 465
    },
    {
      "epoch": 0.2215935879302216,
      "grad_norm": 7.35960578918457,
      "learning_rate": 4.998304154218955e-05,
      "loss": 0.3806,
      "num_input_tokens_seen": 405744,
      "step": 470
    },
    {
      "epoch": 0.22395096652522395,
      "grad_norm": 1.4894665479660034,
      "learning_rate": 4.998267806914812e-05,
      "loss": 0.2587,
      "num_input_tokens_seen": 410016,
      "step": 475
    },
    {
      "epoch": 0.2263083451202263,
      "grad_norm": 1.8842252492904663,
      "learning_rate": 4.998231074346378e-05,
      "loss": 0.4347,
      "num_input_tokens_seen": 413680,
      "step": 480
    },
    {
      "epoch": 0.22866572371522867,
      "grad_norm": 2.2352328300476074,
      "learning_rate": 4.998193956519317e-05,
      "loss": 0.3335,
      "num_input_tokens_seen": 418064,
      "step": 485
    },
    {
      "epoch": 0.23102310231023102,
      "grad_norm": 0.7479366064071655,
      "learning_rate": 4.9981564534393545e-05,
      "loss": 0.2052,
      "num_input_tokens_seen": 422400,
      "step": 490
    },
    {
      "epoch": 0.2333804809052334,
      "grad_norm": 5.417942523956299,
      "learning_rate": 4.998118565112272e-05,
      "loss": 0.481,
      "num_input_tokens_seen": 426688,
      "step": 495
    },
    {
      "epoch": 0.23573785950023574,
      "grad_norm": 2.065774917602539,
      "learning_rate": 4.998080291543914e-05,
      "loss": 0.2447,
      "num_input_tokens_seen": 431264,
      "step": 500
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 0.8569350838661194,
      "learning_rate": 4.9980416327401826e-05,
      "loss": 0.2779,
      "num_input_tokens_seen": 435184,
      "step": 505
    },
    {
      "epoch": 0.24045261669024046,
      "grad_norm": 0.2228672206401825,
      "learning_rate": 4.998002588707038e-05,
      "loss": 0.2902,
      "num_input_tokens_seen": 439376,
      "step": 510
    },
    {
      "epoch": 0.2428099952852428,
      "grad_norm": 2.355104684829712,
      "learning_rate": 4.997963159450503e-05,
      "loss": 0.3055,
      "num_input_tokens_seen": 443472,
      "step": 515
    },
    {
      "epoch": 0.24516737388024518,
      "grad_norm": 1.6579084396362305,
      "learning_rate": 4.9979233449766575e-05,
      "loss": 0.2467,
      "num_input_tokens_seen": 447904,
      "step": 520
    },
    {
      "epoch": 0.24752475247524752,
      "grad_norm": 1.195425271987915,
      "learning_rate": 4.997883145291641e-05,
      "loss": 0.2467,
      "num_input_tokens_seen": 452928,
      "step": 525
    },
    {
      "epoch": 0.24988213107024987,
      "grad_norm": 1.1657607555389404,
      "learning_rate": 4.9978425604016536e-05,
      "loss": 0.2282,
      "num_input_tokens_seen": 456768,
      "step": 530
    },
    {
      "epoch": 0.2522395096652522,
      "grad_norm": 1.9048701524734497,
      "learning_rate": 4.9978015903129536e-05,
      "loss": 0.1188,
      "num_input_tokens_seen": 461040,
      "step": 535
    },
    {
      "epoch": 0.2545968882602546,
      "grad_norm": 0.9923669695854187,
      "learning_rate": 4.997760235031859e-05,
      "loss": 0.2032,
      "num_input_tokens_seen": 465088,
      "step": 540
    },
    {
      "epoch": 0.25695426685525696,
      "grad_norm": 2.9903905391693115,
      "learning_rate": 4.9977184945647473e-05,
      "loss": 0.2822,
      "num_input_tokens_seen": 469568,
      "step": 545
    },
    {
      "epoch": 0.2593116454502593,
      "grad_norm": 3.4362478256225586,
      "learning_rate": 4.997676368918055e-05,
      "loss": 0.4165,
      "num_input_tokens_seen": 473264,
      "step": 550
    },
    {
      "epoch": 0.26166902404526166,
      "grad_norm": 1.5086959600448608,
      "learning_rate": 4.9976338580982794e-05,
      "loss": 0.3489,
      "num_input_tokens_seen": 477792,
      "step": 555
    },
    {
      "epoch": 0.264026402640264,
      "grad_norm": 0.7849045395851135,
      "learning_rate": 4.9975909621119755e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 481856,
      "step": 560
    },
    {
      "epoch": 0.2663837812352664,
      "grad_norm": 1.0226699113845825,
      "learning_rate": 4.997547680965758e-05,
      "loss": 0.1499,
      "num_input_tokens_seen": 487024,
      "step": 565
    },
    {
      "epoch": 0.26874115983026875,
      "grad_norm": 3.0884180068969727,
      "learning_rate": 4.997504014666302e-05,
      "loss": 0.1726,
      "num_input_tokens_seen": 490816,
      "step": 570
    },
    {
      "epoch": 0.2710985384252711,
      "grad_norm": 0.9204435348510742,
      "learning_rate": 4.997459963220342e-05,
      "loss": 0.2994,
      "num_input_tokens_seen": 495696,
      "step": 575
    },
    {
      "epoch": 0.27345591702027344,
      "grad_norm": 7.373989582061768,
      "learning_rate": 4.997415526634671e-05,
      "loss": 0.3038,
      "num_input_tokens_seen": 500800,
      "step": 580
    },
    {
      "epoch": 0.2758132956152758,
      "grad_norm": 1.907460331916809,
      "learning_rate": 4.99737070491614e-05,
      "loss": 0.2099,
      "num_input_tokens_seen": 504864,
      "step": 585
    },
    {
      "epoch": 0.2781706742102782,
      "grad_norm": 3.021040678024292,
      "learning_rate": 4.997325498071663e-05,
      "loss": 0.2433,
      "num_input_tokens_seen": 509104,
      "step": 590
    },
    {
      "epoch": 0.28052805280528054,
      "grad_norm": 0.8773831129074097,
      "learning_rate": 4.997279906108211e-05,
      "loss": 0.1713,
      "num_input_tokens_seen": 512720,
      "step": 595
    },
    {
      "epoch": 0.2828854314002829,
      "grad_norm": 2.8579413890838623,
      "learning_rate": 4.9972339290328155e-05,
      "loss": 0.523,
      "num_input_tokens_seen": 517536,
      "step": 600
    },
    {
      "epoch": 0.2828854314002829,
      "eval_loss": 0.2801620066165924,
      "eval_runtime": 22.1627,
      "eval_samples_per_second": 42.549,
      "eval_steps_per_second": 21.297,
      "num_input_tokens_seen": 517536,
      "step": 600
    },
    {
      "epoch": 0.28524280999528523,
      "grad_norm": 0.40670862793922424,
      "learning_rate": 4.9971875668525646e-05,
      "loss": 0.1376,
      "num_input_tokens_seen": 521488,
      "step": 605
    },
    {
      "epoch": 0.2876001885902876,
      "grad_norm": 2.0429904460906982,
      "learning_rate": 4.997140819574609e-05,
      "loss": 0.1192,
      "num_input_tokens_seen": 525808,
      "step": 610
    },
    {
      "epoch": 0.28995756718529,
      "grad_norm": 1.2519134283065796,
      "learning_rate": 4.997093687206159e-05,
      "loss": 0.2178,
      "num_input_tokens_seen": 529840,
      "step": 615
    },
    {
      "epoch": 0.2923149457802923,
      "grad_norm": 0.7289308309555054,
      "learning_rate": 4.997046169754482e-05,
      "loss": 0.3011,
      "num_input_tokens_seen": 534096,
      "step": 620
    },
    {
      "epoch": 0.29467232437529467,
      "grad_norm": 1.3785525560379028,
      "learning_rate": 4.996998267226905e-05,
      "loss": 0.3059,
      "num_input_tokens_seen": 538752,
      "step": 625
    },
    {
      "epoch": 0.297029702970297,
      "grad_norm": 1.8853416442871094,
      "learning_rate": 4.996949979630817e-05,
      "loss": 0.2491,
      "num_input_tokens_seen": 543968,
      "step": 630
    },
    {
      "epoch": 0.29938708156529936,
      "grad_norm": 1.4006366729736328,
      "learning_rate": 4.996901306973663e-05,
      "loss": 0.2189,
      "num_input_tokens_seen": 548784,
      "step": 635
    },
    {
      "epoch": 0.30174446016030176,
      "grad_norm": 1.6035689115524292,
      "learning_rate": 4.996852249262949e-05,
      "loss": 0.2963,
      "num_input_tokens_seen": 552880,
      "step": 640
    },
    {
      "epoch": 0.3041018387553041,
      "grad_norm": 2.816256046295166,
      "learning_rate": 4.996802806506241e-05,
      "loss": 0.3683,
      "num_input_tokens_seen": 557184,
      "step": 645
    },
    {
      "epoch": 0.30645921735030646,
      "grad_norm": 1.6280863285064697,
      "learning_rate": 4.996752978711164e-05,
      "loss": 0.274,
      "num_input_tokens_seen": 561840,
      "step": 650
    },
    {
      "epoch": 0.3088165959453088,
      "grad_norm": 2.836310386657715,
      "learning_rate": 4.996702765885401e-05,
      "loss": 0.3431,
      "num_input_tokens_seen": 566736,
      "step": 655
    },
    {
      "epoch": 0.31117397454031115,
      "grad_norm": 1.560479998588562,
      "learning_rate": 4.9966521680366964e-05,
      "loss": 0.2247,
      "num_input_tokens_seen": 570528,
      "step": 660
    },
    {
      "epoch": 0.31353135313531355,
      "grad_norm": 3.290661096572876,
      "learning_rate": 4.9966011851728524e-05,
      "loss": 0.2809,
      "num_input_tokens_seen": 574720,
      "step": 665
    },
    {
      "epoch": 0.3158887317303159,
      "grad_norm": 1.8162204027175903,
      "learning_rate": 4.996549817301731e-05,
      "loss": 0.393,
      "num_input_tokens_seen": 578912,
      "step": 670
    },
    {
      "epoch": 0.31824611032531824,
      "grad_norm": 1.262295126914978,
      "learning_rate": 4.9964980644312544e-05,
      "loss": 0.2894,
      "num_input_tokens_seen": 584544,
      "step": 675
    },
    {
      "epoch": 0.3206034889203206,
      "grad_norm": 3.2502832412719727,
      "learning_rate": 4.996445926569403e-05,
      "loss": 0.2984,
      "num_input_tokens_seen": 588752,
      "step": 680
    },
    {
      "epoch": 0.32296086751532294,
      "grad_norm": 1.0783370733261108,
      "learning_rate": 4.996393403724218e-05,
      "loss": 0.1654,
      "num_input_tokens_seen": 593024,
      "step": 685
    },
    {
      "epoch": 0.32531824611032534,
      "grad_norm": 1.6566026210784912,
      "learning_rate": 4.9963404959037985e-05,
      "loss": 0.1897,
      "num_input_tokens_seen": 597280,
      "step": 690
    },
    {
      "epoch": 0.3276756247053277,
      "grad_norm": 1.782347321510315,
      "learning_rate": 4.996287203116303e-05,
      "loss": 0.2667,
      "num_input_tokens_seen": 601904,
      "step": 695
    },
    {
      "epoch": 0.33003300330033003,
      "grad_norm": 0.6407699584960938,
      "learning_rate": 4.996233525369951e-05,
      "loss": 0.2477,
      "num_input_tokens_seen": 606368,
      "step": 700
    },
    {
      "epoch": 0.3323903818953324,
      "grad_norm": 1.1520178318023682,
      "learning_rate": 4.99617946267302e-05,
      "loss": 0.2093,
      "num_input_tokens_seen": 610096,
      "step": 705
    },
    {
      "epoch": 0.3347477604903347,
      "grad_norm": 2.1915571689605713,
      "learning_rate": 4.996125015033846e-05,
      "loss": 0.2612,
      "num_input_tokens_seen": 614288,
      "step": 710
    },
    {
      "epoch": 0.3371051390853371,
      "grad_norm": 1.2935360670089722,
      "learning_rate": 4.996070182460827e-05,
      "loss": 0.2288,
      "num_input_tokens_seen": 619344,
      "step": 715
    },
    {
      "epoch": 0.33946251768033947,
      "grad_norm": 2.77652645111084,
      "learning_rate": 4.996014964962418e-05,
      "loss": 0.2582,
      "num_input_tokens_seen": 624240,
      "step": 720
    },
    {
      "epoch": 0.3418198962753418,
      "grad_norm": 1.8685349225997925,
      "learning_rate": 4.9959593625471344e-05,
      "loss": 0.2348,
      "num_input_tokens_seen": 629312,
      "step": 725
    },
    {
      "epoch": 0.34417727487034416,
      "grad_norm": 1.4558298587799072,
      "learning_rate": 4.995903375223552e-05,
      "loss": 0.1597,
      "num_input_tokens_seen": 633952,
      "step": 730
    },
    {
      "epoch": 0.3465346534653465,
      "grad_norm": 2.2438430786132812,
      "learning_rate": 4.995847003000302e-05,
      "loss": 0.263,
      "num_input_tokens_seen": 638352,
      "step": 735
    },
    {
      "epoch": 0.3488920320603489,
      "grad_norm": 0.8623434901237488,
      "learning_rate": 4.9957902458860804e-05,
      "loss": 0.2592,
      "num_input_tokens_seen": 642656,
      "step": 740
    },
    {
      "epoch": 0.35124941065535126,
      "grad_norm": 1.0298453569412231,
      "learning_rate": 4.995733103889639e-05,
      "loss": 0.1582,
      "num_input_tokens_seen": 647696,
      "step": 745
    },
    {
      "epoch": 0.3536067892503536,
      "grad_norm": 4.065328121185303,
      "learning_rate": 4.99567557701979e-05,
      "loss": 0.2918,
      "num_input_tokens_seen": 652832,
      "step": 750
    },
    {
      "epoch": 0.35596416784535595,
      "grad_norm": 0.31295451521873474,
      "learning_rate": 4.995617665285403e-05,
      "loss": 0.2709,
      "num_input_tokens_seen": 656768,
      "step": 755
    },
    {
      "epoch": 0.3583215464403583,
      "grad_norm": 0.6371302008628845,
      "learning_rate": 4.99555936869541e-05,
      "loss": 0.215,
      "num_input_tokens_seen": 660624,
      "step": 760
    },
    {
      "epoch": 0.3606789250353607,
      "grad_norm": 1.9334808588027954,
      "learning_rate": 4.995500687258803e-05,
      "loss": 0.1429,
      "num_input_tokens_seen": 665488,
      "step": 765
    },
    {
      "epoch": 0.36303630363036304,
      "grad_norm": 1.1226927042007446,
      "learning_rate": 4.995441620984628e-05,
      "loss": 0.1665,
      "num_input_tokens_seen": 669504,
      "step": 770
    },
    {
      "epoch": 0.3653936822253654,
      "grad_norm": 3.8185677528381348,
      "learning_rate": 4.995382169881996e-05,
      "loss": 0.3739,
      "num_input_tokens_seen": 674032,
      "step": 775
    },
    {
      "epoch": 0.36775106082036774,
      "grad_norm": 1.3331282138824463,
      "learning_rate": 4.9953223339600755e-05,
      "loss": 0.2049,
      "num_input_tokens_seen": 678384,
      "step": 780
    },
    {
      "epoch": 0.3701084394153701,
      "grad_norm": 1.704132318496704,
      "learning_rate": 4.995262113228091e-05,
      "loss": 0.1895,
      "num_input_tokens_seen": 682944,
      "step": 785
    },
    {
      "epoch": 0.3724658180103725,
      "grad_norm": 0.8047317862510681,
      "learning_rate": 4.995201507695332e-05,
      "loss": 0.2037,
      "num_input_tokens_seen": 687072,
      "step": 790
    },
    {
      "epoch": 0.37482319660537483,
      "grad_norm": 2.829162359237671,
      "learning_rate": 4.995140517371144e-05,
      "loss": 0.3012,
      "num_input_tokens_seen": 691632,
      "step": 795
    },
    {
      "epoch": 0.3771805752003772,
      "grad_norm": 1.022626519203186,
      "learning_rate": 4.995079142264932e-05,
      "loss": 0.2411,
      "num_input_tokens_seen": 696016,
      "step": 800
    },
    {
      "epoch": 0.3771805752003772,
      "eval_loss": 0.27237120270729065,
      "eval_runtime": 22.1737,
      "eval_samples_per_second": 42.528,
      "eval_steps_per_second": 21.287,
      "num_input_tokens_seen": 696016,
      "step": 800
    },
    {
      "epoch": 0.3795379537953795,
      "grad_norm": 3.017665147781372,
      "learning_rate": 4.995017382386162e-05,
      "loss": 0.1633,
      "num_input_tokens_seen": 700240,
      "step": 805
    },
    {
      "epoch": 0.38189533239038187,
      "grad_norm": 0.9114702939987183,
      "learning_rate": 4.994955237744356e-05,
      "loss": 0.2364,
      "num_input_tokens_seen": 704448,
      "step": 810
    },
    {
      "epoch": 0.38425271098538427,
      "grad_norm": 0.47185760736465454,
      "learning_rate": 4.994892708349101e-05,
      "loss": 0.2006,
      "num_input_tokens_seen": 707968,
      "step": 815
    },
    {
      "epoch": 0.3866100895803866,
      "grad_norm": 2.7477738857269287,
      "learning_rate": 4.994829794210035e-05,
      "loss": 0.3071,
      "num_input_tokens_seen": 712272,
      "step": 820
    },
    {
      "epoch": 0.38896746817538896,
      "grad_norm": 1.056610345840454,
      "learning_rate": 4.994766495336864e-05,
      "loss": 0.2703,
      "num_input_tokens_seen": 716784,
      "step": 825
    },
    {
      "epoch": 0.3913248467703913,
      "grad_norm": 2.4203360080718994,
      "learning_rate": 4.994702811739348e-05,
      "loss": 0.3913,
      "num_input_tokens_seen": 721296,
      "step": 830
    },
    {
      "epoch": 0.39368222536539366,
      "grad_norm": 0.7927408218383789,
      "learning_rate": 4.994638743427308e-05,
      "loss": 0.3097,
      "num_input_tokens_seen": 725536,
      "step": 835
    },
    {
      "epoch": 0.39603960396039606,
      "grad_norm": 0.8012939095497131,
      "learning_rate": 4.994574290410624e-05,
      "loss": 0.2592,
      "num_input_tokens_seen": 730384,
      "step": 840
    },
    {
      "epoch": 0.3983969825553984,
      "grad_norm": 0.6067067384719849,
      "learning_rate": 4.9945094526992364e-05,
      "loss": 0.3206,
      "num_input_tokens_seen": 734960,
      "step": 845
    },
    {
      "epoch": 0.40075436115040075,
      "grad_norm": 1.3486987352371216,
      "learning_rate": 4.994444230303142e-05,
      "loss": 0.243,
      "num_input_tokens_seen": 739040,
      "step": 850
    },
    {
      "epoch": 0.4031117397454031,
      "grad_norm": 1.8352265357971191,
      "learning_rate": 4.994378623232402e-05,
      "loss": 0.3216,
      "num_input_tokens_seen": 743184,
      "step": 855
    },
    {
      "epoch": 0.40546911834040544,
      "grad_norm": 0.7229653000831604,
      "learning_rate": 4.99431263149713e-05,
      "loss": 0.1753,
      "num_input_tokens_seen": 747056,
      "step": 860
    },
    {
      "epoch": 0.40782649693540785,
      "grad_norm": 0.9458274245262146,
      "learning_rate": 4.9942462551075056e-05,
      "loss": 0.2682,
      "num_input_tokens_seen": 751184,
      "step": 865
    },
    {
      "epoch": 0.4101838755304102,
      "grad_norm": 1.287966251373291,
      "learning_rate": 4.994179494073764e-05,
      "loss": 0.2144,
      "num_input_tokens_seen": 755392,
      "step": 870
    },
    {
      "epoch": 0.41254125412541254,
      "grad_norm": 1.3812793493270874,
      "learning_rate": 4.9941123484062e-05,
      "loss": 0.2116,
      "num_input_tokens_seen": 759648,
      "step": 875
    },
    {
      "epoch": 0.4148986327204149,
      "grad_norm": 1.0676442384719849,
      "learning_rate": 4.99404481811517e-05,
      "loss": 0.2238,
      "num_input_tokens_seen": 763792,
      "step": 880
    },
    {
      "epoch": 0.41725601131541723,
      "grad_norm": 1.0318809747695923,
      "learning_rate": 4.9939769032110864e-05,
      "loss": 0.1735,
      "num_input_tokens_seen": 768944,
      "step": 885
    },
    {
      "epoch": 0.41961338991041963,
      "grad_norm": 0.9856290817260742,
      "learning_rate": 4.993908603704423e-05,
      "loss": 0.199,
      "num_input_tokens_seen": 773360,
      "step": 890
    },
    {
      "epoch": 0.421970768505422,
      "grad_norm": 1.0585646629333496,
      "learning_rate": 4.9938399196057126e-05,
      "loss": 0.2945,
      "num_input_tokens_seen": 777808,
      "step": 895
    },
    {
      "epoch": 0.4243281471004243,
      "grad_norm": 2.987017869949341,
      "learning_rate": 4.993770850925547e-05,
      "loss": 0.2816,
      "num_input_tokens_seen": 784416,
      "step": 900
    },
    {
      "epoch": 0.42668552569542667,
      "grad_norm": 2.0195350646972656,
      "learning_rate": 4.993701397674577e-05,
      "loss": 0.2649,
      "num_input_tokens_seen": 788464,
      "step": 905
    },
    {
      "epoch": 0.429042904290429,
      "grad_norm": 4.003117084503174,
      "learning_rate": 4.993631559863515e-05,
      "loss": 0.1489,
      "num_input_tokens_seen": 793232,
      "step": 910
    },
    {
      "epoch": 0.4314002828854314,
      "grad_norm": 1.2080997228622437,
      "learning_rate": 4.9935613375031283e-05,
      "loss": 0.2555,
      "num_input_tokens_seen": 797872,
      "step": 915
    },
    {
      "epoch": 0.43375766148043376,
      "grad_norm": 1.4982812404632568,
      "learning_rate": 4.993490730604248e-05,
      "loss": 0.1823,
      "num_input_tokens_seen": 802000,
      "step": 920
    },
    {
      "epoch": 0.4361150400754361,
      "grad_norm": 0.805120050907135,
      "learning_rate": 4.993419739177761e-05,
      "loss": 0.1543,
      "num_input_tokens_seen": 806192,
      "step": 925
    },
    {
      "epoch": 0.43847241867043846,
      "grad_norm": 1.2698570489883423,
      "learning_rate": 4.9933483632346164e-05,
      "loss": 0.2402,
      "num_input_tokens_seen": 810032,
      "step": 930
    },
    {
      "epoch": 0.4408297972654408,
      "grad_norm": 1.791339635848999,
      "learning_rate": 4.993276602785821e-05,
      "loss": 0.2876,
      "num_input_tokens_seen": 813776,
      "step": 935
    },
    {
      "epoch": 0.4431871758604432,
      "grad_norm": 0.8778189420700073,
      "learning_rate": 4.993204457842441e-05,
      "loss": 0.1399,
      "num_input_tokens_seen": 818272,
      "step": 940
    },
    {
      "epoch": 0.44554455445544555,
      "grad_norm": 1.7652504444122314,
      "learning_rate": 4.993131928415602e-05,
      "loss": 0.186,
      "num_input_tokens_seen": 823504,
      "step": 945
    },
    {
      "epoch": 0.4479019330504479,
      "grad_norm": 2.162152051925659,
      "learning_rate": 4.993059014516489e-05,
      "loss": 0.284,
      "num_input_tokens_seen": 827280,
      "step": 950
    },
    {
      "epoch": 0.45025931164545024,
      "grad_norm": 1.9402828216552734,
      "learning_rate": 4.9929857161563464e-05,
      "loss": 0.2634,
      "num_input_tokens_seen": 831072,
      "step": 955
    },
    {
      "epoch": 0.4526166902404526,
      "grad_norm": 1.9261986017227173,
      "learning_rate": 4.992912033346477e-05,
      "loss": 0.2338,
      "num_input_tokens_seen": 834928,
      "step": 960
    },
    {
      "epoch": 0.454974068835455,
      "grad_norm": 1.2357189655303955,
      "learning_rate": 4.992837966098245e-05,
      "loss": 0.221,
      "num_input_tokens_seen": 838800,
      "step": 965
    },
    {
      "epoch": 0.45733144743045734,
      "grad_norm": 1.2866970300674438,
      "learning_rate": 4.992763514423071e-05,
      "loss": 0.347,
      "num_input_tokens_seen": 842992,
      "step": 970
    },
    {
      "epoch": 0.4596888260254597,
      "grad_norm": 4.084360599517822,
      "learning_rate": 4.992688678332437e-05,
      "loss": 0.2999,
      "num_input_tokens_seen": 847216,
      "step": 975
    },
    {
      "epoch": 0.46204620462046203,
      "grad_norm": 1.550920844078064,
      "learning_rate": 4.992613457837884e-05,
      "loss": 0.156,
      "num_input_tokens_seen": 851152,
      "step": 980
    },
    {
      "epoch": 0.4644035832154644,
      "grad_norm": 2.8920774459838867,
      "learning_rate": 4.992537852951011e-05,
      "loss": 0.1982,
      "num_input_tokens_seen": 855520,
      "step": 985
    },
    {
      "epoch": 0.4667609618104668,
      "grad_norm": 2.7516210079193115,
      "learning_rate": 4.9924618636834785e-05,
      "loss": 0.3078,
      "num_input_tokens_seen": 860288,
      "step": 990
    },
    {
      "epoch": 0.4691183404054691,
      "grad_norm": 1.0556129217147827,
      "learning_rate": 4.9923854900470046e-05,
      "loss": 0.2197,
      "num_input_tokens_seen": 864848,
      "step": 995
    },
    {
      "epoch": 0.47147571900047147,
      "grad_norm": 0.9165362119674683,
      "learning_rate": 4.992308732053367e-05,
      "loss": 0.1891,
      "num_input_tokens_seen": 868992,
      "step": 1000
    },
    {
      "epoch": 0.47147571900047147,
      "eval_loss": 0.25863251090049744,
      "eval_runtime": 22.1572,
      "eval_samples_per_second": 42.56,
      "eval_steps_per_second": 21.302,
      "num_input_tokens_seen": 868992,
      "step": 1000
    },
    {
      "epoch": 0.4738330975954738,
      "grad_norm": 2.4298970699310303,
      "learning_rate": 4.992231589714402e-05,
      "loss": 0.258,
      "num_input_tokens_seen": 872576,
      "step": 1005
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 2.2335283756256104,
      "learning_rate": 4.992154063042007e-05,
      "loss": 0.3045,
      "num_input_tokens_seen": 876752,
      "step": 1010
    },
    {
      "epoch": 0.47854785478547857,
      "grad_norm": 0.9493876695632935,
      "learning_rate": 4.992076152048136e-05,
      "loss": 0.3059,
      "num_input_tokens_seen": 880720,
      "step": 1015
    },
    {
      "epoch": 0.4809052333804809,
      "grad_norm": 1.2394379377365112,
      "learning_rate": 4.991997856744807e-05,
      "loss": 0.2493,
      "num_input_tokens_seen": 884528,
      "step": 1020
    },
    {
      "epoch": 0.48326261197548326,
      "grad_norm": 0.8788845539093018,
      "learning_rate": 4.9919191771440905e-05,
      "loss": 0.2457,
      "num_input_tokens_seen": 889360,
      "step": 1025
    },
    {
      "epoch": 0.4856199905704856,
      "grad_norm": 1.4016131162643433,
      "learning_rate": 4.991840113258122e-05,
      "loss": 0.2084,
      "num_input_tokens_seen": 893504,
      "step": 1030
    },
    {
      "epoch": 0.48797736916548795,
      "grad_norm": 2.3777902126312256,
      "learning_rate": 4.9917606650990933e-05,
      "loss": 0.2281,
      "num_input_tokens_seen": 897392,
      "step": 1035
    },
    {
      "epoch": 0.49033474776049035,
      "grad_norm": 1.1311967372894287,
      "learning_rate": 4.9916808326792566e-05,
      "loss": 0.166,
      "num_input_tokens_seen": 901952,
      "step": 1040
    },
    {
      "epoch": 0.4926921263554927,
      "grad_norm": 1.5846495628356934,
      "learning_rate": 4.9916006160109235e-05,
      "loss": 0.2432,
      "num_input_tokens_seen": 906720,
      "step": 1045
    },
    {
      "epoch": 0.49504950495049505,
      "grad_norm": 0.9230917096138,
      "learning_rate": 4.991520015106464e-05,
      "loss": 0.318,
      "num_input_tokens_seen": 911456,
      "step": 1050
    },
    {
      "epoch": 0.4974068835454974,
      "grad_norm": 3.483808994293213,
      "learning_rate": 4.991439029978308e-05,
      "loss": 0.1928,
      "num_input_tokens_seen": 915760,
      "step": 1055
    },
    {
      "epoch": 0.49976426214049974,
      "grad_norm": 2.6172831058502197,
      "learning_rate": 4.9913576606389434e-05,
      "loss": 0.291,
      "num_input_tokens_seen": 920016,
      "step": 1060
    },
    {
      "epoch": 0.5021216407355021,
      "grad_norm": 0.7803964018821716,
      "learning_rate": 4.991275907100919e-05,
      "loss": 0.1882,
      "num_input_tokens_seen": 924720,
      "step": 1065
    },
    {
      "epoch": 0.5044790193305044,
      "grad_norm": 0.8884827494621277,
      "learning_rate": 4.9911937693768434e-05,
      "loss": 0.1893,
      "num_input_tokens_seen": 928208,
      "step": 1070
    },
    {
      "epoch": 0.5068363979255068,
      "grad_norm": 1.9393093585968018,
      "learning_rate": 4.991111247479382e-05,
      "loss": 0.1837,
      "num_input_tokens_seen": 932064,
      "step": 1075
    },
    {
      "epoch": 0.5091937765205092,
      "grad_norm": 3.075648307800293,
      "learning_rate": 4.9910283414212605e-05,
      "loss": 0.3537,
      "num_input_tokens_seen": 935712,
      "step": 1080
    },
    {
      "epoch": 0.5115511551155115,
      "grad_norm": 1.0196256637573242,
      "learning_rate": 4.990945051215265e-05,
      "loss": 0.1532,
      "num_input_tokens_seen": 940096,
      "step": 1085
    },
    {
      "epoch": 0.5139085337105139,
      "grad_norm": 1.3363890647888184,
      "learning_rate": 4.99086137687424e-05,
      "loss": 0.2261,
      "num_input_tokens_seen": 943872,
      "step": 1090
    },
    {
      "epoch": 0.5162659123055162,
      "grad_norm": 1.8724160194396973,
      "learning_rate": 4.9907773184110874e-05,
      "loss": 0.2014,
      "num_input_tokens_seen": 949296,
      "step": 1095
    },
    {
      "epoch": 0.5186232909005186,
      "grad_norm": 1.863816499710083,
      "learning_rate": 4.9906928758387715e-05,
      "loss": 0.2016,
      "num_input_tokens_seen": 953040,
      "step": 1100
    },
    {
      "epoch": 0.520980669495521,
      "grad_norm": 1.4532032012939453,
      "learning_rate": 4.9906080491703146e-05,
      "loss": 0.2607,
      "num_input_tokens_seen": 957600,
      "step": 1105
    },
    {
      "epoch": 0.5233380480905233,
      "grad_norm": 0.9944967031478882,
      "learning_rate": 4.990522838418797e-05,
      "loss": 0.2063,
      "num_input_tokens_seen": 962496,
      "step": 1110
    },
    {
      "epoch": 0.5256954266855257,
      "grad_norm": 1.4057248830795288,
      "learning_rate": 4.9904372435973604e-05,
      "loss": 0.2596,
      "num_input_tokens_seen": 966688,
      "step": 1115
    },
    {
      "epoch": 0.528052805280528,
      "grad_norm": 1.4562263488769531,
      "learning_rate": 4.990351264719203e-05,
      "loss": 0.2304,
      "num_input_tokens_seen": 969984,
      "step": 1120
    },
    {
      "epoch": 0.5304101838755304,
      "grad_norm": 0.780207633972168,
      "learning_rate": 4.990264901797586e-05,
      "loss": 0.1552,
      "num_input_tokens_seen": 975280,
      "step": 1125
    },
    {
      "epoch": 0.5327675624705328,
      "grad_norm": 3.126469850540161,
      "learning_rate": 4.990178154845826e-05,
      "loss": 0.2075,
      "num_input_tokens_seen": 980064,
      "step": 1130
    },
    {
      "epoch": 0.5351249410655351,
      "grad_norm": 2.0905776023864746,
      "learning_rate": 4.9900910238773014e-05,
      "loss": 0.2707,
      "num_input_tokens_seen": 985024,
      "step": 1135
    },
    {
      "epoch": 0.5374823196605375,
      "grad_norm": 1.2073808908462524,
      "learning_rate": 4.990003508905448e-05,
      "loss": 0.2505,
      "num_input_tokens_seen": 989504,
      "step": 1140
    },
    {
      "epoch": 0.5398396982555398,
      "grad_norm": 3.2708380222320557,
      "learning_rate": 4.989915609943763e-05,
      "loss": 0.2285,
      "num_input_tokens_seen": 994448,
      "step": 1145
    },
    {
      "epoch": 0.5421970768505422,
      "grad_norm": 3.3755247592926025,
      "learning_rate": 4.9898273270058e-05,
      "loss": 0.2124,
      "num_input_tokens_seen": 998976,
      "step": 1150
    },
    {
      "epoch": 0.5445544554455446,
      "grad_norm": 1.7800930738449097,
      "learning_rate": 4.989738660105174e-05,
      "loss": 0.2934,
      "num_input_tokens_seen": 1003136,
      "step": 1155
    },
    {
      "epoch": 0.5469118340405469,
      "grad_norm": 1.3684277534484863,
      "learning_rate": 4.989649609255559e-05,
      "loss": 0.2754,
      "num_input_tokens_seen": 1006512,
      "step": 1160
    },
    {
      "epoch": 0.5492692126355493,
      "grad_norm": 1.515900731086731,
      "learning_rate": 4.989560174470687e-05,
      "loss": 0.1918,
      "num_input_tokens_seen": 1010576,
      "step": 1165
    },
    {
      "epoch": 0.5516265912305516,
      "grad_norm": 1.5553590059280396,
      "learning_rate": 4.989470355764351e-05,
      "loss": 0.2381,
      "num_input_tokens_seen": 1014928,
      "step": 1170
    },
    {
      "epoch": 0.553983969825554,
      "grad_norm": 2.0138468742370605,
      "learning_rate": 4.9893801531504e-05,
      "loss": 0.2259,
      "num_input_tokens_seen": 1019680,
      "step": 1175
    },
    {
      "epoch": 0.5563413484205564,
      "grad_norm": 1.521052598953247,
      "learning_rate": 4.9892895666427475e-05,
      "loss": 0.2538,
      "num_input_tokens_seen": 1024112,
      "step": 1180
    },
    {
      "epoch": 0.5586987270155587,
      "grad_norm": 3.820831537246704,
      "learning_rate": 4.9891985962553606e-05,
      "loss": 0.2549,
      "num_input_tokens_seen": 1028768,
      "step": 1185
    },
    {
      "epoch": 0.5610561056105611,
      "grad_norm": 2.6627018451690674,
      "learning_rate": 4.989107242002269e-05,
      "loss": 0.4376,
      "num_input_tokens_seen": 1032288,
      "step": 1190
    },
    {
      "epoch": 0.5634134842055634,
      "grad_norm": 1.0672571659088135,
      "learning_rate": 4.989015503897561e-05,
      "loss": 0.2737,
      "num_input_tokens_seen": 1036368,
      "step": 1195
    },
    {
      "epoch": 0.5657708628005658,
      "grad_norm": 0.7677374482154846,
      "learning_rate": 4.988923381955383e-05,
      "loss": 0.1707,
      "num_input_tokens_seen": 1040544,
      "step": 1200
    },
    {
      "epoch": 0.5657708628005658,
      "eval_loss": 0.24491755664348602,
      "eval_runtime": 22.1273,
      "eval_samples_per_second": 42.617,
      "eval_steps_per_second": 21.331,
      "num_input_tokens_seen": 1040544,
      "step": 1200
    },
    {
      "epoch": 0.5681282413955682,
      "grad_norm": 2.415074348449707,
      "learning_rate": 4.988830876189942e-05,
      "loss": 0.281,
      "num_input_tokens_seen": 1044576,
      "step": 1205
    },
    {
      "epoch": 0.5704856199905705,
      "grad_norm": 1.3753410577774048,
      "learning_rate": 4.988737986615503e-05,
      "loss": 0.2106,
      "num_input_tokens_seen": 1049408,
      "step": 1210
    },
    {
      "epoch": 0.5728429985855729,
      "grad_norm": 1.0319716930389404,
      "learning_rate": 4.988644713246391e-05,
      "loss": 0.1741,
      "num_input_tokens_seen": 1053424,
      "step": 1215
    },
    {
      "epoch": 0.5752003771805752,
      "grad_norm": 1.328358769416809,
      "learning_rate": 4.988551056096991e-05,
      "loss": 0.3099,
      "num_input_tokens_seen": 1058128,
      "step": 1220
    },
    {
      "epoch": 0.5775577557755776,
      "grad_norm": 1.2121775150299072,
      "learning_rate": 4.988457015181743e-05,
      "loss": 0.2369,
      "num_input_tokens_seen": 1062112,
      "step": 1225
    },
    {
      "epoch": 0.57991513437058,
      "grad_norm": 1.3732377290725708,
      "learning_rate": 4.988362590515153e-05,
      "loss": 0.2542,
      "num_input_tokens_seen": 1066544,
      "step": 1230
    },
    {
      "epoch": 0.5822725129655822,
      "grad_norm": 1.4912837743759155,
      "learning_rate": 4.9882677821117805e-05,
      "loss": 0.1533,
      "num_input_tokens_seen": 1071280,
      "step": 1235
    },
    {
      "epoch": 0.5846298915605846,
      "grad_norm": 2.000551700592041,
      "learning_rate": 4.988172589986246e-05,
      "loss": 0.288,
      "num_input_tokens_seen": 1075568,
      "step": 1240
    },
    {
      "epoch": 0.5869872701555869,
      "grad_norm": 0.8694466948509216,
      "learning_rate": 4.9880770141532304e-05,
      "loss": 0.1793,
      "num_input_tokens_seen": 1079648,
      "step": 1245
    },
    {
      "epoch": 0.5893446487505893,
      "grad_norm": 2.3243117332458496,
      "learning_rate": 4.987981054627472e-05,
      "loss": 0.2935,
      "num_input_tokens_seen": 1084080,
      "step": 1250
    },
    {
      "epoch": 0.5917020273455917,
      "grad_norm": 0.7803671360015869,
      "learning_rate": 4.987884711423769e-05,
      "loss": 0.1977,
      "num_input_tokens_seen": 1087680,
      "step": 1255
    },
    {
      "epoch": 0.594059405940594,
      "grad_norm": 0.8718262910842896,
      "learning_rate": 4.9877879845569784e-05,
      "loss": 0.2033,
      "num_input_tokens_seen": 1092080,
      "step": 1260
    },
    {
      "epoch": 0.5964167845355964,
      "grad_norm": 1.4951746463775635,
      "learning_rate": 4.9876908740420175e-05,
      "loss": 0.1818,
      "num_input_tokens_seen": 1096288,
      "step": 1265
    },
    {
      "epoch": 0.5987741631305987,
      "grad_norm": 0.8434339761734009,
      "learning_rate": 4.987593379893861e-05,
      "loss": 0.2264,
      "num_input_tokens_seen": 1099776,
      "step": 1270
    },
    {
      "epoch": 0.6011315417256011,
      "grad_norm": 1.8133784532546997,
      "learning_rate": 4.987495502127545e-05,
      "loss": 0.2422,
      "num_input_tokens_seen": 1103440,
      "step": 1275
    },
    {
      "epoch": 0.6034889203206035,
      "grad_norm": 1.6589186191558838,
      "learning_rate": 4.987397240758162e-05,
      "loss": 0.2876,
      "num_input_tokens_seen": 1107344,
      "step": 1280
    },
    {
      "epoch": 0.6058462989156058,
      "grad_norm": 2.028730630874634,
      "learning_rate": 4.9872985958008664e-05,
      "loss": 0.283,
      "num_input_tokens_seen": 1111728,
      "step": 1285
    },
    {
      "epoch": 0.6082036775106082,
      "grad_norm": 1.3455874919891357,
      "learning_rate": 4.987199567270871e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 1115376,
      "step": 1290
    },
    {
      "epoch": 0.6105610561056105,
      "grad_norm": 2.452136993408203,
      "learning_rate": 4.9871001551834444e-05,
      "loss": 0.1991,
      "num_input_tokens_seen": 1119184,
      "step": 1295
    },
    {
      "epoch": 0.6129184347006129,
      "grad_norm": 1.2062773704528809,
      "learning_rate": 4.98700035955392e-05,
      "loss": 0.2129,
      "num_input_tokens_seen": 1123600,
      "step": 1300
    },
    {
      "epoch": 0.6152758132956153,
      "grad_norm": 2.3022289276123047,
      "learning_rate": 4.986900180397686e-05,
      "loss": 0.2943,
      "num_input_tokens_seen": 1127408,
      "step": 1305
    },
    {
      "epoch": 0.6176331918906176,
      "grad_norm": 1.1147509813308716,
      "learning_rate": 4.9867996177301926e-05,
      "loss": 0.1681,
      "num_input_tokens_seen": 1131872,
      "step": 1310
    },
    {
      "epoch": 0.61999057048562,
      "grad_norm": 1.9456377029418945,
      "learning_rate": 4.9866986715669464e-05,
      "loss": 0.3529,
      "num_input_tokens_seen": 1138416,
      "step": 1315
    },
    {
      "epoch": 0.6223479490806223,
      "grad_norm": 1.8181740045547485,
      "learning_rate": 4.9865973419235155e-05,
      "loss": 0.1793,
      "num_input_tokens_seen": 1142144,
      "step": 1320
    },
    {
      "epoch": 0.6247053276756247,
      "grad_norm": 0.7284897565841675,
      "learning_rate": 4.986495628815526e-05,
      "loss": 0.2555,
      "num_input_tokens_seen": 1146544,
      "step": 1325
    },
    {
      "epoch": 0.6270627062706271,
      "grad_norm": 2.4954071044921875,
      "learning_rate": 4.986393532258663e-05,
      "loss": 0.2308,
      "num_input_tokens_seen": 1150784,
      "step": 1330
    },
    {
      "epoch": 0.6294200848656294,
      "grad_norm": 0.8742075562477112,
      "learning_rate": 4.986291052268671e-05,
      "loss": 0.2181,
      "num_input_tokens_seen": 1155120,
      "step": 1335
    },
    {
      "epoch": 0.6317774634606318,
      "grad_norm": 0.7873311042785645,
      "learning_rate": 4.986188188861355e-05,
      "loss": 0.2222,
      "num_input_tokens_seen": 1160256,
      "step": 1340
    },
    {
      "epoch": 0.6341348420556341,
      "grad_norm": 0.7422676682472229,
      "learning_rate": 4.9860849420525766e-05,
      "loss": 0.2585,
      "num_input_tokens_seen": 1164544,
      "step": 1345
    },
    {
      "epoch": 0.6364922206506365,
      "grad_norm": 2.2914233207702637,
      "learning_rate": 4.9859813118582575e-05,
      "loss": 0.2227,
      "num_input_tokens_seen": 1168464,
      "step": 1350
    },
    {
      "epoch": 0.6388495992456389,
      "grad_norm": 2.7506697177886963,
      "learning_rate": 4.98587729829438e-05,
      "loss": 0.2239,
      "num_input_tokens_seen": 1171888,
      "step": 1355
    },
    {
      "epoch": 0.6412069778406412,
      "grad_norm": 0.6760467290878296,
      "learning_rate": 4.985772901376983e-05,
      "loss": 0.2218,
      "num_input_tokens_seen": 1175920,
      "step": 1360
    },
    {
      "epoch": 0.6435643564356436,
      "grad_norm": 1.7387597560882568,
      "learning_rate": 4.9856681211221666e-05,
      "loss": 0.215,
      "num_input_tokens_seen": 1179792,
      "step": 1365
    },
    {
      "epoch": 0.6459217350306459,
      "grad_norm": 0.8078663349151611,
      "learning_rate": 4.985562957546089e-05,
      "loss": 0.3207,
      "num_input_tokens_seen": 1185264,
      "step": 1370
    },
    {
      "epoch": 0.6482791136256483,
      "grad_norm": 0.9883490800857544,
      "learning_rate": 4.9854574106649686e-05,
      "loss": 0.2463,
      "num_input_tokens_seen": 1189472,
      "step": 1375
    },
    {
      "epoch": 0.6506364922206507,
      "grad_norm": 1.8638012409210205,
      "learning_rate": 4.985351480495081e-05,
      "loss": 0.2194,
      "num_input_tokens_seen": 1193104,
      "step": 1380
    },
    {
      "epoch": 0.652993870815653,
      "grad_norm": 1.8103914260864258,
      "learning_rate": 4.985245167052762e-05,
      "loss": 0.2439,
      "num_input_tokens_seen": 1197856,
      "step": 1385
    },
    {
      "epoch": 0.6553512494106554,
      "grad_norm": 1.808821678161621,
      "learning_rate": 4.9851384703544066e-05,
      "loss": 0.1996,
      "num_input_tokens_seen": 1203328,
      "step": 1390
    },
    {
      "epoch": 0.6577086280056577,
      "grad_norm": 1.0076441764831543,
      "learning_rate": 4.985031390416469e-05,
      "loss": 0.3225,
      "num_input_tokens_seen": 1207376,
      "step": 1395
    },
    {
      "epoch": 0.6600660066006601,
      "grad_norm": 3.5843911170959473,
      "learning_rate": 4.984923927255461e-05,
      "loss": 0.3671,
      "num_input_tokens_seen": 1211680,
      "step": 1400
    },
    {
      "epoch": 0.6600660066006601,
      "eval_loss": 0.23904502391815186,
      "eval_runtime": 22.178,
      "eval_samples_per_second": 42.52,
      "eval_steps_per_second": 21.282,
      "num_input_tokens_seen": 1211680,
      "step": 1400
    },
    {
      "epoch": 0.6624233851956625,
      "grad_norm": 2.044574022293091,
      "learning_rate": 4.984816080887958e-05,
      "loss": 0.2736,
      "num_input_tokens_seen": 1215600,
      "step": 1405
    },
    {
      "epoch": 0.6647807637906648,
      "grad_norm": 1.527148962020874,
      "learning_rate": 4.9847078513305875e-05,
      "loss": 0.2104,
      "num_input_tokens_seen": 1220256,
      "step": 1410
    },
    {
      "epoch": 0.6671381423856672,
      "grad_norm": 0.8457903265953064,
      "learning_rate": 4.984599238600043e-05,
      "loss": 0.1965,
      "num_input_tokens_seen": 1224464,
      "step": 1415
    },
    {
      "epoch": 0.6694955209806694,
      "grad_norm": 2.1995668411254883,
      "learning_rate": 4.9844902427130716e-05,
      "loss": 0.3176,
      "num_input_tokens_seen": 1228464,
      "step": 1420
    },
    {
      "epoch": 0.6718528995756718,
      "grad_norm": 1.0541527271270752,
      "learning_rate": 4.984380863686482e-05,
      "loss": 0.1523,
      "num_input_tokens_seen": 1232640,
      "step": 1425
    },
    {
      "epoch": 0.6742102781706742,
      "grad_norm": 1.2533800601959229,
      "learning_rate": 4.984271101537143e-05,
      "loss": 0.2306,
      "num_input_tokens_seen": 1237248,
      "step": 1430
    },
    {
      "epoch": 0.6765676567656765,
      "grad_norm": 1.0373866558074951,
      "learning_rate": 4.9841609562819816e-05,
      "loss": 0.2001,
      "num_input_tokens_seen": 1241056,
      "step": 1435
    },
    {
      "epoch": 0.6789250353606789,
      "grad_norm": 0.7156577110290527,
      "learning_rate": 4.984050427937983e-05,
      "loss": 0.1685,
      "num_input_tokens_seen": 1245296,
      "step": 1440
    },
    {
      "epoch": 0.6812824139556812,
      "grad_norm": 1.8790751695632935,
      "learning_rate": 4.983939516522191e-05,
      "loss": 0.153,
      "num_input_tokens_seen": 1249600,
      "step": 1445
    },
    {
      "epoch": 0.6836397925506836,
      "grad_norm": 0.7536019682884216,
      "learning_rate": 4.983828222051711e-05,
      "loss": 0.2054,
      "num_input_tokens_seen": 1253616,
      "step": 1450
    },
    {
      "epoch": 0.685997171145686,
      "grad_norm": 1.7218577861785889,
      "learning_rate": 4.983716544543705e-05,
      "loss": 0.312,
      "num_input_tokens_seen": 1257392,
      "step": 1455
    },
    {
      "epoch": 0.6883545497406883,
      "grad_norm": 3.25358247756958,
      "learning_rate": 4.983604484015395e-05,
      "loss": 0.2346,
      "num_input_tokens_seen": 1261776,
      "step": 1460
    },
    {
      "epoch": 0.6907119283356907,
      "grad_norm": 1.1682939529418945,
      "learning_rate": 4.983492040484064e-05,
      "loss": 0.1636,
      "num_input_tokens_seen": 1265888,
      "step": 1465
    },
    {
      "epoch": 0.693069306930693,
      "grad_norm": 0.9088102579116821,
      "learning_rate": 4.98337921396705e-05,
      "loss": 0.19,
      "num_input_tokens_seen": 1269712,
      "step": 1470
    },
    {
      "epoch": 0.6954266855256954,
      "grad_norm": 0.709406852722168,
      "learning_rate": 4.983266004481753e-05,
      "loss": 0.1858,
      "num_input_tokens_seen": 1274464,
      "step": 1475
    },
    {
      "epoch": 0.6977840641206978,
      "grad_norm": 1.1896185874938965,
      "learning_rate": 4.9831524120456316e-05,
      "loss": 0.1846,
      "num_input_tokens_seen": 1278944,
      "step": 1480
    },
    {
      "epoch": 0.7001414427157001,
      "grad_norm": 2.4044058322906494,
      "learning_rate": 4.9830384366762026e-05,
      "loss": 0.2436,
      "num_input_tokens_seen": 1282608,
      "step": 1485
    },
    {
      "epoch": 0.7024988213107025,
      "grad_norm": 0.8826987147331238,
      "learning_rate": 4.9829240783910436e-05,
      "loss": 0.1813,
      "num_input_tokens_seen": 1286480,
      "step": 1490
    },
    {
      "epoch": 0.7048561999057048,
      "grad_norm": 2.7571372985839844,
      "learning_rate": 4.982809337207789e-05,
      "loss": 0.2373,
      "num_input_tokens_seen": 1290560,
      "step": 1495
    },
    {
      "epoch": 0.7072135785007072,
      "grad_norm": 0.8249913454055786,
      "learning_rate": 4.9826942131441337e-05,
      "loss": 0.1937,
      "num_input_tokens_seen": 1294240,
      "step": 1500
    },
    {
      "epoch": 0.7095709570957096,
      "grad_norm": 0.4897255301475525,
      "learning_rate": 4.9825787062178315e-05,
      "loss": 0.3611,
      "num_input_tokens_seen": 1298512,
      "step": 1505
    },
    {
      "epoch": 0.7119283356907119,
      "grad_norm": 0.6323021054267883,
      "learning_rate": 4.9824628164466945e-05,
      "loss": 0.2113,
      "num_input_tokens_seen": 1302736,
      "step": 1510
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 1.6747448444366455,
      "learning_rate": 4.982346543848595e-05,
      "loss": 0.1777,
      "num_input_tokens_seen": 1307376,
      "step": 1515
    },
    {
      "epoch": 0.7166430928807166,
      "grad_norm": 0.947551429271698,
      "learning_rate": 4.9822298884414626e-05,
      "loss": 0.1211,
      "num_input_tokens_seen": 1311760,
      "step": 1520
    },
    {
      "epoch": 0.719000471475719,
      "grad_norm": 0.732955276966095,
      "learning_rate": 4.982112850243288e-05,
      "loss": 0.1685,
      "num_input_tokens_seen": 1315200,
      "step": 1525
    },
    {
      "epoch": 0.7213578500707214,
      "grad_norm": 1.3198305368423462,
      "learning_rate": 4.98199542927212e-05,
      "loss": 0.2228,
      "num_input_tokens_seen": 1319008,
      "step": 1530
    },
    {
      "epoch": 0.7237152286657237,
      "grad_norm": 1.1603840589523315,
      "learning_rate": 4.981877625546066e-05,
      "loss": 0.2588,
      "num_input_tokens_seen": 1323696,
      "step": 1535
    },
    {
      "epoch": 0.7260726072607261,
      "grad_norm": 2.7100141048431396,
      "learning_rate": 4.981759439083293e-05,
      "loss": 0.2116,
      "num_input_tokens_seen": 1327264,
      "step": 1540
    },
    {
      "epoch": 0.7284299858557284,
      "grad_norm": 1.5490788221359253,
      "learning_rate": 4.981640869902027e-05,
      "loss": 0.2015,
      "num_input_tokens_seen": 1331200,
      "step": 1545
    },
    {
      "epoch": 0.7307873644507308,
      "grad_norm": 2.5623066425323486,
      "learning_rate": 4.9815219180205517e-05,
      "loss": 0.2577,
      "num_input_tokens_seen": 1336144,
      "step": 1550
    },
    {
      "epoch": 0.7331447430457332,
      "grad_norm": 2.405183792114258,
      "learning_rate": 4.9814025834572126e-05,
      "loss": 0.2815,
      "num_input_tokens_seen": 1340688,
      "step": 1555
    },
    {
      "epoch": 0.7355021216407355,
      "grad_norm": 1.3060182332992554,
      "learning_rate": 4.981282866230411e-05,
      "loss": 0.2366,
      "num_input_tokens_seen": 1344960,
      "step": 1560
    },
    {
      "epoch": 0.7378595002357379,
      "grad_norm": 0.9142674207687378,
      "learning_rate": 4.981162766358611e-05,
      "loss": 0.2506,
      "num_input_tokens_seen": 1349968,
      "step": 1565
    },
    {
      "epoch": 0.7402168788307402,
      "grad_norm": 1.0979033708572388,
      "learning_rate": 4.9810422838603316e-05,
      "loss": 0.2193,
      "num_input_tokens_seen": 1354240,
      "step": 1570
    },
    {
      "epoch": 0.7425742574257426,
      "grad_norm": 0.8448873162269592,
      "learning_rate": 4.9809214187541533e-05,
      "loss": 0.1605,
      "num_input_tokens_seen": 1359232,
      "step": 1575
    },
    {
      "epoch": 0.744931636020745,
      "grad_norm": 0.7138634324073792,
      "learning_rate": 4.980800171058715e-05,
      "loss": 0.2599,
      "num_input_tokens_seen": 1363408,
      "step": 1580
    },
    {
      "epoch": 0.7472890146157473,
      "grad_norm": 1.1225407123565674,
      "learning_rate": 4.980678540792715e-05,
      "loss": 0.204,
      "num_input_tokens_seen": 1368176,
      "step": 1585
    },
    {
      "epoch": 0.7496463932107497,
      "grad_norm": 3.225644826889038,
      "learning_rate": 4.980556527974909e-05,
      "loss": 0.33,
      "num_input_tokens_seen": 1372752,
      "step": 1590
    },
    {
      "epoch": 0.752003771805752,
      "grad_norm": 1.391526699066162,
      "learning_rate": 4.980434132624114e-05,
      "loss": 0.1735,
      "num_input_tokens_seen": 1376640,
      "step": 1595
    },
    {
      "epoch": 0.7543611504007544,
      "grad_norm": 0.9788062572479248,
      "learning_rate": 4.980311354759205e-05,
      "loss": 0.1354,
      "num_input_tokens_seen": 1381792,
      "step": 1600
    },
    {
      "epoch": 0.7543611504007544,
      "eval_loss": 0.23937085270881653,
      "eval_runtime": 22.1571,
      "eval_samples_per_second": 42.56,
      "eval_steps_per_second": 21.302,
      "num_input_tokens_seen": 1381792,
      "step": 1600
    },
    {
      "epoch": 0.7567185289957568,
      "grad_norm": 1.191501259803772,
      "learning_rate": 4.980188194399116e-05,
      "loss": 0.1755,
      "num_input_tokens_seen": 1385968,
      "step": 1605
    },
    {
      "epoch": 0.759075907590759,
      "grad_norm": 0.9974446296691895,
      "learning_rate": 4.9800646515628384e-05,
      "loss": 0.1706,
      "num_input_tokens_seen": 1390256,
      "step": 1610
    },
    {
      "epoch": 0.7614332861857614,
      "grad_norm": 1.0779097080230713,
      "learning_rate": 4.979940726269426e-05,
      "loss": 0.2472,
      "num_input_tokens_seen": 1394272,
      "step": 1615
    },
    {
      "epoch": 0.7637906647807637,
      "grad_norm": 1.6718465089797974,
      "learning_rate": 4.979816418537988e-05,
      "loss": 0.2372,
      "num_input_tokens_seen": 1398672,
      "step": 1620
    },
    {
      "epoch": 0.7661480433757661,
      "grad_norm": 2.1119191646575928,
      "learning_rate": 4.979691728387696e-05,
      "loss": 0.2017,
      "num_input_tokens_seen": 1403168,
      "step": 1625
    },
    {
      "epoch": 0.7685054219707685,
      "grad_norm": 1.2721500396728516,
      "learning_rate": 4.979566655837776e-05,
      "loss": 0.1727,
      "num_input_tokens_seen": 1407792,
      "step": 1630
    },
    {
      "epoch": 0.7708628005657708,
      "grad_norm": 0.6625446081161499,
      "learning_rate": 4.9794412009075184e-05,
      "loss": 0.1785,
      "num_input_tokens_seen": 1412080,
      "step": 1635
    },
    {
      "epoch": 0.7732201791607732,
      "grad_norm": 0.9371739029884338,
      "learning_rate": 4.979315363616269e-05,
      "loss": 0.1926,
      "num_input_tokens_seen": 1416752,
      "step": 1640
    },
    {
      "epoch": 0.7755775577557755,
      "grad_norm": 0.5679917335510254,
      "learning_rate": 4.979189143983434e-05,
      "loss": 0.2486,
      "num_input_tokens_seen": 1421728,
      "step": 1645
    },
    {
      "epoch": 0.7779349363507779,
      "grad_norm": 0.728076159954071,
      "learning_rate": 4.979062542028478e-05,
      "loss": 0.2132,
      "num_input_tokens_seen": 1425744,
      "step": 1650
    },
    {
      "epoch": 0.7802923149457803,
      "grad_norm": 2.09423828125,
      "learning_rate": 4.978935557770923e-05,
      "loss": 0.1557,
      "num_input_tokens_seen": 1430528,
      "step": 1655
    },
    {
      "epoch": 0.7826496935407826,
      "grad_norm": 1.475656509399414,
      "learning_rate": 4.978808191230353e-05,
      "loss": 0.2558,
      "num_input_tokens_seen": 1435376,
      "step": 1660
    },
    {
      "epoch": 0.785007072135785,
      "grad_norm": 4.4559645652771,
      "learning_rate": 4.9786804424264085e-05,
      "loss": 0.2833,
      "num_input_tokens_seen": 1440272,
      "step": 1665
    },
    {
      "epoch": 0.7873644507307873,
      "grad_norm": 1.6104339361190796,
      "learning_rate": 4.978552311378792e-05,
      "loss": 0.3398,
      "num_input_tokens_seen": 1444384,
      "step": 1670
    },
    {
      "epoch": 0.7897218293257897,
      "grad_norm": 0.7869780659675598,
      "learning_rate": 4.978423798107261e-05,
      "loss": 0.2487,
      "num_input_tokens_seen": 1449056,
      "step": 1675
    },
    {
      "epoch": 0.7920792079207921,
      "grad_norm": 1.27755868434906,
      "learning_rate": 4.978294902631635e-05,
      "loss": 0.1942,
      "num_input_tokens_seen": 1453104,
      "step": 1680
    },
    {
      "epoch": 0.7944365865157944,
      "grad_norm": 1.9019169807434082,
      "learning_rate": 4.9781656249717914e-05,
      "loss": 0.2361,
      "num_input_tokens_seen": 1457520,
      "step": 1685
    },
    {
      "epoch": 0.7967939651107968,
      "grad_norm": 0.6511656641960144,
      "learning_rate": 4.9780359651476645e-05,
      "loss": 0.1609,
      "num_input_tokens_seen": 1461984,
      "step": 1690
    },
    {
      "epoch": 0.7991513437057991,
      "grad_norm": 1.8626477718353271,
      "learning_rate": 4.977905923179251e-05,
      "loss": 0.2465,
      "num_input_tokens_seen": 1465952,
      "step": 1695
    },
    {
      "epoch": 0.8015087223008015,
      "grad_norm": 1.5253952741622925,
      "learning_rate": 4.977775499086606e-05,
      "loss": 0.2497,
      "num_input_tokens_seen": 1470480,
      "step": 1700
    },
    {
      "epoch": 0.8038661008958039,
      "grad_norm": 0.4835023581981659,
      "learning_rate": 4.97764469288984e-05,
      "loss": 0.2062,
      "num_input_tokens_seen": 1474736,
      "step": 1705
    },
    {
      "epoch": 0.8062234794908062,
      "grad_norm": 2.709240436553955,
      "learning_rate": 4.977513504609127e-05,
      "loss": 0.276,
      "num_input_tokens_seen": 1478944,
      "step": 1710
    },
    {
      "epoch": 0.8085808580858086,
      "grad_norm": 1.4659289121627808,
      "learning_rate": 4.9773819342646965e-05,
      "loss": 0.2137,
      "num_input_tokens_seen": 1484288,
      "step": 1715
    },
    {
      "epoch": 0.8109382366808109,
      "grad_norm": 0.7182785868644714,
      "learning_rate": 4.97724998187684e-05,
      "loss": 0.2184,
      "num_input_tokens_seen": 1489008,
      "step": 1720
    },
    {
      "epoch": 0.8132956152758133,
      "grad_norm": 1.163142442703247,
      "learning_rate": 4.9771176474659045e-05,
      "loss": 0.286,
      "num_input_tokens_seen": 1494672,
      "step": 1725
    },
    {
      "epoch": 0.8156529938708157,
      "grad_norm": 1.4184327125549316,
      "learning_rate": 4.976984931052299e-05,
      "loss": 0.2293,
      "num_input_tokens_seen": 1499072,
      "step": 1730
    },
    {
      "epoch": 0.818010372465818,
      "grad_norm": 0.9214722514152527,
      "learning_rate": 4.976851832656489e-05,
      "loss": 0.1583,
      "num_input_tokens_seen": 1502816,
      "step": 1735
    },
    {
      "epoch": 0.8203677510608204,
      "grad_norm": 1.7758578062057495,
      "learning_rate": 4.9767183522990004e-05,
      "loss": 0.2245,
      "num_input_tokens_seen": 1506736,
      "step": 1740
    },
    {
      "epoch": 0.8227251296558227,
      "grad_norm": 1.282928705215454,
      "learning_rate": 4.9765844900004176e-05,
      "loss": 0.2202,
      "num_input_tokens_seen": 1510992,
      "step": 1745
    },
    {
      "epoch": 0.8250825082508251,
      "grad_norm": 1.1504408121109009,
      "learning_rate": 4.9764502457813834e-05,
      "loss": 0.1886,
      "num_input_tokens_seen": 1515056,
      "step": 1750
    },
    {
      "epoch": 0.8274398868458275,
      "grad_norm": 0.9731258153915405,
      "learning_rate": 4.9763156196626005e-05,
      "loss": 0.2109,
      "num_input_tokens_seen": 1520048,
      "step": 1755
    },
    {
      "epoch": 0.8297972654408298,
      "grad_norm": 1.8562312126159668,
      "learning_rate": 4.97618061166483e-05,
      "loss": 0.2113,
      "num_input_tokens_seen": 1524176,
      "step": 1760
    },
    {
      "epoch": 0.8321546440358322,
      "grad_norm": 0.9318592548370361,
      "learning_rate": 4.9760452218088915e-05,
      "loss": 0.3031,
      "num_input_tokens_seen": 1528784,
      "step": 1765
    },
    {
      "epoch": 0.8345120226308345,
      "grad_norm": 1.806275486946106,
      "learning_rate": 4.975909450115663e-05,
      "loss": 0.2762,
      "num_input_tokens_seen": 1533328,
      "step": 1770
    },
    {
      "epoch": 0.8368694012258369,
      "grad_norm": 1.975741982460022,
      "learning_rate": 4.975773296606084e-05,
      "loss": 0.2587,
      "num_input_tokens_seen": 1538064,
      "step": 1775
    },
    {
      "epoch": 0.8392267798208393,
      "grad_norm": 1.424389362335205,
      "learning_rate": 4.97563676130115e-05,
      "loss": 0.2426,
      "num_input_tokens_seen": 1541952,
      "step": 1780
    },
    {
      "epoch": 0.8415841584158416,
      "grad_norm": 1.7898634672164917,
      "learning_rate": 4.9754998442219166e-05,
      "loss": 0.1778,
      "num_input_tokens_seen": 1546032,
      "step": 1785
    },
    {
      "epoch": 0.843941537010844,
      "grad_norm": 0.7613176703453064,
      "learning_rate": 4.9753625453894984e-05,
      "loss": 0.1752,
      "num_input_tokens_seen": 1550480,
      "step": 1790
    },
    {
      "epoch": 0.8462989156058462,
      "grad_norm": 1.2258604764938354,
      "learning_rate": 4.975224864825068e-05,
      "loss": 0.1726,
      "num_input_tokens_seen": 1554848,
      "step": 1795
    },
    {
      "epoch": 0.8486562942008486,
      "grad_norm": 1.4619604349136353,
      "learning_rate": 4.9750868025498576e-05,
      "loss": 0.2328,
      "num_input_tokens_seen": 1559456,
      "step": 1800
    },
    {
      "epoch": 0.8486562942008486,
      "eval_loss": 0.2352200150489807,
      "eval_runtime": 22.1452,
      "eval_samples_per_second": 42.583,
      "eval_steps_per_second": 21.314,
      "num_input_tokens_seen": 1559456,
      "step": 1800
    },
    {
      "epoch": 0.851013672795851,
      "grad_norm": 1.1367679834365845,
      "learning_rate": 4.974948358585158e-05,
      "loss": 0.2173,
      "num_input_tokens_seen": 1563360,
      "step": 1805
    },
    {
      "epoch": 0.8533710513908533,
      "grad_norm": 0.863335371017456,
      "learning_rate": 4.9748095329523205e-05,
      "loss": 0.1531,
      "num_input_tokens_seen": 1567232,
      "step": 1810
    },
    {
      "epoch": 0.8557284299858557,
      "grad_norm": 1.1062192916870117,
      "learning_rate": 4.974670325672752e-05,
      "loss": 0.3052,
      "num_input_tokens_seen": 1571392,
      "step": 1815
    },
    {
      "epoch": 0.858085808580858,
      "grad_norm": 1.2048113346099854,
      "learning_rate": 4.974530736767921e-05,
      "loss": 0.2455,
      "num_input_tokens_seen": 1575600,
      "step": 1820
    },
    {
      "epoch": 0.8604431871758604,
      "grad_norm": 2.5278494358062744,
      "learning_rate": 4.9743907662593524e-05,
      "loss": 0.322,
      "num_input_tokens_seen": 1580640,
      "step": 1825
    },
    {
      "epoch": 0.8628005657708628,
      "grad_norm": 1.3441838026046753,
      "learning_rate": 4.974250414168633e-05,
      "loss": 0.166,
      "num_input_tokens_seen": 1584656,
      "step": 1830
    },
    {
      "epoch": 0.8651579443658651,
      "grad_norm": 1.4045212268829346,
      "learning_rate": 4.974109680517407e-05,
      "loss": 0.2722,
      "num_input_tokens_seen": 1588976,
      "step": 1835
    },
    {
      "epoch": 0.8675153229608675,
      "grad_norm": 0.5986180901527405,
      "learning_rate": 4.973968565327376e-05,
      "loss": 0.216,
      "num_input_tokens_seen": 1592928,
      "step": 1840
    },
    {
      "epoch": 0.8698727015558698,
      "grad_norm": 0.5707735419273376,
      "learning_rate": 4.973827068620303e-05,
      "loss": 0.1707,
      "num_input_tokens_seen": 1596880,
      "step": 1845
    },
    {
      "epoch": 0.8722300801508722,
      "grad_norm": 1.365468978881836,
      "learning_rate": 4.973685190418008e-05,
      "loss": 0.2016,
      "num_input_tokens_seen": 1601440,
      "step": 1850
    },
    {
      "epoch": 0.8745874587458746,
      "grad_norm": 1.2912050485610962,
      "learning_rate": 4.97354293074237e-05,
      "loss": 0.2171,
      "num_input_tokens_seen": 1606032,
      "step": 1855
    },
    {
      "epoch": 0.8769448373408769,
      "grad_norm": 2.057765007019043,
      "learning_rate": 4.9734002896153276e-05,
      "loss": 0.228,
      "num_input_tokens_seen": 1610720,
      "step": 1860
    },
    {
      "epoch": 0.8793022159358793,
      "grad_norm": 1.58944571018219,
      "learning_rate": 4.973257267058877e-05,
      "loss": 0.2054,
      "num_input_tokens_seen": 1615520,
      "step": 1865
    },
    {
      "epoch": 0.8816595945308816,
      "grad_norm": 0.5646544694900513,
      "learning_rate": 4.973113863095076e-05,
      "loss": 0.293,
      "num_input_tokens_seen": 1620416,
      "step": 1870
    },
    {
      "epoch": 0.884016973125884,
      "grad_norm": 0.7927470803260803,
      "learning_rate": 4.9729700777460384e-05,
      "loss": 0.212,
      "num_input_tokens_seen": 1624368,
      "step": 1875
    },
    {
      "epoch": 0.8863743517208864,
      "grad_norm": 1.6710447072982788,
      "learning_rate": 4.972825911033937e-05,
      "loss": 0.2972,
      "num_input_tokens_seen": 1628656,
      "step": 1880
    },
    {
      "epoch": 0.8887317303158887,
      "grad_norm": 1.2268353700637817,
      "learning_rate": 4.9726813629810056e-05,
      "loss": 0.2334,
      "num_input_tokens_seen": 1633072,
      "step": 1885
    },
    {
      "epoch": 0.8910891089108911,
      "grad_norm": 3.282341480255127,
      "learning_rate": 4.9725364336095326e-05,
      "loss": 0.2906,
      "num_input_tokens_seen": 1637328,
      "step": 1890
    },
    {
      "epoch": 0.8934464875058934,
      "grad_norm": 1.073418140411377,
      "learning_rate": 4.972391122941871e-05,
      "loss": 0.1532,
      "num_input_tokens_seen": 1642304,
      "step": 1895
    },
    {
      "epoch": 0.8958038661008958,
      "grad_norm": 1.2532000541687012,
      "learning_rate": 4.972245431000428e-05,
      "loss": 0.2125,
      "num_input_tokens_seen": 1646656,
      "step": 1900
    },
    {
      "epoch": 0.8981612446958982,
      "grad_norm": 0.6340424418449402,
      "learning_rate": 4.972099357807671e-05,
      "loss": 0.1558,
      "num_input_tokens_seen": 1650720,
      "step": 1905
    },
    {
      "epoch": 0.9005186232909005,
      "grad_norm": 1.403266429901123,
      "learning_rate": 4.971952903386127e-05,
      "loss": 0.3016,
      "num_input_tokens_seen": 1655488,
      "step": 1910
    },
    {
      "epoch": 0.9028760018859029,
      "grad_norm": 0.44591474533081055,
      "learning_rate": 4.971806067758381e-05,
      "loss": 0.2291,
      "num_input_tokens_seen": 1660016,
      "step": 1915
    },
    {
      "epoch": 0.9052333804809052,
      "grad_norm": 0.8210023045539856,
      "learning_rate": 4.971658850947076e-05,
      "loss": 0.1532,
      "num_input_tokens_seen": 1664448,
      "step": 1920
    },
    {
      "epoch": 0.9075907590759076,
      "grad_norm": 2.554323673248291,
      "learning_rate": 4.9715112529749165e-05,
      "loss": 0.2247,
      "num_input_tokens_seen": 1669312,
      "step": 1925
    },
    {
      "epoch": 0.90994813767091,
      "grad_norm": 0.7805220484733582,
      "learning_rate": 4.9713632738646624e-05,
      "loss": 0.1884,
      "num_input_tokens_seen": 1673424,
      "step": 1930
    },
    {
      "epoch": 0.9123055162659123,
      "grad_norm": 2.1348607540130615,
      "learning_rate": 4.971214913639134e-05,
      "loss": 0.2721,
      "num_input_tokens_seen": 1677760,
      "step": 1935
    },
    {
      "epoch": 0.9146628948609147,
      "grad_norm": 0.8901115655899048,
      "learning_rate": 4.9710661723212104e-05,
      "loss": 0.2414,
      "num_input_tokens_seen": 1682064,
      "step": 1940
    },
    {
      "epoch": 0.917020273455917,
      "grad_norm": 0.6565680503845215,
      "learning_rate": 4.9709170499338295e-05,
      "loss": 0.1919,
      "num_input_tokens_seen": 1686752,
      "step": 1945
    },
    {
      "epoch": 0.9193776520509194,
      "grad_norm": 0.765413761138916,
      "learning_rate": 4.9707675464999895e-05,
      "loss": 0.1588,
      "num_input_tokens_seen": 1691232,
      "step": 1950
    },
    {
      "epoch": 0.9217350306459218,
      "grad_norm": 0.6776729822158813,
      "learning_rate": 4.970617662042743e-05,
      "loss": 0.268,
      "num_input_tokens_seen": 1695440,
      "step": 1955
    },
    {
      "epoch": 0.9240924092409241,
      "grad_norm": 0.6181092858314514,
      "learning_rate": 4.970467396585206e-05,
      "loss": 0.1952,
      "num_input_tokens_seen": 1699344,
      "step": 1960
    },
    {
      "epoch": 0.9264497878359265,
      "grad_norm": 1.7263247966766357,
      "learning_rate": 4.97031675015055e-05,
      "loss": 0.2643,
      "num_input_tokens_seen": 1704560,
      "step": 1965
    },
    {
      "epoch": 0.9288071664309288,
      "grad_norm": 2.2068843841552734,
      "learning_rate": 4.9701657227620075e-05,
      "loss": 0.1738,
      "num_input_tokens_seen": 1709312,
      "step": 1970
    },
    {
      "epoch": 0.9311645450259312,
      "grad_norm": 0.6493487358093262,
      "learning_rate": 4.9700143144428685e-05,
      "loss": 0.1927,
      "num_input_tokens_seen": 1714064,
      "step": 1975
    },
    {
      "epoch": 0.9335219236209336,
      "grad_norm": 0.7149948477745056,
      "learning_rate": 4.969862525216482e-05,
      "loss": 0.2612,
      "num_input_tokens_seen": 1718688,
      "step": 1980
    },
    {
      "epoch": 0.9358793022159358,
      "grad_norm": 0.8569145798683167,
      "learning_rate": 4.9697103551062556e-05,
      "loss": 0.1733,
      "num_input_tokens_seen": 1723744,
      "step": 1985
    },
    {
      "epoch": 0.9382366808109383,
      "grad_norm": 0.9474877715110779,
      "learning_rate": 4.9695578041356565e-05,
      "loss": 0.2014,
      "num_input_tokens_seen": 1727824,
      "step": 1990
    },
    {
      "epoch": 0.9405940594059405,
      "grad_norm": 0.7371253967285156,
      "learning_rate": 4.969404872328209e-05,
      "loss": 0.1237,
      "num_input_tokens_seen": 1731728,
      "step": 1995
    },
    {
      "epoch": 0.9429514380009429,
      "grad_norm": 1.8017866611480713,
      "learning_rate": 4.969251559707498e-05,
      "loss": 0.2788,
      "num_input_tokens_seen": 1735840,
      "step": 2000
    },
    {
      "epoch": 0.9429514380009429,
      "eval_loss": 0.2355743944644928,
      "eval_runtime": 22.1956,
      "eval_samples_per_second": 42.486,
      "eval_steps_per_second": 21.265,
      "num_input_tokens_seen": 1735840,
      "step": 2000
    },
    {
      "epoch": 0.9453088165959453,
      "grad_norm": 2.085132598876953,
      "learning_rate": 4.9690978662971674e-05,
      "loss": 0.3085,
      "num_input_tokens_seen": 1739856,
      "step": 2005
    },
    {
      "epoch": 0.9476661951909476,
      "grad_norm": 1.903670072555542,
      "learning_rate": 4.968943792120916e-05,
      "loss": 0.2409,
      "num_input_tokens_seen": 1745168,
      "step": 2010
    },
    {
      "epoch": 0.95002357378595,
      "grad_norm": 1.5028831958770752,
      "learning_rate": 4.9687893372025046e-05,
      "loss": 0.2971,
      "num_input_tokens_seen": 1749936,
      "step": 2015
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 1.570249319076538,
      "learning_rate": 4.9686345015657535e-05,
      "loss": 0.1983,
      "num_input_tokens_seen": 1754432,
      "step": 2020
    },
    {
      "epoch": 0.9547383309759547,
      "grad_norm": 0.7205819487571716,
      "learning_rate": 4.968479285234538e-05,
      "loss": 0.2091,
      "num_input_tokens_seen": 1758960,
      "step": 2025
    },
    {
      "epoch": 0.9570957095709571,
      "grad_norm": 0.7245256900787354,
      "learning_rate": 4.9683236882327974e-05,
      "loss": 0.1508,
      "num_input_tokens_seen": 1762704,
      "step": 2030
    },
    {
      "epoch": 0.9594530881659594,
      "grad_norm": 1.3722597360610962,
      "learning_rate": 4.968167710584526e-05,
      "loss": 0.2405,
      "num_input_tokens_seen": 1767632,
      "step": 2035
    },
    {
      "epoch": 0.9618104667609618,
      "grad_norm": 0.7442454099655151,
      "learning_rate": 4.968011352313775e-05,
      "loss": 0.2233,
      "num_input_tokens_seen": 1772032,
      "step": 2040
    },
    {
      "epoch": 0.9641678453559641,
      "grad_norm": 1.1451423168182373,
      "learning_rate": 4.967854613444659e-05,
      "loss": 0.189,
      "num_input_tokens_seen": 1776112,
      "step": 2045
    },
    {
      "epoch": 0.9665252239509665,
      "grad_norm": 0.7528586983680725,
      "learning_rate": 4.967697494001349e-05,
      "loss": 0.2425,
      "num_input_tokens_seen": 1780480,
      "step": 2050
    },
    {
      "epoch": 0.9688826025459689,
      "grad_norm": 0.7649926543235779,
      "learning_rate": 4.9675399940080736e-05,
      "loss": 0.2109,
      "num_input_tokens_seen": 1784672,
      "step": 2055
    },
    {
      "epoch": 0.9712399811409712,
      "grad_norm": 1.2557573318481445,
      "learning_rate": 4.9673821134891226e-05,
      "loss": 0.2106,
      "num_input_tokens_seen": 1789056,
      "step": 2060
    },
    {
      "epoch": 0.9735973597359736,
      "grad_norm": 1.6368476152420044,
      "learning_rate": 4.967223852468842e-05,
      "loss": 0.2531,
      "num_input_tokens_seen": 1793776,
      "step": 2065
    },
    {
      "epoch": 0.9759547383309759,
      "grad_norm": 2.272239923477173,
      "learning_rate": 4.967065210971639e-05,
      "loss": 0.3074,
      "num_input_tokens_seen": 1798928,
      "step": 2070
    },
    {
      "epoch": 0.9783121169259783,
      "grad_norm": 1.8925212621688843,
      "learning_rate": 4.966906189021977e-05,
      "loss": 0.2953,
      "num_input_tokens_seen": 1803120,
      "step": 2075
    },
    {
      "epoch": 0.9806694955209807,
      "grad_norm": 1.5512921810150146,
      "learning_rate": 4.966746786644379e-05,
      "loss": 0.2564,
      "num_input_tokens_seen": 1807664,
      "step": 2080
    },
    {
      "epoch": 0.983026874115983,
      "grad_norm": 0.5901377201080322,
      "learning_rate": 4.966587003863429e-05,
      "loss": 0.1881,
      "num_input_tokens_seen": 1811344,
      "step": 2085
    },
    {
      "epoch": 0.9853842527109854,
      "grad_norm": 1.0474426746368408,
      "learning_rate": 4.966426840703765e-05,
      "loss": 0.2586,
      "num_input_tokens_seen": 1815472,
      "step": 2090
    },
    {
      "epoch": 0.9877416313059877,
      "grad_norm": 1.5125291347503662,
      "learning_rate": 4.9662662971900875e-05,
      "loss": 0.2674,
      "num_input_tokens_seen": 1819744,
      "step": 2095
    },
    {
      "epoch": 0.9900990099009901,
      "grad_norm": 1.6254551410675049,
      "learning_rate": 4.9661053733471534e-05,
      "loss": 0.2438,
      "num_input_tokens_seen": 1823792,
      "step": 2100
    },
    {
      "epoch": 0.9924563884959925,
      "grad_norm": 1.4071815013885498,
      "learning_rate": 4.965944069199781e-05,
      "loss": 0.2607,
      "num_input_tokens_seen": 1827632,
      "step": 2105
    },
    {
      "epoch": 0.9948137670909948,
      "grad_norm": 1.89168381690979,
      "learning_rate": 4.965782384772842e-05,
      "loss": 0.207,
      "num_input_tokens_seen": 1831376,
      "step": 2110
    },
    {
      "epoch": 0.9971711456859972,
      "grad_norm": 0.7630700469017029,
      "learning_rate": 4.9656203200912734e-05,
      "loss": 0.1847,
      "num_input_tokens_seen": 1835840,
      "step": 2115
    },
    {
      "epoch": 0.9995285242809995,
      "grad_norm": 1.5064918994903564,
      "learning_rate": 4.965457875180067e-05,
      "loss": 0.2081,
      "num_input_tokens_seen": 1840144,
      "step": 2120
    },
    {
      "epoch": 1.0018859028760019,
      "grad_norm": 0.9992051124572754,
      "learning_rate": 4.9652950500642724e-05,
      "loss": 0.2816,
      "num_input_tokens_seen": 1844720,
      "step": 2125
    },
    {
      "epoch": 1.0042432814710043,
      "grad_norm": 2.067429542541504,
      "learning_rate": 4.965131844769001e-05,
      "loss": 0.2126,
      "num_input_tokens_seen": 1848592,
      "step": 2130
    },
    {
      "epoch": 1.0066006600660067,
      "grad_norm": 1.359427571296692,
      "learning_rate": 4.96496825931942e-05,
      "loss": 0.224,
      "num_input_tokens_seen": 1853344,
      "step": 2135
    },
    {
      "epoch": 1.0089580386610089,
      "grad_norm": 1.003531813621521,
      "learning_rate": 4.9648042937407566e-05,
      "loss": 0.2689,
      "num_input_tokens_seen": 1858352,
      "step": 2140
    },
    {
      "epoch": 1.0113154172560113,
      "grad_norm": 0.731319010257721,
      "learning_rate": 4.964639948058297e-05,
      "loss": 0.1971,
      "num_input_tokens_seen": 1863120,
      "step": 2145
    },
    {
      "epoch": 1.0136727958510137,
      "grad_norm": 1.3148454427719116,
      "learning_rate": 4.9644752222973846e-05,
      "loss": 0.1486,
      "num_input_tokens_seen": 1868640,
      "step": 2150
    },
    {
      "epoch": 1.016030174446016,
      "grad_norm": 1.3683695793151855,
      "learning_rate": 4.964310116483422e-05,
      "loss": 0.1415,
      "num_input_tokens_seen": 1873088,
      "step": 2155
    },
    {
      "epoch": 1.0183875530410185,
      "grad_norm": 0.7409855127334595,
      "learning_rate": 4.964144630641872e-05,
      "loss": 0.2266,
      "num_input_tokens_seen": 1877552,
      "step": 2160
    },
    {
      "epoch": 1.0207449316360206,
      "grad_norm": 2.367750883102417,
      "learning_rate": 4.9639787647982525e-05,
      "loss": 0.2872,
      "num_input_tokens_seen": 1882240,
      "step": 2165
    },
    {
      "epoch": 1.023102310231023,
      "grad_norm": 1.8922744989395142,
      "learning_rate": 4.963812518978143e-05,
      "loss": 0.2371,
      "num_input_tokens_seen": 1887120,
      "step": 2170
    },
    {
      "epoch": 1.0254596888260255,
      "grad_norm": 0.6507037281990051,
      "learning_rate": 4.963645893207182e-05,
      "loss": 0.1883,
      "num_input_tokens_seen": 1891312,
      "step": 2175
    },
    {
      "epoch": 1.0278170674210279,
      "grad_norm": 1.4713044166564941,
      "learning_rate": 4.963478887511063e-05,
      "loss": 0.1729,
      "num_input_tokens_seen": 1895360,
      "step": 2180
    },
    {
      "epoch": 1.0301744460160303,
      "grad_norm": 3.6618893146514893,
      "learning_rate": 4.963311501915542e-05,
      "loss": 0.2639,
      "num_input_tokens_seen": 1898704,
      "step": 2185
    },
    {
      "epoch": 1.0325318246110324,
      "grad_norm": 0.4606444537639618,
      "learning_rate": 4.963143736446432e-05,
      "loss": 0.2122,
      "num_input_tokens_seen": 1903328,
      "step": 2190
    },
    {
      "epoch": 1.0348892032060348,
      "grad_norm": 0.5474140048027039,
      "learning_rate": 4.962975591129603e-05,
      "loss": 0.1988,
      "num_input_tokens_seen": 1907024,
      "step": 2195
    },
    {
      "epoch": 1.0372465818010372,
      "grad_norm": 1.572401762008667,
      "learning_rate": 4.962807065990986e-05,
      "loss": 0.2671,
      "num_input_tokens_seen": 1910848,
      "step": 2200
    },
    {
      "epoch": 1.0372465818010372,
      "eval_loss": 0.2288239747285843,
      "eval_runtime": 22.1552,
      "eval_samples_per_second": 42.563,
      "eval_steps_per_second": 21.304,
      "num_input_tokens_seen": 1910848,
      "step": 2200
    },
    {
      "epoch": 1.0396039603960396,
      "grad_norm": 0.5859895944595337,
      "learning_rate": 4.9626381610565714e-05,
      "loss": 0.2543,
      "num_input_tokens_seen": 1914672,
      "step": 2205
    },
    {
      "epoch": 1.041961338991042,
      "grad_norm": 2.2224879264831543,
      "learning_rate": 4.9624688763524043e-05,
      "loss": 0.2093,
      "num_input_tokens_seen": 1918880,
      "step": 2210
    },
    {
      "epoch": 1.0443187175860442,
      "grad_norm": 1.0996782779693604,
      "learning_rate": 4.962299211904591e-05,
      "loss": 0.2337,
      "num_input_tokens_seen": 1923024,
      "step": 2215
    },
    {
      "epoch": 1.0466760961810466,
      "grad_norm": 1.0060921907424927,
      "learning_rate": 4.962129167739296e-05,
      "loss": 0.2445,
      "num_input_tokens_seen": 1927360,
      "step": 2220
    },
    {
      "epoch": 1.049033474776049,
      "grad_norm": 0.8287821412086487,
      "learning_rate": 4.961958743882742e-05,
      "loss": 0.2215,
      "num_input_tokens_seen": 1931776,
      "step": 2225
    },
    {
      "epoch": 1.0513908533710514,
      "grad_norm": 0.9771628379821777,
      "learning_rate": 4.961787940361211e-05,
      "loss": 0.1992,
      "num_input_tokens_seen": 1935376,
      "step": 2230
    },
    {
      "epoch": 1.0537482319660538,
      "grad_norm": 0.95954430103302,
      "learning_rate": 4.961616757201043e-05,
      "loss": 0.1289,
      "num_input_tokens_seen": 1939552,
      "step": 2235
    },
    {
      "epoch": 1.056105610561056,
      "grad_norm": 1.0297051668167114,
      "learning_rate": 4.961445194428637e-05,
      "loss": 0.218,
      "num_input_tokens_seen": 1944192,
      "step": 2240
    },
    {
      "epoch": 1.0584629891560584,
      "grad_norm": 0.44284993410110474,
      "learning_rate": 4.9612732520704486e-05,
      "loss": 0.1644,
      "num_input_tokens_seen": 1948656,
      "step": 2245
    },
    {
      "epoch": 1.0608203677510608,
      "grad_norm": 2.4091708660125732,
      "learning_rate": 4.961100930152994e-05,
      "loss": 0.1729,
      "num_input_tokens_seen": 1953344,
      "step": 2250
    },
    {
      "epoch": 1.0631777463460632,
      "grad_norm": 1.2394448518753052,
      "learning_rate": 4.960928228702849e-05,
      "loss": 0.2447,
      "num_input_tokens_seen": 1957392,
      "step": 2255
    },
    {
      "epoch": 1.0655351249410656,
      "grad_norm": 1.0420091152191162,
      "learning_rate": 4.960755147746645e-05,
      "loss": 0.2185,
      "num_input_tokens_seen": 1961440,
      "step": 2260
    },
    {
      "epoch": 1.0678925035360678,
      "grad_norm": 0.7688211798667908,
      "learning_rate": 4.9605816873110736e-05,
      "loss": 0.2116,
      "num_input_tokens_seen": 1965504,
      "step": 2265
    },
    {
      "epoch": 1.0702498821310702,
      "grad_norm": 0.4416621923446655,
      "learning_rate": 4.960407847422883e-05,
      "loss": 0.2512,
      "num_input_tokens_seen": 1969152,
      "step": 2270
    },
    {
      "epoch": 1.0726072607260726,
      "grad_norm": 0.7483367919921875,
      "learning_rate": 4.960233628108885e-05,
      "loss": 0.1813,
      "num_input_tokens_seen": 1973952,
      "step": 2275
    },
    {
      "epoch": 1.074964639321075,
      "grad_norm": 0.7397369742393494,
      "learning_rate": 4.960059029395942e-05,
      "loss": 0.3193,
      "num_input_tokens_seen": 1977520,
      "step": 2280
    },
    {
      "epoch": 1.0773220179160774,
      "grad_norm": 0.6446831822395325,
      "learning_rate": 4.959884051310983e-05,
      "loss": 0.2147,
      "num_input_tokens_seen": 1981664,
      "step": 2285
    },
    {
      "epoch": 1.0796793965110796,
      "grad_norm": 1.271741271018982,
      "learning_rate": 4.959708693880991e-05,
      "loss": 0.1872,
      "num_input_tokens_seen": 1986304,
      "step": 2290
    },
    {
      "epoch": 1.082036775106082,
      "grad_norm": 0.6644223928451538,
      "learning_rate": 4.9595329571330074e-05,
      "loss": 0.1849,
      "num_input_tokens_seen": 1990400,
      "step": 2295
    },
    {
      "epoch": 1.0843941537010844,
      "grad_norm": 0.4015240967273712,
      "learning_rate": 4.9593568410941326e-05,
      "loss": 0.2006,
      "num_input_tokens_seen": 1994368,
      "step": 2300
    },
    {
      "epoch": 1.0867515322960868,
      "grad_norm": 1.593414068222046,
      "learning_rate": 4.959180345791528e-05,
      "loss": 0.2144,
      "num_input_tokens_seen": 1999104,
      "step": 2305
    },
    {
      "epoch": 1.0891089108910892,
      "grad_norm": 1.0523892641067505,
      "learning_rate": 4.9590034712524086e-05,
      "loss": 0.294,
      "num_input_tokens_seen": 2003248,
      "step": 2310
    },
    {
      "epoch": 1.0914662894860914,
      "grad_norm": 0.8189030289649963,
      "learning_rate": 4.958826217504053e-05,
      "loss": 0.1432,
      "num_input_tokens_seen": 2007040,
      "step": 2315
    },
    {
      "epoch": 1.0938236680810938,
      "grad_norm": 1.4127891063690186,
      "learning_rate": 4.958648584573795e-05,
      "loss": 0.2355,
      "num_input_tokens_seen": 2010720,
      "step": 2320
    },
    {
      "epoch": 1.0961810466760962,
      "grad_norm": 0.638702929019928,
      "learning_rate": 4.958470572489028e-05,
      "loss": 0.1719,
      "num_input_tokens_seen": 2014752,
      "step": 2325
    },
    {
      "epoch": 1.0985384252710986,
      "grad_norm": 1.4768091440200806,
      "learning_rate": 4.958292181277203e-05,
      "loss": 0.2055,
      "num_input_tokens_seen": 2019632,
      "step": 2330
    },
    {
      "epoch": 1.100895803866101,
      "grad_norm": 1.6569145917892456,
      "learning_rate": 4.958113410965832e-05,
      "loss": 0.2116,
      "num_input_tokens_seen": 2023712,
      "step": 2335
    },
    {
      "epoch": 1.1032531824611032,
      "grad_norm": 0.9050070643424988,
      "learning_rate": 4.957934261582481e-05,
      "loss": 0.1689,
      "num_input_tokens_seen": 2027680,
      "step": 2340
    },
    {
      "epoch": 1.1056105610561056,
      "grad_norm": 1.7313966751098633,
      "learning_rate": 4.95775473315478e-05,
      "loss": 0.2994,
      "num_input_tokens_seen": 2032704,
      "step": 2345
    },
    {
      "epoch": 1.107967939651108,
      "grad_norm": 1.2064248323440552,
      "learning_rate": 4.9575748257104124e-05,
      "loss": 0.2374,
      "num_input_tokens_seen": 2036160,
      "step": 2350
    },
    {
      "epoch": 1.1103253182461104,
      "grad_norm": 1.2235831022262573,
      "learning_rate": 4.9573945392771224e-05,
      "loss": 0.2115,
      "num_input_tokens_seen": 2040240,
      "step": 2355
    },
    {
      "epoch": 1.1126826968411128,
      "grad_norm": 1.3807519674301147,
      "learning_rate": 4.9572138738827134e-05,
      "loss": 0.2221,
      "num_input_tokens_seen": 2046224,
      "step": 2360
    },
    {
      "epoch": 1.115040075436115,
      "grad_norm": 0.9015531539916992,
      "learning_rate": 4.957032829555046e-05,
      "loss": 0.3071,
      "num_input_tokens_seen": 2050400,
      "step": 2365
    },
    {
      "epoch": 1.1173974540311173,
      "grad_norm": 1.0194159746170044,
      "learning_rate": 4.956851406322039e-05,
      "loss": 0.2798,
      "num_input_tokens_seen": 2055904,
      "step": 2370
    },
    {
      "epoch": 1.1197548326261197,
      "grad_norm": 0.9910465478897095,
      "learning_rate": 4.9566696042116704e-05,
      "loss": 0.1876,
      "num_input_tokens_seen": 2060208,
      "step": 2375
    },
    {
      "epoch": 1.1221122112211221,
      "grad_norm": 1.3440930843353271,
      "learning_rate": 4.9564874232519766e-05,
      "loss": 0.2543,
      "num_input_tokens_seen": 2064128,
      "step": 2380
    },
    {
      "epoch": 1.1244695898161245,
      "grad_norm": 0.8332763314247131,
      "learning_rate": 4.9563048634710516e-05,
      "loss": 0.1994,
      "num_input_tokens_seen": 2068448,
      "step": 2385
    },
    {
      "epoch": 1.1268269684111267,
      "grad_norm": 1.384052038192749,
      "learning_rate": 4.956121924897049e-05,
      "loss": 0.1938,
      "num_input_tokens_seen": 2072784,
      "step": 2390
    },
    {
      "epoch": 1.1291843470061291,
      "grad_norm": 1.1407479047775269,
      "learning_rate": 4.955938607558181e-05,
      "loss": 0.1787,
      "num_input_tokens_seen": 2076656,
      "step": 2395
    },
    {
      "epoch": 1.1315417256011315,
      "grad_norm": 1.045023798942566,
      "learning_rate": 4.955754911482715e-05,
      "loss": 0.2308,
      "num_input_tokens_seen": 2081696,
      "step": 2400
    },
    {
      "epoch": 1.1315417256011315,
      "eval_loss": 0.2234393209218979,
      "eval_runtime": 22.1733,
      "eval_samples_per_second": 42.529,
      "eval_steps_per_second": 21.287,
      "num_input_tokens_seen": 2081696,
      "step": 2400
    },
    {
      "epoch": 1.133899104196134,
      "grad_norm": 0.769090473651886,
      "learning_rate": 4.9555708366989804e-05,
      "loss": 0.2247,
      "num_input_tokens_seen": 2085824,
      "step": 2405
    },
    {
      "epoch": 1.1362564827911363,
      "grad_norm": 0.46435117721557617,
      "learning_rate": 4.9553863832353655e-05,
      "loss": 0.1918,
      "num_input_tokens_seen": 2090032,
      "step": 2410
    },
    {
      "epoch": 1.1386138613861387,
      "grad_norm": 0.7540394067764282,
      "learning_rate": 4.955201551120313e-05,
      "loss": 0.2167,
      "num_input_tokens_seen": 2093968,
      "step": 2415
    },
    {
      "epoch": 1.140971239981141,
      "grad_norm": 0.7453724145889282,
      "learning_rate": 4.955016340382328e-05,
      "loss": 0.2905,
      "num_input_tokens_seen": 2098352,
      "step": 2420
    },
    {
      "epoch": 1.1433286185761433,
      "grad_norm": 0.6741037964820862,
      "learning_rate": 4.954830751049972e-05,
      "loss": 0.2323,
      "num_input_tokens_seen": 2102624,
      "step": 2425
    },
    {
      "epoch": 1.1456859971711457,
      "grad_norm": 0.6348397731781006,
      "learning_rate": 4.954644783151864e-05,
      "loss": 0.2836,
      "num_input_tokens_seen": 2106672,
      "step": 2430
    },
    {
      "epoch": 1.1480433757661481,
      "grad_norm": 0.637897789478302,
      "learning_rate": 4.954458436716684e-05,
      "loss": 0.1352,
      "num_input_tokens_seen": 2111472,
      "step": 2435
    },
    {
      "epoch": 1.1504007543611503,
      "grad_norm": 1.1868523359298706,
      "learning_rate": 4.954271711773168e-05,
      "loss": 0.246,
      "num_input_tokens_seen": 2116624,
      "step": 2440
    },
    {
      "epoch": 1.1527581329561527,
      "grad_norm": 0.8872718214988708,
      "learning_rate": 4.9540846083501115e-05,
      "loss": 0.1926,
      "num_input_tokens_seen": 2121456,
      "step": 2445
    },
    {
      "epoch": 1.155115511551155,
      "grad_norm": 0.7805548310279846,
      "learning_rate": 4.953897126476369e-05,
      "loss": 0.1856,
      "num_input_tokens_seen": 2125808,
      "step": 2450
    },
    {
      "epoch": 1.1574728901461575,
      "grad_norm": 1.4534517526626587,
      "learning_rate": 4.9537092661808514e-05,
      "loss": 0.2028,
      "num_input_tokens_seen": 2131184,
      "step": 2455
    },
    {
      "epoch": 1.15983026874116,
      "grad_norm": 1.5156935453414917,
      "learning_rate": 4.9535210274925306e-05,
      "loss": 0.21,
      "num_input_tokens_seen": 2135600,
      "step": 2460
    },
    {
      "epoch": 1.1621876473361623,
      "grad_norm": 1.080426573753357,
      "learning_rate": 4.953332410440435e-05,
      "loss": 0.2063,
      "num_input_tokens_seen": 2140848,
      "step": 2465
    },
    {
      "epoch": 1.1645450259311645,
      "grad_norm": 2.0126283168792725,
      "learning_rate": 4.9531434150536496e-05,
      "loss": 0.1714,
      "num_input_tokens_seen": 2144624,
      "step": 2470
    },
    {
      "epoch": 1.166902404526167,
      "grad_norm": 1.0352797508239746,
      "learning_rate": 4.952954041361322e-05,
      "loss": 0.1772,
      "num_input_tokens_seen": 2148752,
      "step": 2475
    },
    {
      "epoch": 1.1692597831211693,
      "grad_norm": 0.9237061738967896,
      "learning_rate": 4.952764289392655e-05,
      "loss": 0.2278,
      "num_input_tokens_seen": 2152624,
      "step": 2480
    },
    {
      "epoch": 1.1716171617161717,
      "grad_norm": 1.9642796516418457,
      "learning_rate": 4.952574159176912e-05,
      "loss": 0.1773,
      "num_input_tokens_seen": 2157200,
      "step": 2485
    },
    {
      "epoch": 1.1739745403111739,
      "grad_norm": 1.1122689247131348,
      "learning_rate": 4.952383650743413e-05,
      "loss": 0.2153,
      "num_input_tokens_seen": 2161664,
      "step": 2490
    },
    {
      "epoch": 1.1763319189061763,
      "grad_norm": 1.3984203338623047,
      "learning_rate": 4.952192764121536e-05,
      "loss": 0.1824,
      "num_input_tokens_seen": 2166464,
      "step": 2495
    },
    {
      "epoch": 1.1786892975011787,
      "grad_norm": 1.7451406717300415,
      "learning_rate": 4.9520014993407185e-05,
      "loss": 0.1426,
      "num_input_tokens_seen": 2170304,
      "step": 2500
    },
    {
      "epoch": 1.181046676096181,
      "grad_norm": 1.5566778182983398,
      "learning_rate": 4.951809856430456e-05,
      "loss": 0.3006,
      "num_input_tokens_seen": 2175696,
      "step": 2505
    },
    {
      "epoch": 1.1834040546911835,
      "grad_norm": 1.7959206104278564,
      "learning_rate": 4.951617835420303e-05,
      "loss": 0.1582,
      "num_input_tokens_seen": 2180240,
      "step": 2510
    },
    {
      "epoch": 1.1857614332861859,
      "grad_norm": 3.3532297611236572,
      "learning_rate": 4.951425436339869e-05,
      "loss": 0.2423,
      "num_input_tokens_seen": 2184736,
      "step": 2515
    },
    {
      "epoch": 1.188118811881188,
      "grad_norm": 1.7361382246017456,
      "learning_rate": 4.9512326592188274e-05,
      "loss": 0.2041,
      "num_input_tokens_seen": 2189072,
      "step": 2520
    },
    {
      "epoch": 1.1904761904761905,
      "grad_norm": 1.2550708055496216,
      "learning_rate": 4.9510395040869054e-05,
      "loss": 0.2487,
      "num_input_tokens_seen": 2193680,
      "step": 2525
    },
    {
      "epoch": 1.1928335690711929,
      "grad_norm": 0.9398348331451416,
      "learning_rate": 4.9508459709738905e-05,
      "loss": 0.2934,
      "num_input_tokens_seen": 2197968,
      "step": 2530
    },
    {
      "epoch": 1.1951909476661953,
      "grad_norm": 1.186946988105774,
      "learning_rate": 4.950652059909627e-05,
      "loss": 0.2264,
      "num_input_tokens_seen": 2202096,
      "step": 2535
    },
    {
      "epoch": 1.1975483262611974,
      "grad_norm": 0.7050871253013611,
      "learning_rate": 4.95045777092402e-05,
      "loss": 0.2058,
      "num_input_tokens_seen": 2205728,
      "step": 2540
    },
    {
      "epoch": 1.1999057048561999,
      "grad_norm": 1.635620355606079,
      "learning_rate": 4.950263104047031e-05,
      "loss": 0.1934,
      "num_input_tokens_seen": 2210128,
      "step": 2545
    },
    {
      "epoch": 1.2022630834512023,
      "grad_norm": 1.224055290222168,
      "learning_rate": 4.9500680593086775e-05,
      "loss": 0.2044,
      "num_input_tokens_seen": 2214336,
      "step": 2550
    },
    {
      "epoch": 1.2046204620462047,
      "grad_norm": 0.7304818630218506,
      "learning_rate": 4.94987263673904e-05,
      "loss": 0.1949,
      "num_input_tokens_seen": 2217968,
      "step": 2555
    },
    {
      "epoch": 1.206977840641207,
      "grad_norm": 1.029937505722046,
      "learning_rate": 4.949676836368256e-05,
      "loss": 0.1416,
      "num_input_tokens_seen": 2221888,
      "step": 2560
    },
    {
      "epoch": 1.2093352192362095,
      "grad_norm": 1.411577820777893,
      "learning_rate": 4.949480658226518e-05,
      "loss": 0.1933,
      "num_input_tokens_seen": 2226000,
      "step": 2565
    },
    {
      "epoch": 1.2116925978312116,
      "grad_norm": 1.04108726978302,
      "learning_rate": 4.949284102344082e-05,
      "loss": 0.2348,
      "num_input_tokens_seen": 2230160,
      "step": 2570
    },
    {
      "epoch": 1.214049976426214,
      "grad_norm": 1.3631505966186523,
      "learning_rate": 4.9490871687512565e-05,
      "loss": 0.2419,
      "num_input_tokens_seen": 2234192,
      "step": 2575
    },
    {
      "epoch": 1.2164073550212164,
      "grad_norm": 0.6250306963920593,
      "learning_rate": 4.948889857478413e-05,
      "loss": 0.2129,
      "num_input_tokens_seen": 2238400,
      "step": 2580
    },
    {
      "epoch": 1.2187647336162188,
      "grad_norm": 1.2595652341842651,
      "learning_rate": 4.948692168555978e-05,
      "loss": 0.2381,
      "num_input_tokens_seen": 2242608,
      "step": 2585
    },
    {
      "epoch": 1.221122112211221,
      "grad_norm": 0.8371504545211792,
      "learning_rate": 4.94849410201444e-05,
      "loss": 0.2133,
      "num_input_tokens_seen": 2247104,
      "step": 2590
    },
    {
      "epoch": 1.2234794908062234,
      "grad_norm": 0.9940767288208008,
      "learning_rate": 4.948295657884341e-05,
      "loss": 0.1994,
      "num_input_tokens_seen": 2251328,
      "step": 2595
    },
    {
      "epoch": 1.2258368694012258,
      "grad_norm": 0.5490002036094666,
      "learning_rate": 4.9480968361962835e-05,
      "loss": 0.2644,
      "num_input_tokens_seen": 2255952,
      "step": 2600
    },
    {
      "epoch": 1.2258368694012258,
      "eval_loss": 0.22386981546878815,
      "eval_runtime": 22.1854,
      "eval_samples_per_second": 42.505,
      "eval_steps_per_second": 21.275,
      "num_input_tokens_seen": 2255952,
      "step": 2600
    },
    {
      "epoch": 1.2281942479962282,
      "grad_norm": 0.7931286096572876,
      "learning_rate": 4.9478976369809305e-05,
      "loss": 0.1627,
      "num_input_tokens_seen": 2260160,
      "step": 2605
    },
    {
      "epoch": 1.2305516265912306,
      "grad_norm": 0.6786996722221375,
      "learning_rate": 4.947698060268999e-05,
      "loss": 0.1969,
      "num_input_tokens_seen": 2264016,
      "step": 2610
    },
    {
      "epoch": 1.232909005186233,
      "grad_norm": 1.1479096412658691,
      "learning_rate": 4.9474981060912665e-05,
      "loss": 0.1942,
      "num_input_tokens_seen": 2267856,
      "step": 2615
    },
    {
      "epoch": 1.2352663837812352,
      "grad_norm": 2.0785529613494873,
      "learning_rate": 4.94729777447857e-05,
      "loss": 0.2668,
      "num_input_tokens_seen": 2271808,
      "step": 2620
    },
    {
      "epoch": 1.2376237623762376,
      "grad_norm": 1.1667317152023315,
      "learning_rate": 4.947097065461801e-05,
      "loss": 0.1898,
      "num_input_tokens_seen": 2276208,
      "step": 2625
    },
    {
      "epoch": 1.23998114097124,
      "grad_norm": 0.8903427720069885,
      "learning_rate": 4.9468959790719125e-05,
      "loss": 0.1624,
      "num_input_tokens_seen": 2279968,
      "step": 2630
    },
    {
      "epoch": 1.2423385195662424,
      "grad_norm": 0.9771891236305237,
      "learning_rate": 4.9466945153399146e-05,
      "loss": 0.2395,
      "num_input_tokens_seen": 2284208,
      "step": 2635
    },
    {
      "epoch": 1.2446958981612446,
      "grad_norm": 1.3387407064437866,
      "learning_rate": 4.9464926742968755e-05,
      "loss": 0.1923,
      "num_input_tokens_seen": 2288944,
      "step": 2640
    },
    {
      "epoch": 1.247053276756247,
      "grad_norm": 1.2967472076416016,
      "learning_rate": 4.946290455973921e-05,
      "loss": 0.2377,
      "num_input_tokens_seen": 2293008,
      "step": 2645
    },
    {
      "epoch": 1.2494106553512494,
      "grad_norm": 2.4240176677703857,
      "learning_rate": 4.9460878604022365e-05,
      "loss": 0.2569,
      "num_input_tokens_seen": 2298240,
      "step": 2650
    },
    {
      "epoch": 1.2517680339462518,
      "grad_norm": 0.7176705598831177,
      "learning_rate": 4.945884887613065e-05,
      "loss": 0.1775,
      "num_input_tokens_seen": 2302832,
      "step": 2655
    },
    {
      "epoch": 1.2541254125412542,
      "grad_norm": 0.8490504026412964,
      "learning_rate": 4.9456815376377055e-05,
      "loss": 0.2178,
      "num_input_tokens_seen": 2306576,
      "step": 2660
    },
    {
      "epoch": 1.2564827911362566,
      "grad_norm": 0.8513333797454834,
      "learning_rate": 4.9454778105075195e-05,
      "loss": 0.2485,
      "num_input_tokens_seen": 2310784,
      "step": 2665
    },
    {
      "epoch": 1.2588401697312588,
      "grad_norm": 1.7787768840789795,
      "learning_rate": 4.945273706253924e-05,
      "loss": 0.3157,
      "num_input_tokens_seen": 2314912,
      "step": 2670
    },
    {
      "epoch": 1.2611975483262612,
      "grad_norm": 1.6174514293670654,
      "learning_rate": 4.9450692249083925e-05,
      "loss": 0.186,
      "num_input_tokens_seen": 2319344,
      "step": 2675
    },
    {
      "epoch": 1.2635549269212636,
      "grad_norm": 1.5588632822036743,
      "learning_rate": 4.9448643665024605e-05,
      "loss": 0.2643,
      "num_input_tokens_seen": 2323568,
      "step": 2680
    },
    {
      "epoch": 1.265912305516266,
      "grad_norm": 0.9553709626197815,
      "learning_rate": 4.944659131067719e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 2328048,
      "step": 2685
    },
    {
      "epoch": 1.2682696841112682,
      "grad_norm": 2.907095432281494,
      "learning_rate": 4.944453518635818e-05,
      "loss": 0.2793,
      "num_input_tokens_seen": 2331712,
      "step": 2690
    },
    {
      "epoch": 1.2706270627062706,
      "grad_norm": 1.9502835273742676,
      "learning_rate": 4.944247529238465e-05,
      "loss": 0.2431,
      "num_input_tokens_seen": 2335808,
      "step": 2695
    },
    {
      "epoch": 1.272984441301273,
      "grad_norm": 2.5764660835266113,
      "learning_rate": 4.944041162907427e-05,
      "loss": 0.2239,
      "num_input_tokens_seen": 2340032,
      "step": 2700
    },
    {
      "epoch": 1.2753418198962754,
      "grad_norm": 1.5685383081436157,
      "learning_rate": 4.943834419674529e-05,
      "loss": 0.2497,
      "num_input_tokens_seen": 2344096,
      "step": 2705
    },
    {
      "epoch": 1.2776991984912778,
      "grad_norm": 1.4056397676467896,
      "learning_rate": 4.9436272995716506e-05,
      "loss": 0.1915,
      "num_input_tokens_seen": 2348096,
      "step": 2710
    },
    {
      "epoch": 1.2800565770862802,
      "grad_norm": 0.675119161605835,
      "learning_rate": 4.943419802630735e-05,
      "loss": 0.2595,
      "num_input_tokens_seen": 2352304,
      "step": 2715
    },
    {
      "epoch": 1.2824139556812824,
      "grad_norm": 0.7126198410987854,
      "learning_rate": 4.94321192888378e-05,
      "loss": 0.2135,
      "num_input_tokens_seen": 2356736,
      "step": 2720
    },
    {
      "epoch": 1.2847713342762848,
      "grad_norm": 1.591935157775879,
      "learning_rate": 4.943003678362842e-05,
      "loss": 0.1795,
      "num_input_tokens_seen": 2361088,
      "step": 2725
    },
    {
      "epoch": 1.2871287128712872,
      "grad_norm": 1.8206456899642944,
      "learning_rate": 4.942795051100036e-05,
      "loss": 0.2705,
      "num_input_tokens_seen": 2364976,
      "step": 2730
    },
    {
      "epoch": 1.2894860914662896,
      "grad_norm": 0.684511125087738,
      "learning_rate": 4.942586047127536e-05,
      "loss": 0.2099,
      "num_input_tokens_seen": 2369792,
      "step": 2735
    },
    {
      "epoch": 1.2918434700612917,
      "grad_norm": 1.0367815494537354,
      "learning_rate": 4.942376666477571e-05,
      "loss": 0.194,
      "num_input_tokens_seen": 2374064,
      "step": 2740
    },
    {
      "epoch": 1.2942008486562941,
      "grad_norm": 1.425480604171753,
      "learning_rate": 4.9421669091824304e-05,
      "loss": 0.2442,
      "num_input_tokens_seen": 2378624,
      "step": 2745
    },
    {
      "epoch": 1.2965582272512965,
      "grad_norm": 0.9652962684631348,
      "learning_rate": 4.9419567752744634e-05,
      "loss": 0.1791,
      "num_input_tokens_seen": 2383088,
      "step": 2750
    },
    {
      "epoch": 1.298915605846299,
      "grad_norm": 0.869178056716919,
      "learning_rate": 4.941746264786074e-05,
      "loss": 0.2917,
      "num_input_tokens_seen": 2387952,
      "step": 2755
    },
    {
      "epoch": 1.3012729844413014,
      "grad_norm": 0.981084942817688,
      "learning_rate": 4.9415353777497254e-05,
      "loss": 0.1922,
      "num_input_tokens_seen": 2391984,
      "step": 2760
    },
    {
      "epoch": 1.3036303630363038,
      "grad_norm": 0.9991363286972046,
      "learning_rate": 4.9413241141979394e-05,
      "loss": 0.2546,
      "num_input_tokens_seen": 2396944,
      "step": 2765
    },
    {
      "epoch": 1.305987741631306,
      "grad_norm": 1.7044949531555176,
      "learning_rate": 4.9411124741632956e-05,
      "loss": 0.2306,
      "num_input_tokens_seen": 2401200,
      "step": 2770
    },
    {
      "epoch": 1.3083451202263083,
      "grad_norm": 1.3100314140319824,
      "learning_rate": 4.940900457678431e-05,
      "loss": 0.2412,
      "num_input_tokens_seen": 2405936,
      "step": 2775
    },
    {
      "epoch": 1.3107024988213107,
      "grad_norm": 0.9882826209068298,
      "learning_rate": 4.9406880647760425e-05,
      "loss": 0.1845,
      "num_input_tokens_seen": 2410256,
      "step": 2780
    },
    {
      "epoch": 1.3130598774163131,
      "grad_norm": 1.4160106182098389,
      "learning_rate": 4.9404752954888824e-05,
      "loss": 0.2067,
      "num_input_tokens_seen": 2415136,
      "step": 2785
    },
    {
      "epoch": 1.3154172560113153,
      "grad_norm": 1.544751524925232,
      "learning_rate": 4.940262149849762e-05,
      "loss": 0.187,
      "num_input_tokens_seen": 2418976,
      "step": 2790
    },
    {
      "epoch": 1.3177746346063177,
      "grad_norm": 0.4134797751903534,
      "learning_rate": 4.9400486278915526e-05,
      "loss": 0.1667,
      "num_input_tokens_seen": 2422768,
      "step": 2795
    },
    {
      "epoch": 1.3201320132013201,
      "grad_norm": 2.035200834274292,
      "learning_rate": 4.939834729647181e-05,
      "loss": 0.2028,
      "num_input_tokens_seen": 2427152,
      "step": 2800
    },
    {
      "epoch": 1.3201320132013201,
      "eval_loss": 0.2222793996334076,
      "eval_runtime": 22.1962,
      "eval_samples_per_second": 42.485,
      "eval_steps_per_second": 21.265,
      "num_input_tokens_seen": 2427152,
      "step": 2800
    },
    {
      "epoch": 1.3224893917963225,
      "grad_norm": 0.8405352830886841,
      "learning_rate": 4.9396204551496326e-05,
      "loss": 0.1882,
      "num_input_tokens_seen": 2431568,
      "step": 2805
    },
    {
      "epoch": 1.324846770391325,
      "grad_norm": 1.4003384113311768,
      "learning_rate": 4.939405804431952e-05,
      "loss": 0.2233,
      "num_input_tokens_seen": 2435680,
      "step": 2810
    },
    {
      "epoch": 1.3272041489863273,
      "grad_norm": 0.41833218932151794,
      "learning_rate": 4.9391907775272414e-05,
      "loss": 0.1922,
      "num_input_tokens_seen": 2439456,
      "step": 2815
    },
    {
      "epoch": 1.3295615275813295,
      "grad_norm": 1.7987420558929443,
      "learning_rate": 4.9389753744686604e-05,
      "loss": 0.2634,
      "num_input_tokens_seen": 2444048,
      "step": 2820
    },
    {
      "epoch": 1.331918906176332,
      "grad_norm": 1.0409022569656372,
      "learning_rate": 4.938759595289426e-05,
      "loss": 0.2092,
      "num_input_tokens_seen": 2449792,
      "step": 2825
    },
    {
      "epoch": 1.3342762847713343,
      "grad_norm": 2.117506504058838,
      "learning_rate": 4.938543440022815e-05,
      "loss": 0.2437,
      "num_input_tokens_seen": 2454688,
      "step": 2830
    },
    {
      "epoch": 1.3366336633663367,
      "grad_norm": 1.5998446941375732,
      "learning_rate": 4.938326908702161e-05,
      "loss": 0.2335,
      "num_input_tokens_seen": 2458480,
      "step": 2835
    },
    {
      "epoch": 1.338991041961339,
      "grad_norm": 0.6569865942001343,
      "learning_rate": 4.9381100013608554e-05,
      "loss": 0.2447,
      "num_input_tokens_seen": 2462656,
      "step": 2840
    },
    {
      "epoch": 1.3413484205563413,
      "grad_norm": 1.622472882270813,
      "learning_rate": 4.9378927180323485e-05,
      "loss": 0.2846,
      "num_input_tokens_seen": 2466960,
      "step": 2845
    },
    {
      "epoch": 1.3437057991513437,
      "grad_norm": 0.861919105052948,
      "learning_rate": 4.937675058750148e-05,
      "loss": 0.2204,
      "num_input_tokens_seen": 2472096,
      "step": 2850
    },
    {
      "epoch": 1.346063177746346,
      "grad_norm": 1.6995302438735962,
      "learning_rate": 4.937457023547819e-05,
      "loss": 0.166,
      "num_input_tokens_seen": 2476624,
      "step": 2855
    },
    {
      "epoch": 1.3484205563413485,
      "grad_norm": 1.6526336669921875,
      "learning_rate": 4.9372386124589876e-05,
      "loss": 0.2006,
      "num_input_tokens_seen": 2480400,
      "step": 2860
    },
    {
      "epoch": 1.350777934936351,
      "grad_norm": 0.40795740485191345,
      "learning_rate": 4.937019825517333e-05,
      "loss": 0.1975,
      "num_input_tokens_seen": 2484416,
      "step": 2865
    },
    {
      "epoch": 1.353135313531353,
      "grad_norm": 1.042680025100708,
      "learning_rate": 4.9368006627565954e-05,
      "loss": 0.2059,
      "num_input_tokens_seen": 2489040,
      "step": 2870
    },
    {
      "epoch": 1.3554926921263555,
      "grad_norm": 1.060367226600647,
      "learning_rate": 4.936581124210573e-05,
      "loss": 0.1829,
      "num_input_tokens_seen": 2493328,
      "step": 2875
    },
    {
      "epoch": 1.3578500707213579,
      "grad_norm": 3.0204577445983887,
      "learning_rate": 4.9363612099131216e-05,
      "loss": 0.2311,
      "num_input_tokens_seen": 2496928,
      "step": 2880
    },
    {
      "epoch": 1.3602074493163603,
      "grad_norm": 0.8274238705635071,
      "learning_rate": 4.936140919898155e-05,
      "loss": 0.2381,
      "num_input_tokens_seen": 2502416,
      "step": 2885
    },
    {
      "epoch": 1.3625648279113625,
      "grad_norm": 1.5213754177093506,
      "learning_rate": 4.9359202541996426e-05,
      "loss": 0.1764,
      "num_input_tokens_seen": 2507296,
      "step": 2890
    },
    {
      "epoch": 1.3649222065063649,
      "grad_norm": 1.9365146160125732,
      "learning_rate": 4.935699212851616e-05,
      "loss": 0.2117,
      "num_input_tokens_seen": 2511264,
      "step": 2895
    },
    {
      "epoch": 1.3672795851013673,
      "grad_norm": 1.0133110284805298,
      "learning_rate": 4.935477795888162e-05,
      "loss": 0.2106,
      "num_input_tokens_seen": 2515616,
      "step": 2900
    },
    {
      "epoch": 1.3696369636963697,
      "grad_norm": 1.8066432476043701,
      "learning_rate": 4.935256003343426e-05,
      "loss": 0.2379,
      "num_input_tokens_seen": 2520320,
      "step": 2905
    },
    {
      "epoch": 1.371994342291372,
      "grad_norm": 2.467456340789795,
      "learning_rate": 4.93503383525161e-05,
      "loss": 0.2361,
      "num_input_tokens_seen": 2524624,
      "step": 2910
    },
    {
      "epoch": 1.3743517208863745,
      "grad_norm": 1.8648476600646973,
      "learning_rate": 4.934811291646977e-05,
      "loss": 0.2832,
      "num_input_tokens_seen": 2529632,
      "step": 2915
    },
    {
      "epoch": 1.3767090994813767,
      "grad_norm": 1.7148252725601196,
      "learning_rate": 4.934588372563845e-05,
      "loss": 0.2031,
      "num_input_tokens_seen": 2533664,
      "step": 2920
    },
    {
      "epoch": 1.379066478076379,
      "grad_norm": 1.7632367610931396,
      "learning_rate": 4.93436507803659e-05,
      "loss": 0.2194,
      "num_input_tokens_seen": 2537792,
      "step": 2925
    },
    {
      "epoch": 1.3814238566713815,
      "grad_norm": 0.504602313041687,
      "learning_rate": 4.934141408099649e-05,
      "loss": 0.1618,
      "num_input_tokens_seen": 2542368,
      "step": 2930
    },
    {
      "epoch": 1.3837812352663839,
      "grad_norm": 1.0280206203460693,
      "learning_rate": 4.9339173627875135e-05,
      "loss": 0.2056,
      "num_input_tokens_seen": 2546512,
      "step": 2935
    },
    {
      "epoch": 1.386138613861386,
      "grad_norm": 1.2389938831329346,
      "learning_rate": 4.9336929421347335e-05,
      "loss": 0.2057,
      "num_input_tokens_seen": 2550400,
      "step": 2940
    },
    {
      "epoch": 1.3884959924563884,
      "grad_norm": 1.6766395568847656,
      "learning_rate": 4.933468146175918e-05,
      "loss": 0.2703,
      "num_input_tokens_seen": 2554480,
      "step": 2945
    },
    {
      "epoch": 1.3908533710513908,
      "grad_norm": 1.423099398612976,
      "learning_rate": 4.933242974945734e-05,
      "loss": 0.2201,
      "num_input_tokens_seen": 2558576,
      "step": 2950
    },
    {
      "epoch": 1.3932107496463932,
      "grad_norm": 0.8246122598648071,
      "learning_rate": 4.933017428478906e-05,
      "loss": 0.2845,
      "num_input_tokens_seen": 2562768,
      "step": 2955
    },
    {
      "epoch": 1.3955681282413956,
      "grad_norm": 1.5553464889526367,
      "learning_rate": 4.932791506810214e-05,
      "loss": 0.1859,
      "num_input_tokens_seen": 2566832,
      "step": 2960
    },
    {
      "epoch": 1.397925506836398,
      "grad_norm": 1.084437608718872,
      "learning_rate": 4.932565209974499e-05,
      "loss": 0.1871,
      "num_input_tokens_seen": 2571824,
      "step": 2965
    },
    {
      "epoch": 1.4002828854314002,
      "grad_norm": 1.7354919910430908,
      "learning_rate": 4.93233853800666e-05,
      "loss": 0.2193,
      "num_input_tokens_seen": 2576160,
      "step": 2970
    },
    {
      "epoch": 1.4026402640264026,
      "grad_norm": 0.689007580280304,
      "learning_rate": 4.932111490941651e-05,
      "loss": 0.2076,
      "num_input_tokens_seen": 2580208,
      "step": 2975
    },
    {
      "epoch": 1.404997642621405,
      "grad_norm": 0.6745365262031555,
      "learning_rate": 4.9318840688144876e-05,
      "loss": 0.2927,
      "num_input_tokens_seen": 2583952,
      "step": 2980
    },
    {
      "epoch": 1.4073550212164074,
      "grad_norm": 1.2342101335525513,
      "learning_rate": 4.9316562716602387e-05,
      "loss": 0.2215,
      "num_input_tokens_seen": 2587856,
      "step": 2985
    },
    {
      "epoch": 1.4097123998114096,
      "grad_norm": 2.1965889930725098,
      "learning_rate": 4.9314280995140346e-05,
      "loss": 0.27,
      "num_input_tokens_seen": 2592784,
      "step": 2990
    },
    {
      "epoch": 1.412069778406412,
      "grad_norm": 1.194669485092163,
      "learning_rate": 4.931199552411063e-05,
      "loss": 0.183,
      "num_input_tokens_seen": 2596864,
      "step": 2995
    },
    {
      "epoch": 1.4144271570014144,
      "grad_norm": 0.7765443325042725,
      "learning_rate": 4.930970630386568e-05,
      "loss": 0.1915,
      "num_input_tokens_seen": 2601296,
      "step": 3000
    },
    {
      "epoch": 1.4144271570014144,
      "eval_loss": 0.22042587399482727,
      "eval_runtime": 22.1877,
      "eval_samples_per_second": 42.501,
      "eval_steps_per_second": 21.273,
      "num_input_tokens_seen": 2601296,
      "step": 3000
    },
    {
      "epoch": 1.4167845355964168,
      "grad_norm": 1.0664104223251343,
      "learning_rate": 4.9307413334758524e-05,
      "loss": 0.2171,
      "num_input_tokens_seen": 2604832,
      "step": 3005
    },
    {
      "epoch": 1.4191419141914192,
      "grad_norm": 1.6415225267410278,
      "learning_rate": 4.930511661714276e-05,
      "loss": 0.1987,
      "num_input_tokens_seen": 2608896,
      "step": 3010
    },
    {
      "epoch": 1.4214992927864216,
      "grad_norm": 0.633159875869751,
      "learning_rate": 4.9302816151372576e-05,
      "loss": 0.2287,
      "num_input_tokens_seen": 2612960,
      "step": 3015
    },
    {
      "epoch": 1.4238566713814238,
      "grad_norm": 1.4326668977737427,
      "learning_rate": 4.930051193780274e-05,
      "loss": 0.219,
      "num_input_tokens_seen": 2616976,
      "step": 3020
    },
    {
      "epoch": 1.4262140499764262,
      "grad_norm": 1.034281611442566,
      "learning_rate": 4.929820397678858e-05,
      "loss": 0.1529,
      "num_input_tokens_seen": 2620928,
      "step": 3025
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.8107935190200806,
      "learning_rate": 4.9295892268686015e-05,
      "loss": 0.1971,
      "num_input_tokens_seen": 2625328,
      "step": 3030
    },
    {
      "epoch": 1.430928807166431,
      "grad_norm": 0.8473072052001953,
      "learning_rate": 4.9293576813851536e-05,
      "loss": 0.1965,
      "num_input_tokens_seen": 2629824,
      "step": 3035
    },
    {
      "epoch": 1.4332861857614332,
      "grad_norm": 1.5543066263198853,
      "learning_rate": 4.929125761264223e-05,
      "loss": 0.2388,
      "num_input_tokens_seen": 2633568,
      "step": 3040
    },
    {
      "epoch": 1.4356435643564356,
      "grad_norm": 1.5309889316558838,
      "learning_rate": 4.928893466541573e-05,
      "loss": 0.1455,
      "num_input_tokens_seen": 2638400,
      "step": 3045
    },
    {
      "epoch": 1.438000942951438,
      "grad_norm": 2.127978563308716,
      "learning_rate": 4.928660797253027e-05,
      "loss": 0.229,
      "num_input_tokens_seen": 2643200,
      "step": 3050
    },
    {
      "epoch": 1.4403583215464404,
      "grad_norm": 0.6690800786018372,
      "learning_rate": 4.928427753434467e-05,
      "loss": 0.1969,
      "num_input_tokens_seen": 2647120,
      "step": 3055
    },
    {
      "epoch": 1.4427157001414428,
      "grad_norm": 3.3050992488861084,
      "learning_rate": 4.9281943351218286e-05,
      "loss": 0.2449,
      "num_input_tokens_seen": 2651088,
      "step": 3060
    },
    {
      "epoch": 1.4450730787364452,
      "grad_norm": 1.6236156225204468,
      "learning_rate": 4.9279605423511095e-05,
      "loss": 0.3636,
      "num_input_tokens_seen": 2655616,
      "step": 3065
    },
    {
      "epoch": 1.4474304573314474,
      "grad_norm": 1.6346572637557983,
      "learning_rate": 4.927726375158363e-05,
      "loss": 0.2608,
      "num_input_tokens_seen": 2659680,
      "step": 3070
    },
    {
      "epoch": 1.4497878359264498,
      "grad_norm": 2.0732409954071045,
      "learning_rate": 4.9274918335797004e-05,
      "loss": 0.1895,
      "num_input_tokens_seen": 2664048,
      "step": 3075
    },
    {
      "epoch": 1.4521452145214522,
      "grad_norm": 0.6998698711395264,
      "learning_rate": 4.927256917651292e-05,
      "loss": 0.203,
      "num_input_tokens_seen": 2668464,
      "step": 3080
    },
    {
      "epoch": 1.4545025931164546,
      "grad_norm": 2.2789816856384277,
      "learning_rate": 4.927021627409364e-05,
      "loss": 0.1917,
      "num_input_tokens_seen": 2672528,
      "step": 3085
    },
    {
      "epoch": 1.4568599717114568,
      "grad_norm": 1.5912320613861084,
      "learning_rate": 4.9267859628902005e-05,
      "loss": 0.1449,
      "num_input_tokens_seen": 2677152,
      "step": 3090
    },
    {
      "epoch": 1.4592173503064592,
      "grad_norm": 1.3258192539215088,
      "learning_rate": 4.9265499241301454e-05,
      "loss": 0.2052,
      "num_input_tokens_seen": 2681600,
      "step": 3095
    },
    {
      "epoch": 1.4615747289014616,
      "grad_norm": 1.6244860887527466,
      "learning_rate": 4.926313511165598e-05,
      "loss": 0.2114,
      "num_input_tokens_seen": 2686096,
      "step": 3100
    },
    {
      "epoch": 1.463932107496464,
      "grad_norm": 0.9535281658172607,
      "learning_rate": 4.926076724033016e-05,
      "loss": 0.2187,
      "num_input_tokens_seen": 2690048,
      "step": 3105
    },
    {
      "epoch": 1.4662894860914664,
      "grad_norm": 0.4665777385234833,
      "learning_rate": 4.9258395627689146e-05,
      "loss": 0.1709,
      "num_input_tokens_seen": 2693824,
      "step": 3110
    },
    {
      "epoch": 1.4686468646864688,
      "grad_norm": 0.7909863591194153,
      "learning_rate": 4.925602027409868e-05,
      "loss": 0.1559,
      "num_input_tokens_seen": 2698448,
      "step": 3115
    },
    {
      "epoch": 1.471004243281471,
      "grad_norm": 0.838658332824707,
      "learning_rate": 4.925364117992507e-05,
      "loss": 0.2911,
      "num_input_tokens_seen": 2702480,
      "step": 3120
    },
    {
      "epoch": 1.4733616218764733,
      "grad_norm": 0.5353749394416809,
      "learning_rate": 4.92512583455352e-05,
      "loss": 0.1748,
      "num_input_tokens_seen": 2706640,
      "step": 3125
    },
    {
      "epoch": 1.4757190004714758,
      "grad_norm": 0.7654573321342468,
      "learning_rate": 4.9248871771296536e-05,
      "loss": 0.2323,
      "num_input_tokens_seen": 2711056,
      "step": 3130
    },
    {
      "epoch": 1.4780763790664782,
      "grad_norm": 0.994553804397583,
      "learning_rate": 4.924648145757711e-05,
      "loss": 0.2122,
      "num_input_tokens_seen": 2715024,
      "step": 3135
    },
    {
      "epoch": 1.4804337576614803,
      "grad_norm": 0.4541115164756775,
      "learning_rate": 4.924408740474554e-05,
      "loss": 0.1789,
      "num_input_tokens_seen": 2719120,
      "step": 3140
    },
    {
      "epoch": 1.4827911362564827,
      "grad_norm": 1.1137585639953613,
      "learning_rate": 4.924168961317103e-05,
      "loss": 0.1955,
      "num_input_tokens_seen": 2722992,
      "step": 3145
    },
    {
      "epoch": 1.4851485148514851,
      "grad_norm": 1.3682467937469482,
      "learning_rate": 4.9239288083223334e-05,
      "loss": 0.2085,
      "num_input_tokens_seen": 2727408,
      "step": 3150
    },
    {
      "epoch": 1.4875058934464875,
      "grad_norm": 1.8794903755187988,
      "learning_rate": 4.9236882815272803e-05,
      "loss": 0.2138,
      "num_input_tokens_seen": 2731008,
      "step": 3155
    },
    {
      "epoch": 1.48986327204149,
      "grad_norm": 0.8094249367713928,
      "learning_rate": 4.9234473809690365e-05,
      "loss": 0.233,
      "num_input_tokens_seen": 2735456,
      "step": 3160
    },
    {
      "epoch": 1.4922206506364923,
      "grad_norm": 0.9327897429466248,
      "learning_rate": 4.923206106684752e-05,
      "loss": 0.2389,
      "num_input_tokens_seen": 2740544,
      "step": 3165
    },
    {
      "epoch": 1.4945780292314945,
      "grad_norm": 1.811686635017395,
      "learning_rate": 4.922964458711634e-05,
      "loss": 0.1674,
      "num_input_tokens_seen": 2745424,
      "step": 3170
    },
    {
      "epoch": 1.496935407826497,
      "grad_norm": 1.3492577075958252,
      "learning_rate": 4.9227224370869474e-05,
      "loss": 0.2102,
      "num_input_tokens_seen": 2749536,
      "step": 3175
    },
    {
      "epoch": 1.4992927864214993,
      "grad_norm": 1.1172691583633423,
      "learning_rate": 4.9224800418480155e-05,
      "loss": 0.1627,
      "num_input_tokens_seen": 2755152,
      "step": 3180
    },
    {
      "epoch": 1.5016501650165015,
      "grad_norm": 1.9311751127243042,
      "learning_rate": 4.9222372730322176e-05,
      "loss": 0.1938,
      "num_input_tokens_seen": 2760736,
      "step": 3185
    },
    {
      "epoch": 1.504007543611504,
      "grad_norm": 2.7909083366394043,
      "learning_rate": 4.921994130676993e-05,
      "loss": 0.2241,
      "num_input_tokens_seen": 2765776,
      "step": 3190
    },
    {
      "epoch": 1.5063649222065063,
      "grad_norm": 2.4430203437805176,
      "learning_rate": 4.9217506148198366e-05,
      "loss": 0.2663,
      "num_input_tokens_seen": 2770272,
      "step": 3195
    },
    {
      "epoch": 1.5087223008015087,
      "grad_norm": 0.7173449397087097,
      "learning_rate": 4.921506725498302e-05,
      "loss": 0.1827,
      "num_input_tokens_seen": 2774672,
      "step": 3200
    },
    {
      "epoch": 1.5087223008015087,
      "eval_loss": 0.22187475860118866,
      "eval_runtime": 22.1751,
      "eval_samples_per_second": 42.525,
      "eval_steps_per_second": 21.285,
      "num_input_tokens_seen": 2774672,
      "step": 3200
    },
    {
      "epoch": 1.511079679396511,
      "grad_norm": 1.4367737770080566,
      "learning_rate": 4.9212624627499994e-05,
      "loss": 0.2326,
      "num_input_tokens_seen": 2779232,
      "step": 3205
    },
    {
      "epoch": 1.5134370579915135,
      "grad_norm": 0.9908407926559448,
      "learning_rate": 4.921017826612597e-05,
      "loss": 0.2274,
      "num_input_tokens_seen": 2784320,
      "step": 3210
    },
    {
      "epoch": 1.515794436586516,
      "grad_norm": 0.6985713243484497,
      "learning_rate": 4.9207728171238223e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 2788400,
      "step": 3215
    },
    {
      "epoch": 1.5181518151815183,
      "grad_norm": 0.7129409313201904,
      "learning_rate": 4.920527434321458e-05,
      "loss": 0.2111,
      "num_input_tokens_seen": 2792896,
      "step": 3220
    },
    {
      "epoch": 1.5205091937765205,
      "grad_norm": 0.5871591567993164,
      "learning_rate": 4.920281678243345e-05,
      "loss": 0.2052,
      "num_input_tokens_seen": 2796640,
      "step": 3225
    },
    {
      "epoch": 1.522866572371523,
      "grad_norm": 1.1540210247039795,
      "learning_rate": 4.920035548927381e-05,
      "loss": 0.1974,
      "num_input_tokens_seen": 2800752,
      "step": 3230
    },
    {
      "epoch": 1.525223950966525,
      "grad_norm": 1.9563833475112915,
      "learning_rate": 4.919789046411525e-05,
      "loss": 0.2063,
      "num_input_tokens_seen": 2804464,
      "step": 3235
    },
    {
      "epoch": 1.5275813295615275,
      "grad_norm": 1.247664213180542,
      "learning_rate": 4.919542170733787e-05,
      "loss": 0.2549,
      "num_input_tokens_seen": 2809008,
      "step": 3240
    },
    {
      "epoch": 1.5299387081565299,
      "grad_norm": 0.6438462138175964,
      "learning_rate": 4.919294921932242e-05,
      "loss": 0.1807,
      "num_input_tokens_seen": 2813344,
      "step": 3245
    },
    {
      "epoch": 1.5322960867515323,
      "grad_norm": 0.4978686273097992,
      "learning_rate": 4.919047300045016e-05,
      "loss": 0.1979,
      "num_input_tokens_seen": 2817520,
      "step": 3250
    },
    {
      "epoch": 1.5346534653465347,
      "grad_norm": 2.151916742324829,
      "learning_rate": 4.918799305110299e-05,
      "loss": 0.2032,
      "num_input_tokens_seen": 2822160,
      "step": 3255
    },
    {
      "epoch": 1.537010843941537,
      "grad_norm": 0.8116099834442139,
      "learning_rate": 4.918550937166331e-05,
      "loss": 0.2097,
      "num_input_tokens_seen": 2826624,
      "step": 3260
    },
    {
      "epoch": 1.5393682225365395,
      "grad_norm": 2.1142125129699707,
      "learning_rate": 4.918302196251415e-05,
      "loss": 0.2068,
      "num_input_tokens_seen": 2830448,
      "step": 3265
    },
    {
      "epoch": 1.541725601131542,
      "grad_norm": 0.5997825860977173,
      "learning_rate": 4.91805308240391e-05,
      "loss": 0.1912,
      "num_input_tokens_seen": 2834192,
      "step": 3270
    },
    {
      "epoch": 1.544082979726544,
      "grad_norm": 0.5110817551612854,
      "learning_rate": 4.9178035956622326e-05,
      "loss": 0.1889,
      "num_input_tokens_seen": 2838736,
      "step": 3275
    },
    {
      "epoch": 1.5464403583215465,
      "grad_norm": 0.8729378581047058,
      "learning_rate": 4.917553736064857e-05,
      "loss": 0.1745,
      "num_input_tokens_seen": 2842896,
      "step": 3280
    },
    {
      "epoch": 1.5487977369165487,
      "grad_norm": 0.7493543028831482,
      "learning_rate": 4.917303503650314e-05,
      "loss": 0.2173,
      "num_input_tokens_seen": 2846496,
      "step": 3285
    },
    {
      "epoch": 1.551155115511551,
      "grad_norm": 0.6653828024864197,
      "learning_rate": 4.917052898457194e-05,
      "loss": 0.2134,
      "num_input_tokens_seen": 2851568,
      "step": 3290
    },
    {
      "epoch": 1.5535124941065535,
      "grad_norm": 0.9907928705215454,
      "learning_rate": 4.916801920524141e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 2855984,
      "step": 3295
    },
    {
      "epoch": 1.5558698727015559,
      "grad_norm": 0.9479832649230957,
      "learning_rate": 4.916550569889862e-05,
      "loss": 0.1857,
      "num_input_tokens_seen": 2860912,
      "step": 3300
    },
    {
      "epoch": 1.5582272512965583,
      "grad_norm": 1.0640774965286255,
      "learning_rate": 4.916298846593116e-05,
      "loss": 0.2512,
      "num_input_tokens_seen": 2865312,
      "step": 3305
    },
    {
      "epoch": 1.5605846298915607,
      "grad_norm": 1.0898514986038208,
      "learning_rate": 4.916046750672722e-05,
      "loss": 0.2012,
      "num_input_tokens_seen": 2869200,
      "step": 3310
    },
    {
      "epoch": 1.562942008486563,
      "grad_norm": 1.133689045906067,
      "learning_rate": 4.915794282167559e-05,
      "loss": 0.184,
      "num_input_tokens_seen": 2872640,
      "step": 3315
    },
    {
      "epoch": 1.5652993870815655,
      "grad_norm": 0.338389128446579,
      "learning_rate": 4.915541441116558e-05,
      "loss": 0.1126,
      "num_input_tokens_seen": 2876848,
      "step": 3320
    },
    {
      "epoch": 1.5676567656765676,
      "grad_norm": 1.3553270101547241,
      "learning_rate": 4.915288227558711e-05,
      "loss": 0.2191,
      "num_input_tokens_seen": 2881904,
      "step": 3325
    },
    {
      "epoch": 1.57001414427157,
      "grad_norm": 1.0770081281661987,
      "learning_rate": 4.915034641533066e-05,
      "loss": 0.1912,
      "num_input_tokens_seen": 2886192,
      "step": 3330
    },
    {
      "epoch": 1.5723715228665722,
      "grad_norm": 1.4156404733657837,
      "learning_rate": 4.914780683078731e-05,
      "loss": 0.1776,
      "num_input_tokens_seen": 2890928,
      "step": 3335
    },
    {
      "epoch": 1.5747289014615746,
      "grad_norm": 1.2132524251937866,
      "learning_rate": 4.9145263522348695e-05,
      "loss": 0.1551,
      "num_input_tokens_seen": 2895488,
      "step": 3340
    },
    {
      "epoch": 1.577086280056577,
      "grad_norm": 0.9356024265289307,
      "learning_rate": 4.9142716490407e-05,
      "loss": 0.1803,
      "num_input_tokens_seen": 2899712,
      "step": 3345
    },
    {
      "epoch": 1.5794436586515794,
      "grad_norm": 1.5086532831192017,
      "learning_rate": 4.914016573535504e-05,
      "loss": 0.1904,
      "num_input_tokens_seen": 2903728,
      "step": 3350
    },
    {
      "epoch": 1.5818010372465818,
      "grad_norm": 0.7060988545417786,
      "learning_rate": 4.9137611257586154e-05,
      "loss": 0.2233,
      "num_input_tokens_seen": 2907792,
      "step": 3355
    },
    {
      "epoch": 1.5841584158415842,
      "grad_norm": 0.5120238661766052,
      "learning_rate": 4.9135053057494274e-05,
      "loss": 0.1209,
      "num_input_tokens_seen": 2912352,
      "step": 3360
    },
    {
      "epoch": 1.5865157944365866,
      "grad_norm": 0.8688015937805176,
      "learning_rate": 4.913249113547392e-05,
      "loss": 0.1413,
      "num_input_tokens_seen": 2916432,
      "step": 3365
    },
    {
      "epoch": 1.588873173031589,
      "grad_norm": 1.4597872495651245,
      "learning_rate": 4.912992549192016e-05,
      "loss": 0.2282,
      "num_input_tokens_seen": 2920416,
      "step": 3370
    },
    {
      "epoch": 1.5912305516265912,
      "grad_norm": 2.2095775604248047,
      "learning_rate": 4.9127356127228665e-05,
      "loss": 0.2107,
      "num_input_tokens_seen": 2924048,
      "step": 3375
    },
    {
      "epoch": 1.5935879302215936,
      "grad_norm": 1.644931435585022,
      "learning_rate": 4.912478304179564e-05,
      "loss": 0.2342,
      "num_input_tokens_seen": 2927760,
      "step": 3380
    },
    {
      "epoch": 1.5959453088165958,
      "grad_norm": 2.1685397624969482,
      "learning_rate": 4.9122206236017896e-05,
      "loss": 0.2669,
      "num_input_tokens_seen": 2932448,
      "step": 3385
    },
    {
      "epoch": 1.5983026874115982,
      "grad_norm": 2.453504800796509,
      "learning_rate": 4.911962571029282e-05,
      "loss": 0.2196,
      "num_input_tokens_seen": 2936800,
      "step": 3390
    },
    {
      "epoch": 1.6006600660066006,
      "grad_norm": 0.501107931137085,
      "learning_rate": 4.9117041465018353e-05,
      "loss": 0.1528,
      "num_input_tokens_seen": 2941136,
      "step": 3395
    },
    {
      "epoch": 1.603017444601603,
      "grad_norm": 1.102338433265686,
      "learning_rate": 4.911445350059302e-05,
      "loss": 0.2717,
      "num_input_tokens_seen": 2944896,
      "step": 3400
    },
    {
      "epoch": 1.603017444601603,
      "eval_loss": 0.22567209601402283,
      "eval_runtime": 22.1698,
      "eval_samples_per_second": 42.535,
      "eval_steps_per_second": 21.29,
      "num_input_tokens_seen": 2944896,
      "step": 3400
    },
    {
      "epoch": 1.6053748231966054,
      "grad_norm": 0.4999842047691345,
      "learning_rate": 4.9111861817415905e-05,
      "loss": 0.2487,
      "num_input_tokens_seen": 2949328,
      "step": 3405
    },
    {
      "epoch": 1.6077322017916078,
      "grad_norm": 0.5615550875663757,
      "learning_rate": 4.91092664158867e-05,
      "loss": 0.1766,
      "num_input_tokens_seen": 2953440,
      "step": 3410
    },
    {
      "epoch": 1.6100895803866102,
      "grad_norm": 1.8979085683822632,
      "learning_rate": 4.910666729640563e-05,
      "loss": 0.2951,
      "num_input_tokens_seen": 2958048,
      "step": 3415
    },
    {
      "epoch": 1.6124469589816126,
      "grad_norm": 0.9022587537765503,
      "learning_rate": 4.910406445937353e-05,
      "loss": 0.2011,
      "num_input_tokens_seen": 2962064,
      "step": 3420
    },
    {
      "epoch": 1.6148043375766148,
      "grad_norm": 1.022475004196167,
      "learning_rate": 4.9101457905191774e-05,
      "loss": 0.1719,
      "num_input_tokens_seen": 2966208,
      "step": 3425
    },
    {
      "epoch": 1.6171617161716172,
      "grad_norm": 1.0496426820755005,
      "learning_rate": 4.909884763426233e-05,
      "loss": 0.2751,
      "num_input_tokens_seen": 2970720,
      "step": 3430
    },
    {
      "epoch": 1.6195190947666194,
      "grad_norm": 1.5422712564468384,
      "learning_rate": 4.9096233646987736e-05,
      "loss": 0.2284,
      "num_input_tokens_seen": 2974704,
      "step": 3435
    },
    {
      "epoch": 1.6218764733616218,
      "grad_norm": 0.6532423496246338,
      "learning_rate": 4.9093615943771104e-05,
      "loss": 0.1321,
      "num_input_tokens_seen": 2978496,
      "step": 3440
    },
    {
      "epoch": 1.6242338519566242,
      "grad_norm": 0.6693040132522583,
      "learning_rate": 4.909099452501611e-05,
      "loss": 0.1382,
      "num_input_tokens_seen": 2983088,
      "step": 3445
    },
    {
      "epoch": 1.6265912305516266,
      "grad_norm": 0.6776031255722046,
      "learning_rate": 4.908836939112702e-05,
      "loss": 0.1772,
      "num_input_tokens_seen": 2987632,
      "step": 3450
    },
    {
      "epoch": 1.628948609146629,
      "grad_norm": 0.41002345085144043,
      "learning_rate": 4.908574054250865e-05,
      "loss": 0.1519,
      "num_input_tokens_seen": 2991520,
      "step": 3455
    },
    {
      "epoch": 1.6313059877416314,
      "grad_norm": 1.93821120262146,
      "learning_rate": 4.9083107979566414e-05,
      "loss": 0.2202,
      "num_input_tokens_seen": 2995888,
      "step": 3460
    },
    {
      "epoch": 1.6336633663366338,
      "grad_norm": 0.6884393692016602,
      "learning_rate": 4.908047170270628e-05,
      "loss": 0.1593,
      "num_input_tokens_seen": 2999952,
      "step": 3465
    },
    {
      "epoch": 1.6360207449316362,
      "grad_norm": 0.8018652200698853,
      "learning_rate": 4.9077831712334784e-05,
      "loss": 0.1814,
      "num_input_tokens_seen": 3003440,
      "step": 3470
    },
    {
      "epoch": 1.6383781235266384,
      "grad_norm": 1.588897705078125,
      "learning_rate": 4.907518800885907e-05,
      "loss": 0.1953,
      "num_input_tokens_seen": 3008240,
      "step": 3475
    },
    {
      "epoch": 1.6407355021216408,
      "grad_norm": 0.7397635579109192,
      "learning_rate": 4.907254059268681e-05,
      "loss": 0.2199,
      "num_input_tokens_seen": 3012672,
      "step": 3480
    },
    {
      "epoch": 1.643092880716643,
      "grad_norm": 0.5949789881706238,
      "learning_rate": 4.906988946422628e-05,
      "loss": 0.1737,
      "num_input_tokens_seen": 3018048,
      "step": 3485
    },
    {
      "epoch": 1.6454502593116453,
      "grad_norm": 0.40007320046424866,
      "learning_rate": 4.9067234623886315e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 3021616,
      "step": 3490
    },
    {
      "epoch": 1.6478076379066477,
      "grad_norm": 0.8408975005149841,
      "learning_rate": 4.9064576072076316e-05,
      "loss": 0.1921,
      "num_input_tokens_seen": 3026240,
      "step": 3495
    },
    {
      "epoch": 1.6501650165016502,
      "grad_norm": 1.3969001770019531,
      "learning_rate": 4.906191380920628e-05,
      "loss": 0.2083,
      "num_input_tokens_seen": 3030512,
      "step": 3500
    },
    {
      "epoch": 1.6525223950966526,
      "grad_norm": 1.141810417175293,
      "learning_rate": 4.905924783568675e-05,
      "loss": 0.1942,
      "num_input_tokens_seen": 3034656,
      "step": 3505
    },
    {
      "epoch": 1.654879773691655,
      "grad_norm": 1.9612503051757812,
      "learning_rate": 4.905657815192886e-05,
      "loss": 0.1715,
      "num_input_tokens_seen": 3038592,
      "step": 3510
    },
    {
      "epoch": 1.6572371522866574,
      "grad_norm": 1.6983704566955566,
      "learning_rate": 4.90539047583443e-05,
      "loss": 0.2978,
      "num_input_tokens_seen": 3042576,
      "step": 3515
    },
    {
      "epoch": 1.6595945308816598,
      "grad_norm": 2.1362645626068115,
      "learning_rate": 4.905122765534534e-05,
      "loss": 0.2478,
      "num_input_tokens_seen": 3046704,
      "step": 3520
    },
    {
      "epoch": 1.661951909476662,
      "grad_norm": 3.986985683441162,
      "learning_rate": 4.9048546843344846e-05,
      "loss": 0.2853,
      "num_input_tokens_seen": 3051648,
      "step": 3525
    },
    {
      "epoch": 1.6643092880716643,
      "grad_norm": 1.0915991067886353,
      "learning_rate": 4.9045862322756206e-05,
      "loss": 0.1843,
      "num_input_tokens_seen": 3056384,
      "step": 3530
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.7035347819328308,
      "learning_rate": 4.904317409399342e-05,
      "loss": 0.2411,
      "num_input_tokens_seen": 3060368,
      "step": 3535
    },
    {
      "epoch": 1.669024045261669,
      "grad_norm": 0.7979409694671631,
      "learning_rate": 4.904048215747104e-05,
      "loss": 0.1979,
      "num_input_tokens_seen": 3063968,
      "step": 3540
    },
    {
      "epoch": 1.6713814238566713,
      "grad_norm": 1.2365413904190063,
      "learning_rate": 4.90377865136042e-05,
      "loss": 0.1146,
      "num_input_tokens_seen": 3067984,
      "step": 3545
    },
    {
      "epoch": 1.6737388024516737,
      "grad_norm": 1.2099214792251587,
      "learning_rate": 4.90350871628086e-05,
      "loss": 0.1843,
      "num_input_tokens_seen": 3072288,
      "step": 3550
    },
    {
      "epoch": 1.6760961810466761,
      "grad_norm": 1.7522472143173218,
      "learning_rate": 4.903238410550052e-05,
      "loss": 0.1664,
      "num_input_tokens_seen": 3076448,
      "step": 3555
    },
    {
      "epoch": 1.6784535596416785,
      "grad_norm": 0.9109477400779724,
      "learning_rate": 4.90296773420968e-05,
      "loss": 0.2038,
      "num_input_tokens_seen": 3081072,
      "step": 3560
    },
    {
      "epoch": 1.680810938236681,
      "grad_norm": 0.8906397819519043,
      "learning_rate": 4.902696687301486e-05,
      "loss": 0.1625,
      "num_input_tokens_seen": 3086160,
      "step": 3565
    },
    {
      "epoch": 1.6831683168316833,
      "grad_norm": 0.8464270830154419,
      "learning_rate": 4.902425269867268e-05,
      "loss": 0.1888,
      "num_input_tokens_seen": 3090608,
      "step": 3570
    },
    {
      "epoch": 1.6855256954266855,
      "grad_norm": 0.5130507946014404,
      "learning_rate": 4.902153481948883e-05,
      "loss": 0.1388,
      "num_input_tokens_seen": 3095504,
      "step": 3575
    },
    {
      "epoch": 1.687883074021688,
      "grad_norm": 1.1484390497207642,
      "learning_rate": 4.901881323588244e-05,
      "loss": 0.2128,
      "num_input_tokens_seen": 3100336,
      "step": 3580
    },
    {
      "epoch": 1.69024045261669,
      "grad_norm": 0.7858063578605652,
      "learning_rate": 4.90160879482732e-05,
      "loss": 0.3022,
      "num_input_tokens_seen": 3104624,
      "step": 3585
    },
    {
      "epoch": 1.6925978312116925,
      "grad_norm": 1.4622080326080322,
      "learning_rate": 4.9013358957081405e-05,
      "loss": 0.2451,
      "num_input_tokens_seen": 3109184,
      "step": 3590
    },
    {
      "epoch": 1.694955209806695,
      "grad_norm": 2.75152325630188,
      "learning_rate": 4.901062626272789e-05,
      "loss": 0.2162,
      "num_input_tokens_seen": 3113248,
      "step": 3595
    },
    {
      "epoch": 1.6973125884016973,
      "grad_norm": 2.102067232131958,
      "learning_rate": 4.900788986563406e-05,
      "loss": 0.2092,
      "num_input_tokens_seen": 3117216,
      "step": 3600
    },
    {
      "epoch": 1.6973125884016973,
      "eval_loss": 0.2248712033033371,
      "eval_runtime": 22.166,
      "eval_samples_per_second": 42.543,
      "eval_steps_per_second": 21.294,
      "num_input_tokens_seen": 3117216,
      "step": 3600
    },
    {
      "epoch": 1.6996699669966997,
      "grad_norm": 3.5060160160064697,
      "learning_rate": 4.9005149766221915e-05,
      "loss": 0.1906,
      "num_input_tokens_seen": 3122032,
      "step": 3605
    },
    {
      "epoch": 1.702027345591702,
      "grad_norm": 0.6463086009025574,
      "learning_rate": 4.9002405964914e-05,
      "loss": 0.1724,
      "num_input_tokens_seen": 3126112,
      "step": 3610
    },
    {
      "epoch": 1.7043847241867045,
      "grad_norm": 0.9116499423980713,
      "learning_rate": 4.899965846213346e-05,
      "loss": 0.2316,
      "num_input_tokens_seen": 3131088,
      "step": 3615
    },
    {
      "epoch": 1.706742102781707,
      "grad_norm": 0.9323167204856873,
      "learning_rate": 4.899690725830399e-05,
      "loss": 0.218,
      "num_input_tokens_seen": 3135632,
      "step": 3620
    },
    {
      "epoch": 1.709099481376709,
      "grad_norm": 1.9242866039276123,
      "learning_rate": 4.899415235384985e-05,
      "loss": 0.2673,
      "num_input_tokens_seen": 3139712,
      "step": 3625
    },
    {
      "epoch": 1.7114568599717115,
      "grad_norm": 1.044283151626587,
      "learning_rate": 4.899139374919589e-05,
      "loss": 0.2548,
      "num_input_tokens_seen": 3143680,
      "step": 3630
    },
    {
      "epoch": 1.7138142385667137,
      "grad_norm": 1.6571027040481567,
      "learning_rate": 4.898863144476752e-05,
      "loss": 0.2383,
      "num_input_tokens_seen": 3148416,
      "step": 3635
    },
    {
      "epoch": 1.716171617161716,
      "grad_norm": 1.0178560018539429,
      "learning_rate": 4.898586544099072e-05,
      "loss": 0.2661,
      "num_input_tokens_seen": 3152592,
      "step": 3640
    },
    {
      "epoch": 1.7185289957567185,
      "grad_norm": 0.5017998218536377,
      "learning_rate": 4.898309573829204e-05,
      "loss": 0.1702,
      "num_input_tokens_seen": 3156816,
      "step": 3645
    },
    {
      "epoch": 1.7208863743517209,
      "grad_norm": 0.7736288905143738,
      "learning_rate": 4.898032233709862e-05,
      "loss": 0.2413,
      "num_input_tokens_seen": 3160672,
      "step": 3650
    },
    {
      "epoch": 1.7232437529467233,
      "grad_norm": 0.5876834988594055,
      "learning_rate": 4.8977545237838123e-05,
      "loss": 0.2092,
      "num_input_tokens_seen": 3164176,
      "step": 3655
    },
    {
      "epoch": 1.7256011315417257,
      "grad_norm": 1.274348258972168,
      "learning_rate": 4.8974764440938836e-05,
      "loss": 0.2497,
      "num_input_tokens_seen": 3169184,
      "step": 3660
    },
    {
      "epoch": 1.727958510136728,
      "grad_norm": 1.5610657930374146,
      "learning_rate": 4.897197994682959e-05,
      "loss": 0.2709,
      "num_input_tokens_seen": 3173056,
      "step": 3665
    },
    {
      "epoch": 1.7303158887317305,
      "grad_norm": 0.5560976266860962,
      "learning_rate": 4.8969191755939786e-05,
      "loss": 0.1905,
      "num_input_tokens_seen": 3177904,
      "step": 3670
    },
    {
      "epoch": 1.7326732673267327,
      "grad_norm": 0.4927385449409485,
      "learning_rate": 4.8966399868699396e-05,
      "loss": 0.1662,
      "num_input_tokens_seen": 3182448,
      "step": 3675
    },
    {
      "epoch": 1.735030645921735,
      "grad_norm": 1.3702868223190308,
      "learning_rate": 4.8963604285538965e-05,
      "loss": 0.1851,
      "num_input_tokens_seen": 3187056,
      "step": 3680
    },
    {
      "epoch": 1.7373880245167372,
      "grad_norm": 1.3009549379348755,
      "learning_rate": 4.8960805006889604e-05,
      "loss": 0.2298,
      "num_input_tokens_seen": 3191120,
      "step": 3685
    },
    {
      "epoch": 1.7397454031117396,
      "grad_norm": 1.2072112560272217,
      "learning_rate": 4.8958002033183004e-05,
      "loss": 0.2228,
      "num_input_tokens_seen": 3194992,
      "step": 3690
    },
    {
      "epoch": 1.742102781706742,
      "grad_norm": 1.4296863079071045,
      "learning_rate": 4.8955195364851414e-05,
      "loss": 0.1591,
      "num_input_tokens_seen": 3199040,
      "step": 3695
    },
    {
      "epoch": 1.7444601603017444,
      "grad_norm": 1.2572522163391113,
      "learning_rate": 4.895238500232766e-05,
      "loss": 0.2203,
      "num_input_tokens_seen": 3203392,
      "step": 3700
    },
    {
      "epoch": 1.7468175388967468,
      "grad_norm": 0.7061545848846436,
      "learning_rate": 4.8949570946045143e-05,
      "loss": 0.2302,
      "num_input_tokens_seen": 3207296,
      "step": 3705
    },
    {
      "epoch": 1.7491749174917492,
      "grad_norm": 0.9574533104896545,
      "learning_rate": 4.89467531964378e-05,
      "loss": 0.1982,
      "num_input_tokens_seen": 3211696,
      "step": 3710
    },
    {
      "epoch": 1.7515322960867516,
      "grad_norm": 1.0355396270751953,
      "learning_rate": 4.894393175394019e-05,
      "loss": 0.2226,
      "num_input_tokens_seen": 3215920,
      "step": 3715
    },
    {
      "epoch": 1.753889674681754,
      "grad_norm": 0.709676206111908,
      "learning_rate": 4.8941106618987406e-05,
      "loss": 0.2156,
      "num_input_tokens_seen": 3219680,
      "step": 3720
    },
    {
      "epoch": 1.7562470532767562,
      "grad_norm": 1.3436939716339111,
      "learning_rate": 4.893827779201512e-05,
      "loss": 0.1891,
      "num_input_tokens_seen": 3223440,
      "step": 3725
    },
    {
      "epoch": 1.7586044318717586,
      "grad_norm": 0.35068660974502563,
      "learning_rate": 4.893544527345957e-05,
      "loss": 0.2587,
      "num_input_tokens_seen": 3227680,
      "step": 3730
    },
    {
      "epoch": 1.7609618104667608,
      "grad_norm": 0.4759823977947235,
      "learning_rate": 4.8932609063757563e-05,
      "loss": 0.2133,
      "num_input_tokens_seen": 3231728,
      "step": 3735
    },
    {
      "epoch": 1.7633191890617632,
      "grad_norm": 1.8089812994003296,
      "learning_rate": 4.8929769163346484e-05,
      "loss": 0.1837,
      "num_input_tokens_seen": 3236112,
      "step": 3740
    },
    {
      "epoch": 1.7656765676567656,
      "grad_norm": 1.1992077827453613,
      "learning_rate": 4.892692557266429e-05,
      "loss": 0.1495,
      "num_input_tokens_seen": 3240704,
      "step": 3745
    },
    {
      "epoch": 1.768033946251768,
      "grad_norm": 0.7794426679611206,
      "learning_rate": 4.8924078292149464e-05,
      "loss": 0.177,
      "num_input_tokens_seen": 3244608,
      "step": 3750
    },
    {
      "epoch": 1.7703913248467704,
      "grad_norm": 2.195241689682007,
      "learning_rate": 4.892122732224114e-05,
      "loss": 0.2209,
      "num_input_tokens_seen": 3249104,
      "step": 3755
    },
    {
      "epoch": 1.7727487034417728,
      "grad_norm": 1.0362707376480103,
      "learning_rate": 4.8918372663378944e-05,
      "loss": 0.1609,
      "num_input_tokens_seen": 3252656,
      "step": 3760
    },
    {
      "epoch": 1.7751060820367752,
      "grad_norm": 1.8428312540054321,
      "learning_rate": 4.89155143160031e-05,
      "loss": 0.1567,
      "num_input_tokens_seen": 3256304,
      "step": 3765
    },
    {
      "epoch": 1.7774634606317776,
      "grad_norm": 1.6199904680252075,
      "learning_rate": 4.891265228055441e-05,
      "loss": 0.2899,
      "num_input_tokens_seen": 3261104,
      "step": 3770
    },
    {
      "epoch": 1.7798208392267798,
      "grad_norm": 0.956888735294342,
      "learning_rate": 4.890978655747424e-05,
      "loss": 0.1883,
      "num_input_tokens_seen": 3264896,
      "step": 3775
    },
    {
      "epoch": 1.7821782178217822,
      "grad_norm": 1.47179114818573,
      "learning_rate": 4.89069171472045e-05,
      "loss": 0.2446,
      "num_input_tokens_seen": 3270368,
      "step": 3780
    },
    {
      "epoch": 1.7845355964167844,
      "grad_norm": 0.5386603474617004,
      "learning_rate": 4.890404405018772e-05,
      "loss": 0.168,
      "num_input_tokens_seen": 3275200,
      "step": 3785
    },
    {
      "epoch": 1.7868929750117868,
      "grad_norm": 2.038933038711548,
      "learning_rate": 4.8901167266866934e-05,
      "loss": 0.2469,
      "num_input_tokens_seen": 3279552,
      "step": 3790
    },
    {
      "epoch": 1.7892503536067892,
      "grad_norm": 1.5007820129394531,
      "learning_rate": 4.88982867976858e-05,
      "loss": 0.1615,
      "num_input_tokens_seen": 3283712,
      "step": 3795
    },
    {
      "epoch": 1.7916077322017916,
      "grad_norm": 1.764037013053894,
      "learning_rate": 4.889540264308852e-05,
      "loss": 0.3069,
      "num_input_tokens_seen": 3287952,
      "step": 3800
    },
    {
      "epoch": 1.7916077322017916,
      "eval_loss": 0.22051014006137848,
      "eval_runtime": 22.2004,
      "eval_samples_per_second": 42.477,
      "eval_steps_per_second": 21.261,
      "num_input_tokens_seen": 3287952,
      "step": 3800
    },
    {
      "epoch": 1.793965110796794,
      "grad_norm": 0.9292309284210205,
      "learning_rate": 4.889251480351986e-05,
      "loss": 0.1565,
      "num_input_tokens_seen": 3292448,
      "step": 3805
    },
    {
      "epoch": 1.7963224893917964,
      "grad_norm": 1.9692481756210327,
      "learning_rate": 4.888962327942517e-05,
      "loss": 0.2723,
      "num_input_tokens_seen": 3297120,
      "step": 3810
    },
    {
      "epoch": 1.7986798679867988,
      "grad_norm": 0.8906758427619934,
      "learning_rate": 4.8886728071250356e-05,
      "loss": 0.1957,
      "num_input_tokens_seen": 3300880,
      "step": 3815
    },
    {
      "epoch": 1.8010372465818012,
      "grad_norm": 1.1353442668914795,
      "learning_rate": 4.8883829179441884e-05,
      "loss": 0.2008,
      "num_input_tokens_seen": 3304864,
      "step": 3820
    },
    {
      "epoch": 1.8033946251768034,
      "grad_norm": 2.7353062629699707,
      "learning_rate": 4.888092660444682e-05,
      "loss": 0.3243,
      "num_input_tokens_seen": 3309168,
      "step": 3825
    },
    {
      "epoch": 1.8057520037718058,
      "grad_norm": 1.5472811460494995,
      "learning_rate": 4.887802034671276e-05,
      "loss": 0.1772,
      "num_input_tokens_seen": 3313808,
      "step": 3830
    },
    {
      "epoch": 1.808109382366808,
      "grad_norm": 0.6223844885826111,
      "learning_rate": 4.88751104066879e-05,
      "loss": 0.2,
      "num_input_tokens_seen": 3318656,
      "step": 3835
    },
    {
      "epoch": 1.8104667609618104,
      "grad_norm": 0.7398841977119446,
      "learning_rate": 4.887219678482098e-05,
      "loss": 0.2343,
      "num_input_tokens_seen": 3322656,
      "step": 3840
    },
    {
      "epoch": 1.8128241395568128,
      "grad_norm": 1.5346544981002808,
      "learning_rate": 4.8869279481561316e-05,
      "loss": 0.1908,
      "num_input_tokens_seen": 3327024,
      "step": 3845
    },
    {
      "epoch": 1.8151815181518152,
      "grad_norm": 1.1816033124923706,
      "learning_rate": 4.88663584973588e-05,
      "loss": 0.2046,
      "num_input_tokens_seen": 3332512,
      "step": 3850
    },
    {
      "epoch": 1.8175388967468176,
      "grad_norm": 0.5998262166976929,
      "learning_rate": 4.8863433832663874e-05,
      "loss": 0.1318,
      "num_input_tokens_seen": 3336992,
      "step": 3855
    },
    {
      "epoch": 1.81989627534182,
      "grad_norm": 1.2128833532333374,
      "learning_rate": 4.886050548792757e-05,
      "loss": 0.2056,
      "num_input_tokens_seen": 3342160,
      "step": 3860
    },
    {
      "epoch": 1.8222536539368224,
      "grad_norm": 1.0361239910125732,
      "learning_rate": 4.8857573463601465e-05,
      "loss": 0.1501,
      "num_input_tokens_seen": 3346688,
      "step": 3865
    },
    {
      "epoch": 1.8246110325318248,
      "grad_norm": 0.6653880476951599,
      "learning_rate": 4.885463776013772e-05,
      "loss": 0.1564,
      "num_input_tokens_seen": 3350768,
      "step": 3870
    },
    {
      "epoch": 1.826968411126827,
      "grad_norm": 2.463707208633423,
      "learning_rate": 4.8851698377989056e-05,
      "loss": 0.2333,
      "num_input_tokens_seen": 3355072,
      "step": 3875
    },
    {
      "epoch": 1.8293257897218294,
      "grad_norm": 0.8850653171539307,
      "learning_rate": 4.884875531760876e-05,
      "loss": 0.1672,
      "num_input_tokens_seen": 3359664,
      "step": 3880
    },
    {
      "epoch": 1.8316831683168315,
      "grad_norm": 0.9456601738929749,
      "learning_rate": 4.88458085794507e-05,
      "loss": 0.1995,
      "num_input_tokens_seen": 3363744,
      "step": 3885
    },
    {
      "epoch": 1.834040546911834,
      "grad_norm": 1.3457549810409546,
      "learning_rate": 4.884285816396929e-05,
      "loss": 0.2444,
      "num_input_tokens_seen": 3368464,
      "step": 3890
    },
    {
      "epoch": 1.8363979255068363,
      "grad_norm": 0.5263032913208008,
      "learning_rate": 4.8839904071619526e-05,
      "loss": 0.1985,
      "num_input_tokens_seen": 3373520,
      "step": 3895
    },
    {
      "epoch": 1.8387553041018387,
      "grad_norm": 1.3329246044158936,
      "learning_rate": 4.8836946302856955e-05,
      "loss": 0.1744,
      "num_input_tokens_seen": 3377808,
      "step": 3900
    },
    {
      "epoch": 1.8411126826968411,
      "grad_norm": 1.340865135192871,
      "learning_rate": 4.8833984858137715e-05,
      "loss": 0.1871,
      "num_input_tokens_seen": 3382080,
      "step": 3905
    },
    {
      "epoch": 1.8434700612918435,
      "grad_norm": 0.7680668830871582,
      "learning_rate": 4.8831019737918494e-05,
      "loss": 0.2048,
      "num_input_tokens_seen": 3386736,
      "step": 3910
    },
    {
      "epoch": 1.845827439886846,
      "grad_norm": 0.9414777159690857,
      "learning_rate": 4.882805094265655e-05,
      "loss": 0.1673,
      "num_input_tokens_seen": 3391184,
      "step": 3915
    },
    {
      "epoch": 1.8481848184818483,
      "grad_norm": 1.3914657831192017,
      "learning_rate": 4.8825078472809706e-05,
      "loss": 0.23,
      "num_input_tokens_seen": 3395328,
      "step": 3920
    },
    {
      "epoch": 1.8505421970768505,
      "grad_norm": 1.30333411693573,
      "learning_rate": 4.882210232883635e-05,
      "loss": 0.2047,
      "num_input_tokens_seen": 3400192,
      "step": 3925
    },
    {
      "epoch": 1.852899575671853,
      "grad_norm": 1.4540553092956543,
      "learning_rate": 4.881912251119546e-05,
      "loss": 0.169,
      "num_input_tokens_seen": 3404544,
      "step": 3930
    },
    {
      "epoch": 1.855256954266855,
      "grad_norm": 0.9666699171066284,
      "learning_rate": 4.881613902034654e-05,
      "loss": 0.1865,
      "num_input_tokens_seen": 3408912,
      "step": 3935
    },
    {
      "epoch": 1.8576143328618575,
      "grad_norm": 0.759465754032135,
      "learning_rate": 4.88131518567497e-05,
      "loss": 0.1715,
      "num_input_tokens_seen": 3413008,
      "step": 3940
    },
    {
      "epoch": 1.85997171145686,
      "grad_norm": 1.100219488143921,
      "learning_rate": 4.881016102086558e-05,
      "loss": 0.1769,
      "num_input_tokens_seen": 3417056,
      "step": 3945
    },
    {
      "epoch": 1.8623290900518623,
      "grad_norm": 1.2949961423873901,
      "learning_rate": 4.8807166513155425e-05,
      "loss": 0.2002,
      "num_input_tokens_seen": 3421744,
      "step": 3950
    },
    {
      "epoch": 1.8646864686468647,
      "grad_norm": 1.3849873542785645,
      "learning_rate": 4.8804168334081004e-05,
      "loss": 0.1908,
      "num_input_tokens_seen": 3425696,
      "step": 3955
    },
    {
      "epoch": 1.8670438472418671,
      "grad_norm": 2.763005495071411,
      "learning_rate": 4.880116648410468e-05,
      "loss": 0.2925,
      "num_input_tokens_seen": 3431088,
      "step": 3960
    },
    {
      "epoch": 1.8694012258368695,
      "grad_norm": 1.5294806957244873,
      "learning_rate": 4.879816096368939e-05,
      "loss": 0.1459,
      "num_input_tokens_seen": 3435872,
      "step": 3965
    },
    {
      "epoch": 1.871758604431872,
      "grad_norm": 0.44467371702194214,
      "learning_rate": 4.879515177329861e-05,
      "loss": 0.1731,
      "num_input_tokens_seen": 3440016,
      "step": 3970
    },
    {
      "epoch": 1.874115983026874,
      "grad_norm": 0.80282062292099,
      "learning_rate": 4.8792138913396394e-05,
      "loss": 0.1637,
      "num_input_tokens_seen": 3444096,
      "step": 3975
    },
    {
      "epoch": 1.8764733616218765,
      "grad_norm": 1.0951766967773438,
      "learning_rate": 4.8789122384447374e-05,
      "loss": 0.263,
      "num_input_tokens_seen": 3448448,
      "step": 3980
    },
    {
      "epoch": 1.8788307402168787,
      "grad_norm": 0.692842423915863,
      "learning_rate": 4.878610218691673e-05,
      "loss": 0.1713,
      "num_input_tokens_seen": 3452320,
      "step": 3985
    },
    {
      "epoch": 1.881188118811881,
      "grad_norm": 0.6000655889511108,
      "learning_rate": 4.87830783212702e-05,
      "loss": 0.1967,
      "num_input_tokens_seen": 3456576,
      "step": 3990
    },
    {
      "epoch": 1.8835454974068835,
      "grad_norm": 1.1105213165283203,
      "learning_rate": 4.878005078797413e-05,
      "loss": 0.1944,
      "num_input_tokens_seen": 3460288,
      "step": 3995
    },
    {
      "epoch": 1.8859028760018859,
      "grad_norm": 1.000645399093628,
      "learning_rate": 4.877701958749539e-05,
      "loss": 0.1938,
      "num_input_tokens_seen": 3464640,
      "step": 4000
    },
    {
      "epoch": 1.8859028760018859,
      "eval_loss": 0.22389943897724152,
      "eval_runtime": 22.1697,
      "eval_samples_per_second": 42.536,
      "eval_steps_per_second": 21.29,
      "num_input_tokens_seen": 3464640,
      "step": 4000
    },
    {
      "epoch": 1.8882602545968883,
      "grad_norm": 0.6504820585250854,
      "learning_rate": 4.877398472030142e-05,
      "loss": 0.2148,
      "num_input_tokens_seen": 3468608,
      "step": 4005
    },
    {
      "epoch": 1.8906176331918907,
      "grad_norm": 2.6595921516418457,
      "learning_rate": 4.877094618686024e-05,
      "loss": 0.2716,
      "num_input_tokens_seen": 3472704,
      "step": 4010
    },
    {
      "epoch": 1.892975011786893,
      "grad_norm": 1.267816424369812,
      "learning_rate": 4.876790398764045e-05,
      "loss": 0.1997,
      "num_input_tokens_seen": 3477024,
      "step": 4015
    },
    {
      "epoch": 1.8953323903818955,
      "grad_norm": 1.2344194650650024,
      "learning_rate": 4.8764858123111167e-05,
      "loss": 0.1547,
      "num_input_tokens_seen": 3481328,
      "step": 4020
    },
    {
      "epoch": 1.8976897689768977,
      "grad_norm": 0.7321279644966125,
      "learning_rate": 4.876180859374212e-05,
      "loss": 0.162,
      "num_input_tokens_seen": 3485440,
      "step": 4025
    },
    {
      "epoch": 1.9000471475719,
      "grad_norm": 1.2521965503692627,
      "learning_rate": 4.875875540000357e-05,
      "loss": 0.2513,
      "num_input_tokens_seen": 3490576,
      "step": 4030
    },
    {
      "epoch": 1.9024045261669023,
      "grad_norm": 1.3612655401229858,
      "learning_rate": 4.8755698542366376e-05,
      "loss": 0.1845,
      "num_input_tokens_seen": 3495568,
      "step": 4035
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 0.5952543020248413,
      "learning_rate": 4.875263802130193e-05,
      "loss": 0.1892,
      "num_input_tokens_seen": 3499280,
      "step": 4040
    },
    {
      "epoch": 1.907119283356907,
      "grad_norm": 0.958967387676239,
      "learning_rate": 4.8749573837282207e-05,
      "loss": 0.2487,
      "num_input_tokens_seen": 3504048,
      "step": 4045
    },
    {
      "epoch": 1.9094766619519095,
      "grad_norm": 1.725858211517334,
      "learning_rate": 4.874650599077974e-05,
      "loss": 0.197,
      "num_input_tokens_seen": 3508000,
      "step": 4050
    },
    {
      "epoch": 1.9118340405469119,
      "grad_norm": 0.8714926838874817,
      "learning_rate": 4.874343448226764e-05,
      "loss": 0.2049,
      "num_input_tokens_seen": 3511904,
      "step": 4055
    },
    {
      "epoch": 1.9141914191419143,
      "grad_norm": 1.083170771598816,
      "learning_rate": 4.874035931221955e-05,
      "loss": 0.2,
      "num_input_tokens_seen": 3516736,
      "step": 4060
    },
    {
      "epoch": 1.9165487977369167,
      "grad_norm": 0.6152244210243225,
      "learning_rate": 4.8737280481109724e-05,
      "loss": 0.1816,
      "num_input_tokens_seen": 3520800,
      "step": 4065
    },
    {
      "epoch": 1.918906176331919,
      "grad_norm": 1.0183947086334229,
      "learning_rate": 4.873419798941294e-05,
      "loss": 0.1318,
      "num_input_tokens_seen": 3524784,
      "step": 4070
    },
    {
      "epoch": 1.9212635549269212,
      "grad_norm": 1.0302813053131104,
      "learning_rate": 4.873111183760458e-05,
      "loss": 0.2534,
      "num_input_tokens_seen": 3528992,
      "step": 4075
    },
    {
      "epoch": 1.9236209335219236,
      "grad_norm": 0.44073739647865295,
      "learning_rate": 4.8728022026160537e-05,
      "loss": 0.2636,
      "num_input_tokens_seen": 3533344,
      "step": 4080
    },
    {
      "epoch": 1.9259783121169258,
      "grad_norm": 1.0513882637023926,
      "learning_rate": 4.872492855555732e-05,
      "loss": 0.188,
      "num_input_tokens_seen": 3538112,
      "step": 4085
    },
    {
      "epoch": 1.9283356907119282,
      "grad_norm": 0.43003350496292114,
      "learning_rate": 4.8721831426271956e-05,
      "loss": 0.1675,
      "num_input_tokens_seen": 3542608,
      "step": 4090
    },
    {
      "epoch": 1.9306930693069306,
      "grad_norm": 0.7563297152519226,
      "learning_rate": 4.87187306387821e-05,
      "loss": 0.1969,
      "num_input_tokens_seen": 3547344,
      "step": 4095
    },
    {
      "epoch": 1.933050447901933,
      "grad_norm": 1.3683894872665405,
      "learning_rate": 4.87156261935659e-05,
      "loss": 0.2506,
      "num_input_tokens_seen": 3551376,
      "step": 4100
    },
    {
      "epoch": 1.9354078264969354,
      "grad_norm": 1.513708233833313,
      "learning_rate": 4.871251809110211e-05,
      "loss": 0.3068,
      "num_input_tokens_seen": 3555744,
      "step": 4105
    },
    {
      "epoch": 1.9377652050919378,
      "grad_norm": 1.0064501762390137,
      "learning_rate": 4.8709406331870044e-05,
      "loss": 0.2555,
      "num_input_tokens_seen": 3560256,
      "step": 4110
    },
    {
      "epoch": 1.9401225836869402,
      "grad_norm": 1.669312596321106,
      "learning_rate": 4.8706290916349574e-05,
      "loss": 0.2905,
      "num_input_tokens_seen": 3564096,
      "step": 4115
    },
    {
      "epoch": 1.9424799622819426,
      "grad_norm": 1.3084686994552612,
      "learning_rate": 4.8703171845021134e-05,
      "loss": 0.2726,
      "num_input_tokens_seen": 3567888,
      "step": 4120
    },
    {
      "epoch": 1.9448373408769448,
      "grad_norm": 2.076873540878296,
      "learning_rate": 4.870004911836572e-05,
      "loss": 0.2441,
      "num_input_tokens_seen": 3571776,
      "step": 4125
    },
    {
      "epoch": 1.9471947194719472,
      "grad_norm": 0.9883481860160828,
      "learning_rate": 4.869692273686489e-05,
      "loss": 0.234,
      "num_input_tokens_seen": 3576528,
      "step": 4130
    },
    {
      "epoch": 1.9495520980669494,
      "grad_norm": 1.0115721225738525,
      "learning_rate": 4.869379270100079e-05,
      "loss": 0.1436,
      "num_input_tokens_seen": 3580352,
      "step": 4135
    },
    {
      "epoch": 1.9519094766619518,
      "grad_norm": 1.570620059967041,
      "learning_rate": 4.86906590112561e-05,
      "loss": 0.2604,
      "num_input_tokens_seen": 3586272,
      "step": 4140
    },
    {
      "epoch": 1.9542668552569542,
      "grad_norm": 3.5202832221984863,
      "learning_rate": 4.8687521668114064e-05,
      "loss": 0.1966,
      "num_input_tokens_seen": 3590144,
      "step": 4145
    },
    {
      "epoch": 1.9566242338519566,
      "grad_norm": 0.8870676755905151,
      "learning_rate": 4.868438067205853e-05,
      "loss": 0.1569,
      "num_input_tokens_seen": 3594656,
      "step": 4150
    },
    {
      "epoch": 1.958981612446959,
      "grad_norm": 1.6622470617294312,
      "learning_rate": 4.8681236023573844e-05,
      "loss": 0.1924,
      "num_input_tokens_seen": 3599536,
      "step": 4155
    },
    {
      "epoch": 1.9613389910419614,
      "grad_norm": 0.7699506878852844,
      "learning_rate": 4.867808772314497e-05,
      "loss": 0.2487,
      "num_input_tokens_seen": 3605136,
      "step": 4160
    },
    {
      "epoch": 1.9636963696369638,
      "grad_norm": 0.8963794708251953,
      "learning_rate": 4.867493577125741e-05,
      "loss": 0.2322,
      "num_input_tokens_seen": 3609728,
      "step": 4165
    },
    {
      "epoch": 1.9660537482319662,
      "grad_norm": 1.6131166219711304,
      "learning_rate": 4.867178016839725e-05,
      "loss": 0.2482,
      "num_input_tokens_seen": 3614320,
      "step": 4170
    },
    {
      "epoch": 1.9684111268269684,
      "grad_norm": 1.7619181871414185,
      "learning_rate": 4.8668620915051094e-05,
      "loss": 0.2141,
      "num_input_tokens_seen": 3618400,
      "step": 4175
    },
    {
      "epoch": 1.9707685054219708,
      "grad_norm": 0.4945549964904785,
      "learning_rate": 4.866545801170616e-05,
      "loss": 0.2341,
      "num_input_tokens_seen": 3622080,
      "step": 4180
    },
    {
      "epoch": 1.973125884016973,
      "grad_norm": 1.401039719581604,
      "learning_rate": 4.86622914588502e-05,
      "loss": 0.2087,
      "num_input_tokens_seen": 3626656,
      "step": 4185
    },
    {
      "epoch": 1.9754832626119754,
      "grad_norm": 1.3825645446777344,
      "learning_rate": 4.865912125697154e-05,
      "loss": 0.1947,
      "num_input_tokens_seen": 3630912,
      "step": 4190
    },
    {
      "epoch": 1.9778406412069778,
      "grad_norm": 1.3922350406646729,
      "learning_rate": 4.865594740655907e-05,
      "loss": 0.3013,
      "num_input_tokens_seen": 3634992,
      "step": 4195
    },
    {
      "epoch": 1.9801980198019802,
      "grad_norm": 0.5182673931121826,
      "learning_rate": 4.865276990810222e-05,
      "loss": 0.178,
      "num_input_tokens_seen": 3638880,
      "step": 4200
    },
    {
      "epoch": 1.9801980198019802,
      "eval_loss": 0.22056177258491516,
      "eval_runtime": 22.1388,
      "eval_samples_per_second": 42.595,
      "eval_steps_per_second": 21.32,
      "num_input_tokens_seen": 3638880,
      "step": 4200
    },
    {
      "epoch": 1.9825553983969826,
      "grad_norm": 0.7442673444747925,
      "learning_rate": 4.8649588762091016e-05,
      "loss": 0.1532,
      "num_input_tokens_seen": 3643232,
      "step": 4205
    },
    {
      "epoch": 1.984912776991985,
      "grad_norm": 1.8096239566802979,
      "learning_rate": 4.8646403969016016e-05,
      "loss": 0.2355,
      "num_input_tokens_seen": 3647728,
      "step": 4210
    },
    {
      "epoch": 1.9872701555869874,
      "grad_norm": 0.5421813130378723,
      "learning_rate": 4.864321552936838e-05,
      "loss": 0.1627,
      "num_input_tokens_seen": 3651248,
      "step": 4215
    },
    {
      "epoch": 1.9896275341819898,
      "grad_norm": 0.5422051548957825,
      "learning_rate": 4.864002344363978e-05,
      "loss": 0.1591,
      "num_input_tokens_seen": 3655712,
      "step": 4220
    },
    {
      "epoch": 1.991984912776992,
      "grad_norm": 1.925114393234253,
      "learning_rate": 4.863682771232248e-05,
      "loss": 0.2256,
      "num_input_tokens_seen": 3660336,
      "step": 4225
    },
    {
      "epoch": 1.9943422913719944,
      "grad_norm": 1.3278346061706543,
      "learning_rate": 4.8633628335909324e-05,
      "loss": 0.2119,
      "num_input_tokens_seen": 3664880,
      "step": 4230
    },
    {
      "epoch": 1.9966996699669965,
      "grad_norm": 0.831552267074585,
      "learning_rate": 4.8630425314893676e-05,
      "loss": 0.2442,
      "num_input_tokens_seen": 3669216,
      "step": 4235
    },
    {
      "epoch": 1.999057048561999,
      "grad_norm": 1.416364312171936,
      "learning_rate": 4.862721864976948e-05,
      "loss": 0.2576,
      "num_input_tokens_seen": 3673840,
      "step": 4240
    },
    {
      "epoch": 2.0014144271570014,
      "grad_norm": 0.541938304901123,
      "learning_rate": 4.862400834103125e-05,
      "loss": 0.1781,
      "num_input_tokens_seen": 3678416,
      "step": 4245
    },
    {
      "epoch": 2.0037718057520038,
      "grad_norm": 1.753805160522461,
      "learning_rate": 4.862079438917406e-05,
      "loss": 0.2268,
      "num_input_tokens_seen": 3682656,
      "step": 4250
    },
    {
      "epoch": 2.006129184347006,
      "grad_norm": 0.7978367805480957,
      "learning_rate": 4.8617576794693536e-05,
      "loss": 0.1704,
      "num_input_tokens_seen": 3687472,
      "step": 4255
    },
    {
      "epoch": 2.0084865629420086,
      "grad_norm": 1.99647057056427,
      "learning_rate": 4.8614355558085875e-05,
      "loss": 0.1671,
      "num_input_tokens_seen": 3691168,
      "step": 4260
    },
    {
      "epoch": 2.010843941537011,
      "grad_norm": 0.7308064103126526,
      "learning_rate": 4.861113067984783e-05,
      "loss": 0.1876,
      "num_input_tokens_seen": 3695312,
      "step": 4265
    },
    {
      "epoch": 2.0132013201320134,
      "grad_norm": 1.4476858377456665,
      "learning_rate": 4.860790216047671e-05,
      "loss": 0.1619,
      "num_input_tokens_seen": 3699744,
      "step": 4270
    },
    {
      "epoch": 2.0155586987270158,
      "grad_norm": 1.5779153108596802,
      "learning_rate": 4.860467000047041e-05,
      "loss": 0.277,
      "num_input_tokens_seen": 3704416,
      "step": 4275
    },
    {
      "epoch": 2.0179160773220177,
      "grad_norm": 0.6920443177223206,
      "learning_rate": 4.860143420032737e-05,
      "loss": 0.1858,
      "num_input_tokens_seen": 3709344,
      "step": 4280
    },
    {
      "epoch": 2.02027345591702,
      "grad_norm": 1.631819248199463,
      "learning_rate": 4.859819476054657e-05,
      "loss": 0.1806,
      "num_input_tokens_seen": 3714368,
      "step": 4285
    },
    {
      "epoch": 2.0226308345120225,
      "grad_norm": 0.9499823451042175,
      "learning_rate": 4.859495168162758e-05,
      "loss": 0.1972,
      "num_input_tokens_seen": 3718688,
      "step": 4290
    },
    {
      "epoch": 2.024988213107025,
      "grad_norm": 0.7351589798927307,
      "learning_rate": 4.859170496407054e-05,
      "loss": 0.257,
      "num_input_tokens_seen": 3723120,
      "step": 4295
    },
    {
      "epoch": 2.0273455917020273,
      "grad_norm": 0.5024363994598389,
      "learning_rate": 4.8588454608376114e-05,
      "loss": 0.1782,
      "num_input_tokens_seen": 3727312,
      "step": 4300
    },
    {
      "epoch": 2.0297029702970297,
      "grad_norm": 1.3905606269836426,
      "learning_rate": 4.8585200615045555e-05,
      "loss": 0.2141,
      "num_input_tokens_seen": 3731696,
      "step": 4305
    },
    {
      "epoch": 2.032060348892032,
      "grad_norm": 0.9193349480628967,
      "learning_rate": 4.8581942984580674e-05,
      "loss": 0.2193,
      "num_input_tokens_seen": 3735968,
      "step": 4310
    },
    {
      "epoch": 2.0344177274870345,
      "grad_norm": 1.0417513847351074,
      "learning_rate": 4.857868171748384e-05,
      "loss": 0.2031,
      "num_input_tokens_seen": 3740256,
      "step": 4315
    },
    {
      "epoch": 2.036775106082037,
      "grad_norm": 1.7688523530960083,
      "learning_rate": 4.8575416814257976e-05,
      "loss": 0.171,
      "num_input_tokens_seen": 3744752,
      "step": 4320
    },
    {
      "epoch": 2.0391324846770393,
      "grad_norm": 0.6291093230247498,
      "learning_rate": 4.857214827540657e-05,
      "loss": 0.2146,
      "num_input_tokens_seen": 3749248,
      "step": 4325
    },
    {
      "epoch": 2.0414898632720413,
      "grad_norm": 1.8121201992034912,
      "learning_rate": 4.856887610143367e-05,
      "loss": 0.1667,
      "num_input_tokens_seen": 3753440,
      "step": 4330
    },
    {
      "epoch": 2.0438472418670437,
      "grad_norm": 0.63355553150177,
      "learning_rate": 4.8565600292843896e-05,
      "loss": 0.2469,
      "num_input_tokens_seen": 3758272,
      "step": 4335
    },
    {
      "epoch": 2.046204620462046,
      "grad_norm": 1.994083285331726,
      "learning_rate": 4.856232085014241e-05,
      "loss": 0.1936,
      "num_input_tokens_seen": 3762128,
      "step": 4340
    },
    {
      "epoch": 2.0485619990570485,
      "grad_norm": 0.6318020820617676,
      "learning_rate": 4.855903777383495e-05,
      "loss": 0.1697,
      "num_input_tokens_seen": 3766432,
      "step": 4345
    },
    {
      "epoch": 2.050919377652051,
      "grad_norm": 1.3628604412078857,
      "learning_rate": 4.85557510644278e-05,
      "loss": 0.2544,
      "num_input_tokens_seen": 3770768,
      "step": 4350
    },
    {
      "epoch": 2.0532767562470533,
      "grad_norm": 1.7040085792541504,
      "learning_rate": 4.855246072242782e-05,
      "loss": 0.2135,
      "num_input_tokens_seen": 3774752,
      "step": 4355
    },
    {
      "epoch": 2.0556341348420557,
      "grad_norm": 1.3417015075683594,
      "learning_rate": 4.8549166748342414e-05,
      "loss": 0.1875,
      "num_input_tokens_seen": 3778576,
      "step": 4360
    },
    {
      "epoch": 2.057991513437058,
      "grad_norm": 1.5162183046340942,
      "learning_rate": 4.8545869142679556e-05,
      "loss": 0.2035,
      "num_input_tokens_seen": 3782384,
      "step": 4365
    },
    {
      "epoch": 2.0603488920320605,
      "grad_norm": 0.47645533084869385,
      "learning_rate": 4.8542567905947776e-05,
      "loss": 0.2305,
      "num_input_tokens_seen": 3786512,
      "step": 4370
    },
    {
      "epoch": 2.062706270627063,
      "grad_norm": 1.3546844720840454,
      "learning_rate": 4.853926303865618e-05,
      "loss": 0.2601,
      "num_input_tokens_seen": 3790416,
      "step": 4375
    },
    {
      "epoch": 2.065063649222065,
      "grad_norm": 0.4865717589855194,
      "learning_rate": 4.853595454131441e-05,
      "loss": 0.2005,
      "num_input_tokens_seen": 3794992,
      "step": 4380
    },
    {
      "epoch": 2.0674210278170673,
      "grad_norm": 0.875678300857544,
      "learning_rate": 4.8532642414432674e-05,
      "loss": 0.1412,
      "num_input_tokens_seen": 3799376,
      "step": 4385
    },
    {
      "epoch": 2.0697784064120697,
      "grad_norm": 0.7629839181900024,
      "learning_rate": 4.8529326658521754e-05,
      "loss": 0.1716,
      "num_input_tokens_seen": 3803824,
      "step": 4390
    },
    {
      "epoch": 2.072135785007072,
      "grad_norm": 1.526282548904419,
      "learning_rate": 4.8526007274092965e-05,
      "loss": 0.178,
      "num_input_tokens_seen": 3808576,
      "step": 4395
    },
    {
      "epoch": 2.0744931636020745,
      "grad_norm": 1.0130016803741455,
      "learning_rate": 4.852268426165822e-05,
      "loss": 0.2565,
      "num_input_tokens_seen": 3812624,
      "step": 4400
    },
    {
      "epoch": 2.0744931636020745,
      "eval_loss": 0.21659815311431885,
      "eval_runtime": 22.1949,
      "eval_samples_per_second": 42.487,
      "eval_steps_per_second": 21.266,
      "num_input_tokens_seen": 3812624,
      "step": 4400
    },
    {
      "epoch": 2.076850542197077,
      "grad_norm": 0.7543633580207825,
      "learning_rate": 4.851935762172995e-05,
      "loss": 0.1704,
      "num_input_tokens_seen": 3816576,
      "step": 4405
    },
    {
      "epoch": 2.0792079207920793,
      "grad_norm": 0.7802368998527527,
      "learning_rate": 4.8516027354821175e-05,
      "loss": 0.244,
      "num_input_tokens_seen": 3821328,
      "step": 4410
    },
    {
      "epoch": 2.0815652993870817,
      "grad_norm": 1.244262456893921,
      "learning_rate": 4.851269346144546e-05,
      "loss": 0.1935,
      "num_input_tokens_seen": 3825488,
      "step": 4415
    },
    {
      "epoch": 2.083922677982084,
      "grad_norm": 0.9237592220306396,
      "learning_rate": 4.850935594211693e-05,
      "loss": 0.2202,
      "num_input_tokens_seen": 3829712,
      "step": 4420
    },
    {
      "epoch": 2.0862800565770865,
      "grad_norm": 2.1337759494781494,
      "learning_rate": 4.850601479735029e-05,
      "loss": 0.2325,
      "num_input_tokens_seen": 3833936,
      "step": 4425
    },
    {
      "epoch": 2.0886374351720884,
      "grad_norm": 1.491941213607788,
      "learning_rate": 4.850267002766076e-05,
      "loss": 0.1783,
      "num_input_tokens_seen": 3838816,
      "step": 4430
    },
    {
      "epoch": 2.090994813767091,
      "grad_norm": 0.5460907816886902,
      "learning_rate": 4.849932163356417e-05,
      "loss": 0.16,
      "num_input_tokens_seen": 3843136,
      "step": 4435
    },
    {
      "epoch": 2.0933521923620932,
      "grad_norm": 0.7348231077194214,
      "learning_rate": 4.8495969615576864e-05,
      "loss": 0.1395,
      "num_input_tokens_seen": 3848144,
      "step": 4440
    },
    {
      "epoch": 2.0957095709570956,
      "grad_norm": 0.49035120010375977,
      "learning_rate": 4.849261397421577e-05,
      "loss": 0.2086,
      "num_input_tokens_seen": 3853584,
      "step": 4445
    },
    {
      "epoch": 2.098066949552098,
      "grad_norm": 1.2966687679290771,
      "learning_rate": 4.848925470999839e-05,
      "loss": 0.1935,
      "num_input_tokens_seen": 3857808,
      "step": 4450
    },
    {
      "epoch": 2.1004243281471005,
      "grad_norm": 0.6298763751983643,
      "learning_rate": 4.848589182344273e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 3861840,
      "step": 4455
    },
    {
      "epoch": 2.102781706742103,
      "grad_norm": 0.6601333022117615,
      "learning_rate": 4.848252531506742e-05,
      "loss": 0.1673,
      "num_input_tokens_seen": 3866128,
      "step": 4460
    },
    {
      "epoch": 2.1051390853371053,
      "grad_norm": 0.5212008953094482,
      "learning_rate": 4.847915518539161e-05,
      "loss": 0.2031,
      "num_input_tokens_seen": 3870816,
      "step": 4465
    },
    {
      "epoch": 2.1074964639321077,
      "grad_norm": 1.6701128482818604,
      "learning_rate": 4.847578143493501e-05,
      "loss": 0.2466,
      "num_input_tokens_seen": 3875600,
      "step": 4470
    },
    {
      "epoch": 2.10985384252711,
      "grad_norm": 0.6385613679885864,
      "learning_rate": 4.847240406421789e-05,
      "loss": 0.2034,
      "num_input_tokens_seen": 3879552,
      "step": 4475
    },
    {
      "epoch": 2.112211221122112,
      "grad_norm": 0.5201774835586548,
      "learning_rate": 4.84690230737611e-05,
      "loss": 0.1265,
      "num_input_tokens_seen": 3883808,
      "step": 4480
    },
    {
      "epoch": 2.1145685997171144,
      "grad_norm": 1.5857723951339722,
      "learning_rate": 4.846563846408602e-05,
      "loss": 0.2324,
      "num_input_tokens_seen": 3888752,
      "step": 4485
    },
    {
      "epoch": 2.116925978312117,
      "grad_norm": 0.46428781747817993,
      "learning_rate": 4.84622502357146e-05,
      "loss": 0.2742,
      "num_input_tokens_seen": 3893376,
      "step": 4490
    },
    {
      "epoch": 2.119283356907119,
      "grad_norm": 1.9103113412857056,
      "learning_rate": 4.8458858389169345e-05,
      "loss": 0.2315,
      "num_input_tokens_seen": 3896800,
      "step": 4495
    },
    {
      "epoch": 2.1216407355021216,
      "grad_norm": 1.1005498170852661,
      "learning_rate": 4.8455462924973334e-05,
      "loss": 0.1764,
      "num_input_tokens_seen": 3900832,
      "step": 4500
    },
    {
      "epoch": 2.123998114097124,
      "grad_norm": 1.1681346893310547,
      "learning_rate": 4.845206384365018e-05,
      "loss": 0.153,
      "num_input_tokens_seen": 3904880,
      "step": 4505
    },
    {
      "epoch": 2.1263554926921264,
      "grad_norm": 1.7681854963302612,
      "learning_rate": 4.844866114572405e-05,
      "loss": 0.1812,
      "num_input_tokens_seen": 3909488,
      "step": 4510
    },
    {
      "epoch": 2.128712871287129,
      "grad_norm": 1.4868360757827759,
      "learning_rate": 4.8445254831719706e-05,
      "loss": 0.2262,
      "num_input_tokens_seen": 3913824,
      "step": 4515
    },
    {
      "epoch": 2.1310702498821312,
      "grad_norm": 1.9704465866088867,
      "learning_rate": 4.8441844902162434e-05,
      "loss": 0.2681,
      "num_input_tokens_seen": 3917872,
      "step": 4520
    },
    {
      "epoch": 2.1334276284771336,
      "grad_norm": 0.8310115933418274,
      "learning_rate": 4.843843135757809e-05,
      "loss": 0.1823,
      "num_input_tokens_seen": 3922384,
      "step": 4525
    },
    {
      "epoch": 2.1357850070721356,
      "grad_norm": 0.6368492841720581,
      "learning_rate": 4.843501419849308e-05,
      "loss": 0.2225,
      "num_input_tokens_seen": 3927216,
      "step": 4530
    },
    {
      "epoch": 2.138142385667138,
      "grad_norm": 0.9823870658874512,
      "learning_rate": 4.8431593425434386e-05,
      "loss": 0.2196,
      "num_input_tokens_seen": 3931584,
      "step": 4535
    },
    {
      "epoch": 2.1404997642621404,
      "grad_norm": 2.0544285774230957,
      "learning_rate": 4.8428169038929526e-05,
      "loss": 0.1571,
      "num_input_tokens_seen": 3936224,
      "step": 4540
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 2.3630824089050293,
      "learning_rate": 4.8424741039506575e-05,
      "loss": 0.205,
      "num_input_tokens_seen": 3940880,
      "step": 4545
    },
    {
      "epoch": 2.145214521452145,
      "grad_norm": 1.0749551057815552,
      "learning_rate": 4.842130942769419e-05,
      "loss": 0.1953,
      "num_input_tokens_seen": 3945136,
      "step": 4550
    },
    {
      "epoch": 2.1475719000471476,
      "grad_norm": 0.8309681415557861,
      "learning_rate": 4.841787420402156e-05,
      "loss": 0.1864,
      "num_input_tokens_seen": 3948992,
      "step": 4555
    },
    {
      "epoch": 2.14992927864215,
      "grad_norm": 1.6883985996246338,
      "learning_rate": 4.841443536901844e-05,
      "loss": 0.2208,
      "num_input_tokens_seen": 3953200,
      "step": 4560
    },
    {
      "epoch": 2.1522866572371524,
      "grad_norm": 0.8596568703651428,
      "learning_rate": 4.841099292321514e-05,
      "loss": 0.1774,
      "num_input_tokens_seen": 3957136,
      "step": 4565
    },
    {
      "epoch": 2.154644035832155,
      "grad_norm": 0.7402194738388062,
      "learning_rate": 4.8407546867142525e-05,
      "loss": 0.2532,
      "num_input_tokens_seen": 3961232,
      "step": 4570
    },
    {
      "epoch": 2.157001414427157,
      "grad_norm": 1.8702244758605957,
      "learning_rate": 4.840409720133203e-05,
      "loss": 0.2328,
      "num_input_tokens_seen": 3964784,
      "step": 4575
    },
    {
      "epoch": 2.159358793022159,
      "grad_norm": 1.7025638818740845,
      "learning_rate": 4.8400643926315634e-05,
      "loss": 0.2324,
      "num_input_tokens_seen": 3969648,
      "step": 4580
    },
    {
      "epoch": 2.1617161716171616,
      "grad_norm": 1.260834813117981,
      "learning_rate": 4.839718704262587e-05,
      "loss": 0.1576,
      "num_input_tokens_seen": 3973680,
      "step": 4585
    },
    {
      "epoch": 2.164073550212164,
      "grad_norm": 1.1026105880737305,
      "learning_rate": 4.839372655079585e-05,
      "loss": 0.1984,
      "num_input_tokens_seen": 3977840,
      "step": 4590
    },
    {
      "epoch": 2.1664309288071664,
      "grad_norm": 1.3099825382232666,
      "learning_rate": 4.83902624513592e-05,
      "loss": 0.2462,
      "num_input_tokens_seen": 3981920,
      "step": 4595
    },
    {
      "epoch": 2.1687883074021688,
      "grad_norm": 1.9627060890197754,
      "learning_rate": 4.838679474485014e-05,
      "loss": 0.3385,
      "num_input_tokens_seen": 3986544,
      "step": 4600
    },
    {
      "epoch": 2.1687883074021688,
      "eval_loss": 0.21835185587406158,
      "eval_runtime": 22.1618,
      "eval_samples_per_second": 42.551,
      "eval_steps_per_second": 21.298,
      "num_input_tokens_seen": 3986544,
      "step": 4600
    },
    {
      "epoch": 2.171145685997171,
      "grad_norm": 0.7111688852310181,
      "learning_rate": 4.838332343180343e-05,
      "loss": 0.213,
      "num_input_tokens_seen": 3990896,
      "step": 4605
    },
    {
      "epoch": 2.1735030645921736,
      "grad_norm": 0.810009241104126,
      "learning_rate": 4.83798485127544e-05,
      "loss": 0.1947,
      "num_input_tokens_seen": 3995296,
      "step": 4610
    },
    {
      "epoch": 2.175860443187176,
      "grad_norm": 1.059201955795288,
      "learning_rate": 4.837636998823892e-05,
      "loss": 0.1348,
      "num_input_tokens_seen": 3999728,
      "step": 4615
    },
    {
      "epoch": 2.1782178217821784,
      "grad_norm": 0.5995899438858032,
      "learning_rate": 4.8372887858793414e-05,
      "loss": 0.1699,
      "num_input_tokens_seen": 4004160,
      "step": 4620
    },
    {
      "epoch": 2.1805752003771808,
      "grad_norm": 0.6272031664848328,
      "learning_rate": 4.836940212495489e-05,
      "loss": 0.2062,
      "num_input_tokens_seen": 4008928,
      "step": 4625
    },
    {
      "epoch": 2.1829325789721827,
      "grad_norm": 0.8270480632781982,
      "learning_rate": 4.836591278726087e-05,
      "loss": 0.2167,
      "num_input_tokens_seen": 4012928,
      "step": 4630
    },
    {
      "epoch": 2.185289957567185,
      "grad_norm": 1.0714101791381836,
      "learning_rate": 4.836241984624947e-05,
      "loss": 0.149,
      "num_input_tokens_seen": 4017376,
      "step": 4635
    },
    {
      "epoch": 2.1876473361621875,
      "grad_norm": 0.5038570761680603,
      "learning_rate": 4.8358923302459336e-05,
      "loss": 0.2277,
      "num_input_tokens_seen": 4021904,
      "step": 4640
    },
    {
      "epoch": 2.19000471475719,
      "grad_norm": 0.6612584590911865,
      "learning_rate": 4.835542315642968e-05,
      "loss": 0.1379,
      "num_input_tokens_seen": 4026368,
      "step": 4645
    },
    {
      "epoch": 2.1923620933521923,
      "grad_norm": 0.6052960753440857,
      "learning_rate": 4.8351919408700274e-05,
      "loss": 0.2102,
      "num_input_tokens_seen": 4030464,
      "step": 4650
    },
    {
      "epoch": 2.1947194719471947,
      "grad_norm": 1.0964645147323608,
      "learning_rate": 4.834841205981144e-05,
      "loss": 0.1993,
      "num_input_tokens_seen": 4034560,
      "step": 4655
    },
    {
      "epoch": 2.197076850542197,
      "grad_norm": 1.0052169561386108,
      "learning_rate": 4.8344901110304054e-05,
      "loss": 0.2598,
      "num_input_tokens_seen": 4038432,
      "step": 4660
    },
    {
      "epoch": 2.1994342291371995,
      "grad_norm": 1.1425986289978027,
      "learning_rate": 4.8341386560719534e-05,
      "loss": 0.1925,
      "num_input_tokens_seen": 4042432,
      "step": 4665
    },
    {
      "epoch": 2.201791607732202,
      "grad_norm": 0.8683399558067322,
      "learning_rate": 4.833786841159989e-05,
      "loss": 0.1696,
      "num_input_tokens_seen": 4046416,
      "step": 4670
    },
    {
      "epoch": 2.2041489863272044,
      "grad_norm": 3.211498737335205,
      "learning_rate": 4.833434666348765e-05,
      "loss": 0.278,
      "num_input_tokens_seen": 4050208,
      "step": 4675
    },
    {
      "epoch": 2.2065063649222063,
      "grad_norm": 0.9292176961898804,
      "learning_rate": 4.833082131692592e-05,
      "loss": 0.2137,
      "num_input_tokens_seen": 4054928,
      "step": 4680
    },
    {
      "epoch": 2.2088637435172087,
      "grad_norm": 1.3190561532974243,
      "learning_rate": 4.832729237245835e-05,
      "loss": 0.2656,
      "num_input_tokens_seen": 4059360,
      "step": 4685
    },
    {
      "epoch": 2.211221122112211,
      "grad_norm": 1.6224479675292969,
      "learning_rate": 4.8323759830629145e-05,
      "loss": 0.2134,
      "num_input_tokens_seen": 4063424,
      "step": 4690
    },
    {
      "epoch": 2.2135785007072135,
      "grad_norm": 1.042535662651062,
      "learning_rate": 4.8320223691983066e-05,
      "loss": 0.2254,
      "num_input_tokens_seen": 4067248,
      "step": 4695
    },
    {
      "epoch": 2.215935879302216,
      "grad_norm": 1.4689944982528687,
      "learning_rate": 4.831668395706544e-05,
      "loss": 0.1917,
      "num_input_tokens_seen": 4070992,
      "step": 4700
    },
    {
      "epoch": 2.2182932578972183,
      "grad_norm": 0.8297308087348938,
      "learning_rate": 4.8313140626422125e-05,
      "loss": 0.125,
      "num_input_tokens_seen": 4075712,
      "step": 4705
    },
    {
      "epoch": 2.2206506364922207,
      "grad_norm": 0.8800947666168213,
      "learning_rate": 4.830959370059956e-05,
      "loss": 0.2369,
      "num_input_tokens_seen": 4079744,
      "step": 4710
    },
    {
      "epoch": 2.223008015087223,
      "grad_norm": 1.1348685026168823,
      "learning_rate": 4.830604318014472e-05,
      "loss": 0.2203,
      "num_input_tokens_seen": 4083776,
      "step": 4715
    },
    {
      "epoch": 2.2253653936822255,
      "grad_norm": 0.4717133343219757,
      "learning_rate": 4.830248906560514e-05,
      "loss": 0.1931,
      "num_input_tokens_seen": 4087232,
      "step": 4720
    },
    {
      "epoch": 2.227722772277228,
      "grad_norm": 0.8618420362472534,
      "learning_rate": 4.829893135752891e-05,
      "loss": 0.1934,
      "num_input_tokens_seen": 4091840,
      "step": 4725
    },
    {
      "epoch": 2.23008015087223,
      "grad_norm": 1.8446707725524902,
      "learning_rate": 4.829537005646466e-05,
      "loss": 0.2265,
      "num_input_tokens_seen": 4096176,
      "step": 4730
    },
    {
      "epoch": 2.2324375294672323,
      "grad_norm": 0.6377031207084656,
      "learning_rate": 4.8291805162961615e-05,
      "loss": 0.1935,
      "num_input_tokens_seen": 4100080,
      "step": 4735
    },
    {
      "epoch": 2.2347949080622347,
      "grad_norm": 0.9078042507171631,
      "learning_rate": 4.82882366775695e-05,
      "loss": 0.1934,
      "num_input_tokens_seen": 4106304,
      "step": 4740
    },
    {
      "epoch": 2.237152286657237,
      "grad_norm": 0.4740830063819885,
      "learning_rate": 4.828466460083864e-05,
      "loss": 0.2313,
      "num_input_tokens_seen": 4111664,
      "step": 4745
    },
    {
      "epoch": 2.2395096652522395,
      "grad_norm": 0.7048792839050293,
      "learning_rate": 4.8281088933319877e-05,
      "loss": 0.2214,
      "num_input_tokens_seen": 4115936,
      "step": 4750
    },
    {
      "epoch": 2.241867043847242,
      "grad_norm": 1.121044397354126,
      "learning_rate": 4.827750967556464e-05,
      "loss": 0.244,
      "num_input_tokens_seen": 4120320,
      "step": 4755
    },
    {
      "epoch": 2.2442244224422443,
      "grad_norm": 0.6351733207702637,
      "learning_rate": 4.827392682812488e-05,
      "loss": 0.2284,
      "num_input_tokens_seen": 4124928,
      "step": 4760
    },
    {
      "epoch": 2.2465818010372467,
      "grad_norm": 0.5901702642440796,
      "learning_rate": 4.827034039155312e-05,
      "loss": 0.1739,
      "num_input_tokens_seen": 4128672,
      "step": 4765
    },
    {
      "epoch": 2.248939179632249,
      "grad_norm": 0.6615360975265503,
      "learning_rate": 4.8266750366402445e-05,
      "loss": 0.1776,
      "num_input_tokens_seen": 4132848,
      "step": 4770
    },
    {
      "epoch": 2.251296558227251,
      "grad_norm": 2.0059421062469482,
      "learning_rate": 4.8263156753226476e-05,
      "loss": 0.2124,
      "num_input_tokens_seen": 4136416,
      "step": 4775
    },
    {
      "epoch": 2.2536539368222535,
      "grad_norm": 0.9871169328689575,
      "learning_rate": 4.8259559552579394e-05,
      "loss": 0.1722,
      "num_input_tokens_seen": 4141040,
      "step": 4780
    },
    {
      "epoch": 2.256011315417256,
      "grad_norm": 1.260871410369873,
      "learning_rate": 4.825595876501593e-05,
      "loss": 0.1779,
      "num_input_tokens_seen": 4145600,
      "step": 4785
    },
    {
      "epoch": 2.2583686940122583,
      "grad_norm": 1.934314250946045,
      "learning_rate": 4.825235439109137e-05,
      "loss": 0.198,
      "num_input_tokens_seen": 4150336,
      "step": 4790
    },
    {
      "epoch": 2.2607260726072607,
      "grad_norm": 0.42774203419685364,
      "learning_rate": 4.824874643136156e-05,
      "loss": 0.2701,
      "num_input_tokens_seen": 4154192,
      "step": 4795
    },
    {
      "epoch": 2.263083451202263,
      "grad_norm": 0.894166886806488,
      "learning_rate": 4.824513488638288e-05,
      "loss": 0.1578,
      "num_input_tokens_seen": 4158272,
      "step": 4800
    },
    {
      "epoch": 2.263083451202263,
      "eval_loss": 0.21478000283241272,
      "eval_runtime": 22.1602,
      "eval_samples_per_second": 42.554,
      "eval_steps_per_second": 21.299,
      "num_input_tokens_seen": 4158272,
      "step": 4800
    },
    {
      "epoch": 2.2654408297972655,
      "grad_norm": 2.161294937133789,
      "learning_rate": 4.8241519756712293e-05,
      "loss": 0.309,
      "num_input_tokens_seen": 4162816,
      "step": 4805
    },
    {
      "epoch": 2.267798208392268,
      "grad_norm": 1.8676716089248657,
      "learning_rate": 4.8237901042907285e-05,
      "loss": 0.2484,
      "num_input_tokens_seen": 4166432,
      "step": 4810
    },
    {
      "epoch": 2.2701555869872703,
      "grad_norm": 0.7124615907669067,
      "learning_rate": 4.823427874552591e-05,
      "loss": 0.1827,
      "num_input_tokens_seen": 4170416,
      "step": 4815
    },
    {
      "epoch": 2.2725129655822727,
      "grad_norm": 1.0617220401763916,
      "learning_rate": 4.823065286512677e-05,
      "loss": 0.2238,
      "num_input_tokens_seen": 4174528,
      "step": 4820
    },
    {
      "epoch": 2.274870344177275,
      "grad_norm": 1.098046898841858,
      "learning_rate": 4.8227023402269025e-05,
      "loss": 0.1518,
      "num_input_tokens_seen": 4178944,
      "step": 4825
    },
    {
      "epoch": 2.2772277227722775,
      "grad_norm": 0.9970951080322266,
      "learning_rate": 4.822339035751239e-05,
      "loss": 0.1919,
      "num_input_tokens_seen": 4183344,
      "step": 4830
    },
    {
      "epoch": 2.2795851013672794,
      "grad_norm": 2.3114354610443115,
      "learning_rate": 4.8219753731417104e-05,
      "loss": 0.2368,
      "num_input_tokens_seen": 4187376,
      "step": 4835
    },
    {
      "epoch": 2.281942479962282,
      "grad_norm": 1.4461513757705688,
      "learning_rate": 4.821611352454401e-05,
      "loss": 0.2744,
      "num_input_tokens_seen": 4191680,
      "step": 4840
    },
    {
      "epoch": 2.2842998585572842,
      "grad_norm": 0.7595301866531372,
      "learning_rate": 4.8212469737454444e-05,
      "loss": 0.1626,
      "num_input_tokens_seen": 4196080,
      "step": 4845
    },
    {
      "epoch": 2.2866572371522866,
      "grad_norm": 0.8175957202911377,
      "learning_rate": 4.820882237071035e-05,
      "loss": 0.1726,
      "num_input_tokens_seen": 4200720,
      "step": 4850
    },
    {
      "epoch": 2.289014615747289,
      "grad_norm": 1.393265724182129,
      "learning_rate": 4.820517142487417e-05,
      "loss": 0.2358,
      "num_input_tokens_seen": 4205120,
      "step": 4855
    },
    {
      "epoch": 2.2913719943422914,
      "grad_norm": 0.8202688694000244,
      "learning_rate": 4.8201516900508956e-05,
      "loss": 0.2913,
      "num_input_tokens_seen": 4208320,
      "step": 4860
    },
    {
      "epoch": 2.293729372937294,
      "grad_norm": 0.5910454988479614,
      "learning_rate": 4.819785879817827e-05,
      "loss": 0.2127,
      "num_input_tokens_seen": 4212976,
      "step": 4865
    },
    {
      "epoch": 2.2960867515322962,
      "grad_norm": 0.9789940118789673,
      "learning_rate": 4.8194197118446226e-05,
      "loss": 0.154,
      "num_input_tokens_seen": 4218096,
      "step": 4870
    },
    {
      "epoch": 2.298444130127298,
      "grad_norm": 1.33669912815094,
      "learning_rate": 4.819053186187752e-05,
      "loss": 0.1506,
      "num_input_tokens_seen": 4222176,
      "step": 4875
    },
    {
      "epoch": 2.3008015087223006,
      "grad_norm": 2.195762872695923,
      "learning_rate": 4.818686302903736e-05,
      "loss": 0.205,
      "num_input_tokens_seen": 4226064,
      "step": 4880
    },
    {
      "epoch": 2.303158887317303,
      "grad_norm": 0.8657389879226685,
      "learning_rate": 4.818319062049154e-05,
      "loss": 0.2567,
      "num_input_tokens_seen": 4229744,
      "step": 4885
    },
    {
      "epoch": 2.3055162659123054,
      "grad_norm": 0.43348509073257446,
      "learning_rate": 4.817951463680639e-05,
      "loss": 0.1449,
      "num_input_tokens_seen": 4233744,
      "step": 4890
    },
    {
      "epoch": 2.307873644507308,
      "grad_norm": 1.0541759729385376,
      "learning_rate": 4.817583507854879e-05,
      "loss": 0.2121,
      "num_input_tokens_seen": 4238336,
      "step": 4895
    },
    {
      "epoch": 2.31023102310231,
      "grad_norm": 0.5092988014221191,
      "learning_rate": 4.817215194628617e-05,
      "loss": 0.1736,
      "num_input_tokens_seen": 4242176,
      "step": 4900
    },
    {
      "epoch": 2.3125884016973126,
      "grad_norm": 0.8653293251991272,
      "learning_rate": 4.816846524058653e-05,
      "loss": 0.2192,
      "num_input_tokens_seen": 4246832,
      "step": 4905
    },
    {
      "epoch": 2.314945780292315,
      "grad_norm": 0.7564467191696167,
      "learning_rate": 4.816477496201839e-05,
      "loss": 0.1713,
      "num_input_tokens_seen": 4251424,
      "step": 4910
    },
    {
      "epoch": 2.3173031588873174,
      "grad_norm": 0.5495389103889465,
      "learning_rate": 4.8161081111150845e-05,
      "loss": 0.1849,
      "num_input_tokens_seen": 4256128,
      "step": 4915
    },
    {
      "epoch": 2.31966053748232,
      "grad_norm": 2.088517189025879,
      "learning_rate": 4.815738368855354e-05,
      "loss": 0.1974,
      "num_input_tokens_seen": 4260384,
      "step": 4920
    },
    {
      "epoch": 2.322017916077322,
      "grad_norm": 0.7003379464149475,
      "learning_rate": 4.815368269479664e-05,
      "loss": 0.2019,
      "num_input_tokens_seen": 4264480,
      "step": 4925
    },
    {
      "epoch": 2.3243752946723246,
      "grad_norm": 0.9408512711524963,
      "learning_rate": 4.814997813045092e-05,
      "loss": 0.1858,
      "num_input_tokens_seen": 4267776,
      "step": 4930
    },
    {
      "epoch": 2.3267326732673266,
      "grad_norm": 0.6995845437049866,
      "learning_rate": 4.814626999608764e-05,
      "loss": 0.1981,
      "num_input_tokens_seen": 4271856,
      "step": 4935
    },
    {
      "epoch": 2.329090051862329,
      "grad_norm": 2.6708340644836426,
      "learning_rate": 4.814255829227865e-05,
      "loss": 0.295,
      "num_input_tokens_seen": 4275568,
      "step": 4940
    },
    {
      "epoch": 2.3314474304573314,
      "grad_norm": 1.1850035190582275,
      "learning_rate": 4.813884301959635e-05,
      "loss": 0.2108,
      "num_input_tokens_seen": 4279632,
      "step": 4945
    },
    {
      "epoch": 2.333804809052334,
      "grad_norm": 0.954846978187561,
      "learning_rate": 4.813512417861368e-05,
      "loss": 0.1734,
      "num_input_tokens_seen": 4284016,
      "step": 4950
    },
    {
      "epoch": 2.336162187647336,
      "grad_norm": 0.9899218678474426,
      "learning_rate": 4.813140176990411e-05,
      "loss": 0.1799,
      "num_input_tokens_seen": 4288032,
      "step": 4955
    },
    {
      "epoch": 2.3385195662423386,
      "grad_norm": 0.7605062127113342,
      "learning_rate": 4.8127675794041714e-05,
      "loss": 0.1937,
      "num_input_tokens_seen": 4292496,
      "step": 4960
    },
    {
      "epoch": 2.340876944837341,
      "grad_norm": 0.474362850189209,
      "learning_rate": 4.812394625160107e-05,
      "loss": 0.2297,
      "num_input_tokens_seen": 4296752,
      "step": 4965
    },
    {
      "epoch": 2.3432343234323434,
      "grad_norm": 1.0148601531982422,
      "learning_rate": 4.812021314315732e-05,
      "loss": 0.2089,
      "num_input_tokens_seen": 4302048,
      "step": 4970
    },
    {
      "epoch": 2.3455917020273453,
      "grad_norm": 1.8083875179290771,
      "learning_rate": 4.811647646928616e-05,
      "loss": 0.2178,
      "num_input_tokens_seen": 4307088,
      "step": 4975
    },
    {
      "epoch": 2.3479490806223478,
      "grad_norm": 0.9391839504241943,
      "learning_rate": 4.8112736230563814e-05,
      "loss": 0.1938,
      "num_input_tokens_seen": 4311184,
      "step": 4980
    },
    {
      "epoch": 2.35030645921735,
      "grad_norm": 0.65658038854599,
      "learning_rate": 4.81089924275671e-05,
      "loss": 0.2501,
      "num_input_tokens_seen": 4315104,
      "step": 4985
    },
    {
      "epoch": 2.3526638378123526,
      "grad_norm": 0.5473071932792664,
      "learning_rate": 4.810524506087335e-05,
      "loss": 0.1488,
      "num_input_tokens_seen": 4319424,
      "step": 4990
    },
    {
      "epoch": 2.355021216407355,
      "grad_norm": 1.246085286140442,
      "learning_rate": 4.810149413106044e-05,
      "loss": 0.2462,
      "num_input_tokens_seen": 4323712,
      "step": 4995
    },
    {
      "epoch": 2.3573785950023574,
      "grad_norm": 1.042881965637207,
      "learning_rate": 4.809773963870684e-05,
      "loss": 0.2113,
      "num_input_tokens_seen": 4328240,
      "step": 5000
    },
    {
      "epoch": 2.3573785950023574,
      "eval_loss": 0.2148771733045578,
      "eval_runtime": 22.1698,
      "eval_samples_per_second": 42.535,
      "eval_steps_per_second": 21.29,
      "num_input_tokens_seen": 4328240,
      "step": 5000
    },
    {
      "epoch": 2.3597359735973598,
      "grad_norm": 1.2957212924957275,
      "learning_rate": 4.809398158439151e-05,
      "loss": 0.1674,
      "num_input_tokens_seen": 4331856,
      "step": 5005
    },
    {
      "epoch": 2.362093352192362,
      "grad_norm": 0.6568636894226074,
      "learning_rate": 4.8090219968694005e-05,
      "loss": 0.144,
      "num_input_tokens_seen": 4336288,
      "step": 5010
    },
    {
      "epoch": 2.3644507307873646,
      "grad_norm": 1.1532766819000244,
      "learning_rate": 4.808645479219442e-05,
      "loss": 0.2504,
      "num_input_tokens_seen": 4340016,
      "step": 5015
    },
    {
      "epoch": 2.366808109382367,
      "grad_norm": 0.7467333674430847,
      "learning_rate": 4.8082686055473375e-05,
      "loss": 0.2713,
      "num_input_tokens_seen": 4345280,
      "step": 5020
    },
    {
      "epoch": 2.3691654879773694,
      "grad_norm": 1.7982455492019653,
      "learning_rate": 4.8078913759112066e-05,
      "loss": 0.2361,
      "num_input_tokens_seen": 4350496,
      "step": 5025
    },
    {
      "epoch": 2.3715228665723718,
      "grad_norm": 1.3505867719650269,
      "learning_rate": 4.807513790369223e-05,
      "loss": 0.1735,
      "num_input_tokens_seen": 4355024,
      "step": 5030
    },
    {
      "epoch": 2.3738802451673737,
      "grad_norm": 1.4179223775863647,
      "learning_rate": 4.8071358489796145e-05,
      "loss": 0.2266,
      "num_input_tokens_seen": 4359136,
      "step": 5035
    },
    {
      "epoch": 2.376237623762376,
      "grad_norm": 0.5825254917144775,
      "learning_rate": 4.806757551800665e-05,
      "loss": 0.1985,
      "num_input_tokens_seen": 4363264,
      "step": 5040
    },
    {
      "epoch": 2.3785950023573785,
      "grad_norm": 0.49773943424224854,
      "learning_rate": 4.806378898890713e-05,
      "loss": 0.1514,
      "num_input_tokens_seen": 4367840,
      "step": 5045
    },
    {
      "epoch": 2.380952380952381,
      "grad_norm": 0.834923267364502,
      "learning_rate": 4.80599989030815e-05,
      "loss": 0.2334,
      "num_input_tokens_seen": 4372848,
      "step": 5050
    },
    {
      "epoch": 2.3833097595473833,
      "grad_norm": 1.056357502937317,
      "learning_rate": 4.805620526111426e-05,
      "loss": 0.1773,
      "num_input_tokens_seen": 4376624,
      "step": 5055
    },
    {
      "epoch": 2.3856671381423857,
      "grad_norm": 1.5586864948272705,
      "learning_rate": 4.805240806359042e-05,
      "loss": 0.2429,
      "num_input_tokens_seen": 4380944,
      "step": 5060
    },
    {
      "epoch": 2.388024516737388,
      "grad_norm": 0.7111282348632812,
      "learning_rate": 4.804860731109557e-05,
      "loss": 0.1939,
      "num_input_tokens_seen": 4384880,
      "step": 5065
    },
    {
      "epoch": 2.3903818953323905,
      "grad_norm": 1.427431583404541,
      "learning_rate": 4.804480300421581e-05,
      "loss": 0.2138,
      "num_input_tokens_seen": 4389184,
      "step": 5070
    },
    {
      "epoch": 2.3927392739273925,
      "grad_norm": 0.6212351322174072,
      "learning_rate": 4.804099514353784e-05,
      "loss": 0.2118,
      "num_input_tokens_seen": 4393424,
      "step": 5075
    },
    {
      "epoch": 2.395096652522395,
      "grad_norm": 1.379512071609497,
      "learning_rate": 4.8037183729648867e-05,
      "loss": 0.2513,
      "num_input_tokens_seen": 4397680,
      "step": 5080
    },
    {
      "epoch": 2.3974540311173973,
      "grad_norm": 0.6681508421897888,
      "learning_rate": 4.803336876313666e-05,
      "loss": 0.1388,
      "num_input_tokens_seen": 4402288,
      "step": 5085
    },
    {
      "epoch": 2.3998114097123997,
      "grad_norm": 2.193281888961792,
      "learning_rate": 4.802955024458953e-05,
      "loss": 0.2493,
      "num_input_tokens_seen": 4406528,
      "step": 5090
    },
    {
      "epoch": 2.402168788307402,
      "grad_norm": 0.9636310935020447,
      "learning_rate": 4.802572817459634e-05,
      "loss": 0.2733,
      "num_input_tokens_seen": 4411392,
      "step": 5095
    },
    {
      "epoch": 2.4045261669024045,
      "grad_norm": 0.7646747827529907,
      "learning_rate": 4.802190255374651e-05,
      "loss": 0.1783,
      "num_input_tokens_seen": 4415600,
      "step": 5100
    },
    {
      "epoch": 2.406883545497407,
      "grad_norm": 0.7686719298362732,
      "learning_rate": 4.801807338263e-05,
      "loss": 0.1946,
      "num_input_tokens_seen": 4420176,
      "step": 5105
    },
    {
      "epoch": 2.4092409240924093,
      "grad_norm": 0.4586082100868225,
      "learning_rate": 4.8014240661837306e-05,
      "loss": 0.1742,
      "num_input_tokens_seen": 4423968,
      "step": 5110
    },
    {
      "epoch": 2.4115983026874117,
      "grad_norm": 1.756581425666809,
      "learning_rate": 4.80104043919595e-05,
      "loss": 0.2585,
      "num_input_tokens_seen": 4429840,
      "step": 5115
    },
    {
      "epoch": 2.413955681282414,
      "grad_norm": 1.693402886390686,
      "learning_rate": 4.800656457358815e-05,
      "loss": 0.1864,
      "num_input_tokens_seen": 4434592,
      "step": 5120
    },
    {
      "epoch": 2.4163130598774165,
      "grad_norm": 1.539862871170044,
      "learning_rate": 4.800272120731544e-05,
      "loss": 0.1856,
      "num_input_tokens_seen": 4439632,
      "step": 5125
    },
    {
      "epoch": 2.418670438472419,
      "grad_norm": 1.170190691947937,
      "learning_rate": 4.799887429373404e-05,
      "loss": 0.1444,
      "num_input_tokens_seen": 4444176,
      "step": 5130
    },
    {
      "epoch": 2.421027817067421,
      "grad_norm": 0.5127860307693481,
      "learning_rate": 4.79950238334372e-05,
      "loss": 0.3661,
      "num_input_tokens_seen": 4448640,
      "step": 5135
    },
    {
      "epoch": 2.4233851956624233,
      "grad_norm": 1.7185500860214233,
      "learning_rate": 4.799116982701872e-05,
      "loss": 0.2286,
      "num_input_tokens_seen": 4453456,
      "step": 5140
    },
    {
      "epoch": 2.4257425742574257,
      "grad_norm": 0.8512595295906067,
      "learning_rate": 4.7987312275072926e-05,
      "loss": 0.2039,
      "num_input_tokens_seen": 4456736,
      "step": 5145
    },
    {
      "epoch": 2.428099952852428,
      "grad_norm": 0.9830265045166016,
      "learning_rate": 4.79834511781947e-05,
      "loss": 0.2039,
      "num_input_tokens_seen": 4462368,
      "step": 5150
    },
    {
      "epoch": 2.4304573314474305,
      "grad_norm": 0.500109076499939,
      "learning_rate": 4.797958653697947e-05,
      "loss": 0.1679,
      "num_input_tokens_seen": 4468560,
      "step": 5155
    },
    {
      "epoch": 2.432814710042433,
      "grad_norm": 0.7842316031455994,
      "learning_rate": 4.7975718352023225e-05,
      "loss": 0.1306,
      "num_input_tokens_seen": 4473184,
      "step": 5160
    },
    {
      "epoch": 2.4351720886374353,
      "grad_norm": 1.1558912992477417,
      "learning_rate": 4.7971846623922476e-05,
      "loss": 0.3387,
      "num_input_tokens_seen": 4477344,
      "step": 5165
    },
    {
      "epoch": 2.4375294672324377,
      "grad_norm": 0.7586527466773987,
      "learning_rate": 4.7967971353274294e-05,
      "loss": 0.1748,
      "num_input_tokens_seen": 4481904,
      "step": 5170
    },
    {
      "epoch": 2.4398868458274396,
      "grad_norm": 0.4469292461872101,
      "learning_rate": 4.79640925406763e-05,
      "loss": 0.2192,
      "num_input_tokens_seen": 4485648,
      "step": 5175
    },
    {
      "epoch": 2.442244224422442,
      "grad_norm": 1.1267647743225098,
      "learning_rate": 4.796021018672664e-05,
      "loss": 0.1698,
      "num_input_tokens_seen": 4489376,
      "step": 5180
    },
    {
      "epoch": 2.4446016030174444,
      "grad_norm": 1.332962989807129,
      "learning_rate": 4.795632429202405e-05,
      "loss": 0.1979,
      "num_input_tokens_seen": 4493872,
      "step": 5185
    },
    {
      "epoch": 2.446958981612447,
      "grad_norm": 0.3723798394203186,
      "learning_rate": 4.795243485716775e-05,
      "loss": 0.1852,
      "num_input_tokens_seen": 4498016,
      "step": 5190
    },
    {
      "epoch": 2.4493163602074493,
      "grad_norm": 1.2830655574798584,
      "learning_rate": 4.794854188275757e-05,
      "loss": 0.2354,
      "num_input_tokens_seen": 4503504,
      "step": 5195
    },
    {
      "epoch": 2.4516737388024517,
      "grad_norm": 1.8345791101455688,
      "learning_rate": 4.794464536939384e-05,
      "loss": 0.2212,
      "num_input_tokens_seen": 4507760,
      "step": 5200
    },
    {
      "epoch": 2.4516737388024517,
      "eval_loss": 0.21647019684314728,
      "eval_runtime": 22.1699,
      "eval_samples_per_second": 42.535,
      "eval_steps_per_second": 21.29,
      "num_input_tokens_seen": 4507760,
      "step": 5200
    },
    {
      "epoch": 2.454031117397454,
      "grad_norm": 0.8449183106422424,
      "learning_rate": 4.794074531767745e-05,
      "loss": 0.1887,
      "num_input_tokens_seen": 4512336,
      "step": 5205
    },
    {
      "epoch": 2.4563884959924565,
      "grad_norm": 0.9606835842132568,
      "learning_rate": 4.7936841728209834e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 4515888,
      "step": 5210
    },
    {
      "epoch": 2.458745874587459,
      "grad_norm": 0.9217950701713562,
      "learning_rate": 4.7932934601593e-05,
      "loss": 0.1512,
      "num_input_tokens_seen": 4520320,
      "step": 5215
    },
    {
      "epoch": 2.4611032531824613,
      "grad_norm": 0.5521701574325562,
      "learning_rate": 4.792902393842943e-05,
      "loss": 0.1969,
      "num_input_tokens_seen": 4524544,
      "step": 5220
    },
    {
      "epoch": 2.4634606317774637,
      "grad_norm": 0.8759423494338989,
      "learning_rate": 4.792510973932225e-05,
      "loss": 0.1791,
      "num_input_tokens_seen": 4528464,
      "step": 5225
    },
    {
      "epoch": 2.465818010372466,
      "grad_norm": 1.247436761856079,
      "learning_rate": 4.7921192004875036e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 4532512,
      "step": 5230
    },
    {
      "epoch": 2.468175388967468,
      "grad_norm": 1.082160472869873,
      "learning_rate": 4.791727073569198e-05,
      "loss": 0.274,
      "num_input_tokens_seen": 4536992,
      "step": 5235
    },
    {
      "epoch": 2.4705327675624704,
      "grad_norm": 1.9717564582824707,
      "learning_rate": 4.7913345932377775e-05,
      "loss": 0.2264,
      "num_input_tokens_seen": 4542336,
      "step": 5240
    },
    {
      "epoch": 2.472890146157473,
      "grad_norm": 1.6691335439682007,
      "learning_rate": 4.790941759553769e-05,
      "loss": 0.2029,
      "num_input_tokens_seen": 4547808,
      "step": 5245
    },
    {
      "epoch": 2.4752475247524752,
      "grad_norm": 0.6202927231788635,
      "learning_rate": 4.79054857257775e-05,
      "loss": 0.1743,
      "num_input_tokens_seen": 4552768,
      "step": 5250
    },
    {
      "epoch": 2.4776049033474776,
      "grad_norm": 0.9641494750976562,
      "learning_rate": 4.790155032370357e-05,
      "loss": 0.2236,
      "num_input_tokens_seen": 4557072,
      "step": 5255
    },
    {
      "epoch": 2.47996228194248,
      "grad_norm": 1.7311524152755737,
      "learning_rate": 4.789761138992278e-05,
      "loss": 0.2532,
      "num_input_tokens_seen": 4561488,
      "step": 5260
    },
    {
      "epoch": 2.4823196605374824,
      "grad_norm": 0.9307465553283691,
      "learning_rate": 4.7893668925042565e-05,
      "loss": 0.2444,
      "num_input_tokens_seen": 4565952,
      "step": 5265
    },
    {
      "epoch": 2.484677039132485,
      "grad_norm": 1.5544403791427612,
      "learning_rate": 4.78897229296709e-05,
      "loss": 0.2465,
      "num_input_tokens_seen": 4570944,
      "step": 5270
    },
    {
      "epoch": 2.487034417727487,
      "grad_norm": 0.7613455057144165,
      "learning_rate": 4.7885773404416315e-05,
      "loss": 0.1881,
      "num_input_tokens_seen": 4575936,
      "step": 5275
    },
    {
      "epoch": 2.489391796322489,
      "grad_norm": 0.8036441802978516,
      "learning_rate": 4.788182034988786e-05,
      "loss": 0.2127,
      "num_input_tokens_seen": 4580560,
      "step": 5280
    },
    {
      "epoch": 2.4917491749174916,
      "grad_norm": 0.48713016510009766,
      "learning_rate": 4.787786376669516e-05,
      "loss": 0.1694,
      "num_input_tokens_seen": 4584800,
      "step": 5285
    },
    {
      "epoch": 2.494106553512494,
      "grad_norm": 1.9142760038375854,
      "learning_rate": 4.787390365544837e-05,
      "loss": 0.2506,
      "num_input_tokens_seen": 4590368,
      "step": 5290
    },
    {
      "epoch": 2.4964639321074964,
      "grad_norm": 1.1564679145812988,
      "learning_rate": 4.786994001675818e-05,
      "loss": 0.1788,
      "num_input_tokens_seen": 4594480,
      "step": 5295
    },
    {
      "epoch": 2.498821310702499,
      "grad_norm": 1.1404708623886108,
      "learning_rate": 4.786597285123584e-05,
      "loss": 0.1836,
      "num_input_tokens_seen": 4598880,
      "step": 5300
    },
    {
      "epoch": 2.501178689297501,
      "grad_norm": 0.5961135625839233,
      "learning_rate": 4.7862002159493135e-05,
      "loss": 0.1887,
      "num_input_tokens_seen": 4602960,
      "step": 5305
    },
    {
      "epoch": 2.5035360678925036,
      "grad_norm": 0.848267138004303,
      "learning_rate": 4.785802794214239e-05,
      "loss": 0.1916,
      "num_input_tokens_seen": 4607296,
      "step": 5310
    },
    {
      "epoch": 2.505893446487506,
      "grad_norm": 0.7229263782501221,
      "learning_rate": 4.7854050199796495e-05,
      "loss": 0.2588,
      "num_input_tokens_seen": 4611872,
      "step": 5315
    },
    {
      "epoch": 2.5082508250825084,
      "grad_norm": 1.3625962734222412,
      "learning_rate": 4.7850068933068845e-05,
      "loss": 0.2426,
      "num_input_tokens_seen": 4615568,
      "step": 5320
    },
    {
      "epoch": 2.510608203677511,
      "grad_norm": 1.0637716054916382,
      "learning_rate": 4.7846084142573425e-05,
      "loss": 0.1365,
      "num_input_tokens_seen": 4619520,
      "step": 5325
    },
    {
      "epoch": 2.512965582272513,
      "grad_norm": 0.48227882385253906,
      "learning_rate": 4.7842095828924725e-05,
      "loss": 0.1925,
      "num_input_tokens_seen": 4623232,
      "step": 5330
    },
    {
      "epoch": 2.515322960867515,
      "grad_norm": 1.239507794380188,
      "learning_rate": 4.783810399273779e-05,
      "loss": 0.2615,
      "num_input_tokens_seen": 4627312,
      "step": 5335
    },
    {
      "epoch": 2.5176803394625176,
      "grad_norm": 0.3303660452365875,
      "learning_rate": 4.7834108634628226e-05,
      "loss": 0.1206,
      "num_input_tokens_seen": 4631552,
      "step": 5340
    },
    {
      "epoch": 2.52003771805752,
      "grad_norm": 1.1052513122558594,
      "learning_rate": 4.783010975521216e-05,
      "loss": 0.2147,
      "num_input_tokens_seen": 4635568,
      "step": 5345
    },
    {
      "epoch": 2.5223950966525224,
      "grad_norm": 1.1505780220031738,
      "learning_rate": 4.782610735510626e-05,
      "loss": 0.2159,
      "num_input_tokens_seen": 4639968,
      "step": 5350
    },
    {
      "epoch": 2.5247524752475248,
      "grad_norm": 1.6446223258972168,
      "learning_rate": 4.782210143492776e-05,
      "loss": 0.2278,
      "num_input_tokens_seen": 4643536,
      "step": 5355
    },
    {
      "epoch": 2.527109853842527,
      "grad_norm": 1.8511713743209839,
      "learning_rate": 4.781809199529442e-05,
      "loss": 0.283,
      "num_input_tokens_seen": 4648256,
      "step": 5360
    },
    {
      "epoch": 2.5294672324375296,
      "grad_norm": 1.7124625444412231,
      "learning_rate": 4.781407903682454e-05,
      "loss": 0.2069,
      "num_input_tokens_seen": 4651712,
      "step": 5365
    },
    {
      "epoch": 2.531824611032532,
      "grad_norm": 0.5598857402801514,
      "learning_rate": 4.781006256013698e-05,
      "loss": 0.1915,
      "num_input_tokens_seen": 4656128,
      "step": 5370
    },
    {
      "epoch": 2.534181989627534,
      "grad_norm": 1.0287702083587646,
      "learning_rate": 4.7806042565851115e-05,
      "loss": 0.2088,
      "num_input_tokens_seen": 4660080,
      "step": 5375
    },
    {
      "epoch": 2.5365393682225363,
      "grad_norm": 1.183178186416626,
      "learning_rate": 4.7802019054586895e-05,
      "loss": 0.2287,
      "num_input_tokens_seen": 4664800,
      "step": 5380
    },
    {
      "epoch": 2.5388967468175387,
      "grad_norm": 0.5070058703422546,
      "learning_rate": 4.779799202696479e-05,
      "loss": 0.1815,
      "num_input_tokens_seen": 4669040,
      "step": 5385
    },
    {
      "epoch": 2.541254125412541,
      "grad_norm": 1.1563645601272583,
      "learning_rate": 4.779396148360581e-05,
      "loss": 0.2649,
      "num_input_tokens_seen": 4673712,
      "step": 5390
    },
    {
      "epoch": 2.5436115040075435,
      "grad_norm": 0.4372324049472809,
      "learning_rate": 4.7789927425131517e-05,
      "loss": 0.2036,
      "num_input_tokens_seen": 4677344,
      "step": 5395
    },
    {
      "epoch": 2.545968882602546,
      "grad_norm": 0.7512884140014648,
      "learning_rate": 4.778588985216403e-05,
      "loss": 0.1806,
      "num_input_tokens_seen": 4681664,
      "step": 5400
    },
    {
      "epoch": 2.545968882602546,
      "eval_loss": 0.21467305719852448,
      "eval_runtime": 22.1536,
      "eval_samples_per_second": 42.566,
      "eval_steps_per_second": 21.306,
      "num_input_tokens_seen": 4681664,
      "step": 5400
    },
    {
      "epoch": 2.5483262611975483,
      "grad_norm": 0.7342270016670227,
      "learning_rate": 4.778184876532598e-05,
      "loss": 0.1757,
      "num_input_tokens_seen": 4685776,
      "step": 5405
    },
    {
      "epoch": 2.5506836397925507,
      "grad_norm": 1.1888705492019653,
      "learning_rate": 4.7777804165240556e-05,
      "loss": 0.148,
      "num_input_tokens_seen": 4690416,
      "step": 5410
    },
    {
      "epoch": 2.553041018387553,
      "grad_norm": 1.3159059286117554,
      "learning_rate": 4.7773756052531485e-05,
      "loss": 0.1479,
      "num_input_tokens_seen": 4695328,
      "step": 5415
    },
    {
      "epoch": 2.5553983969825556,
      "grad_norm": 0.7820848822593689,
      "learning_rate": 4.7769704427823035e-05,
      "loss": 0.1385,
      "num_input_tokens_seen": 4698960,
      "step": 5420
    },
    {
      "epoch": 2.557755775577558,
      "grad_norm": 0.6508060097694397,
      "learning_rate": 4.776564929174003e-05,
      "loss": 0.1992,
      "num_input_tokens_seen": 4703024,
      "step": 5425
    },
    {
      "epoch": 2.5601131541725604,
      "grad_norm": 0.7073059678077698,
      "learning_rate": 4.7761590644907806e-05,
      "loss": 0.1864,
      "num_input_tokens_seen": 4707040,
      "step": 5430
    },
    {
      "epoch": 2.5624705327675623,
      "grad_norm": 0.8110402822494507,
      "learning_rate": 4.7757528487952263e-05,
      "loss": 0.2025,
      "num_input_tokens_seen": 4710896,
      "step": 5435
    },
    {
      "epoch": 2.5648279113625647,
      "grad_norm": 0.9862064123153687,
      "learning_rate": 4.7753462821499836e-05,
      "loss": 0.1337,
      "num_input_tokens_seen": 4715456,
      "step": 5440
    },
    {
      "epoch": 2.567185289957567,
      "grad_norm": 0.40715476870536804,
      "learning_rate": 4.774939364617751e-05,
      "loss": 0.2614,
      "num_input_tokens_seen": 4720128,
      "step": 5445
    },
    {
      "epoch": 2.5695426685525695,
      "grad_norm": 1.0804418325424194,
      "learning_rate": 4.7745320962612795e-05,
      "loss": 0.1453,
      "num_input_tokens_seen": 4725024,
      "step": 5450
    },
    {
      "epoch": 2.571900047147572,
      "grad_norm": 0.5931711792945862,
      "learning_rate": 4.7741244771433756e-05,
      "loss": 0.2204,
      "num_input_tokens_seen": 4730592,
      "step": 5455
    },
    {
      "epoch": 2.5742574257425743,
      "grad_norm": 0.9243809580802917,
      "learning_rate": 4.7737165073268985e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 4734640,
      "step": 5460
    },
    {
      "epoch": 2.5766148043375767,
      "grad_norm": 1.3692289590835571,
      "learning_rate": 4.7733081868747626e-05,
      "loss": 0.1734,
      "num_input_tokens_seen": 4738928,
      "step": 5465
    },
    {
      "epoch": 2.578972182932579,
      "grad_norm": 1.1478575468063354,
      "learning_rate": 4.772899515849936e-05,
      "loss": 0.1995,
      "num_input_tokens_seen": 4743520,
      "step": 5470
    },
    {
      "epoch": 2.581329561527581,
      "grad_norm": 0.49210137128829956,
      "learning_rate": 4.7724904943154414e-05,
      "loss": 0.2149,
      "num_input_tokens_seen": 4746976,
      "step": 5475
    },
    {
      "epoch": 2.5836869401225835,
      "grad_norm": 0.9169567227363586,
      "learning_rate": 4.772081122334354e-05,
      "loss": 0.2038,
      "num_input_tokens_seen": 4751152,
      "step": 5480
    },
    {
      "epoch": 2.586044318717586,
      "grad_norm": 0.9211050868034363,
      "learning_rate": 4.771671399969806e-05,
      "loss": 0.2255,
      "num_input_tokens_seen": 4756624,
      "step": 5485
    },
    {
      "epoch": 2.5884016973125883,
      "grad_norm": 0.9821653962135315,
      "learning_rate": 4.7712613272849794e-05,
      "loss": 0.1265,
      "num_input_tokens_seen": 4760416,
      "step": 5490
    },
    {
      "epoch": 2.5907590759075907,
      "grad_norm": 1.0790847539901733,
      "learning_rate": 4.770850904343114e-05,
      "loss": 0.1648,
      "num_input_tokens_seen": 4765360,
      "step": 5495
    },
    {
      "epoch": 2.593116454502593,
      "grad_norm": 0.628298282623291,
      "learning_rate": 4.770440131207502e-05,
      "loss": 0.1889,
      "num_input_tokens_seen": 4769264,
      "step": 5500
    },
    {
      "epoch": 2.5954738330975955,
      "grad_norm": 1.4077736139297485,
      "learning_rate": 4.7700290079414896e-05,
      "loss": 0.2406,
      "num_input_tokens_seen": 4774064,
      "step": 5505
    },
    {
      "epoch": 2.597831211692598,
      "grad_norm": 0.5564115047454834,
      "learning_rate": 4.769617534608477e-05,
      "loss": 0.1871,
      "num_input_tokens_seen": 4777968,
      "step": 5510
    },
    {
      "epoch": 2.6001885902876003,
      "grad_norm": 0.7230802178382874,
      "learning_rate": 4.7692057112719193e-05,
      "loss": 0.1767,
      "num_input_tokens_seen": 4781952,
      "step": 5515
    },
    {
      "epoch": 2.6025459688826027,
      "grad_norm": 0.9353978037834167,
      "learning_rate": 4.7687935379953234e-05,
      "loss": 0.2714,
      "num_input_tokens_seen": 4786448,
      "step": 5520
    },
    {
      "epoch": 2.604903347477605,
      "grad_norm": 0.5753369927406311,
      "learning_rate": 4.7683810148422534e-05,
      "loss": 0.179,
      "num_input_tokens_seen": 4790512,
      "step": 5525
    },
    {
      "epoch": 2.6072607260726075,
      "grad_norm": 0.8246042728424072,
      "learning_rate": 4.767968141876324e-05,
      "loss": 0.2254,
      "num_input_tokens_seen": 4794832,
      "step": 5530
    },
    {
      "epoch": 2.6096181046676095,
      "grad_norm": 0.697572648525238,
      "learning_rate": 4.767554919161207e-05,
      "loss": 0.1986,
      "num_input_tokens_seen": 4798688,
      "step": 5535
    },
    {
      "epoch": 2.611975483262612,
      "grad_norm": 1.1378977298736572,
      "learning_rate": 4.767141346760624e-05,
      "loss": 0.1955,
      "num_input_tokens_seen": 4802880,
      "step": 5540
    },
    {
      "epoch": 2.6143328618576143,
      "grad_norm": 1.285915493965149,
      "learning_rate": 4.766727424738356e-05,
      "loss": 0.2498,
      "num_input_tokens_seen": 4807776,
      "step": 5545
    },
    {
      "epoch": 2.6166902404526167,
      "grad_norm": 1.1687713861465454,
      "learning_rate": 4.7663131531582325e-05,
      "loss": 0.1849,
      "num_input_tokens_seen": 4812320,
      "step": 5550
    },
    {
      "epoch": 2.619047619047619,
      "grad_norm": 0.8772350549697876,
      "learning_rate": 4.765898532084142e-05,
      "loss": 0.1684,
      "num_input_tokens_seen": 4816464,
      "step": 5555
    },
    {
      "epoch": 2.6214049976426215,
      "grad_norm": 0.631001353263855,
      "learning_rate": 4.765483561580022e-05,
      "loss": 0.2101,
      "num_input_tokens_seen": 4821168,
      "step": 5560
    },
    {
      "epoch": 2.623762376237624,
      "grad_norm": 0.9778459668159485,
      "learning_rate": 4.7650682417098666e-05,
      "loss": 0.1348,
      "num_input_tokens_seen": 4826176,
      "step": 5565
    },
    {
      "epoch": 2.6261197548326263,
      "grad_norm": 0.9118787050247192,
      "learning_rate": 4.7646525725377244e-05,
      "loss": 0.1673,
      "num_input_tokens_seen": 4829856,
      "step": 5570
    },
    {
      "epoch": 2.6284771334276282,
      "grad_norm": 1.234635829925537,
      "learning_rate": 4.764236554127696e-05,
      "loss": 0.2349,
      "num_input_tokens_seen": 4834288,
      "step": 5575
    },
    {
      "epoch": 2.6308345120226306,
      "grad_norm": 0.6128241419792175,
      "learning_rate": 4.7638201865439356e-05,
      "loss": 0.1251,
      "num_input_tokens_seen": 4838704,
      "step": 5580
    },
    {
      "epoch": 2.633191890617633,
      "grad_norm": 0.6883321404457092,
      "learning_rate": 4.7634034698506545e-05,
      "loss": 0.1694,
      "num_input_tokens_seen": 4843008,
      "step": 5585
    },
    {
      "epoch": 2.6355492692126354,
      "grad_norm": 1.178260087966919,
      "learning_rate": 4.762986404112115e-05,
      "loss": 0.2873,
      "num_input_tokens_seen": 4847760,
      "step": 5590
    },
    {
      "epoch": 2.637906647807638,
      "grad_norm": 0.6994194984436035,
      "learning_rate": 4.762568989392633e-05,
      "loss": 0.2016,
      "num_input_tokens_seen": 4852352,
      "step": 5595
    },
    {
      "epoch": 2.6402640264026402,
      "grad_norm": 0.7232382297515869,
      "learning_rate": 4.76215122575658e-05,
      "loss": 0.2273,
      "num_input_tokens_seen": 4856928,
      "step": 5600
    },
    {
      "epoch": 2.6402640264026402,
      "eval_loss": 0.21606530249118805,
      "eval_runtime": 22.1522,
      "eval_samples_per_second": 42.569,
      "eval_steps_per_second": 21.307,
      "num_input_tokens_seen": 4856928,
      "step": 5600
    },
    {
      "epoch": 2.6426214049976426,
      "grad_norm": 0.578803539276123,
      "learning_rate": 4.7617331132683795e-05,
      "loss": 0.1576,
      "num_input_tokens_seen": 4861408,
      "step": 5605
    },
    {
      "epoch": 2.644978783592645,
      "grad_norm": 0.9494000673294067,
      "learning_rate": 4.7613146519925105e-05,
      "loss": 0.2063,
      "num_input_tokens_seen": 4865600,
      "step": 5610
    },
    {
      "epoch": 2.6473361621876474,
      "grad_norm": 1.1658477783203125,
      "learning_rate": 4.7608958419935045e-05,
      "loss": 0.2025,
      "num_input_tokens_seen": 4869744,
      "step": 5615
    },
    {
      "epoch": 2.64969354078265,
      "grad_norm": 0.9670729041099548,
      "learning_rate": 4.760476683335948e-05,
      "loss": 0.206,
      "num_input_tokens_seen": 4874448,
      "step": 5620
    },
    {
      "epoch": 2.6520509193776522,
      "grad_norm": 0.5327708721160889,
      "learning_rate": 4.760057176084479e-05,
      "loss": 0.2441,
      "num_input_tokens_seen": 4878128,
      "step": 5625
    },
    {
      "epoch": 2.6544082979726547,
      "grad_norm": 0.423520565032959,
      "learning_rate": 4.759637320303793e-05,
      "loss": 0.1891,
      "num_input_tokens_seen": 4882048,
      "step": 5630
    },
    {
      "epoch": 2.6567656765676566,
      "grad_norm": 1.1646618843078613,
      "learning_rate": 4.759217116058635e-05,
      "loss": 0.1442,
      "num_input_tokens_seen": 4886432,
      "step": 5635
    },
    {
      "epoch": 2.659123055162659,
      "grad_norm": 0.9949005246162415,
      "learning_rate": 4.758796563413807e-05,
      "loss": 0.1473,
      "num_input_tokens_seen": 4891008,
      "step": 5640
    },
    {
      "epoch": 2.6614804337576614,
      "grad_norm": 0.8142032623291016,
      "learning_rate": 4.758375662434163e-05,
      "loss": 0.1824,
      "num_input_tokens_seen": 4894960,
      "step": 5645
    },
    {
      "epoch": 2.663837812352664,
      "grad_norm": 0.5975801944732666,
      "learning_rate": 4.7579544131846114e-05,
      "loss": 0.2167,
      "num_input_tokens_seen": 4899952,
      "step": 5650
    },
    {
      "epoch": 2.666195190947666,
      "grad_norm": 1.378229022026062,
      "learning_rate": 4.757532815730114e-05,
      "loss": 0.3221,
      "num_input_tokens_seen": 4903712,
      "step": 5655
    },
    {
      "epoch": 2.6685525695426686,
      "grad_norm": 1.1303865909576416,
      "learning_rate": 4.7571108701356865e-05,
      "loss": 0.1184,
      "num_input_tokens_seen": 4907488,
      "step": 5660
    },
    {
      "epoch": 2.670909948137671,
      "grad_norm": 2.4807846546173096,
      "learning_rate": 4.756688576466398e-05,
      "loss": 0.2039,
      "num_input_tokens_seen": 4911648,
      "step": 5665
    },
    {
      "epoch": 2.6732673267326734,
      "grad_norm": 0.6327922344207764,
      "learning_rate": 4.756265934787372e-05,
      "loss": 0.1859,
      "num_input_tokens_seen": 4916560,
      "step": 5670
    },
    {
      "epoch": 2.6756247053276754,
      "grad_norm": 1.1640691757202148,
      "learning_rate": 4.755842945163785e-05,
      "loss": 0.2068,
      "num_input_tokens_seen": 4920608,
      "step": 5675
    },
    {
      "epoch": 2.677982083922678,
      "grad_norm": 1.1764832735061646,
      "learning_rate": 4.755419607660867e-05,
      "loss": 0.2187,
      "num_input_tokens_seen": 4924672,
      "step": 5680
    },
    {
      "epoch": 2.68033946251768,
      "grad_norm": 0.8170294165611267,
      "learning_rate": 4.7549959223439016e-05,
      "loss": 0.1413,
      "num_input_tokens_seen": 4928816,
      "step": 5685
    },
    {
      "epoch": 2.6826968411126826,
      "grad_norm": 0.828630268573761,
      "learning_rate": 4.754571889278228e-05,
      "loss": 0.1694,
      "num_input_tokens_seen": 4933728,
      "step": 5690
    },
    {
      "epoch": 2.685054219707685,
      "grad_norm": 0.8392783999443054,
      "learning_rate": 4.754147508529235e-05,
      "loss": 0.1907,
      "num_input_tokens_seen": 4937184,
      "step": 5695
    },
    {
      "epoch": 2.6874115983026874,
      "grad_norm": 1.0740350484848022,
      "learning_rate": 4.75372278016237e-05,
      "loss": 0.2463,
      "num_input_tokens_seen": 4942016,
      "step": 5700
    },
    {
      "epoch": 2.68976897689769,
      "grad_norm": 0.5248028635978699,
      "learning_rate": 4.753297704243129e-05,
      "loss": 0.1344,
      "num_input_tokens_seen": 4945632,
      "step": 5705
    },
    {
      "epoch": 2.692126355492692,
      "grad_norm": 1.4073601961135864,
      "learning_rate": 4.752872280837066e-05,
      "loss": 0.1605,
      "num_input_tokens_seen": 4949408,
      "step": 5710
    },
    {
      "epoch": 2.6944837340876946,
      "grad_norm": 1.1858537197113037,
      "learning_rate": 4.752446510009786e-05,
      "loss": 0.2926,
      "num_input_tokens_seen": 4953616,
      "step": 5715
    },
    {
      "epoch": 2.696841112682697,
      "grad_norm": 1.1754405498504639,
      "learning_rate": 4.7520203918269476e-05,
      "loss": 0.1974,
      "num_input_tokens_seen": 4958304,
      "step": 5720
    },
    {
      "epoch": 2.6991984912776994,
      "grad_norm": 1.6892673969268799,
      "learning_rate": 4.751593926354265e-05,
      "loss": 0.219,
      "num_input_tokens_seen": 4962576,
      "step": 5725
    },
    {
      "epoch": 2.701555869872702,
      "grad_norm": 0.7460927963256836,
      "learning_rate": 4.751167113657503e-05,
      "loss": 0.2491,
      "num_input_tokens_seen": 4966864,
      "step": 5730
    },
    {
      "epoch": 2.7039132484677038,
      "grad_norm": 1.1830999851226807,
      "learning_rate": 4.7507399538024834e-05,
      "loss": 0.2418,
      "num_input_tokens_seen": 4971024,
      "step": 5735
    },
    {
      "epoch": 2.706270627062706,
      "grad_norm": 1.246556043624878,
      "learning_rate": 4.750312446855077e-05,
      "loss": 0.2271,
      "num_input_tokens_seen": 4975392,
      "step": 5740
    },
    {
      "epoch": 2.7086280056577086,
      "grad_norm": 0.912562906742096,
      "learning_rate": 4.749884592881212e-05,
      "loss": 0.2372,
      "num_input_tokens_seen": 4979056,
      "step": 5745
    },
    {
      "epoch": 2.710985384252711,
      "grad_norm": 0.6695653796195984,
      "learning_rate": 4.74945639194687e-05,
      "loss": 0.1958,
      "num_input_tokens_seen": 4983296,
      "step": 5750
    },
    {
      "epoch": 2.7133427628477134,
      "grad_norm": 0.6756101846694946,
      "learning_rate": 4.749027844118083e-05,
      "loss": 0.1524,
      "num_input_tokens_seen": 4987456,
      "step": 5755
    },
    {
      "epoch": 2.7157001414427158,
      "grad_norm": 0.7881581783294678,
      "learning_rate": 4.7485989494609395e-05,
      "loss": 0.1888,
      "num_input_tokens_seen": 4991200,
      "step": 5760
    },
    {
      "epoch": 2.718057520037718,
      "grad_norm": 0.8113067150115967,
      "learning_rate": 4.748169708041581e-05,
      "loss": 0.146,
      "num_input_tokens_seen": 4995008,
      "step": 5765
    },
    {
      "epoch": 2.7204148986327206,
      "grad_norm": 1.2879751920700073,
      "learning_rate": 4.7477401199262004e-05,
      "loss": 0.2031,
      "num_input_tokens_seen": 4999056,
      "step": 5770
    },
    {
      "epoch": 2.7227722772277225,
      "grad_norm": 1.191232681274414,
      "learning_rate": 4.747310185181048e-05,
      "loss": 0.2289,
      "num_input_tokens_seen": 5003616,
      "step": 5775
    },
    {
      "epoch": 2.725129655822725,
      "grad_norm": 0.9092208743095398,
      "learning_rate": 4.746879903872422e-05,
      "loss": 0.2754,
      "num_input_tokens_seen": 5008816,
      "step": 5780
    },
    {
      "epoch": 2.7274870344177273,
      "grad_norm": 1.1915440559387207,
      "learning_rate": 4.746449276066679e-05,
      "loss": 0.2091,
      "num_input_tokens_seen": 5013008,
      "step": 5785
    },
    {
      "epoch": 2.7298444130127297,
      "grad_norm": 0.6812341213226318,
      "learning_rate": 4.746018301830227e-05,
      "loss": 0.1717,
      "num_input_tokens_seen": 5017072,
      "step": 5790
    },
    {
      "epoch": 2.732201791607732,
      "grad_norm": 1.2620118856430054,
      "learning_rate": 4.7455869812295275e-05,
      "loss": 0.1811,
      "num_input_tokens_seen": 5021024,
      "step": 5795
    },
    {
      "epoch": 2.7345591702027345,
      "grad_norm": 1.6013376712799072,
      "learning_rate": 4.7451553143310964e-05,
      "loss": 0.2041,
      "num_input_tokens_seen": 5024976,
      "step": 5800
    },
    {
      "epoch": 2.7345591702027345,
      "eval_loss": 0.2148492932319641,
      "eval_runtime": 22.1325,
      "eval_samples_per_second": 42.607,
      "eval_steps_per_second": 21.326,
      "num_input_tokens_seen": 5024976,
      "step": 5800
    },
    {
      "epoch": 2.736916548797737,
      "grad_norm": 1.23733389377594,
      "learning_rate": 4.744723301201501e-05,
      "loss": 0.2173,
      "num_input_tokens_seen": 5030048,
      "step": 5805
    },
    {
      "epoch": 2.7392739273927393,
      "grad_norm": 0.7346053719520569,
      "learning_rate": 4.744290941907364e-05,
      "loss": 0.2144,
      "num_input_tokens_seen": 5034192,
      "step": 5810
    },
    {
      "epoch": 2.7416313059877417,
      "grad_norm": 0.5920982956886292,
      "learning_rate": 4.7438582365153594e-05,
      "loss": 0.1666,
      "num_input_tokens_seen": 5037888,
      "step": 5815
    },
    {
      "epoch": 2.743988684582744,
      "grad_norm": 1.1275402307510376,
      "learning_rate": 4.743425185092217e-05,
      "loss": 0.2226,
      "num_input_tokens_seen": 5042496,
      "step": 5820
    },
    {
      "epoch": 2.7463460631777465,
      "grad_norm": 1.3347396850585938,
      "learning_rate": 4.742991787704719e-05,
      "loss": 0.1851,
      "num_input_tokens_seen": 5047424,
      "step": 5825
    },
    {
      "epoch": 2.748703441772749,
      "grad_norm": 0.6115074753761292,
      "learning_rate": 4.7425580444196994e-05,
      "loss": 0.1213,
      "num_input_tokens_seen": 5051536,
      "step": 5830
    },
    {
      "epoch": 2.751060820367751,
      "grad_norm": 1.707585334777832,
      "learning_rate": 4.742123955304048e-05,
      "loss": 0.1975,
      "num_input_tokens_seen": 5055648,
      "step": 5835
    },
    {
      "epoch": 2.7534181989627533,
      "grad_norm": 0.6722565293312073,
      "learning_rate": 4.741689520424706e-05,
      "loss": 0.1487,
      "num_input_tokens_seen": 5060176,
      "step": 5840
    },
    {
      "epoch": 2.7557755775577557,
      "grad_norm": 0.8675293326377869,
      "learning_rate": 4.741254739848669e-05,
      "loss": 0.2491,
      "num_input_tokens_seen": 5065312,
      "step": 5845
    },
    {
      "epoch": 2.758132956152758,
      "grad_norm": 0.7638822197914124,
      "learning_rate": 4.740819613642987e-05,
      "loss": 0.2202,
      "num_input_tokens_seen": 5069712,
      "step": 5850
    },
    {
      "epoch": 2.7604903347477605,
      "grad_norm": 1.0857590436935425,
      "learning_rate": 4.74038414187476e-05,
      "loss": 0.2579,
      "num_input_tokens_seen": 5073936,
      "step": 5855
    },
    {
      "epoch": 2.762847713342763,
      "grad_norm": 1.5449930429458618,
      "learning_rate": 4.739948324611144e-05,
      "loss": 0.1964,
      "num_input_tokens_seen": 5078112,
      "step": 5860
    },
    {
      "epoch": 2.7652050919377653,
      "grad_norm": 1.6986759901046753,
      "learning_rate": 4.7395121619193465e-05,
      "loss": 0.1849,
      "num_input_tokens_seen": 5082000,
      "step": 5865
    },
    {
      "epoch": 2.7675624705327677,
      "grad_norm": 1.8591128587722778,
      "learning_rate": 4.7390756538666313e-05,
      "loss": 0.1727,
      "num_input_tokens_seen": 5086064,
      "step": 5870
    },
    {
      "epoch": 2.7699198491277697,
      "grad_norm": 0.7621943950653076,
      "learning_rate": 4.738638800520311e-05,
      "loss": 0.2246,
      "num_input_tokens_seen": 5090256,
      "step": 5875
    },
    {
      "epoch": 2.772277227722772,
      "grad_norm": 0.9351010322570801,
      "learning_rate": 4.738201601947757e-05,
      "loss": 0.23,
      "num_input_tokens_seen": 5094880,
      "step": 5880
    },
    {
      "epoch": 2.7746346063177745,
      "grad_norm": 0.8584136962890625,
      "learning_rate": 4.7377640582163876e-05,
      "loss": 0.2218,
      "num_input_tokens_seen": 5099136,
      "step": 5885
    },
    {
      "epoch": 2.776991984912777,
      "grad_norm": 1.2089670896530151,
      "learning_rate": 4.7373261693936786e-05,
      "loss": 0.1941,
      "num_input_tokens_seen": 5103152,
      "step": 5890
    },
    {
      "epoch": 2.7793493635077793,
      "grad_norm": 0.544792890548706,
      "learning_rate": 4.7368879355471595e-05,
      "loss": 0.2128,
      "num_input_tokens_seen": 5107776,
      "step": 5895
    },
    {
      "epoch": 2.7817067421027817,
      "grad_norm": 0.522911787033081,
      "learning_rate": 4.736449356744409e-05,
      "loss": 0.183,
      "num_input_tokens_seen": 5112480,
      "step": 5900
    },
    {
      "epoch": 2.784064120697784,
      "grad_norm": 1.5892670154571533,
      "learning_rate": 4.736010433053064e-05,
      "loss": 0.2095,
      "num_input_tokens_seen": 5116672,
      "step": 5905
    },
    {
      "epoch": 2.7864214992927865,
      "grad_norm": 1.3768303394317627,
      "learning_rate": 4.73557116454081e-05,
      "loss": 0.2167,
      "num_input_tokens_seen": 5121072,
      "step": 5910
    },
    {
      "epoch": 2.788778877887789,
      "grad_norm": 1.5236703157424927,
      "learning_rate": 4.735131551275389e-05,
      "loss": 0.2473,
      "num_input_tokens_seen": 5125568,
      "step": 5915
    },
    {
      "epoch": 2.7911362564827913,
      "grad_norm": 0.6119250059127808,
      "learning_rate": 4.734691593324594e-05,
      "loss": 0.2131,
      "num_input_tokens_seen": 5130560,
      "step": 5920
    },
    {
      "epoch": 2.7934936350777937,
      "grad_norm": 1.7895631790161133,
      "learning_rate": 4.734251290756272e-05,
      "loss": 0.2117,
      "num_input_tokens_seen": 5134976,
      "step": 5925
    },
    {
      "epoch": 2.795851013672796,
      "grad_norm": 0.5614873766899109,
      "learning_rate": 4.7338106436383246e-05,
      "loss": 0.1505,
      "num_input_tokens_seen": 5139072,
      "step": 5930
    },
    {
      "epoch": 2.798208392267798,
      "grad_norm": 0.8728413581848145,
      "learning_rate": 4.733369652038703e-05,
      "loss": 0.2543,
      "num_input_tokens_seen": 5143776,
      "step": 5935
    },
    {
      "epoch": 2.8005657708628005,
      "grad_norm": 1.081817626953125,
      "learning_rate": 4.7329283160254156e-05,
      "loss": 0.1991,
      "num_input_tokens_seen": 5148080,
      "step": 5940
    },
    {
      "epoch": 2.802923149457803,
      "grad_norm": 0.4236457645893097,
      "learning_rate": 4.732486635666521e-05,
      "loss": 0.1632,
      "num_input_tokens_seen": 5153120,
      "step": 5945
    },
    {
      "epoch": 2.8052805280528053,
      "grad_norm": 0.8143977522850037,
      "learning_rate": 4.732044611030132e-05,
      "loss": 0.1653,
      "num_input_tokens_seen": 5157648,
      "step": 5950
    },
    {
      "epoch": 2.8076379066478077,
      "grad_norm": 0.6420667767524719,
      "learning_rate": 4.731602242184414e-05,
      "loss": 0.1343,
      "num_input_tokens_seen": 5162448,
      "step": 5955
    },
    {
      "epoch": 2.80999528524281,
      "grad_norm": 2.305966377258301,
      "learning_rate": 4.7311595291975864e-05,
      "loss": 0.2238,
      "num_input_tokens_seen": 5167216,
      "step": 5960
    },
    {
      "epoch": 2.8123526638378125,
      "grad_norm": 0.7257430553436279,
      "learning_rate": 4.7307164721379216e-05,
      "loss": 0.161,
      "num_input_tokens_seen": 5171568,
      "step": 5965
    },
    {
      "epoch": 2.814710042432815,
      "grad_norm": 0.8762022852897644,
      "learning_rate": 4.730273071073743e-05,
      "loss": 0.1877,
      "num_input_tokens_seen": 5175888,
      "step": 5970
    },
    {
      "epoch": 2.817067421027817,
      "grad_norm": 1.956469178199768,
      "learning_rate": 4.729829326073429e-05,
      "loss": 0.1935,
      "num_input_tokens_seen": 5179856,
      "step": 5975
    },
    {
      "epoch": 2.8194247996228192,
      "grad_norm": 0.7557887434959412,
      "learning_rate": 4.7293852372054126e-05,
      "loss": 0.1676,
      "num_input_tokens_seen": 5184144,
      "step": 5980
    },
    {
      "epoch": 2.8217821782178216,
      "grad_norm": 1.0982574224472046,
      "learning_rate": 4.728940804538176e-05,
      "loss": 0.3035,
      "num_input_tokens_seen": 5189376,
      "step": 5985
    },
    {
      "epoch": 2.824139556812824,
      "grad_norm": 0.9428257942199707,
      "learning_rate": 4.7284960281402556e-05,
      "loss": 0.1188,
      "num_input_tokens_seen": 5193568,
      "step": 5990
    },
    {
      "epoch": 2.8264969354078264,
      "grad_norm": 1.7640178203582764,
      "learning_rate": 4.728050908080244e-05,
      "loss": 0.1839,
      "num_input_tokens_seen": 5198368,
      "step": 5995
    },
    {
      "epoch": 2.828854314002829,
      "grad_norm": 1.1390774250030518,
      "learning_rate": 4.727605444426782e-05,
      "loss": 0.2044,
      "num_input_tokens_seen": 5202368,
      "step": 6000
    },
    {
      "epoch": 2.828854314002829,
      "eval_loss": 0.21614356338977814,
      "eval_runtime": 22.1522,
      "eval_samples_per_second": 42.569,
      "eval_steps_per_second": 21.307,
      "num_input_tokens_seen": 5202368,
      "step": 6000
    },
    {
      "epoch": 2.8312116925978312,
      "grad_norm": 1.1916038990020752,
      "learning_rate": 4.727159637248567e-05,
      "loss": 0.1757,
      "num_input_tokens_seen": 5206368,
      "step": 6005
    },
    {
      "epoch": 2.8335690711928336,
      "grad_norm": 0.5948582291603088,
      "learning_rate": 4.7267134866143474e-05,
      "loss": 0.1302,
      "num_input_tokens_seen": 5210160,
      "step": 6010
    },
    {
      "epoch": 2.835926449787836,
      "grad_norm": 0.6187908053398132,
      "learning_rate": 4.726266992592926e-05,
      "loss": 0.1355,
      "num_input_tokens_seen": 5214704,
      "step": 6015
    },
    {
      "epoch": 2.8382838283828384,
      "grad_norm": 1.8855262994766235,
      "learning_rate": 4.725820155253157e-05,
      "loss": 0.2571,
      "num_input_tokens_seen": 5218896,
      "step": 6020
    },
    {
      "epoch": 2.840641206977841,
      "grad_norm": 0.5643291473388672,
      "learning_rate": 4.725372974663948e-05,
      "loss": 0.1316,
      "num_input_tokens_seen": 5223232,
      "step": 6025
    },
    {
      "epoch": 2.8429985855728432,
      "grad_norm": 0.795806884765625,
      "learning_rate": 4.724925450894262e-05,
      "loss": 0.24,
      "num_input_tokens_seen": 5227536,
      "step": 6030
    },
    {
      "epoch": 2.845355964167845,
      "grad_norm": 0.5750205516815186,
      "learning_rate": 4.72447758401311e-05,
      "loss": 0.1317,
      "num_input_tokens_seen": 5232512,
      "step": 6035
    },
    {
      "epoch": 2.8477133427628476,
      "grad_norm": 0.8301213979721069,
      "learning_rate": 4.7240293740895616e-05,
      "loss": 0.1877,
      "num_input_tokens_seen": 5237472,
      "step": 6040
    },
    {
      "epoch": 2.85007072135785,
      "grad_norm": 1.1219377517700195,
      "learning_rate": 4.723580821192733e-05,
      "loss": 0.2173,
      "num_input_tokens_seen": 5241904,
      "step": 6045
    },
    {
      "epoch": 2.8524280999528524,
      "grad_norm": 1.9350464344024658,
      "learning_rate": 4.7231319253917996e-05,
      "loss": 0.2229,
      "num_input_tokens_seen": 5245712,
      "step": 6050
    },
    {
      "epoch": 2.854785478547855,
      "grad_norm": 0.9870319366455078,
      "learning_rate": 4.722682686755986e-05,
      "loss": 0.2267,
      "num_input_tokens_seen": 5250704,
      "step": 6055
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.6597947478294373,
      "learning_rate": 4.722233105354569e-05,
      "loss": 0.213,
      "num_input_tokens_seen": 5255056,
      "step": 6060
    },
    {
      "epoch": 2.8595002357378596,
      "grad_norm": 2.541451930999756,
      "learning_rate": 4.7217831812568815e-05,
      "loss": 0.2158,
      "num_input_tokens_seen": 5259792,
      "step": 6065
    },
    {
      "epoch": 2.861857614332862,
      "grad_norm": 1.1620270013809204,
      "learning_rate": 4.721332914532307e-05,
      "loss": 0.196,
      "num_input_tokens_seen": 5263968,
      "step": 6070
    },
    {
      "epoch": 2.864214992927864,
      "grad_norm": 0.696197509765625,
      "learning_rate": 4.720882305250281e-05,
      "loss": 0.1987,
      "num_input_tokens_seen": 5268112,
      "step": 6075
    },
    {
      "epoch": 2.8665723715228664,
      "grad_norm": 1.3188515901565552,
      "learning_rate": 4.720431353480295e-05,
      "loss": 0.2154,
      "num_input_tokens_seen": 5272576,
      "step": 6080
    },
    {
      "epoch": 2.8689297501178688,
      "grad_norm": 1.6559395790100098,
      "learning_rate": 4.719980059291891e-05,
      "loss": 0.2791,
      "num_input_tokens_seen": 5276608,
      "step": 6085
    },
    {
      "epoch": 2.871287128712871,
      "grad_norm": 0.6863065361976624,
      "learning_rate": 4.7195284227546634e-05,
      "loss": 0.1471,
      "num_input_tokens_seen": 5281344,
      "step": 6090
    },
    {
      "epoch": 2.8736445073078736,
      "grad_norm": 0.7328031063079834,
      "learning_rate": 4.7190764439382604e-05,
      "loss": 0.1689,
      "num_input_tokens_seen": 5285088,
      "step": 6095
    },
    {
      "epoch": 2.876001885902876,
      "grad_norm": 1.5273709297180176,
      "learning_rate": 4.7186241229123826e-05,
      "loss": 0.1581,
      "num_input_tokens_seen": 5289648,
      "step": 6100
    },
    {
      "epoch": 2.8783592644978784,
      "grad_norm": 1.6989362239837646,
      "learning_rate": 4.718171459746785e-05,
      "loss": 0.1836,
      "num_input_tokens_seen": 5294688,
      "step": 6105
    },
    {
      "epoch": 2.880716643092881,
      "grad_norm": 1.222464919090271,
      "learning_rate": 4.717718454511273e-05,
      "loss": 0.2109,
      "num_input_tokens_seen": 5299264,
      "step": 6110
    },
    {
      "epoch": 2.883074021687883,
      "grad_norm": 1.904478669166565,
      "learning_rate": 4.7172651072757056e-05,
      "loss": 0.2415,
      "num_input_tokens_seen": 5303664,
      "step": 6115
    },
    {
      "epoch": 2.8854314002828856,
      "grad_norm": 0.42386630177497864,
      "learning_rate": 4.7168114181099945e-05,
      "loss": 0.2162,
      "num_input_tokens_seen": 5308448,
      "step": 6120
    },
    {
      "epoch": 2.887788778877888,
      "grad_norm": 1.8674676418304443,
      "learning_rate": 4.716357387084105e-05,
      "loss": 0.2445,
      "num_input_tokens_seen": 5312080,
      "step": 6125
    },
    {
      "epoch": 2.8901461574728904,
      "grad_norm": 0.8883599638938904,
      "learning_rate": 4.715903014268054e-05,
      "loss": 0.2118,
      "num_input_tokens_seen": 5315936,
      "step": 6130
    },
    {
      "epoch": 2.8925035360678923,
      "grad_norm": 0.7560530304908752,
      "learning_rate": 4.715448299731911e-05,
      "loss": 0.1573,
      "num_input_tokens_seen": 5320640,
      "step": 6135
    },
    {
      "epoch": 2.8948609146628947,
      "grad_norm": 0.4940895736217499,
      "learning_rate": 4.7149932435457986e-05,
      "loss": 0.208,
      "num_input_tokens_seen": 5324816,
      "step": 6140
    },
    {
      "epoch": 2.897218293257897,
      "grad_norm": 2.6361801624298096,
      "learning_rate": 4.714537845779894e-05,
      "loss": 0.2904,
      "num_input_tokens_seen": 5328512,
      "step": 6145
    },
    {
      "epoch": 2.8995756718528995,
      "grad_norm": 1.7376792430877686,
      "learning_rate": 4.714082106504423e-05,
      "loss": 0.1561,
      "num_input_tokens_seen": 5332848,
      "step": 6150
    },
    {
      "epoch": 2.901933050447902,
      "grad_norm": 1.9113662242889404,
      "learning_rate": 4.713626025789667e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 5337920,
      "step": 6155
    },
    {
      "epoch": 2.9042904290429044,
      "grad_norm": 0.836287260055542,
      "learning_rate": 4.7131696037059606e-05,
      "loss": 0.1731,
      "num_input_tokens_seen": 5342432,
      "step": 6160
    },
    {
      "epoch": 2.9066478076379068,
      "grad_norm": 1.0180597305297852,
      "learning_rate": 4.712712840323689e-05,
      "loss": 0.2365,
      "num_input_tokens_seen": 5346464,
      "step": 6165
    },
    {
      "epoch": 2.909005186232909,
      "grad_norm": 0.7450299263000488,
      "learning_rate": 4.71225573571329e-05,
      "loss": 0.2621,
      "num_input_tokens_seen": 5350944,
      "step": 6170
    },
    {
      "epoch": 2.911362564827911,
      "grad_norm": 1.0935964584350586,
      "learning_rate": 4.711798289945256e-05,
      "loss": 0.1762,
      "num_input_tokens_seen": 5354912,
      "step": 6175
    },
    {
      "epoch": 2.9137199434229135,
      "grad_norm": 1.0636024475097656,
      "learning_rate": 4.71134050309013e-05,
      "loss": 0.2331,
      "num_input_tokens_seen": 5359456,
      "step": 6180
    },
    {
      "epoch": 2.916077322017916,
      "grad_norm": 1.1808507442474365,
      "learning_rate": 4.710882375218509e-05,
      "loss": 0.1648,
      "num_input_tokens_seen": 5364336,
      "step": 6185
    },
    {
      "epoch": 2.9184347006129183,
      "grad_norm": 1.0397711992263794,
      "learning_rate": 4.7104239064010424e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 5368416,
      "step": 6190
    },
    {
      "epoch": 2.9207920792079207,
      "grad_norm": 0.46748360991477966,
      "learning_rate": 4.709965096708432e-05,
      "loss": 0.2157,
      "num_input_tokens_seen": 5372848,
      "step": 6195
    },
    {
      "epoch": 2.923149457802923,
      "grad_norm": 0.4934569001197815,
      "learning_rate": 4.709505946211431e-05,
      "loss": 0.1957,
      "num_input_tokens_seen": 5377360,
      "step": 6200
    },
    {
      "epoch": 2.923149457802923,
      "eval_loss": 0.21630728244781494,
      "eval_runtime": 22.1827,
      "eval_samples_per_second": 42.511,
      "eval_steps_per_second": 21.278,
      "num_input_tokens_seen": 5377360,
      "step": 6200
    },
    {
      "epoch": 2.9255068363979255,
      "grad_norm": 0.8423487544059753,
      "learning_rate": 4.709046454980846e-05,
      "loss": 0.1526,
      "num_input_tokens_seen": 5381136,
      "step": 6205
    },
    {
      "epoch": 2.927864214992928,
      "grad_norm": 0.9127216935157776,
      "learning_rate": 4.708586623087538e-05,
      "loss": 0.2384,
      "num_input_tokens_seen": 5385296,
      "step": 6210
    },
    {
      "epoch": 2.9302215935879303,
      "grad_norm": 0.9227361679077148,
      "learning_rate": 4.708126450602418e-05,
      "loss": 0.1908,
      "num_input_tokens_seen": 5389728,
      "step": 6215
    },
    {
      "epoch": 2.9325789721829327,
      "grad_norm": 0.8829623460769653,
      "learning_rate": 4.7076659375964495e-05,
      "loss": 0.165,
      "num_input_tokens_seen": 5393984,
      "step": 6220
    },
    {
      "epoch": 2.934936350777935,
      "grad_norm": 0.37931060791015625,
      "learning_rate": 4.707205084140651e-05,
      "loss": 0.1451,
      "num_input_tokens_seen": 5398048,
      "step": 6225
    },
    {
      "epoch": 2.9372937293729375,
      "grad_norm": 2.0882580280303955,
      "learning_rate": 4.7067438903060904e-05,
      "loss": 0.1965,
      "num_input_tokens_seen": 5402384,
      "step": 6230
    },
    {
      "epoch": 2.9396511079679395,
      "grad_norm": 1.0189132690429688,
      "learning_rate": 4.70628235616389e-05,
      "loss": 0.2382,
      "num_input_tokens_seen": 5406752,
      "step": 6235
    },
    {
      "epoch": 2.942008486562942,
      "grad_norm": 0.5234609842300415,
      "learning_rate": 4.7058204817852256e-05,
      "loss": 0.1997,
      "num_input_tokens_seen": 5411120,
      "step": 6240
    },
    {
      "epoch": 2.9443658651579443,
      "grad_norm": 1.2153555154800415,
      "learning_rate": 4.705358267241322e-05,
      "loss": 0.1844,
      "num_input_tokens_seen": 5414992,
      "step": 6245
    },
    {
      "epoch": 2.9467232437529467,
      "grad_norm": 1.4922072887420654,
      "learning_rate": 4.704895712603459e-05,
      "loss": 0.2316,
      "num_input_tokens_seen": 5419808,
      "step": 6250
    },
    {
      "epoch": 2.949080622347949,
      "grad_norm": 0.5180653929710388,
      "learning_rate": 4.704432817942969e-05,
      "loss": 0.1424,
      "num_input_tokens_seen": 5423840,
      "step": 6255
    },
    {
      "epoch": 2.9514380009429515,
      "grad_norm": 1.6094239950180054,
      "learning_rate": 4.703969583331236e-05,
      "loss": 0.2006,
      "num_input_tokens_seen": 5428208,
      "step": 6260
    },
    {
      "epoch": 2.953795379537954,
      "grad_norm": 1.178348422050476,
      "learning_rate": 4.7035060088396965e-05,
      "loss": 0.3109,
      "num_input_tokens_seen": 5433744,
      "step": 6265
    },
    {
      "epoch": 2.9561527581329563,
      "grad_norm": 0.7275745868682861,
      "learning_rate": 4.703042094539839e-05,
      "loss": 0.2294,
      "num_input_tokens_seen": 5437568,
      "step": 6270
    },
    {
      "epoch": 2.9585101367279583,
      "grad_norm": 1.4806724786758423,
      "learning_rate": 4.702577840503206e-05,
      "loss": 0.1875,
      "num_input_tokens_seen": 5442016,
      "step": 6275
    },
    {
      "epoch": 2.9608675153229607,
      "grad_norm": 0.7763034105300903,
      "learning_rate": 4.70211324680139e-05,
      "loss": 0.1649,
      "num_input_tokens_seen": 5445984,
      "step": 6280
    },
    {
      "epoch": 2.963224893917963,
      "grad_norm": 0.6184839606285095,
      "learning_rate": 4.7016483135060386e-05,
      "loss": 0.2058,
      "num_input_tokens_seen": 5450192,
      "step": 6285
    },
    {
      "epoch": 2.9655822725129655,
      "grad_norm": 1.4289072751998901,
      "learning_rate": 4.701183040688849e-05,
      "loss": 0.1482,
      "num_input_tokens_seen": 5455088,
      "step": 6290
    },
    {
      "epoch": 2.967939651107968,
      "grad_norm": 1.7712610960006714,
      "learning_rate": 4.700717428421573e-05,
      "loss": 0.1968,
      "num_input_tokens_seen": 5459152,
      "step": 6295
    },
    {
      "epoch": 2.9702970297029703,
      "grad_norm": 2.348703384399414,
      "learning_rate": 4.700251476776014e-05,
      "loss": 0.2867,
      "num_input_tokens_seen": 5462448,
      "step": 6300
    },
    {
      "epoch": 2.9726544082979727,
      "grad_norm": 1.8640990257263184,
      "learning_rate": 4.699785185824026e-05,
      "loss": 0.2015,
      "num_input_tokens_seen": 5466768,
      "step": 6305
    },
    {
      "epoch": 2.975011786892975,
      "grad_norm": 0.6836298108100891,
      "learning_rate": 4.699318555637519e-05,
      "loss": 0.2183,
      "num_input_tokens_seen": 5470448,
      "step": 6310
    },
    {
      "epoch": 2.9773691654879775,
      "grad_norm": 0.5324947834014893,
      "learning_rate": 4.6988515862884525e-05,
      "loss": 0.2016,
      "num_input_tokens_seen": 5475104,
      "step": 6315
    },
    {
      "epoch": 2.97972654408298,
      "grad_norm": 0.5939590930938721,
      "learning_rate": 4.698384277848838e-05,
      "loss": 0.1634,
      "num_input_tokens_seen": 5479632,
      "step": 6320
    },
    {
      "epoch": 2.9820839226779823,
      "grad_norm": 1.2299081087112427,
      "learning_rate": 4.6979166303907425e-05,
      "loss": 0.1497,
      "num_input_tokens_seen": 5483776,
      "step": 6325
    },
    {
      "epoch": 2.9844413012729847,
      "grad_norm": 1.6281503438949585,
      "learning_rate": 4.697448643986281e-05,
      "loss": 0.1958,
      "num_input_tokens_seen": 5487952,
      "step": 6330
    },
    {
      "epoch": 2.9867986798679866,
      "grad_norm": 0.8204582929611206,
      "learning_rate": 4.696980318707624e-05,
      "loss": 0.1977,
      "num_input_tokens_seen": 5491840,
      "step": 6335
    },
    {
      "epoch": 2.989156058462989,
      "grad_norm": 0.819779098033905,
      "learning_rate": 4.6965116546269924e-05,
      "loss": 0.2181,
      "num_input_tokens_seen": 5496976,
      "step": 6340
    },
    {
      "epoch": 2.9915134370579914,
      "grad_norm": 0.5795984864234924,
      "learning_rate": 4.6960426518166615e-05,
      "loss": 0.1808,
      "num_input_tokens_seen": 5500800,
      "step": 6345
    },
    {
      "epoch": 2.993870815652994,
      "grad_norm": 1.0978055000305176,
      "learning_rate": 4.6955733103489556e-05,
      "loss": 0.2134,
      "num_input_tokens_seen": 5504688,
      "step": 6350
    },
    {
      "epoch": 2.9962281942479962,
      "grad_norm": 0.6574112772941589,
      "learning_rate": 4.695103630296255e-05,
      "loss": 0.198,
      "num_input_tokens_seen": 5508864,
      "step": 6355
    },
    {
      "epoch": 2.9985855728429986,
      "grad_norm": 0.8596549034118652,
      "learning_rate": 4.694633611730988e-05,
      "loss": 0.2051,
      "num_input_tokens_seen": 5512784,
      "step": 6360
    },
    {
      "epoch": 3.000942951438001,
      "grad_norm": 1.1129478216171265,
      "learning_rate": 4.694163254725639e-05,
      "loss": 0.1827,
      "num_input_tokens_seen": 5516928,
      "step": 6365
    },
    {
      "epoch": 3.0033003300330035,
      "grad_norm": 0.5702999830245972,
      "learning_rate": 4.693692559352743e-05,
      "loss": 0.2233,
      "num_input_tokens_seen": 5521024,
      "step": 6370
    },
    {
      "epoch": 3.005657708628006,
      "grad_norm": 2.027930736541748,
      "learning_rate": 4.693221525684886e-05,
      "loss": 0.1899,
      "num_input_tokens_seen": 5526416,
      "step": 6375
    },
    {
      "epoch": 3.008015087223008,
      "grad_norm": 1.4407075643539429,
      "learning_rate": 4.6927501537947084e-05,
      "loss": 0.199,
      "num_input_tokens_seen": 5531440,
      "step": 6380
    },
    {
      "epoch": 3.01037246581801,
      "grad_norm": 1.6307483911514282,
      "learning_rate": 4.692278443754901e-05,
      "loss": 0.2278,
      "num_input_tokens_seen": 5535920,
      "step": 6385
    },
    {
      "epoch": 3.0127298444130126,
      "grad_norm": 1.7686395645141602,
      "learning_rate": 4.691806395638208e-05,
      "loss": 0.2344,
      "num_input_tokens_seen": 5540432,
      "step": 6390
    },
    {
      "epoch": 3.015087223008015,
      "grad_norm": 0.3188937306404114,
      "learning_rate": 4.6913340095174255e-05,
      "loss": 0.1376,
      "num_input_tokens_seen": 5546240,
      "step": 6395
    },
    {
      "epoch": 3.0174446016030174,
      "grad_norm": 1.8880008459091187,
      "learning_rate": 4.690861285465399e-05,
      "loss": 0.1939,
      "num_input_tokens_seen": 5550480,
      "step": 6400
    },
    {
      "epoch": 3.0174446016030174,
      "eval_loss": 0.2152761071920395,
      "eval_runtime": 22.17,
      "eval_samples_per_second": 42.535,
      "eval_steps_per_second": 21.29,
      "num_input_tokens_seen": 5550480,
      "step": 6400
    },
    {
      "epoch": 3.01980198019802,
      "grad_norm": 1.7155089378356934,
      "learning_rate": 4.690388223555031e-05,
      "loss": 0.2076,
      "num_input_tokens_seen": 5554144,
      "step": 6405
    },
    {
      "epoch": 3.022159358793022,
      "grad_norm": 0.5357860326766968,
      "learning_rate": 4.689914823859273e-05,
      "loss": 0.2039,
      "num_input_tokens_seen": 5558128,
      "step": 6410
    },
    {
      "epoch": 3.0245167373880246,
      "grad_norm": 0.764172375202179,
      "learning_rate": 4.689441086451129e-05,
      "loss": 0.1829,
      "num_input_tokens_seen": 5562896,
      "step": 6415
    },
    {
      "epoch": 3.026874115983027,
      "grad_norm": 0.7307031154632568,
      "learning_rate": 4.688967011403655e-05,
      "loss": 0.243,
      "num_input_tokens_seen": 5567184,
      "step": 6420
    },
    {
      "epoch": 3.0292314945780294,
      "grad_norm": 0.48321405053138733,
      "learning_rate": 4.68849259878996e-05,
      "loss": 0.2075,
      "num_input_tokens_seen": 5571984,
      "step": 6425
    },
    {
      "epoch": 3.0315888731730314,
      "grad_norm": 0.7918155193328857,
      "learning_rate": 4.6880178486832036e-05,
      "loss": 0.2042,
      "num_input_tokens_seen": 5576528,
      "step": 6430
    },
    {
      "epoch": 3.033946251768034,
      "grad_norm": 1.5321390628814697,
      "learning_rate": 4.687542761156598e-05,
      "loss": 0.2273,
      "num_input_tokens_seen": 5581632,
      "step": 6435
    },
    {
      "epoch": 3.036303630363036,
      "grad_norm": 1.190408706665039,
      "learning_rate": 4.6870673362834096e-05,
      "loss": 0.1736,
      "num_input_tokens_seen": 5586720,
      "step": 6440
    },
    {
      "epoch": 3.0386610089580386,
      "grad_norm": 0.7143381237983704,
      "learning_rate": 4.6865915741369526e-05,
      "loss": 0.1391,
      "num_input_tokens_seen": 5590256,
      "step": 6445
    },
    {
      "epoch": 3.041018387553041,
      "grad_norm": 1.0550860166549683,
      "learning_rate": 4.686115474790597e-05,
      "loss": 0.1934,
      "num_input_tokens_seen": 5594912,
      "step": 6450
    },
    {
      "epoch": 3.0433757661480434,
      "grad_norm": 0.9228099584579468,
      "learning_rate": 4.685639038317762e-05,
      "loss": 0.2291,
      "num_input_tokens_seen": 5599248,
      "step": 6455
    },
    {
      "epoch": 3.045733144743046,
      "grad_norm": 1.0951104164123535,
      "learning_rate": 4.685162264791921e-05,
      "loss": 0.2372,
      "num_input_tokens_seen": 5603840,
      "step": 6460
    },
    {
      "epoch": 3.048090523338048,
      "grad_norm": 1.5756505727767944,
      "learning_rate": 4.684685154286599e-05,
      "loss": 0.2074,
      "num_input_tokens_seen": 5608688,
      "step": 6465
    },
    {
      "epoch": 3.0504479019330506,
      "grad_norm": 0.8257424831390381,
      "learning_rate": 4.684207706875371e-05,
      "loss": 0.209,
      "num_input_tokens_seen": 5612896,
      "step": 6470
    },
    {
      "epoch": 3.052805280528053,
      "grad_norm": 1.6899330615997314,
      "learning_rate": 4.683729922631866e-05,
      "loss": 0.2352,
      "num_input_tokens_seen": 5617232,
      "step": 6475
    },
    {
      "epoch": 3.055162659123055,
      "grad_norm": 0.9218118786811829,
      "learning_rate": 4.683251801629765e-05,
      "loss": 0.1906,
      "num_input_tokens_seen": 5621216,
      "step": 6480
    },
    {
      "epoch": 3.0575200377180574,
      "grad_norm": 0.4963802099227905,
      "learning_rate": 4.6827733439428e-05,
      "loss": 0.2331,
      "num_input_tokens_seen": 5625664,
      "step": 6485
    },
    {
      "epoch": 3.0598774163130598,
      "grad_norm": 0.9776608943939209,
      "learning_rate": 4.682294549644754e-05,
      "loss": 0.2094,
      "num_input_tokens_seen": 5629472,
      "step": 6490
    },
    {
      "epoch": 3.062234794908062,
      "grad_norm": 1.3240665197372437,
      "learning_rate": 4.681815418809464e-05,
      "loss": 0.2031,
      "num_input_tokens_seen": 5633936,
      "step": 6495
    },
    {
      "epoch": 3.0645921735030646,
      "grad_norm": 1.0820789337158203,
      "learning_rate": 4.681335951510819e-05,
      "loss": 0.1963,
      "num_input_tokens_seen": 5638112,
      "step": 6500
    },
    {
      "epoch": 3.066949552098067,
      "grad_norm": 0.5462324023246765,
      "learning_rate": 4.6808561478227576e-05,
      "loss": 0.1608,
      "num_input_tokens_seen": 5642112,
      "step": 6505
    },
    {
      "epoch": 3.0693069306930694,
      "grad_norm": 1.2959402799606323,
      "learning_rate": 4.680376007819271e-05,
      "loss": 0.1649,
      "num_input_tokens_seen": 5646384,
      "step": 6510
    },
    {
      "epoch": 3.0716643092880718,
      "grad_norm": 1.0798271894454956,
      "learning_rate": 4.679895531574405e-05,
      "loss": 0.2128,
      "num_input_tokens_seen": 5650912,
      "step": 6515
    },
    {
      "epoch": 3.074021687883074,
      "grad_norm": 0.5697314739227295,
      "learning_rate": 4.679414719162253e-05,
      "loss": 0.1376,
      "num_input_tokens_seen": 5655472,
      "step": 6520
    },
    {
      "epoch": 3.0763790664780766,
      "grad_norm": 1.1794296503067017,
      "learning_rate": 4.6789335706569635e-05,
      "loss": 0.1855,
      "num_input_tokens_seen": 5659296,
      "step": 6525
    },
    {
      "epoch": 3.0787364450730785,
      "grad_norm": 0.43731486797332764,
      "learning_rate": 4.678452086132734e-05,
      "loss": 0.1741,
      "num_input_tokens_seen": 5663920,
      "step": 6530
    },
    {
      "epoch": 3.081093823668081,
      "grad_norm": 1.9594507217407227,
      "learning_rate": 4.677970265663818e-05,
      "loss": 0.1876,
      "num_input_tokens_seen": 5667840,
      "step": 6535
    },
    {
      "epoch": 3.0834512022630833,
      "grad_norm": 1.4631396532058716,
      "learning_rate": 4.677488109324517e-05,
      "loss": 0.1426,
      "num_input_tokens_seen": 5672656,
      "step": 6540
    },
    {
      "epoch": 3.0858085808580857,
      "grad_norm": 0.6862918138504028,
      "learning_rate": 4.6770056171891846e-05,
      "loss": 0.2392,
      "num_input_tokens_seen": 5677040,
      "step": 6545
    },
    {
      "epoch": 3.088165959453088,
      "grad_norm": 1.1463441848754883,
      "learning_rate": 4.6765227893322286e-05,
      "loss": 0.1924,
      "num_input_tokens_seen": 5680528,
      "step": 6550
    },
    {
      "epoch": 3.0905233380480905,
      "grad_norm": 1.0818536281585693,
      "learning_rate": 4.676039625828107e-05,
      "loss": 0.1786,
      "num_input_tokens_seen": 5685248,
      "step": 6555
    },
    {
      "epoch": 3.092880716643093,
      "grad_norm": 0.9310617446899414,
      "learning_rate": 4.675556126751328e-05,
      "loss": 0.1933,
      "num_input_tokens_seen": 5688976,
      "step": 6560
    },
    {
      "epoch": 3.0952380952380953,
      "grad_norm": 1.2195324897766113,
      "learning_rate": 4.6750722921764556e-05,
      "loss": 0.1638,
      "num_input_tokens_seen": 5692960,
      "step": 6565
    },
    {
      "epoch": 3.0975954738330977,
      "grad_norm": 0.6135625839233398,
      "learning_rate": 4.674588122178102e-05,
      "loss": 0.1808,
      "num_input_tokens_seen": 5697248,
      "step": 6570
    },
    {
      "epoch": 3.0999528524281,
      "grad_norm": 1.6157267093658447,
      "learning_rate": 4.674103616830931e-05,
      "loss": 0.1602,
      "num_input_tokens_seen": 5701328,
      "step": 6575
    },
    {
      "epoch": 3.102310231023102,
      "grad_norm": 0.8030937314033508,
      "learning_rate": 4.673618776209663e-05,
      "loss": 0.1699,
      "num_input_tokens_seen": 5706528,
      "step": 6580
    },
    {
      "epoch": 3.1046676096181045,
      "grad_norm": 1.2519348859786987,
      "learning_rate": 4.673133600389063e-05,
      "loss": 0.2045,
      "num_input_tokens_seen": 5710752,
      "step": 6585
    },
    {
      "epoch": 3.107024988213107,
      "grad_norm": 0.5498633980751038,
      "learning_rate": 4.672648089443953e-05,
      "loss": 0.2002,
      "num_input_tokens_seen": 5715520,
      "step": 6590
    },
    {
      "epoch": 3.1093823668081093,
      "grad_norm": 1.2372210025787354,
      "learning_rate": 4.672162243449204e-05,
      "loss": 0.1994,
      "num_input_tokens_seen": 5719760,
      "step": 6595
    },
    {
      "epoch": 3.1117397454031117,
      "grad_norm": 2.696047306060791,
      "learning_rate": 4.67167606247974e-05,
      "loss": 0.2674,
      "num_input_tokens_seen": 5724080,
      "step": 6600
    },
    {
      "epoch": 3.1117397454031117,
      "eval_loss": 0.21502578258514404,
      "eval_runtime": 22.1751,
      "eval_samples_per_second": 42.525,
      "eval_steps_per_second": 21.285,
      "num_input_tokens_seen": 5724080,
      "step": 6600
    },
    {
      "epoch": 3.114097123998114,
      "grad_norm": 0.6930589079856873,
      "learning_rate": 4.671189546610536e-05,
      "loss": 0.1285,
      "num_input_tokens_seen": 5728128,
      "step": 6605
    },
    {
      "epoch": 3.1164545025931165,
      "grad_norm": 0.5578516721725464,
      "learning_rate": 4.67070269591662e-05,
      "loss": 0.2012,
      "num_input_tokens_seen": 5732576,
      "step": 6610
    },
    {
      "epoch": 3.118811881188119,
      "grad_norm": 1.3748961687088013,
      "learning_rate": 4.670215510473068e-05,
      "loss": 0.1769,
      "num_input_tokens_seen": 5735888,
      "step": 6615
    },
    {
      "epoch": 3.1211692597831213,
      "grad_norm": 1.768334150314331,
      "learning_rate": 4.669727990355013e-05,
      "loss": 0.1675,
      "num_input_tokens_seen": 5739968,
      "step": 6620
    },
    {
      "epoch": 3.1235266383781237,
      "grad_norm": 0.954632043838501,
      "learning_rate": 4.669240135637635e-05,
      "loss": 0.2215,
      "num_input_tokens_seen": 5743824,
      "step": 6625
    },
    {
      "epoch": 3.1258840169731257,
      "grad_norm": 2.677161693572998,
      "learning_rate": 4.6687519463961675e-05,
      "loss": 0.292,
      "num_input_tokens_seen": 5748000,
      "step": 6630
    },
    {
      "epoch": 3.128241395568128,
      "grad_norm": 0.4749905467033386,
      "learning_rate": 4.668263422705896e-05,
      "loss": 0.1488,
      "num_input_tokens_seen": 5752384,
      "step": 6635
    },
    {
      "epoch": 3.1305987741631305,
      "grad_norm": 1.683872938156128,
      "learning_rate": 4.667774564642156e-05,
      "loss": 0.2043,
      "num_input_tokens_seen": 5757216,
      "step": 6640
    },
    {
      "epoch": 3.132956152758133,
      "grad_norm": 0.7426241636276245,
      "learning_rate": 4.6672853722803365e-05,
      "loss": 0.1687,
      "num_input_tokens_seen": 5762368,
      "step": 6645
    },
    {
      "epoch": 3.1353135313531353,
      "grad_norm": 0.297019362449646,
      "learning_rate": 4.666795845695877e-05,
      "loss": 0.2242,
      "num_input_tokens_seen": 5766592,
      "step": 6650
    },
    {
      "epoch": 3.1376709099481377,
      "grad_norm": 0.5431145429611206,
      "learning_rate": 4.666305984964269e-05,
      "loss": 0.1436,
      "num_input_tokens_seen": 5770624,
      "step": 6655
    },
    {
      "epoch": 3.14002828854314,
      "grad_norm": 0.583001971244812,
      "learning_rate": 4.6658157901610535e-05,
      "loss": 0.224,
      "num_input_tokens_seen": 5775296,
      "step": 6660
    },
    {
      "epoch": 3.1423856671381425,
      "grad_norm": 0.6226017475128174,
      "learning_rate": 4.665325261361826e-05,
      "loss": 0.2694,
      "num_input_tokens_seen": 5779280,
      "step": 6665
    },
    {
      "epoch": 3.144743045733145,
      "grad_norm": 1.010037899017334,
      "learning_rate": 4.664834398642232e-05,
      "loss": 0.1883,
      "num_input_tokens_seen": 5783856,
      "step": 6670
    },
    {
      "epoch": 3.1471004243281473,
      "grad_norm": 0.7547823786735535,
      "learning_rate": 4.6643432020779686e-05,
      "loss": 0.1393,
      "num_input_tokens_seen": 5788336,
      "step": 6675
    },
    {
      "epoch": 3.1494578029231493,
      "grad_norm": 0.5626434087753296,
      "learning_rate": 4.663851671744786e-05,
      "loss": 0.2063,
      "num_input_tokens_seen": 5793040,
      "step": 6680
    },
    {
      "epoch": 3.1518151815181517,
      "grad_norm": 0.6840236186981201,
      "learning_rate": 4.6633598077184815e-05,
      "loss": 0.2438,
      "num_input_tokens_seen": 5798576,
      "step": 6685
    },
    {
      "epoch": 3.154172560113154,
      "grad_norm": 1.7111215591430664,
      "learning_rate": 4.662867610074908e-05,
      "loss": 0.2607,
      "num_input_tokens_seen": 5803760,
      "step": 6690
    },
    {
      "epoch": 3.1565299387081565,
      "grad_norm": 1.225968599319458,
      "learning_rate": 4.6623750788899696e-05,
      "loss": 0.2713,
      "num_input_tokens_seen": 5808192,
      "step": 6695
    },
    {
      "epoch": 3.158887317303159,
      "grad_norm": 0.8442971706390381,
      "learning_rate": 4.6618822142396195e-05,
      "loss": 0.2186,
      "num_input_tokens_seen": 5812400,
      "step": 6700
    },
    {
      "epoch": 3.1612446958981613,
      "grad_norm": 0.800540030002594,
      "learning_rate": 4.661389016199864e-05,
      "loss": 0.2103,
      "num_input_tokens_seen": 5817216,
      "step": 6705
    },
    {
      "epoch": 3.1636020744931637,
      "grad_norm": 0.6143836975097656,
      "learning_rate": 4.660895484846761e-05,
      "loss": 0.2174,
      "num_input_tokens_seen": 5820944,
      "step": 6710
    },
    {
      "epoch": 3.165959453088166,
      "grad_norm": 0.35425233840942383,
      "learning_rate": 4.660401620256418e-05,
      "loss": 0.1871,
      "num_input_tokens_seen": 5824912,
      "step": 6715
    },
    {
      "epoch": 3.1683168316831685,
      "grad_norm": 1.1348215341567993,
      "learning_rate": 4.659907422504997e-05,
      "loss": 0.1956,
      "num_input_tokens_seen": 5828944,
      "step": 6720
    },
    {
      "epoch": 3.170674210278171,
      "grad_norm": 1.50424325466156,
      "learning_rate": 4.6594128916687074e-05,
      "loss": 0.2443,
      "num_input_tokens_seen": 5832608,
      "step": 6725
    },
    {
      "epoch": 3.173031588873173,
      "grad_norm": 0.7212063074111938,
      "learning_rate": 4.658918027823813e-05,
      "loss": 0.2078,
      "num_input_tokens_seen": 5836608,
      "step": 6730
    },
    {
      "epoch": 3.1753889674681752,
      "grad_norm": 1.0809381008148193,
      "learning_rate": 4.658422831046628e-05,
      "loss": 0.1851,
      "num_input_tokens_seen": 5840416,
      "step": 6735
    },
    {
      "epoch": 3.1777463460631776,
      "grad_norm": 1.0143598318099976,
      "learning_rate": 4.657927301413518e-05,
      "loss": 0.1956,
      "num_input_tokens_seen": 5844032,
      "step": 6740
    },
    {
      "epoch": 3.18010372465818,
      "grad_norm": 0.4378543198108673,
      "learning_rate": 4.657431439000901e-05,
      "loss": 0.2197,
      "num_input_tokens_seen": 5848336,
      "step": 6745
    },
    {
      "epoch": 3.1824611032531824,
      "grad_norm": 0.7392013072967529,
      "learning_rate": 4.656935243885243e-05,
      "loss": 0.1903,
      "num_input_tokens_seen": 5852976,
      "step": 6750
    },
    {
      "epoch": 3.184818481848185,
      "grad_norm": 1.35980224609375,
      "learning_rate": 4.656438716143066e-05,
      "loss": 0.2498,
      "num_input_tokens_seen": 5857552,
      "step": 6755
    },
    {
      "epoch": 3.1871758604431872,
      "grad_norm": 0.5861925482749939,
      "learning_rate": 4.6559418558509384e-05,
      "loss": 0.1637,
      "num_input_tokens_seen": 5862336,
      "step": 6760
    },
    {
      "epoch": 3.1895332390381896,
      "grad_norm": 1.572373390197754,
      "learning_rate": 4.6554446630854833e-05,
      "loss": 0.1697,
      "num_input_tokens_seen": 5866368,
      "step": 6765
    },
    {
      "epoch": 3.191890617633192,
      "grad_norm": 0.6973896026611328,
      "learning_rate": 4.654947137923374e-05,
      "loss": 0.1965,
      "num_input_tokens_seen": 5871600,
      "step": 6770
    },
    {
      "epoch": 3.1942479962281944,
      "grad_norm": 0.975184977054596,
      "learning_rate": 4.654449280441335e-05,
      "loss": 0.1964,
      "num_input_tokens_seen": 5876560,
      "step": 6775
    },
    {
      "epoch": 3.1966053748231964,
      "grad_norm": 1.149117350578308,
      "learning_rate": 4.653951090716143e-05,
      "loss": 0.2412,
      "num_input_tokens_seen": 5880256,
      "step": 6780
    },
    {
      "epoch": 3.198962753418199,
      "grad_norm": 1.006313681602478,
      "learning_rate": 4.653452568824625e-05,
      "loss": 0.1593,
      "num_input_tokens_seen": 5884096,
      "step": 6785
    },
    {
      "epoch": 3.201320132013201,
      "grad_norm": 0.8185628652572632,
      "learning_rate": 4.6529537148436585e-05,
      "loss": 0.2185,
      "num_input_tokens_seen": 5888256,
      "step": 6790
    },
    {
      "epoch": 3.2036775106082036,
      "grad_norm": 0.5432038307189941,
      "learning_rate": 4.6524545288501734e-05,
      "loss": 0.1908,
      "num_input_tokens_seen": 5892768,
      "step": 6795
    },
    {
      "epoch": 3.206034889203206,
      "grad_norm": 0.7089331746101379,
      "learning_rate": 4.6519550109211506e-05,
      "loss": 0.1837,
      "num_input_tokens_seen": 5896688,
      "step": 6800
    },
    {
      "epoch": 3.206034889203206,
      "eval_loss": 0.2133202701807022,
      "eval_runtime": 22.1613,
      "eval_samples_per_second": 42.552,
      "eval_steps_per_second": 21.298,
      "num_input_tokens_seen": 5896688,
      "step": 6800
    },
    {
      "epoch": 3.2083922677982084,
      "grad_norm": 0.9046180248260498,
      "learning_rate": 4.651455161133622e-05,
      "loss": 0.1744,
      "num_input_tokens_seen": 5901520,
      "step": 6805
    },
    {
      "epoch": 3.210749646393211,
      "grad_norm": 1.1667699813842773,
      "learning_rate": 4.6509549795646704e-05,
      "loss": 0.1791,
      "num_input_tokens_seen": 5905920,
      "step": 6810
    },
    {
      "epoch": 3.213107024988213,
      "grad_norm": 0.7543178200721741,
      "learning_rate": 4.6504544662914306e-05,
      "loss": 0.1551,
      "num_input_tokens_seen": 5909840,
      "step": 6815
    },
    {
      "epoch": 3.2154644035832156,
      "grad_norm": 1.4581403732299805,
      "learning_rate": 4.6499536213910876e-05,
      "loss": 0.2026,
      "num_input_tokens_seen": 5914752,
      "step": 6820
    },
    {
      "epoch": 3.217821782178218,
      "grad_norm": 0.9164820313453674,
      "learning_rate": 4.6494524449408786e-05,
      "loss": 0.2351,
      "num_input_tokens_seen": 5919424,
      "step": 6825
    },
    {
      "epoch": 3.22017916077322,
      "grad_norm": 0.6268575191497803,
      "learning_rate": 4.6489509370180903e-05,
      "loss": 0.2037,
      "num_input_tokens_seen": 5924432,
      "step": 6830
    },
    {
      "epoch": 3.2225365393682224,
      "grad_norm": 1.0165094137191772,
      "learning_rate": 4.648449097700063e-05,
      "loss": 0.2163,
      "num_input_tokens_seen": 5928848,
      "step": 6835
    },
    {
      "epoch": 3.2248939179632248,
      "grad_norm": 0.656306266784668,
      "learning_rate": 4.647946927064185e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 5933152,
      "step": 6840
    },
    {
      "epoch": 3.227251296558227,
      "grad_norm": 1.759229302406311,
      "learning_rate": 4.647444425187898e-05,
      "loss": 0.2698,
      "num_input_tokens_seen": 5937120,
      "step": 6845
    },
    {
      "epoch": 3.2296086751532296,
      "grad_norm": 0.9243722558021545,
      "learning_rate": 4.646941592148695e-05,
      "loss": 0.1918,
      "num_input_tokens_seen": 5941248,
      "step": 6850
    },
    {
      "epoch": 3.231966053748232,
      "grad_norm": 1.0959793329238892,
      "learning_rate": 4.646438428024117e-05,
      "loss": 0.196,
      "num_input_tokens_seen": 5945472,
      "step": 6855
    },
    {
      "epoch": 3.2343234323432344,
      "grad_norm": 0.8084420561790466,
      "learning_rate": 4.64593493289176e-05,
      "loss": 0.2433,
      "num_input_tokens_seen": 5949888,
      "step": 6860
    },
    {
      "epoch": 3.236680810938237,
      "grad_norm": 1.327669620513916,
      "learning_rate": 4.64543110682927e-05,
      "loss": 0.2473,
      "num_input_tokens_seen": 5953632,
      "step": 6865
    },
    {
      "epoch": 3.239038189533239,
      "grad_norm": 0.5772126913070679,
      "learning_rate": 4.644926949914341e-05,
      "loss": 0.1974,
      "num_input_tokens_seen": 5957808,
      "step": 6870
    },
    {
      "epoch": 3.2413955681282416,
      "grad_norm": 0.7239676713943481,
      "learning_rate": 4.644422462224722e-05,
      "loss": 0.1785,
      "num_input_tokens_seen": 5961360,
      "step": 6875
    },
    {
      "epoch": 3.2437529467232435,
      "grad_norm": 1.0365188121795654,
      "learning_rate": 4.643917643838211e-05,
      "loss": 0.1846,
      "num_input_tokens_seen": 5966528,
      "step": 6880
    },
    {
      "epoch": 3.246110325318246,
      "grad_norm": 0.9796733856201172,
      "learning_rate": 4.6434124948326564e-05,
      "loss": 0.189,
      "num_input_tokens_seen": 5970400,
      "step": 6885
    },
    {
      "epoch": 3.2484677039132484,
      "grad_norm": 0.999091386795044,
      "learning_rate": 4.6429070152859594e-05,
      "loss": 0.1323,
      "num_input_tokens_seen": 5975616,
      "step": 6890
    },
    {
      "epoch": 3.2508250825082508,
      "grad_norm": 1.4964356422424316,
      "learning_rate": 4.6424012052760714e-05,
      "loss": 0.2022,
      "num_input_tokens_seen": 5979856,
      "step": 6895
    },
    {
      "epoch": 3.253182461103253,
      "grad_norm": 1.827711820602417,
      "learning_rate": 4.6418950648809945e-05,
      "loss": 0.2022,
      "num_input_tokens_seen": 5985008,
      "step": 6900
    },
    {
      "epoch": 3.2555398396982556,
      "grad_norm": 1.6295924186706543,
      "learning_rate": 4.641388594178782e-05,
      "loss": 0.2594,
      "num_input_tokens_seen": 5990464,
      "step": 6905
    },
    {
      "epoch": 3.257897218293258,
      "grad_norm": 0.8076730370521545,
      "learning_rate": 4.640881793247538e-05,
      "loss": 0.2041,
      "num_input_tokens_seen": 5995024,
      "step": 6910
    },
    {
      "epoch": 3.2602545968882604,
      "grad_norm": 0.7767298817634583,
      "learning_rate": 4.6403746621654173e-05,
      "loss": 0.3373,
      "num_input_tokens_seen": 5999488,
      "step": 6915
    },
    {
      "epoch": 3.2626119754832628,
      "grad_norm": 0.8800544142723083,
      "learning_rate": 4.639867201010626e-05,
      "loss": 0.241,
      "num_input_tokens_seen": 6003536,
      "step": 6920
    },
    {
      "epoch": 3.264969354078265,
      "grad_norm": 1.1187702417373657,
      "learning_rate": 4.6393594098614204e-05,
      "loss": 0.2031,
      "num_input_tokens_seen": 6008432,
      "step": 6925
    },
    {
      "epoch": 3.2673267326732676,
      "grad_norm": 0.4007323384284973,
      "learning_rate": 4.63885128879611e-05,
      "loss": 0.1383,
      "num_input_tokens_seen": 6012016,
      "step": 6930
    },
    {
      "epoch": 3.2696841112682695,
      "grad_norm": 1.105604887008667,
      "learning_rate": 4.638342837893052e-05,
      "loss": 0.2181,
      "num_input_tokens_seen": 6017088,
      "step": 6935
    },
    {
      "epoch": 3.272041489863272,
      "grad_norm": 1.2768769264221191,
      "learning_rate": 4.6378340572306565e-05,
      "loss": 0.2326,
      "num_input_tokens_seen": 6021360,
      "step": 6940
    },
    {
      "epoch": 3.2743988684582743,
      "grad_norm": 0.9717500805854797,
      "learning_rate": 4.6373249468873833e-05,
      "loss": 0.172,
      "num_input_tokens_seen": 6025504,
      "step": 6945
    },
    {
      "epoch": 3.2767562470532767,
      "grad_norm": 1.000753402709961,
      "learning_rate": 4.636815506941744e-05,
      "loss": 0.2333,
      "num_input_tokens_seen": 6029824,
      "step": 6950
    },
    {
      "epoch": 3.279113625648279,
      "grad_norm": 0.6670440435409546,
      "learning_rate": 4.6363057374723004e-05,
      "loss": 0.1719,
      "num_input_tokens_seen": 6034432,
      "step": 6955
    },
    {
      "epoch": 3.2814710042432815,
      "grad_norm": 1.1376651525497437,
      "learning_rate": 4.635795638557666e-05,
      "loss": 0.194,
      "num_input_tokens_seen": 6038448,
      "step": 6960
    },
    {
      "epoch": 3.283828382838284,
      "grad_norm": 1.012152075767517,
      "learning_rate": 4.635285210276504e-05,
      "loss": 0.2239,
      "num_input_tokens_seen": 6042256,
      "step": 6965
    },
    {
      "epoch": 3.2861857614332863,
      "grad_norm": 0.797825276851654,
      "learning_rate": 4.6347744527075295e-05,
      "loss": 0.2814,
      "num_input_tokens_seen": 6046064,
      "step": 6970
    },
    {
      "epoch": 3.2885431400282887,
      "grad_norm": 1.7127622365951538,
      "learning_rate": 4.634263365929506e-05,
      "loss": 0.167,
      "num_input_tokens_seen": 6050064,
      "step": 6975
    },
    {
      "epoch": 3.2909005186232907,
      "grad_norm": 1.2590041160583496,
      "learning_rate": 4.6337519500212515e-05,
      "loss": 0.1987,
      "num_input_tokens_seen": 6054176,
      "step": 6980
    },
    {
      "epoch": 3.293257897218293,
      "grad_norm": 1.571418285369873,
      "learning_rate": 4.633240205061632e-05,
      "loss": 0.1518,
      "num_input_tokens_seen": 6058368,
      "step": 6985
    },
    {
      "epoch": 3.2956152758132955,
      "grad_norm": 2.225250005722046,
      "learning_rate": 4.632728131129565e-05,
      "loss": 0.2369,
      "num_input_tokens_seen": 6062272,
      "step": 6990
    },
    {
      "epoch": 3.297972654408298,
      "grad_norm": 1.2845070362091064,
      "learning_rate": 4.632215728304018e-05,
      "loss": 0.1823,
      "num_input_tokens_seen": 6065888,
      "step": 6995
    },
    {
      "epoch": 3.3003300330033003,
      "grad_norm": 1.5594967603683472,
      "learning_rate": 4.63170299666401e-05,
      "loss": 0.2225,
      "num_input_tokens_seen": 6070544,
      "step": 7000
    },
    {
      "epoch": 3.3003300330033003,
      "eval_loss": 0.216164693236351,
      "eval_runtime": 22.1847,
      "eval_samples_per_second": 42.507,
      "eval_steps_per_second": 21.276,
      "num_input_tokens_seen": 6070544,
      "step": 7000
    },
    {
      "epoch": 3.3026874115983027,
      "grad_norm": 1.139223337173462,
      "learning_rate": 4.631189936288612e-05,
      "loss": 0.1899,
      "num_input_tokens_seen": 6075024,
      "step": 7005
    },
    {
      "epoch": 3.305044790193305,
      "grad_norm": 1.1729024648666382,
      "learning_rate": 4.630676547256944e-05,
      "loss": 0.1982,
      "num_input_tokens_seen": 6079120,
      "step": 7010
    },
    {
      "epoch": 3.3074021687883075,
      "grad_norm": 1.4162888526916504,
      "learning_rate": 4.630162829648176e-05,
      "loss": 0.2176,
      "num_input_tokens_seen": 6082512,
      "step": 7015
    },
    {
      "epoch": 3.30975954738331,
      "grad_norm": 2.7713592052459717,
      "learning_rate": 4.629648783541531e-05,
      "loss": 0.2814,
      "num_input_tokens_seen": 6086992,
      "step": 7020
    },
    {
      "epoch": 3.3121169259783123,
      "grad_norm": 1.537378191947937,
      "learning_rate": 4.6291344090162804e-05,
      "loss": 0.2245,
      "num_input_tokens_seen": 6091984,
      "step": 7025
    },
    {
      "epoch": 3.3144743045733147,
      "grad_norm": 0.887427031993866,
      "learning_rate": 4.628619706151748e-05,
      "loss": 0.186,
      "num_input_tokens_seen": 6095632,
      "step": 7030
    },
    {
      "epoch": 3.3168316831683167,
      "grad_norm": 0.6890794634819031,
      "learning_rate": 4.628104675027306e-05,
      "loss": 0.185,
      "num_input_tokens_seen": 6100656,
      "step": 7035
    },
    {
      "epoch": 3.319189061763319,
      "grad_norm": 1.5001062154769897,
      "learning_rate": 4.6275893157223805e-05,
      "loss": 0.2841,
      "num_input_tokens_seen": 6105248,
      "step": 7040
    },
    {
      "epoch": 3.3215464403583215,
      "grad_norm": 0.4696716368198395,
      "learning_rate": 4.627073628316445e-05,
      "loss": 0.19,
      "num_input_tokens_seen": 6109184,
      "step": 7045
    },
    {
      "epoch": 3.323903818953324,
      "grad_norm": 0.764578104019165,
      "learning_rate": 4.626557612889026e-05,
      "loss": 0.2019,
      "num_input_tokens_seen": 6113696,
      "step": 7050
    },
    {
      "epoch": 3.3262611975483263,
      "grad_norm": 1.2442392110824585,
      "learning_rate": 4.626041269519699e-05,
      "loss": 0.1679,
      "num_input_tokens_seen": 6118080,
      "step": 7055
    },
    {
      "epoch": 3.3286185761433287,
      "grad_norm": 1.0614804029464722,
      "learning_rate": 4.6255245982880905e-05,
      "loss": 0.2103,
      "num_input_tokens_seen": 6122576,
      "step": 7060
    },
    {
      "epoch": 3.330975954738331,
      "grad_norm": 0.48078614473342896,
      "learning_rate": 4.625007599273879e-05,
      "loss": 0.2102,
      "num_input_tokens_seen": 6126816,
      "step": 7065
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.6536986231803894,
      "learning_rate": 4.6244902725567895e-05,
      "loss": 0.2502,
      "num_input_tokens_seen": 6131184,
      "step": 7070
    },
    {
      "epoch": 3.335690711928336,
      "grad_norm": 0.5642781257629395,
      "learning_rate": 4.6239726182166024e-05,
      "loss": 0.1893,
      "num_input_tokens_seen": 6135424,
      "step": 7075
    },
    {
      "epoch": 3.338048090523338,
      "grad_norm": 0.9410204887390137,
      "learning_rate": 4.623454636333147e-05,
      "loss": 0.2017,
      "num_input_tokens_seen": 6139744,
      "step": 7080
    },
    {
      "epoch": 3.3404054691183402,
      "grad_norm": 1.0842959880828857,
      "learning_rate": 4.622936326986301e-05,
      "loss": 0.1382,
      "num_input_tokens_seen": 6143584,
      "step": 7085
    },
    {
      "epoch": 3.3427628477133426,
      "grad_norm": 0.9739652872085571,
      "learning_rate": 4.6224176902559946e-05,
      "loss": 0.263,
      "num_input_tokens_seen": 6148512,
      "step": 7090
    },
    {
      "epoch": 3.345120226308345,
      "grad_norm": 0.6961067318916321,
      "learning_rate": 4.621898726222209e-05,
      "loss": 0.1645,
      "num_input_tokens_seen": 6152928,
      "step": 7095
    },
    {
      "epoch": 3.3474776049033474,
      "grad_norm": 1.5148227214813232,
      "learning_rate": 4.6213794349649744e-05,
      "loss": 0.1794,
      "num_input_tokens_seen": 6156992,
      "step": 7100
    },
    {
      "epoch": 3.34983498349835,
      "grad_norm": 1.082588791847229,
      "learning_rate": 4.6208598165643715e-05,
      "loss": 0.2505,
      "num_input_tokens_seen": 6160880,
      "step": 7105
    },
    {
      "epoch": 3.3521923620933523,
      "grad_norm": 0.9969108700752258,
      "learning_rate": 4.620339871100533e-05,
      "loss": 0.1906,
      "num_input_tokens_seen": 6165120,
      "step": 7110
    },
    {
      "epoch": 3.3545497406883547,
      "grad_norm": 1.267396092414856,
      "learning_rate": 4.6198195986536394e-05,
      "loss": 0.195,
      "num_input_tokens_seen": 6169648,
      "step": 7115
    },
    {
      "epoch": 3.356907119283357,
      "grad_norm": 1.2175137996673584,
      "learning_rate": 4.619298999303926e-05,
      "loss": 0.2211,
      "num_input_tokens_seen": 6173552,
      "step": 7120
    },
    {
      "epoch": 3.3592644978783595,
      "grad_norm": 2.8550286293029785,
      "learning_rate": 4.618778073131673e-05,
      "loss": 0.1971,
      "num_input_tokens_seen": 6178144,
      "step": 7125
    },
    {
      "epoch": 3.361621876473362,
      "grad_norm": 0.6237748265266418,
      "learning_rate": 4.618256820217215e-05,
      "loss": 0.1572,
      "num_input_tokens_seen": 6182448,
      "step": 7130
    },
    {
      "epoch": 3.363979255068364,
      "grad_norm": 1.083996295928955,
      "learning_rate": 4.617735240640936e-05,
      "loss": 0.2453,
      "num_input_tokens_seen": 6186768,
      "step": 7135
    },
    {
      "epoch": 3.366336633663366,
      "grad_norm": 1.9300684928894043,
      "learning_rate": 4.6172133344832705e-05,
      "loss": 0.1822,
      "num_input_tokens_seen": 6191120,
      "step": 7140
    },
    {
      "epoch": 3.3686940122583686,
      "grad_norm": 1.0566307306289673,
      "learning_rate": 4.6166911018247004e-05,
      "loss": 0.184,
      "num_input_tokens_seen": 6194688,
      "step": 7145
    },
    {
      "epoch": 3.371051390853371,
      "grad_norm": 0.7359094023704529,
      "learning_rate": 4.616168542745764e-05,
      "loss": 0.1892,
      "num_input_tokens_seen": 6199520,
      "step": 7150
    },
    {
      "epoch": 3.3734087694483734,
      "grad_norm": 1.9663019180297852,
      "learning_rate": 4.6156456573270446e-05,
      "loss": 0.2442,
      "num_input_tokens_seen": 6203456,
      "step": 7155
    },
    {
      "epoch": 3.375766148043376,
      "grad_norm": 0.962123453617096,
      "learning_rate": 4.615122445649177e-05,
      "loss": 0.2038,
      "num_input_tokens_seen": 6208640,
      "step": 7160
    },
    {
      "epoch": 3.3781235266383782,
      "grad_norm": 0.9263039827346802,
      "learning_rate": 4.6145989077928486e-05,
      "loss": 0.156,
      "num_input_tokens_seen": 6213392,
      "step": 7165
    },
    {
      "epoch": 3.3804809052333806,
      "grad_norm": 0.9770099520683289,
      "learning_rate": 4.6140750438387953e-05,
      "loss": 0.1789,
      "num_input_tokens_seen": 6217952,
      "step": 7170
    },
    {
      "epoch": 3.382838283828383,
      "grad_norm": 0.8110692501068115,
      "learning_rate": 4.613550853867803e-05,
      "loss": 0.1903,
      "num_input_tokens_seen": 6222256,
      "step": 7175
    },
    {
      "epoch": 3.385195662423385,
      "grad_norm": 2.522508382797241,
      "learning_rate": 4.613026337960708e-05,
      "loss": 0.2088,
      "num_input_tokens_seen": 6226496,
      "step": 7180
    },
    {
      "epoch": 3.3875530410183874,
      "grad_norm": 0.49233734607696533,
      "learning_rate": 4.612501496198398e-05,
      "loss": 0.1054,
      "num_input_tokens_seen": 6231312,
      "step": 7185
    },
    {
      "epoch": 3.38991041961339,
      "grad_norm": 0.4605604410171509,
      "learning_rate": 4.61197632866181e-05,
      "loss": 0.179,
      "num_input_tokens_seen": 6236048,
      "step": 7190
    },
    {
      "epoch": 3.392267798208392,
      "grad_norm": 2.3613500595092773,
      "learning_rate": 4.611450835431931e-05,
      "loss": 0.2521,
      "num_input_tokens_seen": 6239840,
      "step": 7195
    },
    {
      "epoch": 3.3946251768033946,
      "grad_norm": 0.4671664237976074,
      "learning_rate": 4.6109250165898e-05,
      "loss": 0.1862,
      "num_input_tokens_seen": 6244624,
      "step": 7200
    },
    {
      "epoch": 3.3946251768033946,
      "eval_loss": 0.21195362508296967,
      "eval_runtime": 22.1747,
      "eval_samples_per_second": 42.526,
      "eval_steps_per_second": 21.285,
      "num_input_tokens_seen": 6244624,
      "step": 7200
    },
    {
      "epoch": 3.396982555398397,
      "grad_norm": 1.69088876247406,
      "learning_rate": 4.610398872216503e-05,
      "loss": 0.2423,
      "num_input_tokens_seen": 6248816,
      "step": 7205
    },
    {
      "epoch": 3.3993399339933994,
      "grad_norm": 1.0374560356140137,
      "learning_rate": 4.6098724023931796e-05,
      "loss": 0.2273,
      "num_input_tokens_seen": 6252736,
      "step": 7210
    },
    {
      "epoch": 3.401697312588402,
      "grad_norm": 0.40098220109939575,
      "learning_rate": 4.609345607201017e-05,
      "loss": 0.1395,
      "num_input_tokens_seen": 6256704,
      "step": 7215
    },
    {
      "epoch": 3.404054691183404,
      "grad_norm": 2.18868088722229,
      "learning_rate": 4.608818486721254e-05,
      "loss": 0.2376,
      "num_input_tokens_seen": 6260096,
      "step": 7220
    },
    {
      "epoch": 3.4064120697784066,
      "grad_norm": 0.8261026740074158,
      "learning_rate": 4.608291041035179e-05,
      "loss": 0.1762,
      "num_input_tokens_seen": 6264032,
      "step": 7225
    },
    {
      "epoch": 3.408769448373409,
      "grad_norm": 2.0944197177886963,
      "learning_rate": 4.607763270224132e-05,
      "loss": 0.2168,
      "num_input_tokens_seen": 6268096,
      "step": 7230
    },
    {
      "epoch": 3.411126826968411,
      "grad_norm": 0.8753376007080078,
      "learning_rate": 4.6072351743695e-05,
      "loss": 0.2253,
      "num_input_tokens_seen": 6272848,
      "step": 7235
    },
    {
      "epoch": 3.4134842055634134,
      "grad_norm": 1.0935451984405518,
      "learning_rate": 4.606706753552723e-05,
      "loss": 0.2042,
      "num_input_tokens_seen": 6277616,
      "step": 7240
    },
    {
      "epoch": 3.4158415841584158,
      "grad_norm": 1.2242608070373535,
      "learning_rate": 4.6061780078552906e-05,
      "loss": 0.1772,
      "num_input_tokens_seen": 6282768,
      "step": 7245
    },
    {
      "epoch": 3.418198962753418,
      "grad_norm": 0.6733000874519348,
      "learning_rate": 4.605648937358742e-05,
      "loss": 0.163,
      "num_input_tokens_seen": 6287680,
      "step": 7250
    },
    {
      "epoch": 3.4205563413484206,
      "grad_norm": 0.5449214577674866,
      "learning_rate": 4.605119542144665e-05,
      "loss": 0.1601,
      "num_input_tokens_seen": 6292032,
      "step": 7255
    },
    {
      "epoch": 3.422913719943423,
      "grad_norm": 0.6617842316627502,
      "learning_rate": 4.604589822294701e-05,
      "loss": 0.1759,
      "num_input_tokens_seen": 6296176,
      "step": 7260
    },
    {
      "epoch": 3.4252710985384254,
      "grad_norm": 0.7599902153015137,
      "learning_rate": 4.604059777890537e-05,
      "loss": 0.2129,
      "num_input_tokens_seen": 6300896,
      "step": 7265
    },
    {
      "epoch": 3.4276284771334278,
      "grad_norm": 0.7047543525695801,
      "learning_rate": 4.6035294090139145e-05,
      "loss": 0.1873,
      "num_input_tokens_seen": 6304912,
      "step": 7270
    },
    {
      "epoch": 3.42998585572843,
      "grad_norm": 0.8289438486099243,
      "learning_rate": 4.6029987157466226e-05,
      "loss": 0.2037,
      "num_input_tokens_seen": 6309328,
      "step": 7275
    },
    {
      "epoch": 3.432343234323432,
      "grad_norm": 0.8483309149742126,
      "learning_rate": 4.602467698170502e-05,
      "loss": 0.1861,
      "num_input_tokens_seen": 6314192,
      "step": 7280
    },
    {
      "epoch": 3.4347006129184345,
      "grad_norm": 0.40862196683883667,
      "learning_rate": 4.601936356367439e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 6318464,
      "step": 7285
    },
    {
      "epoch": 3.437057991513437,
      "grad_norm": 0.5448909401893616,
      "learning_rate": 4.601404690419377e-05,
      "loss": 0.172,
      "num_input_tokens_seen": 6322560,
      "step": 7290
    },
    {
      "epoch": 3.4394153701084393,
      "grad_norm": 0.8649393320083618,
      "learning_rate": 4.600872700408303e-05,
      "loss": 0.1962,
      "num_input_tokens_seen": 6326320,
      "step": 7295
    },
    {
      "epoch": 3.4417727487034417,
      "grad_norm": 1.297441005706787,
      "learning_rate": 4.600340386416258e-05,
      "loss": 0.2084,
      "num_input_tokens_seen": 6330192,
      "step": 7300
    },
    {
      "epoch": 3.444130127298444,
      "grad_norm": 0.49141842126846313,
      "learning_rate": 4.5998077485253296e-05,
      "loss": 0.1861,
      "num_input_tokens_seen": 6334288,
      "step": 7305
    },
    {
      "epoch": 3.4464875058934465,
      "grad_norm": 1.335660457611084,
      "learning_rate": 4.59927478681766e-05,
      "loss": 0.2091,
      "num_input_tokens_seen": 6338400,
      "step": 7310
    },
    {
      "epoch": 3.448844884488449,
      "grad_norm": 1.9414987564086914,
      "learning_rate": 4.5987415013754366e-05,
      "loss": 0.2076,
      "num_input_tokens_seen": 6342976,
      "step": 7315
    },
    {
      "epoch": 3.4512022630834513,
      "grad_norm": 0.7349819540977478,
      "learning_rate": 4.598207892280899e-05,
      "loss": 0.1978,
      "num_input_tokens_seen": 6346944,
      "step": 7320
    },
    {
      "epoch": 3.4535596416784538,
      "grad_norm": 0.475187212228775,
      "learning_rate": 4.597673959616337e-05,
      "loss": 0.2141,
      "num_input_tokens_seen": 6351424,
      "step": 7325
    },
    {
      "epoch": 3.455917020273456,
      "grad_norm": 1.1603418588638306,
      "learning_rate": 4.597139703464089e-05,
      "loss": 0.14,
      "num_input_tokens_seen": 6355392,
      "step": 7330
    },
    {
      "epoch": 3.458274398868458,
      "grad_norm": 1.3917814493179321,
      "learning_rate": 4.596605123906545e-05,
      "loss": 0.1177,
      "num_input_tokens_seen": 6359040,
      "step": 7335
    },
    {
      "epoch": 3.4606317774634605,
      "grad_norm": 0.7596669793128967,
      "learning_rate": 4.596070221026143e-05,
      "loss": 0.2065,
      "num_input_tokens_seen": 6363280,
      "step": 7340
    },
    {
      "epoch": 3.462989156058463,
      "grad_norm": 1.0628666877746582,
      "learning_rate": 4.595534994905372e-05,
      "loss": 0.1843,
      "num_input_tokens_seen": 6368224,
      "step": 7345
    },
    {
      "epoch": 3.4653465346534653,
      "grad_norm": 0.796264111995697,
      "learning_rate": 4.594999445626771e-05,
      "loss": 0.1725,
      "num_input_tokens_seen": 6372240,
      "step": 7350
    },
    {
      "epoch": 3.4677039132484677,
      "grad_norm": 1.639210820198059,
      "learning_rate": 4.5944635732729276e-05,
      "loss": 0.1935,
      "num_input_tokens_seen": 6376352,
      "step": 7355
    },
    {
      "epoch": 3.47006129184347,
      "grad_norm": 0.8009605407714844,
      "learning_rate": 4.5939273779264804e-05,
      "loss": 0.1782,
      "num_input_tokens_seen": 6380864,
      "step": 7360
    },
    {
      "epoch": 3.4724186704384725,
      "grad_norm": 0.8490871787071228,
      "learning_rate": 4.593390859670118e-05,
      "loss": 0.138,
      "num_input_tokens_seen": 6385984,
      "step": 7365
    },
    {
      "epoch": 3.474776049033475,
      "grad_norm": 0.5506072044372559,
      "learning_rate": 4.5928540185865776e-05,
      "loss": 0.1523,
      "num_input_tokens_seen": 6390864,
      "step": 7370
    },
    {
      "epoch": 3.4771334276284773,
      "grad_norm": 1.4297171831130981,
      "learning_rate": 4.592316854758648e-05,
      "loss": 0.1995,
      "num_input_tokens_seen": 6394768,
      "step": 7375
    },
    {
      "epoch": 3.4794908062234793,
      "grad_norm": 0.6858320236206055,
      "learning_rate": 4.5917793682691646e-05,
      "loss": 0.1979,
      "num_input_tokens_seen": 6399344,
      "step": 7380
    },
    {
      "epoch": 3.4818481848184817,
      "grad_norm": 2.4059786796569824,
      "learning_rate": 4.5912415592010164e-05,
      "loss": 0.2093,
      "num_input_tokens_seen": 6403904,
      "step": 7385
    },
    {
      "epoch": 3.484205563413484,
      "grad_norm": 0.9392485618591309,
      "learning_rate": 4.5907034276371386e-05,
      "loss": 0.2029,
      "num_input_tokens_seen": 6408240,
      "step": 7390
    },
    {
      "epoch": 3.4865629420084865,
      "grad_norm": 0.9676254391670227,
      "learning_rate": 4.5901649736605196e-05,
      "loss": 0.1308,
      "num_input_tokens_seen": 6412352,
      "step": 7395
    },
    {
      "epoch": 3.488920320603489,
      "grad_norm": 0.9833754301071167,
      "learning_rate": 4.589626197354195e-05,
      "loss": 0.2273,
      "num_input_tokens_seen": 6416176,
      "step": 7400
    },
    {
      "epoch": 3.488920320603489,
      "eval_loss": 0.21399004757404327,
      "eval_runtime": 22.1722,
      "eval_samples_per_second": 42.531,
      "eval_steps_per_second": 21.288,
      "num_input_tokens_seen": 6416176,
      "step": 7400
    },
    {
      "epoch": 3.4912776991984913,
      "grad_norm": 0.7872740030288696,
      "learning_rate": 4.5890870988012504e-05,
      "loss": 0.2494,
      "num_input_tokens_seen": 6420416,
      "step": 7405
    },
    {
      "epoch": 3.4936350777934937,
      "grad_norm": 0.6351850628852844,
      "learning_rate": 4.5885476780848226e-05,
      "loss": 0.2017,
      "num_input_tokens_seen": 6424592,
      "step": 7410
    },
    {
      "epoch": 3.495992456388496,
      "grad_norm": 1.3293472528457642,
      "learning_rate": 4.5880079352880964e-05,
      "loss": 0.1674,
      "num_input_tokens_seen": 6429328,
      "step": 7415
    },
    {
      "epoch": 3.4983498349834985,
      "grad_norm": 0.8355581760406494,
      "learning_rate": 4.5874678704943065e-05,
      "loss": 0.1673,
      "num_input_tokens_seen": 6433776,
      "step": 7420
    },
    {
      "epoch": 3.500707213578501,
      "grad_norm": 1.0496388673782349,
      "learning_rate": 4.5869274837867394e-05,
      "loss": 0.2672,
      "num_input_tokens_seen": 6437840,
      "step": 7425
    },
    {
      "epoch": 3.5030645921735033,
      "grad_norm": 0.807055652141571,
      "learning_rate": 4.5863867752487275e-05,
      "loss": 0.174,
      "num_input_tokens_seen": 6441744,
      "step": 7430
    },
    {
      "epoch": 3.5054219707685053,
      "grad_norm": 0.5731155276298523,
      "learning_rate": 4.5858457449636554e-05,
      "loss": 0.1191,
      "num_input_tokens_seen": 6445440,
      "step": 7435
    },
    {
      "epoch": 3.5077793493635077,
      "grad_norm": 0.5158705115318298,
      "learning_rate": 4.5853043930149574e-05,
      "loss": 0.1529,
      "num_input_tokens_seen": 6449888,
      "step": 7440
    },
    {
      "epoch": 3.51013672795851,
      "grad_norm": 2.0033671855926514,
      "learning_rate": 4.584762719486117e-05,
      "loss": 0.1725,
      "num_input_tokens_seen": 6453808,
      "step": 7445
    },
    {
      "epoch": 3.5124941065535125,
      "grad_norm": 0.8149189352989197,
      "learning_rate": 4.584220724460665e-05,
      "loss": 0.253,
      "num_input_tokens_seen": 6457696,
      "step": 7450
    },
    {
      "epoch": 3.514851485148515,
      "grad_norm": 1.1353051662445068,
      "learning_rate": 4.5836784080221865e-05,
      "loss": 0.1873,
      "num_input_tokens_seen": 6462960,
      "step": 7455
    },
    {
      "epoch": 3.5172088637435173,
      "grad_norm": 1.0420113801956177,
      "learning_rate": 4.583135770254312e-05,
      "loss": 0.1526,
      "num_input_tokens_seen": 6467536,
      "step": 7460
    },
    {
      "epoch": 3.5195662423385197,
      "grad_norm": 1.929486632347107,
      "learning_rate": 4.5825928112407236e-05,
      "loss": 0.177,
      "num_input_tokens_seen": 6471616,
      "step": 7465
    },
    {
      "epoch": 3.521923620933522,
      "grad_norm": 1.2955513000488281,
      "learning_rate": 4.582049531065152e-05,
      "loss": 0.2244,
      "num_input_tokens_seen": 6475824,
      "step": 7470
    },
    {
      "epoch": 3.524280999528524,
      "grad_norm": 0.5153872966766357,
      "learning_rate": 4.5815059298113783e-05,
      "loss": 0.1943,
      "num_input_tokens_seen": 6479808,
      "step": 7475
    },
    {
      "epoch": 3.5266383781235264,
      "grad_norm": 1.8609962463378906,
      "learning_rate": 4.580962007563232e-05,
      "loss": 0.2371,
      "num_input_tokens_seen": 6484048,
      "step": 7480
    },
    {
      "epoch": 3.528995756718529,
      "grad_norm": 0.7201302647590637,
      "learning_rate": 4.5804177644045935e-05,
      "loss": 0.1981,
      "num_input_tokens_seen": 6488784,
      "step": 7485
    },
    {
      "epoch": 3.5313531353135312,
      "grad_norm": 0.9637472033500671,
      "learning_rate": 4.579873200419391e-05,
      "loss": 0.2058,
      "num_input_tokens_seen": 6493264,
      "step": 7490
    },
    {
      "epoch": 3.5337105139085336,
      "grad_norm": 0.48033902049064636,
      "learning_rate": 4.5793283156916046e-05,
      "loss": 0.2129,
      "num_input_tokens_seen": 6498368,
      "step": 7495
    },
    {
      "epoch": 3.536067892503536,
      "grad_norm": 0.6490579843521118,
      "learning_rate": 4.578783110305261e-05,
      "loss": 0.1925,
      "num_input_tokens_seen": 6502416,
      "step": 7500
    },
    {
      "epoch": 3.5384252710985384,
      "grad_norm": 0.6331693530082703,
      "learning_rate": 4.578237584344438e-05,
      "loss": 0.241,
      "num_input_tokens_seen": 6507264,
      "step": 7505
    },
    {
      "epoch": 3.540782649693541,
      "grad_norm": 0.972976803779602,
      "learning_rate": 4.577691737893263e-05,
      "loss": 0.2588,
      "num_input_tokens_seen": 6511280,
      "step": 7510
    },
    {
      "epoch": 3.5431400282885432,
      "grad_norm": 0.44402143359184265,
      "learning_rate": 4.577145571035912e-05,
      "loss": 0.1374,
      "num_input_tokens_seen": 6515312,
      "step": 7515
    },
    {
      "epoch": 3.5454974068835456,
      "grad_norm": 1.1518335342407227,
      "learning_rate": 4.576599083856611e-05,
      "loss": 0.1482,
      "num_input_tokens_seen": 6519312,
      "step": 7520
    },
    {
      "epoch": 3.547854785478548,
      "grad_norm": 2.395240545272827,
      "learning_rate": 4.576052276439635e-05,
      "loss": 0.2099,
      "num_input_tokens_seen": 6524416,
      "step": 7525
    },
    {
      "epoch": 3.5502121640735504,
      "grad_norm": 1.1083687543869019,
      "learning_rate": 4.575505148869308e-05,
      "loss": 0.1787,
      "num_input_tokens_seen": 6528720,
      "step": 7530
    },
    {
      "epoch": 3.5525695426685524,
      "grad_norm": 0.35240232944488525,
      "learning_rate": 4.574957701230006e-05,
      "loss": 0.1709,
      "num_input_tokens_seen": 6533296,
      "step": 7535
    },
    {
      "epoch": 3.554926921263555,
      "grad_norm": 1.4332597255706787,
      "learning_rate": 4.57440993360615e-05,
      "loss": 0.2678,
      "num_input_tokens_seen": 6537792,
      "step": 7540
    },
    {
      "epoch": 3.557284299858557,
      "grad_norm": 0.5777317881584167,
      "learning_rate": 4.5738618460822134e-05,
      "loss": 0.1929,
      "num_input_tokens_seen": 6542208,
      "step": 7545
    },
    {
      "epoch": 3.5596416784535596,
      "grad_norm": 0.8312135338783264,
      "learning_rate": 4.573313438742719e-05,
      "loss": 0.2039,
      "num_input_tokens_seen": 6546496,
      "step": 7550
    },
    {
      "epoch": 3.561999057048562,
      "grad_norm": 1.6120115518569946,
      "learning_rate": 4.5727647116722374e-05,
      "loss": 0.1538,
      "num_input_tokens_seen": 6550288,
      "step": 7555
    },
    {
      "epoch": 3.5643564356435644,
      "grad_norm": 1.1120628118515015,
      "learning_rate": 4.5722156649553884e-05,
      "loss": 0.1738,
      "num_input_tokens_seen": 6554416,
      "step": 7560
    },
    {
      "epoch": 3.566713814238567,
      "grad_norm": 0.8085107803344727,
      "learning_rate": 4.571666298676843e-05,
      "loss": 0.2203,
      "num_input_tokens_seen": 6559232,
      "step": 7565
    },
    {
      "epoch": 3.569071192833569,
      "grad_norm": 0.7729193568229675,
      "learning_rate": 4.571116612921321e-05,
      "loss": 0.1778,
      "num_input_tokens_seen": 6563008,
      "step": 7570
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 0.9996755719184875,
      "learning_rate": 4.57056660777359e-05,
      "loss": 0.2394,
      "num_input_tokens_seen": 6566848,
      "step": 7575
    },
    {
      "epoch": 3.5737859500235736,
      "grad_norm": 1.8785228729248047,
      "learning_rate": 4.5700162833184666e-05,
      "loss": 0.2296,
      "num_input_tokens_seen": 6571664,
      "step": 7580
    },
    {
      "epoch": 3.576143328618576,
      "grad_norm": 1.267157793045044,
      "learning_rate": 4.5694656396408195e-05,
      "loss": 0.1582,
      "num_input_tokens_seen": 6575616,
      "step": 7585
    },
    {
      "epoch": 3.5785007072135784,
      "grad_norm": 0.5381425619125366,
      "learning_rate": 4.5689146768255646e-05,
      "loss": 0.1569,
      "num_input_tokens_seen": 6579280,
      "step": 7590
    },
    {
      "epoch": 3.580858085808581,
      "grad_norm": 1.0524078607559204,
      "learning_rate": 4.568363394957667e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 6583472,
      "step": 7595
    },
    {
      "epoch": 3.583215464403583,
      "grad_norm": 0.8667198419570923,
      "learning_rate": 4.567811794122141e-05,
      "loss": 0.2755,
      "num_input_tokens_seen": 6587616,
      "step": 7600
    },
    {
      "epoch": 3.583215464403583,
      "eval_loss": 0.20970885455608368,
      "eval_runtime": 22.1259,
      "eval_samples_per_second": 42.62,
      "eval_steps_per_second": 21.332,
      "num_input_tokens_seen": 6587616,
      "step": 7600
    },
    {
      "epoch": 3.5855728429985856,
      "grad_norm": 0.6716144680976868,
      "learning_rate": 4.56725987440405e-05,
      "loss": 0.1661,
      "num_input_tokens_seen": 6591680,
      "step": 7605
    },
    {
      "epoch": 3.587930221593588,
      "grad_norm": 0.36621859669685364,
      "learning_rate": 4.566707635888508e-05,
      "loss": 0.1585,
      "num_input_tokens_seen": 6595344,
      "step": 7610
    },
    {
      "epoch": 3.5902876001885904,
      "grad_norm": 0.8901990652084351,
      "learning_rate": 4.566155078660677e-05,
      "loss": 0.1502,
      "num_input_tokens_seen": 6598832,
      "step": 7615
    },
    {
      "epoch": 3.592644978783593,
      "grad_norm": 0.4199020564556122,
      "learning_rate": 4.565602202805768e-05,
      "loss": 0.155,
      "num_input_tokens_seen": 6603056,
      "step": 7620
    },
    {
      "epoch": 3.595002357378595,
      "grad_norm": 1.6395764350891113,
      "learning_rate": 4.56504900840904e-05,
      "loss": 0.1999,
      "num_input_tokens_seen": 6607040,
      "step": 7625
    },
    {
      "epoch": 3.5973597359735976,
      "grad_norm": 1.6091945171356201,
      "learning_rate": 4.564495495555805e-05,
      "loss": 0.1448,
      "num_input_tokens_seen": 6611840,
      "step": 7630
    },
    {
      "epoch": 3.5997171145685996,
      "grad_norm": 0.6012479662895203,
      "learning_rate": 4.5639416643314204e-05,
      "loss": 0.1861,
      "num_input_tokens_seen": 6616608,
      "step": 7635
    },
    {
      "epoch": 3.602074493163602,
      "grad_norm": 0.9418091177940369,
      "learning_rate": 4.5633875148212946e-05,
      "loss": 0.1381,
      "num_input_tokens_seen": 6620624,
      "step": 7640
    },
    {
      "epoch": 3.6044318717586044,
      "grad_norm": 0.9522907733917236,
      "learning_rate": 4.562833047110883e-05,
      "loss": 0.1743,
      "num_input_tokens_seen": 6624656,
      "step": 7645
    },
    {
      "epoch": 3.6067892503536068,
      "grad_norm": 0.7535807490348816,
      "learning_rate": 4.5622782612856923e-05,
      "loss": 0.2492,
      "num_input_tokens_seen": 6628976,
      "step": 7650
    },
    {
      "epoch": 3.609146628948609,
      "grad_norm": 0.7164435982704163,
      "learning_rate": 4.561723157431278e-05,
      "loss": 0.2161,
      "num_input_tokens_seen": 6633744,
      "step": 7655
    },
    {
      "epoch": 3.6115040075436116,
      "grad_norm": 1.4436842203140259,
      "learning_rate": 4.5611677356332435e-05,
      "loss": 0.1907,
      "num_input_tokens_seen": 6638256,
      "step": 7660
    },
    {
      "epoch": 3.613861386138614,
      "grad_norm": 0.43960198760032654,
      "learning_rate": 4.560611995977242e-05,
      "loss": 0.2372,
      "num_input_tokens_seen": 6643408,
      "step": 7665
    },
    {
      "epoch": 3.6162187647336164,
      "grad_norm": 1.347912311553955,
      "learning_rate": 4.560055938548975e-05,
      "loss": 0.2498,
      "num_input_tokens_seen": 6647744,
      "step": 7670
    },
    {
      "epoch": 3.6185761433286183,
      "grad_norm": 1.0439788103103638,
      "learning_rate": 4.5594995634341944e-05,
      "loss": 0.1478,
      "num_input_tokens_seen": 6652448,
      "step": 7675
    },
    {
      "epoch": 3.6209335219236207,
      "grad_norm": 1.1942938566207886,
      "learning_rate": 4.5589428707187e-05,
      "loss": 0.1712,
      "num_input_tokens_seen": 6656576,
      "step": 7680
    },
    {
      "epoch": 3.623290900518623,
      "grad_norm": 0.9535844922065735,
      "learning_rate": 4.55838586048834e-05,
      "loss": 0.1974,
      "num_input_tokens_seen": 6661088,
      "step": 7685
    },
    {
      "epoch": 3.6256482791136255,
      "grad_norm": 0.6580312848091125,
      "learning_rate": 4.557828532829013e-05,
      "loss": 0.2638,
      "num_input_tokens_seen": 6665008,
      "step": 7690
    },
    {
      "epoch": 3.628005657708628,
      "grad_norm": 1.00593101978302,
      "learning_rate": 4.557270887826667e-05,
      "loss": 0.1641,
      "num_input_tokens_seen": 6668880,
      "step": 7695
    },
    {
      "epoch": 3.6303630363036303,
      "grad_norm": 0.5226604342460632,
      "learning_rate": 4.556712925567296e-05,
      "loss": 0.1763,
      "num_input_tokens_seen": 6673984,
      "step": 7700
    },
    {
      "epoch": 3.6327204148986327,
      "grad_norm": 0.5475869178771973,
      "learning_rate": 4.5561546461369454e-05,
      "loss": 0.189,
      "num_input_tokens_seen": 6678080,
      "step": 7705
    },
    {
      "epoch": 3.635077793493635,
      "grad_norm": 1.7706961631774902,
      "learning_rate": 4.55559604962171e-05,
      "loss": 0.2006,
      "num_input_tokens_seen": 6682416,
      "step": 7710
    },
    {
      "epoch": 3.6374351720886375,
      "grad_norm": 0.6134690642356873,
      "learning_rate": 4.55503713610773e-05,
      "loss": 0.1575,
      "num_input_tokens_seen": 6687072,
      "step": 7715
    },
    {
      "epoch": 3.63979255068364,
      "grad_norm": 0.8684471249580383,
      "learning_rate": 4.5544779056812e-05,
      "loss": 0.1725,
      "num_input_tokens_seen": 6691824,
      "step": 7720
    },
    {
      "epoch": 3.6421499292786423,
      "grad_norm": 0.45054304599761963,
      "learning_rate": 4.553918358428358e-05,
      "loss": 0.1472,
      "num_input_tokens_seen": 6696128,
      "step": 7725
    },
    {
      "epoch": 3.6445073078736447,
      "grad_norm": 0.3194052278995514,
      "learning_rate": 4.553358494435494e-05,
      "loss": 0.2501,
      "num_input_tokens_seen": 6700192,
      "step": 7730
    },
    {
      "epoch": 3.6468646864686467,
      "grad_norm": 1.2225801944732666,
      "learning_rate": 4.5527983137889464e-05,
      "loss": 0.1685,
      "num_input_tokens_seen": 6704064,
      "step": 7735
    },
    {
      "epoch": 3.649222065063649,
      "grad_norm": 0.9763994216918945,
      "learning_rate": 4.5522378165751015e-05,
      "loss": 0.191,
      "num_input_tokens_seen": 6708944,
      "step": 7740
    },
    {
      "epoch": 3.6515794436586515,
      "grad_norm": 0.7968204617500305,
      "learning_rate": 4.5516770028803954e-05,
      "loss": 0.1615,
      "num_input_tokens_seen": 6713184,
      "step": 7745
    },
    {
      "epoch": 3.653936822253654,
      "grad_norm": 0.8931065797805786,
      "learning_rate": 4.5511158727913116e-05,
      "loss": 0.2218,
      "num_input_tokens_seen": 6717424,
      "step": 7750
    },
    {
      "epoch": 3.6562942008486563,
      "grad_norm": 1.6325647830963135,
      "learning_rate": 4.5505544263943856e-05,
      "loss": 0.1918,
      "num_input_tokens_seen": 6720960,
      "step": 7755
    },
    {
      "epoch": 3.6586515794436587,
      "grad_norm": 0.9073262810707092,
      "learning_rate": 4.549992663776197e-05,
      "loss": 0.2117,
      "num_input_tokens_seen": 6724800,
      "step": 7760
    },
    {
      "epoch": 3.661008958038661,
      "grad_norm": 0.7387428283691406,
      "learning_rate": 4.5494305850233786e-05,
      "loss": 0.19,
      "num_input_tokens_seen": 6728928,
      "step": 7765
    },
    {
      "epoch": 3.6633663366336635,
      "grad_norm": 0.389989972114563,
      "learning_rate": 4.5488681902226094e-05,
      "loss": 0.162,
      "num_input_tokens_seen": 6732736,
      "step": 7770
    },
    {
      "epoch": 3.6657237152286655,
      "grad_norm": 0.5857774615287781,
      "learning_rate": 4.5483054794606174e-05,
      "loss": 0.2681,
      "num_input_tokens_seen": 6737328,
      "step": 7775
    },
    {
      "epoch": 3.668081093823668,
      "grad_norm": 0.5380121469497681,
      "learning_rate": 4.547742452824179e-05,
      "loss": 0.1563,
      "num_input_tokens_seen": 6741760,
      "step": 7780
    },
    {
      "epoch": 3.6704384724186703,
      "grad_norm": 0.8536801338195801,
      "learning_rate": 4.5471791104001215e-05,
      "loss": 0.1946,
      "num_input_tokens_seen": 6745952,
      "step": 7785
    },
    {
      "epoch": 3.6727958510136727,
      "grad_norm": 0.8310866951942444,
      "learning_rate": 4.546615452275319e-05,
      "loss": 0.1933,
      "num_input_tokens_seen": 6750096,
      "step": 7790
    },
    {
      "epoch": 3.675153229608675,
      "grad_norm": 0.5218557715415955,
      "learning_rate": 4.5460514785366944e-05,
      "loss": 0.2022,
      "num_input_tokens_seen": 6754704,
      "step": 7795
    },
    {
      "epoch": 3.6775106082036775,
      "grad_norm": 0.513979434967041,
      "learning_rate": 4.545487189271219e-05,
      "loss": 0.2027,
      "num_input_tokens_seen": 6759696,
      "step": 7800
    },
    {
      "epoch": 3.6775106082036775,
      "eval_loss": 0.21043811738491058,
      "eval_runtime": 22.1739,
      "eval_samples_per_second": 42.527,
      "eval_steps_per_second": 21.286,
      "num_input_tokens_seen": 6759696,
      "step": 7800
    },
    {
      "epoch": 3.67986798679868,
      "grad_norm": 0.6438459157943726,
      "learning_rate": 4.544922584565914e-05,
      "loss": 0.2876,
      "num_input_tokens_seen": 6763664,
      "step": 7805
    },
    {
      "epoch": 3.6822253653936823,
      "grad_norm": 1.5222790241241455,
      "learning_rate": 4.544357664507848e-05,
      "loss": 0.2258,
      "num_input_tokens_seen": 6767296,
      "step": 7810
    },
    {
      "epoch": 3.6845827439886847,
      "grad_norm": 1.0104933977127075,
      "learning_rate": 4.54379242918414e-05,
      "loss": 0.1627,
      "num_input_tokens_seen": 6771728,
      "step": 7815
    },
    {
      "epoch": 3.686940122583687,
      "grad_norm": 1.6066476106643677,
      "learning_rate": 4.543226878681955e-05,
      "loss": 0.2216,
      "num_input_tokens_seen": 6775968,
      "step": 7820
    },
    {
      "epoch": 3.6892975011786895,
      "grad_norm": 0.9402990341186523,
      "learning_rate": 4.5426610130885087e-05,
      "loss": 0.1235,
      "num_input_tokens_seen": 6780384,
      "step": 7825
    },
    {
      "epoch": 3.691654879773692,
      "grad_norm": 0.58922278881073,
      "learning_rate": 4.542094832491064e-05,
      "loss": 0.1561,
      "num_input_tokens_seen": 6784592,
      "step": 7830
    },
    {
      "epoch": 3.694012258368694,
      "grad_norm": 0.8559759259223938,
      "learning_rate": 4.541528336976934e-05,
      "loss": 0.1881,
      "num_input_tokens_seen": 6788592,
      "step": 7835
    },
    {
      "epoch": 3.6963696369636962,
      "grad_norm": 1.6552640199661255,
      "learning_rate": 4.540961526633479e-05,
      "loss": 0.2359,
      "num_input_tokens_seen": 6792384,
      "step": 7840
    },
    {
      "epoch": 3.6987270155586986,
      "grad_norm": 0.45177409052848816,
      "learning_rate": 4.540394401548108e-05,
      "loss": 0.245,
      "num_input_tokens_seen": 6796480,
      "step": 7845
    },
    {
      "epoch": 3.701084394153701,
      "grad_norm": 1.568807601928711,
      "learning_rate": 4.539826961808279e-05,
      "loss": 0.1748,
      "num_input_tokens_seen": 6800400,
      "step": 7850
    },
    {
      "epoch": 3.7034417727487035,
      "grad_norm": 0.5026620030403137,
      "learning_rate": 4.5392592075014994e-05,
      "loss": 0.2327,
      "num_input_tokens_seen": 6804896,
      "step": 7855
    },
    {
      "epoch": 3.705799151343706,
      "grad_norm": 1.2100327014923096,
      "learning_rate": 4.538691138715322e-05,
      "loss": 0.1978,
      "num_input_tokens_seen": 6809216,
      "step": 7860
    },
    {
      "epoch": 3.7081565299387083,
      "grad_norm": 1.219085931777954,
      "learning_rate": 4.5381227555373516e-05,
      "loss": 0.2346,
      "num_input_tokens_seen": 6813904,
      "step": 7865
    },
    {
      "epoch": 3.7105139085337107,
      "grad_norm": 0.645618200302124,
      "learning_rate": 4.537554058055239e-05,
      "loss": 0.1647,
      "num_input_tokens_seen": 6818128,
      "step": 7870
    },
    {
      "epoch": 3.7128712871287126,
      "grad_norm": 0.7006823420524597,
      "learning_rate": 4.5369850463566865e-05,
      "loss": 0.2617,
      "num_input_tokens_seen": 6823136,
      "step": 7875
    },
    {
      "epoch": 3.715228665723715,
      "grad_norm": 0.5351584553718567,
      "learning_rate": 4.5364157205294404e-05,
      "loss": 0.2246,
      "num_input_tokens_seen": 6827760,
      "step": 7880
    },
    {
      "epoch": 3.7175860443187174,
      "grad_norm": 1.865959644317627,
      "learning_rate": 4.5358460806612996e-05,
      "loss": 0.2094,
      "num_input_tokens_seen": 6832144,
      "step": 7885
    },
    {
      "epoch": 3.71994342291372,
      "grad_norm": 1.6844885349273682,
      "learning_rate": 4.535276126840109e-05,
      "loss": 0.1925,
      "num_input_tokens_seen": 6835760,
      "step": 7890
    },
    {
      "epoch": 3.7223008015087222,
      "grad_norm": 1.223912000656128,
      "learning_rate": 4.5347058591537626e-05,
      "loss": 0.1459,
      "num_input_tokens_seen": 6840128,
      "step": 7895
    },
    {
      "epoch": 3.7246581801037246,
      "grad_norm": 0.7349305748939514,
      "learning_rate": 4.534135277690203e-05,
      "loss": 0.1595,
      "num_input_tokens_seen": 6843920,
      "step": 7900
    },
    {
      "epoch": 3.727015558698727,
      "grad_norm": 1.0810683965682983,
      "learning_rate": 4.533564382537421e-05,
      "loss": 0.2369,
      "num_input_tokens_seen": 6848304,
      "step": 7905
    },
    {
      "epoch": 3.7293729372937294,
      "grad_norm": 1.2927629947662354,
      "learning_rate": 4.532993173783456e-05,
      "loss": 0.2127,
      "num_input_tokens_seen": 6852544,
      "step": 7910
    },
    {
      "epoch": 3.731730315888732,
      "grad_norm": 0.647264838218689,
      "learning_rate": 4.5324216515163954e-05,
      "loss": 0.1968,
      "num_input_tokens_seen": 6857696,
      "step": 7915
    },
    {
      "epoch": 3.7340876944837342,
      "grad_norm": 0.8690697550773621,
      "learning_rate": 4.531849815824375e-05,
      "loss": 0.1755,
      "num_input_tokens_seen": 6862496,
      "step": 7920
    },
    {
      "epoch": 3.7364450730787366,
      "grad_norm": 2.0014917850494385,
      "learning_rate": 4.5312776667955795e-05,
      "loss": 0.1855,
      "num_input_tokens_seen": 6867056,
      "step": 7925
    },
    {
      "epoch": 3.738802451673739,
      "grad_norm": 0.955735981464386,
      "learning_rate": 4.5307052045182405e-05,
      "loss": 0.1857,
      "num_input_tokens_seen": 6872144,
      "step": 7930
    },
    {
      "epoch": 3.741159830268741,
      "grad_norm": 0.7357672452926636,
      "learning_rate": 4.53013242908064e-05,
      "loss": 0.2102,
      "num_input_tokens_seen": 6876464,
      "step": 7935
    },
    {
      "epoch": 3.7435172088637434,
      "grad_norm": 0.610407292842865,
      "learning_rate": 4.529559340571107e-05,
      "loss": 0.2203,
      "num_input_tokens_seen": 6882096,
      "step": 7940
    },
    {
      "epoch": 3.745874587458746,
      "grad_norm": 2.2754745483398438,
      "learning_rate": 4.528985939078018e-05,
      "loss": 0.2374,
      "num_input_tokens_seen": 6886048,
      "step": 7945
    },
    {
      "epoch": 3.748231966053748,
      "grad_norm": 1.1751790046691895,
      "learning_rate": 4.5284122246898e-05,
      "loss": 0.1862,
      "num_input_tokens_seen": 6890976,
      "step": 7950
    },
    {
      "epoch": 3.7505893446487506,
      "grad_norm": 1.3189547061920166,
      "learning_rate": 4.527838197494926e-05,
      "loss": 0.1497,
      "num_input_tokens_seen": 6895440,
      "step": 7955
    },
    {
      "epoch": 3.752946723243753,
      "grad_norm": 1.0813920497894287,
      "learning_rate": 4.527263857581918e-05,
      "loss": 0.1481,
      "num_input_tokens_seen": 6899824,
      "step": 7960
    },
    {
      "epoch": 3.7553041018387554,
      "grad_norm": 0.6903287172317505,
      "learning_rate": 4.526689205039347e-05,
      "loss": 0.2186,
      "num_input_tokens_seen": 6903632,
      "step": 7965
    },
    {
      "epoch": 3.757661480433758,
      "grad_norm": 0.42760804295539856,
      "learning_rate": 4.5261142399558324e-05,
      "loss": 0.1679,
      "num_input_tokens_seen": 6907648,
      "step": 7970
    },
    {
      "epoch": 3.7600188590287598,
      "grad_norm": 0.6370038390159607,
      "learning_rate": 4.525538962420041e-05,
      "loss": 0.1426,
      "num_input_tokens_seen": 6912080,
      "step": 7975
    },
    {
      "epoch": 3.762376237623762,
      "grad_norm": 1.1530015468597412,
      "learning_rate": 4.524963372520685e-05,
      "loss": 0.1783,
      "num_input_tokens_seen": 6916400,
      "step": 7980
    },
    {
      "epoch": 3.7647336162187646,
      "grad_norm": 0.91339111328125,
      "learning_rate": 4.524387470346531e-05,
      "loss": 0.251,
      "num_input_tokens_seen": 6920064,
      "step": 7985
    },
    {
      "epoch": 3.767090994813767,
      "grad_norm": 0.9691147208213806,
      "learning_rate": 4.5238112559863885e-05,
      "loss": 0.1339,
      "num_input_tokens_seen": 6924256,
      "step": 7990
    },
    {
      "epoch": 3.7694483734087694,
      "grad_norm": 0.6004630923271179,
      "learning_rate": 4.5232347295291175e-05,
      "loss": 0.2481,
      "num_input_tokens_seen": 6928624,
      "step": 7995
    },
    {
      "epoch": 3.7718057520037718,
      "grad_norm": 0.6028829216957092,
      "learning_rate": 4.522657891063626e-05,
      "loss": 0.2504,
      "num_input_tokens_seen": 6932384,
      "step": 8000
    },
    {
      "epoch": 3.7718057520037718,
      "eval_loss": 0.21393226087093353,
      "eval_runtime": 22.1712,
      "eval_samples_per_second": 42.533,
      "eval_steps_per_second": 21.289,
      "num_input_tokens_seen": 6932384,
      "step": 8000
    },
    {
      "epoch": 3.774163130598774,
      "grad_norm": 0.6598201990127563,
      "learning_rate": 4.52208074067887e-05,
      "loss": 0.2235,
      "num_input_tokens_seen": 6937344,
      "step": 8005
    },
    {
      "epoch": 3.7765205091937766,
      "grad_norm": 1.057503342628479,
      "learning_rate": 4.5215032784638516e-05,
      "loss": 0.2028,
      "num_input_tokens_seen": 6942160,
      "step": 8010
    },
    {
      "epoch": 3.778877887788779,
      "grad_norm": 0.5480987429618835,
      "learning_rate": 4.5209255045076245e-05,
      "loss": 0.1767,
      "num_input_tokens_seen": 6946544,
      "step": 8015
    },
    {
      "epoch": 3.7812352663837814,
      "grad_norm": 1.2081804275512695,
      "learning_rate": 4.5203474188992875e-05,
      "loss": 0.1591,
      "num_input_tokens_seen": 6951264,
      "step": 8020
    },
    {
      "epoch": 3.783592644978784,
      "grad_norm": 1.6549479961395264,
      "learning_rate": 4.51976902172799e-05,
      "loss": 0.1741,
      "num_input_tokens_seen": 6955280,
      "step": 8025
    },
    {
      "epoch": 3.785950023573786,
      "grad_norm": 0.671352207660675,
      "learning_rate": 4.519190313082927e-05,
      "loss": 0.1817,
      "num_input_tokens_seen": 6960096,
      "step": 8030
    },
    {
      "epoch": 3.7883074021687886,
      "grad_norm": 0.5247141718864441,
      "learning_rate": 4.518611293053343e-05,
      "loss": 0.136,
      "num_input_tokens_seen": 6963936,
      "step": 8035
    },
    {
      "epoch": 3.7906647807637905,
      "grad_norm": 1.3971874713897705,
      "learning_rate": 4.51803196172853e-05,
      "loss": 0.198,
      "num_input_tokens_seen": 6968640,
      "step": 8040
    },
    {
      "epoch": 3.793022159358793,
      "grad_norm": 0.7247575521469116,
      "learning_rate": 4.517452319197828e-05,
      "loss": 0.2304,
      "num_input_tokens_seen": 6972752,
      "step": 8045
    },
    {
      "epoch": 3.7953795379537953,
      "grad_norm": 0.680764377117157,
      "learning_rate": 4.5168723655506265e-05,
      "loss": 0.1308,
      "num_input_tokens_seen": 6976576,
      "step": 8050
    },
    {
      "epoch": 3.7977369165487977,
      "grad_norm": 0.8835184574127197,
      "learning_rate": 4.51629210087636e-05,
      "loss": 0.2453,
      "num_input_tokens_seen": 6981408,
      "step": 8055
    },
    {
      "epoch": 3.8000942951438,
      "grad_norm": 0.5972124338150024,
      "learning_rate": 4.515711525264513e-05,
      "loss": 0.1896,
      "num_input_tokens_seen": 6985840,
      "step": 8060
    },
    {
      "epoch": 3.8024516737388026,
      "grad_norm": 0.6029956340789795,
      "learning_rate": 4.5151306388046175e-05,
      "loss": 0.1273,
      "num_input_tokens_seen": 6989632,
      "step": 8065
    },
    {
      "epoch": 3.804809052333805,
      "grad_norm": 0.830504834651947,
      "learning_rate": 4.514549441586255e-05,
      "loss": 0.1436,
      "num_input_tokens_seen": 6993312,
      "step": 8070
    },
    {
      "epoch": 3.807166430928807,
      "grad_norm": 1.1361472606658936,
      "learning_rate": 4.513967933699051e-05,
      "loss": 0.2077,
      "num_input_tokens_seen": 6997824,
      "step": 8075
    },
    {
      "epoch": 3.8095238095238093,
      "grad_norm": 0.8150410056114197,
      "learning_rate": 4.513386115232684e-05,
      "loss": 0.1984,
      "num_input_tokens_seen": 7002272,
      "step": 8080
    },
    {
      "epoch": 3.8118811881188117,
      "grad_norm": 0.8571754097938538,
      "learning_rate": 4.5128039862768745e-05,
      "loss": 0.1415,
      "num_input_tokens_seen": 7006672,
      "step": 8085
    },
    {
      "epoch": 3.814238566713814,
      "grad_norm": 1.0224876403808594,
      "learning_rate": 4.512221546921397e-05,
      "loss": 0.3148,
      "num_input_tokens_seen": 7011024,
      "step": 8090
    },
    {
      "epoch": 3.8165959453088165,
      "grad_norm": 0.7584603428840637,
      "learning_rate": 4.5116387972560694e-05,
      "loss": 0.1822,
      "num_input_tokens_seen": 7015264,
      "step": 8095
    },
    {
      "epoch": 3.818953323903819,
      "grad_norm": 0.635654866695404,
      "learning_rate": 4.511055737370759e-05,
      "loss": 0.1505,
      "num_input_tokens_seen": 7019616,
      "step": 8100
    },
    {
      "epoch": 3.8213107024988213,
      "grad_norm": 0.6252183318138123,
      "learning_rate": 4.510472367355383e-05,
      "loss": 0.1449,
      "num_input_tokens_seen": 7023984,
      "step": 8105
    },
    {
      "epoch": 3.8236680810938237,
      "grad_norm": 0.6952454447746277,
      "learning_rate": 4.509888687299901e-05,
      "loss": 0.1494,
      "num_input_tokens_seen": 7027648,
      "step": 8110
    },
    {
      "epoch": 3.826025459688826,
      "grad_norm": 1.2558313608169556,
      "learning_rate": 4.5093046972943266e-05,
      "loss": 0.1649,
      "num_input_tokens_seen": 7031744,
      "step": 8115
    },
    {
      "epoch": 3.8283828382838285,
      "grad_norm": 1.0722911357879639,
      "learning_rate": 4.508720397428717e-05,
      "loss": 0.178,
      "num_input_tokens_seen": 7036624,
      "step": 8120
    },
    {
      "epoch": 3.830740216878831,
      "grad_norm": 1.1286125183105469,
      "learning_rate": 4.508135787793178e-05,
      "loss": 0.2075,
      "num_input_tokens_seen": 7040960,
      "step": 8125
    },
    {
      "epoch": 3.8330975954738333,
      "grad_norm": 0.5681092739105225,
      "learning_rate": 4.5075508684778664e-05,
      "loss": 0.196,
      "num_input_tokens_seen": 7045024,
      "step": 8130
    },
    {
      "epoch": 3.8354549740688357,
      "grad_norm": 1.3077887296676636,
      "learning_rate": 4.506965639572982e-05,
      "loss": 0.1551,
      "num_input_tokens_seen": 7049360,
      "step": 8135
    },
    {
      "epoch": 3.8378123526638377,
      "grad_norm": 1.012286901473999,
      "learning_rate": 4.506380101168774e-05,
      "loss": 0.3712,
      "num_input_tokens_seen": 7053968,
      "step": 8140
    },
    {
      "epoch": 3.84016973125884,
      "grad_norm": 0.8536804914474487,
      "learning_rate": 4.505794253355542e-05,
      "loss": 0.2112,
      "num_input_tokens_seen": 7057520,
      "step": 8145
    },
    {
      "epoch": 3.8425271098538425,
      "grad_norm": 0.8060280680656433,
      "learning_rate": 4.5052080962236286e-05,
      "loss": 0.2607,
      "num_input_tokens_seen": 7061728,
      "step": 8150
    },
    {
      "epoch": 3.844884488448845,
      "grad_norm": 0.8489245176315308,
      "learning_rate": 4.504621629863428e-05,
      "loss": 0.1939,
      "num_input_tokens_seen": 7065680,
      "step": 8155
    },
    {
      "epoch": 3.8472418670438473,
      "grad_norm": 1.7807382345199585,
      "learning_rate": 4.504034854365381e-05,
      "loss": 0.1929,
      "num_input_tokens_seen": 7069904,
      "step": 8160
    },
    {
      "epoch": 3.8495992456388497,
      "grad_norm": 1.7104440927505493,
      "learning_rate": 4.503447769819974e-05,
      "loss": 0.2564,
      "num_input_tokens_seen": 7074224,
      "step": 8165
    },
    {
      "epoch": 3.851956624233852,
      "grad_norm": 1.2748651504516602,
      "learning_rate": 4.502860376317745e-05,
      "loss": 0.2917,
      "num_input_tokens_seen": 7077968,
      "step": 8170
    },
    {
      "epoch": 3.854314002828854,
      "grad_norm": 0.5699766278266907,
      "learning_rate": 4.502272673949276e-05,
      "loss": 0.1931,
      "num_input_tokens_seen": 7081552,
      "step": 8175
    },
    {
      "epoch": 3.8566713814238565,
      "grad_norm": 0.8374987244606018,
      "learning_rate": 4.501684662805199e-05,
      "loss": 0.162,
      "num_input_tokens_seen": 7086160,
      "step": 8180
    },
    {
      "epoch": 3.859028760018859,
      "grad_norm": 0.951883852481842,
      "learning_rate": 4.5010963429761924e-05,
      "loss": 0.2087,
      "num_input_tokens_seen": 7090416,
      "step": 8185
    },
    {
      "epoch": 3.8613861386138613,
      "grad_norm": 1.7666349411010742,
      "learning_rate": 4.500507714552982e-05,
      "loss": 0.1682,
      "num_input_tokens_seen": 7094304,
      "step": 8190
    },
    {
      "epoch": 3.8637435172088637,
      "grad_norm": 0.5736773014068604,
      "learning_rate": 4.499918777626342e-05,
      "loss": 0.1731,
      "num_input_tokens_seen": 7099264,
      "step": 8195
    },
    {
      "epoch": 3.866100895803866,
      "grad_norm": 0.8573030233383179,
      "learning_rate": 4.499329532287093e-05,
      "loss": 0.1801,
      "num_input_tokens_seen": 7103328,
      "step": 8200
    },
    {
      "epoch": 3.866100895803866,
      "eval_loss": 0.21035586297512054,
      "eval_runtime": 22.182,
      "eval_samples_per_second": 42.512,
      "eval_steps_per_second": 21.279,
      "num_input_tokens_seen": 7103328,
      "step": 8200
    },
    {
      "epoch": 3.8684582743988685,
      "grad_norm": 0.34275195002555847,
      "learning_rate": 4.4987399786261064e-05,
      "loss": 0.1761,
      "num_input_tokens_seen": 7107472,
      "step": 8205
    },
    {
      "epoch": 3.870815652993871,
      "grad_norm": 2.459608554840088,
      "learning_rate": 4.498150116734297e-05,
      "loss": 0.3759,
      "num_input_tokens_seen": 7111776,
      "step": 8210
    },
    {
      "epoch": 3.8731730315888733,
      "grad_norm": 2.0004465579986572,
      "learning_rate": 4.4975599467026294e-05,
      "loss": 0.1805,
      "num_input_tokens_seen": 7115616,
      "step": 8215
    },
    {
      "epoch": 3.8755304101838757,
      "grad_norm": 0.4231348931789398,
      "learning_rate": 4.496969468622114e-05,
      "loss": 0.1901,
      "num_input_tokens_seen": 7119808,
      "step": 8220
    },
    {
      "epoch": 3.877887788778878,
      "grad_norm": 0.5926587581634521,
      "learning_rate": 4.496378682583813e-05,
      "loss": 0.1936,
      "num_input_tokens_seen": 7124064,
      "step": 8225
    },
    {
      "epoch": 3.8802451673738805,
      "grad_norm": 0.976767361164093,
      "learning_rate": 4.495787588678829e-05,
      "loss": 0.2171,
      "num_input_tokens_seen": 7128848,
      "step": 8230
    },
    {
      "epoch": 3.882602545968883,
      "grad_norm": 0.7780059576034546,
      "learning_rate": 4.4951961869983196e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 7134128,
      "step": 8235
    },
    {
      "epoch": 3.884959924563885,
      "grad_norm": 0.6089810729026794,
      "learning_rate": 4.494604477633485e-05,
      "loss": 0.1943,
      "num_input_tokens_seen": 7138912,
      "step": 8240
    },
    {
      "epoch": 3.8873173031588872,
      "grad_norm": 0.5358609557151794,
      "learning_rate": 4.4940124606755734e-05,
      "loss": 0.1941,
      "num_input_tokens_seen": 7143792,
      "step": 8245
    },
    {
      "epoch": 3.8896746817538896,
      "grad_norm": 1.9514752626419067,
      "learning_rate": 4.493420136215882e-05,
      "loss": 0.1757,
      "num_input_tokens_seen": 7148368,
      "step": 8250
    },
    {
      "epoch": 3.892032060348892,
      "grad_norm": 1.5683848857879639,
      "learning_rate": 4.492827504345756e-05,
      "loss": 0.2183,
      "num_input_tokens_seen": 7154208,
      "step": 8255
    },
    {
      "epoch": 3.8943894389438944,
      "grad_norm": 1.4283537864685059,
      "learning_rate": 4.492234565156584e-05,
      "loss": 0.23,
      "num_input_tokens_seen": 7158960,
      "step": 8260
    },
    {
      "epoch": 3.896746817538897,
      "grad_norm": 2.1567790508270264,
      "learning_rate": 4.491641318739807e-05,
      "loss": 0.2562,
      "num_input_tokens_seen": 7164688,
      "step": 8265
    },
    {
      "epoch": 3.8991041961338992,
      "grad_norm": 0.5006362795829773,
      "learning_rate": 4.4910477651869096e-05,
      "loss": 0.1454,
      "num_input_tokens_seen": 7168944,
      "step": 8270
    },
    {
      "epoch": 3.901461574728901,
      "grad_norm": 1.2448405027389526,
      "learning_rate": 4.4904539045894254e-05,
      "loss": 0.2427,
      "num_input_tokens_seen": 7173648,
      "step": 8275
    },
    {
      "epoch": 3.9038189533239036,
      "grad_norm": 0.6253890991210938,
      "learning_rate": 4.4898597370389364e-05,
      "loss": 0.1785,
      "num_input_tokens_seen": 7177888,
      "step": 8280
    },
    {
      "epoch": 3.906176331918906,
      "grad_norm": 1.6377402544021606,
      "learning_rate": 4.489265262627069e-05,
      "loss": 0.2091,
      "num_input_tokens_seen": 7181984,
      "step": 8285
    },
    {
      "epoch": 3.9085337105139084,
      "grad_norm": 0.8052113056182861,
      "learning_rate": 4.488670481445499e-05,
      "loss": 0.1858,
      "num_input_tokens_seen": 7186528,
      "step": 8290
    },
    {
      "epoch": 3.910891089108911,
      "grad_norm": 0.5575518608093262,
      "learning_rate": 4.488075393585951e-05,
      "loss": 0.1859,
      "num_input_tokens_seen": 7191296,
      "step": 8295
    },
    {
      "epoch": 3.913248467703913,
      "grad_norm": 0.5439773201942444,
      "learning_rate": 4.487479999140193e-05,
      "loss": 0.1731,
      "num_input_tokens_seen": 7194944,
      "step": 8300
    },
    {
      "epoch": 3.9156058462989156,
      "grad_norm": 0.5284773111343384,
      "learning_rate": 4.4868842982000425e-05,
      "loss": 0.1936,
      "num_input_tokens_seen": 7198784,
      "step": 8305
    },
    {
      "epoch": 3.917963224893918,
      "grad_norm": 0.9521035552024841,
      "learning_rate": 4.486288290857365e-05,
      "loss": 0.1738,
      "num_input_tokens_seen": 7202752,
      "step": 8310
    },
    {
      "epoch": 3.9203206034889204,
      "grad_norm": 1.2669373750686646,
      "learning_rate": 4.4856919772040715e-05,
      "loss": 0.2037,
      "num_input_tokens_seen": 7207680,
      "step": 8315
    },
    {
      "epoch": 3.922677982083923,
      "grad_norm": 0.49530693888664246,
      "learning_rate": 4.485095357332122e-05,
      "loss": 0.187,
      "num_input_tokens_seen": 7211280,
      "step": 8320
    },
    {
      "epoch": 3.9250353606789252,
      "grad_norm": 0.6910849213600159,
      "learning_rate": 4.484498431333521e-05,
      "loss": 0.1888,
      "num_input_tokens_seen": 7215696,
      "step": 8325
    },
    {
      "epoch": 3.9273927392739276,
      "grad_norm": 0.5856147408485413,
      "learning_rate": 4.4839011993003245e-05,
      "loss": 0.1868,
      "num_input_tokens_seen": 7219104,
      "step": 8330
    },
    {
      "epoch": 3.92975011786893,
      "grad_norm": 1.7562493085861206,
      "learning_rate": 4.4833036613246305e-05,
      "loss": 0.2095,
      "num_input_tokens_seen": 7223504,
      "step": 8335
    },
    {
      "epoch": 3.932107496463932,
      "grad_norm": 0.7656682729721069,
      "learning_rate": 4.482705817498589e-05,
      "loss": 0.1405,
      "num_input_tokens_seen": 7227424,
      "step": 8340
    },
    {
      "epoch": 3.9344648750589344,
      "grad_norm": 1.5675843954086304,
      "learning_rate": 4.4821076679143934e-05,
      "loss": 0.1595,
      "num_input_tokens_seen": 7231632,
      "step": 8345
    },
    {
      "epoch": 3.936822253653937,
      "grad_norm": 0.6866617202758789,
      "learning_rate": 4.481509212664288e-05,
      "loss": 0.1406,
      "num_input_tokens_seen": 7235712,
      "step": 8350
    },
    {
      "epoch": 3.939179632248939,
      "grad_norm": 0.7972464561462402,
      "learning_rate": 4.480910451840559e-05,
      "loss": 0.1634,
      "num_input_tokens_seen": 7240384,
      "step": 8355
    },
    {
      "epoch": 3.9415370108439416,
      "grad_norm": 2.054877519607544,
      "learning_rate": 4.480311385535546e-05,
      "loss": 0.1999,
      "num_input_tokens_seen": 7244320,
      "step": 8360
    },
    {
      "epoch": 3.943894389438944,
      "grad_norm": 1.3108683824539185,
      "learning_rate": 4.47971201384163e-05,
      "loss": 0.2192,
      "num_input_tokens_seen": 7248528,
      "step": 8365
    },
    {
      "epoch": 3.9462517680339464,
      "grad_norm": 0.7375291585922241,
      "learning_rate": 4.4791123368512446e-05,
      "loss": 0.1804,
      "num_input_tokens_seen": 7252352,
      "step": 8370
    },
    {
      "epoch": 3.9486091466289484,
      "grad_norm": 0.9549609422683716,
      "learning_rate": 4.478512354656864e-05,
      "loss": 0.1517,
      "num_input_tokens_seen": 7257152,
      "step": 8375
    },
    {
      "epoch": 3.9509665252239508,
      "grad_norm": 0.779682993888855,
      "learning_rate": 4.477912067351016e-05,
      "loss": 0.2185,
      "num_input_tokens_seen": 7261136,
      "step": 8380
    },
    {
      "epoch": 3.953323903818953,
      "grad_norm": 0.48216769099235535,
      "learning_rate": 4.477311475026271e-05,
      "loss": 0.1625,
      "num_input_tokens_seen": 7265040,
      "step": 8385
    },
    {
      "epoch": 3.9556812824139556,
      "grad_norm": 0.9196189641952515,
      "learning_rate": 4.476710577775248e-05,
      "loss": 0.1831,
      "num_input_tokens_seen": 7269104,
      "step": 8390
    },
    {
      "epoch": 3.958038661008958,
      "grad_norm": 1.7936992645263672,
      "learning_rate": 4.476109375690612e-05,
      "loss": 0.2598,
      "num_input_tokens_seen": 7272512,
      "step": 8395
    },
    {
      "epoch": 3.9603960396039604,
      "grad_norm": 1.8590306043624878,
      "learning_rate": 4.4755078688650784e-05,
      "loss": 0.2825,
      "num_input_tokens_seen": 7276304,
      "step": 8400
    },
    {
      "epoch": 3.9603960396039604,
      "eval_loss": 0.20955069363117218,
      "eval_runtime": 22.2065,
      "eval_samples_per_second": 42.465,
      "eval_steps_per_second": 21.255,
      "num_input_tokens_seen": 7276304,
      "step": 8400
    },
    {
      "epoch": 3.9627534181989628,
      "grad_norm": 1.756956934928894,
      "learning_rate": 4.474906057391406e-05,
      "loss": 0.2092,
      "num_input_tokens_seen": 7281440,
      "step": 8405
    },
    {
      "epoch": 3.965110796793965,
      "grad_norm": 0.7134620547294617,
      "learning_rate": 4.4743039413624e-05,
      "loss": 0.1976,
      "num_input_tokens_seen": 7286192,
      "step": 8410
    },
    {
      "epoch": 3.9674681753889676,
      "grad_norm": 0.6521754264831543,
      "learning_rate": 4.473701520870916e-05,
      "loss": 0.1975,
      "num_input_tokens_seen": 7290224,
      "step": 8415
    },
    {
      "epoch": 3.96982555398397,
      "grad_norm": 0.8053073883056641,
      "learning_rate": 4.4730987960098544e-05,
      "loss": 0.2061,
      "num_input_tokens_seen": 7294352,
      "step": 8420
    },
    {
      "epoch": 3.9721829325789724,
      "grad_norm": 1.1950461864471436,
      "learning_rate": 4.4724957668721635e-05,
      "loss": 0.1247,
      "num_input_tokens_seen": 7299472,
      "step": 8425
    },
    {
      "epoch": 3.9745403111739748,
      "grad_norm": 0.6342334747314453,
      "learning_rate": 4.471892433550836e-05,
      "loss": 0.1268,
      "num_input_tokens_seen": 7303280,
      "step": 8430
    },
    {
      "epoch": 3.976897689768977,
      "grad_norm": 1.6607422828674316,
      "learning_rate": 4.471288796138916e-05,
      "loss": 0.1841,
      "num_input_tokens_seen": 7307504,
      "step": 8435
    },
    {
      "epoch": 3.979255068363979,
      "grad_norm": 0.8609625697135925,
      "learning_rate": 4.470684854729491e-05,
      "loss": 0.3054,
      "num_input_tokens_seen": 7312192,
      "step": 8440
    },
    {
      "epoch": 3.9816124469589815,
      "grad_norm": 1.451356291770935,
      "learning_rate": 4.4700806094156955e-05,
      "loss": 0.1726,
      "num_input_tokens_seen": 7316352,
      "step": 8445
    },
    {
      "epoch": 3.983969825553984,
      "grad_norm": 1.219629168510437,
      "learning_rate": 4.469476060290713e-05,
      "loss": 0.1895,
      "num_input_tokens_seen": 7321056,
      "step": 8450
    },
    {
      "epoch": 3.9863272041489863,
      "grad_norm": 1.6338404417037964,
      "learning_rate": 4.468871207447772e-05,
      "loss": 0.1973,
      "num_input_tokens_seen": 7324880,
      "step": 8455
    },
    {
      "epoch": 3.9886845827439887,
      "grad_norm": 0.8004659414291382,
      "learning_rate": 4.4682660509801486e-05,
      "loss": 0.1729,
      "num_input_tokens_seen": 7328976,
      "step": 8460
    },
    {
      "epoch": 3.991041961338991,
      "grad_norm": 1.0490727424621582,
      "learning_rate": 4.467660590981165e-05,
      "loss": 0.2054,
      "num_input_tokens_seen": 7334784,
      "step": 8465
    },
    {
      "epoch": 3.9933993399339935,
      "grad_norm": 0.5129048824310303,
      "learning_rate": 4.467054827544191e-05,
      "loss": 0.1825,
      "num_input_tokens_seen": 7338528,
      "step": 8470
    },
    {
      "epoch": 3.9957567185289955,
      "grad_norm": 0.9373007416725159,
      "learning_rate": 4.4664487607626434e-05,
      "loss": 0.1761,
      "num_input_tokens_seen": 7342560,
      "step": 8475
    },
    {
      "epoch": 3.998114097123998,
      "grad_norm": 1.2412158250808716,
      "learning_rate": 4.4658423907299845e-05,
      "loss": 0.2025,
      "num_input_tokens_seen": 7346944,
      "step": 8480
    },
    {
      "epoch": 4.000471475719,
      "grad_norm": 0.4596331715583801,
      "learning_rate": 4.465235717539725e-05,
      "loss": 0.1958,
      "num_input_tokens_seen": 7350464,
      "step": 8485
    },
    {
      "epoch": 4.002828854314003,
      "grad_norm": 0.9100537300109863,
      "learning_rate": 4.464628741285421e-05,
      "loss": 0.1819,
      "num_input_tokens_seen": 7354608,
      "step": 8490
    },
    {
      "epoch": 4.005186232909005,
      "grad_norm": 0.6474422216415405,
      "learning_rate": 4.4640214620606754e-05,
      "loss": 0.1992,
      "num_input_tokens_seen": 7358480,
      "step": 8495
    },
    {
      "epoch": 4.0075436115040075,
      "grad_norm": 0.6303736567497253,
      "learning_rate": 4.46341387995914e-05,
      "loss": 0.1688,
      "num_input_tokens_seen": 7362608,
      "step": 8500
    },
    {
      "epoch": 4.00990099009901,
      "grad_norm": 1.0986418724060059,
      "learning_rate": 4.4628059950745106e-05,
      "loss": 0.1538,
      "num_input_tokens_seen": 7367008,
      "step": 8505
    },
    {
      "epoch": 4.012258368694012,
      "grad_norm": 0.7816317677497864,
      "learning_rate": 4.4621978075005297e-05,
      "loss": 0.1766,
      "num_input_tokens_seen": 7371376,
      "step": 8510
    },
    {
      "epoch": 4.014615747289015,
      "grad_norm": 1.5728886127471924,
      "learning_rate": 4.461589317330989e-05,
      "loss": 0.2502,
      "num_input_tokens_seen": 7375424,
      "step": 8515
    },
    {
      "epoch": 4.016973125884017,
      "grad_norm": 0.6112728118896484,
      "learning_rate": 4.460980524659724e-05,
      "loss": 0.1565,
      "num_input_tokens_seen": 7379472,
      "step": 8520
    },
    {
      "epoch": 4.0193305044790195,
      "grad_norm": 1.9487841129302979,
      "learning_rate": 4.46037142958062e-05,
      "loss": 0.1737,
      "num_input_tokens_seen": 7383408,
      "step": 8525
    },
    {
      "epoch": 4.021687883074022,
      "grad_norm": 1.1714510917663574,
      "learning_rate": 4.4597620321876046e-05,
      "loss": 0.2065,
      "num_input_tokens_seen": 7387280,
      "step": 8530
    },
    {
      "epoch": 4.024045261669024,
      "grad_norm": 1.0632878541946411,
      "learning_rate": 4.459152332574656e-05,
      "loss": 0.226,
      "num_input_tokens_seen": 7391808,
      "step": 8535
    },
    {
      "epoch": 4.026402640264027,
      "grad_norm": 0.520340621471405,
      "learning_rate": 4.4585423308357985e-05,
      "loss": 0.2261,
      "num_input_tokens_seen": 7395728,
      "step": 8540
    },
    {
      "epoch": 4.028760018859029,
      "grad_norm": 1.1290801763534546,
      "learning_rate": 4.457932027065102e-05,
      "loss": 0.1383,
      "num_input_tokens_seen": 7399760,
      "step": 8545
    },
    {
      "epoch": 4.0311173974540315,
      "grad_norm": 0.7640106678009033,
      "learning_rate": 4.45732142135668e-05,
      "loss": 0.153,
      "num_input_tokens_seen": 7404896,
      "step": 8550
    },
    {
      "epoch": 4.033474776049033,
      "grad_norm": 0.44223079085350037,
      "learning_rate": 4.4567105138046986e-05,
      "loss": 0.1394,
      "num_input_tokens_seen": 7408928,
      "step": 8555
    },
    {
      "epoch": 4.035832154644035,
      "grad_norm": 1.1240384578704834,
      "learning_rate": 4.456099304503365e-05,
      "loss": 0.1608,
      "num_input_tokens_seen": 7413392,
      "step": 8560
    },
    {
      "epoch": 4.038189533239038,
      "grad_norm": 2.233980894088745,
      "learning_rate": 4.455487793546939e-05,
      "loss": 0.2447,
      "num_input_tokens_seen": 7417424,
      "step": 8565
    },
    {
      "epoch": 4.04054691183404,
      "grad_norm": 2.081073760986328,
      "learning_rate": 4.454875981029719e-05,
      "loss": 0.2629,
      "num_input_tokens_seen": 7422128,
      "step": 8570
    },
    {
      "epoch": 4.042904290429043,
      "grad_norm": 1.5326086282730103,
      "learning_rate": 4.454263867046057e-05,
      "loss": 0.2815,
      "num_input_tokens_seen": 7426512,
      "step": 8575
    },
    {
      "epoch": 4.045261669024045,
      "grad_norm": 0.5676333904266357,
      "learning_rate": 4.4536514516903484e-05,
      "loss": 0.1026,
      "num_input_tokens_seen": 7430496,
      "step": 8580
    },
    {
      "epoch": 4.0476190476190474,
      "grad_norm": 0.9336457848548889,
      "learning_rate": 4.453038735057034e-05,
      "loss": 0.2688,
      "num_input_tokens_seen": 7435552,
      "step": 8585
    },
    {
      "epoch": 4.04997642621405,
      "grad_norm": 0.5199158787727356,
      "learning_rate": 4.4524257172406034e-05,
      "loss": 0.1412,
      "num_input_tokens_seen": 7439360,
      "step": 8590
    },
    {
      "epoch": 4.052333804809052,
      "grad_norm": 0.7456449866294861,
      "learning_rate": 4.451812398335592e-05,
      "loss": 0.1698,
      "num_input_tokens_seen": 7443728,
      "step": 8595
    },
    {
      "epoch": 4.054691183404055,
      "grad_norm": 0.6619675159454346,
      "learning_rate": 4.4511987784365805e-05,
      "loss": 0.1634,
      "num_input_tokens_seen": 7448112,
      "step": 8600
    },
    {
      "epoch": 4.054691183404055,
      "eval_loss": 0.2127109169960022,
      "eval_runtime": 22.1726,
      "eval_samples_per_second": 42.53,
      "eval_steps_per_second": 21.288,
      "num_input_tokens_seen": 7448112,
      "step": 8600
    },
    {
      "epoch": 4.057048561999057,
      "grad_norm": 1.6459053754806519,
      "learning_rate": 4.450584857638197e-05,
      "loss": 0.1777,
      "num_input_tokens_seen": 7452768,
      "step": 8605
    },
    {
      "epoch": 4.0594059405940595,
      "grad_norm": 1.3644555807113647,
      "learning_rate": 4.449970636035116e-05,
      "loss": 0.1747,
      "num_input_tokens_seen": 7456800,
      "step": 8610
    },
    {
      "epoch": 4.061763319189062,
      "grad_norm": 1.2578604221343994,
      "learning_rate": 4.4493561137220574e-05,
      "loss": 0.2241,
      "num_input_tokens_seen": 7460960,
      "step": 8615
    },
    {
      "epoch": 4.064120697784064,
      "grad_norm": 1.1483500003814697,
      "learning_rate": 4.44874129079379e-05,
      "loss": 0.1506,
      "num_input_tokens_seen": 7465536,
      "step": 8620
    },
    {
      "epoch": 4.066478076379067,
      "grad_norm": 0.7799763679504395,
      "learning_rate": 4.4481261673451255e-05,
      "loss": 0.1799,
      "num_input_tokens_seen": 7470400,
      "step": 8625
    },
    {
      "epoch": 4.068835454974069,
      "grad_norm": 0.5778266787528992,
      "learning_rate": 4.4475107434709245e-05,
      "loss": 0.1622,
      "num_input_tokens_seen": 7474432,
      "step": 8630
    },
    {
      "epoch": 4.0711928335690715,
      "grad_norm": 0.4570435583591461,
      "learning_rate": 4.446895019266093e-05,
      "loss": 0.1766,
      "num_input_tokens_seen": 7478576,
      "step": 8635
    },
    {
      "epoch": 4.073550212164074,
      "grad_norm": 1.5450479984283447,
      "learning_rate": 4.446278994825583e-05,
      "loss": 0.1752,
      "num_input_tokens_seen": 7483120,
      "step": 8640
    },
    {
      "epoch": 4.075907590759076,
      "grad_norm": 0.7025929689407349,
      "learning_rate": 4.445662670244394e-05,
      "loss": 0.1722,
      "num_input_tokens_seen": 7487120,
      "step": 8645
    },
    {
      "epoch": 4.078264969354079,
      "grad_norm": 1.659471035003662,
      "learning_rate": 4.44504604561757e-05,
      "loss": 0.2361,
      "num_input_tokens_seen": 7491776,
      "step": 8650
    },
    {
      "epoch": 4.08062234794908,
      "grad_norm": 0.9302463531494141,
      "learning_rate": 4.4444291210402035e-05,
      "loss": 0.175,
      "num_input_tokens_seen": 7496320,
      "step": 8655
    },
    {
      "epoch": 4.082979726544083,
      "grad_norm": 0.457186758518219,
      "learning_rate": 4.443811896607431e-05,
      "loss": 0.1487,
      "num_input_tokens_seen": 7500144,
      "step": 8660
    },
    {
      "epoch": 4.085337105139085,
      "grad_norm": 0.4994775354862213,
      "learning_rate": 4.443194372414436e-05,
      "loss": 0.2061,
      "num_input_tokens_seen": 7504432,
      "step": 8665
    },
    {
      "epoch": 4.087694483734087,
      "grad_norm": 0.8967015147209167,
      "learning_rate": 4.442576548556449e-05,
      "loss": 0.2596,
      "num_input_tokens_seen": 7508864,
      "step": 8670
    },
    {
      "epoch": 4.09005186232909,
      "grad_norm": 0.38222962617874146,
      "learning_rate": 4.441958425128747e-05,
      "loss": 0.1196,
      "num_input_tokens_seen": 7512544,
      "step": 8675
    },
    {
      "epoch": 4.092409240924092,
      "grad_norm": 1.4722692966461182,
      "learning_rate": 4.4413400022266515e-05,
      "loss": 0.2698,
      "num_input_tokens_seen": 7517328,
      "step": 8680
    },
    {
      "epoch": 4.094766619519095,
      "grad_norm": 0.999476432800293,
      "learning_rate": 4.4407212799455313e-05,
      "loss": 0.1998,
      "num_input_tokens_seen": 7521264,
      "step": 8685
    },
    {
      "epoch": 4.097123998114097,
      "grad_norm": 0.4128602147102356,
      "learning_rate": 4.4401022583808003e-05,
      "loss": 0.1973,
      "num_input_tokens_seen": 7525824,
      "step": 8690
    },
    {
      "epoch": 4.099481376709099,
      "grad_norm": 1.6738300323486328,
      "learning_rate": 4.439482937627921e-05,
      "loss": 0.2208,
      "num_input_tokens_seen": 7530672,
      "step": 8695
    },
    {
      "epoch": 4.101838755304102,
      "grad_norm": 0.5655991435050964,
      "learning_rate": 4.4388633177824004e-05,
      "loss": 0.2289,
      "num_input_tokens_seen": 7536336,
      "step": 8700
    },
    {
      "epoch": 4.104196133899104,
      "grad_norm": 1.4113680124282837,
      "learning_rate": 4.4382433989397895e-05,
      "loss": 0.2173,
      "num_input_tokens_seen": 7540256,
      "step": 8705
    },
    {
      "epoch": 4.106553512494107,
      "grad_norm": 0.685195803642273,
      "learning_rate": 4.4376231811956895e-05,
      "loss": 0.1557,
      "num_input_tokens_seen": 7544160,
      "step": 8710
    },
    {
      "epoch": 4.108910891089109,
      "grad_norm": 0.6996685266494751,
      "learning_rate": 4.437002664645745e-05,
      "loss": 0.1984,
      "num_input_tokens_seen": 7547904,
      "step": 8715
    },
    {
      "epoch": 4.111268269684111,
      "grad_norm": 0.8376425504684448,
      "learning_rate": 4.436381849385649e-05,
      "loss": 0.1832,
      "num_input_tokens_seen": 7551840,
      "step": 8720
    },
    {
      "epoch": 4.113625648279114,
      "grad_norm": 1.57170832157135,
      "learning_rate": 4.435760735511136e-05,
      "loss": 0.2206,
      "num_input_tokens_seen": 7556896,
      "step": 8725
    },
    {
      "epoch": 4.115983026874116,
      "grad_norm": 0.47246965765953064,
      "learning_rate": 4.435139323117992e-05,
      "loss": 0.1556,
      "num_input_tokens_seen": 7561712,
      "step": 8730
    },
    {
      "epoch": 4.118340405469119,
      "grad_norm": 1.5908442735671997,
      "learning_rate": 4.434517612302046e-05,
      "loss": 0.2071,
      "num_input_tokens_seen": 7566528,
      "step": 8735
    },
    {
      "epoch": 4.120697784064121,
      "grad_norm": 1.1046239137649536,
      "learning_rate": 4.433895603159174e-05,
      "loss": 0.1905,
      "num_input_tokens_seen": 7571664,
      "step": 8740
    },
    {
      "epoch": 4.123055162659123,
      "grad_norm": 1.0495867729187012,
      "learning_rate": 4.433273295785296e-05,
      "loss": 0.2034,
      "num_input_tokens_seen": 7576048,
      "step": 8745
    },
    {
      "epoch": 4.125412541254126,
      "grad_norm": 1.1620335578918457,
      "learning_rate": 4.432650690276382e-05,
      "loss": 0.1448,
      "num_input_tokens_seen": 7581424,
      "step": 8750
    },
    {
      "epoch": 4.127769919849127,
      "grad_norm": 0.5116179585456848,
      "learning_rate": 4.4320277867284435e-05,
      "loss": 0.1856,
      "num_input_tokens_seen": 7585232,
      "step": 8755
    },
    {
      "epoch": 4.13012729844413,
      "grad_norm": 0.9281818270683289,
      "learning_rate": 4.431404585237541e-05,
      "loss": 0.2998,
      "num_input_tokens_seen": 7588976,
      "step": 8760
    },
    {
      "epoch": 4.132484677039132,
      "grad_norm": 0.597327470779419,
      "learning_rate": 4.43078108589978e-05,
      "loss": 0.185,
      "num_input_tokens_seen": 7593120,
      "step": 8765
    },
    {
      "epoch": 4.1348420556341345,
      "grad_norm": 0.9714121222496033,
      "learning_rate": 4.4301572888113116e-05,
      "loss": 0.2882,
      "num_input_tokens_seen": 7597488,
      "step": 8770
    },
    {
      "epoch": 4.137199434229137,
      "grad_norm": 0.8261313438415527,
      "learning_rate": 4.4295331940683337e-05,
      "loss": 0.2302,
      "num_input_tokens_seen": 7602464,
      "step": 8775
    },
    {
      "epoch": 4.139556812824139,
      "grad_norm": 1.6532809734344482,
      "learning_rate": 4.428908801767089e-05,
      "loss": 0.1859,
      "num_input_tokens_seen": 7606032,
      "step": 8780
    },
    {
      "epoch": 4.141914191419142,
      "grad_norm": 1.3443496227264404,
      "learning_rate": 4.428284112003868e-05,
      "loss": 0.1575,
      "num_input_tokens_seen": 7610416,
      "step": 8785
    },
    {
      "epoch": 4.144271570014144,
      "grad_norm": 0.6156585812568665,
      "learning_rate": 4.4276591248750033e-05,
      "loss": 0.1252,
      "num_input_tokens_seen": 7614896,
      "step": 8790
    },
    {
      "epoch": 4.1466289486091465,
      "grad_norm": 1.459492802619934,
      "learning_rate": 4.4270338404768774e-05,
      "loss": 0.1675,
      "num_input_tokens_seen": 7619440,
      "step": 8795
    },
    {
      "epoch": 4.148986327204149,
      "grad_norm": 1.5900040864944458,
      "learning_rate": 4.426408258905917e-05,
      "loss": 0.1915,
      "num_input_tokens_seen": 7623632,
      "step": 8800
    },
    {
      "epoch": 4.148986327204149,
      "eval_loss": 0.20982177555561066,
      "eval_runtime": 22.1514,
      "eval_samples_per_second": 42.571,
      "eval_steps_per_second": 21.308,
      "num_input_tokens_seen": 7623632,
      "step": 8800
    },
    {
      "epoch": 4.151343705799151,
      "grad_norm": 0.4203936755657196,
      "learning_rate": 4.425782380258594e-05,
      "loss": 0.2332,
      "num_input_tokens_seen": 7627584,
      "step": 8805
    },
    {
      "epoch": 4.153701084394154,
      "grad_norm": 0.9989956617355347,
      "learning_rate": 4.425156204631427e-05,
      "loss": 0.2166,
      "num_input_tokens_seen": 7631888,
      "step": 8810
    },
    {
      "epoch": 4.156058462989156,
      "grad_norm": 0.7304065823554993,
      "learning_rate": 4.424529732120981e-05,
      "loss": 0.2034,
      "num_input_tokens_seen": 7636096,
      "step": 8815
    },
    {
      "epoch": 4.158415841584159,
      "grad_norm": 0.6814261078834534,
      "learning_rate": 4.423902962823864e-05,
      "loss": 0.1832,
      "num_input_tokens_seen": 7639424,
      "step": 8820
    },
    {
      "epoch": 4.160773220179161,
      "grad_norm": 0.3505866825580597,
      "learning_rate": 4.423275896836733e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 7643392,
      "step": 8825
    },
    {
      "epoch": 4.163130598774163,
      "grad_norm": 1.1846238374710083,
      "learning_rate": 4.42264853425629e-05,
      "loss": 0.16,
      "num_input_tokens_seen": 7648496,
      "step": 8830
    },
    {
      "epoch": 4.165487977369166,
      "grad_norm": 1.0322810411453247,
      "learning_rate": 4.4220208751792816e-05,
      "loss": 0.1636,
      "num_input_tokens_seen": 7652768,
      "step": 8835
    },
    {
      "epoch": 4.167845355964168,
      "grad_norm": 1.6484352350234985,
      "learning_rate": 4.421392919702499e-05,
      "loss": 0.1757,
      "num_input_tokens_seen": 7657312,
      "step": 8840
    },
    {
      "epoch": 4.170202734559171,
      "grad_norm": 1.1627254486083984,
      "learning_rate": 4.4207646679227846e-05,
      "loss": 0.2317,
      "num_input_tokens_seen": 7661824,
      "step": 8845
    },
    {
      "epoch": 4.172560113154173,
      "grad_norm": 1.2400572299957275,
      "learning_rate": 4.42013611993702e-05,
      "loss": 0.1872,
      "num_input_tokens_seen": 7665680,
      "step": 8850
    },
    {
      "epoch": 4.174917491749175,
      "grad_norm": 0.5668044090270996,
      "learning_rate": 4.419507275842135e-05,
      "loss": 0.2018,
      "num_input_tokens_seen": 7670240,
      "step": 8855
    },
    {
      "epoch": 4.177274870344177,
      "grad_norm": 0.9388474225997925,
      "learning_rate": 4.418878135735106e-05,
      "loss": 0.187,
      "num_input_tokens_seen": 7674256,
      "step": 8860
    },
    {
      "epoch": 4.179632248939179,
      "grad_norm": 0.9388694167137146,
      "learning_rate": 4.418248699712955e-05,
      "loss": 0.2455,
      "num_input_tokens_seen": 7678768,
      "step": 8865
    },
    {
      "epoch": 4.181989627534182,
      "grad_norm": 1.6882585287094116,
      "learning_rate": 4.417618967872748e-05,
      "loss": 0.1378,
      "num_input_tokens_seen": 7682800,
      "step": 8870
    },
    {
      "epoch": 4.184347006129184,
      "grad_norm": 1.351457118988037,
      "learning_rate": 4.4169889403115985e-05,
      "loss": 0.1814,
      "num_input_tokens_seen": 7687296,
      "step": 8875
    },
    {
      "epoch": 4.1867043847241865,
      "grad_norm": 0.7981986999511719,
      "learning_rate": 4.4163586171266627e-05,
      "loss": 0.1659,
      "num_input_tokens_seen": 7691568,
      "step": 8880
    },
    {
      "epoch": 4.189061763319189,
      "grad_norm": 0.21288730204105377,
      "learning_rate": 4.415727998415147e-05,
      "loss": 0.1526,
      "num_input_tokens_seen": 7695616,
      "step": 8885
    },
    {
      "epoch": 4.191419141914191,
      "grad_norm": 0.8860719799995422,
      "learning_rate": 4.4150970842742985e-05,
      "loss": 0.205,
      "num_input_tokens_seen": 7700144,
      "step": 8890
    },
    {
      "epoch": 4.193776520509194,
      "grad_norm": 1.0566365718841553,
      "learning_rate": 4.4144658748014134e-05,
      "loss": 0.2118,
      "num_input_tokens_seen": 7705216,
      "step": 8895
    },
    {
      "epoch": 4.196133899104196,
      "grad_norm": 2.100452184677124,
      "learning_rate": 4.413834370093831e-05,
      "loss": 0.2287,
      "num_input_tokens_seen": 7709936,
      "step": 8900
    },
    {
      "epoch": 4.1984912776991985,
      "grad_norm": 1.1625210046768188,
      "learning_rate": 4.413202570248939e-05,
      "loss": 0.1963,
      "num_input_tokens_seen": 7714416,
      "step": 8905
    },
    {
      "epoch": 4.200848656294201,
      "grad_norm": 1.2532678842544556,
      "learning_rate": 4.412570475364167e-05,
      "loss": 0.2151,
      "num_input_tokens_seen": 7718208,
      "step": 8910
    },
    {
      "epoch": 4.203206034889203,
      "grad_norm": 0.7042484283447266,
      "learning_rate": 4.411938085536994e-05,
      "loss": 0.1428,
      "num_input_tokens_seen": 7723008,
      "step": 8915
    },
    {
      "epoch": 4.205563413484206,
      "grad_norm": 1.3538317680358887,
      "learning_rate": 4.41130540086494e-05,
      "loss": 0.2182,
      "num_input_tokens_seen": 7726864,
      "step": 8920
    },
    {
      "epoch": 4.207920792079208,
      "grad_norm": 0.6284123063087463,
      "learning_rate": 4.4106724214455754e-05,
      "loss": 0.1917,
      "num_input_tokens_seen": 7731872,
      "step": 8925
    },
    {
      "epoch": 4.2102781706742105,
      "grad_norm": 0.7771755456924438,
      "learning_rate": 4.4100391473765115e-05,
      "loss": 0.1758,
      "num_input_tokens_seen": 7736864,
      "step": 8930
    },
    {
      "epoch": 4.212635549269213,
      "grad_norm": 0.9300211071968079,
      "learning_rate": 4.409405578755408e-05,
      "loss": 0.1639,
      "num_input_tokens_seen": 7740992,
      "step": 8935
    },
    {
      "epoch": 4.214992927864215,
      "grad_norm": 1.3030860424041748,
      "learning_rate": 4.4087717156799705e-05,
      "loss": 0.1894,
      "num_input_tokens_seen": 7745168,
      "step": 8940
    },
    {
      "epoch": 4.217350306459218,
      "grad_norm": 0.5832027196884155,
      "learning_rate": 4.408137558247946e-05,
      "loss": 0.2727,
      "num_input_tokens_seen": 7749008,
      "step": 8945
    },
    {
      "epoch": 4.21970768505422,
      "grad_norm": 1.868158221244812,
      "learning_rate": 4.4075031065571306e-05,
      "loss": 0.2157,
      "num_input_tokens_seen": 7754176,
      "step": 8950
    },
    {
      "epoch": 4.222065063649222,
      "grad_norm": 1.7625998258590698,
      "learning_rate": 4.406868360705366e-05,
      "loss": 0.2136,
      "num_input_tokens_seen": 7759216,
      "step": 8955
    },
    {
      "epoch": 4.224422442244224,
      "grad_norm": 1.797856330871582,
      "learning_rate": 4.406233320790536e-05,
      "loss": 0.1994,
      "num_input_tokens_seen": 7763376,
      "step": 8960
    },
    {
      "epoch": 4.226779820839226,
      "grad_norm": 1.1122708320617676,
      "learning_rate": 4.4055979869105734e-05,
      "loss": 0.1718,
      "num_input_tokens_seen": 7767584,
      "step": 8965
    },
    {
      "epoch": 4.229137199434229,
      "grad_norm": 1.0794519186019897,
      "learning_rate": 4.404962359163454e-05,
      "loss": 0.1597,
      "num_input_tokens_seen": 7772192,
      "step": 8970
    },
    {
      "epoch": 4.231494578029231,
      "grad_norm": 0.7589421272277832,
      "learning_rate": 4.404326437647199e-05,
      "loss": 0.1747,
      "num_input_tokens_seen": 7776000,
      "step": 8975
    },
    {
      "epoch": 4.233851956624234,
      "grad_norm": 0.7238977551460266,
      "learning_rate": 4.403690222459877e-05,
      "loss": 0.201,
      "num_input_tokens_seen": 7780272,
      "step": 8980
    },
    {
      "epoch": 4.236209335219236,
      "grad_norm": 0.7309803366661072,
      "learning_rate": 4.4030537136995984e-05,
      "loss": 0.1664,
      "num_input_tokens_seen": 7784992,
      "step": 8985
    },
    {
      "epoch": 4.238566713814238,
      "grad_norm": 1.7061916589736938,
      "learning_rate": 4.402416911464523e-05,
      "loss": 0.2101,
      "num_input_tokens_seen": 7789888,
      "step": 8990
    },
    {
      "epoch": 4.240924092409241,
      "grad_norm": 1.7366790771484375,
      "learning_rate": 4.4017798158528516e-05,
      "loss": 0.179,
      "num_input_tokens_seen": 7794464,
      "step": 8995
    },
    {
      "epoch": 4.243281471004243,
      "grad_norm": 1.7199375629425049,
      "learning_rate": 4.401142426962834e-05,
      "loss": 0.1694,
      "num_input_tokens_seen": 7799248,
      "step": 9000
    },
    {
      "epoch": 4.243281471004243,
      "eval_loss": 0.2103806585073471,
      "eval_runtime": 22.1432,
      "eval_samples_per_second": 42.586,
      "eval_steps_per_second": 21.316,
      "num_input_tokens_seen": 7799248,
      "step": 9000
    },
    {
      "epoch": 4.245638849599246,
      "grad_norm": 0.6850101947784424,
      "learning_rate": 4.400504744892763e-05,
      "loss": 0.1969,
      "num_input_tokens_seen": 7803184,
      "step": 9005
    },
    {
      "epoch": 4.247996228194248,
      "grad_norm": 0.8215040564537048,
      "learning_rate": 4.399866769740975e-05,
      "loss": 0.1787,
      "num_input_tokens_seen": 7808064,
      "step": 9010
    },
    {
      "epoch": 4.2503536067892504,
      "grad_norm": 1.2527358531951904,
      "learning_rate": 4.399228501605859e-05,
      "loss": 0.1942,
      "num_input_tokens_seen": 7812352,
      "step": 9015
    },
    {
      "epoch": 4.252710985384253,
      "grad_norm": 0.7275928258895874,
      "learning_rate": 4.398589940585839e-05,
      "loss": 0.1914,
      "num_input_tokens_seen": 7816400,
      "step": 9020
    },
    {
      "epoch": 4.255068363979255,
      "grad_norm": 0.7668024897575378,
      "learning_rate": 4.3979510867793917e-05,
      "loss": 0.1203,
      "num_input_tokens_seen": 7819936,
      "step": 9025
    },
    {
      "epoch": 4.257425742574258,
      "grad_norm": 0.4961075484752655,
      "learning_rate": 4.3973119402850346e-05,
      "loss": 0.2428,
      "num_input_tokens_seen": 7823984,
      "step": 9030
    },
    {
      "epoch": 4.25978312116926,
      "grad_norm": 0.7828457355499268,
      "learning_rate": 4.396672501201334e-05,
      "loss": 0.1847,
      "num_input_tokens_seen": 7828624,
      "step": 9035
    },
    {
      "epoch": 4.2621404997642625,
      "grad_norm": 0.8275213241577148,
      "learning_rate": 4.396032769626899e-05,
      "loss": 0.1935,
      "num_input_tokens_seen": 7833136,
      "step": 9040
    },
    {
      "epoch": 4.264497878359265,
      "grad_norm": 0.9313285946846008,
      "learning_rate": 4.395392745660384e-05,
      "loss": 0.1715,
      "num_input_tokens_seen": 7837360,
      "step": 9045
    },
    {
      "epoch": 4.266855256954267,
      "grad_norm": 0.9366320967674255,
      "learning_rate": 4.394752429400488e-05,
      "loss": 0.1976,
      "num_input_tokens_seen": 7841408,
      "step": 9050
    },
    {
      "epoch": 4.26921263554927,
      "grad_norm": 1.017607569694519,
      "learning_rate": 4.394111820945957e-05,
      "loss": 0.2331,
      "num_input_tokens_seen": 7845056,
      "step": 9055
    },
    {
      "epoch": 4.271570014144271,
      "grad_norm": 0.7917237877845764,
      "learning_rate": 4.393470920395579e-05,
      "loss": 0.1627,
      "num_input_tokens_seen": 7850144,
      "step": 9060
    },
    {
      "epoch": 4.273927392739274,
      "grad_norm": 0.4721064269542694,
      "learning_rate": 4.392829727848192e-05,
      "loss": 0.2543,
      "num_input_tokens_seen": 7855024,
      "step": 9065
    },
    {
      "epoch": 4.276284771334276,
      "grad_norm": 3.299891471862793,
      "learning_rate": 4.392188243402673e-05,
      "loss": 0.3178,
      "num_input_tokens_seen": 7859680,
      "step": 9070
    },
    {
      "epoch": 4.278642149929278,
      "grad_norm": 0.7230488657951355,
      "learning_rate": 4.391546467157949e-05,
      "loss": 0.1883,
      "num_input_tokens_seen": 7863776,
      "step": 9075
    },
    {
      "epoch": 4.280999528524281,
      "grad_norm": 2.38116192817688,
      "learning_rate": 4.390904399212988e-05,
      "loss": 0.224,
      "num_input_tokens_seen": 7868192,
      "step": 9080
    },
    {
      "epoch": 4.283356907119283,
      "grad_norm": 0.9892973303794861,
      "learning_rate": 4.390262039666807e-05,
      "loss": 0.1819,
      "num_input_tokens_seen": 7872720,
      "step": 9085
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.8535193800926208,
      "learning_rate": 4.389619388618464e-05,
      "loss": 0.1425,
      "num_input_tokens_seen": 7876880,
      "step": 9090
    },
    {
      "epoch": 4.288071664309288,
      "grad_norm": 0.5350276231765747,
      "learning_rate": 4.3889764461670655e-05,
      "loss": 0.2063,
      "num_input_tokens_seen": 7881184,
      "step": 9095
    },
    {
      "epoch": 4.29042904290429,
      "grad_norm": 0.6196175813674927,
      "learning_rate": 4.38833321241176e-05,
      "loss": 0.2373,
      "num_input_tokens_seen": 7886144,
      "step": 9100
    },
    {
      "epoch": 4.292786421499293,
      "grad_norm": 1.9640042781829834,
      "learning_rate": 4.3876896874517434e-05,
      "loss": 0.2235,
      "num_input_tokens_seen": 7890272,
      "step": 9105
    },
    {
      "epoch": 4.295143800094295,
      "grad_norm": 1.4006714820861816,
      "learning_rate": 4.3870458713862554e-05,
      "loss": 0.1984,
      "num_input_tokens_seen": 7894112,
      "step": 9110
    },
    {
      "epoch": 4.297501178689298,
      "grad_norm": 0.37977272272109985,
      "learning_rate": 4.386401764314579e-05,
      "loss": 0.1257,
      "num_input_tokens_seen": 7897920,
      "step": 9115
    },
    {
      "epoch": 4.2998585572843,
      "grad_norm": 0.7274892926216125,
      "learning_rate": 4.385757366336045e-05,
      "loss": 0.1753,
      "num_input_tokens_seen": 7901888,
      "step": 9120
    },
    {
      "epoch": 4.302215935879302,
      "grad_norm": 0.5104596614837646,
      "learning_rate": 4.385112677550027e-05,
      "loss": 0.2098,
      "num_input_tokens_seen": 7906512,
      "step": 9125
    },
    {
      "epoch": 4.304573314474305,
      "grad_norm": 1.2793627977371216,
      "learning_rate": 4.384467698055945e-05,
      "loss": 0.1788,
      "num_input_tokens_seen": 7910848,
      "step": 9130
    },
    {
      "epoch": 4.306930693069307,
      "grad_norm": 2.132533550262451,
      "learning_rate": 4.383822427953261e-05,
      "loss": 0.1988,
      "num_input_tokens_seen": 7915536,
      "step": 9135
    },
    {
      "epoch": 4.30928807166431,
      "grad_norm": 0.9035170674324036,
      "learning_rate": 4.3831768673414864e-05,
      "loss": 0.2342,
      "num_input_tokens_seen": 7919904,
      "step": 9140
    },
    {
      "epoch": 4.311645450259312,
      "grad_norm": 0.7126007676124573,
      "learning_rate": 4.382531016320173e-05,
      "loss": 0.1733,
      "num_input_tokens_seen": 7925472,
      "step": 9145
    },
    {
      "epoch": 4.314002828854314,
      "grad_norm": 1.486680507659912,
      "learning_rate": 4.3818848749889184e-05,
      "loss": 0.2362,
      "num_input_tokens_seen": 7929152,
      "step": 9150
    },
    {
      "epoch": 4.316360207449316,
      "grad_norm": 1.1772900819778442,
      "learning_rate": 4.381238443447368e-05,
      "loss": 0.1377,
      "num_input_tokens_seen": 7933408,
      "step": 9155
    },
    {
      "epoch": 4.318717586044318,
      "grad_norm": 1.265820026397705,
      "learning_rate": 4.380591721795208e-05,
      "loss": 0.2451,
      "num_input_tokens_seen": 7937584,
      "step": 9160
    },
    {
      "epoch": 4.321074964639321,
      "grad_norm": 2.2722411155700684,
      "learning_rate": 4.3799447101321723e-05,
      "loss": 0.2047,
      "num_input_tokens_seen": 7941616,
      "step": 9165
    },
    {
      "epoch": 4.323432343234323,
      "grad_norm": 0.841168999671936,
      "learning_rate": 4.379297408558036e-05,
      "loss": 0.1873,
      "num_input_tokens_seen": 7946288,
      "step": 9170
    },
    {
      "epoch": 4.3257897218293255,
      "grad_norm": 0.7411676049232483,
      "learning_rate": 4.378649817172624e-05,
      "loss": 0.1706,
      "num_input_tokens_seen": 7950496,
      "step": 9175
    },
    {
      "epoch": 4.328147100424328,
      "grad_norm": 0.6269524693489075,
      "learning_rate": 4.378001936075801e-05,
      "loss": 0.1626,
      "num_input_tokens_seen": 7955056,
      "step": 9180
    },
    {
      "epoch": 4.33050447901933,
      "grad_norm": 1.117876410484314,
      "learning_rate": 4.377353765367479e-05,
      "loss": 0.2464,
      "num_input_tokens_seen": 7959520,
      "step": 9185
    },
    {
      "epoch": 4.332861857614333,
      "grad_norm": 1.061100959777832,
      "learning_rate": 4.376705305147614e-05,
      "loss": 0.2327,
      "num_input_tokens_seen": 7964944,
      "step": 9190
    },
    {
      "epoch": 4.335219236209335,
      "grad_norm": 0.40878087282180786,
      "learning_rate": 4.376056555516206e-05,
      "loss": 0.1828,
      "num_input_tokens_seen": 7969600,
      "step": 9195
    },
    {
      "epoch": 4.3375766148043375,
      "grad_norm": 0.4654250144958496,
      "learning_rate": 4.375407516573302e-05,
      "loss": 0.1458,
      "num_input_tokens_seen": 7974368,
      "step": 9200
    },
    {
      "epoch": 4.3375766148043375,
      "eval_loss": 0.20827005803585052,
      "eval_runtime": 22.1771,
      "eval_samples_per_second": 42.521,
      "eval_steps_per_second": 21.283,
      "num_input_tokens_seen": 7974368,
      "step": 9200
    },
    {
      "epoch": 4.33993399339934,
      "grad_norm": 0.7894243001937866,
      "learning_rate": 4.3747581884189913e-05,
      "loss": 0.1496,
      "num_input_tokens_seen": 7979056,
      "step": 9205
    },
    {
      "epoch": 4.342291371994342,
      "grad_norm": 0.7545067071914673,
      "learning_rate": 4.374108571153408e-05,
      "loss": 0.139,
      "num_input_tokens_seen": 7982976,
      "step": 9210
    },
    {
      "epoch": 4.344648750589345,
      "grad_norm": 1.3651808500289917,
      "learning_rate": 4.3734586648767316e-05,
      "loss": 0.2403,
      "num_input_tokens_seen": 7987152,
      "step": 9215
    },
    {
      "epoch": 4.347006129184347,
      "grad_norm": 0.8428252339363098,
      "learning_rate": 4.372808469689186e-05,
      "loss": 0.1957,
      "num_input_tokens_seen": 7991024,
      "step": 9220
    },
    {
      "epoch": 4.3493635077793495,
      "grad_norm": 0.8157060742378235,
      "learning_rate": 4.372157985691039e-05,
      "loss": 0.1525,
      "num_input_tokens_seen": 7995344,
      "step": 9225
    },
    {
      "epoch": 4.351720886374352,
      "grad_norm": 1.1351996660232544,
      "learning_rate": 4.371507212982603e-05,
      "loss": 0.2616,
      "num_input_tokens_seen": 7999584,
      "step": 9230
    },
    {
      "epoch": 4.354078264969354,
      "grad_norm": 0.8023806214332581,
      "learning_rate": 4.370856151664236e-05,
      "loss": 0.1494,
      "num_input_tokens_seen": 8003920,
      "step": 9235
    },
    {
      "epoch": 4.356435643564357,
      "grad_norm": 0.5107640027999878,
      "learning_rate": 4.3702048018363404e-05,
      "loss": 0.1842,
      "num_input_tokens_seen": 8008032,
      "step": 9240
    },
    {
      "epoch": 4.358793022159359,
      "grad_norm": 1.8975212574005127,
      "learning_rate": 4.369553163599362e-05,
      "loss": 0.2445,
      "num_input_tokens_seen": 8013088,
      "step": 9245
    },
    {
      "epoch": 4.3611504007543616,
      "grad_norm": 0.567416250705719,
      "learning_rate": 4.3689012370537904e-05,
      "loss": 0.1684,
      "num_input_tokens_seen": 8018240,
      "step": 9250
    },
    {
      "epoch": 4.363507779349364,
      "grad_norm": 0.5334466695785522,
      "learning_rate": 4.368249022300164e-05,
      "loss": 0.1913,
      "num_input_tokens_seen": 8022032,
      "step": 9255
    },
    {
      "epoch": 4.3658651579443655,
      "grad_norm": 1.2478824853897095,
      "learning_rate": 4.367596519439059e-05,
      "loss": 0.2023,
      "num_input_tokens_seen": 8026400,
      "step": 9260
    },
    {
      "epoch": 4.368222536539368,
      "grad_norm": 0.39663854241371155,
      "learning_rate": 4.366943728571101e-05,
      "loss": 0.1312,
      "num_input_tokens_seen": 8030672,
      "step": 9265
    },
    {
      "epoch": 4.37057991513437,
      "grad_norm": 1.286133050918579,
      "learning_rate": 4.366290649796959e-05,
      "loss": 0.1611,
      "num_input_tokens_seen": 8035136,
      "step": 9270
    },
    {
      "epoch": 4.372937293729373,
      "grad_norm": 1.066510558128357,
      "learning_rate": 4.3656372832173456e-05,
      "loss": 0.2047,
      "num_input_tokens_seen": 8039152,
      "step": 9275
    },
    {
      "epoch": 4.375294672324375,
      "grad_norm": 0.9776200652122498,
      "learning_rate": 4.364983628933017e-05,
      "loss": 0.1666,
      "num_input_tokens_seen": 8043344,
      "step": 9280
    },
    {
      "epoch": 4.3776520509193775,
      "grad_norm": 1.511682391166687,
      "learning_rate": 4.364329687044777e-05,
      "loss": 0.1465,
      "num_input_tokens_seen": 8047536,
      "step": 9285
    },
    {
      "epoch": 4.38000942951438,
      "grad_norm": 2.794677257537842,
      "learning_rate": 4.36367545765347e-05,
      "loss": 0.2502,
      "num_input_tokens_seen": 8051920,
      "step": 9290
    },
    {
      "epoch": 4.382366808109382,
      "grad_norm": 1.2960549592971802,
      "learning_rate": 4.363020940859988e-05,
      "loss": 0.1568,
      "num_input_tokens_seen": 8056416,
      "step": 9295
    },
    {
      "epoch": 4.384724186704385,
      "grad_norm": 0.690558671951294,
      "learning_rate": 4.362366136765263e-05,
      "loss": 0.2795,
      "num_input_tokens_seen": 8060576,
      "step": 9300
    },
    {
      "epoch": 4.387081565299387,
      "grad_norm": 0.49181345105171204,
      "learning_rate": 4.361711045470278e-05,
      "loss": 0.1471,
      "num_input_tokens_seen": 8065024,
      "step": 9305
    },
    {
      "epoch": 4.3894389438943895,
      "grad_norm": 0.6351017355918884,
      "learning_rate": 4.3610556670760524e-05,
      "loss": 0.2289,
      "num_input_tokens_seen": 8070416,
      "step": 9310
    },
    {
      "epoch": 4.391796322489392,
      "grad_norm": 0.42849817872047424,
      "learning_rate": 4.360400001683657e-05,
      "loss": 0.1815,
      "num_input_tokens_seen": 8074608,
      "step": 9315
    },
    {
      "epoch": 4.394153701084394,
      "grad_norm": 1.1668438911437988,
      "learning_rate": 4.3597440493942e-05,
      "loss": 0.2246,
      "num_input_tokens_seen": 8078976,
      "step": 9320
    },
    {
      "epoch": 4.396511079679397,
      "grad_norm": 0.633296549320221,
      "learning_rate": 4.3590878103088405e-05,
      "loss": 0.2858,
      "num_input_tokens_seen": 8082880,
      "step": 9325
    },
    {
      "epoch": 4.398868458274399,
      "grad_norm": 1.0687471628189087,
      "learning_rate": 4.358431284528779e-05,
      "loss": 0.1395,
      "num_input_tokens_seen": 8086144,
      "step": 9330
    },
    {
      "epoch": 4.4012258368694015,
      "grad_norm": 0.47236835956573486,
      "learning_rate": 4.357774472155257e-05,
      "loss": 0.1915,
      "num_input_tokens_seen": 8090960,
      "step": 9335
    },
    {
      "epoch": 4.403583215464404,
      "grad_norm": 0.8085957169532776,
      "learning_rate": 4.3571173732895664e-05,
      "loss": 0.2136,
      "num_input_tokens_seen": 8095072,
      "step": 9340
    },
    {
      "epoch": 4.405940594059406,
      "grad_norm": 0.7891285419464111,
      "learning_rate": 4.356459988033039e-05,
      "loss": 0.2696,
      "num_input_tokens_seen": 8099248,
      "step": 9345
    },
    {
      "epoch": 4.408297972654409,
      "grad_norm": 0.5486921072006226,
      "learning_rate": 4.355802316487051e-05,
      "loss": 0.1167,
      "num_input_tokens_seen": 8103616,
      "step": 9350
    },
    {
      "epoch": 4.41065535124941,
      "grad_norm": 0.6996266841888428,
      "learning_rate": 4.355144358753025e-05,
      "loss": 0.174,
      "num_input_tokens_seen": 8107648,
      "step": 9355
    },
    {
      "epoch": 4.413012729844413,
      "grad_norm": 0.6512516736984253,
      "learning_rate": 4.354486114932425e-05,
      "loss": 0.1754,
      "num_input_tokens_seen": 8112080,
      "step": 9360
    },
    {
      "epoch": 4.415370108439415,
      "grad_norm": 1.2840393781661987,
      "learning_rate": 4.353827585126762e-05,
      "loss": 0.2137,
      "num_input_tokens_seen": 8116784,
      "step": 9365
    },
    {
      "epoch": 4.417727487034417,
      "grad_norm": 0.4319985806941986,
      "learning_rate": 4.353168769437588e-05,
      "loss": 0.1744,
      "num_input_tokens_seen": 8120880,
      "step": 9370
    },
    {
      "epoch": 4.42008486562942,
      "grad_norm": 0.7903768420219421,
      "learning_rate": 4.3525096679665014e-05,
      "loss": 0.2199,
      "num_input_tokens_seen": 8125120,
      "step": 9375
    },
    {
      "epoch": 4.422442244224422,
      "grad_norm": 0.4572511315345764,
      "learning_rate": 4.351850280815144e-05,
      "loss": 0.1648,
      "num_input_tokens_seen": 8129088,
      "step": 9380
    },
    {
      "epoch": 4.424799622819425,
      "grad_norm": 0.45572641491889954,
      "learning_rate": 4.3511906080852014e-05,
      "loss": 0.1635,
      "num_input_tokens_seen": 8134192,
      "step": 9385
    },
    {
      "epoch": 4.427157001414427,
      "grad_norm": 1.0523982048034668,
      "learning_rate": 4.350530649878404e-05,
      "loss": 0.1768,
      "num_input_tokens_seen": 8138080,
      "step": 9390
    },
    {
      "epoch": 4.429514380009429,
      "grad_norm": 1.313226580619812,
      "learning_rate": 4.3498704062965246e-05,
      "loss": 0.1207,
      "num_input_tokens_seen": 8142304,
      "step": 9395
    },
    {
      "epoch": 4.431871758604432,
      "grad_norm": 1.5648599863052368,
      "learning_rate": 4.3492098774413815e-05,
      "loss": 0.2236,
      "num_input_tokens_seen": 8146384,
      "step": 9400
    },
    {
      "epoch": 4.431871758604432,
      "eval_loss": 0.2106235921382904,
      "eval_runtime": 22.197,
      "eval_samples_per_second": 42.483,
      "eval_steps_per_second": 21.264,
      "num_input_tokens_seen": 8146384,
      "step": 9400
    },
    {
      "epoch": 4.434229137199434,
      "grad_norm": 0.46746090054512024,
      "learning_rate": 4.3485490634148375e-05,
      "loss": 0.2418,
      "num_input_tokens_seen": 8150448,
      "step": 9405
    },
    {
      "epoch": 4.436586515794437,
      "grad_norm": 0.9495583772659302,
      "learning_rate": 4.347887964318797e-05,
      "loss": 0.1779,
      "num_input_tokens_seen": 8154784,
      "step": 9410
    },
    {
      "epoch": 4.438943894389439,
      "grad_norm": 0.7529160380363464,
      "learning_rate": 4.34722658025521e-05,
      "loss": 0.1927,
      "num_input_tokens_seen": 8159008,
      "step": 9415
    },
    {
      "epoch": 4.441301272984441,
      "grad_norm": 0.6734161376953125,
      "learning_rate": 4.346564911326071e-05,
      "loss": 0.1891,
      "num_input_tokens_seen": 8163488,
      "step": 9420
    },
    {
      "epoch": 4.443658651579444,
      "grad_norm": 0.9095292091369629,
      "learning_rate": 4.345902957633418e-05,
      "loss": 0.1708,
      "num_input_tokens_seen": 8167440,
      "step": 9425
    },
    {
      "epoch": 4.446016030174446,
      "grad_norm": 1.1342276334762573,
      "learning_rate": 4.345240719279331e-05,
      "loss": 0.1821,
      "num_input_tokens_seen": 8172336,
      "step": 9430
    },
    {
      "epoch": 4.448373408769449,
      "grad_norm": 0.5088895559310913,
      "learning_rate": 4.3445781963659374e-05,
      "loss": 0.1535,
      "num_input_tokens_seen": 8176160,
      "step": 9435
    },
    {
      "epoch": 4.450730787364451,
      "grad_norm": 0.6074459552764893,
      "learning_rate": 4.3439153889954045e-05,
      "loss": 0.1587,
      "num_input_tokens_seen": 8179952,
      "step": 9440
    },
    {
      "epoch": 4.4530881659594534,
      "grad_norm": 1.1213204860687256,
      "learning_rate": 4.343252297269946e-05,
      "loss": 0.2102,
      "num_input_tokens_seen": 8184048,
      "step": 9445
    },
    {
      "epoch": 4.455445544554456,
      "grad_norm": 1.5581624507904053,
      "learning_rate": 4.342588921291821e-05,
      "loss": 0.2122,
      "num_input_tokens_seen": 8188368,
      "step": 9450
    },
    {
      "epoch": 4.457802923149458,
      "grad_norm": 0.7548031806945801,
      "learning_rate": 4.341925261163328e-05,
      "loss": 0.1877,
      "num_input_tokens_seen": 8194000,
      "step": 9455
    },
    {
      "epoch": 4.46016030174446,
      "grad_norm": 1.5569041967391968,
      "learning_rate": 4.341261316986813e-05,
      "loss": 0.185,
      "num_input_tokens_seen": 8201072,
      "step": 9460
    },
    {
      "epoch": 4.462517680339462,
      "grad_norm": 0.9434026479721069,
      "learning_rate": 4.340597088864664e-05,
      "loss": 0.1414,
      "num_input_tokens_seen": 8205680,
      "step": 9465
    },
    {
      "epoch": 4.464875058934465,
      "grad_norm": 1.8076915740966797,
      "learning_rate": 4.339932576899313e-05,
      "loss": 0.319,
      "num_input_tokens_seen": 8210464,
      "step": 9470
    },
    {
      "epoch": 4.467232437529467,
      "grad_norm": 1.3679090738296509,
      "learning_rate": 4.3392677811932375e-05,
      "loss": 0.1646,
      "num_input_tokens_seen": 8214752,
      "step": 9475
    },
    {
      "epoch": 4.469589816124469,
      "grad_norm": 0.8756098747253418,
      "learning_rate": 4.338602701848956e-05,
      "loss": 0.1915,
      "num_input_tokens_seen": 8218848,
      "step": 9480
    },
    {
      "epoch": 4.471947194719472,
      "grad_norm": 2.9448323249816895,
      "learning_rate": 4.337937338969033e-05,
      "loss": 0.2366,
      "num_input_tokens_seen": 8223200,
      "step": 9485
    },
    {
      "epoch": 4.474304573314474,
      "grad_norm": 1.1859560012817383,
      "learning_rate": 4.337271692656075e-05,
      "loss": 0.1975,
      "num_input_tokens_seen": 8227280,
      "step": 9490
    },
    {
      "epoch": 4.476661951909477,
      "grad_norm": 1.44984769821167,
      "learning_rate": 4.336605763012733e-05,
      "loss": 0.2103,
      "num_input_tokens_seen": 8232304,
      "step": 9495
    },
    {
      "epoch": 4.479019330504479,
      "grad_norm": 0.7893577218055725,
      "learning_rate": 4.3359395501417026e-05,
      "loss": 0.1347,
      "num_input_tokens_seen": 8236592,
      "step": 9500
    },
    {
      "epoch": 4.481376709099481,
      "grad_norm": 0.9365079998970032,
      "learning_rate": 4.335273054145722e-05,
      "loss": 0.1446,
      "num_input_tokens_seen": 8240800,
      "step": 9505
    },
    {
      "epoch": 4.483734087694484,
      "grad_norm": 1.0757715702056885,
      "learning_rate": 4.334606275127572e-05,
      "loss": 0.1912,
      "num_input_tokens_seen": 8244816,
      "step": 9510
    },
    {
      "epoch": 4.486091466289486,
      "grad_norm": 0.8285229802131653,
      "learning_rate": 4.33393921319008e-05,
      "loss": 0.1659,
      "num_input_tokens_seen": 8249136,
      "step": 9515
    },
    {
      "epoch": 4.488448844884489,
      "grad_norm": 0.401743620634079,
      "learning_rate": 4.3332718684361146e-05,
      "loss": 0.1569,
      "num_input_tokens_seen": 8253536,
      "step": 9520
    },
    {
      "epoch": 4.490806223479491,
      "grad_norm": 0.8039323091506958,
      "learning_rate": 4.332604240968588e-05,
      "loss": 0.1524,
      "num_input_tokens_seen": 8257504,
      "step": 9525
    },
    {
      "epoch": 4.493163602074493,
      "grad_norm": 1.3556413650512695,
      "learning_rate": 4.331936330890459e-05,
      "loss": 0.196,
      "num_input_tokens_seen": 8261728,
      "step": 9530
    },
    {
      "epoch": 4.495520980669496,
      "grad_norm": 1.2395981550216675,
      "learning_rate": 4.331268138304725e-05,
      "loss": 0.1669,
      "num_input_tokens_seen": 8266592,
      "step": 9535
    },
    {
      "epoch": 4.497878359264498,
      "grad_norm": 1.3264271020889282,
      "learning_rate": 4.330599663314431e-05,
      "loss": 0.1872,
      "num_input_tokens_seen": 8270112,
      "step": 9540
    },
    {
      "epoch": 4.500235737859501,
      "grad_norm": 1.1527959108352661,
      "learning_rate": 4.329930906022665e-05,
      "loss": 0.2534,
      "num_input_tokens_seen": 8274656,
      "step": 9545
    },
    {
      "epoch": 4.502593116454502,
      "grad_norm": 0.7623805403709412,
      "learning_rate": 4.3292618665325564e-05,
      "loss": 0.1661,
      "num_input_tokens_seen": 8279040,
      "step": 9550
    },
    {
      "epoch": 4.5049504950495045,
      "grad_norm": 2.329829692840576,
      "learning_rate": 4.3285925449472796e-05,
      "loss": 0.1734,
      "num_input_tokens_seen": 8283088,
      "step": 9555
    },
    {
      "epoch": 4.507307873644507,
      "grad_norm": 0.9864648580551147,
      "learning_rate": 4.327922941370054e-05,
      "loss": 0.2455,
      "num_input_tokens_seen": 8287664,
      "step": 9560
    },
    {
      "epoch": 4.509665252239509,
      "grad_norm": 1.9561561346054077,
      "learning_rate": 4.3272530559041384e-05,
      "loss": 0.1818,
      "num_input_tokens_seen": 8291872,
      "step": 9565
    },
    {
      "epoch": 4.512022630834512,
      "grad_norm": 1.1663981676101685,
      "learning_rate": 4.32658288865284e-05,
      "loss": 0.1714,
      "num_input_tokens_seen": 8295744,
      "step": 9570
    },
    {
      "epoch": 4.514380009429514,
      "grad_norm": 1.477092981338501,
      "learning_rate": 4.325912439719505e-05,
      "loss": 0.1765,
      "num_input_tokens_seen": 8300016,
      "step": 9575
    },
    {
      "epoch": 4.5167373880245165,
      "grad_norm": 0.5661135315895081,
      "learning_rate": 4.3252417092075266e-05,
      "loss": 0.2271,
      "num_input_tokens_seen": 8303840,
      "step": 9580
    },
    {
      "epoch": 4.519094766619519,
      "grad_norm": 0.5931970477104187,
      "learning_rate": 4.3245706972203385e-05,
      "loss": 0.1711,
      "num_input_tokens_seen": 8307472,
      "step": 9585
    },
    {
      "epoch": 4.521452145214521,
      "grad_norm": 1.4069417715072632,
      "learning_rate": 4.323899403861421e-05,
      "loss": 0.1947,
      "num_input_tokens_seen": 8312128,
      "step": 9590
    },
    {
      "epoch": 4.523809523809524,
      "grad_norm": 0.6652243137359619,
      "learning_rate": 4.3232278292342935e-05,
      "loss": 0.1689,
      "num_input_tokens_seen": 8316672,
      "step": 9595
    },
    {
      "epoch": 4.526166902404526,
      "grad_norm": 2.7529900074005127,
      "learning_rate": 4.322555973442524e-05,
      "loss": 0.2027,
      "num_input_tokens_seen": 8321456,
      "step": 9600
    },
    {
      "epoch": 4.526166902404526,
      "eval_loss": 0.2086411714553833,
      "eval_runtime": 22.1548,
      "eval_samples_per_second": 42.564,
      "eval_steps_per_second": 21.305,
      "num_input_tokens_seen": 8321456,
      "step": 9600
    },
    {
      "epoch": 4.5285242809995285,
      "grad_norm": 0.8507729172706604,
      "learning_rate": 4.3218838365897184e-05,
      "loss": 0.2716,
      "num_input_tokens_seen": 8326384,
      "step": 9605
    },
    {
      "epoch": 4.530881659594531,
      "grad_norm": 0.7397441864013672,
      "learning_rate": 4.3212114187795306e-05,
      "loss": 0.1531,
      "num_input_tokens_seen": 8329920,
      "step": 9610
    },
    {
      "epoch": 4.533239038189533,
      "grad_norm": 1.7536988258361816,
      "learning_rate": 4.320538720115656e-05,
      "loss": 0.1698,
      "num_input_tokens_seen": 8334832,
      "step": 9615
    },
    {
      "epoch": 4.535596416784536,
      "grad_norm": 2.213845729827881,
      "learning_rate": 4.319865740701831e-05,
      "loss": 0.3136,
      "num_input_tokens_seen": 8339584,
      "step": 9620
    },
    {
      "epoch": 4.537953795379538,
      "grad_norm": 0.5655922889709473,
      "learning_rate": 4.3191924806418396e-05,
      "loss": 0.2023,
      "num_input_tokens_seen": 8343552,
      "step": 9625
    },
    {
      "epoch": 4.5403111739745405,
      "grad_norm": 0.7798997759819031,
      "learning_rate": 4.318518940039507e-05,
      "loss": 0.1871,
      "num_input_tokens_seen": 8348160,
      "step": 9630
    },
    {
      "epoch": 4.542668552569543,
      "grad_norm": 0.5627073049545288,
      "learning_rate": 4.3178451189987e-05,
      "loss": 0.1421,
      "num_input_tokens_seen": 8352080,
      "step": 9635
    },
    {
      "epoch": 4.545025931164545,
      "grad_norm": 0.5941051840782166,
      "learning_rate": 4.3171710176233315e-05,
      "loss": 0.1827,
      "num_input_tokens_seen": 8356576,
      "step": 9640
    },
    {
      "epoch": 4.547383309759548,
      "grad_norm": 0.6316375136375427,
      "learning_rate": 4.316496636017355e-05,
      "loss": 0.1594,
      "num_input_tokens_seen": 8361328,
      "step": 9645
    },
    {
      "epoch": 4.54974068835455,
      "grad_norm": 0.8581926822662354,
      "learning_rate": 4.315821974284771e-05,
      "loss": 0.2408,
      "num_input_tokens_seen": 8365744,
      "step": 9650
    },
    {
      "epoch": 4.5520980669495525,
      "grad_norm": 0.48930034041404724,
      "learning_rate": 4.315147032529619e-05,
      "loss": 0.2237,
      "num_input_tokens_seen": 8369984,
      "step": 9655
    },
    {
      "epoch": 4.554455445544555,
      "grad_norm": 1.1400030851364136,
      "learning_rate": 4.3144718108559845e-05,
      "loss": 0.2428,
      "num_input_tokens_seen": 8374432,
      "step": 9660
    },
    {
      "epoch": 4.5568128241395565,
      "grad_norm": 0.7909103631973267,
      "learning_rate": 4.3137963093679945e-05,
      "loss": 0.1955,
      "num_input_tokens_seen": 8378144,
      "step": 9665
    },
    {
      "epoch": 4.559170202734559,
      "grad_norm": 0.8026726841926575,
      "learning_rate": 4.31312052816982e-05,
      "loss": 0.1615,
      "num_input_tokens_seen": 8383056,
      "step": 9670
    },
    {
      "epoch": 4.561527581329561,
      "grad_norm": 1.5930348634719849,
      "learning_rate": 4.312444467365675e-05,
      "loss": 0.1369,
      "num_input_tokens_seen": 8387696,
      "step": 9675
    },
    {
      "epoch": 4.563884959924564,
      "grad_norm": 0.8849079012870789,
      "learning_rate": 4.311768127059816e-05,
      "loss": 0.1434,
      "num_input_tokens_seen": 8391712,
      "step": 9680
    },
    {
      "epoch": 4.566242338519566,
      "grad_norm": 1.1999173164367676,
      "learning_rate": 4.3110915073565444e-05,
      "loss": 0.1754,
      "num_input_tokens_seen": 8395360,
      "step": 9685
    },
    {
      "epoch": 4.5685997171145685,
      "grad_norm": 0.9217119812965393,
      "learning_rate": 4.310414608360203e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 8399152,
      "step": 9690
    },
    {
      "epoch": 4.570957095709571,
      "grad_norm": 0.6550322771072388,
      "learning_rate": 4.309737430175177e-05,
      "loss": 0.1692,
      "num_input_tokens_seen": 8403232,
      "step": 9695
    },
    {
      "epoch": 4.573314474304573,
      "grad_norm": 0.5669682025909424,
      "learning_rate": 4.309059972905897e-05,
      "loss": 0.2043,
      "num_input_tokens_seen": 8407712,
      "step": 9700
    },
    {
      "epoch": 4.575671852899576,
      "grad_norm": 0.6171478629112244,
      "learning_rate": 4.308382236656836e-05,
      "loss": 0.1411,
      "num_input_tokens_seen": 8412192,
      "step": 9705
    },
    {
      "epoch": 4.578029231494578,
      "grad_norm": 0.5644235014915466,
      "learning_rate": 4.307704221532507e-05,
      "loss": 0.2341,
      "num_input_tokens_seen": 8416144,
      "step": 9710
    },
    {
      "epoch": 4.5803866100895805,
      "grad_norm": 1.0652797222137451,
      "learning_rate": 4.307025927637471e-05,
      "loss": 0.1701,
      "num_input_tokens_seen": 8420096,
      "step": 9715
    },
    {
      "epoch": 4.582743988684583,
      "grad_norm": 0.6085250973701477,
      "learning_rate": 4.306347355076328e-05,
      "loss": 0.1348,
      "num_input_tokens_seen": 8423664,
      "step": 9720
    },
    {
      "epoch": 4.585101367279585,
      "grad_norm": 1.5494868755340576,
      "learning_rate": 4.305668503953724e-05,
      "loss": 0.2046,
      "num_input_tokens_seen": 8428112,
      "step": 9725
    },
    {
      "epoch": 4.587458745874588,
      "grad_norm": 2.1313576698303223,
      "learning_rate": 4.3049893743743436e-05,
      "loss": 0.1901,
      "num_input_tokens_seen": 8432224,
      "step": 9730
    },
    {
      "epoch": 4.58981612446959,
      "grad_norm": 0.6200114488601685,
      "learning_rate": 4.304309966442919e-05,
      "loss": 0.1669,
      "num_input_tokens_seen": 8436816,
      "step": 9735
    },
    {
      "epoch": 4.5921735030645925,
      "grad_norm": 0.5643750429153442,
      "learning_rate": 4.303630280264224e-05,
      "loss": 0.2274,
      "num_input_tokens_seen": 8441200,
      "step": 9740
    },
    {
      "epoch": 4.594530881659595,
      "grad_norm": 1.308505654335022,
      "learning_rate": 4.302950315943074e-05,
      "loss": 0.1958,
      "num_input_tokens_seen": 8446096,
      "step": 9745
    },
    {
      "epoch": 4.596888260254596,
      "grad_norm": 0.7698128819465637,
      "learning_rate": 4.3022700735843275e-05,
      "loss": 0.1274,
      "num_input_tokens_seen": 8450800,
      "step": 9750
    },
    {
      "epoch": 4.599245638849599,
      "grad_norm": 0.7885860800743103,
      "learning_rate": 4.301589553292887e-05,
      "loss": 0.2359,
      "num_input_tokens_seen": 8455296,
      "step": 9755
    },
    {
      "epoch": 4.601603017444601,
      "grad_norm": 1.2310646772384644,
      "learning_rate": 4.300908755173697e-05,
      "loss": 0.2076,
      "num_input_tokens_seen": 8459472,
      "step": 9760
    },
    {
      "epoch": 4.603960396039604,
      "grad_norm": 0.8475798964500427,
      "learning_rate": 4.300227679331745e-05,
      "loss": 0.2129,
      "num_input_tokens_seen": 8463280,
      "step": 9765
    },
    {
      "epoch": 4.606317774634606,
      "grad_norm": 1.0400522947311401,
      "learning_rate": 4.299546325872063e-05,
      "loss": 0.2666,
      "num_input_tokens_seen": 8467360,
      "step": 9770
    },
    {
      "epoch": 4.608675153229608,
      "grad_norm": 0.7246336340904236,
      "learning_rate": 4.2988646948997225e-05,
      "loss": 0.1924,
      "num_input_tokens_seen": 8471152,
      "step": 9775
    },
    {
      "epoch": 4.611032531824611,
      "grad_norm": 0.46658629179000854,
      "learning_rate": 4.29818278651984e-05,
      "loss": 0.2099,
      "num_input_tokens_seen": 8474912,
      "step": 9780
    },
    {
      "epoch": 4.613389910419613,
      "grad_norm": 0.9995042681694031,
      "learning_rate": 4.297500600837574e-05,
      "loss": 0.1996,
      "num_input_tokens_seen": 8478464,
      "step": 9785
    },
    {
      "epoch": 4.615747289014616,
      "grad_norm": 0.4806329309940338,
      "learning_rate": 4.2968181379581276e-05,
      "loss": 0.1471,
      "num_input_tokens_seen": 8482656,
      "step": 9790
    },
    {
      "epoch": 4.618104667609618,
      "grad_norm": 0.9212090373039246,
      "learning_rate": 4.296135397986743e-05,
      "loss": 0.2862,
      "num_input_tokens_seen": 8486848,
      "step": 9795
    },
    {
      "epoch": 4.62046204620462,
      "grad_norm": 0.9558032155036926,
      "learning_rate": 4.295452381028709e-05,
      "loss": 0.1654,
      "num_input_tokens_seen": 8490096,
      "step": 9800
    },
    {
      "epoch": 4.62046204620462,
      "eval_loss": 0.2078966349363327,
      "eval_runtime": 22.1343,
      "eval_samples_per_second": 42.603,
      "eval_steps_per_second": 21.324,
      "num_input_tokens_seen": 8490096,
      "step": 9800
    },
    {
      "epoch": 4.622819424799623,
      "grad_norm": 0.42294901609420776,
      "learning_rate": 4.294769087189354e-05,
      "loss": 0.1635,
      "num_input_tokens_seen": 8494800,
      "step": 9805
    },
    {
      "epoch": 4.625176803394625,
      "grad_norm": 0.5393435955047607,
      "learning_rate": 4.294085516574052e-05,
      "loss": 0.1492,
      "num_input_tokens_seen": 8498928,
      "step": 9810
    },
    {
      "epoch": 4.627534181989628,
      "grad_norm": 0.9515250325202942,
      "learning_rate": 4.2934016692882176e-05,
      "loss": 0.1797,
      "num_input_tokens_seen": 8503264,
      "step": 9815
    },
    {
      "epoch": 4.62989156058463,
      "grad_norm": 1.555909514427185,
      "learning_rate": 4.292717545437308e-05,
      "loss": 0.153,
      "num_input_tokens_seen": 8506624,
      "step": 9820
    },
    {
      "epoch": 4.632248939179632,
      "grad_norm": 1.5866687297821045,
      "learning_rate": 4.292033145126825e-05,
      "loss": 0.2355,
      "num_input_tokens_seen": 8511440,
      "step": 9825
    },
    {
      "epoch": 4.634606317774635,
      "grad_norm": 0.545413076877594,
      "learning_rate": 4.29134846846231e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 8515584,
      "step": 9830
    },
    {
      "epoch": 4.636963696369637,
      "grad_norm": 1.155220866203308,
      "learning_rate": 4.29066351554935e-05,
      "loss": 0.2798,
      "num_input_tokens_seen": 8519280,
      "step": 9835
    },
    {
      "epoch": 4.63932107496464,
      "grad_norm": 0.9606001973152161,
      "learning_rate": 4.289978286493574e-05,
      "loss": 0.1799,
      "num_input_tokens_seen": 8522816,
      "step": 9840
    },
    {
      "epoch": 4.641678453559642,
      "grad_norm": 0.5262511968612671,
      "learning_rate": 4.28929278140065e-05,
      "loss": 0.1572,
      "num_input_tokens_seen": 8526720,
      "step": 9845
    },
    {
      "epoch": 4.644035832154644,
      "grad_norm": 0.7166857719421387,
      "learning_rate": 4.288607000376295e-05,
      "loss": 0.163,
      "num_input_tokens_seen": 8531808,
      "step": 9850
    },
    {
      "epoch": 4.646393210749647,
      "grad_norm": 2.6292762756347656,
      "learning_rate": 4.2879209435262624e-05,
      "loss": 0.2079,
      "num_input_tokens_seen": 8536000,
      "step": 9855
    },
    {
      "epoch": 4.648750589344649,
      "grad_norm": 0.5023476481437683,
      "learning_rate": 4.287234610956353e-05,
      "loss": 0.2206,
      "num_input_tokens_seen": 8539776,
      "step": 9860
    },
    {
      "epoch": 4.651107967939651,
      "grad_norm": 0.928207516670227,
      "learning_rate": 4.2865480027724056e-05,
      "loss": 0.1319,
      "num_input_tokens_seen": 8544384,
      "step": 9865
    },
    {
      "epoch": 4.653465346534653,
      "grad_norm": 1.393520474433899,
      "learning_rate": 4.285861119080306e-05,
      "loss": 0.1548,
      "num_input_tokens_seen": 8548464,
      "step": 9870
    },
    {
      "epoch": 4.655822725129656,
      "grad_norm": 1.529066801071167,
      "learning_rate": 4.2851739599859784e-05,
      "loss": 0.2224,
      "num_input_tokens_seen": 8552496,
      "step": 9875
    },
    {
      "epoch": 4.658180103724658,
      "grad_norm": 0.73691725730896,
      "learning_rate": 4.2844865255953934e-05,
      "loss": 0.1448,
      "num_input_tokens_seen": 8557104,
      "step": 9880
    },
    {
      "epoch": 4.66053748231966,
      "grad_norm": 1.0698862075805664,
      "learning_rate": 4.2837988160145605e-05,
      "loss": 0.2687,
      "num_input_tokens_seen": 8561568,
      "step": 9885
    },
    {
      "epoch": 4.662894860914663,
      "grad_norm": 0.8033753037452698,
      "learning_rate": 4.2831108313495336e-05,
      "loss": 0.1811,
      "num_input_tokens_seen": 8565760,
      "step": 9890
    },
    {
      "epoch": 4.665252239509665,
      "grad_norm": 1.355417251586914,
      "learning_rate": 4.282422571706408e-05,
      "loss": 0.2173,
      "num_input_tokens_seen": 8571344,
      "step": 9895
    },
    {
      "epoch": 4.667609618104668,
      "grad_norm": 0.5635940432548523,
      "learning_rate": 4.281734037191323e-05,
      "loss": 0.148,
      "num_input_tokens_seen": 8576432,
      "step": 9900
    },
    {
      "epoch": 4.66996699669967,
      "grad_norm": 0.4922243356704712,
      "learning_rate": 4.281045227910459e-05,
      "loss": 0.2008,
      "num_input_tokens_seen": 8580640,
      "step": 9905
    },
    {
      "epoch": 4.672324375294672,
      "grad_norm": 0.796310305595398,
      "learning_rate": 4.280356143970038e-05,
      "loss": 0.2127,
      "num_input_tokens_seen": 8585392,
      "step": 9910
    },
    {
      "epoch": 4.674681753889675,
      "grad_norm": 0.9703190326690674,
      "learning_rate": 4.279666785476327e-05,
      "loss": 0.1898,
      "num_input_tokens_seen": 8589488,
      "step": 9915
    },
    {
      "epoch": 4.677039132484677,
      "grad_norm": 1.0608679056167603,
      "learning_rate": 4.2789771525356325e-05,
      "loss": 0.212,
      "num_input_tokens_seen": 8593936,
      "step": 9920
    },
    {
      "epoch": 4.67939651107968,
      "grad_norm": 0.6974827647209167,
      "learning_rate": 4.2782872452543056e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 8598976,
      "step": 9925
    },
    {
      "epoch": 4.681753889674682,
      "grad_norm": 0.6548259854316711,
      "learning_rate": 4.2775970637387376e-05,
      "loss": 0.1387,
      "num_input_tokens_seen": 8603312,
      "step": 9930
    },
    {
      "epoch": 4.684111268269684,
      "grad_norm": 0.6815323233604431,
      "learning_rate": 4.276906608095363e-05,
      "loss": 0.1607,
      "num_input_tokens_seen": 8607680,
      "step": 9935
    },
    {
      "epoch": 4.686468646864687,
      "grad_norm": 1.06612229347229,
      "learning_rate": 4.276215878430661e-05,
      "loss": 0.1617,
      "num_input_tokens_seen": 8611216,
      "step": 9940
    },
    {
      "epoch": 4.688826025459689,
      "grad_norm": 0.9430149793624878,
      "learning_rate": 4.275524874851149e-05,
      "loss": 0.1913,
      "num_input_tokens_seen": 8615744,
      "step": 9945
    },
    {
      "epoch": 4.691183404054691,
      "grad_norm": 0.43036192655563354,
      "learning_rate": 4.274833597463388e-05,
      "loss": 0.1483,
      "num_input_tokens_seen": 8620416,
      "step": 9950
    },
    {
      "epoch": 4.693540782649693,
      "grad_norm": 0.46112731099128723,
      "learning_rate": 4.2741420463739824e-05,
      "loss": 0.1768,
      "num_input_tokens_seen": 8624480,
      "step": 9955
    },
    {
      "epoch": 4.6958981612446955,
      "grad_norm": 0.7042108178138733,
      "learning_rate": 4.273450221689578e-05,
      "loss": 0.1264,
      "num_input_tokens_seen": 8629120,
      "step": 9960
    },
    {
      "epoch": 4.698255539839698,
      "grad_norm": 1.2474446296691895,
      "learning_rate": 4.272758123516863e-05,
      "loss": 0.1527,
      "num_input_tokens_seen": 8633744,
      "step": 9965
    },
    {
      "epoch": 4.7006129184347,
      "grad_norm": 0.7793437838554382,
      "learning_rate": 4.272065751962567e-05,
      "loss": 0.2211,
      "num_input_tokens_seen": 8638912,
      "step": 9970
    },
    {
      "epoch": 4.702970297029703,
      "grad_norm": 1.500207543373108,
      "learning_rate": 4.271373107133464e-05,
      "loss": 0.2453,
      "num_input_tokens_seen": 8643360,
      "step": 9975
    },
    {
      "epoch": 4.705327675624705,
      "grad_norm": 0.6248324513435364,
      "learning_rate": 4.270680189136366e-05,
      "loss": 0.2592,
      "num_input_tokens_seen": 8647680,
      "step": 9980
    },
    {
      "epoch": 4.7076850542197075,
      "grad_norm": 0.3533123731613159,
      "learning_rate": 4.269986998078132e-05,
      "loss": 0.1575,
      "num_input_tokens_seen": 8652288,
      "step": 9985
    },
    {
      "epoch": 4.71004243281471,
      "grad_norm": 0.9798794388771057,
      "learning_rate": 4.2692935340656595e-05,
      "loss": 0.2151,
      "num_input_tokens_seen": 8656768,
      "step": 9990
    },
    {
      "epoch": 4.712399811409712,
      "grad_norm": 0.832027018070221,
      "learning_rate": 4.26859979720589e-05,
      "loss": 0.2092,
      "num_input_tokens_seen": 8661920,
      "step": 9995
    },
    {
      "epoch": 4.714757190004715,
      "grad_norm": 0.6555613279342651,
      "learning_rate": 4.267905787605806e-05,
      "loss": 0.1937,
      "num_input_tokens_seen": 8665904,
      "step": 10000
    },
    {
      "epoch": 4.714757190004715,
      "eval_loss": 0.21520103514194489,
      "eval_runtime": 22.1807,
      "eval_samples_per_second": 42.514,
      "eval_steps_per_second": 21.28,
      "num_input_tokens_seen": 8665904,
      "step": 10000
    },
    {
      "epoch": 4.717114568599717,
      "grad_norm": 0.5170387625694275,
      "learning_rate": 4.267211505372433e-05,
      "loss": 0.2048,
      "num_input_tokens_seen": 8669696,
      "step": 10005
    },
    {
      "epoch": 4.7194719471947195,
      "grad_norm": 0.9957901835441589,
      "learning_rate": 4.266516950612837e-05,
      "loss": 0.1736,
      "num_input_tokens_seen": 8673440,
      "step": 10010
    },
    {
      "epoch": 4.721829325789722,
      "grad_norm": 0.9731206893920898,
      "learning_rate": 4.265822123434128e-05,
      "loss": 0.1663,
      "num_input_tokens_seen": 8677504,
      "step": 10015
    },
    {
      "epoch": 4.724186704384724,
      "grad_norm": 1.752611517906189,
      "learning_rate": 4.265127023943457e-05,
      "loss": 0.151,
      "num_input_tokens_seen": 8682544,
      "step": 10020
    },
    {
      "epoch": 4.726544082979727,
      "grad_norm": 1.0590320825576782,
      "learning_rate": 4.2644316522480176e-05,
      "loss": 0.1802,
      "num_input_tokens_seen": 8686576,
      "step": 10025
    },
    {
      "epoch": 4.728901461574729,
      "grad_norm": 2.682206630706787,
      "learning_rate": 4.263736008455044e-05,
      "loss": 0.2793,
      "num_input_tokens_seen": 8690032,
      "step": 10030
    },
    {
      "epoch": 4.7312588401697315,
      "grad_norm": 0.7191716432571411,
      "learning_rate": 4.2630400926718125e-05,
      "loss": 0.1976,
      "num_input_tokens_seen": 8694320,
      "step": 10035
    },
    {
      "epoch": 4.733616218764734,
      "grad_norm": 1.1386569738388062,
      "learning_rate": 4.262343905005644e-05,
      "loss": 0.1582,
      "num_input_tokens_seen": 8698736,
      "step": 10040
    },
    {
      "epoch": 4.735973597359736,
      "grad_norm": 0.8861280083656311,
      "learning_rate": 4.261647445563897e-05,
      "loss": 0.2502,
      "num_input_tokens_seen": 8702816,
      "step": 10045
    },
    {
      "epoch": 4.738330975954739,
      "grad_norm": 0.34061819314956665,
      "learning_rate": 4.260950714453976e-05,
      "loss": 0.1649,
      "num_input_tokens_seen": 8707568,
      "step": 10050
    },
    {
      "epoch": 4.740688354549741,
      "grad_norm": 0.4383425712585449,
      "learning_rate": 4.2602537117833266e-05,
      "loss": 0.1366,
      "num_input_tokens_seen": 8711792,
      "step": 10055
    },
    {
      "epoch": 4.7430457331447435,
      "grad_norm": 1.217005968093872,
      "learning_rate": 4.259556437659433e-05,
      "loss": 0.2281,
      "num_input_tokens_seen": 8715872,
      "step": 10060
    },
    {
      "epoch": 4.745403111739745,
      "grad_norm": 0.8488222360610962,
      "learning_rate": 4.258858892189825e-05,
      "loss": 0.1945,
      "num_input_tokens_seen": 8719776,
      "step": 10065
    },
    {
      "epoch": 4.7477604903347475,
      "grad_norm": 1.100300908088684,
      "learning_rate": 4.2581610754820725e-05,
      "loss": 0.2347,
      "num_input_tokens_seen": 8723888,
      "step": 10070
    },
    {
      "epoch": 4.75011786892975,
      "grad_norm": 1.0789536237716675,
      "learning_rate": 4.2574629876437876e-05,
      "loss": 0.2049,
      "num_input_tokens_seen": 8728048,
      "step": 10075
    },
    {
      "epoch": 4.752475247524752,
      "grad_norm": 0.5745972990989685,
      "learning_rate": 4.256764628782625e-05,
      "loss": 0.1089,
      "num_input_tokens_seen": 8733280,
      "step": 10080
    },
    {
      "epoch": 4.754832626119755,
      "grad_norm": 0.8584667444229126,
      "learning_rate": 4.256065999006279e-05,
      "loss": 0.2408,
      "num_input_tokens_seen": 8737872,
      "step": 10085
    },
    {
      "epoch": 4.757190004714757,
      "grad_norm": 0.38669297099113464,
      "learning_rate": 4.2553670984224885e-05,
      "loss": 0.2643,
      "num_input_tokens_seen": 8742208,
      "step": 10090
    },
    {
      "epoch": 4.7595473833097595,
      "grad_norm": 0.86231529712677,
      "learning_rate": 4.254667927139032e-05,
      "loss": 0.183,
      "num_input_tokens_seen": 8746144,
      "step": 10095
    },
    {
      "epoch": 4.761904761904762,
      "grad_norm": 0.6928648352622986,
      "learning_rate": 4.2539684852637295e-05,
      "loss": 0.1648,
      "num_input_tokens_seen": 8750000,
      "step": 10100
    },
    {
      "epoch": 4.764262140499764,
      "grad_norm": 0.80072420835495,
      "learning_rate": 4.253268772904446e-05,
      "loss": 0.1882,
      "num_input_tokens_seen": 8754080,
      "step": 10105
    },
    {
      "epoch": 4.766619519094767,
      "grad_norm": 1.839860439300537,
      "learning_rate": 4.252568790169085e-05,
      "loss": 0.2383,
      "num_input_tokens_seen": 8758176,
      "step": 10110
    },
    {
      "epoch": 4.768976897689769,
      "grad_norm": 1.1762959957122803,
      "learning_rate": 4.251868537165592e-05,
      "loss": 0.1957,
      "num_input_tokens_seen": 8762496,
      "step": 10115
    },
    {
      "epoch": 4.7713342762847715,
      "grad_norm": 0.6920431852340698,
      "learning_rate": 4.251168014001955e-05,
      "loss": 0.2136,
      "num_input_tokens_seen": 8766992,
      "step": 10120
    },
    {
      "epoch": 4.773691654879774,
      "grad_norm": 1.3751274347305298,
      "learning_rate": 4.250467220786204e-05,
      "loss": 0.2481,
      "num_input_tokens_seen": 8771360,
      "step": 10125
    },
    {
      "epoch": 4.776049033474776,
      "grad_norm": 1.2523950338363647,
      "learning_rate": 4.249766157626409e-05,
      "loss": 0.2285,
      "num_input_tokens_seen": 8775952,
      "step": 10130
    },
    {
      "epoch": 4.778406412069779,
      "grad_norm": 1.2643886804580688,
      "learning_rate": 4.249064824630684e-05,
      "loss": 0.2027,
      "num_input_tokens_seen": 8780464,
      "step": 10135
    },
    {
      "epoch": 4.780763790664781,
      "grad_norm": 1.0170972347259521,
      "learning_rate": 4.248363221907183e-05,
      "loss": 0.1862,
      "num_input_tokens_seen": 8785264,
      "step": 10140
    },
    {
      "epoch": 4.7831211692597835,
      "grad_norm": 0.9839297533035278,
      "learning_rate": 4.2476613495641026e-05,
      "loss": 0.192,
      "num_input_tokens_seen": 8789904,
      "step": 10145
    },
    {
      "epoch": 4.785478547854785,
      "grad_norm": 2.315812826156616,
      "learning_rate": 4.246959207709679e-05,
      "loss": 0.2467,
      "num_input_tokens_seen": 8794832,
      "step": 10150
    },
    {
      "epoch": 4.787835926449787,
      "grad_norm": 1.0475941896438599,
      "learning_rate": 4.246256796452192e-05,
      "loss": 0.1721,
      "num_input_tokens_seen": 8799680,
      "step": 10155
    },
    {
      "epoch": 4.79019330504479,
      "grad_norm": 0.9276164174079895,
      "learning_rate": 4.245554115899962e-05,
      "loss": 0.158,
      "num_input_tokens_seen": 8804256,
      "step": 10160
    },
    {
      "epoch": 4.792550683639792,
      "grad_norm": 0.7762619853019714,
      "learning_rate": 4.2448511661613514e-05,
      "loss": 0.2129,
      "num_input_tokens_seen": 8808496,
      "step": 10165
    },
    {
      "epoch": 4.794908062234795,
      "grad_norm": 3.122493267059326,
      "learning_rate": 4.2441479473447635e-05,
      "loss": 0.2259,
      "num_input_tokens_seen": 8812880,
      "step": 10170
    },
    {
      "epoch": 4.797265440829797,
      "grad_norm": 1.9571244716644287,
      "learning_rate": 4.243444459558644e-05,
      "loss": 0.3078,
      "num_input_tokens_seen": 8816720,
      "step": 10175
    },
    {
      "epoch": 4.799622819424799,
      "grad_norm": 0.5821974873542786,
      "learning_rate": 4.24274070291148e-05,
      "loss": 0.1976,
      "num_input_tokens_seen": 8821040,
      "step": 10180
    },
    {
      "epoch": 4.801980198019802,
      "grad_norm": 1.560286521911621,
      "learning_rate": 4.242036677511798e-05,
      "loss": 0.1703,
      "num_input_tokens_seen": 8825536,
      "step": 10185
    },
    {
      "epoch": 4.804337576614804,
      "grad_norm": 1.0950140953063965,
      "learning_rate": 4.241332383468169e-05,
      "loss": 0.1667,
      "num_input_tokens_seen": 8829584,
      "step": 10190
    },
    {
      "epoch": 4.806694955209807,
      "grad_norm": 0.9579285979270935,
      "learning_rate": 4.2406278208892034e-05,
      "loss": 0.2181,
      "num_input_tokens_seen": 8833904,
      "step": 10195
    },
    {
      "epoch": 4.809052333804809,
      "grad_norm": 0.6980276107788086,
      "learning_rate": 4.2399229898835536e-05,
      "loss": 0.1727,
      "num_input_tokens_seen": 8837712,
      "step": 10200
    },
    {
      "epoch": 4.809052333804809,
      "eval_loss": 0.2099882811307907,
      "eval_runtime": 22.1489,
      "eval_samples_per_second": 42.575,
      "eval_steps_per_second": 21.31,
      "num_input_tokens_seen": 8837712,
      "step": 10200
    },
    {
      "epoch": 4.811409712399811,
      "grad_norm": 0.7861825823783875,
      "learning_rate": 4.239217890559914e-05,
      "loss": 0.1871,
      "num_input_tokens_seen": 8842464,
      "step": 10205
    },
    {
      "epoch": 4.813767090994814,
      "grad_norm": 1.1993577480316162,
      "learning_rate": 4.238512523027019e-05,
      "loss": 0.2496,
      "num_input_tokens_seen": 8846768,
      "step": 10210
    },
    {
      "epoch": 4.816124469589816,
      "grad_norm": 0.6256890296936035,
      "learning_rate": 4.237806887393645e-05,
      "loss": 0.1594,
      "num_input_tokens_seen": 8850944,
      "step": 10215
    },
    {
      "epoch": 4.818481848184819,
      "grad_norm": 1.0298224687576294,
      "learning_rate": 4.237100983768611e-05,
      "loss": 0.2081,
      "num_input_tokens_seen": 8855264,
      "step": 10220
    },
    {
      "epoch": 4.820839226779821,
      "grad_norm": 0.4396386742591858,
      "learning_rate": 4.2363948122607756e-05,
      "loss": 0.1679,
      "num_input_tokens_seen": 8859424,
      "step": 10225
    },
    {
      "epoch": 4.823196605374823,
      "grad_norm": 0.9639380574226379,
      "learning_rate": 4.235688372979039e-05,
      "loss": 0.1972,
      "num_input_tokens_seen": 8864272,
      "step": 10230
    },
    {
      "epoch": 4.825553983969826,
      "grad_norm": 0.795805037021637,
      "learning_rate": 4.234981666032343e-05,
      "loss": 0.1899,
      "num_input_tokens_seen": 8868304,
      "step": 10235
    },
    {
      "epoch": 4.827911362564828,
      "grad_norm": 1.9897929430007935,
      "learning_rate": 4.2342746915296704e-05,
      "loss": 0.2173,
      "num_input_tokens_seen": 8872832,
      "step": 10240
    },
    {
      "epoch": 4.830268741159831,
      "grad_norm": 0.9272876381874084,
      "learning_rate": 4.233567449580047e-05,
      "loss": 0.1914,
      "num_input_tokens_seen": 8877472,
      "step": 10245
    },
    {
      "epoch": 4.832626119754833,
      "grad_norm": 1.4856572151184082,
      "learning_rate": 4.232859940292537e-05,
      "loss": 0.2298,
      "num_input_tokens_seen": 8882304,
      "step": 10250
    },
    {
      "epoch": 4.834983498349835,
      "grad_norm": 0.7799663543701172,
      "learning_rate": 4.232152163776248e-05,
      "loss": 0.2246,
      "num_input_tokens_seen": 8886272,
      "step": 10255
    },
    {
      "epoch": 4.837340876944838,
      "grad_norm": 1.756327509880066,
      "learning_rate": 4.231444120140328e-05,
      "loss": 0.2045,
      "num_input_tokens_seen": 8890368,
      "step": 10260
    },
    {
      "epoch": 4.839698255539839,
      "grad_norm": 1.1572613716125488,
      "learning_rate": 4.230735809493967e-05,
      "loss": 0.1997,
      "num_input_tokens_seen": 8894480,
      "step": 10265
    },
    {
      "epoch": 4.842055634134842,
      "grad_norm": 2.164384365081787,
      "learning_rate": 4.2300272319463926e-05,
      "loss": 0.2125,
      "num_input_tokens_seen": 8898560,
      "step": 10270
    },
    {
      "epoch": 4.844413012729844,
      "grad_norm": 0.631945788860321,
      "learning_rate": 4.2293183876068786e-05,
      "loss": 0.1985,
      "num_input_tokens_seen": 8902608,
      "step": 10275
    },
    {
      "epoch": 4.8467703913248465,
      "grad_norm": 0.9145972728729248,
      "learning_rate": 4.228609276584737e-05,
      "loss": 0.1743,
      "num_input_tokens_seen": 8907040,
      "step": 10280
    },
    {
      "epoch": 4.849127769919849,
      "grad_norm": 0.8193127512931824,
      "learning_rate": 4.227899898989323e-05,
      "loss": 0.2,
      "num_input_tokens_seen": 8910704,
      "step": 10285
    },
    {
      "epoch": 4.851485148514851,
      "grad_norm": 0.8380143642425537,
      "learning_rate": 4.2271902549300293e-05,
      "loss": 0.2635,
      "num_input_tokens_seen": 8915072,
      "step": 10290
    },
    {
      "epoch": 4.853842527109854,
      "grad_norm": 0.6678992509841919,
      "learning_rate": 4.226480344516294e-05,
      "loss": 0.2036,
      "num_input_tokens_seen": 8918976,
      "step": 10295
    },
    {
      "epoch": 4.856199905704856,
      "grad_norm": 1.2111320495605469,
      "learning_rate": 4.2257701678575925e-05,
      "loss": 0.2126,
      "num_input_tokens_seen": 8923568,
      "step": 10300
    },
    {
      "epoch": 4.858557284299859,
      "grad_norm": 0.45676788687705994,
      "learning_rate": 4.225059725063444e-05,
      "loss": 0.1955,
      "num_input_tokens_seen": 8927968,
      "step": 10305
    },
    {
      "epoch": 4.860914662894861,
      "grad_norm": 0.6736120581626892,
      "learning_rate": 4.2243490162434074e-05,
      "loss": 0.1567,
      "num_input_tokens_seen": 8932480,
      "step": 10310
    },
    {
      "epoch": 4.863272041489863,
      "grad_norm": 0.5128206610679626,
      "learning_rate": 4.223638041507083e-05,
      "loss": 0.1511,
      "num_input_tokens_seen": 8936624,
      "step": 10315
    },
    {
      "epoch": 4.865629420084866,
      "grad_norm": 0.5427549481391907,
      "learning_rate": 4.2229268009641124e-05,
      "loss": 0.1709,
      "num_input_tokens_seen": 8941264,
      "step": 10320
    },
    {
      "epoch": 4.867986798679868,
      "grad_norm": 1.4386613368988037,
      "learning_rate": 4.222215294724177e-05,
      "loss": 0.1578,
      "num_input_tokens_seen": 8945600,
      "step": 10325
    },
    {
      "epoch": 4.870344177274871,
      "grad_norm": 1.9236745834350586,
      "learning_rate": 4.2215035228970005e-05,
      "loss": 0.2399,
      "num_input_tokens_seen": 8949616,
      "step": 10330
    },
    {
      "epoch": 4.872701555869873,
      "grad_norm": 0.8603480458259583,
      "learning_rate": 4.2207914855923464e-05,
      "loss": 0.2142,
      "num_input_tokens_seen": 8954672,
      "step": 10335
    },
    {
      "epoch": 4.875058934464875,
      "grad_norm": 1.3765954971313477,
      "learning_rate": 4.220079182920021e-05,
      "loss": 0.1779,
      "num_input_tokens_seen": 8958576,
      "step": 10340
    },
    {
      "epoch": 4.877416313059878,
      "grad_norm": 0.5780720710754395,
      "learning_rate": 4.2193666149898705e-05,
      "loss": 0.1821,
      "num_input_tokens_seen": 8962880,
      "step": 10345
    },
    {
      "epoch": 4.879773691654879,
      "grad_norm": 1.0254771709442139,
      "learning_rate": 4.21865378191178e-05,
      "loss": 0.1893,
      "num_input_tokens_seen": 8967952,
      "step": 10350
    },
    {
      "epoch": 4.882131070249882,
      "grad_norm": 0.7489106059074402,
      "learning_rate": 4.217940683795678e-05,
      "loss": 0.2002,
      "num_input_tokens_seen": 8972160,
      "step": 10355
    },
    {
      "epoch": 4.884488448844884,
      "grad_norm": 2.145120143890381,
      "learning_rate": 4.217227320751534e-05,
      "loss": 0.2513,
      "num_input_tokens_seen": 8975888,
      "step": 10360
    },
    {
      "epoch": 4.8868458274398865,
      "grad_norm": 0.8164795637130737,
      "learning_rate": 4.216513692889358e-05,
      "loss": 0.191,
      "num_input_tokens_seen": 8980752,
      "step": 10365
    },
    {
      "epoch": 4.889203206034889,
      "grad_norm": 1.2523095607757568,
      "learning_rate": 4.215799800319199e-05,
      "loss": 0.2496,
      "num_input_tokens_seen": 8984704,
      "step": 10370
    },
    {
      "epoch": 4.891560584629891,
      "grad_norm": 0.5639967918395996,
      "learning_rate": 4.2150856431511485e-05,
      "loss": 0.1802,
      "num_input_tokens_seen": 8988432,
      "step": 10375
    },
    {
      "epoch": 4.893917963224894,
      "grad_norm": 2.6392722129821777,
      "learning_rate": 4.214371221495339e-05,
      "loss": 0.1952,
      "num_input_tokens_seen": 8992704,
      "step": 10380
    },
    {
      "epoch": 4.896275341819896,
      "grad_norm": 1.608940839767456,
      "learning_rate": 4.213656535461942e-05,
      "loss": 0.1682,
      "num_input_tokens_seen": 8996560,
      "step": 10385
    },
    {
      "epoch": 4.8986327204148985,
      "grad_norm": 0.7743380665779114,
      "learning_rate": 4.2129415851611734e-05,
      "loss": 0.1913,
      "num_input_tokens_seen": 9001280,
      "step": 10390
    },
    {
      "epoch": 4.900990099009901,
      "grad_norm": 0.7090429067611694,
      "learning_rate": 4.2122263707032855e-05,
      "loss": 0.2858,
      "num_input_tokens_seen": 9005616,
      "step": 10395
    },
    {
      "epoch": 4.903347477604903,
      "grad_norm": 2.0189762115478516,
      "learning_rate": 4.211510892198574e-05,
      "loss": 0.2062,
      "num_input_tokens_seen": 9010400,
      "step": 10400
    },
    {
      "epoch": 4.903347477604903,
      "eval_loss": 0.20720769464969635,
      "eval_runtime": 22.1377,
      "eval_samples_per_second": 42.597,
      "eval_steps_per_second": 21.321,
      "num_input_tokens_seen": 9010400,
      "step": 10400
    },
    {
      "epoch": 4.905704856199906,
      "grad_norm": 1.6947910785675049,
      "learning_rate": 4.210795149757375e-05,
      "loss": 0.302,
      "num_input_tokens_seen": 9014832,
      "step": 10405
    },
    {
      "epoch": 4.908062234794908,
      "grad_norm": 1.2959092855453491,
      "learning_rate": 4.210079143490065e-05,
      "loss": 0.2385,
      "num_input_tokens_seen": 9019648,
      "step": 10410
    },
    {
      "epoch": 4.9104196133899105,
      "grad_norm": 1.321946620941162,
      "learning_rate": 4.2093628735070604e-05,
      "loss": 0.1364,
      "num_input_tokens_seen": 9024000,
      "step": 10415
    },
    {
      "epoch": 4.912776991984913,
      "grad_norm": 1.4613758325576782,
      "learning_rate": 4.208646339918819e-05,
      "loss": 0.2318,
      "num_input_tokens_seen": 9028352,
      "step": 10420
    },
    {
      "epoch": 4.915134370579915,
      "grad_norm": 0.5185139179229736,
      "learning_rate": 4.2079295428358414e-05,
      "loss": 0.1716,
      "num_input_tokens_seen": 9032080,
      "step": 10425
    },
    {
      "epoch": 4.917491749174918,
      "grad_norm": 0.8096476793289185,
      "learning_rate": 4.207212482368664e-05,
      "loss": 0.2727,
      "num_input_tokens_seen": 9036864,
      "step": 10430
    },
    {
      "epoch": 4.91984912776992,
      "grad_norm": 1.2091165781021118,
      "learning_rate": 4.206495158627867e-05,
      "loss": 0.2247,
      "num_input_tokens_seen": 9040800,
      "step": 10435
    },
    {
      "epoch": 4.9222065063649225,
      "grad_norm": 0.8294252157211304,
      "learning_rate": 4.205777571724073e-05,
      "loss": 0.2164,
      "num_input_tokens_seen": 9044864,
      "step": 10440
    },
    {
      "epoch": 4.924563884959925,
      "grad_norm": 0.4582131505012512,
      "learning_rate": 4.20505972176794e-05,
      "loss": 0.192,
      "num_input_tokens_seen": 9049024,
      "step": 10445
    },
    {
      "epoch": 4.926921263554927,
      "grad_norm": 0.8007526397705078,
      "learning_rate": 4.204341608870171e-05,
      "loss": 0.1313,
      "num_input_tokens_seen": 9053248,
      "step": 10450
    },
    {
      "epoch": 4.92927864214993,
      "grad_norm": 0.5522972941398621,
      "learning_rate": 4.203623233141508e-05,
      "loss": 0.1749,
      "num_input_tokens_seen": 9057376,
      "step": 10455
    },
    {
      "epoch": 4.931636020744932,
      "grad_norm": 0.9926797151565552,
      "learning_rate": 4.2029045946927334e-05,
      "loss": 0.1758,
      "num_input_tokens_seen": 9061024,
      "step": 10460
    },
    {
      "epoch": 4.933993399339934,
      "grad_norm": 0.8600942492485046,
      "learning_rate": 4.20218569363467e-05,
      "loss": 0.2162,
      "num_input_tokens_seen": 9065664,
      "step": 10465
    },
    {
      "epoch": 4.936350777934936,
      "grad_norm": 2.050795793533325,
      "learning_rate": 4.2014665300781834e-05,
      "loss": 0.2405,
      "num_input_tokens_seen": 9069968,
      "step": 10470
    },
    {
      "epoch": 4.938708156529938,
      "grad_norm": 0.9181129336357117,
      "learning_rate": 4.200747104134174e-05,
      "loss": 0.19,
      "num_input_tokens_seen": 9074432,
      "step": 10475
    },
    {
      "epoch": 4.941065535124941,
      "grad_norm": 0.48956847190856934,
      "learning_rate": 4.200027415913588e-05,
      "loss": 0.1739,
      "num_input_tokens_seen": 9078816,
      "step": 10480
    },
    {
      "epoch": 4.943422913719943,
      "grad_norm": 2.1236915588378906,
      "learning_rate": 4.1993074655274126e-05,
      "loss": 0.248,
      "num_input_tokens_seen": 9083712,
      "step": 10485
    },
    {
      "epoch": 4.945780292314946,
      "grad_norm": 0.6372154355049133,
      "learning_rate": 4.198587253086669e-05,
      "loss": 0.2121,
      "num_input_tokens_seen": 9088880,
      "step": 10490
    },
    {
      "epoch": 4.948137670909948,
      "grad_norm": 0.9355289340019226,
      "learning_rate": 4.197866778702426e-05,
      "loss": 0.2467,
      "num_input_tokens_seen": 9093760,
      "step": 10495
    },
    {
      "epoch": 4.9504950495049505,
      "grad_norm": 0.6371662020683289,
      "learning_rate": 4.197146042485789e-05,
      "loss": 0.1692,
      "num_input_tokens_seen": 9098944,
      "step": 10500
    },
    {
      "epoch": 4.952852428099953,
      "grad_norm": 0.7682996988296509,
      "learning_rate": 4.1964250445479046e-05,
      "loss": 0.1347,
      "num_input_tokens_seen": 9102640,
      "step": 10505
    },
    {
      "epoch": 4.955209806694955,
      "grad_norm": 0.9456664323806763,
      "learning_rate": 4.19570378499996e-05,
      "loss": 0.2284,
      "num_input_tokens_seen": 9107072,
      "step": 10510
    },
    {
      "epoch": 4.957567185289958,
      "grad_norm": 0.8235449194908142,
      "learning_rate": 4.194982263953182e-05,
      "loss": 0.1794,
      "num_input_tokens_seen": 9111216,
      "step": 10515
    },
    {
      "epoch": 4.95992456388496,
      "grad_norm": 0.655134379863739,
      "learning_rate": 4.194260481518838e-05,
      "loss": 0.1634,
      "num_input_tokens_seen": 9116352,
      "step": 10520
    },
    {
      "epoch": 4.9622819424799625,
      "grad_norm": 0.5501133799552917,
      "learning_rate": 4.1935384378082366e-05,
      "loss": 0.1686,
      "num_input_tokens_seen": 9120528,
      "step": 10525
    },
    {
      "epoch": 4.964639321074965,
      "grad_norm": 0.8234023451805115,
      "learning_rate": 4.1928161329327267e-05,
      "loss": 0.1918,
      "num_input_tokens_seen": 9124560,
      "step": 10530
    },
    {
      "epoch": 4.966996699669967,
      "grad_norm": 0.6785542368888855,
      "learning_rate": 4.1920935670036945e-05,
      "loss": 0.1731,
      "num_input_tokens_seen": 9129120,
      "step": 10535
    },
    {
      "epoch": 4.96935407826497,
      "grad_norm": 0.36810746788978577,
      "learning_rate": 4.1913707401325705e-05,
      "loss": 0.1657,
      "num_input_tokens_seen": 9132960,
      "step": 10540
    },
    {
      "epoch": 4.971711456859972,
      "grad_norm": 0.8953893780708313,
      "learning_rate": 4.1906476524308235e-05,
      "loss": 0.2027,
      "num_input_tokens_seen": 9137216,
      "step": 10545
    },
    {
      "epoch": 4.974068835454974,
      "grad_norm": 1.4004720449447632,
      "learning_rate": 4.189924304009962e-05,
      "loss": 0.2294,
      "num_input_tokens_seen": 9141776,
      "step": 10550
    },
    {
      "epoch": 4.976426214049976,
      "grad_norm": 0.9181898236274719,
      "learning_rate": 4.189200694981537e-05,
      "loss": 0.1429,
      "num_input_tokens_seen": 9145696,
      "step": 10555
    },
    {
      "epoch": 4.978783592644978,
      "grad_norm": 1.2034872770309448,
      "learning_rate": 4.188476825457136e-05,
      "loss": 0.2261,
      "num_input_tokens_seen": 9149904,
      "step": 10560
    },
    {
      "epoch": 4.981140971239981,
      "grad_norm": 0.6583842635154724,
      "learning_rate": 4.18775269554839e-05,
      "loss": 0.279,
      "num_input_tokens_seen": 9154544,
      "step": 10565
    },
    {
      "epoch": 4.983498349834983,
      "grad_norm": 0.4035191237926483,
      "learning_rate": 4.187028305366969e-05,
      "loss": 0.1489,
      "num_input_tokens_seen": 9158768,
      "step": 10570
    },
    {
      "epoch": 4.985855728429986,
      "grad_norm": 1.4351283311843872,
      "learning_rate": 4.1863036550245824e-05,
      "loss": 0.2181,
      "num_input_tokens_seen": 9163232,
      "step": 10575
    },
    {
      "epoch": 4.988213107024988,
      "grad_norm": 1.4319971799850464,
      "learning_rate": 4.1855787446329806e-05,
      "loss": 0.1326,
      "num_input_tokens_seen": 9167456,
      "step": 10580
    },
    {
      "epoch": 4.99057048561999,
      "grad_norm": 1.9607025384902954,
      "learning_rate": 4.184853574303955e-05,
      "loss": 0.2767,
      "num_input_tokens_seen": 9172288,
      "step": 10585
    },
    {
      "epoch": 4.992927864214993,
      "grad_norm": 0.8571035861968994,
      "learning_rate": 4.184128144149334e-05,
      "loss": 0.2192,
      "num_input_tokens_seen": 9176736,
      "step": 10590
    },
    {
      "epoch": 4.995285242809995,
      "grad_norm": 1.3640042543411255,
      "learning_rate": 4.1834024542809896e-05,
      "loss": 0.1658,
      "num_input_tokens_seen": 9180880,
      "step": 10595
    },
    {
      "epoch": 4.997642621404998,
      "grad_norm": 0.94045490026474,
      "learning_rate": 4.1826765048108315e-05,
      "loss": 0.1884,
      "num_input_tokens_seen": 9185584,
      "step": 10600
    },
    {
      "epoch": 4.997642621404998,
      "eval_loss": 0.20559388399124146,
      "eval_runtime": 22.1973,
      "eval_samples_per_second": 42.483,
      "eval_steps_per_second": 21.264,
      "num_input_tokens_seen": 9185584,
      "step": 10600
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.7138654589653015,
      "learning_rate": 4.181950295850811e-05,
      "loss": 0.1714,
      "num_input_tokens_seen": 9189824,
      "step": 10605
    },
    {
      "epoch": 5.002357378595002,
      "grad_norm": 0.9641549587249756,
      "learning_rate": 4.181223827512918e-05,
      "loss": 0.184,
      "num_input_tokens_seen": 9194368,
      "step": 10610
    },
    {
      "epoch": 5.004714757190005,
      "grad_norm": 0.6832592487335205,
      "learning_rate": 4.180497099909183e-05,
      "loss": 0.2018,
      "num_input_tokens_seen": 9199328,
      "step": 10615
    },
    {
      "epoch": 5.007072135785007,
      "grad_norm": 0.8170561194419861,
      "learning_rate": 4.179770113151677e-05,
      "loss": 0.1979,
      "num_input_tokens_seen": 9202896,
      "step": 10620
    },
    {
      "epoch": 5.00942951438001,
      "grad_norm": 0.5014097690582275,
      "learning_rate": 4.179042867352511e-05,
      "loss": 0.1798,
      "num_input_tokens_seen": 9207088,
      "step": 10625
    },
    {
      "epoch": 5.011786892975012,
      "grad_norm": 0.6254046559333801,
      "learning_rate": 4.1783153626238334e-05,
      "loss": 0.1551,
      "num_input_tokens_seen": 9211488,
      "step": 10630
    },
    {
      "epoch": 5.014144271570014,
      "grad_norm": 0.5340003371238708,
      "learning_rate": 4.177587599077836e-05,
      "loss": 0.2026,
      "num_input_tokens_seen": 9215024,
      "step": 10635
    },
    {
      "epoch": 5.016501650165017,
      "grad_norm": 1.6181005239486694,
      "learning_rate": 4.1768595768267494e-05,
      "loss": 0.2516,
      "num_input_tokens_seen": 9218928,
      "step": 10640
    },
    {
      "epoch": 5.018859028760019,
      "grad_norm": 0.5438501834869385,
      "learning_rate": 4.176131295982843e-05,
      "loss": 0.1658,
      "num_input_tokens_seen": 9222992,
      "step": 10645
    },
    {
      "epoch": 5.021216407355022,
      "grad_norm": 0.5716044306755066,
      "learning_rate": 4.1754027566584276e-05,
      "loss": 0.1527,
      "num_input_tokens_seen": 9227264,
      "step": 10650
    },
    {
      "epoch": 5.023573785950024,
      "grad_norm": 0.5085797905921936,
      "learning_rate": 4.174673958965852e-05,
      "loss": 0.1088,
      "num_input_tokens_seen": 9231808,
      "step": 10655
    },
    {
      "epoch": 5.0259311645450255,
      "grad_norm": 0.5002150535583496,
      "learning_rate": 4.173944903017507e-05,
      "loss": 0.1816,
      "num_input_tokens_seen": 9237072,
      "step": 10660
    },
    {
      "epoch": 5.028288543140028,
      "grad_norm": 0.7989639043807983,
      "learning_rate": 4.173215588925822e-05,
      "loss": 0.1868,
      "num_input_tokens_seen": 9241808,
      "step": 10665
    },
    {
      "epoch": 5.03064592173503,
      "grad_norm": 0.9384559392929077,
      "learning_rate": 4.172486016803266e-05,
      "loss": 0.1853,
      "num_input_tokens_seen": 9245840,
      "step": 10670
    },
    {
      "epoch": 5.033003300330033,
      "grad_norm": 0.4322076737880707,
      "learning_rate": 4.171756186762349e-05,
      "loss": 0.1793,
      "num_input_tokens_seen": 9249728,
      "step": 10675
    },
    {
      "epoch": 5.035360678925035,
      "grad_norm": 1.2186270952224731,
      "learning_rate": 4.171026098915619e-05,
      "loss": 0.2772,
      "num_input_tokens_seen": 9254080,
      "step": 10680
    },
    {
      "epoch": 5.0377180575200375,
      "grad_norm": 0.8697561621665955,
      "learning_rate": 4.170295753375665e-05,
      "loss": 0.2435,
      "num_input_tokens_seen": 9257952,
      "step": 10685
    },
    {
      "epoch": 5.04007543611504,
      "grad_norm": 0.7626600861549377,
      "learning_rate": 4.169565150255117e-05,
      "loss": 0.1741,
      "num_input_tokens_seen": 9261568,
      "step": 10690
    },
    {
      "epoch": 5.042432814710042,
      "grad_norm": 1.3398765325546265,
      "learning_rate": 4.16883428966664e-05,
      "loss": 0.2441,
      "num_input_tokens_seen": 9266656,
      "step": 10695
    },
    {
      "epoch": 5.044790193305045,
      "grad_norm": 1.3275502920150757,
      "learning_rate": 4.168103171722944e-05,
      "loss": 0.1894,
      "num_input_tokens_seen": 9271424,
      "step": 10700
    },
    {
      "epoch": 5.047147571900047,
      "grad_norm": 3.3179588317871094,
      "learning_rate": 4.167371796536777e-05,
      "loss": 0.2715,
      "num_input_tokens_seen": 9275696,
      "step": 10705
    },
    {
      "epoch": 5.0495049504950495,
      "grad_norm": 0.44503721594810486,
      "learning_rate": 4.166640164220924e-05,
      "loss": 0.1504,
      "num_input_tokens_seen": 9280176,
      "step": 10710
    },
    {
      "epoch": 5.051862329090052,
      "grad_norm": 1.545180082321167,
      "learning_rate": 4.1659082748882144e-05,
      "loss": 0.218,
      "num_input_tokens_seen": 9284528,
      "step": 10715
    },
    {
      "epoch": 5.054219707685054,
      "grad_norm": 0.672044575214386,
      "learning_rate": 4.1651761286515135e-05,
      "loss": 0.2235,
      "num_input_tokens_seen": 9288912,
      "step": 10720
    },
    {
      "epoch": 5.056577086280057,
      "grad_norm": 1.1917270421981812,
      "learning_rate": 4.164443725623728e-05,
      "loss": 0.1623,
      "num_input_tokens_seen": 9293120,
      "step": 10725
    },
    {
      "epoch": 5.058934464875059,
      "grad_norm": 1.1209782361984253,
      "learning_rate": 4.163711065917802e-05,
      "loss": 0.1834,
      "num_input_tokens_seen": 9297088,
      "step": 10730
    },
    {
      "epoch": 5.061291843470062,
      "grad_norm": 1.1665209531784058,
      "learning_rate": 4.1629781496467234e-05,
      "loss": 0.2004,
      "num_input_tokens_seen": 9300784,
      "step": 10735
    },
    {
      "epoch": 5.063649222065064,
      "grad_norm": 1.6208617687225342,
      "learning_rate": 4.1622449769235164e-05,
      "loss": 0.1603,
      "num_input_tokens_seen": 9305200,
      "step": 10740
    },
    {
      "epoch": 5.066006600660066,
      "grad_norm": 0.38687214255332947,
      "learning_rate": 4.161511547861243e-05,
      "loss": 0.1402,
      "num_input_tokens_seen": 9309568,
      "step": 10745
    },
    {
      "epoch": 5.068363979255069,
      "grad_norm": 0.47891828417778015,
      "learning_rate": 4.1607778625730104e-05,
      "loss": 0.1547,
      "num_input_tokens_seen": 9314048,
      "step": 10750
    },
    {
      "epoch": 5.07072135785007,
      "grad_norm": 0.9268267154693604,
      "learning_rate": 4.160043921171961e-05,
      "loss": 0.2257,
      "num_input_tokens_seen": 9318416,
      "step": 10755
    },
    {
      "epoch": 5.073078736445073,
      "grad_norm": 0.8876824975013733,
      "learning_rate": 4.159309723771276e-05,
      "loss": 0.2232,
      "num_input_tokens_seen": 9322880,
      "step": 10760
    },
    {
      "epoch": 5.075436115040075,
      "grad_norm": 2.275129795074463,
      "learning_rate": 4.158575270484181e-05,
      "loss": 0.2376,
      "num_input_tokens_seen": 9327584,
      "step": 10765
    },
    {
      "epoch": 5.0777934936350775,
      "grad_norm": 2.090752124786377,
      "learning_rate": 4.157840561423936e-05,
      "loss": 0.2024,
      "num_input_tokens_seen": 9331984,
      "step": 10770
    },
    {
      "epoch": 5.08015087223008,
      "grad_norm": 0.6983162760734558,
      "learning_rate": 4.1571055967038416e-05,
      "loss": 0.1646,
      "num_input_tokens_seen": 9336384,
      "step": 10775
    },
    {
      "epoch": 5.082508250825082,
      "grad_norm": 0.567505419254303,
      "learning_rate": 4.156370376437241e-05,
      "loss": 0.2378,
      "num_input_tokens_seen": 9340352,
      "step": 10780
    },
    {
      "epoch": 5.084865629420085,
      "grad_norm": 0.8282793164253235,
      "learning_rate": 4.155634900737513e-05,
      "loss": 0.1918,
      "num_input_tokens_seen": 9344656,
      "step": 10785
    },
    {
      "epoch": 5.087223008015087,
      "grad_norm": 0.5430209636688232,
      "learning_rate": 4.1548991697180764e-05,
      "loss": 0.2022,
      "num_input_tokens_seen": 9349920,
      "step": 10790
    },
    {
      "epoch": 5.0895803866100895,
      "grad_norm": 1.7769014835357666,
      "learning_rate": 4.1541631834923914e-05,
      "loss": 0.1989,
      "num_input_tokens_seen": 9354032,
      "step": 10795
    },
    {
      "epoch": 5.091937765205092,
      "grad_norm": 0.7376298904418945,
      "learning_rate": 4.153426942173956e-05,
      "loss": 0.2037,
      "num_input_tokens_seen": 9358160,
      "step": 10800
    },
    {
      "epoch": 5.091937765205092,
      "eval_loss": 0.20674607157707214,
      "eval_runtime": 22.1521,
      "eval_samples_per_second": 42.569,
      "eval_steps_per_second": 21.307,
      "num_input_tokens_seen": 9358160,
      "step": 10800
    },
    {
      "epoch": 5.094295143800094,
      "grad_norm": 0.6556090712547302,
      "learning_rate": 4.152690445876308e-05,
      "loss": 0.214,
      "num_input_tokens_seen": 9362512,
      "step": 10805
    },
    {
      "epoch": 5.096652522395097,
      "grad_norm": 1.506831169128418,
      "learning_rate": 4.1519536947130245e-05,
      "loss": 0.1828,
      "num_input_tokens_seen": 9366928,
      "step": 10810
    },
    {
      "epoch": 5.099009900990099,
      "grad_norm": 1.077331781387329,
      "learning_rate": 4.151216688797722e-05,
      "loss": 0.1365,
      "num_input_tokens_seen": 9371136,
      "step": 10815
    },
    {
      "epoch": 5.1013672795851015,
      "grad_norm": 0.9751758575439453,
      "learning_rate": 4.150479428244054e-05,
      "loss": 0.1603,
      "num_input_tokens_seen": 9375264,
      "step": 10820
    },
    {
      "epoch": 5.103724658180104,
      "grad_norm": 0.8492812514305115,
      "learning_rate": 4.1497419131657176e-05,
      "loss": 0.2047,
      "num_input_tokens_seen": 9380128,
      "step": 10825
    },
    {
      "epoch": 5.106082036775106,
      "grad_norm": 0.8542137742042542,
      "learning_rate": 4.149004143676447e-05,
      "loss": 0.2016,
      "num_input_tokens_seen": 9384512,
      "step": 10830
    },
    {
      "epoch": 5.108439415370109,
      "grad_norm": 1.4984707832336426,
      "learning_rate": 4.148266119890015e-05,
      "loss": 0.1635,
      "num_input_tokens_seen": 9389424,
      "step": 10835
    },
    {
      "epoch": 5.110796793965111,
      "grad_norm": 0.7448017597198486,
      "learning_rate": 4.1475278419202324e-05,
      "loss": 0.2452,
      "num_input_tokens_seen": 9393344,
      "step": 10840
    },
    {
      "epoch": 5.1131541725601135,
      "grad_norm": 0.7040205597877502,
      "learning_rate": 4.146789309880953e-05,
      "loss": 0.192,
      "num_input_tokens_seen": 9397904,
      "step": 10845
    },
    {
      "epoch": 5.115511551155116,
      "grad_norm": 0.8843414187431335,
      "learning_rate": 4.146050523886068e-05,
      "loss": 0.1905,
      "num_input_tokens_seen": 9401776,
      "step": 10850
    },
    {
      "epoch": 5.117868929750118,
      "grad_norm": 0.44692838191986084,
      "learning_rate": 4.1453114840495055e-05,
      "loss": 0.2343,
      "num_input_tokens_seen": 9405936,
      "step": 10855
    },
    {
      "epoch": 5.12022630834512,
      "grad_norm": 1.1773481369018555,
      "learning_rate": 4.1445721904852364e-05,
      "loss": 0.2559,
      "num_input_tokens_seen": 9410640,
      "step": 10860
    },
    {
      "epoch": 5.122583686940122,
      "grad_norm": 1.3487967252731323,
      "learning_rate": 4.143832643307269e-05,
      "loss": 0.1706,
      "num_input_tokens_seen": 9415120,
      "step": 10865
    },
    {
      "epoch": 5.124941065535125,
      "grad_norm": 0.6209455132484436,
      "learning_rate": 4.1430928426296503e-05,
      "loss": 0.17,
      "num_input_tokens_seen": 9419248,
      "step": 10870
    },
    {
      "epoch": 5.127298444130127,
      "grad_norm": 1.6166157722473145,
      "learning_rate": 4.142352788566466e-05,
      "loss": 0.1799,
      "num_input_tokens_seen": 9424032,
      "step": 10875
    },
    {
      "epoch": 5.129655822725129,
      "grad_norm": 0.567707896232605,
      "learning_rate": 4.1416124812318424e-05,
      "loss": 0.2122,
      "num_input_tokens_seen": 9428400,
      "step": 10880
    },
    {
      "epoch": 5.132013201320132,
      "grad_norm": 0.9041414260864258,
      "learning_rate": 4.1408719207399453e-05,
      "loss": 0.1724,
      "num_input_tokens_seen": 9432944,
      "step": 10885
    },
    {
      "epoch": 5.134370579915134,
      "grad_norm": 1.1498706340789795,
      "learning_rate": 4.140131107204978e-05,
      "loss": 0.2041,
      "num_input_tokens_seen": 9437792,
      "step": 10890
    },
    {
      "epoch": 5.136727958510137,
      "grad_norm": 0.4463474154472351,
      "learning_rate": 4.139390040741182e-05,
      "loss": 0.1533,
      "num_input_tokens_seen": 9441600,
      "step": 10895
    },
    {
      "epoch": 5.139085337105139,
      "grad_norm": 0.4720922112464905,
      "learning_rate": 4.1386487214628396e-05,
      "loss": 0.146,
      "num_input_tokens_seen": 9446064,
      "step": 10900
    },
    {
      "epoch": 5.141442715700141,
      "grad_norm": 0.38871437311172485,
      "learning_rate": 4.137907149484272e-05,
      "loss": 0.193,
      "num_input_tokens_seen": 9451536,
      "step": 10905
    },
    {
      "epoch": 5.143800094295144,
      "grad_norm": 0.7186244130134583,
      "learning_rate": 4.137165324919839e-05,
      "loss": 0.2444,
      "num_input_tokens_seen": 9455904,
      "step": 10910
    },
    {
      "epoch": 5.146157472890146,
      "grad_norm": 2.1677799224853516,
      "learning_rate": 4.136423247883939e-05,
      "loss": 0.1885,
      "num_input_tokens_seen": 9461824,
      "step": 10915
    },
    {
      "epoch": 5.148514851485149,
      "grad_norm": 1.2605106830596924,
      "learning_rate": 4.135680918491009e-05,
      "loss": 0.185,
      "num_input_tokens_seen": 9465440,
      "step": 10920
    },
    {
      "epoch": 5.150872230080151,
      "grad_norm": 1.158719539642334,
      "learning_rate": 4.1349383368555265e-05,
      "loss": 0.1792,
      "num_input_tokens_seen": 9470400,
      "step": 10925
    },
    {
      "epoch": 5.1532296086751535,
      "grad_norm": 0.40296173095703125,
      "learning_rate": 4.1341955030920065e-05,
      "loss": 0.1331,
      "num_input_tokens_seen": 9474496,
      "step": 10930
    },
    {
      "epoch": 5.155586987270156,
      "grad_norm": 1.7110768556594849,
      "learning_rate": 4.1334524173150036e-05,
      "loss": 0.1526,
      "num_input_tokens_seen": 9478784,
      "step": 10935
    },
    {
      "epoch": 5.157944365865158,
      "grad_norm": 1.285440444946289,
      "learning_rate": 4.13270907963911e-05,
      "loss": 0.1951,
      "num_input_tokens_seen": 9482448,
      "step": 10940
    },
    {
      "epoch": 5.160301744460161,
      "grad_norm": 1.8510833978652954,
      "learning_rate": 4.131965490178959e-05,
      "loss": 0.1585,
      "num_input_tokens_seen": 9486784,
      "step": 10945
    },
    {
      "epoch": 5.162659123055163,
      "grad_norm": 0.5654150247573853,
      "learning_rate": 4.131221649049222e-05,
      "loss": 0.2227,
      "num_input_tokens_seen": 9491216,
      "step": 10950
    },
    {
      "epoch": 5.165016501650165,
      "grad_norm": 1.1631520986557007,
      "learning_rate": 4.130477556364606e-05,
      "loss": 0.1329,
      "num_input_tokens_seen": 9495536,
      "step": 10955
    },
    {
      "epoch": 5.167373880245167,
      "grad_norm": 1.348781704902649,
      "learning_rate": 4.129733212239861e-05,
      "loss": 0.1717,
      "num_input_tokens_seen": 9499552,
      "step": 10960
    },
    {
      "epoch": 5.169731258840169,
      "grad_norm": 1.3408108949661255,
      "learning_rate": 4.128988616789774e-05,
      "loss": 0.1951,
      "num_input_tokens_seen": 9503712,
      "step": 10965
    },
    {
      "epoch": 5.172088637435172,
      "grad_norm": 1.0112155675888062,
      "learning_rate": 4.1282437701291724e-05,
      "loss": 0.1926,
      "num_input_tokens_seen": 9508544,
      "step": 10970
    },
    {
      "epoch": 5.174446016030174,
      "grad_norm": 1.0225086212158203,
      "learning_rate": 4.1274986723729184e-05,
      "loss": 0.1616,
      "num_input_tokens_seen": 9512704,
      "step": 10975
    },
    {
      "epoch": 5.176803394625177,
      "grad_norm": 0.9412850141525269,
      "learning_rate": 4.126753323635917e-05,
      "loss": 0.1049,
      "num_input_tokens_seen": 9517216,
      "step": 10980
    },
    {
      "epoch": 5.179160773220179,
      "grad_norm": 0.8697872161865234,
      "learning_rate": 4.12600772403311e-05,
      "loss": 0.1993,
      "num_input_tokens_seen": 9522416,
      "step": 10985
    },
    {
      "epoch": 5.181518151815181,
      "grad_norm": 0.7165787816047668,
      "learning_rate": 4.125261873679479e-05,
      "loss": 0.1653,
      "num_input_tokens_seen": 9526400,
      "step": 10990
    },
    {
      "epoch": 5.183875530410184,
      "grad_norm": 1.644213318824768,
      "learning_rate": 4.124515772690042e-05,
      "loss": 0.1894,
      "num_input_tokens_seen": 9531536,
      "step": 10995
    },
    {
      "epoch": 5.186232909005186,
      "grad_norm": 1.0763053894042969,
      "learning_rate": 4.123769421179858e-05,
      "loss": 0.1906,
      "num_input_tokens_seen": 9535520,
      "step": 11000
    },
    {
      "epoch": 5.186232909005186,
      "eval_loss": 0.20994095504283905,
      "eval_runtime": 22.1232,
      "eval_samples_per_second": 42.625,
      "eval_steps_per_second": 21.335,
      "num_input_tokens_seen": 9535520,
      "step": 11000
    },
    {
      "epoch": 5.188590287600189,
      "grad_norm": 0.3829050064086914,
      "learning_rate": 4.1230228192640236e-05,
      "loss": 0.2085,
      "num_input_tokens_seen": 9539696,
      "step": 11005
    },
    {
      "epoch": 5.190947666195191,
      "grad_norm": 2.020225763320923,
      "learning_rate": 4.122275967057675e-05,
      "loss": 0.1934,
      "num_input_tokens_seen": 9543792,
      "step": 11010
    },
    {
      "epoch": 5.193305044790193,
      "grad_norm": 0.5862178802490234,
      "learning_rate": 4.1215288646759846e-05,
      "loss": 0.1362,
      "num_input_tokens_seen": 9548896,
      "step": 11015
    },
    {
      "epoch": 5.195662423385196,
      "grad_norm": 1.2168549299240112,
      "learning_rate": 4.120781512234166e-05,
      "loss": 0.1602,
      "num_input_tokens_seen": 9552944,
      "step": 11020
    },
    {
      "epoch": 5.198019801980198,
      "grad_norm": 1.369523286819458,
      "learning_rate": 4.120033909847471e-05,
      "loss": 0.1849,
      "num_input_tokens_seen": 9557840,
      "step": 11025
    },
    {
      "epoch": 5.200377180575201,
      "grad_norm": 0.9188913702964783,
      "learning_rate": 4.119286057631187e-05,
      "loss": 0.2389,
      "num_input_tokens_seen": 9562064,
      "step": 11030
    },
    {
      "epoch": 5.202734559170203,
      "grad_norm": 1.9588191509246826,
      "learning_rate": 4.118537955700646e-05,
      "loss": 0.2751,
      "num_input_tokens_seen": 9566832,
      "step": 11035
    },
    {
      "epoch": 5.205091937765205,
      "grad_norm": 0.46442633867263794,
      "learning_rate": 4.11778960417121e-05,
      "loss": 0.2059,
      "num_input_tokens_seen": 9571520,
      "step": 11040
    },
    {
      "epoch": 5.207449316360208,
      "grad_norm": 1.12179696559906,
      "learning_rate": 4.117041003158288e-05,
      "loss": 0.2101,
      "num_input_tokens_seen": 9576272,
      "step": 11045
    },
    {
      "epoch": 5.20980669495521,
      "grad_norm": 1.2898526191711426,
      "learning_rate": 4.1162921527773215e-05,
      "loss": 0.176,
      "num_input_tokens_seen": 9580704,
      "step": 11050
    },
    {
      "epoch": 5.212164073550213,
      "grad_norm": 1.6840559244155884,
      "learning_rate": 4.115543053143794e-05,
      "loss": 0.2403,
      "num_input_tokens_seen": 9585712,
      "step": 11055
    },
    {
      "epoch": 5.214521452145214,
      "grad_norm": 0.7435351014137268,
      "learning_rate": 4.114793704373226e-05,
      "loss": 0.1834,
      "num_input_tokens_seen": 9590208,
      "step": 11060
    },
    {
      "epoch": 5.2168788307402165,
      "grad_norm": 0.5914081931114197,
      "learning_rate": 4.114044106581175e-05,
      "loss": 0.1771,
      "num_input_tokens_seen": 9594672,
      "step": 11065
    },
    {
      "epoch": 5.219236209335219,
      "grad_norm": 1.2988853454589844,
      "learning_rate": 4.11329425988324e-05,
      "loss": 0.2644,
      "num_input_tokens_seen": 9599856,
      "step": 11070
    },
    {
      "epoch": 5.221593587930221,
      "grad_norm": 0.5110063552856445,
      "learning_rate": 4.112544164395056e-05,
      "loss": 0.1922,
      "num_input_tokens_seen": 9603440,
      "step": 11075
    },
    {
      "epoch": 5.223950966525224,
      "grad_norm": 1.6210790872573853,
      "learning_rate": 4.111793820232297e-05,
      "loss": 0.2392,
      "num_input_tokens_seen": 9606960,
      "step": 11080
    },
    {
      "epoch": 5.226308345120226,
      "grad_norm": 0.45093926787376404,
      "learning_rate": 4.1110432275106767e-05,
      "loss": 0.1292,
      "num_input_tokens_seen": 9610752,
      "step": 11085
    },
    {
      "epoch": 5.2286657237152285,
      "grad_norm": 1.1028918027877808,
      "learning_rate": 4.110292386345944e-05,
      "loss": 0.1693,
      "num_input_tokens_seen": 9614880,
      "step": 11090
    },
    {
      "epoch": 5.231023102310231,
      "grad_norm": 0.3485613167285919,
      "learning_rate": 4.109541296853891e-05,
      "loss": 0.2421,
      "num_input_tokens_seen": 9619408,
      "step": 11095
    },
    {
      "epoch": 5.233380480905233,
      "grad_norm": 0.8188386559486389,
      "learning_rate": 4.108789959150341e-05,
      "loss": 0.1662,
      "num_input_tokens_seen": 9623216,
      "step": 11100
    },
    {
      "epoch": 5.235737859500236,
      "grad_norm": 1.3992844820022583,
      "learning_rate": 4.108038373351163e-05,
      "loss": 0.1768,
      "num_input_tokens_seen": 9628032,
      "step": 11105
    },
    {
      "epoch": 5.238095238095238,
      "grad_norm": 0.41031649708747864,
      "learning_rate": 4.10728653957226e-05,
      "loss": 0.1783,
      "num_input_tokens_seen": 9633072,
      "step": 11110
    },
    {
      "epoch": 5.2404526166902405,
      "grad_norm": 0.9363765716552734,
      "learning_rate": 4.106534457929575e-05,
      "loss": 0.1577,
      "num_input_tokens_seen": 9637600,
      "step": 11115
    },
    {
      "epoch": 5.242809995285243,
      "grad_norm": 0.31701377034187317,
      "learning_rate": 4.105782128539086e-05,
      "loss": 0.1559,
      "num_input_tokens_seen": 9641504,
      "step": 11120
    },
    {
      "epoch": 5.245167373880245,
      "grad_norm": 0.929710865020752,
      "learning_rate": 4.1050295515168144e-05,
      "loss": 0.2072,
      "num_input_tokens_seen": 9645696,
      "step": 11125
    },
    {
      "epoch": 5.247524752475248,
      "grad_norm": 1.614943027496338,
      "learning_rate": 4.1042767269788155e-05,
      "loss": 0.2303,
      "num_input_tokens_seen": 9650528,
      "step": 11130
    },
    {
      "epoch": 5.24988213107025,
      "grad_norm": 0.7634584903717041,
      "learning_rate": 4.103523655041185e-05,
      "loss": 0.2466,
      "num_input_tokens_seen": 9653888,
      "step": 11135
    },
    {
      "epoch": 5.2522395096652525,
      "grad_norm": 0.6043961048126221,
      "learning_rate": 4.102770335820055e-05,
      "loss": 0.1496,
      "num_input_tokens_seen": 9658768,
      "step": 11140
    },
    {
      "epoch": 5.254596888260255,
      "grad_norm": 0.6269164681434631,
      "learning_rate": 4.1020167694315984e-05,
      "loss": 0.1417,
      "num_input_tokens_seen": 9662976,
      "step": 11145
    },
    {
      "epoch": 5.256954266855257,
      "grad_norm": 1.495835542678833,
      "learning_rate": 4.101262955992023e-05,
      "loss": 0.112,
      "num_input_tokens_seen": 9667824,
      "step": 11150
    },
    {
      "epoch": 5.259311645450259,
      "grad_norm": 1.3638715744018555,
      "learning_rate": 4.100508895617578e-05,
      "loss": 0.1801,
      "num_input_tokens_seen": 9672864,
      "step": 11155
    },
    {
      "epoch": 5.261669024045261,
      "grad_norm": 1.5894666910171509,
      "learning_rate": 4.099754588424547e-05,
      "loss": 0.194,
      "num_input_tokens_seen": 9676944,
      "step": 11160
    },
    {
      "epoch": 5.264026402640264,
      "grad_norm": 1.1149674654006958,
      "learning_rate": 4.0990000345292546e-05,
      "loss": 0.2023,
      "num_input_tokens_seen": 9681024,
      "step": 11165
    },
    {
      "epoch": 5.266383781235266,
      "grad_norm": 1.2653895616531372,
      "learning_rate": 4.098245234048064e-05,
      "loss": 0.2685,
      "num_input_tokens_seen": 9685936,
      "step": 11170
    },
    {
      "epoch": 5.2687411598302685,
      "grad_norm": 1.5140509605407715,
      "learning_rate": 4.0974901870973726e-05,
      "loss": 0.1457,
      "num_input_tokens_seen": 9689376,
      "step": 11175
    },
    {
      "epoch": 5.271098538425271,
      "grad_norm": 1.8035789728164673,
      "learning_rate": 4.096734893793619e-05,
      "loss": 0.2241,
      "num_input_tokens_seen": 9693728,
      "step": 11180
    },
    {
      "epoch": 5.273455917020273,
      "grad_norm": 1.457481026649475,
      "learning_rate": 4.095979354253279e-05,
      "loss": 0.2736,
      "num_input_tokens_seen": 9697952,
      "step": 11185
    },
    {
      "epoch": 5.275813295615276,
      "grad_norm": 0.8751828074455261,
      "learning_rate": 4.0952235685928656e-05,
      "loss": 0.1718,
      "num_input_tokens_seen": 9701840,
      "step": 11190
    },
    {
      "epoch": 5.278170674210278,
      "grad_norm": 0.5193503499031067,
      "learning_rate": 4.094467536928932e-05,
      "loss": 0.1122,
      "num_input_tokens_seen": 9705248,
      "step": 11195
    },
    {
      "epoch": 5.2805280528052805,
      "grad_norm": 2.3148317337036133,
      "learning_rate": 4.093711259378067e-05,
      "loss": 0.2102,
      "num_input_tokens_seen": 9709232,
      "step": 11200
    },
    {
      "epoch": 5.2805280528052805,
      "eval_loss": 0.20765407383441925,
      "eval_runtime": 22.1635,
      "eval_samples_per_second": 42.548,
      "eval_steps_per_second": 21.296,
      "num_input_tokens_seen": 9709232,
      "step": 11200
    },
    {
      "epoch": 5.282885431400283,
      "grad_norm": 0.785564661026001,
      "learning_rate": 4.092954736056897e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 9712928,
      "step": 11205
    },
    {
      "epoch": 5.285242809995285,
      "grad_norm": 1.8761591911315918,
      "learning_rate": 4.09219796708209e-05,
      "loss": 0.2696,
      "num_input_tokens_seen": 9717616,
      "step": 11210
    },
    {
      "epoch": 5.287600188590288,
      "grad_norm": 0.795274555683136,
      "learning_rate": 4.0914409525703464e-05,
      "loss": 0.1998,
      "num_input_tokens_seen": 9721968,
      "step": 11215
    },
    {
      "epoch": 5.28995756718529,
      "grad_norm": 1.234602689743042,
      "learning_rate": 4.090683692638408e-05,
      "loss": 0.1744,
      "num_input_tokens_seen": 9726688,
      "step": 11220
    },
    {
      "epoch": 5.2923149457802925,
      "grad_norm": 0.7403039336204529,
      "learning_rate": 4.089926187403056e-05,
      "loss": 0.1788,
      "num_input_tokens_seen": 9730912,
      "step": 11225
    },
    {
      "epoch": 5.294672324375295,
      "grad_norm": 2.2014448642730713,
      "learning_rate": 4.0891684369811044e-05,
      "loss": 0.1596,
      "num_input_tokens_seen": 9735136,
      "step": 11230
    },
    {
      "epoch": 5.297029702970297,
      "grad_norm": 1.3095670938491821,
      "learning_rate": 4.0884104414894107e-05,
      "loss": 0.2113,
      "num_input_tokens_seen": 9739568,
      "step": 11235
    },
    {
      "epoch": 5.2993870815653,
      "grad_norm": 0.7175862193107605,
      "learning_rate": 4.087652201044864e-05,
      "loss": 0.1662,
      "num_input_tokens_seen": 9744976,
      "step": 11240
    },
    {
      "epoch": 5.301744460160302,
      "grad_norm": 0.7250401973724365,
      "learning_rate": 4.086893715764397e-05,
      "loss": 0.1331,
      "num_input_tokens_seen": 9748864,
      "step": 11245
    },
    {
      "epoch": 5.3041018387553045,
      "grad_norm": 1.1396268606185913,
      "learning_rate": 4.086134985764977e-05,
      "loss": 0.1959,
      "num_input_tokens_seen": 9753840,
      "step": 11250
    },
    {
      "epoch": 5.306459217350307,
      "grad_norm": 0.8099697828292847,
      "learning_rate": 4.0853760111636085e-05,
      "loss": 0.1472,
      "num_input_tokens_seen": 9758048,
      "step": 11255
    },
    {
      "epoch": 5.308816595945308,
      "grad_norm": 1.30785071849823,
      "learning_rate": 4.084616792077337e-05,
      "loss": 0.2105,
      "num_input_tokens_seen": 9762000,
      "step": 11260
    },
    {
      "epoch": 5.311173974540311,
      "grad_norm": 0.6504552960395813,
      "learning_rate": 4.083857328623243e-05,
      "loss": 0.2276,
      "num_input_tokens_seen": 9766112,
      "step": 11265
    },
    {
      "epoch": 5.313531353135313,
      "grad_norm": 0.6929129958152771,
      "learning_rate": 4.083097620918444e-05,
      "loss": 0.1746,
      "num_input_tokens_seen": 9770160,
      "step": 11270
    },
    {
      "epoch": 5.315888731730316,
      "grad_norm": 0.8478732109069824,
      "learning_rate": 4.082337669080097e-05,
      "loss": 0.1901,
      "num_input_tokens_seen": 9774816,
      "step": 11275
    },
    {
      "epoch": 5.318246110325318,
      "grad_norm": 0.4637429118156433,
      "learning_rate": 4.081577473225398e-05,
      "loss": 0.1621,
      "num_input_tokens_seen": 9778672,
      "step": 11280
    },
    {
      "epoch": 5.32060348892032,
      "grad_norm": 0.3697223365306854,
      "learning_rate": 4.080817033471577e-05,
      "loss": 0.1621,
      "num_input_tokens_seen": 9782000,
      "step": 11285
    },
    {
      "epoch": 5.322960867515323,
      "grad_norm": 1.3312121629714966,
      "learning_rate": 4.080056349935903e-05,
      "loss": 0.2442,
      "num_input_tokens_seen": 9786048,
      "step": 11290
    },
    {
      "epoch": 5.325318246110325,
      "grad_norm": 0.7722378969192505,
      "learning_rate": 4.079295422735684e-05,
      "loss": 0.214,
      "num_input_tokens_seen": 9790592,
      "step": 11295
    },
    {
      "epoch": 5.327675624705328,
      "grad_norm": 0.8809412717819214,
      "learning_rate": 4.078534251988264e-05,
      "loss": 0.1724,
      "num_input_tokens_seen": 9794752,
      "step": 11300
    },
    {
      "epoch": 5.33003300330033,
      "grad_norm": 1.0129324197769165,
      "learning_rate": 4.077772837811025e-05,
      "loss": 0.1666,
      "num_input_tokens_seen": 9799664,
      "step": 11305
    },
    {
      "epoch": 5.332390381895332,
      "grad_norm": 0.6636875867843628,
      "learning_rate": 4.0770111803213874e-05,
      "loss": 0.1681,
      "num_input_tokens_seen": 9804048,
      "step": 11310
    },
    {
      "epoch": 5.334747760490335,
      "grad_norm": 1.2742681503295898,
      "learning_rate": 4.076249279636807e-05,
      "loss": 0.1648,
      "num_input_tokens_seen": 9808992,
      "step": 11315
    },
    {
      "epoch": 5.337105139085337,
      "grad_norm": 0.8753578066825867,
      "learning_rate": 4.075487135874781e-05,
      "loss": 0.2176,
      "num_input_tokens_seen": 9813104,
      "step": 11320
    },
    {
      "epoch": 5.33946251768034,
      "grad_norm": 2.1635794639587402,
      "learning_rate": 4.074724749152837e-05,
      "loss": 0.2424,
      "num_input_tokens_seen": 9817456,
      "step": 11325
    },
    {
      "epoch": 5.341819896275342,
      "grad_norm": 1.1817541122436523,
      "learning_rate": 4.07396211958855e-05,
      "loss": 0.2044,
      "num_input_tokens_seen": 9822336,
      "step": 11330
    },
    {
      "epoch": 5.344177274870344,
      "grad_norm": 1.5956143140792847,
      "learning_rate": 4.073199247299523e-05,
      "loss": 0.2068,
      "num_input_tokens_seen": 9827168,
      "step": 11335
    },
    {
      "epoch": 5.346534653465347,
      "grad_norm": 1.048228144645691,
      "learning_rate": 4.072436132403403e-05,
      "loss": 0.1555,
      "num_input_tokens_seen": 9830928,
      "step": 11340
    },
    {
      "epoch": 5.348892032060349,
      "grad_norm": 0.9426819682121277,
      "learning_rate": 4.0716727750178704e-05,
      "loss": 0.1704,
      "num_input_tokens_seen": 9834848,
      "step": 11345
    },
    {
      "epoch": 5.351249410655352,
      "grad_norm": 0.45162898302078247,
      "learning_rate": 4.0709091752606455e-05,
      "loss": 0.1133,
      "num_input_tokens_seen": 9838512,
      "step": 11350
    },
    {
      "epoch": 5.353606789250353,
      "grad_norm": 1.5469090938568115,
      "learning_rate": 4.070145333249484e-05,
      "loss": 0.2082,
      "num_input_tokens_seen": 9842640,
      "step": 11355
    },
    {
      "epoch": 5.355964167845356,
      "grad_norm": 1.4556591510772705,
      "learning_rate": 4.069381249102181e-05,
      "loss": 0.1928,
      "num_input_tokens_seen": 9847904,
      "step": 11360
    },
    {
      "epoch": 5.358321546440358,
      "grad_norm": 0.5777970552444458,
      "learning_rate": 4.0686169229365665e-05,
      "loss": 0.1861,
      "num_input_tokens_seen": 9851712,
      "step": 11365
    },
    {
      "epoch": 5.36067892503536,
      "grad_norm": 1.0813874006271362,
      "learning_rate": 4.067852354870511e-05,
      "loss": 0.1113,
      "num_input_tokens_seen": 9854816,
      "step": 11370
    },
    {
      "epoch": 5.363036303630363,
      "grad_norm": 2.562901735305786,
      "learning_rate": 4.067087545021919e-05,
      "loss": 0.2514,
      "num_input_tokens_seen": 9859664,
      "step": 11375
    },
    {
      "epoch": 5.365393682225365,
      "grad_norm": 1.3654563426971436,
      "learning_rate": 4.066322493508734e-05,
      "loss": 0.1782,
      "num_input_tokens_seen": 9863440,
      "step": 11380
    },
    {
      "epoch": 5.367751060820368,
      "grad_norm": 0.6508436799049377,
      "learning_rate": 4.065557200448937e-05,
      "loss": 0.1393,
      "num_input_tokens_seen": 9867600,
      "step": 11385
    },
    {
      "epoch": 5.37010843941537,
      "grad_norm": 1.040199875831604,
      "learning_rate": 4.064791665960546e-05,
      "loss": 0.166,
      "num_input_tokens_seen": 9872208,
      "step": 11390
    },
    {
      "epoch": 5.372465818010372,
      "grad_norm": 0.9681344032287598,
      "learning_rate": 4.064025890161615e-05,
      "loss": 0.1451,
      "num_input_tokens_seen": 9876544,
      "step": 11395
    },
    {
      "epoch": 5.374823196605375,
      "grad_norm": 1.5451513528823853,
      "learning_rate": 4.0632598731702373e-05,
      "loss": 0.1786,
      "num_input_tokens_seen": 9880896,
      "step": 11400
    },
    {
      "epoch": 5.374823196605375,
      "eval_loss": 0.20955561101436615,
      "eval_runtime": 22.1774,
      "eval_samples_per_second": 42.521,
      "eval_steps_per_second": 21.283,
      "num_input_tokens_seen": 9880896,
      "step": 11400
    },
    {
      "epoch": 5.377180575200377,
      "grad_norm": 0.9525657892227173,
      "learning_rate": 4.0624936151045426e-05,
      "loss": 0.1257,
      "num_input_tokens_seen": 9885840,
      "step": 11405
    },
    {
      "epoch": 5.37953795379538,
      "grad_norm": 1.8727149963378906,
      "learning_rate": 4.061727116082696e-05,
      "loss": 0.1867,
      "num_input_tokens_seen": 9889856,
      "step": 11410
    },
    {
      "epoch": 5.381895332390382,
      "grad_norm": 1.0558555126190186,
      "learning_rate": 4.060960376222903e-05,
      "loss": 0.1731,
      "num_input_tokens_seen": 9894976,
      "step": 11415
    },
    {
      "epoch": 5.384252710985384,
      "grad_norm": 1.0984615087509155,
      "learning_rate": 4.0601933956434034e-05,
      "loss": 0.1167,
      "num_input_tokens_seen": 9898720,
      "step": 11420
    },
    {
      "epoch": 5.386610089580387,
      "grad_norm": 0.3106929361820221,
      "learning_rate": 4.059426174462476e-05,
      "loss": 0.1512,
      "num_input_tokens_seen": 9902144,
      "step": 11425
    },
    {
      "epoch": 5.388967468175389,
      "grad_norm": 0.5089482069015503,
      "learning_rate": 4.058658712798435e-05,
      "loss": 0.0944,
      "num_input_tokens_seen": 9906016,
      "step": 11430
    },
    {
      "epoch": 5.391324846770392,
      "grad_norm": 2.186206102371216,
      "learning_rate": 4.0578910107696336e-05,
      "loss": 0.1949,
      "num_input_tokens_seen": 9910160,
      "step": 11435
    },
    {
      "epoch": 5.393682225365394,
      "grad_norm": 1.0594812631607056,
      "learning_rate": 4.05712306849446e-05,
      "loss": 0.167,
      "num_input_tokens_seen": 9914992,
      "step": 11440
    },
    {
      "epoch": 5.396039603960396,
      "grad_norm": 0.8656353950500488,
      "learning_rate": 4.0563548860913415e-05,
      "loss": 0.1585,
      "num_input_tokens_seen": 9919824,
      "step": 11445
    },
    {
      "epoch": 5.398396982555399,
      "grad_norm": 1.3970370292663574,
      "learning_rate": 4.0555864636787414e-05,
      "loss": 0.2025,
      "num_input_tokens_seen": 9923184,
      "step": 11450
    },
    {
      "epoch": 5.400754361150401,
      "grad_norm": 1.0576937198638916,
      "learning_rate": 4.054817801375159e-05,
      "loss": 0.1985,
      "num_input_tokens_seen": 9926976,
      "step": 11455
    },
    {
      "epoch": 5.403111739745403,
      "grad_norm": 0.5862334966659546,
      "learning_rate": 4.054048899299134e-05,
      "loss": 0.1626,
      "num_input_tokens_seen": 9931280,
      "step": 11460
    },
    {
      "epoch": 5.405469118340405,
      "grad_norm": 1.0070288181304932,
      "learning_rate": 4.0532797575692385e-05,
      "loss": 0.3208,
      "num_input_tokens_seen": 9935376,
      "step": 11465
    },
    {
      "epoch": 5.4078264969354075,
      "grad_norm": 1.6788376569747925,
      "learning_rate": 4.052510376304085e-05,
      "loss": 0.2318,
      "num_input_tokens_seen": 9939616,
      "step": 11470
    },
    {
      "epoch": 5.41018387553041,
      "grad_norm": 0.6627516150474548,
      "learning_rate": 4.051740755622321e-05,
      "loss": 0.1148,
      "num_input_tokens_seen": 9943808,
      "step": 11475
    },
    {
      "epoch": 5.412541254125412,
      "grad_norm": 1.1471949815750122,
      "learning_rate": 4.050970895642632e-05,
      "loss": 0.1909,
      "num_input_tokens_seen": 9948352,
      "step": 11480
    },
    {
      "epoch": 5.414898632720415,
      "grad_norm": 0.5170823335647583,
      "learning_rate": 4.050200796483741e-05,
      "loss": 0.1102,
      "num_input_tokens_seen": 9952704,
      "step": 11485
    },
    {
      "epoch": 5.417256011315417,
      "grad_norm": 1.3119537830352783,
      "learning_rate": 4.049430458264405e-05,
      "loss": 0.2018,
      "num_input_tokens_seen": 9957440,
      "step": 11490
    },
    {
      "epoch": 5.4196133899104195,
      "grad_norm": 1.263456106185913,
      "learning_rate": 4.048659881103422e-05,
      "loss": 0.1822,
      "num_input_tokens_seen": 9961248,
      "step": 11495
    },
    {
      "epoch": 5.421970768505422,
      "grad_norm": 0.5808209180831909,
      "learning_rate": 4.0478890651196235e-05,
      "loss": 0.1551,
      "num_input_tokens_seen": 9965520,
      "step": 11500
    },
    {
      "epoch": 5.424328147100424,
      "grad_norm": 1.2986466884613037,
      "learning_rate": 4.047118010431879e-05,
      "loss": 0.1535,
      "num_input_tokens_seen": 9970192,
      "step": 11505
    },
    {
      "epoch": 5.426685525695427,
      "grad_norm": 1.0570777654647827,
      "learning_rate": 4.046346717159094e-05,
      "loss": 0.1843,
      "num_input_tokens_seen": 9974320,
      "step": 11510
    },
    {
      "epoch": 5.429042904290429,
      "grad_norm": 0.47005653381347656,
      "learning_rate": 4.045575185420214e-05,
      "loss": 0.1456,
      "num_input_tokens_seen": 9978752,
      "step": 11515
    },
    {
      "epoch": 5.4314002828854315,
      "grad_norm": 1.4330693483352661,
      "learning_rate": 4.0448034153342165e-05,
      "loss": 0.2979,
      "num_input_tokens_seen": 9982768,
      "step": 11520
    },
    {
      "epoch": 5.433757661480434,
      "grad_norm": 0.7745643854141235,
      "learning_rate": 4.0440314070201194e-05,
      "loss": 0.1711,
      "num_input_tokens_seen": 9987088,
      "step": 11525
    },
    {
      "epoch": 5.436115040075436,
      "grad_norm": 0.503913938999176,
      "learning_rate": 4.043259160596976e-05,
      "loss": 0.1354,
      "num_input_tokens_seen": 9991600,
      "step": 11530
    },
    {
      "epoch": 5.438472418670439,
      "grad_norm": 0.8275750875473022,
      "learning_rate": 4.0424866761838767e-05,
      "loss": 0.119,
      "num_input_tokens_seen": 9996464,
      "step": 11535
    },
    {
      "epoch": 5.440829797265441,
      "grad_norm": 0.8631144762039185,
      "learning_rate": 4.041713953899948e-05,
      "loss": 0.1518,
      "num_input_tokens_seen": 10001024,
      "step": 11540
    },
    {
      "epoch": 5.4431871758604435,
      "grad_norm": 0.7858003377914429,
      "learning_rate": 4.0409409938643515e-05,
      "loss": 0.1676,
      "num_input_tokens_seen": 10005488,
      "step": 11545
    },
    {
      "epoch": 5.445544554455446,
      "grad_norm": 1.1469264030456543,
      "learning_rate": 4.0401677961962904e-05,
      "loss": 0.2277,
      "num_input_tokens_seen": 10009808,
      "step": 11550
    },
    {
      "epoch": 5.4479019330504475,
      "grad_norm": 0.4443369209766388,
      "learning_rate": 4.039394361015001e-05,
      "loss": 0.2012,
      "num_input_tokens_seen": 10013664,
      "step": 11555
    },
    {
      "epoch": 5.45025931164545,
      "grad_norm": 0.7505351901054382,
      "learning_rate": 4.038620688439755e-05,
      "loss": 0.1992,
      "num_input_tokens_seen": 10018080,
      "step": 11560
    },
    {
      "epoch": 5.452616690240452,
      "grad_norm": 0.6407216191291809,
      "learning_rate": 4.037846778589862e-05,
      "loss": 0.1276,
      "num_input_tokens_seen": 10022272,
      "step": 11565
    },
    {
      "epoch": 5.454974068835455,
      "grad_norm": 1.0628997087478638,
      "learning_rate": 4.0370726315846715e-05,
      "loss": 0.1309,
      "num_input_tokens_seen": 10027568,
      "step": 11570
    },
    {
      "epoch": 5.457331447430457,
      "grad_norm": 1.2210736274719238,
      "learning_rate": 4.036298247543565e-05,
      "loss": 0.2007,
      "num_input_tokens_seen": 10032192,
      "step": 11575
    },
    {
      "epoch": 5.4596888260254595,
      "grad_norm": 1.2974281311035156,
      "learning_rate": 4.035523626585962e-05,
      "loss": 0.2277,
      "num_input_tokens_seen": 10036048,
      "step": 11580
    },
    {
      "epoch": 5.462046204620462,
      "grad_norm": 1.282546877861023,
      "learning_rate": 4.0347487688313194e-05,
      "loss": 0.2642,
      "num_input_tokens_seen": 10041088,
      "step": 11585
    },
    {
      "epoch": 5.464403583215464,
      "grad_norm": 0.5729175806045532,
      "learning_rate": 4.0339736743991296e-05,
      "loss": 0.2142,
      "num_input_tokens_seen": 10044688,
      "step": 11590
    },
    {
      "epoch": 5.466760961810467,
      "grad_norm": 1.6085997819900513,
      "learning_rate": 4.0331983434089227e-05,
      "loss": 0.2305,
      "num_input_tokens_seen": 10048928,
      "step": 11595
    },
    {
      "epoch": 5.469118340405469,
      "grad_norm": 1.2166343927383423,
      "learning_rate": 4.032422775980264e-05,
      "loss": 0.1862,
      "num_input_tokens_seen": 10053056,
      "step": 11600
    },
    {
      "epoch": 5.469118340405469,
      "eval_loss": 0.21005617082118988,
      "eval_runtime": 22.1775,
      "eval_samples_per_second": 42.52,
      "eval_steps_per_second": 21.283,
      "num_input_tokens_seen": 10053056,
      "step": 11600
    },
    {
      "epoch": 5.4714757190004715,
      "grad_norm": 1.1305170059204102,
      "learning_rate": 4.031646972232754e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 10057584,
      "step": 11605
    },
    {
      "epoch": 5.473833097595474,
      "grad_norm": 1.3272002935409546,
      "learning_rate": 4.0308709322860344e-05,
      "loss": 0.2037,
      "num_input_tokens_seen": 10062016,
      "step": 11610
    },
    {
      "epoch": 5.476190476190476,
      "grad_norm": 0.9068134427070618,
      "learning_rate": 4.0300946562597784e-05,
      "loss": 0.1419,
      "num_input_tokens_seen": 10067376,
      "step": 11615
    },
    {
      "epoch": 5.478547854785479,
      "grad_norm": 1.4064759016036987,
      "learning_rate": 4.029318144273698e-05,
      "loss": 0.2486,
      "num_input_tokens_seen": 10071568,
      "step": 11620
    },
    {
      "epoch": 5.480905233380481,
      "grad_norm": 0.6671514511108398,
      "learning_rate": 4.0285413964475415e-05,
      "loss": 0.1067,
      "num_input_tokens_seen": 10074976,
      "step": 11625
    },
    {
      "epoch": 5.4832626119754835,
      "grad_norm": 1.4371196031570435,
      "learning_rate": 4.0277644129010927e-05,
      "loss": 0.1913,
      "num_input_tokens_seen": 10079776,
      "step": 11630
    },
    {
      "epoch": 5.485619990570486,
      "grad_norm": 0.5436498522758484,
      "learning_rate": 4.0269871937541724e-05,
      "loss": 0.1517,
      "num_input_tokens_seen": 10084416,
      "step": 11635
    },
    {
      "epoch": 5.487977369165488,
      "grad_norm": 0.9260357022285461,
      "learning_rate": 4.026209739126637e-05,
      "loss": 0.2019,
      "num_input_tokens_seen": 10089536,
      "step": 11640
    },
    {
      "epoch": 5.490334747760491,
      "grad_norm": 0.4737558662891388,
      "learning_rate": 4.025432049138381e-05,
      "loss": 0.3035,
      "num_input_tokens_seen": 10094112,
      "step": 11645
    },
    {
      "epoch": 5.492692126355493,
      "grad_norm": 1.6712003946304321,
      "learning_rate": 4.0246541239093325e-05,
      "loss": 0.2315,
      "num_input_tokens_seen": 10098336,
      "step": 11650
    },
    {
      "epoch": 5.4950495049504955,
      "grad_norm": 0.4427454173564911,
      "learning_rate": 4.023875963559459e-05,
      "loss": 0.1184,
      "num_input_tokens_seen": 10102144,
      "step": 11655
    },
    {
      "epoch": 5.497406883545497,
      "grad_norm": 0.7522137761116028,
      "learning_rate": 4.023097568208761e-05,
      "loss": 0.2125,
      "num_input_tokens_seen": 10106240,
      "step": 11660
    },
    {
      "epoch": 5.499764262140499,
      "grad_norm": 2.2883005142211914,
      "learning_rate": 4.022318937977277e-05,
      "loss": 0.2839,
      "num_input_tokens_seen": 10111296,
      "step": 11665
    },
    {
      "epoch": 5.502121640735502,
      "grad_norm": 0.9815599322319031,
      "learning_rate": 4.021540072985084e-05,
      "loss": 0.227,
      "num_input_tokens_seen": 10115280,
      "step": 11670
    },
    {
      "epoch": 5.504479019330504,
      "grad_norm": 0.7159522771835327,
      "learning_rate": 4.020760973352289e-05,
      "loss": 0.2702,
      "num_input_tokens_seen": 10118896,
      "step": 11675
    },
    {
      "epoch": 5.506836397925507,
      "grad_norm": 0.6220460534095764,
      "learning_rate": 4.019981639199042e-05,
      "loss": 0.2324,
      "num_input_tokens_seen": 10123360,
      "step": 11680
    },
    {
      "epoch": 5.509193776520509,
      "grad_norm": 0.6515495181083679,
      "learning_rate": 4.0192020706455245e-05,
      "loss": 0.1731,
      "num_input_tokens_seen": 10127584,
      "step": 11685
    },
    {
      "epoch": 5.511551155115511,
      "grad_norm": 1.5716525316238403,
      "learning_rate": 4.018422267811956e-05,
      "loss": 0.2129,
      "num_input_tokens_seen": 10132480,
      "step": 11690
    },
    {
      "epoch": 5.513908533710514,
      "grad_norm": 1.3312780857086182,
      "learning_rate": 4.017642230818592e-05,
      "loss": 0.2655,
      "num_input_tokens_seen": 10137040,
      "step": 11695
    },
    {
      "epoch": 5.516265912305516,
      "grad_norm": 0.8898484110832214,
      "learning_rate": 4.0168619597857246e-05,
      "loss": 0.1802,
      "num_input_tokens_seen": 10141984,
      "step": 11700
    },
    {
      "epoch": 5.518623290900519,
      "grad_norm": 0.8158263564109802,
      "learning_rate": 4.016081454833681e-05,
      "loss": 0.1636,
      "num_input_tokens_seen": 10146880,
      "step": 11705
    },
    {
      "epoch": 5.520980669495521,
      "grad_norm": 1.310509443283081,
      "learning_rate": 4.0153007160828245e-05,
      "loss": 0.1906,
      "num_input_tokens_seen": 10150880,
      "step": 11710
    },
    {
      "epoch": 5.523338048090523,
      "grad_norm": 0.4462392330169678,
      "learning_rate": 4.0145197436535555e-05,
      "loss": 0.1474,
      "num_input_tokens_seen": 10155184,
      "step": 11715
    },
    {
      "epoch": 5.525695426685526,
      "grad_norm": 1.5872396230697632,
      "learning_rate": 4.0137385376663095e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 10159040,
      "step": 11720
    },
    {
      "epoch": 5.528052805280528,
      "grad_norm": 1.4725518226623535,
      "learning_rate": 4.012957098241558e-05,
      "loss": 0.2834,
      "num_input_tokens_seen": 10162880,
      "step": 11725
    },
    {
      "epoch": 5.530410183875531,
      "grad_norm": 1.2138155698776245,
      "learning_rate": 4.0121754254998076e-05,
      "loss": 0.202,
      "num_input_tokens_seen": 10167664,
      "step": 11730
    },
    {
      "epoch": 5.532767562470533,
      "grad_norm": 0.4234617054462433,
      "learning_rate": 4.011393519561606e-05,
      "loss": 0.1945,
      "num_input_tokens_seen": 10171808,
      "step": 11735
    },
    {
      "epoch": 5.535124941065535,
      "grad_norm": 0.8614006638526917,
      "learning_rate": 4.010611380547529e-05,
      "loss": 0.1698,
      "num_input_tokens_seen": 10176064,
      "step": 11740
    },
    {
      "epoch": 5.537482319660538,
      "grad_norm": 0.9512196779251099,
      "learning_rate": 4.009829008578192e-05,
      "loss": 0.2351,
      "num_input_tokens_seen": 10181328,
      "step": 11745
    },
    {
      "epoch": 5.539839698255539,
      "grad_norm": 1.0795565843582153,
      "learning_rate": 4.00904640377425e-05,
      "loss": 0.144,
      "num_input_tokens_seen": 10185088,
      "step": 11750
    },
    {
      "epoch": 5.542197076850542,
      "grad_norm": 1.6764849424362183,
      "learning_rate": 4.0082635662563886e-05,
      "loss": 0.2122,
      "num_input_tokens_seen": 10189440,
      "step": 11755
    },
    {
      "epoch": 5.544554455445544,
      "grad_norm": 1.3432613611221313,
      "learning_rate": 4.007480496145331e-05,
      "loss": 0.245,
      "num_input_tokens_seen": 10194096,
      "step": 11760
    },
    {
      "epoch": 5.5469118340405466,
      "grad_norm": 2.244154214859009,
      "learning_rate": 4.006697193561837e-05,
      "loss": 0.2124,
      "num_input_tokens_seen": 10198240,
      "step": 11765
    },
    {
      "epoch": 5.549269212635549,
      "grad_norm": 0.6187487244606018,
      "learning_rate": 4.005913658626701e-05,
      "loss": 0.1888,
      "num_input_tokens_seen": 10202640,
      "step": 11770
    },
    {
      "epoch": 5.551626591230551,
      "grad_norm": 1.7667067050933838,
      "learning_rate": 4.005129891460754e-05,
      "loss": 0.191,
      "num_input_tokens_seen": 10206672,
      "step": 11775
    },
    {
      "epoch": 5.553983969825554,
      "grad_norm": 0.6310191750526428,
      "learning_rate": 4.004345892184864e-05,
      "loss": 0.1226,
      "num_input_tokens_seen": 10210720,
      "step": 11780
    },
    {
      "epoch": 5.556341348420556,
      "grad_norm": 0.907402515411377,
      "learning_rate": 4.003561660919932e-05,
      "loss": 0.2225,
      "num_input_tokens_seen": 10215152,
      "step": 11785
    },
    {
      "epoch": 5.558698727015559,
      "grad_norm": 0.891759991645813,
      "learning_rate": 4.002777197786897e-05,
      "loss": 0.1521,
      "num_input_tokens_seen": 10219536,
      "step": 11790
    },
    {
      "epoch": 5.561056105610561,
      "grad_norm": 0.8022083640098572,
      "learning_rate": 4.0019925029067326e-05,
      "loss": 0.1937,
      "num_input_tokens_seen": 10225104,
      "step": 11795
    },
    {
      "epoch": 5.563413484205563,
      "grad_norm": 1.9486761093139648,
      "learning_rate": 4.0012075764004495e-05,
      "loss": 0.2268,
      "num_input_tokens_seen": 10229152,
      "step": 11800
    },
    {
      "epoch": 5.563413484205563,
      "eval_loss": 0.2064814418554306,
      "eval_runtime": 22.1499,
      "eval_samples_per_second": 42.574,
      "eval_steps_per_second": 21.309,
      "num_input_tokens_seen": 10229152,
      "step": 11800
    },
    {
      "epoch": 5.565770862800566,
      "grad_norm": 0.6309127807617188,
      "learning_rate": 4.000422418389094e-05,
      "loss": 0.2305,
      "num_input_tokens_seen": 10233136,
      "step": 11805
    },
    {
      "epoch": 5.568128241395568,
      "grad_norm": 1.6735502481460571,
      "learning_rate": 3.999637028993744e-05,
      "loss": 0.2472,
      "num_input_tokens_seen": 10237696,
      "step": 11810
    },
    {
      "epoch": 5.570485619990571,
      "grad_norm": 0.7624267935752869,
      "learning_rate": 3.99885140833552e-05,
      "loss": 0.1871,
      "num_input_tokens_seen": 10241504,
      "step": 11815
    },
    {
      "epoch": 5.572842998585573,
      "grad_norm": 0.9251819252967834,
      "learning_rate": 3.998065556535572e-05,
      "loss": 0.1502,
      "num_input_tokens_seen": 10246064,
      "step": 11820
    },
    {
      "epoch": 5.575200377180575,
      "grad_norm": 0.7428832650184631,
      "learning_rate": 3.9972794737150895e-05,
      "loss": 0.2086,
      "num_input_tokens_seen": 10250736,
      "step": 11825
    },
    {
      "epoch": 5.577557755775578,
      "grad_norm": 0.7631860971450806,
      "learning_rate": 3.996493159995297e-05,
      "loss": 0.1671,
      "num_input_tokens_seen": 10255424,
      "step": 11830
    },
    {
      "epoch": 5.57991513437058,
      "grad_norm": 1.0039814710617065,
      "learning_rate": 3.995706615497453e-05,
      "loss": 0.1778,
      "num_input_tokens_seen": 10259696,
      "step": 11835
    },
    {
      "epoch": 5.582272512965583,
      "grad_norm": 0.4432162642478943,
      "learning_rate": 3.994919840342852e-05,
      "loss": 0.1363,
      "num_input_tokens_seen": 10263712,
      "step": 11840
    },
    {
      "epoch": 5.584629891560585,
      "grad_norm": 0.7305384874343872,
      "learning_rate": 3.994132834652825e-05,
      "loss": 0.2356,
      "num_input_tokens_seen": 10269376,
      "step": 11845
    },
    {
      "epoch": 5.586987270155587,
      "grad_norm": 0.910003662109375,
      "learning_rate": 3.99334559854874e-05,
      "loss": 0.1835,
      "num_input_tokens_seen": 10273488,
      "step": 11850
    },
    {
      "epoch": 5.58934464875059,
      "grad_norm": 1.9988962411880493,
      "learning_rate": 3.9925581321519955e-05,
      "loss": 0.2682,
      "num_input_tokens_seen": 10277248,
      "step": 11855
    },
    {
      "epoch": 5.591702027345592,
      "grad_norm": 0.5470210313796997,
      "learning_rate": 3.991770435584031e-05,
      "loss": 0.1765,
      "num_input_tokens_seen": 10282784,
      "step": 11860
    },
    {
      "epoch": 5.594059405940594,
      "grad_norm": 1.139883041381836,
      "learning_rate": 3.990982508966319e-05,
      "loss": 0.2636,
      "num_input_tokens_seen": 10287856,
      "step": 11865
    },
    {
      "epoch": 5.596416784535596,
      "grad_norm": 0.9648983478546143,
      "learning_rate": 3.990194352420367e-05,
      "loss": 0.1532,
      "num_input_tokens_seen": 10292128,
      "step": 11870
    },
    {
      "epoch": 5.5987741631305985,
      "grad_norm": 0.5635062456130981,
      "learning_rate": 3.9894059660677184e-05,
      "loss": 0.2093,
      "num_input_tokens_seen": 10296912,
      "step": 11875
    },
    {
      "epoch": 5.601131541725601,
      "grad_norm": 0.45848581194877625,
      "learning_rate": 3.9886173500299526e-05,
      "loss": 0.1968,
      "num_input_tokens_seen": 10302112,
      "step": 11880
    },
    {
      "epoch": 5.603488920320603,
      "grad_norm": 1.0511775016784668,
      "learning_rate": 3.987828504428685e-05,
      "loss": 0.178,
      "num_input_tokens_seen": 10306352,
      "step": 11885
    },
    {
      "epoch": 5.605846298915606,
      "grad_norm": 1.3116681575775146,
      "learning_rate": 3.987039429385565e-05,
      "loss": 0.1862,
      "num_input_tokens_seen": 10310464,
      "step": 11890
    },
    {
      "epoch": 5.608203677510608,
      "grad_norm": 0.3981076180934906,
      "learning_rate": 3.986250125022277e-05,
      "loss": 0.1704,
      "num_input_tokens_seen": 10315120,
      "step": 11895
    },
    {
      "epoch": 5.6105610561056105,
      "grad_norm": 0.6342429518699646,
      "learning_rate": 3.985460591460544e-05,
      "loss": 0.2285,
      "num_input_tokens_seen": 10319344,
      "step": 11900
    },
    {
      "epoch": 5.612918434700613,
      "grad_norm": 0.6520922780036926,
      "learning_rate": 3.984670828822118e-05,
      "loss": 0.2024,
      "num_input_tokens_seen": 10323776,
      "step": 11905
    },
    {
      "epoch": 5.615275813295615,
      "grad_norm": 1.4416016340255737,
      "learning_rate": 3.983880837228794e-05,
      "loss": 0.1239,
      "num_input_tokens_seen": 10327552,
      "step": 11910
    },
    {
      "epoch": 5.617633191890618,
      "grad_norm": 0.5859640836715698,
      "learning_rate": 3.983090616802396e-05,
      "loss": 0.213,
      "num_input_tokens_seen": 10331648,
      "step": 11915
    },
    {
      "epoch": 5.61999057048562,
      "grad_norm": 0.6705449223518372,
      "learning_rate": 3.982300167664788e-05,
      "loss": 0.2063,
      "num_input_tokens_seen": 10336112,
      "step": 11920
    },
    {
      "epoch": 5.6223479490806225,
      "grad_norm": 0.6339258551597595,
      "learning_rate": 3.981509489937868e-05,
      "loss": 0.1494,
      "num_input_tokens_seen": 10340640,
      "step": 11925
    },
    {
      "epoch": 5.624705327675625,
      "grad_norm": 0.6421445608139038,
      "learning_rate": 3.9807185837435643e-05,
      "loss": 0.1549,
      "num_input_tokens_seen": 10344896,
      "step": 11930
    },
    {
      "epoch": 5.627062706270627,
      "grad_norm": 0.7637136578559875,
      "learning_rate": 3.9799274492038484e-05,
      "loss": 0.1446,
      "num_input_tokens_seen": 10349136,
      "step": 11935
    },
    {
      "epoch": 5.62942008486563,
      "grad_norm": 0.9574556350708008,
      "learning_rate": 3.979136086440722e-05,
      "loss": 0.2382,
      "num_input_tokens_seen": 10353360,
      "step": 11940
    },
    {
      "epoch": 5.631777463460632,
      "grad_norm": 1.2167555093765259,
      "learning_rate": 3.9783444955762226e-05,
      "loss": 0.2564,
      "num_input_tokens_seen": 10358000,
      "step": 11945
    },
    {
      "epoch": 5.634134842055634,
      "grad_norm": 0.7688062191009521,
      "learning_rate": 3.977552676732424e-05,
      "loss": 0.199,
      "num_input_tokens_seen": 10361872,
      "step": 11950
    },
    {
      "epoch": 5.636492220650636,
      "grad_norm": 0.6827632784843445,
      "learning_rate": 3.976760630031435e-05,
      "loss": 0.1704,
      "num_input_tokens_seen": 10366416,
      "step": 11955
    },
    {
      "epoch": 5.6388495992456384,
      "grad_norm": 1.5930429697036743,
      "learning_rate": 3.975968355595398e-05,
      "loss": 0.2311,
      "num_input_tokens_seen": 10370672,
      "step": 11960
    },
    {
      "epoch": 5.641206977840641,
      "grad_norm": 1.0291342735290527,
      "learning_rate": 3.9751758535464935e-05,
      "loss": 0.214,
      "num_input_tokens_seen": 10375088,
      "step": 11965
    },
    {
      "epoch": 5.643564356435643,
      "grad_norm": 1.450624942779541,
      "learning_rate": 3.9743831240069326e-05,
      "loss": 0.1558,
      "num_input_tokens_seen": 10379712,
      "step": 11970
    },
    {
      "epoch": 5.645921735030646,
      "grad_norm": 0.6603943705558777,
      "learning_rate": 3.9735901670989675e-05,
      "loss": 0.2045,
      "num_input_tokens_seen": 10383392,
      "step": 11975
    },
    {
      "epoch": 5.648279113625648,
      "grad_norm": 0.8388243317604065,
      "learning_rate": 3.97279698294488e-05,
      "loss": 0.1701,
      "num_input_tokens_seen": 10387328,
      "step": 11980
    },
    {
      "epoch": 5.6506364922206505,
      "grad_norm": 0.5037867426872253,
      "learning_rate": 3.9720035716669876e-05,
      "loss": 0.2122,
      "num_input_tokens_seen": 10391760,
      "step": 11985
    },
    {
      "epoch": 5.652993870815653,
      "grad_norm": 1.0044033527374268,
      "learning_rate": 3.9712099333876474e-05,
      "loss": 0.2347,
      "num_input_tokens_seen": 10396480,
      "step": 11990
    },
    {
      "epoch": 5.655351249410655,
      "grad_norm": 1.020058035850525,
      "learning_rate": 3.9704160682292475e-05,
      "loss": 0.138,
      "num_input_tokens_seen": 10400448,
      "step": 11995
    },
    {
      "epoch": 5.657708628005658,
      "grad_norm": 1.8662233352661133,
      "learning_rate": 3.9696219763142106e-05,
      "loss": 0.1881,
      "num_input_tokens_seen": 10404384,
      "step": 12000
    },
    {
      "epoch": 5.657708628005658,
      "eval_loss": 0.20576860010623932,
      "eval_runtime": 22.1243,
      "eval_samples_per_second": 42.623,
      "eval_steps_per_second": 21.334,
      "num_input_tokens_seen": 10404384,
      "step": 12000
    },
    {
      "epoch": 5.66006600660066,
      "grad_norm": 0.3013932406902313,
      "learning_rate": 3.968827657764997e-05,
      "loss": 0.1365,
      "num_input_tokens_seen": 10408064,
      "step": 12005
    },
    {
      "epoch": 5.6624233851956625,
      "grad_norm": 2.1379785537719727,
      "learning_rate": 3.9680331127041e-05,
      "loss": 0.1745,
      "num_input_tokens_seen": 10412800,
      "step": 12010
    },
    {
      "epoch": 5.664780763790665,
      "grad_norm": 0.7541766166687012,
      "learning_rate": 3.9672383412540495e-05,
      "loss": 0.1931,
      "num_input_tokens_seen": 10416640,
      "step": 12015
    },
    {
      "epoch": 5.667138142385667,
      "grad_norm": 1.7811561822891235,
      "learning_rate": 3.966443343537407e-05,
      "loss": 0.2349,
      "num_input_tokens_seen": 10420736,
      "step": 12020
    },
    {
      "epoch": 5.66949552098067,
      "grad_norm": 0.489525705575943,
      "learning_rate": 3.965648119676772e-05,
      "loss": 0.1738,
      "num_input_tokens_seen": 10424976,
      "step": 12025
    },
    {
      "epoch": 5.671852899575672,
      "grad_norm": 0.578550398349762,
      "learning_rate": 3.96485266979478e-05,
      "loss": 0.1809,
      "num_input_tokens_seen": 10428768,
      "step": 12030
    },
    {
      "epoch": 5.6742102781706745,
      "grad_norm": 0.808254063129425,
      "learning_rate": 3.9640569940140974e-05,
      "loss": 0.1812,
      "num_input_tokens_seen": 10432928,
      "step": 12035
    },
    {
      "epoch": 5.676567656765677,
      "grad_norm": 1.6419689655303955,
      "learning_rate": 3.963261092457428e-05,
      "loss": 0.2946,
      "num_input_tokens_seen": 10437168,
      "step": 12040
    },
    {
      "epoch": 5.678925035360679,
      "grad_norm": 1.0443404912948608,
      "learning_rate": 3.962464965247509e-05,
      "loss": 0.1868,
      "num_input_tokens_seen": 10441632,
      "step": 12045
    },
    {
      "epoch": 5.681282413955682,
      "grad_norm": 1.0185644626617432,
      "learning_rate": 3.9616686125071135e-05,
      "loss": 0.1021,
      "num_input_tokens_seen": 10445216,
      "step": 12050
    },
    {
      "epoch": 5.683639792550684,
      "grad_norm": 0.8743000626564026,
      "learning_rate": 3.9608720343590506e-05,
      "loss": 0.1759,
      "num_input_tokens_seen": 10450032,
      "step": 12055
    },
    {
      "epoch": 5.6859971711456865,
      "grad_norm": 0.5286688804626465,
      "learning_rate": 3.960075230926161e-05,
      "loss": 0.1391,
      "num_input_tokens_seen": 10453856,
      "step": 12060
    },
    {
      "epoch": 5.688354549740688,
      "grad_norm": 0.4934365451335907,
      "learning_rate": 3.959278202331322e-05,
      "loss": 0.2586,
      "num_input_tokens_seen": 10458832,
      "step": 12065
    },
    {
      "epoch": 5.69071192833569,
      "grad_norm": 0.6535439491271973,
      "learning_rate": 3.958480948697446e-05,
      "loss": 0.2149,
      "num_input_tokens_seen": 10462768,
      "step": 12070
    },
    {
      "epoch": 5.693069306930693,
      "grad_norm": 0.9538576006889343,
      "learning_rate": 3.95768347014748e-05,
      "loss": 0.2231,
      "num_input_tokens_seen": 10467360,
      "step": 12075
    },
    {
      "epoch": 5.695426685525695,
      "grad_norm": 1.7379815578460693,
      "learning_rate": 3.956885766804404e-05,
      "loss": 0.2569,
      "num_input_tokens_seen": 10471920,
      "step": 12080
    },
    {
      "epoch": 5.697784064120698,
      "grad_norm": 0.6854400038719177,
      "learning_rate": 3.956087838791235e-05,
      "loss": 0.1765,
      "num_input_tokens_seen": 10476496,
      "step": 12085
    },
    {
      "epoch": 5.7001414427157,
      "grad_norm": 1.0028282403945923,
      "learning_rate": 3.955289686231022e-05,
      "loss": 0.2106,
      "num_input_tokens_seen": 10480528,
      "step": 12090
    },
    {
      "epoch": 5.702498821310702,
      "grad_norm": 0.7064796686172485,
      "learning_rate": 3.9544913092468504e-05,
      "loss": 0.1986,
      "num_input_tokens_seen": 10485536,
      "step": 12095
    },
    {
      "epoch": 5.704856199905705,
      "grad_norm": 1.0031485557556152,
      "learning_rate": 3.9536927079618425e-05,
      "loss": 0.1927,
      "num_input_tokens_seen": 10490032,
      "step": 12100
    },
    {
      "epoch": 5.707213578500707,
      "grad_norm": 0.6191164255142212,
      "learning_rate": 3.9528938824991494e-05,
      "loss": 0.2163,
      "num_input_tokens_seen": 10493920,
      "step": 12105
    },
    {
      "epoch": 5.70957095709571,
      "grad_norm": 1.0236672163009644,
      "learning_rate": 3.952094832981962e-05,
      "loss": 0.244,
      "num_input_tokens_seen": 10497408,
      "step": 12110
    },
    {
      "epoch": 5.711928335690712,
      "grad_norm": 0.9912799596786499,
      "learning_rate": 3.951295559533503e-05,
      "loss": 0.2484,
      "num_input_tokens_seen": 10501152,
      "step": 12115
    },
    {
      "epoch": 5.714285714285714,
      "grad_norm": 0.7928313612937927,
      "learning_rate": 3.95049606227703e-05,
      "loss": 0.1371,
      "num_input_tokens_seen": 10505360,
      "step": 12120
    },
    {
      "epoch": 5.716643092880717,
      "grad_norm": 0.5163719058036804,
      "learning_rate": 3.949696341335838e-05,
      "loss": 0.1735,
      "num_input_tokens_seen": 10509600,
      "step": 12125
    },
    {
      "epoch": 5.719000471475719,
      "grad_norm": 1.504892349243164,
      "learning_rate": 3.9488963968332503e-05,
      "loss": 0.2397,
      "num_input_tokens_seen": 10513616,
      "step": 12130
    },
    {
      "epoch": 5.721357850070722,
      "grad_norm": 0.44551557302474976,
      "learning_rate": 3.948096228892631e-05,
      "loss": 0.1712,
      "num_input_tokens_seen": 10517728,
      "step": 12135
    },
    {
      "epoch": 5.723715228665724,
      "grad_norm": 2.064192056655884,
      "learning_rate": 3.947295837637375e-05,
      "loss": 0.238,
      "num_input_tokens_seen": 10522352,
      "step": 12140
    },
    {
      "epoch": 5.726072607260726,
      "grad_norm": 1.763988733291626,
      "learning_rate": 3.9464952231909135e-05,
      "loss": 0.1487,
      "num_input_tokens_seen": 10526368,
      "step": 12145
    },
    {
      "epoch": 5.728429985855728,
      "grad_norm": 0.8870442509651184,
      "learning_rate": 3.945694385676711e-05,
      "loss": 0.2029,
      "num_input_tokens_seen": 10531104,
      "step": 12150
    },
    {
      "epoch": 5.73078736445073,
      "grad_norm": 0.974120557308197,
      "learning_rate": 3.944893325218265e-05,
      "loss": 0.164,
      "num_input_tokens_seen": 10534800,
      "step": 12155
    },
    {
      "epoch": 5.733144743045733,
      "grad_norm": 1.7783399820327759,
      "learning_rate": 3.944092041939112e-05,
      "loss": 0.1931,
      "num_input_tokens_seen": 10539056,
      "step": 12160
    },
    {
      "epoch": 5.735502121640735,
      "grad_norm": 0.5263592600822449,
      "learning_rate": 3.943290535962818e-05,
      "loss": 0.1679,
      "num_input_tokens_seen": 10542704,
      "step": 12165
    },
    {
      "epoch": 5.7378595002357375,
      "grad_norm": 0.752558171749115,
      "learning_rate": 3.942488807412985e-05,
      "loss": 0.2361,
      "num_input_tokens_seen": 10546832,
      "step": 12170
    },
    {
      "epoch": 5.74021687883074,
      "grad_norm": 0.8773133754730225,
      "learning_rate": 3.941686856413251e-05,
      "loss": 0.1575,
      "num_input_tokens_seen": 10551328,
      "step": 12175
    },
    {
      "epoch": 5.742574257425742,
      "grad_norm": 0.4749230146408081,
      "learning_rate": 3.9408846830872874e-05,
      "loss": 0.1913,
      "num_input_tokens_seen": 10555056,
      "step": 12180
    },
    {
      "epoch": 5.744931636020745,
      "grad_norm": 0.5459922552108765,
      "learning_rate": 3.940082287558798e-05,
      "loss": 0.1621,
      "num_input_tokens_seen": 10559680,
      "step": 12185
    },
    {
      "epoch": 5.747289014615747,
      "grad_norm": 0.5084032416343689,
      "learning_rate": 3.939279669951522e-05,
      "loss": 0.1782,
      "num_input_tokens_seen": 10564480,
      "step": 12190
    },
    {
      "epoch": 5.7496463932107496,
      "grad_norm": 1.2383524179458618,
      "learning_rate": 3.938476830389234e-05,
      "loss": 0.2393,
      "num_input_tokens_seen": 10568672,
      "step": 12195
    },
    {
      "epoch": 5.752003771805752,
      "grad_norm": 1.3511080741882324,
      "learning_rate": 3.937673768995742e-05,
      "loss": 0.18,
      "num_input_tokens_seen": 10573872,
      "step": 12200
    },
    {
      "epoch": 5.752003771805752,
      "eval_loss": 0.20933309197425842,
      "eval_runtime": 22.1227,
      "eval_samples_per_second": 42.626,
      "eval_steps_per_second": 21.336,
      "num_input_tokens_seen": 10573872,
      "step": 12200
    },
    {
      "epoch": 5.754361150400754,
      "grad_norm": 0.5285685658454895,
      "learning_rate": 3.936870485894888e-05,
      "loss": 0.2344,
      "num_input_tokens_seen": 10578272,
      "step": 12205
    },
    {
      "epoch": 5.756718528995757,
      "grad_norm": 0.987428605556488,
      "learning_rate": 3.9360669812105475e-05,
      "loss": 0.1739,
      "num_input_tokens_seen": 10582000,
      "step": 12210
    },
    {
      "epoch": 5.759075907590759,
      "grad_norm": 0.6149322390556335,
      "learning_rate": 3.9352632550666325e-05,
      "loss": 0.2245,
      "num_input_tokens_seen": 10586720,
      "step": 12215
    },
    {
      "epoch": 5.761433286185762,
      "grad_norm": 0.6888597011566162,
      "learning_rate": 3.9344593075870866e-05,
      "loss": 0.2658,
      "num_input_tokens_seen": 10590960,
      "step": 12220
    },
    {
      "epoch": 5.763790664780764,
      "grad_norm": 1.249773621559143,
      "learning_rate": 3.933655138895889e-05,
      "loss": 0.1995,
      "num_input_tokens_seen": 10595568,
      "step": 12225
    },
    {
      "epoch": 5.766148043375766,
      "grad_norm": 0.9056380391120911,
      "learning_rate": 3.932850749117053e-05,
      "loss": 0.2563,
      "num_input_tokens_seen": 10599392,
      "step": 12230
    },
    {
      "epoch": 5.768505421970769,
      "grad_norm": 0.7715040445327759,
      "learning_rate": 3.932046138374624e-05,
      "loss": 0.2034,
      "num_input_tokens_seen": 10603824,
      "step": 12235
    },
    {
      "epoch": 5.770862800565771,
      "grad_norm": 0.8854569792747498,
      "learning_rate": 3.9312413067926854e-05,
      "loss": 0.1946,
      "num_input_tokens_seen": 10608064,
      "step": 12240
    },
    {
      "epoch": 5.773220179160774,
      "grad_norm": 2.035381317138672,
      "learning_rate": 3.9304362544953506e-05,
      "loss": 0.2161,
      "num_input_tokens_seen": 10613536,
      "step": 12245
    },
    {
      "epoch": 5.775577557755776,
      "grad_norm": 0.9930052757263184,
      "learning_rate": 3.929630981606769e-05,
      "loss": 0.2061,
      "num_input_tokens_seen": 10618176,
      "step": 12250
    },
    {
      "epoch": 5.777934936350778,
      "grad_norm": 1.9576772451400757,
      "learning_rate": 3.928825488251124e-05,
      "loss": 0.1702,
      "num_input_tokens_seen": 10623536,
      "step": 12255
    },
    {
      "epoch": 5.780292314945781,
      "grad_norm": 0.6253674626350403,
      "learning_rate": 3.9280197745526344e-05,
      "loss": 0.2385,
      "num_input_tokens_seen": 10628336,
      "step": 12260
    },
    {
      "epoch": 5.782649693540782,
      "grad_norm": 0.37158897519111633,
      "learning_rate": 3.9272138406355495e-05,
      "loss": 0.1819,
      "num_input_tokens_seen": 10632336,
      "step": 12265
    },
    {
      "epoch": 5.785007072135785,
      "grad_norm": 0.8757267594337463,
      "learning_rate": 3.926407686624154e-05,
      "loss": 0.1664,
      "num_input_tokens_seen": 10636960,
      "step": 12270
    },
    {
      "epoch": 5.787364450730787,
      "grad_norm": 1.1592321395874023,
      "learning_rate": 3.9256013126427684e-05,
      "loss": 0.1947,
      "num_input_tokens_seen": 10640960,
      "step": 12275
    },
    {
      "epoch": 5.7897218293257895,
      "grad_norm": 0.44305238127708435,
      "learning_rate": 3.9247947188157455e-05,
      "loss": 0.1429,
      "num_input_tokens_seen": 10644544,
      "step": 12280
    },
    {
      "epoch": 5.792079207920792,
      "grad_norm": 2.0310134887695312,
      "learning_rate": 3.9239879052674715e-05,
      "loss": 0.1966,
      "num_input_tokens_seen": 10648272,
      "step": 12285
    },
    {
      "epoch": 5.794436586515794,
      "grad_norm": 1.2150582075119019,
      "learning_rate": 3.9231808721223673e-05,
      "loss": 0.1375,
      "num_input_tokens_seen": 10651920,
      "step": 12290
    },
    {
      "epoch": 5.796793965110797,
      "grad_norm": 0.7112054228782654,
      "learning_rate": 3.9223736195048886e-05,
      "loss": 0.1437,
      "num_input_tokens_seen": 10656288,
      "step": 12295
    },
    {
      "epoch": 5.799151343705799,
      "grad_norm": 0.5777669548988342,
      "learning_rate": 3.921566147539523e-05,
      "loss": 0.1936,
      "num_input_tokens_seen": 10660304,
      "step": 12300
    },
    {
      "epoch": 5.8015087223008015,
      "grad_norm": 1.1953169107437134,
      "learning_rate": 3.920758456350792e-05,
      "loss": 0.2382,
      "num_input_tokens_seen": 10664608,
      "step": 12305
    },
    {
      "epoch": 5.803866100895804,
      "grad_norm": 0.8217502236366272,
      "learning_rate": 3.919950546063253e-05,
      "loss": 0.1708,
      "num_input_tokens_seen": 10670064,
      "step": 12310
    },
    {
      "epoch": 5.806223479490806,
      "grad_norm": 0.6086592078208923,
      "learning_rate": 3.919142416801496e-05,
      "loss": 0.2177,
      "num_input_tokens_seen": 10674928,
      "step": 12315
    },
    {
      "epoch": 5.808580858085809,
      "grad_norm": 0.49367856979370117,
      "learning_rate": 3.918334068690144e-05,
      "loss": 0.1885,
      "num_input_tokens_seen": 10679264,
      "step": 12320
    },
    {
      "epoch": 5.810938236680811,
      "grad_norm": 1.8926095962524414,
      "learning_rate": 3.917525501853855e-05,
      "loss": 0.193,
      "num_input_tokens_seen": 10683376,
      "step": 12325
    },
    {
      "epoch": 5.8132956152758135,
      "grad_norm": 2.22643780708313,
      "learning_rate": 3.916716716417319e-05,
      "loss": 0.2044,
      "num_input_tokens_seen": 10686928,
      "step": 12330
    },
    {
      "epoch": 5.815652993870816,
      "grad_norm": 0.7202649712562561,
      "learning_rate": 3.915907712505263e-05,
      "loss": 0.1524,
      "num_input_tokens_seen": 10691632,
      "step": 12335
    },
    {
      "epoch": 5.818010372465818,
      "grad_norm": 0.5169095396995544,
      "learning_rate": 3.915098490242444e-05,
      "loss": 0.2075,
      "num_input_tokens_seen": 10696096,
      "step": 12340
    },
    {
      "epoch": 5.820367751060821,
      "grad_norm": 2.4034838676452637,
      "learning_rate": 3.914289049753654e-05,
      "loss": 0.2612,
      "num_input_tokens_seen": 10700544,
      "step": 12345
    },
    {
      "epoch": 5.822725129655822,
      "grad_norm": 0.7671160101890564,
      "learning_rate": 3.913479391163719e-05,
      "loss": 0.1384,
      "num_input_tokens_seen": 10704944,
      "step": 12350
    },
    {
      "epoch": 5.825082508250825,
      "grad_norm": 0.734845757484436,
      "learning_rate": 3.9126695145975e-05,
      "loss": 0.2138,
      "num_input_tokens_seen": 10709632,
      "step": 12355
    },
    {
      "epoch": 5.827439886845827,
      "grad_norm": 1.5135706663131714,
      "learning_rate": 3.911859420179889e-05,
      "loss": 0.2544,
      "num_input_tokens_seen": 10714336,
      "step": 12360
    },
    {
      "epoch": 5.829797265440829,
      "grad_norm": 1.1408478021621704,
      "learning_rate": 3.911049108035813e-05,
      "loss": 0.1318,
      "num_input_tokens_seen": 10719056,
      "step": 12365
    },
    {
      "epoch": 5.832154644035832,
      "grad_norm": 0.5784863829612732,
      "learning_rate": 3.910238578290232e-05,
      "loss": 0.1517,
      "num_input_tokens_seen": 10722736,
      "step": 12370
    },
    {
      "epoch": 5.834512022630834,
      "grad_norm": 1.492133378982544,
      "learning_rate": 3.90942783106814e-05,
      "loss": 0.2088,
      "num_input_tokens_seen": 10727104,
      "step": 12375
    },
    {
      "epoch": 5.836869401225837,
      "grad_norm": 0.778036892414093,
      "learning_rate": 3.908616866494564e-05,
      "loss": 0.1758,
      "num_input_tokens_seen": 10731504,
      "step": 12380
    },
    {
      "epoch": 5.839226779820839,
      "grad_norm": 0.7190864682197571,
      "learning_rate": 3.907805684694566e-05,
      "loss": 0.1158,
      "num_input_tokens_seen": 10735312,
      "step": 12385
    },
    {
      "epoch": 5.841584158415841,
      "grad_norm": 1.5132633447647095,
      "learning_rate": 3.90699428579324e-05,
      "loss": 0.2096,
      "num_input_tokens_seen": 10738928,
      "step": 12390
    },
    {
      "epoch": 5.843941537010844,
      "grad_norm": 0.5430698990821838,
      "learning_rate": 3.906182669915713e-05,
      "loss": 0.1867,
      "num_input_tokens_seen": 10743568,
      "step": 12395
    },
    {
      "epoch": 5.846298915605846,
      "grad_norm": 1.569865345954895,
      "learning_rate": 3.9053708371871476e-05,
      "loss": 0.2274,
      "num_input_tokens_seen": 10748304,
      "step": 12400
    },
    {
      "epoch": 5.846298915605846,
      "eval_loss": 0.2079291194677353,
      "eval_runtime": 22.1958,
      "eval_samples_per_second": 42.485,
      "eval_steps_per_second": 21.265,
      "num_input_tokens_seen": 10748304,
      "step": 12400
    },
    {
      "epoch": 5.848656294200849,
      "grad_norm": 0.5214832425117493,
      "learning_rate": 3.904558787732738e-05,
      "loss": 0.1538,
      "num_input_tokens_seen": 10752736,
      "step": 12405
    },
    {
      "epoch": 5.851013672795851,
      "grad_norm": 0.5561400651931763,
      "learning_rate": 3.9037465216777135e-05,
      "loss": 0.1586,
      "num_input_tokens_seen": 10757216,
      "step": 12410
    },
    {
      "epoch": 5.8533710513908535,
      "grad_norm": 1.0277453660964966,
      "learning_rate": 3.902934039147334e-05,
      "loss": 0.1835,
      "num_input_tokens_seen": 10761056,
      "step": 12415
    },
    {
      "epoch": 5.855728429985856,
      "grad_norm": 0.793773889541626,
      "learning_rate": 3.902121340266894e-05,
      "loss": 0.1945,
      "num_input_tokens_seen": 10764784,
      "step": 12420
    },
    {
      "epoch": 5.858085808580858,
      "grad_norm": 0.5918493866920471,
      "learning_rate": 3.9013084251617246e-05,
      "loss": 0.2683,
      "num_input_tokens_seen": 10769296,
      "step": 12425
    },
    {
      "epoch": 5.860443187175861,
      "grad_norm": 0.9823734164237976,
      "learning_rate": 3.9004952939571865e-05,
      "loss": 0.1641,
      "num_input_tokens_seen": 10773200,
      "step": 12430
    },
    {
      "epoch": 5.862800565770863,
      "grad_norm": 0.8739505410194397,
      "learning_rate": 3.899681946778673e-05,
      "loss": 0.1854,
      "num_input_tokens_seen": 10777568,
      "step": 12435
    },
    {
      "epoch": 5.8651579443658655,
      "grad_norm": 1.1334154605865479,
      "learning_rate": 3.898868383751615e-05,
      "loss": 0.1619,
      "num_input_tokens_seen": 10782112,
      "step": 12440
    },
    {
      "epoch": 5.867515322960868,
      "grad_norm": 0.9769644141197205,
      "learning_rate": 3.8980546050014724e-05,
      "loss": 0.2153,
      "num_input_tokens_seen": 10786176,
      "step": 12445
    },
    {
      "epoch": 5.86987270155587,
      "grad_norm": 0.49138087034225464,
      "learning_rate": 3.897240610653741e-05,
      "loss": 0.1689,
      "num_input_tokens_seen": 10790592,
      "step": 12450
    },
    {
      "epoch": 5.872230080150873,
      "grad_norm": 0.8302065134048462,
      "learning_rate": 3.896426400833948e-05,
      "loss": 0.1788,
      "num_input_tokens_seen": 10794480,
      "step": 12455
    },
    {
      "epoch": 5.874587458745875,
      "grad_norm": 1.436772346496582,
      "learning_rate": 3.895611975667656e-05,
      "loss": 0.2228,
      "num_input_tokens_seen": 10799120,
      "step": 12460
    },
    {
      "epoch": 5.876944837340877,
      "grad_norm": 1.0956778526306152,
      "learning_rate": 3.8947973352804584e-05,
      "loss": 0.2078,
      "num_input_tokens_seen": 10803936,
      "step": 12465
    },
    {
      "epoch": 5.879302215935879,
      "grad_norm": 1.736181378364563,
      "learning_rate": 3.893982479797984e-05,
      "loss": 0.2428,
      "num_input_tokens_seen": 10807936,
      "step": 12470
    },
    {
      "epoch": 5.881659594530881,
      "grad_norm": 0.6803781986236572,
      "learning_rate": 3.8931674093458926e-05,
      "loss": 0.1729,
      "num_input_tokens_seen": 10812032,
      "step": 12475
    },
    {
      "epoch": 5.884016973125884,
      "grad_norm": 0.7206957936286926,
      "learning_rate": 3.89235212404988e-05,
      "loss": 0.245,
      "num_input_tokens_seen": 10816032,
      "step": 12480
    },
    {
      "epoch": 5.886374351720886,
      "grad_norm": 0.4644705355167389,
      "learning_rate": 3.891536624035672e-05,
      "loss": 0.198,
      "num_input_tokens_seen": 10819840,
      "step": 12485
    },
    {
      "epoch": 5.888731730315889,
      "grad_norm": 0.6384208798408508,
      "learning_rate": 3.8907209094290295e-05,
      "loss": 0.1609,
      "num_input_tokens_seen": 10823920,
      "step": 12490
    },
    {
      "epoch": 5.891089108910891,
      "grad_norm": 1.265516757965088,
      "learning_rate": 3.8899049803557466e-05,
      "loss": 0.1504,
      "num_input_tokens_seen": 10827776,
      "step": 12495
    },
    {
      "epoch": 5.893446487505893,
      "grad_norm": 2.7566027641296387,
      "learning_rate": 3.889088836941648e-05,
      "loss": 0.2409,
      "num_input_tokens_seen": 10833952,
      "step": 12500
    },
    {
      "epoch": 5.895803866100896,
      "grad_norm": 0.5144676566123962,
      "learning_rate": 3.8882724793125946e-05,
      "loss": 0.1893,
      "num_input_tokens_seen": 10837664,
      "step": 12505
    },
    {
      "epoch": 5.898161244695898,
      "grad_norm": 0.9221445918083191,
      "learning_rate": 3.8874559075944794e-05,
      "loss": 0.1923,
      "num_input_tokens_seen": 10841616,
      "step": 12510
    },
    {
      "epoch": 5.900518623290901,
      "grad_norm": 0.9443637132644653,
      "learning_rate": 3.886639121913227e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 10845760,
      "step": 12515
    },
    {
      "epoch": 5.902876001885903,
      "grad_norm": 1.0716705322265625,
      "learning_rate": 3.885822122394797e-05,
      "loss": 0.2229,
      "num_input_tokens_seen": 10850128,
      "step": 12520
    },
    {
      "epoch": 5.905233380480905,
      "grad_norm": 0.8260701894760132,
      "learning_rate": 3.8850049091651794e-05,
      "loss": 0.1914,
      "num_input_tokens_seen": 10854448,
      "step": 12525
    },
    {
      "epoch": 5.907590759075908,
      "grad_norm": 0.9720013737678528,
      "learning_rate": 3.8841874823504e-05,
      "loss": 0.124,
      "num_input_tokens_seen": 10858656,
      "step": 12530
    },
    {
      "epoch": 5.90994813767091,
      "grad_norm": 2.4563536643981934,
      "learning_rate": 3.8833698420765157e-05,
      "loss": 0.211,
      "num_input_tokens_seen": 10863216,
      "step": 12535
    },
    {
      "epoch": 5.912305516265913,
      "grad_norm": 0.8730310797691345,
      "learning_rate": 3.882551988469618e-05,
      "loss": 0.1781,
      "num_input_tokens_seen": 10867168,
      "step": 12540
    },
    {
      "epoch": 5.914662894860915,
      "grad_norm": 1.195899248123169,
      "learning_rate": 3.881733921655829e-05,
      "loss": 0.2386,
      "num_input_tokens_seen": 10871312,
      "step": 12545
    },
    {
      "epoch": 5.9170202734559165,
      "grad_norm": 2.199793577194214,
      "learning_rate": 3.8809156417613054e-05,
      "loss": 0.2354,
      "num_input_tokens_seen": 10876528,
      "step": 12550
    },
    {
      "epoch": 5.919377652050919,
      "grad_norm": 2.455470323562622,
      "learning_rate": 3.8800971489122364e-05,
      "loss": 0.1604,
      "num_input_tokens_seen": 10881120,
      "step": 12555
    },
    {
      "epoch": 5.921735030645921,
      "grad_norm": 0.7265248894691467,
      "learning_rate": 3.8792784432348434e-05,
      "loss": 0.1814,
      "num_input_tokens_seen": 10884832,
      "step": 12560
    },
    {
      "epoch": 5.924092409240924,
      "grad_norm": 0.41171568632125854,
      "learning_rate": 3.878459524855381e-05,
      "loss": 0.2045,
      "num_input_tokens_seen": 10888816,
      "step": 12565
    },
    {
      "epoch": 5.926449787835926,
      "grad_norm": 0.6503592729568481,
      "learning_rate": 3.8776403939001384e-05,
      "loss": 0.1857,
      "num_input_tokens_seen": 10892448,
      "step": 12570
    },
    {
      "epoch": 5.9288071664309285,
      "grad_norm": 0.5585503578186035,
      "learning_rate": 3.876821050495433e-05,
      "loss": 0.1602,
      "num_input_tokens_seen": 10896688,
      "step": 12575
    },
    {
      "epoch": 5.931164545025931,
      "grad_norm": 0.8205670714378357,
      "learning_rate": 3.87600149476762e-05,
      "loss": 0.2255,
      "num_input_tokens_seen": 10900736,
      "step": 12580
    },
    {
      "epoch": 5.933521923620933,
      "grad_norm": 0.46430712938308716,
      "learning_rate": 3.8751817268430843e-05,
      "loss": 0.1987,
      "num_input_tokens_seen": 10905344,
      "step": 12585
    },
    {
      "epoch": 5.935879302215936,
      "grad_norm": 1.0180494785308838,
      "learning_rate": 3.8743617468482464e-05,
      "loss": 0.1632,
      "num_input_tokens_seen": 10909616,
      "step": 12590
    },
    {
      "epoch": 5.938236680810938,
      "grad_norm": 0.9714590311050415,
      "learning_rate": 3.8735415549095535e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 10913840,
      "step": 12595
    },
    {
      "epoch": 5.9405940594059405,
      "grad_norm": 0.6984928250312805,
      "learning_rate": 3.8727211511534934e-05,
      "loss": 0.177,
      "num_input_tokens_seen": 10917920,
      "step": 12600
    },
    {
      "epoch": 5.9405940594059405,
      "eval_loss": 0.20501913130283356,
      "eval_runtime": 22.1511,
      "eval_samples_per_second": 42.571,
      "eval_steps_per_second": 21.308,
      "num_input_tokens_seen": 10917920,
      "step": 12600
    },
    {
      "epoch": 5.942951438000943,
      "grad_norm": 0.8280900120735168,
      "learning_rate": 3.8719005357065804e-05,
      "loss": 0.1788,
      "num_input_tokens_seen": 10922128,
      "step": 12605
    },
    {
      "epoch": 5.945308816595945,
      "grad_norm": 0.38180670142173767,
      "learning_rate": 3.8710797086953645e-05,
      "loss": 0.1534,
      "num_input_tokens_seen": 10926528,
      "step": 12610
    },
    {
      "epoch": 5.947666195190948,
      "grad_norm": 0.7773281335830688,
      "learning_rate": 3.870258670246427e-05,
      "loss": 0.1583,
      "num_input_tokens_seen": 10930816,
      "step": 12615
    },
    {
      "epoch": 5.95002357378595,
      "grad_norm": 1.0304478406906128,
      "learning_rate": 3.869437420486384e-05,
      "loss": 0.2293,
      "num_input_tokens_seen": 10935488,
      "step": 12620
    },
    {
      "epoch": 5.9523809523809526,
      "grad_norm": 0.8969800472259521,
      "learning_rate": 3.8686159595418805e-05,
      "loss": 0.2614,
      "num_input_tokens_seen": 10939616,
      "step": 12625
    },
    {
      "epoch": 5.954738330975955,
      "grad_norm": 0.9453440308570862,
      "learning_rate": 3.867794287539597e-05,
      "loss": 0.182,
      "num_input_tokens_seen": 10943616,
      "step": 12630
    },
    {
      "epoch": 5.957095709570957,
      "grad_norm": 0.6544432640075684,
      "learning_rate": 3.866972404606245e-05,
      "loss": 0.1949,
      "num_input_tokens_seen": 10947456,
      "step": 12635
    },
    {
      "epoch": 5.95945308816596,
      "grad_norm": 0.6957946419715881,
      "learning_rate": 3.866150310868571e-05,
      "loss": 0.2691,
      "num_input_tokens_seen": 10951744,
      "step": 12640
    },
    {
      "epoch": 5.961810466760962,
      "grad_norm": 1.336802363395691,
      "learning_rate": 3.8653280064533506e-05,
      "loss": 0.1719,
      "num_input_tokens_seen": 10956416,
      "step": 12645
    },
    {
      "epoch": 5.964167845355965,
      "grad_norm": 1.5791946649551392,
      "learning_rate": 3.864505491487394e-05,
      "loss": 0.1725,
      "num_input_tokens_seen": 10960672,
      "step": 12650
    },
    {
      "epoch": 5.966525223950967,
      "grad_norm": 0.6411140561103821,
      "learning_rate": 3.8636827660975414e-05,
      "loss": 0.1887,
      "num_input_tokens_seen": 10964896,
      "step": 12655
    },
    {
      "epoch": 5.968882602545969,
      "grad_norm": 0.6747446060180664,
      "learning_rate": 3.862859830410671e-05,
      "loss": 0.1539,
      "num_input_tokens_seen": 10969568,
      "step": 12660
    },
    {
      "epoch": 5.971239981140971,
      "grad_norm": 0.8146329522132874,
      "learning_rate": 3.862036684553688e-05,
      "loss": 0.171,
      "num_input_tokens_seen": 10973712,
      "step": 12665
    },
    {
      "epoch": 5.973597359735973,
      "grad_norm": 1.1278852224349976,
      "learning_rate": 3.8612133286535314e-05,
      "loss": 0.1943,
      "num_input_tokens_seen": 10978464,
      "step": 12670
    },
    {
      "epoch": 5.975954738330976,
      "grad_norm": 2.4154865741729736,
      "learning_rate": 3.860389762837173e-05,
      "loss": 0.2831,
      "num_input_tokens_seen": 10983392,
      "step": 12675
    },
    {
      "epoch": 5.978312116925978,
      "grad_norm": 0.7996242642402649,
      "learning_rate": 3.859565987231618e-05,
      "loss": 0.1491,
      "num_input_tokens_seen": 10986928,
      "step": 12680
    },
    {
      "epoch": 5.9806694955209805,
      "grad_norm": 1.0629855394363403,
      "learning_rate": 3.858742001963902e-05,
      "loss": 0.2549,
      "num_input_tokens_seen": 10990448,
      "step": 12685
    },
    {
      "epoch": 5.983026874115983,
      "grad_norm": 2.1321425437927246,
      "learning_rate": 3.857917807161094e-05,
      "loss": 0.2286,
      "num_input_tokens_seen": 10995056,
      "step": 12690
    },
    {
      "epoch": 5.985384252710985,
      "grad_norm": 0.7703915238380432,
      "learning_rate": 3.857093402950296e-05,
      "loss": 0.1275,
      "num_input_tokens_seen": 10999568,
      "step": 12695
    },
    {
      "epoch": 5.987741631305988,
      "grad_norm": 0.8696743249893188,
      "learning_rate": 3.8562687894586414e-05,
      "loss": 0.2133,
      "num_input_tokens_seen": 11003648,
      "step": 12700
    },
    {
      "epoch": 5.99009900990099,
      "grad_norm": 0.7950209379196167,
      "learning_rate": 3.8554439668132946e-05,
      "loss": 0.2394,
      "num_input_tokens_seen": 11009488,
      "step": 12705
    },
    {
      "epoch": 5.9924563884959925,
      "grad_norm": 0.4374142587184906,
      "learning_rate": 3.854618935141455e-05,
      "loss": 0.1825,
      "num_input_tokens_seen": 11015024,
      "step": 12710
    },
    {
      "epoch": 5.994813767090995,
      "grad_norm": 0.8296253681182861,
      "learning_rate": 3.8537936945703525e-05,
      "loss": 0.1905,
      "num_input_tokens_seen": 11018768,
      "step": 12715
    },
    {
      "epoch": 5.997171145685997,
      "grad_norm": 1.1714545488357544,
      "learning_rate": 3.852968245227249e-05,
      "loss": 0.224,
      "num_input_tokens_seen": 11023872,
      "step": 12720
    },
    {
      "epoch": 5.999528524281,
      "grad_norm": 1.6681524515151978,
      "learning_rate": 3.85214258723944e-05,
      "loss": 0.1521,
      "num_input_tokens_seen": 11027488,
      "step": 12725
    },
    {
      "epoch": 6.001885902876002,
      "grad_norm": 1.6339412927627563,
      "learning_rate": 3.8513167207342524e-05,
      "loss": 0.1924,
      "num_input_tokens_seen": 11031792,
      "step": 12730
    },
    {
      "epoch": 6.0042432814710045,
      "grad_norm": 0.6744508147239685,
      "learning_rate": 3.850490645839044e-05,
      "loss": 0.2051,
      "num_input_tokens_seen": 11036432,
      "step": 12735
    },
    {
      "epoch": 6.006600660066007,
      "grad_norm": 0.6396738290786743,
      "learning_rate": 3.849664362681207e-05,
      "loss": 0.2166,
      "num_input_tokens_seen": 11040832,
      "step": 12740
    },
    {
      "epoch": 6.008958038661009,
      "grad_norm": 0.7886515855789185,
      "learning_rate": 3.848837871388165e-05,
      "loss": 0.1656,
      "num_input_tokens_seen": 11045216,
      "step": 12745
    },
    {
      "epoch": 6.011315417256012,
      "grad_norm": 1.5691858530044556,
      "learning_rate": 3.848011172087371e-05,
      "loss": 0.2252,
      "num_input_tokens_seen": 11048560,
      "step": 12750
    },
    {
      "epoch": 6.013672795851014,
      "grad_norm": 1.1593705415725708,
      "learning_rate": 3.847184264906315e-05,
      "loss": 0.1858,
      "num_input_tokens_seen": 11053232,
      "step": 12755
    },
    {
      "epoch": 6.016030174446016,
      "grad_norm": 1.9410057067871094,
      "learning_rate": 3.846357149972516e-05,
      "loss": 0.2903,
      "num_input_tokens_seen": 11057840,
      "step": 12760
    },
    {
      "epoch": 6.018387553041018,
      "grad_norm": 0.47846975922584534,
      "learning_rate": 3.8455298274135246e-05,
      "loss": 0.1651,
      "num_input_tokens_seen": 11061616,
      "step": 12765
    },
    {
      "epoch": 6.02074493163602,
      "grad_norm": 0.5376976132392883,
      "learning_rate": 3.8447022973569254e-05,
      "loss": 0.1688,
      "num_input_tokens_seen": 11065584,
      "step": 12770
    },
    {
      "epoch": 6.023102310231023,
      "grad_norm": 1.9293546676635742,
      "learning_rate": 3.843874559930332e-05,
      "loss": 0.2453,
      "num_input_tokens_seen": 11070272,
      "step": 12775
    },
    {
      "epoch": 6.025459688826025,
      "grad_norm": 1.2231066226959229,
      "learning_rate": 3.843046615261394e-05,
      "loss": 0.2212,
      "num_input_tokens_seen": 11073904,
      "step": 12780
    },
    {
      "epoch": 6.027817067421028,
      "grad_norm": 0.5702390074729919,
      "learning_rate": 3.842218463477791e-05,
      "loss": 0.178,
      "num_input_tokens_seen": 11078400,
      "step": 12785
    },
    {
      "epoch": 6.03017444601603,
      "grad_norm": 2.071408748626709,
      "learning_rate": 3.841390104707233e-05,
      "loss": 0.2145,
      "num_input_tokens_seen": 11083488,
      "step": 12790
    },
    {
      "epoch": 6.032531824611032,
      "grad_norm": 1.107897400856018,
      "learning_rate": 3.8405615390774643e-05,
      "loss": 0.2198,
      "num_input_tokens_seen": 11088656,
      "step": 12795
    },
    {
      "epoch": 6.034889203206035,
      "grad_norm": 0.7843061089515686,
      "learning_rate": 3.839732766716259e-05,
      "loss": 0.2307,
      "num_input_tokens_seen": 11092736,
      "step": 12800
    },
    {
      "epoch": 6.034889203206035,
      "eval_loss": 0.20661431550979614,
      "eval_runtime": 22.1565,
      "eval_samples_per_second": 42.561,
      "eval_steps_per_second": 21.303,
      "num_input_tokens_seen": 11092736,
      "step": 12800
    },
    {
      "epoch": 6.037246581801037,
      "grad_norm": 0.8765320777893066,
      "learning_rate": 3.838903787751425e-05,
      "loss": 0.1682,
      "num_input_tokens_seen": 11097248,
      "step": 12805
    },
    {
      "epoch": 6.03960396039604,
      "grad_norm": 1.3774080276489258,
      "learning_rate": 3.838074602310802e-05,
      "loss": 0.238,
      "num_input_tokens_seen": 11102208,
      "step": 12810
    },
    {
      "epoch": 6.041961338991042,
      "grad_norm": 0.5840504765510559,
      "learning_rate": 3.837245210522258e-05,
      "loss": 0.1432,
      "num_input_tokens_seen": 11106880,
      "step": 12815
    },
    {
      "epoch": 6.044318717586044,
      "grad_norm": 1.7755153179168701,
      "learning_rate": 3.8364156125136996e-05,
      "loss": 0.1988,
      "num_input_tokens_seen": 11110560,
      "step": 12820
    },
    {
      "epoch": 6.046676096181047,
      "grad_norm": 1.1326093673706055,
      "learning_rate": 3.835585808413059e-05,
      "loss": 0.1483,
      "num_input_tokens_seen": 11114496,
      "step": 12825
    },
    {
      "epoch": 6.049033474776049,
      "grad_norm": 0.26245638728141785,
      "learning_rate": 3.8347557983483024e-05,
      "loss": 0.1611,
      "num_input_tokens_seen": 11118640,
      "step": 12830
    },
    {
      "epoch": 6.051390853371052,
      "grad_norm": 1.4149391651153564,
      "learning_rate": 3.833925582447428e-05,
      "loss": 0.2041,
      "num_input_tokens_seen": 11122720,
      "step": 12835
    },
    {
      "epoch": 6.053748231966054,
      "grad_norm": 1.008602499961853,
      "learning_rate": 3.8330951608384656e-05,
      "loss": 0.1716,
      "num_input_tokens_seen": 11127840,
      "step": 12840
    },
    {
      "epoch": 6.0561056105610565,
      "grad_norm": 0.9435743689537048,
      "learning_rate": 3.832264533649477e-05,
      "loss": 0.129,
      "num_input_tokens_seen": 11131712,
      "step": 12845
    },
    {
      "epoch": 6.058462989156059,
      "grad_norm": 0.4566563665866852,
      "learning_rate": 3.8314337010085555e-05,
      "loss": 0.1536,
      "num_input_tokens_seen": 11135808,
      "step": 12850
    },
    {
      "epoch": 6.060820367751061,
      "grad_norm": 1.21574068069458,
      "learning_rate": 3.830602663043824e-05,
      "loss": 0.1633,
      "num_input_tokens_seen": 11140400,
      "step": 12855
    },
    {
      "epoch": 6.063177746346063,
      "grad_norm": 1.104163646697998,
      "learning_rate": 3.8297714198834414e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 11144880,
      "step": 12860
    },
    {
      "epoch": 6.065535124941065,
      "grad_norm": 0.7742725610733032,
      "learning_rate": 3.828939971655595e-05,
      "loss": 0.1602,
      "num_input_tokens_seen": 11149232,
      "step": 12865
    },
    {
      "epoch": 6.067892503536068,
      "grad_norm": 1.2221332788467407,
      "learning_rate": 3.828108318488505e-05,
      "loss": 0.1764,
      "num_input_tokens_seen": 11153792,
      "step": 12870
    },
    {
      "epoch": 6.07024988213107,
      "grad_norm": 0.3540181517601013,
      "learning_rate": 3.8272764605104216e-05,
      "loss": 0.1445,
      "num_input_tokens_seen": 11158320,
      "step": 12875
    },
    {
      "epoch": 6.072607260726072,
      "grad_norm": 1.6265733242034912,
      "learning_rate": 3.826444397849628e-05,
      "loss": 0.1965,
      "num_input_tokens_seen": 11163424,
      "step": 12880
    },
    {
      "epoch": 6.074964639321075,
      "grad_norm": 1.4488134384155273,
      "learning_rate": 3.825612130634439e-05,
      "loss": 0.2582,
      "num_input_tokens_seen": 11168016,
      "step": 12885
    },
    {
      "epoch": 6.077322017916077,
      "grad_norm": 1.2434360980987549,
      "learning_rate": 3.824779658993202e-05,
      "loss": 0.1841,
      "num_input_tokens_seen": 11172288,
      "step": 12890
    },
    {
      "epoch": 6.07967939651108,
      "grad_norm": 0.48295873403549194,
      "learning_rate": 3.823946983054292e-05,
      "loss": 0.1331,
      "num_input_tokens_seen": 11176112,
      "step": 12895
    },
    {
      "epoch": 6.082036775106082,
      "grad_norm": 1.176013708114624,
      "learning_rate": 3.82311410294612e-05,
      "loss": 0.1608,
      "num_input_tokens_seen": 11180656,
      "step": 12900
    },
    {
      "epoch": 6.084394153701084,
      "grad_norm": 1.2138067483901978,
      "learning_rate": 3.822281018797127e-05,
      "loss": 0.1779,
      "num_input_tokens_seen": 11184400,
      "step": 12905
    },
    {
      "epoch": 6.086751532296087,
      "grad_norm": 1.056015133857727,
      "learning_rate": 3.821447730735783e-05,
      "loss": 0.223,
      "num_input_tokens_seen": 11188704,
      "step": 12910
    },
    {
      "epoch": 6.089108910891089,
      "grad_norm": 0.9798020124435425,
      "learning_rate": 3.820614238890592e-05,
      "loss": 0.208,
      "num_input_tokens_seen": 11192848,
      "step": 12915
    },
    {
      "epoch": 6.091466289486092,
      "grad_norm": 0.6697331070899963,
      "learning_rate": 3.819780543390091e-05,
      "loss": 0.2301,
      "num_input_tokens_seen": 11197536,
      "step": 12920
    },
    {
      "epoch": 6.093823668081094,
      "grad_norm": 0.5134458541870117,
      "learning_rate": 3.818946644362844e-05,
      "loss": 0.206,
      "num_input_tokens_seen": 11201984,
      "step": 12925
    },
    {
      "epoch": 6.096181046676096,
      "grad_norm": 0.5902479290962219,
      "learning_rate": 3.81811254193745e-05,
      "loss": 0.1493,
      "num_input_tokens_seen": 11206736,
      "step": 12930
    },
    {
      "epoch": 6.098538425271099,
      "grad_norm": 1.1646214723587036,
      "learning_rate": 3.8172782362425366e-05,
      "loss": 0.2033,
      "num_input_tokens_seen": 11211072,
      "step": 12935
    },
    {
      "epoch": 6.100895803866101,
      "grad_norm": 1.3462673425674438,
      "learning_rate": 3.816443727406765e-05,
      "loss": 0.2002,
      "num_input_tokens_seen": 11215392,
      "step": 12940
    },
    {
      "epoch": 6.103253182461104,
      "grad_norm": 1.5173336267471313,
      "learning_rate": 3.815609015558829e-05,
      "loss": 0.1518,
      "num_input_tokens_seen": 11220000,
      "step": 12945
    },
    {
      "epoch": 6.105610561056106,
      "grad_norm": 1.6221508979797363,
      "learning_rate": 3.814774100827448e-05,
      "loss": 0.1723,
      "num_input_tokens_seen": 11224128,
      "step": 12950
    },
    {
      "epoch": 6.107967939651108,
      "grad_norm": 0.9224163293838501,
      "learning_rate": 3.813938983341379e-05,
      "loss": 0.2156,
      "num_input_tokens_seen": 11228000,
      "step": 12955
    },
    {
      "epoch": 6.11032531824611,
      "grad_norm": 0.936775267124176,
      "learning_rate": 3.813103663229407e-05,
      "loss": 0.233,
      "num_input_tokens_seen": 11232304,
      "step": 12960
    },
    {
      "epoch": 6.112682696841112,
      "grad_norm": 0.7444096803665161,
      "learning_rate": 3.812268140620349e-05,
      "loss": 0.1732,
      "num_input_tokens_seen": 11237584,
      "step": 12965
    },
    {
      "epoch": 6.115040075436115,
      "grad_norm": 1.5975456237792969,
      "learning_rate": 3.811432415643051e-05,
      "loss": 0.175,
      "num_input_tokens_seen": 11240992,
      "step": 12970
    },
    {
      "epoch": 6.117397454031117,
      "grad_norm": 0.5893133878707886,
      "learning_rate": 3.8105964884263954e-05,
      "loss": 0.1919,
      "num_input_tokens_seen": 11245568,
      "step": 12975
    },
    {
      "epoch": 6.1197548326261195,
      "grad_norm": 1.6421923637390137,
      "learning_rate": 3.809760359099291e-05,
      "loss": 0.1652,
      "num_input_tokens_seen": 11249840,
      "step": 12980
    },
    {
      "epoch": 6.122112211221122,
      "grad_norm": 0.37614279985427856,
      "learning_rate": 3.8089240277906804e-05,
      "loss": 0.1745,
      "num_input_tokens_seen": 11255968,
      "step": 12985
    },
    {
      "epoch": 6.124469589816124,
      "grad_norm": 0.4919806718826294,
      "learning_rate": 3.808087494629535e-05,
      "loss": 0.1456,
      "num_input_tokens_seen": 11260752,
      "step": 12990
    },
    {
      "epoch": 6.126826968411127,
      "grad_norm": 1.1796568632125854,
      "learning_rate": 3.8072507597448595e-05,
      "loss": 0.1992,
      "num_input_tokens_seen": 11264720,
      "step": 12995
    },
    {
      "epoch": 6.129184347006129,
      "grad_norm": 0.7602362632751465,
      "learning_rate": 3.806413823265689e-05,
      "loss": 0.1909,
      "num_input_tokens_seen": 11269264,
      "step": 13000
    },
    {
      "epoch": 6.129184347006129,
      "eval_loss": 0.20676816999912262,
      "eval_runtime": 22.1739,
      "eval_samples_per_second": 42.527,
      "eval_steps_per_second": 21.286,
      "num_input_tokens_seen": 11269264,
      "step": 13000
    },
    {
      "epoch": 6.1315417256011315,
      "grad_norm": 0.6655197739601135,
      "learning_rate": 3.805576685321089e-05,
      "loss": 0.1928,
      "num_input_tokens_seen": 11273904,
      "step": 13005
    },
    {
      "epoch": 6.133899104196134,
      "grad_norm": 1.436972737312317,
      "learning_rate": 3.804739346040158e-05,
      "loss": 0.3449,
      "num_input_tokens_seen": 11279424,
      "step": 13010
    },
    {
      "epoch": 6.136256482791136,
      "grad_norm": 1.102852463722229,
      "learning_rate": 3.8039018055520234e-05,
      "loss": 0.2129,
      "num_input_tokens_seen": 11284192,
      "step": 13015
    },
    {
      "epoch": 6.138613861386139,
      "grad_norm": 1.3184231519699097,
      "learning_rate": 3.803064063985844e-05,
      "loss": 0.2078,
      "num_input_tokens_seen": 11288992,
      "step": 13020
    },
    {
      "epoch": 6.140971239981141,
      "grad_norm": 1.2605044841766357,
      "learning_rate": 3.802226121470811e-05,
      "loss": 0.1974,
      "num_input_tokens_seen": 11293456,
      "step": 13025
    },
    {
      "epoch": 6.1433286185761435,
      "grad_norm": 0.8714604377746582,
      "learning_rate": 3.801387978136145e-05,
      "loss": 0.1483,
      "num_input_tokens_seen": 11297488,
      "step": 13030
    },
    {
      "epoch": 6.145685997171146,
      "grad_norm": 0.5949074625968933,
      "learning_rate": 3.800549634111099e-05,
      "loss": 0.2224,
      "num_input_tokens_seen": 11302096,
      "step": 13035
    },
    {
      "epoch": 6.148043375766148,
      "grad_norm": 0.7974242568016052,
      "learning_rate": 3.799711089524955e-05,
      "loss": 0.1999,
      "num_input_tokens_seen": 11305856,
      "step": 13040
    },
    {
      "epoch": 6.150400754361151,
      "grad_norm": 0.8033232092857361,
      "learning_rate": 3.7988723445070285e-05,
      "loss": 0.1603,
      "num_input_tokens_seen": 11310768,
      "step": 13045
    },
    {
      "epoch": 6.152758132956153,
      "grad_norm": 2.116393566131592,
      "learning_rate": 3.798033399186663e-05,
      "loss": 0.2007,
      "num_input_tokens_seen": 11315152,
      "step": 13050
    },
    {
      "epoch": 6.1551155115511555,
      "grad_norm": 1.8613065481185913,
      "learning_rate": 3.797194253693237e-05,
      "loss": 0.2267,
      "num_input_tokens_seen": 11319648,
      "step": 13055
    },
    {
      "epoch": 6.157472890146157,
      "grad_norm": 1.052126169204712,
      "learning_rate": 3.796354908156153e-05,
      "loss": 0.1828,
      "num_input_tokens_seen": 11323664,
      "step": 13060
    },
    {
      "epoch": 6.1598302687411595,
      "grad_norm": 1.0072948932647705,
      "learning_rate": 3.795515362704853e-05,
      "loss": 0.195,
      "num_input_tokens_seen": 11328384,
      "step": 13065
    },
    {
      "epoch": 6.162187647336162,
      "grad_norm": 0.9186083078384399,
      "learning_rate": 3.794675617468803e-05,
      "loss": 0.2133,
      "num_input_tokens_seen": 11332608,
      "step": 13070
    },
    {
      "epoch": 6.164545025931164,
      "grad_norm": 1.085052251815796,
      "learning_rate": 3.793835672577503e-05,
      "loss": 0.1525,
      "num_input_tokens_seen": 11337088,
      "step": 13075
    },
    {
      "epoch": 6.166902404526167,
      "grad_norm": 0.765601396560669,
      "learning_rate": 3.7929955281604826e-05,
      "loss": 0.2186,
      "num_input_tokens_seen": 11340864,
      "step": 13080
    },
    {
      "epoch": 6.169259783121169,
      "grad_norm": 1.062495231628418,
      "learning_rate": 3.7921551843473036e-05,
      "loss": 0.1761,
      "num_input_tokens_seen": 11345712,
      "step": 13085
    },
    {
      "epoch": 6.1716171617161715,
      "grad_norm": 1.351740837097168,
      "learning_rate": 3.791314641267557e-05,
      "loss": 0.1702,
      "num_input_tokens_seen": 11349296,
      "step": 13090
    },
    {
      "epoch": 6.173974540311174,
      "grad_norm": 1.1895322799682617,
      "learning_rate": 3.790473899050864e-05,
      "loss": 0.2449,
      "num_input_tokens_seen": 11353200,
      "step": 13095
    },
    {
      "epoch": 6.176331918906176,
      "grad_norm": 1.0660685300827026,
      "learning_rate": 3.7896329578268794e-05,
      "loss": 0.1813,
      "num_input_tokens_seen": 11357184,
      "step": 13100
    },
    {
      "epoch": 6.178689297501179,
      "grad_norm": 1.1776554584503174,
      "learning_rate": 3.7887918177252855e-05,
      "loss": 0.1349,
      "num_input_tokens_seen": 11360736,
      "step": 13105
    },
    {
      "epoch": 6.181046676096181,
      "grad_norm": 2.6535680294036865,
      "learning_rate": 3.787950478875798e-05,
      "loss": 0.1828,
      "num_input_tokens_seen": 11365888,
      "step": 13110
    },
    {
      "epoch": 6.1834040546911835,
      "grad_norm": 0.7152304649353027,
      "learning_rate": 3.787108941408162e-05,
      "loss": 0.2529,
      "num_input_tokens_seen": 11370256,
      "step": 13115
    },
    {
      "epoch": 6.185761433286186,
      "grad_norm": 0.9372276067733765,
      "learning_rate": 3.786267205452151e-05,
      "loss": 0.1903,
      "num_input_tokens_seen": 11375248,
      "step": 13120
    },
    {
      "epoch": 6.188118811881188,
      "grad_norm": 1.4611320495605469,
      "learning_rate": 3.785425271137573e-05,
      "loss": 0.1898,
      "num_input_tokens_seen": 11380064,
      "step": 13125
    },
    {
      "epoch": 6.190476190476191,
      "grad_norm": 1.4205466508865356,
      "learning_rate": 3.7845831385942655e-05,
      "loss": 0.1721,
      "num_input_tokens_seen": 11384784,
      "step": 13130
    },
    {
      "epoch": 6.192833569071193,
      "grad_norm": 0.8435620665550232,
      "learning_rate": 3.7837408079520944e-05,
      "loss": 0.1468,
      "num_input_tokens_seen": 11389104,
      "step": 13135
    },
    {
      "epoch": 6.1951909476661955,
      "grad_norm": 0.5091607570648193,
      "learning_rate": 3.782898279340957e-05,
      "loss": 0.1236,
      "num_input_tokens_seen": 11393200,
      "step": 13140
    },
    {
      "epoch": 6.197548326261198,
      "grad_norm": 0.668104350566864,
      "learning_rate": 3.782055552890784e-05,
      "loss": 0.2082,
      "num_input_tokens_seen": 11397200,
      "step": 13145
    },
    {
      "epoch": 6.1999057048562,
      "grad_norm": 1.2269669771194458,
      "learning_rate": 3.781212628731534e-05,
      "loss": 0.1529,
      "num_input_tokens_seen": 11400816,
      "step": 13150
    },
    {
      "epoch": 6.202263083451202,
      "grad_norm": 1.5866721868515015,
      "learning_rate": 3.7803695069931946e-05,
      "loss": 0.1709,
      "num_input_tokens_seen": 11404432,
      "step": 13155
    },
    {
      "epoch": 6.204620462046204,
      "grad_norm": 0.6721461415290833,
      "learning_rate": 3.779526187805789e-05,
      "loss": 0.1693,
      "num_input_tokens_seen": 11408528,
      "step": 13160
    },
    {
      "epoch": 6.206977840641207,
      "grad_norm": 1.3343712091445923,
      "learning_rate": 3.778682671299364e-05,
      "loss": 0.1973,
      "num_input_tokens_seen": 11413424,
      "step": 13165
    },
    {
      "epoch": 6.209335219236209,
      "grad_norm": 1.706442952156067,
      "learning_rate": 3.777838957604003e-05,
      "loss": 0.1849,
      "num_input_tokens_seen": 11416944,
      "step": 13170
    },
    {
      "epoch": 6.211692597831211,
      "grad_norm": 0.5190753936767578,
      "learning_rate": 3.776995046849816e-05,
      "loss": 0.2001,
      "num_input_tokens_seen": 11420848,
      "step": 13175
    },
    {
      "epoch": 6.214049976426214,
      "grad_norm": 0.5779367089271545,
      "learning_rate": 3.776150939166945e-05,
      "loss": 0.1714,
      "num_input_tokens_seen": 11424976,
      "step": 13180
    },
    {
      "epoch": 6.216407355021216,
      "grad_norm": 0.48863789439201355,
      "learning_rate": 3.775306634685562e-05,
      "loss": 0.1397,
      "num_input_tokens_seen": 11429008,
      "step": 13185
    },
    {
      "epoch": 6.218764733616219,
      "grad_norm": 0.5639715194702148,
      "learning_rate": 3.7744621335358696e-05,
      "loss": 0.1604,
      "num_input_tokens_seen": 11432928,
      "step": 13190
    },
    {
      "epoch": 6.221122112211221,
      "grad_norm": 1.3698310852050781,
      "learning_rate": 3.7736174358481e-05,
      "loss": 0.1756,
      "num_input_tokens_seen": 11437120,
      "step": 13195
    },
    {
      "epoch": 6.223479490806223,
      "grad_norm": 2.553093433380127,
      "learning_rate": 3.7727725417525175e-05,
      "loss": 0.2759,
      "num_input_tokens_seen": 11441120,
      "step": 13200
    },
    {
      "epoch": 6.223479490806223,
      "eval_loss": 0.20718315243721008,
      "eval_runtime": 22.1742,
      "eval_samples_per_second": 42.527,
      "eval_steps_per_second": 21.286,
      "num_input_tokens_seen": 11441120,
      "step": 13200
    },
    {
      "epoch": 6.225836869401226,
      "grad_norm": 1.3967981338500977,
      "learning_rate": 3.771927451379414e-05,
      "loss": 0.206,
      "num_input_tokens_seen": 11445408,
      "step": 13205
    },
    {
      "epoch": 6.228194247996228,
      "grad_norm": 1.3453385829925537,
      "learning_rate": 3.7710821648591135e-05,
      "loss": 0.2301,
      "num_input_tokens_seen": 11449024,
      "step": 13210
    },
    {
      "epoch": 6.230551626591231,
      "grad_norm": 0.8518226146697998,
      "learning_rate": 3.7702366823219694e-05,
      "loss": 0.1587,
      "num_input_tokens_seen": 11453728,
      "step": 13215
    },
    {
      "epoch": 6.232909005186233,
      "grad_norm": 0.5258207321166992,
      "learning_rate": 3.769391003898366e-05,
      "loss": 0.1783,
      "num_input_tokens_seen": 11457600,
      "step": 13220
    },
    {
      "epoch": 6.235266383781235,
      "grad_norm": 1.445528268814087,
      "learning_rate": 3.768545129718718e-05,
      "loss": 0.1671,
      "num_input_tokens_seen": 11461696,
      "step": 13225
    },
    {
      "epoch": 6.237623762376238,
      "grad_norm": 0.81913161277771,
      "learning_rate": 3.7676990599134686e-05,
      "loss": 0.2379,
      "num_input_tokens_seen": 11465520,
      "step": 13230
    },
    {
      "epoch": 6.23998114097124,
      "grad_norm": 1.3636438846588135,
      "learning_rate": 3.766852794613095e-05,
      "loss": 0.2179,
      "num_input_tokens_seen": 11470688,
      "step": 13235
    },
    {
      "epoch": 6.242338519566243,
      "grad_norm": 1.791720986366272,
      "learning_rate": 3.766006333948099e-05,
      "loss": 0.2317,
      "num_input_tokens_seen": 11475664,
      "step": 13240
    },
    {
      "epoch": 6.244695898161245,
      "grad_norm": 0.6000313758850098,
      "learning_rate": 3.765159678049017e-05,
      "loss": 0.227,
      "num_input_tokens_seen": 11479984,
      "step": 13245
    },
    {
      "epoch": 6.247053276756247,
      "grad_norm": 1.0001925230026245,
      "learning_rate": 3.7643128270464134e-05,
      "loss": 0.1827,
      "num_input_tokens_seen": 11484928,
      "step": 13250
    },
    {
      "epoch": 6.24941065535125,
      "grad_norm": 0.7851129174232483,
      "learning_rate": 3.763465781070884e-05,
      "loss": 0.1863,
      "num_input_tokens_seen": 11489760,
      "step": 13255
    },
    {
      "epoch": 6.251768033946251,
      "grad_norm": 1.67833411693573,
      "learning_rate": 3.762618540253052e-05,
      "loss": 0.236,
      "num_input_tokens_seen": 11493472,
      "step": 13260
    },
    {
      "epoch": 6.254125412541254,
      "grad_norm": 0.4987073838710785,
      "learning_rate": 3.761771104723576e-05,
      "loss": 0.1309,
      "num_input_tokens_seen": 11497152,
      "step": 13265
    },
    {
      "epoch": 6.256482791136256,
      "grad_norm": 0.9067997932434082,
      "learning_rate": 3.7609234746131386e-05,
      "loss": 0.2038,
      "num_input_tokens_seen": 11501568,
      "step": 13270
    },
    {
      "epoch": 6.258840169731259,
      "grad_norm": 1.7554019689559937,
      "learning_rate": 3.7600756500524556e-05,
      "loss": 0.2569,
      "num_input_tokens_seen": 11505456,
      "step": 13275
    },
    {
      "epoch": 6.261197548326261,
      "grad_norm": 0.8539285063743591,
      "learning_rate": 3.759227631172271e-05,
      "loss": 0.1472,
      "num_input_tokens_seen": 11509680,
      "step": 13280
    },
    {
      "epoch": 6.263554926921263,
      "grad_norm": 0.6618529558181763,
      "learning_rate": 3.758379418103363e-05,
      "loss": 0.1509,
      "num_input_tokens_seen": 11514208,
      "step": 13285
    },
    {
      "epoch": 6.265912305516266,
      "grad_norm": 1.3237017393112183,
      "learning_rate": 3.757531010976534e-05,
      "loss": 0.1574,
      "num_input_tokens_seen": 11518128,
      "step": 13290
    },
    {
      "epoch": 6.268269684111268,
      "grad_norm": 1.5029103755950928,
      "learning_rate": 3.75668240992262e-05,
      "loss": 0.1893,
      "num_input_tokens_seen": 11522384,
      "step": 13295
    },
    {
      "epoch": 6.270627062706271,
      "grad_norm": 0.7994213104248047,
      "learning_rate": 3.7558336150724865e-05,
      "loss": 0.1806,
      "num_input_tokens_seen": 11526848,
      "step": 13300
    },
    {
      "epoch": 6.272984441301273,
      "grad_norm": 0.378937304019928,
      "learning_rate": 3.754984626557028e-05,
      "loss": 0.1637,
      "num_input_tokens_seen": 11531360,
      "step": 13305
    },
    {
      "epoch": 6.275341819896275,
      "grad_norm": 0.5819053649902344,
      "learning_rate": 3.754135444507168e-05,
      "loss": 0.1473,
      "num_input_tokens_seen": 11536464,
      "step": 13310
    },
    {
      "epoch": 6.277699198491278,
      "grad_norm": 1.7988368272781372,
      "learning_rate": 3.753286069053863e-05,
      "loss": 0.1796,
      "num_input_tokens_seen": 11540816,
      "step": 13315
    },
    {
      "epoch": 6.28005657708628,
      "grad_norm": 2.1682186126708984,
      "learning_rate": 3.7524365003280945e-05,
      "loss": 0.2194,
      "num_input_tokens_seen": 11547056,
      "step": 13320
    },
    {
      "epoch": 6.282413955681283,
      "grad_norm": 0.6576429605484009,
      "learning_rate": 3.75158673846088e-05,
      "loss": 0.1731,
      "num_input_tokens_seen": 11550496,
      "step": 13325
    },
    {
      "epoch": 6.284771334276285,
      "grad_norm": 0.6749131679534912,
      "learning_rate": 3.750736783583262e-05,
      "loss": 0.1921,
      "num_input_tokens_seen": 11554240,
      "step": 13330
    },
    {
      "epoch": 6.287128712871287,
      "grad_norm": 0.8273175954818726,
      "learning_rate": 3.7498866358263144e-05,
      "loss": 0.1635,
      "num_input_tokens_seen": 11558064,
      "step": 13335
    },
    {
      "epoch": 6.28948609146629,
      "grad_norm": 0.5181607604026794,
      "learning_rate": 3.74903629532114e-05,
      "loss": 0.1576,
      "num_input_tokens_seen": 11562368,
      "step": 13340
    },
    {
      "epoch": 6.291843470061292,
      "grad_norm": 0.5963915586471558,
      "learning_rate": 3.748185762198873e-05,
      "loss": 0.1956,
      "num_input_tokens_seen": 11566704,
      "step": 13345
    },
    {
      "epoch": 6.294200848656295,
      "grad_norm": 0.5621787309646606,
      "learning_rate": 3.747335036590676e-05,
      "loss": 0.1854,
      "num_input_tokens_seen": 11571408,
      "step": 13350
    },
    {
      "epoch": 6.296558227251296,
      "grad_norm": 0.8821688294410706,
      "learning_rate": 3.7464841186277405e-05,
      "loss": 0.1809,
      "num_input_tokens_seen": 11575408,
      "step": 13355
    },
    {
      "epoch": 6.2989156058462985,
      "grad_norm": 0.6320316791534424,
      "learning_rate": 3.7456330084412896e-05,
      "loss": 0.2849,
      "num_input_tokens_seen": 11580160,
      "step": 13360
    },
    {
      "epoch": 6.301272984441301,
      "grad_norm": 1.3906669616699219,
      "learning_rate": 3.744781706162576e-05,
      "loss": 0.1824,
      "num_input_tokens_seen": 11584144,
      "step": 13365
    },
    {
      "epoch": 6.303630363036303,
      "grad_norm": 1.8216643333435059,
      "learning_rate": 3.743930211922879e-05,
      "loss": 0.1878,
      "num_input_tokens_seen": 11588048,
      "step": 13370
    },
    {
      "epoch": 6.305987741631306,
      "grad_norm": 1.2399638891220093,
      "learning_rate": 3.743078525853513e-05,
      "loss": 0.2638,
      "num_input_tokens_seen": 11592144,
      "step": 13375
    },
    {
      "epoch": 6.308345120226308,
      "grad_norm": 1.1385096311569214,
      "learning_rate": 3.7422266480858154e-05,
      "loss": 0.2709,
      "num_input_tokens_seen": 11596080,
      "step": 13380
    },
    {
      "epoch": 6.3107024988213105,
      "grad_norm": 1.2243093252182007,
      "learning_rate": 3.741374578751158e-05,
      "loss": 0.199,
      "num_input_tokens_seen": 11601232,
      "step": 13385
    },
    {
      "epoch": 6.313059877416313,
      "grad_norm": 0.48648905754089355,
      "learning_rate": 3.740522317980941e-05,
      "loss": 0.1557,
      "num_input_tokens_seen": 11605472,
      "step": 13390
    },
    {
      "epoch": 6.315417256011315,
      "grad_norm": 1.2856255769729614,
      "learning_rate": 3.739669865906593e-05,
      "loss": 0.2011,
      "num_input_tokens_seen": 11610256,
      "step": 13395
    },
    {
      "epoch": 6.317774634606318,
      "grad_norm": 0.9343747496604919,
      "learning_rate": 3.738817222659573e-05,
      "loss": 0.2091,
      "num_input_tokens_seen": 11614176,
      "step": 13400
    },
    {
      "epoch": 6.317774634606318,
      "eval_loss": 0.2065698355436325,
      "eval_runtime": 22.1597,
      "eval_samples_per_second": 42.555,
      "eval_steps_per_second": 21.3,
      "num_input_tokens_seen": 11614176,
      "step": 13400
    },
    {
      "epoch": 6.32013201320132,
      "grad_norm": 1.2955689430236816,
      "learning_rate": 3.73796438837137e-05,
      "loss": 0.1782,
      "num_input_tokens_seen": 11617808,
      "step": 13405
    },
    {
      "epoch": 6.3224893917963225,
      "grad_norm": 0.7582072019577026,
      "learning_rate": 3.7371113631735e-05,
      "loss": 0.1908,
      "num_input_tokens_seen": 11622544,
      "step": 13410
    },
    {
      "epoch": 6.324846770391325,
      "grad_norm": 0.4664658010005951,
      "learning_rate": 3.736258147197512e-05,
      "loss": 0.195,
      "num_input_tokens_seen": 11627168,
      "step": 13415
    },
    {
      "epoch": 6.327204148986327,
      "grad_norm": 0.9612610340118408,
      "learning_rate": 3.735404740574981e-05,
      "loss": 0.1128,
      "num_input_tokens_seen": 11631296,
      "step": 13420
    },
    {
      "epoch": 6.32956152758133,
      "grad_norm": 0.6345967054367065,
      "learning_rate": 3.7345511434375145e-05,
      "loss": 0.1165,
      "num_input_tokens_seen": 11636192,
      "step": 13425
    },
    {
      "epoch": 6.331918906176332,
      "grad_norm": 1.2520480155944824,
      "learning_rate": 3.733697355916748e-05,
      "loss": 0.2099,
      "num_input_tokens_seen": 11640592,
      "step": 13430
    },
    {
      "epoch": 6.3342762847713345,
      "grad_norm": 1.0385510921478271,
      "learning_rate": 3.732843378144345e-05,
      "loss": 0.1799,
      "num_input_tokens_seen": 11645024,
      "step": 13435
    },
    {
      "epoch": 6.336633663366337,
      "grad_norm": 1.5858299732208252,
      "learning_rate": 3.7319892102519995e-05,
      "loss": 0.1939,
      "num_input_tokens_seen": 11649184,
      "step": 13440
    },
    {
      "epoch": 6.338991041961339,
      "grad_norm": 0.4694328010082245,
      "learning_rate": 3.731134852371436e-05,
      "loss": 0.1491,
      "num_input_tokens_seen": 11653584,
      "step": 13445
    },
    {
      "epoch": 6.341348420556342,
      "grad_norm": 1.4280428886413574,
      "learning_rate": 3.730280304634408e-05,
      "loss": 0.1534,
      "num_input_tokens_seen": 11658176,
      "step": 13450
    },
    {
      "epoch": 6.343705799151344,
      "grad_norm": 1.2720320224761963,
      "learning_rate": 3.729425567172696e-05,
      "loss": 0.166,
      "num_input_tokens_seen": 11663008,
      "step": 13455
    },
    {
      "epoch": 6.346063177746346,
      "grad_norm": 2.3222458362579346,
      "learning_rate": 3.728570640118111e-05,
      "loss": 0.2573,
      "num_input_tokens_seen": 11668144,
      "step": 13460
    },
    {
      "epoch": 6.348420556341348,
      "grad_norm": 0.7774571776390076,
      "learning_rate": 3.727715523602494e-05,
      "loss": 0.1972,
      "num_input_tokens_seen": 11672048,
      "step": 13465
    },
    {
      "epoch": 6.3507779349363505,
      "grad_norm": 1.1548665761947632,
      "learning_rate": 3.726860217757715e-05,
      "loss": 0.0997,
      "num_input_tokens_seen": 11676000,
      "step": 13470
    },
    {
      "epoch": 6.353135313531353,
      "grad_norm": 0.9782509803771973,
      "learning_rate": 3.726004722715673e-05,
      "loss": 0.221,
      "num_input_tokens_seen": 11680224,
      "step": 13475
    },
    {
      "epoch": 6.355492692126355,
      "grad_norm": 2.512582302093506,
      "learning_rate": 3.725149038608296e-05,
      "loss": 0.2189,
      "num_input_tokens_seen": 11684240,
      "step": 13480
    },
    {
      "epoch": 6.357850070721358,
      "grad_norm": 1.3277900218963623,
      "learning_rate": 3.7242931655675404e-05,
      "loss": 0.1466,
      "num_input_tokens_seen": 11689632,
      "step": 13485
    },
    {
      "epoch": 6.36020744931636,
      "grad_norm": 1.2812936305999756,
      "learning_rate": 3.7234371037253937e-05,
      "loss": 0.2599,
      "num_input_tokens_seen": 11693328,
      "step": 13490
    },
    {
      "epoch": 6.3625648279113625,
      "grad_norm": 0.6678707599639893,
      "learning_rate": 3.7225808532138705e-05,
      "loss": 0.1645,
      "num_input_tokens_seen": 11697984,
      "step": 13495
    },
    {
      "epoch": 6.364922206506365,
      "grad_norm": 0.5342229008674622,
      "learning_rate": 3.721724414165016e-05,
      "loss": 0.1656,
      "num_input_tokens_seen": 11702480,
      "step": 13500
    },
    {
      "epoch": 6.367279585101367,
      "grad_norm": 1.297614336013794,
      "learning_rate": 3.720867786710904e-05,
      "loss": 0.2272,
      "num_input_tokens_seen": 11706128,
      "step": 13505
    },
    {
      "epoch": 6.36963696369637,
      "grad_norm": 0.8747941851615906,
      "learning_rate": 3.7200109709836366e-05,
      "loss": 0.2229,
      "num_input_tokens_seen": 11709680,
      "step": 13510
    },
    {
      "epoch": 6.371994342291372,
      "grad_norm": 1.4415897130966187,
      "learning_rate": 3.7191539671153465e-05,
      "loss": 0.1973,
      "num_input_tokens_seen": 11714976,
      "step": 13515
    },
    {
      "epoch": 6.3743517208863745,
      "grad_norm": 1.200190782546997,
      "learning_rate": 3.718296775238193e-05,
      "loss": 0.1505,
      "num_input_tokens_seen": 11718704,
      "step": 13520
    },
    {
      "epoch": 6.376709099481377,
      "grad_norm": 1.016242265701294,
      "learning_rate": 3.7174393954843675e-05,
      "loss": 0.1379,
      "num_input_tokens_seen": 11722768,
      "step": 13525
    },
    {
      "epoch": 6.379066478076379,
      "grad_norm": 2.051754951477051,
      "learning_rate": 3.716581827986087e-05,
      "loss": 0.2237,
      "num_input_tokens_seen": 11727136,
      "step": 13530
    },
    {
      "epoch": 6.381423856671382,
      "grad_norm": 0.8109329342842102,
      "learning_rate": 3.7157240728756004e-05,
      "loss": 0.2049,
      "num_input_tokens_seen": 11730400,
      "step": 13535
    },
    {
      "epoch": 6.383781235266384,
      "grad_norm": 0.9528835415840149,
      "learning_rate": 3.714866130285184e-05,
      "loss": 0.1532,
      "num_input_tokens_seen": 11734736,
      "step": 13540
    },
    {
      "epoch": 6.3861386138613865,
      "grad_norm": 1.1648104190826416,
      "learning_rate": 3.714008000347143e-05,
      "loss": 0.2294,
      "num_input_tokens_seen": 11739744,
      "step": 13545
    },
    {
      "epoch": 6.388495992456389,
      "grad_norm": 0.7069249153137207,
      "learning_rate": 3.7131496831938126e-05,
      "loss": 0.1866,
      "num_input_tokens_seen": 11743712,
      "step": 13550
    },
    {
      "epoch": 6.39085337105139,
      "grad_norm": 0.703685462474823,
      "learning_rate": 3.7122911789575565e-05,
      "loss": 0.1638,
      "num_input_tokens_seen": 11748384,
      "step": 13555
    },
    {
      "epoch": 6.393210749646393,
      "grad_norm": 0.6452196836471558,
      "learning_rate": 3.711432487770765e-05,
      "loss": 0.2022,
      "num_input_tokens_seen": 11753120,
      "step": 13560
    },
    {
      "epoch": 6.395568128241395,
      "grad_norm": 0.8301488757133484,
      "learning_rate": 3.710573609765861e-05,
      "loss": 0.1927,
      "num_input_tokens_seen": 11757744,
      "step": 13565
    },
    {
      "epoch": 6.397925506836398,
      "grad_norm": 0.5942070484161377,
      "learning_rate": 3.709714545075292e-05,
      "loss": 0.1942,
      "num_input_tokens_seen": 11761600,
      "step": 13570
    },
    {
      "epoch": 6.4002828854314,
      "grad_norm": 0.8880425691604614,
      "learning_rate": 3.708855293831538e-05,
      "loss": 0.2018,
      "num_input_tokens_seen": 11765472,
      "step": 13575
    },
    {
      "epoch": 6.402640264026402,
      "grad_norm": 0.5999334454536438,
      "learning_rate": 3.707995856167107e-05,
      "loss": 0.1551,
      "num_input_tokens_seen": 11769824,
      "step": 13580
    },
    {
      "epoch": 6.404997642621405,
      "grad_norm": 1.5092194080352783,
      "learning_rate": 3.707136232214534e-05,
      "loss": 0.1545,
      "num_input_tokens_seen": 11773648,
      "step": 13585
    },
    {
      "epoch": 6.407355021216407,
      "grad_norm": 0.6895807385444641,
      "learning_rate": 3.7062764221063844e-05,
      "loss": 0.1751,
      "num_input_tokens_seen": 11777456,
      "step": 13590
    },
    {
      "epoch": 6.40971239981141,
      "grad_norm": 1.7283684015274048,
      "learning_rate": 3.705416425975252e-05,
      "loss": 0.2307,
      "num_input_tokens_seen": 11781792,
      "step": 13595
    },
    {
      "epoch": 6.412069778406412,
      "grad_norm": 0.5530927181243896,
      "learning_rate": 3.704556243953758e-05,
      "loss": 0.1318,
      "num_input_tokens_seen": 11785424,
      "step": 13600
    },
    {
      "epoch": 6.412069778406412,
      "eval_loss": 0.20556892454624176,
      "eval_runtime": 22.1349,
      "eval_samples_per_second": 42.602,
      "eval_steps_per_second": 21.324,
      "num_input_tokens_seen": 11785424,
      "step": 13600
    },
    {
      "epoch": 6.414427157001414,
      "grad_norm": 0.7340503334999084,
      "learning_rate": 3.7036958761745535e-05,
      "loss": 0.1363,
      "num_input_tokens_seen": 11789584,
      "step": 13605
    },
    {
      "epoch": 6.416784535596417,
      "grad_norm": 1.195406198501587,
      "learning_rate": 3.702835322770318e-05,
      "loss": 0.1395,
      "num_input_tokens_seen": 11793728,
      "step": 13610
    },
    {
      "epoch": 6.419141914191419,
      "grad_norm": 0.8353258371353149,
      "learning_rate": 3.701974583873761e-05,
      "loss": 0.1756,
      "num_input_tokens_seen": 11798288,
      "step": 13615
    },
    {
      "epoch": 6.421499292786422,
      "grad_norm": 0.9308305382728577,
      "learning_rate": 3.701113659617618e-05,
      "loss": 0.1816,
      "num_input_tokens_seen": 11802368,
      "step": 13620
    },
    {
      "epoch": 6.423856671381424,
      "grad_norm": 1.218442440032959,
      "learning_rate": 3.7002525501346535e-05,
      "loss": 0.1711,
      "num_input_tokens_seen": 11806656,
      "step": 13625
    },
    {
      "epoch": 6.426214049976426,
      "grad_norm": 1.0987640619277954,
      "learning_rate": 3.699391255557664e-05,
      "loss": 0.2738,
      "num_input_tokens_seen": 11810912,
      "step": 13630
    },
    {
      "epoch": 6.428571428571429,
      "grad_norm": 0.6433650255203247,
      "learning_rate": 3.69852977601947e-05,
      "loss": 0.1808,
      "num_input_tokens_seen": 11815056,
      "step": 13635
    },
    {
      "epoch": 6.430928807166431,
      "grad_norm": 1.9589293003082275,
      "learning_rate": 3.697668111652922e-05,
      "loss": 0.2427,
      "num_input_tokens_seen": 11819824,
      "step": 13640
    },
    {
      "epoch": 6.433286185761434,
      "grad_norm": 0.43151912093162537,
      "learning_rate": 3.6968062625909005e-05,
      "loss": 0.1476,
      "num_input_tokens_seen": 11823888,
      "step": 13645
    },
    {
      "epoch": 6.435643564356436,
      "grad_norm": 1.5660583972930908,
      "learning_rate": 3.6959442289663135e-05,
      "loss": 0.2018,
      "num_input_tokens_seen": 11828256,
      "step": 13650
    },
    {
      "epoch": 6.438000942951438,
      "grad_norm": 0.6728929877281189,
      "learning_rate": 3.695082010912098e-05,
      "loss": 0.1948,
      "num_input_tokens_seen": 11832848,
      "step": 13655
    },
    {
      "epoch": 6.44035832154644,
      "grad_norm": 0.8286604285240173,
      "learning_rate": 3.694219608561217e-05,
      "loss": 0.1563,
      "num_input_tokens_seen": 11837616,
      "step": 13660
    },
    {
      "epoch": 6.442715700141442,
      "grad_norm": 1.0475046634674072,
      "learning_rate": 3.693357022046665e-05,
      "loss": 0.1657,
      "num_input_tokens_seen": 11842144,
      "step": 13665
    },
    {
      "epoch": 6.445073078736445,
      "grad_norm": 1.3885421752929688,
      "learning_rate": 3.6924942515014644e-05,
      "loss": 0.1462,
      "num_input_tokens_seen": 11846560,
      "step": 13670
    },
    {
      "epoch": 6.447430457331447,
      "grad_norm": 1.4441273212432861,
      "learning_rate": 3.691631297058664e-05,
      "loss": 0.2237,
      "num_input_tokens_seen": 11850208,
      "step": 13675
    },
    {
      "epoch": 6.4497878359264496,
      "grad_norm": 0.5419784784317017,
      "learning_rate": 3.6907681588513424e-05,
      "loss": 0.1089,
      "num_input_tokens_seen": 11854624,
      "step": 13680
    },
    {
      "epoch": 6.452145214521452,
      "grad_norm": 0.8798506855964661,
      "learning_rate": 3.689904837012606e-05,
      "loss": 0.2376,
      "num_input_tokens_seen": 11859680,
      "step": 13685
    },
    {
      "epoch": 6.454502593116454,
      "grad_norm": 0.8685320019721985,
      "learning_rate": 3.689041331675591e-05,
      "loss": 0.1541,
      "num_input_tokens_seen": 11863888,
      "step": 13690
    },
    {
      "epoch": 6.456859971711457,
      "grad_norm": 1.4354612827301025,
      "learning_rate": 3.688177642973461e-05,
      "loss": 0.1951,
      "num_input_tokens_seen": 11868640,
      "step": 13695
    },
    {
      "epoch": 6.459217350306459,
      "grad_norm": 2.365914821624756,
      "learning_rate": 3.687313771039406e-05,
      "loss": 0.2199,
      "num_input_tokens_seen": 11872816,
      "step": 13700
    },
    {
      "epoch": 6.461574728901462,
      "grad_norm": 2.002246379852295,
      "learning_rate": 3.686449716006647e-05,
      "loss": 0.2273,
      "num_input_tokens_seen": 11876288,
      "step": 13705
    },
    {
      "epoch": 6.463932107496464,
      "grad_norm": 1.2810865640640259,
      "learning_rate": 3.685585478008432e-05,
      "loss": 0.2332,
      "num_input_tokens_seen": 11880528,
      "step": 13710
    },
    {
      "epoch": 6.466289486091466,
      "grad_norm": 1.141167402267456,
      "learning_rate": 3.6847210571780364e-05,
      "loss": 0.1754,
      "num_input_tokens_seen": 11884512,
      "step": 13715
    },
    {
      "epoch": 6.468646864686469,
      "grad_norm": 1.0862747430801392,
      "learning_rate": 3.683856453648767e-05,
      "loss": 0.265,
      "num_input_tokens_seen": 11889184,
      "step": 13720
    },
    {
      "epoch": 6.471004243281471,
      "grad_norm": 0.5858092904090881,
      "learning_rate": 3.682991667553954e-05,
      "loss": 0.1605,
      "num_input_tokens_seen": 11894064,
      "step": 13725
    },
    {
      "epoch": 6.473361621876474,
      "grad_norm": 0.7114397883415222,
      "learning_rate": 3.6821266990269606e-05,
      "loss": 0.1336,
      "num_input_tokens_seen": 11898992,
      "step": 13730
    },
    {
      "epoch": 6.475719000471476,
      "grad_norm": 0.5629537105560303,
      "learning_rate": 3.681261548201174e-05,
      "loss": 0.2681,
      "num_input_tokens_seen": 11903840,
      "step": 13735
    },
    {
      "epoch": 6.478076379066478,
      "grad_norm": 0.44694748520851135,
      "learning_rate": 3.6803962152100125e-05,
      "loss": 0.1746,
      "num_input_tokens_seen": 11907792,
      "step": 13740
    },
    {
      "epoch": 6.480433757661481,
      "grad_norm": 0.6795097589492798,
      "learning_rate": 3.67953070018692e-05,
      "loss": 0.1934,
      "num_input_tokens_seen": 11911536,
      "step": 13745
    },
    {
      "epoch": 6.482791136256483,
      "grad_norm": 0.4637584984302521,
      "learning_rate": 3.678665003265371e-05,
      "loss": 0.1677,
      "num_input_tokens_seen": 11915920,
      "step": 13750
    },
    {
      "epoch": 6.485148514851485,
      "grad_norm": 0.42765241861343384,
      "learning_rate": 3.677799124578867e-05,
      "loss": 0.1595,
      "num_input_tokens_seen": 11920528,
      "step": 13755
    },
    {
      "epoch": 6.487505893446487,
      "grad_norm": 0.7417716979980469,
      "learning_rate": 3.676933064260937e-05,
      "loss": 0.207,
      "num_input_tokens_seen": 11924784,
      "step": 13760
    },
    {
      "epoch": 6.4898632720414895,
      "grad_norm": 1.4932894706726074,
      "learning_rate": 3.6760668224451365e-05,
      "loss": 0.1812,
      "num_input_tokens_seen": 11928976,
      "step": 13765
    },
    {
      "epoch": 6.492220650636492,
      "grad_norm": 0.8235538005828857,
      "learning_rate": 3.675200399265054e-05,
      "loss": 0.2025,
      "num_input_tokens_seen": 11933984,
      "step": 13770
    },
    {
      "epoch": 6.494578029231494,
      "grad_norm": 1.4214637279510498,
      "learning_rate": 3.6743337948543014e-05,
      "loss": 0.2358,
      "num_input_tokens_seen": 11938400,
      "step": 13775
    },
    {
      "epoch": 6.496935407826497,
      "grad_norm": 1.0226045846939087,
      "learning_rate": 3.6734670093465204e-05,
      "loss": 0.1996,
      "num_input_tokens_seen": 11942928,
      "step": 13780
    },
    {
      "epoch": 6.499292786421499,
      "grad_norm": 0.9882229566574097,
      "learning_rate": 3.672600042875379e-05,
      "loss": 0.1736,
      "num_input_tokens_seen": 11947536,
      "step": 13785
    },
    {
      "epoch": 6.5016501650165015,
      "grad_norm": 0.9304018616676331,
      "learning_rate": 3.671732895574575e-05,
      "loss": 0.1788,
      "num_input_tokens_seen": 11951392,
      "step": 13790
    },
    {
      "epoch": 6.504007543611504,
      "grad_norm": 1.1033076047897339,
      "learning_rate": 3.670865567577834e-05,
      "loss": 0.2378,
      "num_input_tokens_seen": 11955776,
      "step": 13795
    },
    {
      "epoch": 6.506364922206506,
      "grad_norm": 1.4068925380706787,
      "learning_rate": 3.669998059018909e-05,
      "loss": 0.1866,
      "num_input_tokens_seen": 11960752,
      "step": 13800
    },
    {
      "epoch": 6.506364922206506,
      "eval_loss": 0.20391212403774261,
      "eval_runtime": 22.147,
      "eval_samples_per_second": 42.579,
      "eval_steps_per_second": 21.312,
      "num_input_tokens_seen": 11960752,
      "step": 13800
    },
    {
      "epoch": 6.508722300801509,
      "grad_norm": 2.6344149112701416,
      "learning_rate": 3.6691303700315796e-05,
      "loss": 0.2373,
      "num_input_tokens_seen": 11965200,
      "step": 13805
    },
    {
      "epoch": 6.511079679396511,
      "grad_norm": 0.8778369426727295,
      "learning_rate": 3.668262500749655e-05,
      "loss": 0.1629,
      "num_input_tokens_seen": 11969120,
      "step": 13810
    },
    {
      "epoch": 6.5134370579915135,
      "grad_norm": 1.1217942237854004,
      "learning_rate": 3.667394451306971e-05,
      "loss": 0.1859,
      "num_input_tokens_seen": 11972944,
      "step": 13815
    },
    {
      "epoch": 6.515794436586516,
      "grad_norm": 0.6023867130279541,
      "learning_rate": 3.666526221837393e-05,
      "loss": 0.1825,
      "num_input_tokens_seen": 11976752,
      "step": 13820
    },
    {
      "epoch": 6.518151815181518,
      "grad_norm": 1.2859195470809937,
      "learning_rate": 3.665657812474812e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 11980896,
      "step": 13825
    },
    {
      "epoch": 6.520509193776521,
      "grad_norm": 0.26786375045776367,
      "learning_rate": 3.664789223353147e-05,
      "loss": 0.165,
      "num_input_tokens_seen": 11984960,
      "step": 13830
    },
    {
      "epoch": 6.522866572371523,
      "grad_norm": 0.6413196325302124,
      "learning_rate": 3.663920454606347e-05,
      "loss": 0.1813,
      "num_input_tokens_seen": 11989552,
      "step": 13835
    },
    {
      "epoch": 6.5252239509665255,
      "grad_norm": 0.9566588997840881,
      "learning_rate": 3.6630515063683856e-05,
      "loss": 0.1193,
      "num_input_tokens_seen": 11993584,
      "step": 13840
    },
    {
      "epoch": 6.527581329561528,
      "grad_norm": 1.6803317070007324,
      "learning_rate": 3.662182378773267e-05,
      "loss": 0.1497,
      "num_input_tokens_seen": 11997936,
      "step": 13845
    },
    {
      "epoch": 6.52993870815653,
      "grad_norm": 0.5535024404525757,
      "learning_rate": 3.66131307195502e-05,
      "loss": 0.1741,
      "num_input_tokens_seen": 12002848,
      "step": 13850
    },
    {
      "epoch": 6.532296086751533,
      "grad_norm": 0.7558871507644653,
      "learning_rate": 3.6604435860477034e-05,
      "loss": 0.2018,
      "num_input_tokens_seen": 12006912,
      "step": 13855
    },
    {
      "epoch": 6.534653465346535,
      "grad_norm": 0.5141012668609619,
      "learning_rate": 3.6595739211854025e-05,
      "loss": 0.1792,
      "num_input_tokens_seen": 12010976,
      "step": 13860
    },
    {
      "epoch": 6.537010843941537,
      "grad_norm": 0.4673299789428711,
      "learning_rate": 3.658704077502231e-05,
      "loss": 0.2778,
      "num_input_tokens_seen": 12016208,
      "step": 13865
    },
    {
      "epoch": 6.539368222536539,
      "grad_norm": 0.6631970405578613,
      "learning_rate": 3.65783405513233e-05,
      "loss": 0.1358,
      "num_input_tokens_seen": 12019856,
      "step": 13870
    },
    {
      "epoch": 6.5417256011315414,
      "grad_norm": 0.47394251823425293,
      "learning_rate": 3.656963854209867e-05,
      "loss": 0.172,
      "num_input_tokens_seen": 12024480,
      "step": 13875
    },
    {
      "epoch": 6.544082979726544,
      "grad_norm": 1.7876418828964233,
      "learning_rate": 3.656093474869038e-05,
      "loss": 0.1939,
      "num_input_tokens_seen": 12029072,
      "step": 13880
    },
    {
      "epoch": 6.546440358321546,
      "grad_norm": 0.8933086395263672,
      "learning_rate": 3.655222917244068e-05,
      "loss": 0.1769,
      "num_input_tokens_seen": 12032688,
      "step": 13885
    },
    {
      "epoch": 6.548797736916549,
      "grad_norm": 1.3085920810699463,
      "learning_rate": 3.6543521814692054e-05,
      "loss": 0.2157,
      "num_input_tokens_seen": 12036752,
      "step": 13890
    },
    {
      "epoch": 6.551155115511551,
      "grad_norm": 1.191525936126709,
      "learning_rate": 3.653481267678731e-05,
      "loss": 0.1798,
      "num_input_tokens_seen": 12040816,
      "step": 13895
    },
    {
      "epoch": 6.5535124941065535,
      "grad_norm": 1.0043561458587646,
      "learning_rate": 3.652610176006949e-05,
      "loss": 0.1323,
      "num_input_tokens_seen": 12046176,
      "step": 13900
    },
    {
      "epoch": 6.555869872701556,
      "grad_norm": 0.6487537622451782,
      "learning_rate": 3.6517389065881925e-05,
      "loss": 0.1742,
      "num_input_tokens_seen": 12050112,
      "step": 13905
    },
    {
      "epoch": 6.558227251296558,
      "grad_norm": 0.7467843294143677,
      "learning_rate": 3.650867459556824e-05,
      "loss": 0.222,
      "num_input_tokens_seen": 12054656,
      "step": 13910
    },
    {
      "epoch": 6.560584629891561,
      "grad_norm": 0.19831989705562592,
      "learning_rate": 3.64999583504723e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 12058640,
      "step": 13915
    },
    {
      "epoch": 6.562942008486563,
      "grad_norm": 1.0078222751617432,
      "learning_rate": 3.649124033193827e-05,
      "loss": 0.1548,
      "num_input_tokens_seen": 12063888,
      "step": 13920
    },
    {
      "epoch": 6.5652993870815655,
      "grad_norm": 0.37721362709999084,
      "learning_rate": 3.648252054131057e-05,
      "loss": 0.1832,
      "num_input_tokens_seen": 12068064,
      "step": 13925
    },
    {
      "epoch": 6.567656765676568,
      "grad_norm": 2.6221535205841064,
      "learning_rate": 3.647379897993391e-05,
      "loss": 0.2088,
      "num_input_tokens_seen": 12072704,
      "step": 13930
    },
    {
      "epoch": 6.57001414427157,
      "grad_norm": 0.954239010810852,
      "learning_rate": 3.646507564915325e-05,
      "loss": 0.1728,
      "num_input_tokens_seen": 12076976,
      "step": 13935
    },
    {
      "epoch": 6.572371522866573,
      "grad_norm": 0.8282114267349243,
      "learning_rate": 3.645635055031385e-05,
      "loss": 0.1682,
      "num_input_tokens_seen": 12080960,
      "step": 13940
    },
    {
      "epoch": 6.574728901461575,
      "grad_norm": 0.5394594669342041,
      "learning_rate": 3.6447623684761224e-05,
      "loss": 0.171,
      "num_input_tokens_seen": 12085664,
      "step": 13945
    },
    {
      "epoch": 6.5770862800565775,
      "grad_norm": 1.5003211498260498,
      "learning_rate": 3.643889505384117e-05,
      "loss": 0.1747,
      "num_input_tokens_seen": 12090448,
      "step": 13950
    },
    {
      "epoch": 6.579443658651579,
      "grad_norm": 1.4328263998031616,
      "learning_rate": 3.6430164658899744e-05,
      "loss": 0.1618,
      "num_input_tokens_seen": 12094384,
      "step": 13955
    },
    {
      "epoch": 6.581801037246581,
      "grad_norm": 0.9571413397789001,
      "learning_rate": 3.642143250128329e-05,
      "loss": 0.1943,
      "num_input_tokens_seen": 12098720,
      "step": 13960
    },
    {
      "epoch": 6.584158415841584,
      "grad_norm": 0.5148213505744934,
      "learning_rate": 3.641269858233841e-05,
      "loss": 0.231,
      "num_input_tokens_seen": 12103072,
      "step": 13965
    },
    {
      "epoch": 6.586515794436586,
      "grad_norm": 0.48324814438819885,
      "learning_rate": 3.640396290341199e-05,
      "loss": 0.1679,
      "num_input_tokens_seen": 12107840,
      "step": 13970
    },
    {
      "epoch": 6.588873173031589,
      "grad_norm": 0.5230081677436829,
      "learning_rate": 3.639522546585118e-05,
      "loss": 0.1971,
      "num_input_tokens_seen": 12111984,
      "step": 13975
    },
    {
      "epoch": 6.591230551626591,
      "grad_norm": 1.556401014328003,
      "learning_rate": 3.6386486271003404e-05,
      "loss": 0.1787,
      "num_input_tokens_seen": 12116800,
      "step": 13980
    },
    {
      "epoch": 6.593587930221593,
      "grad_norm": 0.9345512390136719,
      "learning_rate": 3.6377745320216346e-05,
      "loss": 0.1831,
      "num_input_tokens_seen": 12120752,
      "step": 13985
    },
    {
      "epoch": 6.595945308816596,
      "grad_norm": 1.5843247175216675,
      "learning_rate": 3.636900261483798e-05,
      "loss": 0.1774,
      "num_input_tokens_seen": 12124848,
      "step": 13990
    },
    {
      "epoch": 6.598302687411598,
      "grad_norm": 1.178699016571045,
      "learning_rate": 3.636025815621654e-05,
      "loss": 0.1904,
      "num_input_tokens_seen": 12128784,
      "step": 13995
    },
    {
      "epoch": 6.600660066006601,
      "grad_norm": 1.1197649240493774,
      "learning_rate": 3.635151194570054e-05,
      "loss": 0.1811,
      "num_input_tokens_seen": 12132672,
      "step": 14000
    },
    {
      "epoch": 6.600660066006601,
      "eval_loss": 0.20753885805606842,
      "eval_runtime": 22.1785,
      "eval_samples_per_second": 42.519,
      "eval_steps_per_second": 21.282,
      "num_input_tokens_seen": 12132672,
      "step": 14000
    },
    {
      "epoch": 6.603017444601603,
      "grad_norm": 1.0952731370925903,
      "learning_rate": 3.634276398463873e-05,
      "loss": 0.1662,
      "num_input_tokens_seen": 12136768,
      "step": 14005
    },
    {
      "epoch": 6.605374823196605,
      "grad_norm": 0.6539545059204102,
      "learning_rate": 3.633401427438018e-05,
      "loss": 0.1554,
      "num_input_tokens_seen": 12141456,
      "step": 14010
    },
    {
      "epoch": 6.607732201791608,
      "grad_norm": 1.2409496307373047,
      "learning_rate": 3.63252628162742e-05,
      "loss": 0.167,
      "num_input_tokens_seen": 12145920,
      "step": 14015
    },
    {
      "epoch": 6.61008958038661,
      "grad_norm": 0.6678215861320496,
      "learning_rate": 3.6316509611670364e-05,
      "loss": 0.1674,
      "num_input_tokens_seen": 12149760,
      "step": 14020
    },
    {
      "epoch": 6.612446958981613,
      "grad_norm": 1.5046857595443726,
      "learning_rate": 3.630775466191854e-05,
      "loss": 0.1384,
      "num_input_tokens_seen": 12153088,
      "step": 14025
    },
    {
      "epoch": 6.614804337576615,
      "grad_norm": 0.9192318320274353,
      "learning_rate": 3.629899796836884e-05,
      "loss": 0.2098,
      "num_input_tokens_seen": 12158048,
      "step": 14030
    },
    {
      "epoch": 6.617161716171617,
      "grad_norm": 0.5388753414154053,
      "learning_rate": 3.6290239532371666e-05,
      "loss": 0.212,
      "num_input_tokens_seen": 12162864,
      "step": 14035
    },
    {
      "epoch": 6.61951909476662,
      "grad_norm": 0.4994087219238281,
      "learning_rate": 3.628147935527767e-05,
      "loss": 0.1617,
      "num_input_tokens_seen": 12168352,
      "step": 14040
    },
    {
      "epoch": 6.621876473361622,
      "grad_norm": 2.167617082595825,
      "learning_rate": 3.627271743843779e-05,
      "loss": 0.2082,
      "num_input_tokens_seen": 12172768,
      "step": 14045
    },
    {
      "epoch": 6.624233851956625,
      "grad_norm": 0.8573362231254578,
      "learning_rate": 3.626395378320321e-05,
      "loss": 0.1706,
      "num_input_tokens_seen": 12177296,
      "step": 14050
    },
    {
      "epoch": 6.626591230551627,
      "grad_norm": 1.5808374881744385,
      "learning_rate": 3.625518839092541e-05,
      "loss": 0.2505,
      "num_input_tokens_seen": 12180576,
      "step": 14055
    },
    {
      "epoch": 6.628948609146629,
      "grad_norm": 1.5934388637542725,
      "learning_rate": 3.624642126295612e-05,
      "loss": 0.1917,
      "num_input_tokens_seen": 12185184,
      "step": 14060
    },
    {
      "epoch": 6.631305987741631,
      "grad_norm": 0.7000859379768372,
      "learning_rate": 3.6237652400647345e-05,
      "loss": 0.131,
      "num_input_tokens_seen": 12189376,
      "step": 14065
    },
    {
      "epoch": 6.633663366336633,
      "grad_norm": 1.0096662044525146,
      "learning_rate": 3.622888180535134e-05,
      "loss": 0.1551,
      "num_input_tokens_seen": 12192992,
      "step": 14070
    },
    {
      "epoch": 6.636020744931636,
      "grad_norm": 0.8714025020599365,
      "learning_rate": 3.6220109478420655e-05,
      "loss": 0.1663,
      "num_input_tokens_seen": 12197520,
      "step": 14075
    },
    {
      "epoch": 6.638378123526638,
      "grad_norm": 1.64369797706604,
      "learning_rate": 3.6211335421208084e-05,
      "loss": 0.1823,
      "num_input_tokens_seen": 12201728,
      "step": 14080
    },
    {
      "epoch": 6.6407355021216405,
      "grad_norm": 1.2205634117126465,
      "learning_rate": 3.62025596350667e-05,
      "loss": 0.1998,
      "num_input_tokens_seen": 12205648,
      "step": 14085
    },
    {
      "epoch": 6.643092880716643,
      "grad_norm": 1.4289071559906006,
      "learning_rate": 3.619378212134984e-05,
      "loss": 0.1234,
      "num_input_tokens_seen": 12209312,
      "step": 14090
    },
    {
      "epoch": 6.645450259311645,
      "grad_norm": 0.46564385294914246,
      "learning_rate": 3.618500288141111e-05,
      "loss": 0.1494,
      "num_input_tokens_seen": 12213728,
      "step": 14095
    },
    {
      "epoch": 6.647807637906648,
      "grad_norm": 1.563502311706543,
      "learning_rate": 3.617622191660438e-05,
      "loss": 0.1949,
      "num_input_tokens_seen": 12218208,
      "step": 14100
    },
    {
      "epoch": 6.65016501650165,
      "grad_norm": 0.8682853579521179,
      "learning_rate": 3.616743922828377e-05,
      "loss": 0.1925,
      "num_input_tokens_seen": 12222608,
      "step": 14105
    },
    {
      "epoch": 6.6525223950966526,
      "grad_norm": 1.414576530456543,
      "learning_rate": 3.615865481780371e-05,
      "loss": 0.1971,
      "num_input_tokens_seen": 12226224,
      "step": 14110
    },
    {
      "epoch": 6.654879773691655,
      "grad_norm": 1.566459059715271,
      "learning_rate": 3.614986868651883e-05,
      "loss": 0.1892,
      "num_input_tokens_seen": 12230192,
      "step": 14115
    },
    {
      "epoch": 6.657237152286657,
      "grad_norm": 3.156360626220703,
      "learning_rate": 3.614108083578409e-05,
      "loss": 0.2288,
      "num_input_tokens_seen": 12234768,
      "step": 14120
    },
    {
      "epoch": 6.65959453088166,
      "grad_norm": 1.2730662822723389,
      "learning_rate": 3.613229126695467e-05,
      "loss": 0.1854,
      "num_input_tokens_seen": 12238960,
      "step": 14125
    },
    {
      "epoch": 6.661951909476662,
      "grad_norm": 1.4974058866500854,
      "learning_rate": 3.612349998138605e-05,
      "loss": 0.2279,
      "num_input_tokens_seen": 12243616,
      "step": 14130
    },
    {
      "epoch": 6.664309288071665,
      "grad_norm": 0.6764591932296753,
      "learning_rate": 3.6114706980433946e-05,
      "loss": 0.1401,
      "num_input_tokens_seen": 12248496,
      "step": 14135
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 3.0150880813598633,
      "learning_rate": 3.610591226545435e-05,
      "loss": 0.1946,
      "num_input_tokens_seen": 12253184,
      "step": 14140
    },
    {
      "epoch": 6.669024045261669,
      "grad_norm": 0.9191609025001526,
      "learning_rate": 3.6097115837803505e-05,
      "loss": 0.1846,
      "num_input_tokens_seen": 12256848,
      "step": 14145
    },
    {
      "epoch": 6.671381423856672,
      "grad_norm": 0.9494617581367493,
      "learning_rate": 3.608831769883795e-05,
      "loss": 0.1965,
      "num_input_tokens_seen": 12261184,
      "step": 14150
    },
    {
      "epoch": 6.673738802451673,
      "grad_norm": 0.6074720621109009,
      "learning_rate": 3.607951784991446e-05,
      "loss": 0.1514,
      "num_input_tokens_seen": 12265504,
      "step": 14155
    },
    {
      "epoch": 6.676096181046676,
      "grad_norm": 1.8235810995101929,
      "learning_rate": 3.6070716292390085e-05,
      "loss": 0.1752,
      "num_input_tokens_seen": 12269248,
      "step": 14160
    },
    {
      "epoch": 6.678453559641678,
      "grad_norm": 0.6765730381011963,
      "learning_rate": 3.606191302762213e-05,
      "loss": 0.1425,
      "num_input_tokens_seen": 12273072,
      "step": 14165
    },
    {
      "epoch": 6.6808109382366805,
      "grad_norm": 0.5704247951507568,
      "learning_rate": 3.605310805696818e-05,
      "loss": 0.2551,
      "num_input_tokens_seen": 12278112,
      "step": 14170
    },
    {
      "epoch": 6.683168316831683,
      "grad_norm": 1.1213911771774292,
      "learning_rate": 3.6044301381786067e-05,
      "loss": 0.2118,
      "num_input_tokens_seen": 12282320,
      "step": 14175
    },
    {
      "epoch": 6.685525695426685,
      "grad_norm": 0.464571475982666,
      "learning_rate": 3.6035493003433883e-05,
      "loss": 0.1751,
      "num_input_tokens_seen": 12286320,
      "step": 14180
    },
    {
      "epoch": 6.687883074021688,
      "grad_norm": 0.4546494781970978,
      "learning_rate": 3.6026682923269994e-05,
      "loss": 0.1154,
      "num_input_tokens_seen": 12290448,
      "step": 14185
    },
    {
      "epoch": 6.69024045261669,
      "grad_norm": 1.2951455116271973,
      "learning_rate": 3.6017871142653034e-05,
      "loss": 0.1688,
      "num_input_tokens_seen": 12294864,
      "step": 14190
    },
    {
      "epoch": 6.6925978312116925,
      "grad_norm": 0.7291257381439209,
      "learning_rate": 3.600905766294189e-05,
      "loss": 0.2297,
      "num_input_tokens_seen": 12299360,
      "step": 14195
    },
    {
      "epoch": 6.694955209806695,
      "grad_norm": 2.5983707904815674,
      "learning_rate": 3.60002424854957e-05,
      "loss": 0.2338,
      "num_input_tokens_seen": 12303424,
      "step": 14200
    },
    {
      "epoch": 6.694955209806695,
      "eval_loss": 0.2037079632282257,
      "eval_runtime": 22.1377,
      "eval_samples_per_second": 42.597,
      "eval_steps_per_second": 21.321,
      "num_input_tokens_seen": 12303424,
      "step": 14200
    },
    {
      "epoch": 6.697312588401697,
      "grad_norm": 1.067902684211731,
      "learning_rate": 3.5991425611673876e-05,
      "loss": 0.1437,
      "num_input_tokens_seen": 12307952,
      "step": 14205
    },
    {
      "epoch": 6.6996699669967,
      "grad_norm": 0.9547677040100098,
      "learning_rate": 3.5982607042836105e-05,
      "loss": 0.1724,
      "num_input_tokens_seen": 12311552,
      "step": 14210
    },
    {
      "epoch": 6.702027345591702,
      "grad_norm": 0.6990019679069519,
      "learning_rate": 3.597378678034231e-05,
      "loss": 0.1567,
      "num_input_tokens_seen": 12314848,
      "step": 14215
    },
    {
      "epoch": 6.7043847241867045,
      "grad_norm": 1.2359848022460938,
      "learning_rate": 3.596496482555269e-05,
      "loss": 0.2051,
      "num_input_tokens_seen": 12318800,
      "step": 14220
    },
    {
      "epoch": 6.706742102781707,
      "grad_norm": 1.1764010190963745,
      "learning_rate": 3.595614117982769e-05,
      "loss": 0.2072,
      "num_input_tokens_seen": 12323520,
      "step": 14225
    },
    {
      "epoch": 6.709099481376709,
      "grad_norm": 0.7293670177459717,
      "learning_rate": 3.594731584452805e-05,
      "loss": 0.2004,
      "num_input_tokens_seen": 12328048,
      "step": 14230
    },
    {
      "epoch": 6.711456859971712,
      "grad_norm": 1.9942954778671265,
      "learning_rate": 3.593848882101472e-05,
      "loss": 0.3375,
      "num_input_tokens_seen": 12332224,
      "step": 14235
    },
    {
      "epoch": 6.713814238566714,
      "grad_norm": 0.7402311563491821,
      "learning_rate": 3.592966011064896e-05,
      "loss": 0.1391,
      "num_input_tokens_seen": 12336368,
      "step": 14240
    },
    {
      "epoch": 6.7161716171617165,
      "grad_norm": 0.5809162855148315,
      "learning_rate": 3.592082971479226e-05,
      "loss": 0.1302,
      "num_input_tokens_seen": 12340544,
      "step": 14245
    },
    {
      "epoch": 6.718528995756719,
      "grad_norm": 0.4190351068973541,
      "learning_rate": 3.5911997634806385e-05,
      "loss": 0.1721,
      "num_input_tokens_seen": 12345072,
      "step": 14250
    },
    {
      "epoch": 6.720886374351721,
      "grad_norm": 1.0432298183441162,
      "learning_rate": 3.5903163872053336e-05,
      "loss": 0.1633,
      "num_input_tokens_seen": 12351008,
      "step": 14255
    },
    {
      "epoch": 6.723243752946724,
      "grad_norm": 0.7073342800140381,
      "learning_rate": 3.58943284278954e-05,
      "loss": 0.1967,
      "num_input_tokens_seen": 12355584,
      "step": 14260
    },
    {
      "epoch": 6.725601131541725,
      "grad_norm": 2.053182363510132,
      "learning_rate": 3.588549130369512e-05,
      "loss": 0.2807,
      "num_input_tokens_seen": 12359680,
      "step": 14265
    },
    {
      "epoch": 6.727958510136728,
      "grad_norm": 0.8254354596138,
      "learning_rate": 3.5876652500815274e-05,
      "loss": 0.2145,
      "num_input_tokens_seen": 12363184,
      "step": 14270
    },
    {
      "epoch": 6.73031588873173,
      "grad_norm": 0.43667107820510864,
      "learning_rate": 3.586781202061894e-05,
      "loss": 0.1654,
      "num_input_tokens_seen": 12367200,
      "step": 14275
    },
    {
      "epoch": 6.732673267326732,
      "grad_norm": 1.3503621816635132,
      "learning_rate": 3.585896986446942e-05,
      "loss": 0.1755,
      "num_input_tokens_seen": 12371808,
      "step": 14280
    },
    {
      "epoch": 6.735030645921735,
      "grad_norm": 0.6678754687309265,
      "learning_rate": 3.585012603373028e-05,
      "loss": 0.2829,
      "num_input_tokens_seen": 12375872,
      "step": 14285
    },
    {
      "epoch": 6.737388024516737,
      "grad_norm": 1.5690444707870483,
      "learning_rate": 3.584128052976535e-05,
      "loss": 0.1691,
      "num_input_tokens_seen": 12379664,
      "step": 14290
    },
    {
      "epoch": 6.73974540311174,
      "grad_norm": 1.6236017942428589,
      "learning_rate": 3.5832433353938724e-05,
      "loss": 0.2937,
      "num_input_tokens_seen": 12383792,
      "step": 14295
    },
    {
      "epoch": 6.742102781706742,
      "grad_norm": 0.7859368920326233,
      "learning_rate": 3.5823584507614746e-05,
      "loss": 0.1864,
      "num_input_tokens_seen": 12389120,
      "step": 14300
    },
    {
      "epoch": 6.7444601603017444,
      "grad_norm": 0.436367005109787,
      "learning_rate": 3.581473399215802e-05,
      "loss": 0.2371,
      "num_input_tokens_seen": 12392992,
      "step": 14305
    },
    {
      "epoch": 6.746817538896747,
      "grad_norm": 0.6052256226539612,
      "learning_rate": 3.580588180893341e-05,
      "loss": 0.201,
      "num_input_tokens_seen": 12397616,
      "step": 14310
    },
    {
      "epoch": 6.749174917491749,
      "grad_norm": 0.6346387267112732,
      "learning_rate": 3.579702795930602e-05,
      "loss": 0.1759,
      "num_input_tokens_seen": 12402096,
      "step": 14315
    },
    {
      "epoch": 6.751532296086752,
      "grad_norm": 0.3424694836139679,
      "learning_rate": 3.578817244464125e-05,
      "loss": 0.2054,
      "num_input_tokens_seen": 12406528,
      "step": 14320
    },
    {
      "epoch": 6.753889674681754,
      "grad_norm": 1.4649310111999512,
      "learning_rate": 3.577931526630471e-05,
      "loss": 0.1409,
      "num_input_tokens_seen": 12411648,
      "step": 14325
    },
    {
      "epoch": 6.7562470532767565,
      "grad_norm": 1.2261227369308472,
      "learning_rate": 3.577045642566229e-05,
      "loss": 0.1995,
      "num_input_tokens_seen": 12416800,
      "step": 14330
    },
    {
      "epoch": 6.758604431871759,
      "grad_norm": 1.309782862663269,
      "learning_rate": 3.576159592408014e-05,
      "loss": 0.2737,
      "num_input_tokens_seen": 12420608,
      "step": 14335
    },
    {
      "epoch": 6.760961810466761,
      "grad_norm": 0.6781545877456665,
      "learning_rate": 3.575273376292466e-05,
      "loss": 0.1509,
      "num_input_tokens_seen": 12424832,
      "step": 14340
    },
    {
      "epoch": 6.763319189061764,
      "grad_norm": 0.4902195632457733,
      "learning_rate": 3.574386994356251e-05,
      "loss": 0.1466,
      "num_input_tokens_seen": 12429024,
      "step": 14345
    },
    {
      "epoch": 6.765676567656766,
      "grad_norm": 1.5174181461334229,
      "learning_rate": 3.573500446736059e-05,
      "loss": 0.1963,
      "num_input_tokens_seen": 12433040,
      "step": 14350
    },
    {
      "epoch": 6.768033946251768,
      "grad_norm": 0.4511356055736542,
      "learning_rate": 3.5726137335686094e-05,
      "loss": 0.1841,
      "num_input_tokens_seen": 12436544,
      "step": 14355
    },
    {
      "epoch": 6.77039132484677,
      "grad_norm": 0.4817649722099304,
      "learning_rate": 3.571726854990642e-05,
      "loss": 0.1951,
      "num_input_tokens_seen": 12440912,
      "step": 14360
    },
    {
      "epoch": 6.772748703441772,
      "grad_norm": 1.290723443031311,
      "learning_rate": 3.570839811138925e-05,
      "loss": 0.2142,
      "num_input_tokens_seen": 12444672,
      "step": 14365
    },
    {
      "epoch": 6.775106082036775,
      "grad_norm": 1.124808430671692,
      "learning_rate": 3.569952602150252e-05,
      "loss": 0.1889,
      "num_input_tokens_seen": 12449344,
      "step": 14370
    },
    {
      "epoch": 6.777463460631777,
      "grad_norm": 1.0810075998306274,
      "learning_rate": 3.569065228161442e-05,
      "loss": 0.2375,
      "num_input_tokens_seen": 12453120,
      "step": 14375
    },
    {
      "epoch": 6.77982083922678,
      "grad_norm": 0.5339473485946655,
      "learning_rate": 3.5681776893093395e-05,
      "loss": 0.1455,
      "num_input_tokens_seen": 12457120,
      "step": 14380
    },
    {
      "epoch": 6.782178217821782,
      "grad_norm": 0.5484223365783691,
      "learning_rate": 3.5672899857308134e-05,
      "loss": 0.197,
      "num_input_tokens_seen": 12461392,
      "step": 14385
    },
    {
      "epoch": 6.784535596416784,
      "grad_norm": 0.6395797729492188,
      "learning_rate": 3.566402117562759e-05,
      "loss": 0.2057,
      "num_input_tokens_seen": 12465408,
      "step": 14390
    },
    {
      "epoch": 6.786892975011787,
      "grad_norm": 0.8397663235664368,
      "learning_rate": 3.565514084942097e-05,
      "loss": 0.211,
      "num_input_tokens_seen": 12470256,
      "step": 14395
    },
    {
      "epoch": 6.789250353606789,
      "grad_norm": 0.8576312065124512,
      "learning_rate": 3.564625888005773e-05,
      "loss": 0.2086,
      "num_input_tokens_seen": 12474592,
      "step": 14400
    },
    {
      "epoch": 6.789250353606789,
      "eval_loss": 0.20526346564292908,
      "eval_runtime": 22.2035,
      "eval_samples_per_second": 42.471,
      "eval_steps_per_second": 21.258,
      "num_input_tokens_seen": 12474592,
      "step": 14400
    },
    {
      "epoch": 6.791607732201792,
      "grad_norm": 0.7197822332382202,
      "learning_rate": 3.563737526890759e-05,
      "loss": 0.2263,
      "num_input_tokens_seen": 12479136,
      "step": 14405
    },
    {
      "epoch": 6.793965110796794,
      "grad_norm": 1.5676854848861694,
      "learning_rate": 3.562849001734049e-05,
      "loss": 0.2208,
      "num_input_tokens_seen": 12484368,
      "step": 14410
    },
    {
      "epoch": 6.796322489391796,
      "grad_norm": 0.8949878215789795,
      "learning_rate": 3.561960312672667e-05,
      "loss": 0.1958,
      "num_input_tokens_seen": 12488832,
      "step": 14415
    },
    {
      "epoch": 6.798679867986799,
      "grad_norm": 0.8885391354560852,
      "learning_rate": 3.5610714598436596e-05,
      "loss": 0.1067,
      "num_input_tokens_seen": 12493440,
      "step": 14420
    },
    {
      "epoch": 6.801037246581801,
      "grad_norm": 0.9532304406166077,
      "learning_rate": 3.5601824433840986e-05,
      "loss": 0.2155,
      "num_input_tokens_seen": 12497264,
      "step": 14425
    },
    {
      "epoch": 6.803394625176804,
      "grad_norm": 0.730601966381073,
      "learning_rate": 3.559293263431082e-05,
      "loss": 0.1815,
      "num_input_tokens_seen": 12501872,
      "step": 14430
    },
    {
      "epoch": 6.805752003771806,
      "grad_norm": 0.5729410648345947,
      "learning_rate": 3.558403920121732e-05,
      "loss": 0.2558,
      "num_input_tokens_seen": 12506080,
      "step": 14435
    },
    {
      "epoch": 6.808109382366808,
      "grad_norm": 1.9035309553146362,
      "learning_rate": 3.557514413593197e-05,
      "loss": 0.2328,
      "num_input_tokens_seen": 12509536,
      "step": 14440
    },
    {
      "epoch": 6.810466760961811,
      "grad_norm": 1.422888994216919,
      "learning_rate": 3.55662474398265e-05,
      "loss": 0.1925,
      "num_input_tokens_seen": 12514624,
      "step": 14445
    },
    {
      "epoch": 6.812824139556813,
      "grad_norm": 1.1334470510482788,
      "learning_rate": 3.555734911427288e-05,
      "loss": 0.2006,
      "num_input_tokens_seen": 12518320,
      "step": 14450
    },
    {
      "epoch": 6.815181518151816,
      "grad_norm": 0.6903905272483826,
      "learning_rate": 3.5548449160643363e-05,
      "loss": 0.1961,
      "num_input_tokens_seen": 12522800,
      "step": 14455
    },
    {
      "epoch": 6.817538896746818,
      "grad_norm": 0.6675986051559448,
      "learning_rate": 3.553954758031043e-05,
      "loss": 0.1946,
      "num_input_tokens_seen": 12526992,
      "step": 14460
    },
    {
      "epoch": 6.8198962753418195,
      "grad_norm": 1.1350698471069336,
      "learning_rate": 3.5530644374646815e-05,
      "loss": 0.181,
      "num_input_tokens_seen": 12530784,
      "step": 14465
    },
    {
      "epoch": 6.822253653936822,
      "grad_norm": 2.0630385875701904,
      "learning_rate": 3.552173954502549e-05,
      "loss": 0.2281,
      "num_input_tokens_seen": 12535248,
      "step": 14470
    },
    {
      "epoch": 6.824611032531824,
      "grad_norm": 1.3260234594345093,
      "learning_rate": 3.55128330928197e-05,
      "loss": 0.1858,
      "num_input_tokens_seen": 12540160,
      "step": 14475
    },
    {
      "epoch": 6.826968411126827,
      "grad_norm": 1.2496392726898193,
      "learning_rate": 3.550392501940294e-05,
      "loss": 0.1653,
      "num_input_tokens_seen": 12544608,
      "step": 14480
    },
    {
      "epoch": 6.829325789721829,
      "grad_norm": 0.7285873889923096,
      "learning_rate": 3.5495015326148945e-05,
      "loss": 0.1691,
      "num_input_tokens_seen": 12549920,
      "step": 14485
    },
    {
      "epoch": 6.8316831683168315,
      "grad_norm": 0.3948063850402832,
      "learning_rate": 3.548610401443169e-05,
      "loss": 0.1627,
      "num_input_tokens_seen": 12554320,
      "step": 14490
    },
    {
      "epoch": 6.834040546911834,
      "grad_norm": 1.4279521703720093,
      "learning_rate": 3.547719108562543e-05,
      "loss": 0.2036,
      "num_input_tokens_seen": 12558304,
      "step": 14495
    },
    {
      "epoch": 6.836397925506836,
      "grad_norm": 0.8297402262687683,
      "learning_rate": 3.546827654110464e-05,
      "loss": 0.1496,
      "num_input_tokens_seen": 12562336,
      "step": 14500
    },
    {
      "epoch": 6.838755304101839,
      "grad_norm": 1.199236273765564,
      "learning_rate": 3.545936038224405e-05,
      "loss": 0.2086,
      "num_input_tokens_seen": 12565728,
      "step": 14505
    },
    {
      "epoch": 6.841112682696841,
      "grad_norm": 0.6499026417732239,
      "learning_rate": 3.545044261041864e-05,
      "loss": 0.2029,
      "num_input_tokens_seen": 12569440,
      "step": 14510
    },
    {
      "epoch": 6.8434700612918435,
      "grad_norm": 0.9033247828483582,
      "learning_rate": 3.5441523227003657e-05,
      "loss": 0.1907,
      "num_input_tokens_seen": 12573920,
      "step": 14515
    },
    {
      "epoch": 6.845827439886846,
      "grad_norm": 2.61633563041687,
      "learning_rate": 3.543260223337459e-05,
      "loss": 0.2781,
      "num_input_tokens_seen": 12578464,
      "step": 14520
    },
    {
      "epoch": 6.848184818481848,
      "grad_norm": 0.4551036059856415,
      "learning_rate": 3.542367963090714e-05,
      "loss": 0.1971,
      "num_input_tokens_seen": 12582736,
      "step": 14525
    },
    {
      "epoch": 6.850542197076851,
      "grad_norm": 1.483716368675232,
      "learning_rate": 3.5414755420977295e-05,
      "loss": 0.2986,
      "num_input_tokens_seen": 12587152,
      "step": 14530
    },
    {
      "epoch": 6.852899575671853,
      "grad_norm": 1.4282945394515991,
      "learning_rate": 3.54058296049613e-05,
      "loss": 0.1723,
      "num_input_tokens_seen": 12591760,
      "step": 14535
    },
    {
      "epoch": 6.8552569542668556,
      "grad_norm": 1.6433141231536865,
      "learning_rate": 3.53969021842356e-05,
      "loss": 0.2224,
      "num_input_tokens_seen": 12596448,
      "step": 14540
    },
    {
      "epoch": 6.857614332861858,
      "grad_norm": 0.5458544492721558,
      "learning_rate": 3.5387973160176926e-05,
      "loss": 0.1618,
      "num_input_tokens_seen": 12600496,
      "step": 14545
    },
    {
      "epoch": 6.85997171145686,
      "grad_norm": 0.5013637542724609,
      "learning_rate": 3.537904253416224e-05,
      "loss": 0.1371,
      "num_input_tokens_seen": 12604448,
      "step": 14550
    },
    {
      "epoch": 6.862329090051862,
      "grad_norm": 0.5971200466156006,
      "learning_rate": 3.537011030756878e-05,
      "loss": 0.2134,
      "num_input_tokens_seen": 12608496,
      "step": 14555
    },
    {
      "epoch": 6.864686468646864,
      "grad_norm": 0.5378842353820801,
      "learning_rate": 3.536117648177399e-05,
      "loss": 0.1984,
      "num_input_tokens_seen": 12612960,
      "step": 14560
    },
    {
      "epoch": 6.867043847241867,
      "grad_norm": 1.3130701780319214,
      "learning_rate": 3.535224105815558e-05,
      "loss": 0.1746,
      "num_input_tokens_seen": 12617456,
      "step": 14565
    },
    {
      "epoch": 6.869401225836869,
      "grad_norm": 1.6939384937286377,
      "learning_rate": 3.5343304038091494e-05,
      "loss": 0.158,
      "num_input_tokens_seen": 12621584,
      "step": 14570
    },
    {
      "epoch": 6.8717586044318715,
      "grad_norm": 0.6672232151031494,
      "learning_rate": 3.5334365422959955e-05,
      "loss": 0.2273,
      "num_input_tokens_seen": 12626752,
      "step": 14575
    },
    {
      "epoch": 6.874115983026874,
      "grad_norm": 0.8653534054756165,
      "learning_rate": 3.5325425214139396e-05,
      "loss": 0.1441,
      "num_input_tokens_seen": 12631136,
      "step": 14580
    },
    {
      "epoch": 6.876473361621876,
      "grad_norm": 0.39460816979408264,
      "learning_rate": 3.531648341300851e-05,
      "loss": 0.1699,
      "num_input_tokens_seen": 12636224,
      "step": 14585
    },
    {
      "epoch": 6.878830740216879,
      "grad_norm": 1.680280327796936,
      "learning_rate": 3.530754002094623e-05,
      "loss": 0.2217,
      "num_input_tokens_seen": 12640592,
      "step": 14590
    },
    {
      "epoch": 6.881188118811881,
      "grad_norm": 0.5356877446174622,
      "learning_rate": 3.529859503933175e-05,
      "loss": 0.1738,
      "num_input_tokens_seen": 12644912,
      "step": 14595
    },
    {
      "epoch": 6.8835454974068835,
      "grad_norm": 0.4889218807220459,
      "learning_rate": 3.52896484695445e-05,
      "loss": 0.2102,
      "num_input_tokens_seen": 12649424,
      "step": 14600
    },
    {
      "epoch": 6.8835454974068835,
      "eval_loss": 0.20267586410045624,
      "eval_runtime": 22.1711,
      "eval_samples_per_second": 42.533,
      "eval_steps_per_second": 21.289,
      "num_input_tokens_seen": 12649424,
      "step": 14600
    },
    {
      "epoch": 6.885902876001886,
      "grad_norm": 1.3869043588638306,
      "learning_rate": 3.528070031296414e-05,
      "loss": 0.1939,
      "num_input_tokens_seen": 12653056,
      "step": 14605
    },
    {
      "epoch": 6.888260254596888,
      "grad_norm": 0.5435015559196472,
      "learning_rate": 3.5271750570970605e-05,
      "loss": 0.1805,
      "num_input_tokens_seen": 12657536,
      "step": 14610
    },
    {
      "epoch": 6.890617633191891,
      "grad_norm": 1.1883960962295532,
      "learning_rate": 3.526279924494405e-05,
      "loss": 0.2287,
      "num_input_tokens_seen": 12661760,
      "step": 14615
    },
    {
      "epoch": 6.892975011786893,
      "grad_norm": 0.3548820912837982,
      "learning_rate": 3.5253846336264874e-05,
      "loss": 0.1684,
      "num_input_tokens_seen": 12665888,
      "step": 14620
    },
    {
      "epoch": 6.8953323903818955,
      "grad_norm": 1.0969996452331543,
      "learning_rate": 3.5244891846313736e-05,
      "loss": 0.1762,
      "num_input_tokens_seen": 12669776,
      "step": 14625
    },
    {
      "epoch": 6.897689768976898,
      "grad_norm": 0.9218679070472717,
      "learning_rate": 3.5235935776471527e-05,
      "loss": 0.2621,
      "num_input_tokens_seen": 12674576,
      "step": 14630
    },
    {
      "epoch": 6.9000471475719,
      "grad_norm": 0.5176541209220886,
      "learning_rate": 3.522697812811939e-05,
      "loss": 0.1635,
      "num_input_tokens_seen": 12678592,
      "step": 14635
    },
    {
      "epoch": 6.902404526166903,
      "grad_norm": 0.9050940871238708,
      "learning_rate": 3.521801890263871e-05,
      "loss": 0.2467,
      "num_input_tokens_seen": 12683296,
      "step": 14640
    },
    {
      "epoch": 6.904761904761905,
      "grad_norm": 1.4949489831924438,
      "learning_rate": 3.5209058101411114e-05,
      "loss": 0.217,
      "num_input_tokens_seen": 12687728,
      "step": 14645
    },
    {
      "epoch": 6.9071192833569075,
      "grad_norm": 1.386551022529602,
      "learning_rate": 3.520009572581845e-05,
      "loss": 0.2273,
      "num_input_tokens_seen": 12691616,
      "step": 14650
    },
    {
      "epoch": 6.90947666195191,
      "grad_norm": 0.9942753314971924,
      "learning_rate": 3.519113177724285e-05,
      "loss": 0.1442,
      "num_input_tokens_seen": 12695456,
      "step": 14655
    },
    {
      "epoch": 6.911834040546912,
      "grad_norm": 0.7322081923484802,
      "learning_rate": 3.5182166257066656e-05,
      "loss": 0.2353,
      "num_input_tokens_seen": 12699312,
      "step": 14660
    },
    {
      "epoch": 6.914191419141914,
      "grad_norm": 0.7228081822395325,
      "learning_rate": 3.517319916667247e-05,
      "loss": 0.1692,
      "num_input_tokens_seen": 12704096,
      "step": 14665
    },
    {
      "epoch": 6.916548797736916,
      "grad_norm": 0.8452159762382507,
      "learning_rate": 3.516423050744313e-05,
      "loss": 0.1942,
      "num_input_tokens_seen": 12708160,
      "step": 14670
    },
    {
      "epoch": 6.918906176331919,
      "grad_norm": 1.3653274774551392,
      "learning_rate": 3.5155260280761704e-05,
      "loss": 0.1473,
      "num_input_tokens_seen": 12712096,
      "step": 14675
    },
    {
      "epoch": 6.921263554926921,
      "grad_norm": 0.9017449021339417,
      "learning_rate": 3.514628848801154e-05,
      "loss": 0.2333,
      "num_input_tokens_seen": 12716304,
      "step": 14680
    },
    {
      "epoch": 6.923620933521923,
      "grad_norm": 2.6838266849517822,
      "learning_rate": 3.5137315130576174e-05,
      "loss": 0.2218,
      "num_input_tokens_seen": 12719920,
      "step": 14685
    },
    {
      "epoch": 6.925978312116926,
      "grad_norm": 0.9014167189598083,
      "learning_rate": 3.512834020983942e-05,
      "loss": 0.1973,
      "num_input_tokens_seen": 12724480,
      "step": 14690
    },
    {
      "epoch": 6.928335690711928,
      "grad_norm": 0.7879084944725037,
      "learning_rate": 3.5119363727185334e-05,
      "loss": 0.1753,
      "num_input_tokens_seen": 12728992,
      "step": 14695
    },
    {
      "epoch": 6.930693069306931,
      "grad_norm": 0.8226804137229919,
      "learning_rate": 3.511038568399819e-05,
      "loss": 0.1664,
      "num_input_tokens_seen": 12733904,
      "step": 14700
    },
    {
      "epoch": 6.933050447901933,
      "grad_norm": 0.4649672210216522,
      "learning_rate": 3.510140608166251e-05,
      "loss": 0.2413,
      "num_input_tokens_seen": 12738368,
      "step": 14705
    },
    {
      "epoch": 6.935407826496935,
      "grad_norm": 0.4917888641357422,
      "learning_rate": 3.509242492156308e-05,
      "loss": 0.201,
      "num_input_tokens_seen": 12742016,
      "step": 14710
    },
    {
      "epoch": 6.937765205091938,
      "grad_norm": 0.5657393932342529,
      "learning_rate": 3.5083442205084896e-05,
      "loss": 0.1323,
      "num_input_tokens_seen": 12746576,
      "step": 14715
    },
    {
      "epoch": 6.94012258368694,
      "grad_norm": 1.3689441680908203,
      "learning_rate": 3.507445793361321e-05,
      "loss": 0.2228,
      "num_input_tokens_seen": 12750432,
      "step": 14720
    },
    {
      "epoch": 6.942479962281943,
      "grad_norm": 0.7084893584251404,
      "learning_rate": 3.5065472108533505e-05,
      "loss": 0.1915,
      "num_input_tokens_seen": 12755312,
      "step": 14725
    },
    {
      "epoch": 6.944837340876945,
      "grad_norm": 0.4967528283596039,
      "learning_rate": 3.5056484731231504e-05,
      "loss": 0.1479,
      "num_input_tokens_seen": 12759312,
      "step": 14730
    },
    {
      "epoch": 6.947194719471947,
      "grad_norm": 0.5524225831031799,
      "learning_rate": 3.504749580309319e-05,
      "loss": 0.2088,
      "num_input_tokens_seen": 12763600,
      "step": 14735
    },
    {
      "epoch": 6.94955209806695,
      "grad_norm": 0.4411320090293884,
      "learning_rate": 3.5038505325504753e-05,
      "loss": 0.1429,
      "num_input_tokens_seen": 12768080,
      "step": 14740
    },
    {
      "epoch": 6.951909476661952,
      "grad_norm": 0.8600887060165405,
      "learning_rate": 3.502951329985264e-05,
      "loss": 0.1917,
      "num_input_tokens_seen": 12772352,
      "step": 14745
    },
    {
      "epoch": 6.954266855256955,
      "grad_norm": 0.5197879076004028,
      "learning_rate": 3.502051972752354e-05,
      "loss": 0.1355,
      "num_input_tokens_seen": 12776560,
      "step": 14750
    },
    {
      "epoch": 6.956624233851956,
      "grad_norm": 0.904850959777832,
      "learning_rate": 3.5011524609904374e-05,
      "loss": 0.1758,
      "num_input_tokens_seen": 12780880,
      "step": 14755
    },
    {
      "epoch": 6.958981612446959,
      "grad_norm": 1.4766634702682495,
      "learning_rate": 3.50025279483823e-05,
      "loss": 0.2132,
      "num_input_tokens_seen": 12785216,
      "step": 14760
    },
    {
      "epoch": 6.961338991041961,
      "grad_norm": 1.3364323377609253,
      "learning_rate": 3.499352974434472e-05,
      "loss": 0.2131,
      "num_input_tokens_seen": 12789984,
      "step": 14765
    },
    {
      "epoch": 6.963696369636963,
      "grad_norm": 0.6343903541564941,
      "learning_rate": 3.498452999917926e-05,
      "loss": 0.1424,
      "num_input_tokens_seen": 12794688,
      "step": 14770
    },
    {
      "epoch": 6.966053748231966,
      "grad_norm": 1.904178500175476,
      "learning_rate": 3.4975528714273795e-05,
      "loss": 0.1453,
      "num_input_tokens_seen": 12799440,
      "step": 14775
    },
    {
      "epoch": 6.968411126826968,
      "grad_norm": 1.0576673746109009,
      "learning_rate": 3.4966525891016454e-05,
      "loss": 0.2054,
      "num_input_tokens_seen": 12803856,
      "step": 14780
    },
    {
      "epoch": 6.970768505421971,
      "grad_norm": 1.4442417621612549,
      "learning_rate": 3.495752153079557e-05,
      "loss": 0.201,
      "num_input_tokens_seen": 12808256,
      "step": 14785
    },
    {
      "epoch": 6.973125884016973,
      "grad_norm": 0.7628756761550903,
      "learning_rate": 3.494851563499974e-05,
      "loss": 0.1204,
      "num_input_tokens_seen": 12812272,
      "step": 14790
    },
    {
      "epoch": 6.975483262611975,
      "grad_norm": 1.3268511295318604,
      "learning_rate": 3.493950820501777e-05,
      "loss": 0.1305,
      "num_input_tokens_seen": 12816640,
      "step": 14795
    },
    {
      "epoch": 6.977840641206978,
      "grad_norm": 0.3275807499885559,
      "learning_rate": 3.493049924223872e-05,
      "loss": 0.0934,
      "num_input_tokens_seen": 12821280,
      "step": 14800
    },
    {
      "epoch": 6.977840641206978,
      "eval_loss": 0.2060907930135727,
      "eval_runtime": 22.1282,
      "eval_samples_per_second": 42.615,
      "eval_steps_per_second": 21.33,
      "num_input_tokens_seen": 12821280,
      "step": 14800
    },
    {
      "epoch": 6.98019801980198,
      "grad_norm": 0.8086602091789246,
      "learning_rate": 3.49214887480519e-05,
      "loss": 0.2445,
      "num_input_tokens_seen": 12826720,
      "step": 14805
    },
    {
      "epoch": 6.982555398396983,
      "grad_norm": 1.7544797658920288,
      "learning_rate": 3.4912476723846834e-05,
      "loss": 0.2039,
      "num_input_tokens_seen": 12831536,
      "step": 14810
    },
    {
      "epoch": 6.984912776991985,
      "grad_norm": 0.5402692556381226,
      "learning_rate": 3.490346317101328e-05,
      "loss": 0.2258,
      "num_input_tokens_seen": 12835536,
      "step": 14815
    },
    {
      "epoch": 6.987270155586987,
      "grad_norm": 0.7749181985855103,
      "learning_rate": 3.4894448090941266e-05,
      "loss": 0.2117,
      "num_input_tokens_seen": 12839744,
      "step": 14820
    },
    {
      "epoch": 6.98962753418199,
      "grad_norm": 1.0471882820129395,
      "learning_rate": 3.488543148502101e-05,
      "loss": 0.19,
      "num_input_tokens_seen": 12843856,
      "step": 14825
    },
    {
      "epoch": 6.991984912776992,
      "grad_norm": 1.3245198726654053,
      "learning_rate": 3.487641335464299e-05,
      "loss": 0.1619,
      "num_input_tokens_seen": 12848464,
      "step": 14830
    },
    {
      "epoch": 6.994342291371995,
      "grad_norm": 0.6384978294372559,
      "learning_rate": 3.4867393701197914e-05,
      "loss": 0.1773,
      "num_input_tokens_seen": 12852928,
      "step": 14835
    },
    {
      "epoch": 6.996699669966997,
      "grad_norm": 0.8024163842201233,
      "learning_rate": 3.485837252607673e-05,
      "loss": 0.1359,
      "num_input_tokens_seen": 12858096,
      "step": 14840
    },
    {
      "epoch": 6.999057048561999,
      "grad_norm": 0.78171306848526,
      "learning_rate": 3.4849349830670615e-05,
      "loss": 0.1491,
      "num_input_tokens_seen": 12861712,
      "step": 14845
    },
    {
      "epoch": 7.001414427157002,
      "grad_norm": 1.0428471565246582,
      "learning_rate": 3.4840325616370976e-05,
      "loss": 0.203,
      "num_input_tokens_seen": 12865632,
      "step": 14850
    },
    {
      "epoch": 7.003771805752004,
      "grad_norm": 0.9340112805366516,
      "learning_rate": 3.483129988456947e-05,
      "loss": 0.221,
      "num_input_tokens_seen": 12870192,
      "step": 14855
    },
    {
      "epoch": 7.006129184347006,
      "grad_norm": 0.5586615800857544,
      "learning_rate": 3.482227263665797e-05,
      "loss": 0.1949,
      "num_input_tokens_seen": 12875152,
      "step": 14860
    },
    {
      "epoch": 7.008486562942008,
      "grad_norm": 0.7354224920272827,
      "learning_rate": 3.48132438740286e-05,
      "loss": 0.1247,
      "num_input_tokens_seen": 12879376,
      "step": 14865
    },
    {
      "epoch": 7.0108439415370105,
      "grad_norm": 2.178635358810425,
      "learning_rate": 3.48042135980737e-05,
      "loss": 0.2257,
      "num_input_tokens_seen": 12884176,
      "step": 14870
    },
    {
      "epoch": 7.013201320132013,
      "grad_norm": 0.6996216773986816,
      "learning_rate": 3.479518181018586e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 12889568,
      "step": 14875
    },
    {
      "epoch": 7.015558698727015,
      "grad_norm": 0.623114287853241,
      "learning_rate": 3.4786148511757886e-05,
      "loss": 0.1511,
      "num_input_tokens_seen": 12893808,
      "step": 14880
    },
    {
      "epoch": 7.017916077322018,
      "grad_norm": 1.1125904321670532,
      "learning_rate": 3.477711370418284e-05,
      "loss": 0.2797,
      "num_input_tokens_seen": 12899184,
      "step": 14885
    },
    {
      "epoch": 7.02027345591702,
      "grad_norm": 0.8747548460960388,
      "learning_rate": 3.476807738885399e-05,
      "loss": 0.2584,
      "num_input_tokens_seen": 12903568,
      "step": 14890
    },
    {
      "epoch": 7.0226308345120225,
      "grad_norm": 1.0131959915161133,
      "learning_rate": 3.475903956716485e-05,
      "loss": 0.1225,
      "num_input_tokens_seen": 12907696,
      "step": 14895
    },
    {
      "epoch": 7.024988213107025,
      "grad_norm": 0.9778690934181213,
      "learning_rate": 3.475000024050917e-05,
      "loss": 0.2626,
      "num_input_tokens_seen": 12912304,
      "step": 14900
    },
    {
      "epoch": 7.027345591702027,
      "grad_norm": 0.7191489934921265,
      "learning_rate": 3.4740959410280926e-05,
      "loss": 0.1512,
      "num_input_tokens_seen": 12916256,
      "step": 14905
    },
    {
      "epoch": 7.02970297029703,
      "grad_norm": 1.6246275901794434,
      "learning_rate": 3.4731917077874324e-05,
      "loss": 0.3056,
      "num_input_tokens_seen": 12920992,
      "step": 14910
    },
    {
      "epoch": 7.032060348892032,
      "grad_norm": 0.5354019403457642,
      "learning_rate": 3.4722873244683816e-05,
      "loss": 0.2403,
      "num_input_tokens_seen": 12925024,
      "step": 14915
    },
    {
      "epoch": 7.0344177274870345,
      "grad_norm": 0.41826310753822327,
      "learning_rate": 3.4713827912104065e-05,
      "loss": 0.1831,
      "num_input_tokens_seen": 12929776,
      "step": 14920
    },
    {
      "epoch": 7.036775106082037,
      "grad_norm": 1.7080188989639282,
      "learning_rate": 3.470478108152998e-05,
      "loss": 0.1974,
      "num_input_tokens_seen": 12934096,
      "step": 14925
    },
    {
      "epoch": 7.039132484677039,
      "grad_norm": 0.4443165063858032,
      "learning_rate": 3.4695732754356695e-05,
      "loss": 0.1581,
      "num_input_tokens_seen": 12938080,
      "step": 14930
    },
    {
      "epoch": 7.041489863272042,
      "grad_norm": 1.6539548635482788,
      "learning_rate": 3.4686682931979576e-05,
      "loss": 0.2061,
      "num_input_tokens_seen": 12941616,
      "step": 14935
    },
    {
      "epoch": 7.043847241867044,
      "grad_norm": 0.8173926472663879,
      "learning_rate": 3.467763161579422e-05,
      "loss": 0.1361,
      "num_input_tokens_seen": 12945440,
      "step": 14940
    },
    {
      "epoch": 7.0462046204620465,
      "grad_norm": 1.2719093561172485,
      "learning_rate": 3.466857880719645e-05,
      "loss": 0.1356,
      "num_input_tokens_seen": 12949104,
      "step": 14945
    },
    {
      "epoch": 7.048561999057049,
      "grad_norm": 0.8004023432731628,
      "learning_rate": 3.465952450758233e-05,
      "loss": 0.1386,
      "num_input_tokens_seen": 12953056,
      "step": 14950
    },
    {
      "epoch": 7.050919377652051,
      "grad_norm": 0.4645894467830658,
      "learning_rate": 3.4650468718348126e-05,
      "loss": 0.1836,
      "num_input_tokens_seen": 12957632,
      "step": 14955
    },
    {
      "epoch": 7.053276756247053,
      "grad_norm": 0.3638777732849121,
      "learning_rate": 3.464141144089038e-05,
      "loss": 0.1325,
      "num_input_tokens_seen": 12961408,
      "step": 14960
    },
    {
      "epoch": 7.055634134842055,
      "grad_norm": 0.621051549911499,
      "learning_rate": 3.463235267660583e-05,
      "loss": 0.2081,
      "num_input_tokens_seen": 12966352,
      "step": 14965
    },
    {
      "epoch": 7.057991513437058,
      "grad_norm": 0.5529163479804993,
      "learning_rate": 3.462329242689145e-05,
      "loss": 0.1499,
      "num_input_tokens_seen": 12970688,
      "step": 14970
    },
    {
      "epoch": 7.06034889203206,
      "grad_norm": 0.6085306406021118,
      "learning_rate": 3.461423069314444e-05,
      "loss": 0.1775,
      "num_input_tokens_seen": 12974736,
      "step": 14975
    },
    {
      "epoch": 7.0627062706270625,
      "grad_norm": 0.9774405360221863,
      "learning_rate": 3.460516747676224e-05,
      "loss": 0.16,
      "num_input_tokens_seen": 12979712,
      "step": 14980
    },
    {
      "epoch": 7.065063649222065,
      "grad_norm": 0.8706043362617493,
      "learning_rate": 3.459610277914251e-05,
      "loss": 0.1551,
      "num_input_tokens_seen": 12983488,
      "step": 14985
    },
    {
      "epoch": 7.067421027817067,
      "grad_norm": 0.6470679640769958,
      "learning_rate": 3.458703660168314e-05,
      "loss": 0.1406,
      "num_input_tokens_seen": 12987936,
      "step": 14990
    },
    {
      "epoch": 7.06977840641207,
      "grad_norm": 1.2241034507751465,
      "learning_rate": 3.457796894578224e-05,
      "loss": 0.2417,
      "num_input_tokens_seen": 12992608,
      "step": 14995
    },
    {
      "epoch": 7.072135785007072,
      "grad_norm": 1.6305922269821167,
      "learning_rate": 3.456889981283817e-05,
      "loss": 0.2323,
      "num_input_tokens_seen": 12996208,
      "step": 15000
    },
    {
      "epoch": 7.072135785007072,
      "eval_loss": 0.20808608829975128,
      "eval_runtime": 22.157,
      "eval_samples_per_second": 42.56,
      "eval_steps_per_second": 21.302,
      "num_input_tokens_seen": 12996208,
      "step": 15000
    },
    {
      "epoch": 7.0744931636020745,
      "grad_norm": 0.5418935418128967,
      "learning_rate": 3.45598292042495e-05,
      "loss": 0.1999,
      "num_input_tokens_seen": 13000144,
      "step": 15005
    },
    {
      "epoch": 7.076850542197077,
      "grad_norm": 1.6133954524993896,
      "learning_rate": 3.4550757121415035e-05,
      "loss": 0.1773,
      "num_input_tokens_seen": 13004544,
      "step": 15010
    },
    {
      "epoch": 7.079207920792079,
      "grad_norm": 1.6315722465515137,
      "learning_rate": 3.454168356573378e-05,
      "loss": 0.1772,
      "num_input_tokens_seen": 13009456,
      "step": 15015
    },
    {
      "epoch": 7.081565299387082,
      "grad_norm": 0.447020024061203,
      "learning_rate": 3.453260853860503e-05,
      "loss": 0.2322,
      "num_input_tokens_seen": 13014432,
      "step": 15020
    },
    {
      "epoch": 7.083922677982084,
      "grad_norm": 1.2041221857070923,
      "learning_rate": 3.452353204142824e-05,
      "loss": 0.204,
      "num_input_tokens_seen": 13018912,
      "step": 15025
    },
    {
      "epoch": 7.0862800565770865,
      "grad_norm": 0.5735824704170227,
      "learning_rate": 3.4514454075603136e-05,
      "loss": 0.1733,
      "num_input_tokens_seen": 13024064,
      "step": 15030
    },
    {
      "epoch": 7.088637435172089,
      "grad_norm": 1.1533604860305786,
      "learning_rate": 3.450537464252964e-05,
      "loss": 0.3092,
      "num_input_tokens_seen": 13028016,
      "step": 15035
    },
    {
      "epoch": 7.090994813767091,
      "grad_norm": 0.5463526248931885,
      "learning_rate": 3.4496293743607925e-05,
      "loss": 0.2508,
      "num_input_tokens_seen": 13032672,
      "step": 15040
    },
    {
      "epoch": 7.093352192362094,
      "grad_norm": 0.8442049622535706,
      "learning_rate": 3.448721138023838e-05,
      "loss": 0.1995,
      "num_input_tokens_seen": 13036544,
      "step": 15045
    },
    {
      "epoch": 7.095709570957096,
      "grad_norm": 0.715582013130188,
      "learning_rate": 3.447812755382162e-05,
      "loss": 0.1909,
      "num_input_tokens_seen": 13040464,
      "step": 15050
    },
    {
      "epoch": 7.0980669495520985,
      "grad_norm": 0.6718910932540894,
      "learning_rate": 3.446904226575847e-05,
      "loss": 0.1773,
      "num_input_tokens_seen": 13044368,
      "step": 15055
    },
    {
      "epoch": 7.1004243281471,
      "grad_norm": 0.581889271736145,
      "learning_rate": 3.445995551745002e-05,
      "loss": 0.1601,
      "num_input_tokens_seen": 13048928,
      "step": 15060
    },
    {
      "epoch": 7.102781706742102,
      "grad_norm": 0.640162467956543,
      "learning_rate": 3.445086731029753e-05,
      "loss": 0.157,
      "num_input_tokens_seen": 13053024,
      "step": 15065
    },
    {
      "epoch": 7.105139085337105,
      "grad_norm": 0.8841479420661926,
      "learning_rate": 3.444177764570255e-05,
      "loss": 0.173,
      "num_input_tokens_seen": 13057888,
      "step": 15070
    },
    {
      "epoch": 7.107496463932107,
      "grad_norm": 1.2936081886291504,
      "learning_rate": 3.44326865250668e-05,
      "loss": 0.2678,
      "num_input_tokens_seen": 13061984,
      "step": 15075
    },
    {
      "epoch": 7.10985384252711,
      "grad_norm": 1.2247928380966187,
      "learning_rate": 3.442359394979225e-05,
      "loss": 0.1921,
      "num_input_tokens_seen": 13066496,
      "step": 15080
    },
    {
      "epoch": 7.112211221122112,
      "grad_norm": 1.2045371532440186,
      "learning_rate": 3.441449992128108e-05,
      "loss": 0.1785,
      "num_input_tokens_seen": 13070288,
      "step": 15085
    },
    {
      "epoch": 7.114568599717114,
      "grad_norm": 1.4346930980682373,
      "learning_rate": 3.440540444093573e-05,
      "loss": 0.1986,
      "num_input_tokens_seen": 13074800,
      "step": 15090
    },
    {
      "epoch": 7.116925978312117,
      "grad_norm": 1.2396531105041504,
      "learning_rate": 3.43963075101588e-05,
      "loss": 0.1603,
      "num_input_tokens_seen": 13079808,
      "step": 15095
    },
    {
      "epoch": 7.119283356907119,
      "grad_norm": 0.4523015022277832,
      "learning_rate": 3.438720913035318e-05,
      "loss": 0.1505,
      "num_input_tokens_seen": 13083824,
      "step": 15100
    },
    {
      "epoch": 7.121640735502122,
      "grad_norm": 0.5963098406791687,
      "learning_rate": 3.437810930292195e-05,
      "loss": 0.1735,
      "num_input_tokens_seen": 13088368,
      "step": 15105
    },
    {
      "epoch": 7.123998114097124,
      "grad_norm": 1.012174129486084,
      "learning_rate": 3.43690080292684e-05,
      "loss": 0.1585,
      "num_input_tokens_seen": 13092624,
      "step": 15110
    },
    {
      "epoch": 7.126355492692126,
      "grad_norm": 1.8839683532714844,
      "learning_rate": 3.435990531079608e-05,
      "loss": 0.2075,
      "num_input_tokens_seen": 13096400,
      "step": 15115
    },
    {
      "epoch": 7.128712871287129,
      "grad_norm": 1.1313998699188232,
      "learning_rate": 3.435080114890874e-05,
      "loss": 0.2294,
      "num_input_tokens_seen": 13100448,
      "step": 15120
    },
    {
      "epoch": 7.131070249882131,
      "grad_norm": 0.9155084490776062,
      "learning_rate": 3.434169554501035e-05,
      "loss": 0.1924,
      "num_input_tokens_seen": 13105232,
      "step": 15125
    },
    {
      "epoch": 7.133427628477134,
      "grad_norm": 0.986707329750061,
      "learning_rate": 3.433258850050511e-05,
      "loss": 0.2215,
      "num_input_tokens_seen": 13110208,
      "step": 15130
    },
    {
      "epoch": 7.135785007072136,
      "grad_norm": 1.251062035560608,
      "learning_rate": 3.4323480016797446e-05,
      "loss": 0.1621,
      "num_input_tokens_seen": 13114800,
      "step": 15135
    },
    {
      "epoch": 7.138142385667138,
      "grad_norm": 1.0603400468826294,
      "learning_rate": 3.4314370095291995e-05,
      "loss": 0.2218,
      "num_input_tokens_seen": 13119792,
      "step": 15140
    },
    {
      "epoch": 7.140499764262141,
      "grad_norm": 0.6834056377410889,
      "learning_rate": 3.430525873739363e-05,
      "loss": 0.1855,
      "num_input_tokens_seen": 13123840,
      "step": 15145
    },
    {
      "epoch": 7.142857142857143,
      "grad_norm": 1.3499430418014526,
      "learning_rate": 3.429614594450743e-05,
      "loss": 0.2629,
      "num_input_tokens_seen": 13127760,
      "step": 15150
    },
    {
      "epoch": 7.145214521452146,
      "grad_norm": 1.3934139013290405,
      "learning_rate": 3.428703171803869e-05,
      "loss": 0.1726,
      "num_input_tokens_seen": 13133360,
      "step": 15155
    },
    {
      "epoch": 7.147571900047147,
      "grad_norm": 1.1963884830474854,
      "learning_rate": 3.4277916059392964e-05,
      "loss": 0.1489,
      "num_input_tokens_seen": 13137392,
      "step": 15160
    },
    {
      "epoch": 7.1499292786421496,
      "grad_norm": 0.802753746509552,
      "learning_rate": 3.426879896997598e-05,
      "loss": 0.2063,
      "num_input_tokens_seen": 13142160,
      "step": 15165
    },
    {
      "epoch": 7.152286657237152,
      "grad_norm": 2.135918617248535,
      "learning_rate": 3.425968045119372e-05,
      "loss": 0.1856,
      "num_input_tokens_seen": 13148080,
      "step": 15170
    },
    {
      "epoch": 7.154644035832154,
      "grad_norm": 0.513687014579773,
      "learning_rate": 3.425056050445237e-05,
      "loss": 0.1548,
      "num_input_tokens_seen": 13151632,
      "step": 15175
    },
    {
      "epoch": 7.157001414427157,
      "grad_norm": 0.877056896686554,
      "learning_rate": 3.4241439131158336e-05,
      "loss": 0.1997,
      "num_input_tokens_seen": 13155424,
      "step": 15180
    },
    {
      "epoch": 7.159358793022159,
      "grad_norm": 1.2392926216125488,
      "learning_rate": 3.423231633271825e-05,
      "loss": 0.1601,
      "num_input_tokens_seen": 13160064,
      "step": 15185
    },
    {
      "epoch": 7.161716171617162,
      "grad_norm": 0.5202606916427612,
      "learning_rate": 3.4223192110538985e-05,
      "loss": 0.2242,
      "num_input_tokens_seen": 13164576,
      "step": 15190
    },
    {
      "epoch": 7.164073550212164,
      "grad_norm": 2.4318172931671143,
      "learning_rate": 3.4214066466027575e-05,
      "loss": 0.2131,
      "num_input_tokens_seen": 13168512,
      "step": 15195
    },
    {
      "epoch": 7.166430928807166,
      "grad_norm": 1.5042245388031006,
      "learning_rate": 3.4204939400591325e-05,
      "loss": 0.2743,
      "num_input_tokens_seen": 13172592,
      "step": 15200
    },
    {
      "epoch": 7.166430928807166,
      "eval_loss": 0.20626972615718842,
      "eval_runtime": 22.1708,
      "eval_samples_per_second": 42.533,
      "eval_steps_per_second": 21.289,
      "num_input_tokens_seen": 13172592,
      "step": 15200
    },
    {
      "epoch": 7.168788307402169,
      "grad_norm": 1.3327484130859375,
      "learning_rate": 3.419581091563775e-05,
      "loss": 0.1849,
      "num_input_tokens_seen": 13177312,
      "step": 15205
    },
    {
      "epoch": 7.171145685997171,
      "grad_norm": 1.4504177570343018,
      "learning_rate": 3.418668101257456e-05,
      "loss": 0.1639,
      "num_input_tokens_seen": 13181936,
      "step": 15210
    },
    {
      "epoch": 7.173503064592174,
      "grad_norm": 1.1016507148742676,
      "learning_rate": 3.417754969280971e-05,
      "loss": 0.2205,
      "num_input_tokens_seen": 13186176,
      "step": 15215
    },
    {
      "epoch": 7.175860443187176,
      "grad_norm": 0.9413275122642517,
      "learning_rate": 3.416841695775137e-05,
      "loss": 0.1816,
      "num_input_tokens_seen": 13189872,
      "step": 15220
    },
    {
      "epoch": 7.178217821782178,
      "grad_norm": 1.0592235326766968,
      "learning_rate": 3.415928280880792e-05,
      "loss": 0.2074,
      "num_input_tokens_seen": 13194272,
      "step": 15225
    },
    {
      "epoch": 7.180575200377181,
      "grad_norm": 1.4372166395187378,
      "learning_rate": 3.4150147247387965e-05,
      "loss": 0.2051,
      "num_input_tokens_seen": 13198128,
      "step": 15230
    },
    {
      "epoch": 7.182932578972183,
      "grad_norm": 0.8472356200218201,
      "learning_rate": 3.4141010274900306e-05,
      "loss": 0.189,
      "num_input_tokens_seen": 13203216,
      "step": 15235
    },
    {
      "epoch": 7.185289957567186,
      "grad_norm": 1.3500980138778687,
      "learning_rate": 3.413187189275399e-05,
      "loss": 0.2156,
      "num_input_tokens_seen": 13207744,
      "step": 15240
    },
    {
      "epoch": 7.187647336162188,
      "grad_norm": 0.8462612628936768,
      "learning_rate": 3.4122732102358265e-05,
      "loss": 0.176,
      "num_input_tokens_seen": 13211904,
      "step": 15245
    },
    {
      "epoch": 7.19000471475719,
      "grad_norm": 1.2599486112594604,
      "learning_rate": 3.411359090512261e-05,
      "loss": 0.1676,
      "num_input_tokens_seen": 13216208,
      "step": 15250
    },
    {
      "epoch": 7.192362093352193,
      "grad_norm": 1.1149035692214966,
      "learning_rate": 3.410444830245672e-05,
      "loss": 0.2011,
      "num_input_tokens_seen": 13220704,
      "step": 15255
    },
    {
      "epoch": 7.194719471947194,
      "grad_norm": 0.8515975475311279,
      "learning_rate": 3.409530429577048e-05,
      "loss": 0.1984,
      "num_input_tokens_seen": 13225136,
      "step": 15260
    },
    {
      "epoch": 7.197076850542197,
      "grad_norm": 1.062036156654358,
      "learning_rate": 3.408615888647402e-05,
      "loss": 0.2177,
      "num_input_tokens_seen": 13229264,
      "step": 15265
    },
    {
      "epoch": 7.199434229137199,
      "grad_norm": 1.0353920459747314,
      "learning_rate": 3.4077012075977675e-05,
      "loss": 0.1979,
      "num_input_tokens_seen": 13233040,
      "step": 15270
    },
    {
      "epoch": 7.2017916077322015,
      "grad_norm": 0.734756350517273,
      "learning_rate": 3.4067863865692e-05,
      "loss": 0.1806,
      "num_input_tokens_seen": 13237344,
      "step": 15275
    },
    {
      "epoch": 7.204148986327204,
      "grad_norm": 2.0769705772399902,
      "learning_rate": 3.4058714257027755e-05,
      "loss": 0.1586,
      "num_input_tokens_seen": 13242640,
      "step": 15280
    },
    {
      "epoch": 7.206506364922206,
      "grad_norm": 0.7757366895675659,
      "learning_rate": 3.404956325139594e-05,
      "loss": 0.2114,
      "num_input_tokens_seen": 13247008,
      "step": 15285
    },
    {
      "epoch": 7.208863743517209,
      "grad_norm": 1.2447998523712158,
      "learning_rate": 3.404041085020775e-05,
      "loss": 0.1701,
      "num_input_tokens_seen": 13251072,
      "step": 15290
    },
    {
      "epoch": 7.211221122112211,
      "grad_norm": 0.5394159555435181,
      "learning_rate": 3.403125705487459e-05,
      "loss": 0.1407,
      "num_input_tokens_seen": 13254960,
      "step": 15295
    },
    {
      "epoch": 7.2135785007072135,
      "grad_norm": 1.1936851739883423,
      "learning_rate": 3.402210186680811e-05,
      "loss": 0.2052,
      "num_input_tokens_seen": 13258992,
      "step": 15300
    },
    {
      "epoch": 7.215935879302216,
      "grad_norm": 0.6116772890090942,
      "learning_rate": 3.4012945287420137e-05,
      "loss": 0.1915,
      "num_input_tokens_seen": 13262880,
      "step": 15305
    },
    {
      "epoch": 7.218293257897218,
      "grad_norm": 0.5661678314208984,
      "learning_rate": 3.400378731812274e-05,
      "loss": 0.1845,
      "num_input_tokens_seen": 13267456,
      "step": 15310
    },
    {
      "epoch": 7.220650636492221,
      "grad_norm": 0.6627749800682068,
      "learning_rate": 3.399462796032817e-05,
      "loss": 0.1659,
      "num_input_tokens_seen": 13271776,
      "step": 15315
    },
    {
      "epoch": 7.223008015087223,
      "grad_norm": 0.7132918238639832,
      "learning_rate": 3.3985467215448954e-05,
      "loss": 0.1412,
      "num_input_tokens_seen": 13275856,
      "step": 15320
    },
    {
      "epoch": 7.2253653936822255,
      "grad_norm": 0.6387122869491577,
      "learning_rate": 3.3976305084897776e-05,
      "loss": 0.175,
      "num_input_tokens_seen": 13279872,
      "step": 15325
    },
    {
      "epoch": 7.227722772277228,
      "grad_norm": 0.7792193293571472,
      "learning_rate": 3.3967141570087544e-05,
      "loss": 0.1332,
      "num_input_tokens_seen": 13283968,
      "step": 15330
    },
    {
      "epoch": 7.23008015087223,
      "grad_norm": 1.6284537315368652,
      "learning_rate": 3.39579766724314e-05,
      "loss": 0.2004,
      "num_input_tokens_seen": 13288160,
      "step": 15335
    },
    {
      "epoch": 7.232437529467233,
      "grad_norm": 1.0074188709259033,
      "learning_rate": 3.3948810393342677e-05,
      "loss": 0.2029,
      "num_input_tokens_seen": 13292944,
      "step": 15340
    },
    {
      "epoch": 7.234794908062235,
      "grad_norm": 0.4260003864765167,
      "learning_rate": 3.3939642734234936e-05,
      "loss": 0.1234,
      "num_input_tokens_seen": 13296864,
      "step": 15345
    },
    {
      "epoch": 7.2371522866572375,
      "grad_norm": 0.6001573204994202,
      "learning_rate": 3.393047369652194e-05,
      "loss": 0.1844,
      "num_input_tokens_seen": 13301744,
      "step": 15350
    },
    {
      "epoch": 7.23950966525224,
      "grad_norm": 1.5188860893249512,
      "learning_rate": 3.3921303281617664e-05,
      "loss": 0.1419,
      "num_input_tokens_seen": 13305488,
      "step": 15355
    },
    {
      "epoch": 7.2418670438472414,
      "grad_norm": 0.6660593152046204,
      "learning_rate": 3.391213149093632e-05,
      "loss": 0.113,
      "num_input_tokens_seen": 13310352,
      "step": 15360
    },
    {
      "epoch": 7.244224422442244,
      "grad_norm": 0.5181620717048645,
      "learning_rate": 3.3902958325892303e-05,
      "loss": 0.1621,
      "num_input_tokens_seen": 13314096,
      "step": 15365
    },
    {
      "epoch": 7.246581801037246,
      "grad_norm": 1.2385796308517456,
      "learning_rate": 3.389378378790023e-05,
      "loss": 0.1689,
      "num_input_tokens_seen": 13318064,
      "step": 15370
    },
    {
      "epoch": 7.248939179632249,
      "grad_norm": 1.0534343719482422,
      "learning_rate": 3.388460787837493e-05,
      "loss": 0.1862,
      "num_input_tokens_seen": 13322304,
      "step": 15375
    },
    {
      "epoch": 7.251296558227251,
      "grad_norm": 0.8035048246383667,
      "learning_rate": 3.387543059873145e-05,
      "loss": 0.2604,
      "num_input_tokens_seen": 13326688,
      "step": 15380
    },
    {
      "epoch": 7.2536539368222535,
      "grad_norm": 0.47383224964141846,
      "learning_rate": 3.386625195038503e-05,
      "loss": 0.1687,
      "num_input_tokens_seen": 13330816,
      "step": 15385
    },
    {
      "epoch": 7.256011315417256,
      "grad_norm": 1.02586030960083,
      "learning_rate": 3.3857071934751136e-05,
      "loss": 0.1979,
      "num_input_tokens_seen": 13335232,
      "step": 15390
    },
    {
      "epoch": 7.258368694012258,
      "grad_norm": 0.894547164440155,
      "learning_rate": 3.384789055324544e-05,
      "loss": 0.1221,
      "num_input_tokens_seen": 13339120,
      "step": 15395
    },
    {
      "epoch": 7.260726072607261,
      "grad_norm": 0.8916916847229004,
      "learning_rate": 3.3838707807283843e-05,
      "loss": 0.2108,
      "num_input_tokens_seen": 13342864,
      "step": 15400
    },
    {
      "epoch": 7.260726072607261,
      "eval_loss": 0.20571482181549072,
      "eval_runtime": 22.1536,
      "eval_samples_per_second": 42.566,
      "eval_steps_per_second": 21.306,
      "num_input_tokens_seen": 13342864,
      "step": 15400
    },
    {
      "epoch": 7.263083451202263,
      "grad_norm": 1.1993509531021118,
      "learning_rate": 3.382952369828243e-05,
      "loss": 0.1855,
      "num_input_tokens_seen": 13347184,
      "step": 15405
    },
    {
      "epoch": 7.2654408297972655,
      "grad_norm": 0.9327906370162964,
      "learning_rate": 3.38203382276575e-05,
      "loss": 0.1934,
      "num_input_tokens_seen": 13352176,
      "step": 15410
    },
    {
      "epoch": 7.267798208392268,
      "grad_norm": 0.6209182739257812,
      "learning_rate": 3.381115139682557e-05,
      "loss": 0.1547,
      "num_input_tokens_seen": 13356480,
      "step": 15415
    },
    {
      "epoch": 7.27015558698727,
      "grad_norm": 0.8423446416854858,
      "learning_rate": 3.3801963207203366e-05,
      "loss": 0.1605,
      "num_input_tokens_seen": 13360720,
      "step": 15420
    },
    {
      "epoch": 7.272512965582273,
      "grad_norm": 1.078455924987793,
      "learning_rate": 3.379277366020782e-05,
      "loss": 0.256,
      "num_input_tokens_seen": 13365552,
      "step": 15425
    },
    {
      "epoch": 7.274870344177275,
      "grad_norm": 1.0542681217193604,
      "learning_rate": 3.3783582757256085e-05,
      "loss": 0.253,
      "num_input_tokens_seen": 13369600,
      "step": 15430
    },
    {
      "epoch": 7.2772277227722775,
      "grad_norm": 1.2937639951705933,
      "learning_rate": 3.3774390499765504e-05,
      "loss": 0.1488,
      "num_input_tokens_seen": 13374576,
      "step": 15435
    },
    {
      "epoch": 7.27958510136728,
      "grad_norm": 0.9305339455604553,
      "learning_rate": 3.376519688915364e-05,
      "loss": 0.2304,
      "num_input_tokens_seen": 13379056,
      "step": 15440
    },
    {
      "epoch": 7.281942479962282,
      "grad_norm": 0.4646015465259552,
      "learning_rate": 3.3756001926838273e-05,
      "loss": 0.1931,
      "num_input_tokens_seen": 13384272,
      "step": 15445
    },
    {
      "epoch": 7.284299858557285,
      "grad_norm": 0.446076899766922,
      "learning_rate": 3.374680561423737e-05,
      "loss": 0.1711,
      "num_input_tokens_seen": 13387632,
      "step": 15450
    },
    {
      "epoch": 7.286657237152287,
      "grad_norm": 0.7124108076095581,
      "learning_rate": 3.373760795276912e-05,
      "loss": 0.1678,
      "num_input_tokens_seen": 13391408,
      "step": 15455
    },
    {
      "epoch": 7.2890146157472895,
      "grad_norm": 0.36179444193840027,
      "learning_rate": 3.372840894385192e-05,
      "loss": 0.1393,
      "num_input_tokens_seen": 13396240,
      "step": 15460
    },
    {
      "epoch": 7.291371994342291,
      "grad_norm": 0.6862141489982605,
      "learning_rate": 3.3719208588904375e-05,
      "loss": 0.1498,
      "num_input_tokens_seen": 13400336,
      "step": 15465
    },
    {
      "epoch": 7.293729372937293,
      "grad_norm": 0.877494215965271,
      "learning_rate": 3.371000688934529e-05,
      "loss": 0.1847,
      "num_input_tokens_seen": 13404384,
      "step": 15470
    },
    {
      "epoch": 7.296086751532296,
      "grad_norm": 0.5769132375717163,
      "learning_rate": 3.370080384659369e-05,
      "loss": 0.1649,
      "num_input_tokens_seen": 13408640,
      "step": 15475
    },
    {
      "epoch": 7.298444130127298,
      "grad_norm": 0.754225492477417,
      "learning_rate": 3.36915994620688e-05,
      "loss": 0.1609,
      "num_input_tokens_seen": 13412528,
      "step": 15480
    },
    {
      "epoch": 7.300801508722301,
      "grad_norm": 1.9170732498168945,
      "learning_rate": 3.3682393737190035e-05,
      "loss": 0.2139,
      "num_input_tokens_seen": 13416960,
      "step": 15485
    },
    {
      "epoch": 7.303158887317303,
      "grad_norm": 1.9104098081588745,
      "learning_rate": 3.3673186673377054e-05,
      "loss": 0.1948,
      "num_input_tokens_seen": 13421824,
      "step": 15490
    },
    {
      "epoch": 7.305516265912305,
      "grad_norm": 1.0177016258239746,
      "learning_rate": 3.366397827204969e-05,
      "loss": 0.1029,
      "num_input_tokens_seen": 13426240,
      "step": 15495
    },
    {
      "epoch": 7.307873644507308,
      "grad_norm": 1.0842493772506714,
      "learning_rate": 3.3654768534628e-05,
      "loss": 0.1873,
      "num_input_tokens_seen": 13430144,
      "step": 15500
    },
    {
      "epoch": 7.31023102310231,
      "grad_norm": 0.8667251467704773,
      "learning_rate": 3.3645557462532245e-05,
      "loss": 0.2541,
      "num_input_tokens_seen": 13434208,
      "step": 15505
    },
    {
      "epoch": 7.312588401697313,
      "grad_norm": 0.5552924871444702,
      "learning_rate": 3.363634505718288e-05,
      "loss": 0.2087,
      "num_input_tokens_seen": 13438896,
      "step": 15510
    },
    {
      "epoch": 7.314945780292315,
      "grad_norm": 2.325636148452759,
      "learning_rate": 3.362713132000057e-05,
      "loss": 0.2235,
      "num_input_tokens_seen": 13443152,
      "step": 15515
    },
    {
      "epoch": 7.317303158887317,
      "grad_norm": 0.7734192609786987,
      "learning_rate": 3.36179162524062e-05,
      "loss": 0.1948,
      "num_input_tokens_seen": 13447216,
      "step": 15520
    },
    {
      "epoch": 7.31966053748232,
      "grad_norm": 0.6679274439811707,
      "learning_rate": 3.3608699855820846e-05,
      "loss": 0.1798,
      "num_input_tokens_seen": 13451360,
      "step": 15525
    },
    {
      "epoch": 7.322017916077322,
      "grad_norm": 1.1421678066253662,
      "learning_rate": 3.359948213166578e-05,
      "loss": 0.2031,
      "num_input_tokens_seen": 13455584,
      "step": 15530
    },
    {
      "epoch": 7.324375294672325,
      "grad_norm": 0.6769078969955444,
      "learning_rate": 3.359026308136252e-05,
      "loss": 0.1473,
      "num_input_tokens_seen": 13460160,
      "step": 15535
    },
    {
      "epoch": 7.326732673267327,
      "grad_norm": 0.6370957493782043,
      "learning_rate": 3.358104270633272e-05,
      "loss": 0.1394,
      "num_input_tokens_seen": 13463728,
      "step": 15540
    },
    {
      "epoch": 7.329090051862329,
      "grad_norm": 0.5695913434028625,
      "learning_rate": 3.357182100799831e-05,
      "loss": 0.1758,
      "num_input_tokens_seen": 13468480,
      "step": 15545
    },
    {
      "epoch": 7.331447430457332,
      "grad_norm": 0.8525389432907104,
      "learning_rate": 3.3562597987781384e-05,
      "loss": 0.1734,
      "num_input_tokens_seen": 13472608,
      "step": 15550
    },
    {
      "epoch": 7.333804809052333,
      "grad_norm": 0.845404863357544,
      "learning_rate": 3.355337364710424e-05,
      "loss": 0.1743,
      "num_input_tokens_seen": 13476944,
      "step": 15555
    },
    {
      "epoch": 7.336162187647336,
      "grad_norm": 0.9708401560783386,
      "learning_rate": 3.354414798738939e-05,
      "loss": 0.2361,
      "num_input_tokens_seen": 13482192,
      "step": 15560
    },
    {
      "epoch": 7.338519566242338,
      "grad_norm": 0.7137155532836914,
      "learning_rate": 3.353492101005955e-05,
      "loss": 0.1705,
      "num_input_tokens_seen": 13486128,
      "step": 15565
    },
    {
      "epoch": 7.3408769448373405,
      "grad_norm": 1.0414886474609375,
      "learning_rate": 3.352569271653763e-05,
      "loss": 0.2254,
      "num_input_tokens_seen": 13491056,
      "step": 15570
    },
    {
      "epoch": 7.343234323432343,
      "grad_norm": 0.8118765950202942,
      "learning_rate": 3.351646310824675e-05,
      "loss": 0.157,
      "num_input_tokens_seen": 13494736,
      "step": 15575
    },
    {
      "epoch": 7.345591702027345,
      "grad_norm": 0.7604871988296509,
      "learning_rate": 3.350723218661023e-05,
      "loss": 0.2689,
      "num_input_tokens_seen": 13498848,
      "step": 15580
    },
    {
      "epoch": 7.347949080622348,
      "grad_norm": 1.0194250345230103,
      "learning_rate": 3.349799995305162e-05,
      "loss": 0.197,
      "num_input_tokens_seen": 13502752,
      "step": 15585
    },
    {
      "epoch": 7.35030645921735,
      "grad_norm": 1.7553609609603882,
      "learning_rate": 3.348876640899461e-05,
      "loss": 0.1705,
      "num_input_tokens_seen": 13507296,
      "step": 15590
    },
    {
      "epoch": 7.3526638378123526,
      "grad_norm": 0.6922000050544739,
      "learning_rate": 3.3479531555863144e-05,
      "loss": 0.1766,
      "num_input_tokens_seen": 13511584,
      "step": 15595
    },
    {
      "epoch": 7.355021216407355,
      "grad_norm": 2.299443006515503,
      "learning_rate": 3.3470295395081344e-05,
      "loss": 0.2056,
      "num_input_tokens_seen": 13515600,
      "step": 15600
    },
    {
      "epoch": 7.355021216407355,
      "eval_loss": 0.20454387366771698,
      "eval_runtime": 22.1742,
      "eval_samples_per_second": 42.527,
      "eval_steps_per_second": 21.286,
      "num_input_tokens_seen": 13515600,
      "step": 15600
    },
    {
      "epoch": 7.357378595002357,
      "grad_norm": 0.5784431099891663,
      "learning_rate": 3.3461057928073556e-05,
      "loss": 0.1774,
      "num_input_tokens_seen": 13520000,
      "step": 15605
    },
    {
      "epoch": 7.35973597359736,
      "grad_norm": 1.6543951034545898,
      "learning_rate": 3.345181915626431e-05,
      "loss": 0.1601,
      "num_input_tokens_seen": 13523840,
      "step": 15610
    },
    {
      "epoch": 7.362093352192362,
      "grad_norm": 0.4882192015647888,
      "learning_rate": 3.344257908107834e-05,
      "loss": 0.1877,
      "num_input_tokens_seen": 13528448,
      "step": 15615
    },
    {
      "epoch": 7.364450730787365,
      "grad_norm": 0.6292494535446167,
      "learning_rate": 3.343333770394058e-05,
      "loss": 0.1225,
      "num_input_tokens_seen": 13533104,
      "step": 15620
    },
    {
      "epoch": 7.366808109382367,
      "grad_norm": 0.7232236862182617,
      "learning_rate": 3.342409502627616e-05,
      "loss": 0.1682,
      "num_input_tokens_seen": 13537808,
      "step": 15625
    },
    {
      "epoch": 7.369165487977369,
      "grad_norm": 0.8292902708053589,
      "learning_rate": 3.341485104951043e-05,
      "loss": 0.2669,
      "num_input_tokens_seen": 13541952,
      "step": 15630
    },
    {
      "epoch": 7.371522866572372,
      "grad_norm": 0.7456992864608765,
      "learning_rate": 3.340560577506892e-05,
      "loss": 0.1648,
      "num_input_tokens_seen": 13546208,
      "step": 15635
    },
    {
      "epoch": 7.373880245167374,
      "grad_norm": 1.6387149095535278,
      "learning_rate": 3.339635920437735e-05,
      "loss": 0.2516,
      "num_input_tokens_seen": 13550480,
      "step": 15640
    },
    {
      "epoch": 7.376237623762377,
      "grad_norm": 0.8320251107215881,
      "learning_rate": 3.338711133886169e-05,
      "loss": 0.1684,
      "num_input_tokens_seen": 13554752,
      "step": 15645
    },
    {
      "epoch": 7.378595002357379,
      "grad_norm": 1.086830973625183,
      "learning_rate": 3.3377862179948064e-05,
      "loss": 0.2008,
      "num_input_tokens_seen": 13559696,
      "step": 15650
    },
    {
      "epoch": 7.380952380952381,
      "grad_norm": 0.6866241693496704,
      "learning_rate": 3.336861172906281e-05,
      "loss": 0.2023,
      "num_input_tokens_seen": 13563856,
      "step": 15655
    },
    {
      "epoch": 7.383309759547384,
      "grad_norm": 1.4996142387390137,
      "learning_rate": 3.335935998763245e-05,
      "loss": 0.1903,
      "num_input_tokens_seen": 13567760,
      "step": 15660
    },
    {
      "epoch": 7.385667138142385,
      "grad_norm": 0.9455329775810242,
      "learning_rate": 3.3350106957083744e-05,
      "loss": 0.1479,
      "num_input_tokens_seen": 13572624,
      "step": 15665
    },
    {
      "epoch": 7.388024516737388,
      "grad_norm": 0.6086110472679138,
      "learning_rate": 3.33408526388436e-05,
      "loss": 0.1971,
      "num_input_tokens_seen": 13577088,
      "step": 15670
    },
    {
      "epoch": 7.39038189533239,
      "grad_norm": 1.0848807096481323,
      "learning_rate": 3.3331597034339166e-05,
      "loss": 0.1312,
      "num_input_tokens_seen": 13580704,
      "step": 15675
    },
    {
      "epoch": 7.3927392739273925,
      "grad_norm": 2.263397455215454,
      "learning_rate": 3.3322340144997764e-05,
      "loss": 0.2242,
      "num_input_tokens_seen": 13584752,
      "step": 15680
    },
    {
      "epoch": 7.395096652522395,
      "grad_norm": 1.4971846342086792,
      "learning_rate": 3.331308197224693e-05,
      "loss": 0.1771,
      "num_input_tokens_seen": 13589504,
      "step": 15685
    },
    {
      "epoch": 7.397454031117397,
      "grad_norm": 1.994798183441162,
      "learning_rate": 3.330382251751438e-05,
      "loss": 0.2094,
      "num_input_tokens_seen": 13593312,
      "step": 15690
    },
    {
      "epoch": 7.3998114097124,
      "grad_norm": 0.8107173442840576,
      "learning_rate": 3.3294561782228054e-05,
      "loss": 0.1451,
      "num_input_tokens_seen": 13597968,
      "step": 15695
    },
    {
      "epoch": 7.402168788307402,
      "grad_norm": 1.005326747894287,
      "learning_rate": 3.328529976781607e-05,
      "loss": 0.1526,
      "num_input_tokens_seen": 13602448,
      "step": 15700
    },
    {
      "epoch": 7.4045261669024045,
      "grad_norm": 1.545971393585205,
      "learning_rate": 3.327603647570673e-05,
      "loss": 0.1647,
      "num_input_tokens_seen": 13606912,
      "step": 15705
    },
    {
      "epoch": 7.406883545497407,
      "grad_norm": 0.8596174120903015,
      "learning_rate": 3.326677190732857e-05,
      "loss": 0.2442,
      "num_input_tokens_seen": 13610944,
      "step": 15710
    },
    {
      "epoch": 7.409240924092409,
      "grad_norm": 0.6689602136611938,
      "learning_rate": 3.325750606411029e-05,
      "loss": 0.2654,
      "num_input_tokens_seen": 13615184,
      "step": 15715
    },
    {
      "epoch": 7.411598302687412,
      "grad_norm": 1.5220597982406616,
      "learning_rate": 3.3248238947480804e-05,
      "loss": 0.2102,
      "num_input_tokens_seen": 13619440,
      "step": 15720
    },
    {
      "epoch": 7.413955681282414,
      "grad_norm": 0.5599352121353149,
      "learning_rate": 3.323897055886922e-05,
      "loss": 0.1745,
      "num_input_tokens_seen": 13623152,
      "step": 15725
    },
    {
      "epoch": 7.4163130598774165,
      "grad_norm": 0.5717639327049255,
      "learning_rate": 3.322970089970484e-05,
      "loss": 0.2412,
      "num_input_tokens_seen": 13627936,
      "step": 15730
    },
    {
      "epoch": 7.418670438472419,
      "grad_norm": 0.6085986495018005,
      "learning_rate": 3.3220429971417165e-05,
      "loss": 0.1598,
      "num_input_tokens_seen": 13631968,
      "step": 15735
    },
    {
      "epoch": 7.421027817067421,
      "grad_norm": 3.1602184772491455,
      "learning_rate": 3.321115777543588e-05,
      "loss": 0.1957,
      "num_input_tokens_seen": 13636576,
      "step": 15740
    },
    {
      "epoch": 7.423385195662424,
      "grad_norm": 0.6531084775924683,
      "learning_rate": 3.320188431319088e-05,
      "loss": 0.2037,
      "num_input_tokens_seen": 13640688,
      "step": 15745
    },
    {
      "epoch": 7.425742574257426,
      "grad_norm": 1.3593131303787231,
      "learning_rate": 3.319260958611224e-05,
      "loss": 0.1919,
      "num_input_tokens_seen": 13644768,
      "step": 15750
    },
    {
      "epoch": 7.428099952852428,
      "grad_norm": 1.3592395782470703,
      "learning_rate": 3.3183333595630256e-05,
      "loss": 0.161,
      "num_input_tokens_seen": 13649040,
      "step": 15755
    },
    {
      "epoch": 7.43045733144743,
      "grad_norm": 0.4989885687828064,
      "learning_rate": 3.317405634317538e-05,
      "loss": 0.1598,
      "num_input_tokens_seen": 13653344,
      "step": 15760
    },
    {
      "epoch": 7.432814710042432,
      "grad_norm": 0.7922552227973938,
      "learning_rate": 3.3164777830178315e-05,
      "loss": 0.1821,
      "num_input_tokens_seen": 13657696,
      "step": 15765
    },
    {
      "epoch": 7.435172088637435,
      "grad_norm": 1.3659909963607788,
      "learning_rate": 3.315549805806989e-05,
      "loss": 0.1557,
      "num_input_tokens_seen": 13661696,
      "step": 15770
    },
    {
      "epoch": 7.437529467232437,
      "grad_norm": 0.5426134467124939,
      "learning_rate": 3.314621702828118e-05,
      "loss": 0.2355,
      "num_input_tokens_seen": 13666240,
      "step": 15775
    },
    {
      "epoch": 7.43988684582744,
      "grad_norm": 0.9254170060157776,
      "learning_rate": 3.313693474224342e-05,
      "loss": 0.1237,
      "num_input_tokens_seen": 13670944,
      "step": 15780
    },
    {
      "epoch": 7.442244224422442,
      "grad_norm": 1.1350444555282593,
      "learning_rate": 3.312765120138809e-05,
      "loss": 0.1451,
      "num_input_tokens_seen": 13676320,
      "step": 15785
    },
    {
      "epoch": 7.4446016030174444,
      "grad_norm": 1.2418251037597656,
      "learning_rate": 3.311836640714679e-05,
      "loss": 0.1996,
      "num_input_tokens_seen": 13680272,
      "step": 15790
    },
    {
      "epoch": 7.446958981612447,
      "grad_norm": 0.9639815092086792,
      "learning_rate": 3.310908036095137e-05,
      "loss": 0.1406,
      "num_input_tokens_seen": 13684880,
      "step": 15795
    },
    {
      "epoch": 7.449316360207449,
      "grad_norm": 3.186826705932617,
      "learning_rate": 3.309979306423386e-05,
      "loss": 0.2515,
      "num_input_tokens_seen": 13688640,
      "step": 15800
    },
    {
      "epoch": 7.449316360207449,
      "eval_loss": 0.20531496405601501,
      "eval_runtime": 22.1863,
      "eval_samples_per_second": 42.504,
      "eval_steps_per_second": 21.274,
      "num_input_tokens_seen": 13688640,
      "step": 15800
    },
    {
      "epoch": 7.451673738802452,
      "grad_norm": 3.053022623062134,
      "learning_rate": 3.309050451842647e-05,
      "loss": 0.3418,
      "num_input_tokens_seen": 13693616,
      "step": 15805
    },
    {
      "epoch": 7.454031117397454,
      "grad_norm": 0.4409962296485901,
      "learning_rate": 3.3081214724961604e-05,
      "loss": 0.1234,
      "num_input_tokens_seen": 13697392,
      "step": 15810
    },
    {
      "epoch": 7.4563884959924565,
      "grad_norm": 0.664632260799408,
      "learning_rate": 3.307192368527188e-05,
      "loss": 0.2063,
      "num_input_tokens_seen": 13702080,
      "step": 15815
    },
    {
      "epoch": 7.458745874587459,
      "grad_norm": 0.5809630751609802,
      "learning_rate": 3.306263140079008e-05,
      "loss": 0.2058,
      "num_input_tokens_seen": 13706192,
      "step": 15820
    },
    {
      "epoch": 7.461103253182461,
      "grad_norm": 0.5875013470649719,
      "learning_rate": 3.30533378729492e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 13710528,
      "step": 15825
    },
    {
      "epoch": 7.463460631777464,
      "grad_norm": 1.6383808851242065,
      "learning_rate": 3.304404310318242e-05,
      "loss": 0.134,
      "num_input_tokens_seen": 13714736,
      "step": 15830
    },
    {
      "epoch": 7.465818010372466,
      "grad_norm": 1.6303606033325195,
      "learning_rate": 3.3034747092923105e-05,
      "loss": 0.1945,
      "num_input_tokens_seen": 13719360,
      "step": 15835
    },
    {
      "epoch": 7.4681753889674685,
      "grad_norm": 1.030829906463623,
      "learning_rate": 3.3025449843604806e-05,
      "loss": 0.0991,
      "num_input_tokens_seen": 13722864,
      "step": 15840
    },
    {
      "epoch": 7.470532767562471,
      "grad_norm": 0.344463586807251,
      "learning_rate": 3.30161513566613e-05,
      "loss": 0.1608,
      "num_input_tokens_seen": 13727424,
      "step": 15845
    },
    {
      "epoch": 7.472890146157473,
      "grad_norm": 1.6585938930511475,
      "learning_rate": 3.3006851633526506e-05,
      "loss": 0.1655,
      "num_input_tokens_seen": 13730912,
      "step": 15850
    },
    {
      "epoch": 7.475247524752476,
      "grad_norm": 1.7737380266189575,
      "learning_rate": 3.2997550675634584e-05,
      "loss": 0.1979,
      "num_input_tokens_seen": 13734720,
      "step": 15855
    },
    {
      "epoch": 7.477604903347478,
      "grad_norm": 0.9712607264518738,
      "learning_rate": 3.2988248484419825e-05,
      "loss": 0.1853,
      "num_input_tokens_seen": 13739520,
      "step": 15860
    },
    {
      "epoch": 7.47996228194248,
      "grad_norm": 1.605780839920044,
      "learning_rate": 3.2978945061316776e-05,
      "loss": 0.2088,
      "num_input_tokens_seen": 13743504,
      "step": 15865
    },
    {
      "epoch": 7.482319660537482,
      "grad_norm": 0.8566784858703613,
      "learning_rate": 3.296964040776013e-05,
      "loss": 0.1479,
      "num_input_tokens_seen": 13747216,
      "step": 15870
    },
    {
      "epoch": 7.484677039132484,
      "grad_norm": 0.6041831374168396,
      "learning_rate": 3.296033452518478e-05,
      "loss": 0.1728,
      "num_input_tokens_seen": 13751872,
      "step": 15875
    },
    {
      "epoch": 7.487034417727487,
      "grad_norm": 1.1174020767211914,
      "learning_rate": 3.2951027415025806e-05,
      "loss": 0.2729,
      "num_input_tokens_seen": 13756848,
      "step": 15880
    },
    {
      "epoch": 7.489391796322489,
      "grad_norm": 0.5851876735687256,
      "learning_rate": 3.294171907871849e-05,
      "loss": 0.2223,
      "num_input_tokens_seen": 13761408,
      "step": 15885
    },
    {
      "epoch": 7.491749174917492,
      "grad_norm": 0.63139808177948,
      "learning_rate": 3.293240951769828e-05,
      "loss": 0.1783,
      "num_input_tokens_seen": 13766096,
      "step": 15890
    },
    {
      "epoch": 7.494106553512494,
      "grad_norm": 0.7132121324539185,
      "learning_rate": 3.2923098733400846e-05,
      "loss": 0.2033,
      "num_input_tokens_seen": 13770896,
      "step": 15895
    },
    {
      "epoch": 7.496463932107496,
      "grad_norm": 1.7688298225402832,
      "learning_rate": 3.291378672726202e-05,
      "loss": 0.2171,
      "num_input_tokens_seen": 13776048,
      "step": 15900
    },
    {
      "epoch": 7.498821310702499,
      "grad_norm": 1.6963119506835938,
      "learning_rate": 3.2904473500717824e-05,
      "loss": 0.2526,
      "num_input_tokens_seen": 13780752,
      "step": 15905
    },
    {
      "epoch": 7.501178689297501,
      "grad_norm": 0.2737867534160614,
      "learning_rate": 3.289515905520449e-05,
      "loss": 0.1232,
      "num_input_tokens_seen": 13784832,
      "step": 15910
    },
    {
      "epoch": 7.503536067892504,
      "grad_norm": 1.2434842586517334,
      "learning_rate": 3.288584339215841e-05,
      "loss": 0.168,
      "num_input_tokens_seen": 13789328,
      "step": 15915
    },
    {
      "epoch": 7.505893446487506,
      "grad_norm": 0.3487618565559387,
      "learning_rate": 3.287652651301617e-05,
      "loss": 0.119,
      "num_input_tokens_seen": 13793504,
      "step": 15920
    },
    {
      "epoch": 7.508250825082508,
      "grad_norm": 0.2868436574935913,
      "learning_rate": 3.286720841921457e-05,
      "loss": 0.1531,
      "num_input_tokens_seen": 13797968,
      "step": 15925
    },
    {
      "epoch": 7.510608203677511,
      "grad_norm": 0.8910343050956726,
      "learning_rate": 3.285788911219056e-05,
      "loss": 0.1471,
      "num_input_tokens_seen": 13801472,
      "step": 15930
    },
    {
      "epoch": 7.512965582272513,
      "grad_norm": 1.3343241214752197,
      "learning_rate": 3.284856859338131e-05,
      "loss": 0.1856,
      "num_input_tokens_seen": 13806832,
      "step": 15935
    },
    {
      "epoch": 7.515322960867516,
      "grad_norm": 0.6511170864105225,
      "learning_rate": 3.283924686422414e-05,
      "loss": 0.18,
      "num_input_tokens_seen": 13810896,
      "step": 15940
    },
    {
      "epoch": 7.517680339462518,
      "grad_norm": 1.062375545501709,
      "learning_rate": 3.282992392615659e-05,
      "loss": 0.2316,
      "num_input_tokens_seen": 13815856,
      "step": 15945
    },
    {
      "epoch": 7.52003771805752,
      "grad_norm": 0.9367988109588623,
      "learning_rate": 3.282059978061638e-05,
      "loss": 0.2585,
      "num_input_tokens_seen": 13819680,
      "step": 15950
    },
    {
      "epoch": 7.522395096652522,
      "grad_norm": 1.401437759399414,
      "learning_rate": 3.28112744290414e-05,
      "loss": 0.2491,
      "num_input_tokens_seen": 13823984,
      "step": 15955
    },
    {
      "epoch": 7.524752475247524,
      "grad_norm": 1.2278416156768799,
      "learning_rate": 3.280194787286974e-05,
      "loss": 0.1543,
      "num_input_tokens_seen": 13828256,
      "step": 15960
    },
    {
      "epoch": 7.527109853842527,
      "grad_norm": 1.658801555633545,
      "learning_rate": 3.2792620113539674e-05,
      "loss": 0.1254,
      "num_input_tokens_seen": 13832112,
      "step": 15965
    },
    {
      "epoch": 7.529467232437529,
      "grad_norm": 1.3792744874954224,
      "learning_rate": 3.278329115248966e-05,
      "loss": 0.2203,
      "num_input_tokens_seen": 13836928,
      "step": 15970
    },
    {
      "epoch": 7.5318246110325315,
      "grad_norm": 0.6474944353103638,
      "learning_rate": 3.277396099115834e-05,
      "loss": 0.2241,
      "num_input_tokens_seen": 13841232,
      "step": 15975
    },
    {
      "epoch": 7.534181989627534,
      "grad_norm": 1.3845995664596558,
      "learning_rate": 3.276462963098454e-05,
      "loss": 0.1936,
      "num_input_tokens_seen": 13845056,
      "step": 15980
    },
    {
      "epoch": 7.536539368222536,
      "grad_norm": 0.8982362747192383,
      "learning_rate": 3.275529707340728e-05,
      "loss": 0.2077,
      "num_input_tokens_seen": 13849648,
      "step": 15985
    },
    {
      "epoch": 7.538896746817539,
      "grad_norm": 1.1162265539169312,
      "learning_rate": 3.274596331986574e-05,
      "loss": 0.1654,
      "num_input_tokens_seen": 13853632,
      "step": 15990
    },
    {
      "epoch": 7.541254125412541,
      "grad_norm": 0.6830479502677917,
      "learning_rate": 3.273662837179932e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 13858288,
      "step": 15995
    },
    {
      "epoch": 7.5436115040075435,
      "grad_norm": 0.9637929201126099,
      "learning_rate": 3.272729223064758e-05,
      "loss": 0.2318,
      "num_input_tokens_seen": 13863312,
      "step": 16000
    },
    {
      "epoch": 7.5436115040075435,
      "eval_loss": 0.2059893012046814,
      "eval_runtime": 22.2196,
      "eval_samples_per_second": 42.44,
      "eval_steps_per_second": 21.243,
      "num_input_tokens_seen": 13863312,
      "step": 16000
    },
    {
      "epoch": 7.545968882602546,
      "grad_norm": 1.7353664636611938,
      "learning_rate": 3.2717954897850264e-05,
      "loss": 0.1652,
      "num_input_tokens_seen": 13867280,
      "step": 16005
    },
    {
      "epoch": 7.548326261197548,
      "grad_norm": 1.4842743873596191,
      "learning_rate": 3.270861637484733e-05,
      "loss": 0.3093,
      "num_input_tokens_seen": 13871360,
      "step": 16010
    },
    {
      "epoch": 7.550683639792551,
      "grad_norm": 0.5020714402198792,
      "learning_rate": 3.2699276663078867e-05,
      "loss": 0.1241,
      "num_input_tokens_seen": 13876032,
      "step": 16015
    },
    {
      "epoch": 7.553041018387553,
      "grad_norm": 0.9980263710021973,
      "learning_rate": 3.268993576398519e-05,
      "loss": 0.185,
      "num_input_tokens_seen": 13880784,
      "step": 16020
    },
    {
      "epoch": 7.5553983969825556,
      "grad_norm": 0.8152327537536621,
      "learning_rate": 3.268059367900678e-05,
      "loss": 0.2083,
      "num_input_tokens_seen": 13885280,
      "step": 16025
    },
    {
      "epoch": 7.557755775577558,
      "grad_norm": 0.6283565163612366,
      "learning_rate": 3.26712504095843e-05,
      "loss": 0.1387,
      "num_input_tokens_seen": 13889328,
      "step": 16030
    },
    {
      "epoch": 7.56011315417256,
      "grad_norm": 0.523647129535675,
      "learning_rate": 3.2661905957158615e-05,
      "loss": 0.1827,
      "num_input_tokens_seen": 13894256,
      "step": 16035
    },
    {
      "epoch": 7.562470532767563,
      "grad_norm": 1.3004494905471802,
      "learning_rate": 3.2652560323170734e-05,
      "loss": 0.2295,
      "num_input_tokens_seen": 13898560,
      "step": 16040
    },
    {
      "epoch": 7.564827911362565,
      "grad_norm": 1.5677598714828491,
      "learning_rate": 3.264321350906189e-05,
      "loss": 0.2458,
      "num_input_tokens_seen": 13902592,
      "step": 16045
    },
    {
      "epoch": 7.567185289957568,
      "grad_norm": 0.6412163972854614,
      "learning_rate": 3.263386551627346e-05,
      "loss": 0.186,
      "num_input_tokens_seen": 13906464,
      "step": 16050
    },
    {
      "epoch": 7.56954266855257,
      "grad_norm": 0.6807870268821716,
      "learning_rate": 3.2624516346247055e-05,
      "loss": 0.1651,
      "num_input_tokens_seen": 13910752,
      "step": 16055
    },
    {
      "epoch": 7.571900047147572,
      "grad_norm": 0.9999604225158691,
      "learning_rate": 3.2615166000424404e-05,
      "loss": 0.1777,
      "num_input_tokens_seen": 13915088,
      "step": 16060
    },
    {
      "epoch": 7.574257425742574,
      "grad_norm": 0.7007350921630859,
      "learning_rate": 3.260581448024745e-05,
      "loss": 0.1896,
      "num_input_tokens_seen": 13919104,
      "step": 16065
    },
    {
      "epoch": 7.576614804337576,
      "grad_norm": 0.9404815435409546,
      "learning_rate": 3.2596461787158335e-05,
      "loss": 0.1788,
      "num_input_tokens_seen": 13923392,
      "step": 16070
    },
    {
      "epoch": 7.578972182932579,
      "grad_norm": 1.6315641403198242,
      "learning_rate": 3.258710792259934e-05,
      "loss": 0.2252,
      "num_input_tokens_seen": 13928240,
      "step": 16075
    },
    {
      "epoch": 7.581329561527581,
      "grad_norm": 1.6217825412750244,
      "learning_rate": 3.257775288801296e-05,
      "loss": 0.1833,
      "num_input_tokens_seen": 13932464,
      "step": 16080
    },
    {
      "epoch": 7.5836869401225835,
      "grad_norm": 1.1543828248977661,
      "learning_rate": 3.256839668484186e-05,
      "loss": 0.1241,
      "num_input_tokens_seen": 13936272,
      "step": 16085
    },
    {
      "epoch": 7.586044318717586,
      "grad_norm": 0.6880226135253906,
      "learning_rate": 3.255903931452888e-05,
      "loss": 0.1453,
      "num_input_tokens_seen": 13940352,
      "step": 16090
    },
    {
      "epoch": 7.588401697312588,
      "grad_norm": 0.5388121604919434,
      "learning_rate": 3.2549680778517045e-05,
      "loss": 0.2055,
      "num_input_tokens_seen": 13945376,
      "step": 16095
    },
    {
      "epoch": 7.590759075907591,
      "grad_norm": 0.9271823763847351,
      "learning_rate": 3.2540321078249556e-05,
      "loss": 0.1928,
      "num_input_tokens_seen": 13949488,
      "step": 16100
    },
    {
      "epoch": 7.593116454502593,
      "grad_norm": 1.6025002002716064,
      "learning_rate": 3.2530960215169795e-05,
      "loss": 0.1324,
      "num_input_tokens_seen": 13954256,
      "step": 16105
    },
    {
      "epoch": 7.5954738330975955,
      "grad_norm": 1.649917483329773,
      "learning_rate": 3.2521598190721345e-05,
      "loss": 0.2005,
      "num_input_tokens_seen": 13958832,
      "step": 16110
    },
    {
      "epoch": 7.597831211692598,
      "grad_norm": 0.9242149591445923,
      "learning_rate": 3.251223500634792e-05,
      "loss": 0.1594,
      "num_input_tokens_seen": 13963008,
      "step": 16115
    },
    {
      "epoch": 7.6001885902876,
      "grad_norm": 1.2197200059890747,
      "learning_rate": 3.2502870663493445e-05,
      "loss": 0.2721,
      "num_input_tokens_seen": 13966944,
      "step": 16120
    },
    {
      "epoch": 7.602545968882603,
      "grad_norm": 2.310964822769165,
      "learning_rate": 3.249350516360203e-05,
      "loss": 0.2065,
      "num_input_tokens_seen": 13971520,
      "step": 16125
    },
    {
      "epoch": 7.604903347477605,
      "grad_norm": 0.814414918422699,
      "learning_rate": 3.248413850811797e-05,
      "loss": 0.2256,
      "num_input_tokens_seen": 13975888,
      "step": 16130
    },
    {
      "epoch": 7.6072607260726075,
      "grad_norm": 1.3000142574310303,
      "learning_rate": 3.2474770698485677e-05,
      "loss": 0.1909,
      "num_input_tokens_seen": 13980304,
      "step": 16135
    },
    {
      "epoch": 7.60961810466761,
      "grad_norm": 0.9954477548599243,
      "learning_rate": 3.246540173614983e-05,
      "loss": 0.1837,
      "num_input_tokens_seen": 13984016,
      "step": 16140
    },
    {
      "epoch": 7.611975483262612,
      "grad_norm": 0.6658660173416138,
      "learning_rate": 3.2456031622555197e-05,
      "loss": 0.15,
      "num_input_tokens_seen": 13988464,
      "step": 16145
    },
    {
      "epoch": 7.614332861857615,
      "grad_norm": 1.0630909204483032,
      "learning_rate": 3.2446660359146794e-05,
      "loss": 0.2653,
      "num_input_tokens_seen": 13992208,
      "step": 16150
    },
    {
      "epoch": 7.616690240452616,
      "grad_norm": 1.0686461925506592,
      "learning_rate": 3.2437287947369786e-05,
      "loss": 0.2276,
      "num_input_tokens_seen": 13996688,
      "step": 16155
    },
    {
      "epoch": 7.619047619047619,
      "grad_norm": 0.755664050579071,
      "learning_rate": 3.2427914388669525e-05,
      "loss": 0.1317,
      "num_input_tokens_seen": 14001056,
      "step": 16160
    },
    {
      "epoch": 7.621404997642621,
      "grad_norm": 1.5293070077896118,
      "learning_rate": 3.241853968449151e-05,
      "loss": 0.1552,
      "num_input_tokens_seen": 14004368,
      "step": 16165
    },
    {
      "epoch": 7.623762376237623,
      "grad_norm": 1.7637882232666016,
      "learning_rate": 3.240916383628144e-05,
      "loss": 0.2263,
      "num_input_tokens_seen": 14007936,
      "step": 16170
    },
    {
      "epoch": 7.626119754832626,
      "grad_norm": 0.6118456721305847,
      "learning_rate": 3.239978684548521e-05,
      "loss": 0.1941,
      "num_input_tokens_seen": 14012112,
      "step": 16175
    },
    {
      "epoch": 7.628477133427628,
      "grad_norm": 1.173054575920105,
      "learning_rate": 3.239040871354885e-05,
      "loss": 0.2155,
      "num_input_tokens_seen": 14016240,
      "step": 16180
    },
    {
      "epoch": 7.630834512022631,
      "grad_norm": 2.8544414043426514,
      "learning_rate": 3.2381029441918596e-05,
      "loss": 0.2074,
      "num_input_tokens_seen": 14020448,
      "step": 16185
    },
    {
      "epoch": 7.633191890617633,
      "grad_norm": 0.7823328375816345,
      "learning_rate": 3.2371649032040845e-05,
      "loss": 0.1303,
      "num_input_tokens_seen": 14024864,
      "step": 16190
    },
    {
      "epoch": 7.635549269212635,
      "grad_norm": 0.5438839197158813,
      "learning_rate": 3.2362267485362174e-05,
      "loss": 0.1581,
      "num_input_tokens_seen": 14029184,
      "step": 16195
    },
    {
      "epoch": 7.637906647807638,
      "grad_norm": 1.8918280601501465,
      "learning_rate": 3.235288480332934e-05,
      "loss": 0.2144,
      "num_input_tokens_seen": 14032992,
      "step": 16200
    },
    {
      "epoch": 7.637906647807638,
      "eval_loss": 0.20301644504070282,
      "eval_runtime": 22.1333,
      "eval_samples_per_second": 42.605,
      "eval_steps_per_second": 21.325,
      "num_input_tokens_seen": 14032992,
      "step": 16200
    },
    {
      "epoch": 7.64026402640264,
      "grad_norm": 0.898449182510376,
      "learning_rate": 3.234350098738927e-05,
      "loss": 0.2237,
      "num_input_tokens_seen": 14037728,
      "step": 16205
    },
    {
      "epoch": 7.642621404997643,
      "grad_norm": 0.6351487040519714,
      "learning_rate": 3.233411603898906e-05,
      "loss": 0.1886,
      "num_input_tokens_seen": 14042192,
      "step": 16210
    },
    {
      "epoch": 7.644978783592645,
      "grad_norm": 0.8796973824501038,
      "learning_rate": 3.232472995957599e-05,
      "loss": 0.1635,
      "num_input_tokens_seen": 14046144,
      "step": 16215
    },
    {
      "epoch": 7.6473361621876474,
      "grad_norm": 0.9292914271354675,
      "learning_rate": 3.231534275059751e-05,
      "loss": 0.232,
      "num_input_tokens_seen": 14050208,
      "step": 16220
    },
    {
      "epoch": 7.64969354078265,
      "grad_norm": 2.7565524578094482,
      "learning_rate": 3.230595441350125e-05,
      "loss": 0.209,
      "num_input_tokens_seen": 14054880,
      "step": 16225
    },
    {
      "epoch": 7.652050919377652,
      "grad_norm": 0.5013763904571533,
      "learning_rate": 3.2296564949735e-05,
      "loss": 0.1428,
      "num_input_tokens_seen": 14059600,
      "step": 16230
    },
    {
      "epoch": 7.654408297972655,
      "grad_norm": 0.8656436204910278,
      "learning_rate": 3.228717436074675e-05,
      "loss": 0.2462,
      "num_input_tokens_seen": 14062976,
      "step": 16235
    },
    {
      "epoch": 7.656765676567657,
      "grad_norm": 0.7591071128845215,
      "learning_rate": 3.227778264798463e-05,
      "loss": 0.12,
      "num_input_tokens_seen": 14067120,
      "step": 16240
    },
    {
      "epoch": 7.6591230551626595,
      "grad_norm": 0.4933767020702362,
      "learning_rate": 3.226838981289698e-05,
      "loss": 0.118,
      "num_input_tokens_seen": 14071696,
      "step": 16245
    },
    {
      "epoch": 7.661480433757662,
      "grad_norm": 0.9496174454689026,
      "learning_rate": 3.225899585693227e-05,
      "loss": 0.1585,
      "num_input_tokens_seen": 14075760,
      "step": 16250
    },
    {
      "epoch": 7.663837812352664,
      "grad_norm": 0.7389936447143555,
      "learning_rate": 3.224960078153918e-05,
      "loss": 0.212,
      "num_input_tokens_seen": 14080016,
      "step": 16255
    },
    {
      "epoch": 7.666195190947667,
      "grad_norm": 2.3415915966033936,
      "learning_rate": 3.224020458816655e-05,
      "loss": 0.2841,
      "num_input_tokens_seen": 14085360,
      "step": 16260
    },
    {
      "epoch": 7.668552569542668,
      "grad_norm": 1.4696134328842163,
      "learning_rate": 3.223080727826337e-05,
      "loss": 0.1498,
      "num_input_tokens_seen": 14089280,
      "step": 16265
    },
    {
      "epoch": 7.670909948137671,
      "grad_norm": 0.32028669118881226,
      "learning_rate": 3.222140885327885e-05,
      "loss": 0.2241,
      "num_input_tokens_seen": 14093728,
      "step": 16270
    },
    {
      "epoch": 7.673267326732673,
      "grad_norm": 1.3232120275497437,
      "learning_rate": 3.221200931466234e-05,
      "loss": 0.1635,
      "num_input_tokens_seen": 14098384,
      "step": 16275
    },
    {
      "epoch": 7.675624705327675,
      "grad_norm": 0.9181324243545532,
      "learning_rate": 3.220260866386336e-05,
      "loss": 0.2277,
      "num_input_tokens_seen": 14102880,
      "step": 16280
    },
    {
      "epoch": 7.677982083922678,
      "grad_norm": 1.351136326789856,
      "learning_rate": 3.21932069023316e-05,
      "loss": 0.1261,
      "num_input_tokens_seen": 14107184,
      "step": 16285
    },
    {
      "epoch": 7.68033946251768,
      "grad_norm": 0.8954136371612549,
      "learning_rate": 3.218380403151695e-05,
      "loss": 0.2053,
      "num_input_tokens_seen": 14111456,
      "step": 16290
    },
    {
      "epoch": 7.682696841112683,
      "grad_norm": 0.9531975984573364,
      "learning_rate": 3.217440005286943e-05,
      "loss": 0.1568,
      "num_input_tokens_seen": 14116192,
      "step": 16295
    },
    {
      "epoch": 7.685054219707685,
      "grad_norm": 0.5861485004425049,
      "learning_rate": 3.216499496783928e-05,
      "loss": 0.1799,
      "num_input_tokens_seen": 14120944,
      "step": 16300
    },
    {
      "epoch": 7.687411598302687,
      "grad_norm": 0.9155855178833008,
      "learning_rate": 3.2155588777876856e-05,
      "loss": 0.1811,
      "num_input_tokens_seen": 14125120,
      "step": 16305
    },
    {
      "epoch": 7.68976897689769,
      "grad_norm": 1.0687296390533447,
      "learning_rate": 3.214618148443273e-05,
      "loss": 0.1195,
      "num_input_tokens_seen": 14129616,
      "step": 16310
    },
    {
      "epoch": 7.692126355492692,
      "grad_norm": 1.1162196397781372,
      "learning_rate": 3.2136773088957595e-05,
      "loss": 0.1225,
      "num_input_tokens_seen": 14133696,
      "step": 16315
    },
    {
      "epoch": 7.694483734087695,
      "grad_norm": 1.1190108060836792,
      "learning_rate": 3.2127363592902374e-05,
      "loss": 0.1857,
      "num_input_tokens_seen": 14138128,
      "step": 16320
    },
    {
      "epoch": 7.696841112682697,
      "grad_norm": 0.9164262413978577,
      "learning_rate": 3.211795299771812e-05,
      "loss": 0.1694,
      "num_input_tokens_seen": 14142608,
      "step": 16325
    },
    {
      "epoch": 7.699198491277699,
      "grad_norm": 0.5430912375450134,
      "learning_rate": 3.210854130485605e-05,
      "loss": 0.2196,
      "num_input_tokens_seen": 14147120,
      "step": 16330
    },
    {
      "epoch": 7.701555869872702,
      "grad_norm": 1.4365328550338745,
      "learning_rate": 3.209912851576759e-05,
      "loss": 0.1974,
      "num_input_tokens_seen": 14151648,
      "step": 16335
    },
    {
      "epoch": 7.703913248467704,
      "grad_norm": 1.0242351293563843,
      "learning_rate": 3.208971463190431e-05,
      "loss": 0.2243,
      "num_input_tokens_seen": 14156512,
      "step": 16340
    },
    {
      "epoch": 7.706270627062707,
      "grad_norm": 0.9059776663780212,
      "learning_rate": 3.208029965471793e-05,
      "loss": 0.18,
      "num_input_tokens_seen": 14160640,
      "step": 16345
    },
    {
      "epoch": 7.708628005657709,
      "grad_norm": 0.9495840668678284,
      "learning_rate": 3.2070883585660364e-05,
      "loss": 0.1544,
      "num_input_tokens_seen": 14165088,
      "step": 16350
    },
    {
      "epoch": 7.7109853842527105,
      "grad_norm": 0.8794978857040405,
      "learning_rate": 3.20614664261837e-05,
      "loss": 0.1959,
      "num_input_tokens_seen": 14169376,
      "step": 16355
    },
    {
      "epoch": 7.713342762847713,
      "grad_norm": 1.1959044933319092,
      "learning_rate": 3.205204817774016e-05,
      "loss": 0.145,
      "num_input_tokens_seen": 14173600,
      "step": 16360
    },
    {
      "epoch": 7.715700141442715,
      "grad_norm": 1.5085724592208862,
      "learning_rate": 3.204262884178218e-05,
      "loss": 0.1762,
      "num_input_tokens_seen": 14177472,
      "step": 16365
    },
    {
      "epoch": 7.718057520037718,
      "grad_norm": 0.9040716886520386,
      "learning_rate": 3.2033208419762314e-05,
      "loss": 0.1983,
      "num_input_tokens_seen": 14181040,
      "step": 16370
    },
    {
      "epoch": 7.72041489863272,
      "grad_norm": 1.1884195804595947,
      "learning_rate": 3.2023786913133344e-05,
      "loss": 0.1819,
      "num_input_tokens_seen": 14184848,
      "step": 16375
    },
    {
      "epoch": 7.7227722772277225,
      "grad_norm": 0.9322276711463928,
      "learning_rate": 3.201436432334816e-05,
      "loss": 0.1282,
      "num_input_tokens_seen": 14189408,
      "step": 16380
    },
    {
      "epoch": 7.725129655822725,
      "grad_norm": 1.8421565294265747,
      "learning_rate": 3.2004940651859844e-05,
      "loss": 0.2521,
      "num_input_tokens_seen": 14193536,
      "step": 16385
    },
    {
      "epoch": 7.727487034417727,
      "grad_norm": 0.9316601157188416,
      "learning_rate": 3.1995515900121655e-05,
      "loss": 0.1833,
      "num_input_tokens_seen": 14198064,
      "step": 16390
    },
    {
      "epoch": 7.72984441301273,
      "grad_norm": 1.533643126487732,
      "learning_rate": 3.1986090069587e-05,
      "loss": 0.1322,
      "num_input_tokens_seen": 14202336,
      "step": 16395
    },
    {
      "epoch": 7.732201791607732,
      "grad_norm": 1.77792227268219,
      "learning_rate": 3.1976663161709466e-05,
      "loss": 0.18,
      "num_input_tokens_seen": 14205936,
      "step": 16400
    },
    {
      "epoch": 7.732201791607732,
      "eval_loss": 0.20439031720161438,
      "eval_runtime": 22.1614,
      "eval_samples_per_second": 42.551,
      "eval_steps_per_second": 21.298,
      "num_input_tokens_seen": 14205936,
      "step": 16400
    },
    {
      "epoch": 7.7345591702027345,
      "grad_norm": 0.7259771227836609,
      "learning_rate": 3.196723517794279e-05,
      "loss": 0.1825,
      "num_input_tokens_seen": 14210128,
      "step": 16405
    },
    {
      "epoch": 7.736916548797737,
      "grad_norm": 0.5076358914375305,
      "learning_rate": 3.19578061197409e-05,
      "loss": 0.1408,
      "num_input_tokens_seen": 14214608,
      "step": 16410
    },
    {
      "epoch": 7.739273927392739,
      "grad_norm": 0.9120587706565857,
      "learning_rate": 3.194837598855787e-05,
      "loss": 0.1697,
      "num_input_tokens_seen": 14219392,
      "step": 16415
    },
    {
      "epoch": 7.741631305987742,
      "grad_norm": 1.242974042892456,
      "learning_rate": 3.193894478584794e-05,
      "loss": 0.1066,
      "num_input_tokens_seen": 14223968,
      "step": 16420
    },
    {
      "epoch": 7.743988684582744,
      "grad_norm": 0.40875545144081116,
      "learning_rate": 3.192951251306553e-05,
      "loss": 0.1692,
      "num_input_tokens_seen": 14228160,
      "step": 16425
    },
    {
      "epoch": 7.7463460631777465,
      "grad_norm": 0.7767103910446167,
      "learning_rate": 3.192007917166521e-05,
      "loss": 0.1349,
      "num_input_tokens_seen": 14233168,
      "step": 16430
    },
    {
      "epoch": 7.748703441772749,
      "grad_norm": 0.7721659541130066,
      "learning_rate": 3.191064476310171e-05,
      "loss": 0.2025,
      "num_input_tokens_seen": 14237200,
      "step": 16435
    },
    {
      "epoch": 7.751060820367751,
      "grad_norm": 2.0111429691314697,
      "learning_rate": 3.1901209288829944e-05,
      "loss": 0.1481,
      "num_input_tokens_seen": 14241712,
      "step": 16440
    },
    {
      "epoch": 7.753418198962754,
      "grad_norm": 1.010788083076477,
      "learning_rate": 3.1891772750304985e-05,
      "loss": 0.2926,
      "num_input_tokens_seen": 14246112,
      "step": 16445
    },
    {
      "epoch": 7.755775577557756,
      "grad_norm": 0.31981006264686584,
      "learning_rate": 3.188233514898206e-05,
      "loss": 0.1534,
      "num_input_tokens_seen": 14249776,
      "step": 16450
    },
    {
      "epoch": 7.7581329561527586,
      "grad_norm": 2.1004600524902344,
      "learning_rate": 3.187289648631657e-05,
      "loss": 0.2175,
      "num_input_tokens_seen": 14253232,
      "step": 16455
    },
    {
      "epoch": 7.760490334747761,
      "grad_norm": 0.5463874340057373,
      "learning_rate": 3.186345676376406e-05,
      "loss": 0.2249,
      "num_input_tokens_seen": 14257152,
      "step": 16460
    },
    {
      "epoch": 7.7628477133427625,
      "grad_norm": 0.5471091866493225,
      "learning_rate": 3.1854015982780275e-05,
      "loss": 0.1709,
      "num_input_tokens_seen": 14261152,
      "step": 16465
    },
    {
      "epoch": 7.765205091937765,
      "grad_norm": 1.1221588850021362,
      "learning_rate": 3.1844574144821084e-05,
      "loss": 0.208,
      "num_input_tokens_seen": 14265632,
      "step": 16470
    },
    {
      "epoch": 7.767562470532767,
      "grad_norm": 1.6001466512680054,
      "learning_rate": 3.1835131251342554e-05,
      "loss": 0.152,
      "num_input_tokens_seen": 14269760,
      "step": 16475
    },
    {
      "epoch": 7.76991984912777,
      "grad_norm": 1.5194593667984009,
      "learning_rate": 3.182568730380089e-05,
      "loss": 0.2709,
      "num_input_tokens_seen": 14273440,
      "step": 16480
    },
    {
      "epoch": 7.772277227722772,
      "grad_norm": 0.5153416991233826,
      "learning_rate": 3.181624230365245e-05,
      "loss": 0.2128,
      "num_input_tokens_seen": 14278048,
      "step": 16485
    },
    {
      "epoch": 7.7746346063177745,
      "grad_norm": 0.6248987317085266,
      "learning_rate": 3.180679625235381e-05,
      "loss": 0.1668,
      "num_input_tokens_seen": 14282496,
      "step": 16490
    },
    {
      "epoch": 7.776991984912777,
      "grad_norm": 0.9418242573738098,
      "learning_rate": 3.1797349151361646e-05,
      "loss": 0.1635,
      "num_input_tokens_seen": 14286608,
      "step": 16495
    },
    {
      "epoch": 7.779349363507779,
      "grad_norm": 0.49499550461769104,
      "learning_rate": 3.178790100213281e-05,
      "loss": 0.1826,
      "num_input_tokens_seen": 14290560,
      "step": 16500
    },
    {
      "epoch": 7.781706742102782,
      "grad_norm": 0.5903633236885071,
      "learning_rate": 3.1778451806124346e-05,
      "loss": 0.2431,
      "num_input_tokens_seen": 14294512,
      "step": 16505
    },
    {
      "epoch": 7.784064120697784,
      "grad_norm": 0.8110105991363525,
      "learning_rate": 3.176900156479342e-05,
      "loss": 0.1908,
      "num_input_tokens_seen": 14298384,
      "step": 16510
    },
    {
      "epoch": 7.7864214992927865,
      "grad_norm": 0.8372633457183838,
      "learning_rate": 3.17595502795974e-05,
      "loss": 0.2461,
      "num_input_tokens_seen": 14302704,
      "step": 16515
    },
    {
      "epoch": 7.788778877887789,
      "grad_norm": 0.6037340760231018,
      "learning_rate": 3.175009795199377e-05,
      "loss": 0.1742,
      "num_input_tokens_seen": 14307552,
      "step": 16520
    },
    {
      "epoch": 7.791136256482791,
      "grad_norm": 1.414579153060913,
      "learning_rate": 3.1740644583440224e-05,
      "loss": 0.2866,
      "num_input_tokens_seen": 14313040,
      "step": 16525
    },
    {
      "epoch": 7.793493635077794,
      "grad_norm": 1.0044883489608765,
      "learning_rate": 3.173119017539457e-05,
      "loss": 0.2476,
      "num_input_tokens_seen": 14318384,
      "step": 16530
    },
    {
      "epoch": 7.795851013672796,
      "grad_norm": 0.755894124507904,
      "learning_rate": 3.172173472931479e-05,
      "loss": 0.1803,
      "num_input_tokens_seen": 14322672,
      "step": 16535
    },
    {
      "epoch": 7.7982083922677985,
      "grad_norm": 1.380680799484253,
      "learning_rate": 3.1712278246659055e-05,
      "loss": 0.2548,
      "num_input_tokens_seen": 14326944,
      "step": 16540
    },
    {
      "epoch": 7.800565770862801,
      "grad_norm": 0.8084297180175781,
      "learning_rate": 3.170282072888566e-05,
      "loss": 0.1463,
      "num_input_tokens_seen": 14331360,
      "step": 16545
    },
    {
      "epoch": 7.802923149457803,
      "grad_norm": 0.3474099934101105,
      "learning_rate": 3.169336217745307e-05,
      "loss": 0.201,
      "num_input_tokens_seen": 14335888,
      "step": 16550
    },
    {
      "epoch": 7.805280528052805,
      "grad_norm": 0.558945894241333,
      "learning_rate": 3.1683902593819924e-05,
      "loss": 0.1342,
      "num_input_tokens_seen": 14339856,
      "step": 16555
    },
    {
      "epoch": 7.807637906647807,
      "grad_norm": 0.8056771755218506,
      "learning_rate": 3.1674441979445e-05,
      "loss": 0.1532,
      "num_input_tokens_seen": 14344160,
      "step": 16560
    },
    {
      "epoch": 7.80999528524281,
      "grad_norm": 0.8014129996299744,
      "learning_rate": 3.166498033578725e-05,
      "loss": 0.2015,
      "num_input_tokens_seen": 14348880,
      "step": 16565
    },
    {
      "epoch": 7.812352663837812,
      "grad_norm": 0.4678465723991394,
      "learning_rate": 3.165551766430578e-05,
      "loss": 0.2112,
      "num_input_tokens_seen": 14353664,
      "step": 16570
    },
    {
      "epoch": 7.814710042432814,
      "grad_norm": 0.8789570331573486,
      "learning_rate": 3.164605396645984e-05,
      "loss": 0.1792,
      "num_input_tokens_seen": 14357184,
      "step": 16575
    },
    {
      "epoch": 7.817067421027817,
      "grad_norm": 1.4044251441955566,
      "learning_rate": 3.163658924370886e-05,
      "loss": 0.1684,
      "num_input_tokens_seen": 14361408,
      "step": 16580
    },
    {
      "epoch": 7.819424799622819,
      "grad_norm": 0.6444565057754517,
      "learning_rate": 3.1627123497512415e-05,
      "loss": 0.1995,
      "num_input_tokens_seen": 14365456,
      "step": 16585
    },
    {
      "epoch": 7.821782178217822,
      "grad_norm": 0.5568722486495972,
      "learning_rate": 3.1617656729330245e-05,
      "loss": 0.151,
      "num_input_tokens_seen": 14369680,
      "step": 16590
    },
    {
      "epoch": 7.824139556812824,
      "grad_norm": 1.3080278635025024,
      "learning_rate": 3.1608188940622255e-05,
      "loss": 0.1947,
      "num_input_tokens_seen": 14374016,
      "step": 16595
    },
    {
      "epoch": 7.826496935407826,
      "grad_norm": 0.4590199291706085,
      "learning_rate": 3.159872013284847e-05,
      "loss": 0.1864,
      "num_input_tokens_seen": 14378336,
      "step": 16600
    },
    {
      "epoch": 7.826496935407826,
      "eval_loss": 0.20352429151535034,
      "eval_runtime": 22.1725,
      "eval_samples_per_second": 42.53,
      "eval_steps_per_second": 21.288,
      "num_input_tokens_seen": 14378336,
      "step": 16600
    },
    {
      "epoch": 7.828854314002829,
      "grad_norm": 0.5407246947288513,
      "learning_rate": 3.1589250307469134e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 14383152,
      "step": 16605
    },
    {
      "epoch": 7.831211692597831,
      "grad_norm": 1.500565528869629,
      "learning_rate": 3.1579779465944586e-05,
      "loss": 0.2527,
      "num_input_tokens_seen": 14387344,
      "step": 16610
    },
    {
      "epoch": 7.833569071192834,
      "grad_norm": 1.171248197555542,
      "learning_rate": 3.1570307609735363e-05,
      "loss": 0.1748,
      "num_input_tokens_seen": 14391088,
      "step": 16615
    },
    {
      "epoch": 7.835926449787836,
      "grad_norm": 0.7232687473297119,
      "learning_rate": 3.156083474030213e-05,
      "loss": 0.1924,
      "num_input_tokens_seen": 14396032,
      "step": 16620
    },
    {
      "epoch": 7.838283828382838,
      "grad_norm": 0.7298116683959961,
      "learning_rate": 3.155136085910573e-05,
      "loss": 0.1322,
      "num_input_tokens_seen": 14400944,
      "step": 16625
    },
    {
      "epoch": 7.840641206977841,
      "grad_norm": 0.9986057281494141,
      "learning_rate": 3.154188596760717e-05,
      "loss": 0.1769,
      "num_input_tokens_seen": 14405856,
      "step": 16630
    },
    {
      "epoch": 7.842998585572843,
      "grad_norm": 1.136035442352295,
      "learning_rate": 3.153241006726757e-05,
      "loss": 0.1414,
      "num_input_tokens_seen": 14410528,
      "step": 16635
    },
    {
      "epoch": 7.845355964167846,
      "grad_norm": 1.408408522605896,
      "learning_rate": 3.152293315954825e-05,
      "loss": 0.1755,
      "num_input_tokens_seen": 14415520,
      "step": 16640
    },
    {
      "epoch": 7.847713342762848,
      "grad_norm": 1.258259654045105,
      "learning_rate": 3.1513455245910666e-05,
      "loss": 0.1404,
      "num_input_tokens_seen": 14420144,
      "step": 16645
    },
    {
      "epoch": 7.8500707213578504,
      "grad_norm": 1.3888750076293945,
      "learning_rate": 3.150397632781643e-05,
      "loss": 0.2634,
      "num_input_tokens_seen": 14424752,
      "step": 16650
    },
    {
      "epoch": 7.852428099952853,
      "grad_norm": 1.0159180164337158,
      "learning_rate": 3.149449640672731e-05,
      "loss": 0.2409,
      "num_input_tokens_seen": 14428800,
      "step": 16655
    },
    {
      "epoch": 7.854785478547855,
      "grad_norm": 0.7951340675354004,
      "learning_rate": 3.148501548410523e-05,
      "loss": 0.1519,
      "num_input_tokens_seen": 14433488,
      "step": 16660
    },
    {
      "epoch": 7.857142857142857,
      "grad_norm": 0.7637442946434021,
      "learning_rate": 3.1475533561412256e-05,
      "loss": 0.1134,
      "num_input_tokens_seen": 14437232,
      "step": 16665
    },
    {
      "epoch": 7.859500235737859,
      "grad_norm": 0.9069541096687317,
      "learning_rate": 3.146605064011065e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 14442208,
      "step": 16670
    },
    {
      "epoch": 7.861857614332862,
      "grad_norm": 1.5802024602890015,
      "learning_rate": 3.145656672166277e-05,
      "loss": 0.1735,
      "num_input_tokens_seen": 14446208,
      "step": 16675
    },
    {
      "epoch": 7.864214992927864,
      "grad_norm": 0.7243556380271912,
      "learning_rate": 3.144708180753116e-05,
      "loss": 0.2224,
      "num_input_tokens_seen": 14450464,
      "step": 16680
    },
    {
      "epoch": 7.866572371522866,
      "grad_norm": 0.746988832950592,
      "learning_rate": 3.143759589917851e-05,
      "loss": 0.2058,
      "num_input_tokens_seen": 14454256,
      "step": 16685
    },
    {
      "epoch": 7.868929750117869,
      "grad_norm": 1.233904242515564,
      "learning_rate": 3.142810899806768e-05,
      "loss": 0.114,
      "num_input_tokens_seen": 14459088,
      "step": 16690
    },
    {
      "epoch": 7.871287128712871,
      "grad_norm": 1.5298652648925781,
      "learning_rate": 3.141862110566166e-05,
      "loss": 0.1573,
      "num_input_tokens_seen": 14462928,
      "step": 16695
    },
    {
      "epoch": 7.873644507307874,
      "grad_norm": 1.5268418788909912,
      "learning_rate": 3.1409132223423606e-05,
      "loss": 0.206,
      "num_input_tokens_seen": 14467824,
      "step": 16700
    },
    {
      "epoch": 7.876001885902876,
      "grad_norm": 1.111716628074646,
      "learning_rate": 3.139964235281682e-05,
      "loss": 0.12,
      "num_input_tokens_seen": 14472000,
      "step": 16705
    },
    {
      "epoch": 7.878359264497878,
      "grad_norm": 0.9822524785995483,
      "learning_rate": 3.139015149530476e-05,
      "loss": 0.1345,
      "num_input_tokens_seen": 14475632,
      "step": 16710
    },
    {
      "epoch": 7.880716643092881,
      "grad_norm": 1.0198801755905151,
      "learning_rate": 3.1380659652351034e-05,
      "loss": 0.1643,
      "num_input_tokens_seen": 14479696,
      "step": 16715
    },
    {
      "epoch": 7.883074021687883,
      "grad_norm": 1.0086276531219482,
      "learning_rate": 3.137116682541941e-05,
      "loss": 0.1802,
      "num_input_tokens_seen": 14483776,
      "step": 16720
    },
    {
      "epoch": 7.885431400282886,
      "grad_norm": 1.0707694292068481,
      "learning_rate": 3.136167301597379e-05,
      "loss": 0.1671,
      "num_input_tokens_seen": 14487472,
      "step": 16725
    },
    {
      "epoch": 7.887788778877888,
      "grad_norm": 2.2638936042785645,
      "learning_rate": 3.1352178225478254e-05,
      "loss": 0.246,
      "num_input_tokens_seen": 14491568,
      "step": 16730
    },
    {
      "epoch": 7.89014615747289,
      "grad_norm": 0.49570995569229126,
      "learning_rate": 3.1342682455396996e-05,
      "loss": 0.2323,
      "num_input_tokens_seen": 14496336,
      "step": 16735
    },
    {
      "epoch": 7.892503536067893,
      "grad_norm": 0.7311133146286011,
      "learning_rate": 3.133318570719441e-05,
      "loss": 0.192,
      "num_input_tokens_seen": 14501200,
      "step": 16740
    },
    {
      "epoch": 7.894860914662895,
      "grad_norm": 1.4333617687225342,
      "learning_rate": 3.132368798233499e-05,
      "loss": 0.1767,
      "num_input_tokens_seen": 14505296,
      "step": 16745
    },
    {
      "epoch": 7.897218293257898,
      "grad_norm": 0.3506915867328644,
      "learning_rate": 3.131418928228342e-05,
      "loss": 0.147,
      "num_input_tokens_seen": 14509536,
      "step": 16750
    },
    {
      "epoch": 7.899575671852899,
      "grad_norm": 3.7079296112060547,
      "learning_rate": 3.1304689608504514e-05,
      "loss": 0.2415,
      "num_input_tokens_seen": 14513920,
      "step": 16755
    },
    {
      "epoch": 7.9019330504479015,
      "grad_norm": 1.6049811840057373,
      "learning_rate": 3.129518896246324e-05,
      "loss": 0.1968,
      "num_input_tokens_seen": 14518016,
      "step": 16760
    },
    {
      "epoch": 7.904290429042904,
      "grad_norm": 0.5403749942779541,
      "learning_rate": 3.128568734562472e-05,
      "loss": 0.151,
      "num_input_tokens_seen": 14521296,
      "step": 16765
    },
    {
      "epoch": 7.906647807637906,
      "grad_norm": 0.8352468013763428,
      "learning_rate": 3.127618475945421e-05,
      "loss": 0.1432,
      "num_input_tokens_seen": 14525600,
      "step": 16770
    },
    {
      "epoch": 7.909005186232909,
      "grad_norm": 0.40717336535453796,
      "learning_rate": 3.126668120541715e-05,
      "loss": 0.1709,
      "num_input_tokens_seen": 14529600,
      "step": 16775
    },
    {
      "epoch": 7.911362564827911,
      "grad_norm": 0.7904708981513977,
      "learning_rate": 3.1257176684979096e-05,
      "loss": 0.2613,
      "num_input_tokens_seen": 14533744,
      "step": 16780
    },
    {
      "epoch": 7.9137199434229135,
      "grad_norm": 2.3911688327789307,
      "learning_rate": 3.124767119960576e-05,
      "loss": 0.1751,
      "num_input_tokens_seen": 14538672,
      "step": 16785
    },
    {
      "epoch": 7.916077322017916,
      "grad_norm": 0.3069595694541931,
      "learning_rate": 3.123816475076301e-05,
      "loss": 0.1218,
      "num_input_tokens_seen": 14542560,
      "step": 16790
    },
    {
      "epoch": 7.918434700612918,
      "grad_norm": 1.3352597951889038,
      "learning_rate": 3.122865733991687e-05,
      "loss": 0.1579,
      "num_input_tokens_seen": 14547376,
      "step": 16795
    },
    {
      "epoch": 7.920792079207921,
      "grad_norm": 0.48408639430999756,
      "learning_rate": 3.1219148968533486e-05,
      "loss": 0.1913,
      "num_input_tokens_seen": 14551456,
      "step": 16800
    },
    {
      "epoch": 7.920792079207921,
      "eval_loss": 0.20315596461296082,
      "eval_runtime": 22.1438,
      "eval_samples_per_second": 42.585,
      "eval_steps_per_second": 21.315,
      "num_input_tokens_seen": 14551456,
      "step": 16800
    },
    {
      "epoch": 7.923149457802923,
      "grad_norm": 0.37427181005477905,
      "learning_rate": 3.120963963807918e-05,
      "loss": 0.1322,
      "num_input_tokens_seen": 14555536,
      "step": 16805
    },
    {
      "epoch": 7.9255068363979255,
      "grad_norm": 0.8846280574798584,
      "learning_rate": 3.12001293500204e-05,
      "loss": 0.1981,
      "num_input_tokens_seen": 14559824,
      "step": 16810
    },
    {
      "epoch": 7.927864214992928,
      "grad_norm": 0.8332262635231018,
      "learning_rate": 3.1190618105823765e-05,
      "loss": 0.1515,
      "num_input_tokens_seen": 14563536,
      "step": 16815
    },
    {
      "epoch": 7.93022159358793,
      "grad_norm": 0.8557077050209045,
      "learning_rate": 3.118110590695603e-05,
      "loss": 0.1667,
      "num_input_tokens_seen": 14567600,
      "step": 16820
    },
    {
      "epoch": 7.932578972182933,
      "grad_norm": 0.6479594111442566,
      "learning_rate": 3.117159275488407e-05,
      "loss": 0.1425,
      "num_input_tokens_seen": 14571920,
      "step": 16825
    },
    {
      "epoch": 7.934936350777935,
      "grad_norm": 1.428499698638916,
      "learning_rate": 3.1162078651074956e-05,
      "loss": 0.1349,
      "num_input_tokens_seen": 14576016,
      "step": 16830
    },
    {
      "epoch": 7.9372937293729375,
      "grad_norm": 0.7519460320472717,
      "learning_rate": 3.1152563596995885e-05,
      "loss": 0.1805,
      "num_input_tokens_seen": 14581152,
      "step": 16835
    },
    {
      "epoch": 7.93965110796794,
      "grad_norm": 1.0656911134719849,
      "learning_rate": 3.1143047594114186e-05,
      "loss": 0.15,
      "num_input_tokens_seen": 14586000,
      "step": 16840
    },
    {
      "epoch": 7.942008486562942,
      "grad_norm": 0.9655153155326843,
      "learning_rate": 3.113353064389734e-05,
      "loss": 0.1718,
      "num_input_tokens_seen": 14589984,
      "step": 16845
    },
    {
      "epoch": 7.944365865157945,
      "grad_norm": 1.6285051107406616,
      "learning_rate": 3.1124012747812993e-05,
      "loss": 0.1668,
      "num_input_tokens_seen": 14594240,
      "step": 16850
    },
    {
      "epoch": 7.946723243752947,
      "grad_norm": 0.4127196669578552,
      "learning_rate": 3.1114493907328936e-05,
      "loss": 0.206,
      "num_input_tokens_seen": 14598880,
      "step": 16855
    },
    {
      "epoch": 7.9490806223479495,
      "grad_norm": 0.9313958287239075,
      "learning_rate": 3.110497412391306e-05,
      "loss": 0.1538,
      "num_input_tokens_seen": 14603888,
      "step": 16860
    },
    {
      "epoch": 7.951438000942951,
      "grad_norm": 1.239810585975647,
      "learning_rate": 3.1095453399033466e-05,
      "loss": 0.212,
      "num_input_tokens_seen": 14608448,
      "step": 16865
    },
    {
      "epoch": 7.9537953795379535,
      "grad_norm": 1.1008589267730713,
      "learning_rate": 3.108593173415835e-05,
      "loss": 0.1934,
      "num_input_tokens_seen": 14612624,
      "step": 16870
    },
    {
      "epoch": 7.956152758132956,
      "grad_norm": 0.3690917193889618,
      "learning_rate": 3.107640913075609e-05,
      "loss": 0.1178,
      "num_input_tokens_seen": 14617136,
      "step": 16875
    },
    {
      "epoch": 7.958510136727958,
      "grad_norm": 1.6222190856933594,
      "learning_rate": 3.106688559029517e-05,
      "loss": 0.2259,
      "num_input_tokens_seen": 14621376,
      "step": 16880
    },
    {
      "epoch": 7.960867515322961,
      "grad_norm": 1.02181077003479,
      "learning_rate": 3.105736111424425e-05,
      "loss": 0.2094,
      "num_input_tokens_seen": 14625424,
      "step": 16885
    },
    {
      "epoch": 7.963224893917963,
      "grad_norm": 0.7573238611221313,
      "learning_rate": 3.1047835704072136e-05,
      "loss": 0.1316,
      "num_input_tokens_seen": 14629488,
      "step": 16890
    },
    {
      "epoch": 7.9655822725129655,
      "grad_norm": 0.22145625948905945,
      "learning_rate": 3.103830936124775e-05,
      "loss": 0.1844,
      "num_input_tokens_seen": 14633920,
      "step": 16895
    },
    {
      "epoch": 7.967939651107968,
      "grad_norm": 1.0543633699417114,
      "learning_rate": 3.102878208724018e-05,
      "loss": 0.1633,
      "num_input_tokens_seen": 14638144,
      "step": 16900
    },
    {
      "epoch": 7.97029702970297,
      "grad_norm": 0.8682079911231995,
      "learning_rate": 3.101925388351865e-05,
      "loss": 0.2895,
      "num_input_tokens_seen": 14642976,
      "step": 16905
    },
    {
      "epoch": 7.972654408297973,
      "grad_norm": 1.657678246498108,
      "learning_rate": 3.1009724751552515e-05,
      "loss": 0.1622,
      "num_input_tokens_seen": 14647136,
      "step": 16910
    },
    {
      "epoch": 7.975011786892975,
      "grad_norm": 0.4770064949989319,
      "learning_rate": 3.100019469281131e-05,
      "loss": 0.1837,
      "num_input_tokens_seen": 14652672,
      "step": 16915
    },
    {
      "epoch": 7.9773691654879775,
      "grad_norm": 0.37323901057243347,
      "learning_rate": 3.0990663708764685e-05,
      "loss": 0.1301,
      "num_input_tokens_seen": 14657952,
      "step": 16920
    },
    {
      "epoch": 7.97972654408298,
      "grad_norm": 0.5429285764694214,
      "learning_rate": 3.098113180088243e-05,
      "loss": 0.3221,
      "num_input_tokens_seen": 14662768,
      "step": 16925
    },
    {
      "epoch": 7.982083922677982,
      "grad_norm": 0.776161253452301,
      "learning_rate": 3.097159897063448e-05,
      "loss": 0.1285,
      "num_input_tokens_seen": 14667488,
      "step": 16930
    },
    {
      "epoch": 7.984441301272985,
      "grad_norm": 0.7908071279525757,
      "learning_rate": 3.096206521949094e-05,
      "loss": 0.2268,
      "num_input_tokens_seen": 14672656,
      "step": 16935
    },
    {
      "epoch": 7.986798679867987,
      "grad_norm": 0.9174708724021912,
      "learning_rate": 3.0952530548922006e-05,
      "loss": 0.1604,
      "num_input_tokens_seen": 14677168,
      "step": 16940
    },
    {
      "epoch": 7.9891560584629895,
      "grad_norm": 0.6390156149864197,
      "learning_rate": 3.0942994960398064e-05,
      "loss": 0.1741,
      "num_input_tokens_seen": 14681904,
      "step": 16945
    },
    {
      "epoch": 7.991513437057992,
      "grad_norm": 0.49997633695602417,
      "learning_rate": 3.093345845538961e-05,
      "loss": 0.173,
      "num_input_tokens_seen": 14685856,
      "step": 16950
    },
    {
      "epoch": 7.993870815652993,
      "grad_norm": 1.9929981231689453,
      "learning_rate": 3.09239210353673e-05,
      "loss": 0.1726,
      "num_input_tokens_seen": 14691056,
      "step": 16955
    },
    {
      "epoch": 7.996228194247996,
      "grad_norm": 1.9004284143447876,
      "learning_rate": 3.0914382701801926e-05,
      "loss": 0.1522,
      "num_input_tokens_seen": 14695488,
      "step": 16960
    },
    {
      "epoch": 7.998585572842998,
      "grad_norm": 0.7050639390945435,
      "learning_rate": 3.090484345616441e-05,
      "loss": 0.2017,
      "num_input_tokens_seen": 14699536,
      "step": 16965
    },
    {
      "epoch": 8.000942951438,
      "grad_norm": 0.852756679058075,
      "learning_rate": 3.0895303299925825e-05,
      "loss": 0.1901,
      "num_input_tokens_seen": 14704272,
      "step": 16970
    },
    {
      "epoch": 8.003300330033003,
      "grad_norm": 1.92157781124115,
      "learning_rate": 3.0885762234557393e-05,
      "loss": 0.2377,
      "num_input_tokens_seen": 14708352,
      "step": 16975
    },
    {
      "epoch": 8.005657708628005,
      "grad_norm": 0.9538938999176025,
      "learning_rate": 3.087622026153045e-05,
      "loss": 0.1672,
      "num_input_tokens_seen": 14712880,
      "step": 16980
    },
    {
      "epoch": 8.008015087223008,
      "grad_norm": 1.247992753982544,
      "learning_rate": 3.086667738231651e-05,
      "loss": 0.1523,
      "num_input_tokens_seen": 14717344,
      "step": 16985
    },
    {
      "epoch": 8.01037246581801,
      "grad_norm": 1.1932709217071533,
      "learning_rate": 3.085713359838718e-05,
      "loss": 0.1216,
      "num_input_tokens_seen": 14721632,
      "step": 16990
    },
    {
      "epoch": 8.012729844413013,
      "grad_norm": 0.8097779154777527,
      "learning_rate": 3.084758891121425e-05,
      "loss": 0.1805,
      "num_input_tokens_seen": 14726080,
      "step": 16995
    },
    {
      "epoch": 8.015087223008015,
      "grad_norm": 0.4648035168647766,
      "learning_rate": 3.083804332226963e-05,
      "loss": 0.1406,
      "num_input_tokens_seen": 14730672,
      "step": 17000
    },
    {
      "epoch": 8.015087223008015,
      "eval_loss": 0.20384275913238525,
      "eval_runtime": 22.1606,
      "eval_samples_per_second": 42.553,
      "eval_steps_per_second": 21.299,
      "num_input_tokens_seen": 14730672,
      "step": 17000
    },
    {
      "epoch": 8.017444601603017,
      "grad_norm": 0.8245071172714233,
      "learning_rate": 3.082849683302536e-05,
      "loss": 0.1649,
      "num_input_tokens_seen": 14734560,
      "step": 17005
    },
    {
      "epoch": 8.01980198019802,
      "grad_norm": 1.4314666986465454,
      "learning_rate": 3.081894944495363e-05,
      "loss": 0.177,
      "num_input_tokens_seen": 14738896,
      "step": 17010
    },
    {
      "epoch": 8.022159358793022,
      "grad_norm": 0.8272209167480469,
      "learning_rate": 3.080940115952677e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 14743120,
      "step": 17015
    },
    {
      "epoch": 8.024516737388025,
      "grad_norm": 0.6788647770881653,
      "learning_rate": 3.0799851978217245e-05,
      "loss": 0.2214,
      "num_input_tokens_seen": 14747504,
      "step": 17020
    },
    {
      "epoch": 8.026874115983027,
      "grad_norm": 0.7257833480834961,
      "learning_rate": 3.0790301902497666e-05,
      "loss": 0.1509,
      "num_input_tokens_seen": 14751856,
      "step": 17025
    },
    {
      "epoch": 8.02923149457803,
      "grad_norm": 0.6580378413200378,
      "learning_rate": 3.078075093384076e-05,
      "loss": 0.3075,
      "num_input_tokens_seen": 14756288,
      "step": 17030
    },
    {
      "epoch": 8.031588873173032,
      "grad_norm": 2.3782215118408203,
      "learning_rate": 3.077119907371942e-05,
      "loss": 0.19,
      "num_input_tokens_seen": 14760416,
      "step": 17035
    },
    {
      "epoch": 8.033946251768034,
      "grad_norm": 0.7762043476104736,
      "learning_rate": 3.076164632360666e-05,
      "loss": 0.2006,
      "num_input_tokens_seen": 14765184,
      "step": 17040
    },
    {
      "epoch": 8.036303630363037,
      "grad_norm": 1.4530055522918701,
      "learning_rate": 3.075209268497563e-05,
      "loss": 0.1741,
      "num_input_tokens_seen": 14769792,
      "step": 17045
    },
    {
      "epoch": 8.038661008958039,
      "grad_norm": 0.40593698620796204,
      "learning_rate": 3.074253815929961e-05,
      "loss": 0.1673,
      "num_input_tokens_seen": 14774560,
      "step": 17050
    },
    {
      "epoch": 8.041018387553041,
      "grad_norm": 0.8668432831764221,
      "learning_rate": 3.0732982748052054e-05,
      "loss": 0.1796,
      "num_input_tokens_seen": 14778400,
      "step": 17055
    },
    {
      "epoch": 8.043375766148044,
      "grad_norm": 0.8173043131828308,
      "learning_rate": 3.072342645270651e-05,
      "loss": 0.1805,
      "num_input_tokens_seen": 14782704,
      "step": 17060
    },
    {
      "epoch": 8.045733144743046,
      "grad_norm": 0.5831515192985535,
      "learning_rate": 3.071386927473668e-05,
      "loss": 0.1891,
      "num_input_tokens_seen": 14786688,
      "step": 17065
    },
    {
      "epoch": 8.048090523338049,
      "grad_norm": 1.0082874298095703,
      "learning_rate": 3.0704311215616404e-05,
      "loss": 0.1863,
      "num_input_tokens_seen": 14791104,
      "step": 17070
    },
    {
      "epoch": 8.050447901933051,
      "grad_norm": 0.6016269326210022,
      "learning_rate": 3.0694752276819656e-05,
      "loss": 0.1568,
      "num_input_tokens_seen": 14795872,
      "step": 17075
    },
    {
      "epoch": 8.052805280528053,
      "grad_norm": 1.434270977973938,
      "learning_rate": 3.068519245982054e-05,
      "loss": 0.1978,
      "num_input_tokens_seen": 14800368,
      "step": 17080
    },
    {
      "epoch": 8.055162659123056,
      "grad_norm": 0.7782105207443237,
      "learning_rate": 3.0675631766093304e-05,
      "loss": 0.1983,
      "num_input_tokens_seen": 14805408,
      "step": 17085
    },
    {
      "epoch": 8.057520037718058,
      "grad_norm": 1.5114779472351074,
      "learning_rate": 3.066607019711232e-05,
      "loss": 0.1969,
      "num_input_tokens_seen": 14809728,
      "step": 17090
    },
    {
      "epoch": 8.05987741631306,
      "grad_norm": 0.5999196171760559,
      "learning_rate": 3.065650775435211e-05,
      "loss": 0.2086,
      "num_input_tokens_seen": 14814016,
      "step": 17095
    },
    {
      "epoch": 8.062234794908063,
      "grad_norm": 0.873447835445404,
      "learning_rate": 3.0646944439287326e-05,
      "loss": 0.1416,
      "num_input_tokens_seen": 14818784,
      "step": 17100
    },
    {
      "epoch": 8.064592173503065,
      "grad_norm": 1.2750012874603271,
      "learning_rate": 3.0637380253392736e-05,
      "loss": 0.1799,
      "num_input_tokens_seen": 14824128,
      "step": 17105
    },
    {
      "epoch": 8.066949552098066,
      "grad_norm": 1.0846185684204102,
      "learning_rate": 3.062781519814327e-05,
      "loss": 0.2393,
      "num_input_tokens_seen": 14828256,
      "step": 17110
    },
    {
      "epoch": 8.069306930693068,
      "grad_norm": 1.2541429996490479,
      "learning_rate": 3.0618249275013985e-05,
      "loss": 0.1316,
      "num_input_tokens_seen": 14832656,
      "step": 17115
    },
    {
      "epoch": 8.07166430928807,
      "grad_norm": 0.5377731323242188,
      "learning_rate": 3.060868248548005e-05,
      "loss": 0.1957,
      "num_input_tokens_seen": 14836544,
      "step": 17120
    },
    {
      "epoch": 8.074021687883073,
      "grad_norm": 0.859316349029541,
      "learning_rate": 3.0599114831016796e-05,
      "loss": 0.2198,
      "num_input_tokens_seen": 14841632,
      "step": 17125
    },
    {
      "epoch": 8.076379066478076,
      "grad_norm": 0.9670841097831726,
      "learning_rate": 3.0589546313099666e-05,
      "loss": 0.2142,
      "num_input_tokens_seen": 14845632,
      "step": 17130
    },
    {
      "epoch": 8.078736445073078,
      "grad_norm": 0.6724987030029297,
      "learning_rate": 3.0579976933204255e-05,
      "loss": 0.1739,
      "num_input_tokens_seen": 14849824,
      "step": 17135
    },
    {
      "epoch": 8.08109382366808,
      "grad_norm": 0.9945793747901917,
      "learning_rate": 3.0570406692806284e-05,
      "loss": 0.2244,
      "num_input_tokens_seen": 14854416,
      "step": 17140
    },
    {
      "epoch": 8.083451202263083,
      "grad_norm": 1.282455563545227,
      "learning_rate": 3.05608355933816e-05,
      "loss": 0.1772,
      "num_input_tokens_seen": 14857728,
      "step": 17145
    },
    {
      "epoch": 8.085808580858085,
      "grad_norm": 1.393537163734436,
      "learning_rate": 3.055126363640618e-05,
      "loss": 0.1887,
      "num_input_tokens_seen": 14861856,
      "step": 17150
    },
    {
      "epoch": 8.088165959453088,
      "grad_norm": 0.9089736938476562,
      "learning_rate": 3.0541690823356146e-05,
      "loss": 0.1958,
      "num_input_tokens_seen": 14865280,
      "step": 17155
    },
    {
      "epoch": 8.09052333804809,
      "grad_norm": 0.643754780292511,
      "learning_rate": 3.053211715570775e-05,
      "loss": 0.1546,
      "num_input_tokens_seen": 14869936,
      "step": 17160
    },
    {
      "epoch": 8.092880716643092,
      "grad_norm": 0.8884620666503906,
      "learning_rate": 3.052254263493736e-05,
      "loss": 0.2049,
      "num_input_tokens_seen": 14873936,
      "step": 17165
    },
    {
      "epoch": 8.095238095238095,
      "grad_norm": 1.4881857633590698,
      "learning_rate": 3.0512967262521498e-05,
      "loss": 0.1775,
      "num_input_tokens_seen": 14878208,
      "step": 17170
    },
    {
      "epoch": 8.097595473833097,
      "grad_norm": 1.7575953006744385,
      "learning_rate": 3.0503391039936803e-05,
      "loss": 0.204,
      "num_input_tokens_seen": 14882672,
      "step": 17175
    },
    {
      "epoch": 8.0999528524281,
      "grad_norm": 0.4097791612148285,
      "learning_rate": 3.0493813968660056e-05,
      "loss": 0.1446,
      "num_input_tokens_seen": 14887632,
      "step": 17180
    },
    {
      "epoch": 8.102310231023102,
      "grad_norm": 1.2213598489761353,
      "learning_rate": 3.0484236050168153e-05,
      "loss": 0.1566,
      "num_input_tokens_seen": 14891664,
      "step": 17185
    },
    {
      "epoch": 8.104667609618105,
      "grad_norm": 0.898615300655365,
      "learning_rate": 3.0474657285938123e-05,
      "loss": 0.203,
      "num_input_tokens_seen": 14896256,
      "step": 17190
    },
    {
      "epoch": 8.107024988213107,
      "grad_norm": 0.6760168671607971,
      "learning_rate": 3.046507767744715e-05,
      "loss": 0.2171,
      "num_input_tokens_seen": 14900352,
      "step": 17195
    },
    {
      "epoch": 8.10938236680811,
      "grad_norm": 0.44241607189178467,
      "learning_rate": 3.045549722617252e-05,
      "loss": 0.2018,
      "num_input_tokens_seen": 14904544,
      "step": 17200
    },
    {
      "epoch": 8.10938236680811,
      "eval_loss": 0.20289117097854614,
      "eval_runtime": 22.1507,
      "eval_samples_per_second": 42.572,
      "eval_steps_per_second": 21.309,
      "num_input_tokens_seen": 14904544,
      "step": 17200
    },
    {
      "epoch": 8.111739745403112,
      "grad_norm": 1.075117588043213,
      "learning_rate": 3.0445915933591658e-05,
      "loss": 0.2582,
      "num_input_tokens_seen": 14908528,
      "step": 17205
    },
    {
      "epoch": 8.114097123998114,
      "grad_norm": 0.5238259434700012,
      "learning_rate": 3.0436333801182114e-05,
      "loss": 0.1411,
      "num_input_tokens_seen": 14912864,
      "step": 17210
    },
    {
      "epoch": 8.116454502593117,
      "grad_norm": 0.5124346613883972,
      "learning_rate": 3.0426750830421596e-05,
      "loss": 0.1074,
      "num_input_tokens_seen": 14917984,
      "step": 17215
    },
    {
      "epoch": 8.118811881188119,
      "grad_norm": 1.2262879610061646,
      "learning_rate": 3.0417167022787897e-05,
      "loss": 0.1916,
      "num_input_tokens_seen": 14921760,
      "step": 17220
    },
    {
      "epoch": 8.121169259783121,
      "grad_norm": 1.1416373252868652,
      "learning_rate": 3.0407582379758966e-05,
      "loss": 0.2014,
      "num_input_tokens_seen": 14926160,
      "step": 17225
    },
    {
      "epoch": 8.123526638378124,
      "grad_norm": 1.8849021196365356,
      "learning_rate": 3.039799690281287e-05,
      "loss": 0.2173,
      "num_input_tokens_seen": 14930368,
      "step": 17230
    },
    {
      "epoch": 8.125884016973126,
      "grad_norm": 0.8091434240341187,
      "learning_rate": 3.0388410593427823e-05,
      "loss": 0.1476,
      "num_input_tokens_seen": 14934672,
      "step": 17235
    },
    {
      "epoch": 8.128241395568129,
      "grad_norm": 0.7051036953926086,
      "learning_rate": 3.0378823453082146e-05,
      "loss": 0.1874,
      "num_input_tokens_seen": 14939696,
      "step": 17240
    },
    {
      "epoch": 8.130598774163131,
      "grad_norm": 0.8088411092758179,
      "learning_rate": 3.03692354832543e-05,
      "loss": 0.1325,
      "num_input_tokens_seen": 14943568,
      "step": 17245
    },
    {
      "epoch": 8.132956152758133,
      "grad_norm": 0.8542206883430481,
      "learning_rate": 3.0359646685422865e-05,
      "loss": 0.1507,
      "num_input_tokens_seen": 14948672,
      "step": 17250
    },
    {
      "epoch": 8.135313531353136,
      "grad_norm": 1.5837312936782837,
      "learning_rate": 3.035005706106656e-05,
      "loss": 0.186,
      "num_input_tokens_seen": 14953824,
      "step": 17255
    },
    {
      "epoch": 8.137670909948138,
      "grad_norm": 1.898747205734253,
      "learning_rate": 3.034046661166422e-05,
      "loss": 0.1578,
      "num_input_tokens_seen": 14957936,
      "step": 17260
    },
    {
      "epoch": 8.14002828854314,
      "grad_norm": 0.8853567242622375,
      "learning_rate": 3.033087533869482e-05,
      "loss": 0.1699,
      "num_input_tokens_seen": 14962224,
      "step": 17265
    },
    {
      "epoch": 8.142385667138143,
      "grad_norm": 0.5848296880722046,
      "learning_rate": 3.0321283243637444e-05,
      "loss": 0.114,
      "num_input_tokens_seen": 14966288,
      "step": 17270
    },
    {
      "epoch": 8.144743045733145,
      "grad_norm": 0.9436250329017639,
      "learning_rate": 3.0311690327971326e-05,
      "loss": 0.1351,
      "num_input_tokens_seen": 14971184,
      "step": 17275
    },
    {
      "epoch": 8.147100424328148,
      "grad_norm": 0.7565488219261169,
      "learning_rate": 3.030209659317581e-05,
      "loss": 0.1651,
      "num_input_tokens_seen": 14976112,
      "step": 17280
    },
    {
      "epoch": 8.14945780292315,
      "grad_norm": 0.6869747638702393,
      "learning_rate": 3.0292502040730362e-05,
      "loss": 0.1571,
      "num_input_tokens_seen": 14980448,
      "step": 17285
    },
    {
      "epoch": 8.151815181518153,
      "grad_norm": 2.285796642303467,
      "learning_rate": 3.0282906672114597e-05,
      "loss": 0.2629,
      "num_input_tokens_seen": 14985024,
      "step": 17290
    },
    {
      "epoch": 8.154172560113155,
      "grad_norm": 0.43825042247772217,
      "learning_rate": 3.027331048880823e-05,
      "loss": 0.2224,
      "num_input_tokens_seen": 14989680,
      "step": 17295
    },
    {
      "epoch": 8.156529938708157,
      "grad_norm": 0.47734296321868896,
      "learning_rate": 3.0263713492291123e-05,
      "loss": 0.2043,
      "num_input_tokens_seen": 14994000,
      "step": 17300
    },
    {
      "epoch": 8.15888731730316,
      "grad_norm": 1.259864330291748,
      "learning_rate": 3.0254115684043242e-05,
      "loss": 0.2772,
      "num_input_tokens_seen": 14998416,
      "step": 17305
    },
    {
      "epoch": 8.16124469589816,
      "grad_norm": 1.1284961700439453,
      "learning_rate": 3.024451706554469e-05,
      "loss": 0.2176,
      "num_input_tokens_seen": 15003776,
      "step": 17310
    },
    {
      "epoch": 8.163602074493163,
      "grad_norm": 1.1698641777038574,
      "learning_rate": 3.0234917638275705e-05,
      "loss": 0.2626,
      "num_input_tokens_seen": 15008224,
      "step": 17315
    },
    {
      "epoch": 8.165959453088165,
      "grad_norm": 1.4702388048171997,
      "learning_rate": 3.0225317403716635e-05,
      "loss": 0.2369,
      "num_input_tokens_seen": 15012528,
      "step": 17320
    },
    {
      "epoch": 8.168316831683168,
      "grad_norm": 1.5382450819015503,
      "learning_rate": 3.0215716363347956e-05,
      "loss": 0.1667,
      "num_input_tokens_seen": 15016656,
      "step": 17325
    },
    {
      "epoch": 8.17067421027817,
      "grad_norm": 1.0605064630508423,
      "learning_rate": 3.0206114518650275e-05,
      "loss": 0.1362,
      "num_input_tokens_seen": 15020752,
      "step": 17330
    },
    {
      "epoch": 8.173031588873172,
      "grad_norm": 1.143837809562683,
      "learning_rate": 3.0196511871104304e-05,
      "loss": 0.246,
      "num_input_tokens_seen": 15025216,
      "step": 17335
    },
    {
      "epoch": 8.175388967468175,
      "grad_norm": 0.6934006214141846,
      "learning_rate": 3.01869084221909e-05,
      "loss": 0.1059,
      "num_input_tokens_seen": 15029184,
      "step": 17340
    },
    {
      "epoch": 8.177746346063177,
      "grad_norm": 0.7784964442253113,
      "learning_rate": 3.0177304173391037e-05,
      "loss": 0.1892,
      "num_input_tokens_seen": 15033008,
      "step": 17345
    },
    {
      "epoch": 8.18010372465818,
      "grad_norm": 0.9929692149162292,
      "learning_rate": 3.01676991261858e-05,
      "loss": 0.1329,
      "num_input_tokens_seen": 15036560,
      "step": 17350
    },
    {
      "epoch": 8.182461103253182,
      "grad_norm": 0.4804447591304779,
      "learning_rate": 3.015809328205642e-05,
      "loss": 0.1557,
      "num_input_tokens_seen": 15040336,
      "step": 17355
    },
    {
      "epoch": 8.184818481848184,
      "grad_norm": 0.7575746774673462,
      "learning_rate": 3.0148486642484248e-05,
      "loss": 0.2257,
      "num_input_tokens_seen": 15044256,
      "step": 17360
    },
    {
      "epoch": 8.187175860443187,
      "grad_norm": 0.6691369414329529,
      "learning_rate": 3.0138879208950722e-05,
      "loss": 0.1512,
      "num_input_tokens_seen": 15047744,
      "step": 17365
    },
    {
      "epoch": 8.18953323903819,
      "grad_norm": 1.0862982273101807,
      "learning_rate": 3.012927098293744e-05,
      "loss": 0.2157,
      "num_input_tokens_seen": 15051616,
      "step": 17370
    },
    {
      "epoch": 8.191890617633192,
      "grad_norm": 1.625288724899292,
      "learning_rate": 3.0119661965926123e-05,
      "loss": 0.2899,
      "num_input_tokens_seen": 15055632,
      "step": 17375
    },
    {
      "epoch": 8.194247996228194,
      "grad_norm": 0.9437656402587891,
      "learning_rate": 3.0110052159398587e-05,
      "loss": 0.1565,
      "num_input_tokens_seen": 15059536,
      "step": 17380
    },
    {
      "epoch": 8.196605374823196,
      "grad_norm": 0.8290880918502808,
      "learning_rate": 3.0100441564836802e-05,
      "loss": 0.1596,
      "num_input_tokens_seen": 15064336,
      "step": 17385
    },
    {
      "epoch": 8.198962753418199,
      "grad_norm": 0.6439509987831116,
      "learning_rate": 3.0090830183722817e-05,
      "loss": 0.1637,
      "num_input_tokens_seen": 15069456,
      "step": 17390
    },
    {
      "epoch": 8.201320132013201,
      "grad_norm": 0.7642310261726379,
      "learning_rate": 3.0081218017538852e-05,
      "loss": 0.1528,
      "num_input_tokens_seen": 15074416,
      "step": 17395
    },
    {
      "epoch": 8.203677510608204,
      "grad_norm": 1.7038936614990234,
      "learning_rate": 3.0071605067767212e-05,
      "loss": 0.133,
      "num_input_tokens_seen": 15078832,
      "step": 17400
    },
    {
      "epoch": 8.203677510608204,
      "eval_loss": 0.20304836332798004,
      "eval_runtime": 22.1934,
      "eval_samples_per_second": 42.49,
      "eval_steps_per_second": 21.268,
      "num_input_tokens_seen": 15078832,
      "step": 17400
    },
    {
      "epoch": 8.206034889203206,
      "grad_norm": 0.5496950149536133,
      "learning_rate": 3.006199133589034e-05,
      "loss": 0.1889,
      "num_input_tokens_seen": 15083120,
      "step": 17405
    },
    {
      "epoch": 8.208392267798208,
      "grad_norm": 0.7335545420646667,
      "learning_rate": 3.005237682339079e-05,
      "loss": 0.2153,
      "num_input_tokens_seen": 15088272,
      "step": 17410
    },
    {
      "epoch": 8.21074964639321,
      "grad_norm": 0.8774899244308472,
      "learning_rate": 3.0042761531751228e-05,
      "loss": 0.1154,
      "num_input_tokens_seen": 15092080,
      "step": 17415
    },
    {
      "epoch": 8.213107024988213,
      "grad_norm": 0.4540136158466339,
      "learning_rate": 3.0033145462454482e-05,
      "loss": 0.1199,
      "num_input_tokens_seen": 15095952,
      "step": 17420
    },
    {
      "epoch": 8.215464403583216,
      "grad_norm": 1.4492688179016113,
      "learning_rate": 3.002352861698345e-05,
      "loss": 0.2056,
      "num_input_tokens_seen": 15100128,
      "step": 17425
    },
    {
      "epoch": 8.217821782178218,
      "grad_norm": 0.7364781498908997,
      "learning_rate": 3.0013910996821178e-05,
      "loss": 0.1674,
      "num_input_tokens_seen": 15104416,
      "step": 17430
    },
    {
      "epoch": 8.22017916077322,
      "grad_norm": 0.6891986131668091,
      "learning_rate": 3.0004292603450817e-05,
      "loss": 0.2311,
      "num_input_tokens_seen": 15108592,
      "step": 17435
    },
    {
      "epoch": 8.222536539368223,
      "grad_norm": 0.6505154967308044,
      "learning_rate": 2.9994673438355653e-05,
      "loss": 0.1089,
      "num_input_tokens_seen": 15112608,
      "step": 17440
    },
    {
      "epoch": 8.224893917963225,
      "grad_norm": 1.557337760925293,
      "learning_rate": 2.9985053503019078e-05,
      "loss": 0.216,
      "num_input_tokens_seen": 15116608,
      "step": 17445
    },
    {
      "epoch": 8.227251296558228,
      "grad_norm": 1.5661083459854126,
      "learning_rate": 2.99754327989246e-05,
      "loss": 0.2373,
      "num_input_tokens_seen": 15120064,
      "step": 17450
    },
    {
      "epoch": 8.22960867515323,
      "grad_norm": 0.8284567594528198,
      "learning_rate": 2.9965811327555864e-05,
      "loss": 0.2381,
      "num_input_tokens_seen": 15124608,
      "step": 17455
    },
    {
      "epoch": 8.231966053748232,
      "grad_norm": 0.5305343270301819,
      "learning_rate": 2.995618909039662e-05,
      "loss": 0.196,
      "num_input_tokens_seen": 15129376,
      "step": 17460
    },
    {
      "epoch": 8.234323432343235,
      "grad_norm": 1.2824877500534058,
      "learning_rate": 2.9946566088930727e-05,
      "loss": 0.2578,
      "num_input_tokens_seen": 15133936,
      "step": 17465
    },
    {
      "epoch": 8.236680810938237,
      "grad_norm": 0.7278766632080078,
      "learning_rate": 2.9936942324642192e-05,
      "loss": 0.2154,
      "num_input_tokens_seen": 15138272,
      "step": 17470
    },
    {
      "epoch": 8.23903818953324,
      "grad_norm": 1.1401786804199219,
      "learning_rate": 2.9927317799015097e-05,
      "loss": 0.2013,
      "num_input_tokens_seen": 15143056,
      "step": 17475
    },
    {
      "epoch": 8.241395568128242,
      "grad_norm": 0.4473118185997009,
      "learning_rate": 2.9917692513533685e-05,
      "loss": 0.1518,
      "num_input_tokens_seen": 15147168,
      "step": 17480
    },
    {
      "epoch": 8.243752946723244,
      "grad_norm": 0.3677109181880951,
      "learning_rate": 2.990806646968229e-05,
      "loss": 0.1747,
      "num_input_tokens_seen": 15151392,
      "step": 17485
    },
    {
      "epoch": 8.246110325318247,
      "grad_norm": 0.3479693531990051,
      "learning_rate": 2.989843966894536e-05,
      "loss": 0.189,
      "num_input_tokens_seen": 15156160,
      "step": 17490
    },
    {
      "epoch": 8.24846770391325,
      "grad_norm": 1.9503811597824097,
      "learning_rate": 2.9888812112807472e-05,
      "loss": 0.1919,
      "num_input_tokens_seen": 15159856,
      "step": 17495
    },
    {
      "epoch": 8.250825082508252,
      "grad_norm": 0.969740629196167,
      "learning_rate": 2.987918380275333e-05,
      "loss": 0.183,
      "num_input_tokens_seen": 15163392,
      "step": 17500
    },
    {
      "epoch": 8.253182461103254,
      "grad_norm": 2.65802001953125,
      "learning_rate": 2.9869554740267724e-05,
      "loss": 0.2135,
      "num_input_tokens_seen": 15167472,
      "step": 17505
    },
    {
      "epoch": 8.255539839698255,
      "grad_norm": 0.7670469284057617,
      "learning_rate": 2.9859924926835585e-05,
      "loss": 0.2284,
      "num_input_tokens_seen": 15171424,
      "step": 17510
    },
    {
      "epoch": 8.257897218293257,
      "grad_norm": 0.7017144560813904,
      "learning_rate": 2.9850294363941944e-05,
      "loss": 0.1925,
      "num_input_tokens_seen": 15177968,
      "step": 17515
    },
    {
      "epoch": 8.26025459688826,
      "grad_norm": 1.226340651512146,
      "learning_rate": 2.9840663053071967e-05,
      "loss": 0.2379,
      "num_input_tokens_seen": 15182432,
      "step": 17520
    },
    {
      "epoch": 8.262611975483262,
      "grad_norm": 0.7728093266487122,
      "learning_rate": 2.983103099571091e-05,
      "loss": 0.1584,
      "num_input_tokens_seen": 15186352,
      "step": 17525
    },
    {
      "epoch": 8.264969354078264,
      "grad_norm": 1.3296812772750854,
      "learning_rate": 2.9821398193344164e-05,
      "loss": 0.1725,
      "num_input_tokens_seen": 15191744,
      "step": 17530
    },
    {
      "epoch": 8.267326732673267,
      "grad_norm": 1.62741219997406,
      "learning_rate": 2.9811764647457226e-05,
      "loss": 0.1618,
      "num_input_tokens_seen": 15196224,
      "step": 17535
    },
    {
      "epoch": 8.269684111268269,
      "grad_norm": 1.0081019401550293,
      "learning_rate": 2.9802130359535714e-05,
      "loss": 0.1926,
      "num_input_tokens_seen": 15201008,
      "step": 17540
    },
    {
      "epoch": 8.272041489863271,
      "grad_norm": 1.2766075134277344,
      "learning_rate": 2.979249533106535e-05,
      "loss": 0.2632,
      "num_input_tokens_seen": 15205424,
      "step": 17545
    },
    {
      "epoch": 8.274398868458274,
      "grad_norm": 0.40695980191230774,
      "learning_rate": 2.9782859563531986e-05,
      "loss": 0.1877,
      "num_input_tokens_seen": 15209504,
      "step": 17550
    },
    {
      "epoch": 8.276756247053276,
      "grad_norm": 1.1683703660964966,
      "learning_rate": 2.977322305842156e-05,
      "loss": 0.1623,
      "num_input_tokens_seen": 15214256,
      "step": 17555
    },
    {
      "epoch": 8.279113625648279,
      "grad_norm": 1.6883426904678345,
      "learning_rate": 2.9763585817220162e-05,
      "loss": 0.238,
      "num_input_tokens_seen": 15218080,
      "step": 17560
    },
    {
      "epoch": 8.281471004243281,
      "grad_norm": 0.9799091815948486,
      "learning_rate": 2.975394784141397e-05,
      "loss": 0.2346,
      "num_input_tokens_seen": 15221856,
      "step": 17565
    },
    {
      "epoch": 8.283828382838283,
      "grad_norm": 0.9040976762771606,
      "learning_rate": 2.974430913248928e-05,
      "loss": 0.168,
      "num_input_tokens_seen": 15226192,
      "step": 17570
    },
    {
      "epoch": 8.286185761433286,
      "grad_norm": 0.5641459822654724,
      "learning_rate": 2.9734669691932497e-05,
      "loss": 0.1492,
      "num_input_tokens_seen": 15230912,
      "step": 17575
    },
    {
      "epoch": 8.288543140028288,
      "grad_norm": 1.5130051374435425,
      "learning_rate": 2.9725029521230147e-05,
      "loss": 0.1681,
      "num_input_tokens_seen": 15236560,
      "step": 17580
    },
    {
      "epoch": 8.29090051862329,
      "grad_norm": 0.7622697353363037,
      "learning_rate": 2.9715388621868873e-05,
      "loss": 0.2147,
      "num_input_tokens_seen": 15241456,
      "step": 17585
    },
    {
      "epoch": 8.293257897218293,
      "grad_norm": 1.3790111541748047,
      "learning_rate": 2.970574699533541e-05,
      "loss": 0.2188,
      "num_input_tokens_seen": 15245744,
      "step": 17590
    },
    {
      "epoch": 8.295615275813295,
      "grad_norm": 1.127516746520996,
      "learning_rate": 2.969610464311662e-05,
      "loss": 0.2049,
      "num_input_tokens_seen": 15249824,
      "step": 17595
    },
    {
      "epoch": 8.297972654408298,
      "grad_norm": 2.6127126216888428,
      "learning_rate": 2.9686461566699487e-05,
      "loss": 0.2306,
      "num_input_tokens_seen": 15254544,
      "step": 17600
    },
    {
      "epoch": 8.297972654408298,
      "eval_loss": 0.20088715851306915,
      "eval_runtime": 22.1351,
      "eval_samples_per_second": 42.602,
      "eval_steps_per_second": 21.324,
      "num_input_tokens_seen": 15254544,
      "step": 17600
    },
    {
      "epoch": 8.3003300330033,
      "grad_norm": 1.8281309604644775,
      "learning_rate": 2.9676817767571086e-05,
      "loss": 0.159,
      "num_input_tokens_seen": 15258352,
      "step": 17605
    },
    {
      "epoch": 8.302687411598303,
      "grad_norm": 0.5101170539855957,
      "learning_rate": 2.966717324721861e-05,
      "loss": 0.1466,
      "num_input_tokens_seen": 15261856,
      "step": 17610
    },
    {
      "epoch": 8.305044790193305,
      "grad_norm": 0.991374135017395,
      "learning_rate": 2.9657528007129366e-05,
      "loss": 0.2616,
      "num_input_tokens_seen": 15266240,
      "step": 17615
    },
    {
      "epoch": 8.307402168788308,
      "grad_norm": 0.4144097566604614,
      "learning_rate": 2.9647882048790777e-05,
      "loss": 0.1633,
      "num_input_tokens_seen": 15270000,
      "step": 17620
    },
    {
      "epoch": 8.30975954738331,
      "grad_norm": 0.585355281829834,
      "learning_rate": 2.963823537369037e-05,
      "loss": 0.1354,
      "num_input_tokens_seen": 15274704,
      "step": 17625
    },
    {
      "epoch": 8.312116925978312,
      "grad_norm": 0.4778047502040863,
      "learning_rate": 2.9628587983315775e-05,
      "loss": 0.1877,
      "num_input_tokens_seen": 15278448,
      "step": 17630
    },
    {
      "epoch": 8.314474304573315,
      "grad_norm": 1.5049360990524292,
      "learning_rate": 2.9618939879154746e-05,
      "loss": 0.1767,
      "num_input_tokens_seen": 15282384,
      "step": 17635
    },
    {
      "epoch": 8.316831683168317,
      "grad_norm": 1.0002375841140747,
      "learning_rate": 2.9609291062695143e-05,
      "loss": 0.1791,
      "num_input_tokens_seen": 15286544,
      "step": 17640
    },
    {
      "epoch": 8.31918906176332,
      "grad_norm": 0.7014461159706116,
      "learning_rate": 2.9599641535424938e-05,
      "loss": 0.1462,
      "num_input_tokens_seen": 15290128,
      "step": 17645
    },
    {
      "epoch": 8.321546440358322,
      "grad_norm": 1.0871398448944092,
      "learning_rate": 2.9589991298832202e-05,
      "loss": 0.2038,
      "num_input_tokens_seen": 15295424,
      "step": 17650
    },
    {
      "epoch": 8.323903818953324,
      "grad_norm": 1.3547879457473755,
      "learning_rate": 2.958034035440513e-05,
      "loss": 0.1862,
      "num_input_tokens_seen": 15299344,
      "step": 17655
    },
    {
      "epoch": 8.326261197548327,
      "grad_norm": 0.33671319484710693,
      "learning_rate": 2.957068870363201e-05,
      "loss": 0.112,
      "num_input_tokens_seen": 15303872,
      "step": 17660
    },
    {
      "epoch": 8.32861857614333,
      "grad_norm": 0.7873387932777405,
      "learning_rate": 2.956103634800126e-05,
      "loss": 0.2507,
      "num_input_tokens_seen": 15307824,
      "step": 17665
    },
    {
      "epoch": 8.330975954738332,
      "grad_norm": 0.39468270540237427,
      "learning_rate": 2.9551383289001384e-05,
      "loss": 0.2732,
      "num_input_tokens_seen": 15311776,
      "step": 17670
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 1.3582651615142822,
      "learning_rate": 2.9541729528121005e-05,
      "loss": 0.2319,
      "num_input_tokens_seen": 15316752,
      "step": 17675
    },
    {
      "epoch": 8.335690711928336,
      "grad_norm": 1.1725274324417114,
      "learning_rate": 2.9532075066848856e-05,
      "loss": 0.1644,
      "num_input_tokens_seen": 15320512,
      "step": 17680
    },
    {
      "epoch": 8.338048090523339,
      "grad_norm": 0.6238380670547485,
      "learning_rate": 2.9522419906673786e-05,
      "loss": 0.1934,
      "num_input_tokens_seen": 15324512,
      "step": 17685
    },
    {
      "epoch": 8.340405469118341,
      "grad_norm": 0.6087464094161987,
      "learning_rate": 2.951276404908474e-05,
      "loss": 0.1775,
      "num_input_tokens_seen": 15329552,
      "step": 17690
    },
    {
      "epoch": 8.342762847713344,
      "grad_norm": 0.6656209826469421,
      "learning_rate": 2.9503107495570752e-05,
      "loss": 0.123,
      "num_input_tokens_seen": 15333744,
      "step": 17695
    },
    {
      "epoch": 8.345120226308346,
      "grad_norm": 2.1558165550231934,
      "learning_rate": 2.9493450247621003e-05,
      "loss": 0.1555,
      "num_input_tokens_seen": 15338000,
      "step": 17700
    },
    {
      "epoch": 8.347477604903348,
      "grad_norm": 0.9235938191413879,
      "learning_rate": 2.948379230672476e-05,
      "loss": 0.1873,
      "num_input_tokens_seen": 15341776,
      "step": 17705
    },
    {
      "epoch": 8.34983498349835,
      "grad_norm": 1.978780746459961,
      "learning_rate": 2.9474133674371396e-05,
      "loss": 0.2226,
      "num_input_tokens_seen": 15346016,
      "step": 17710
    },
    {
      "epoch": 8.352192362093351,
      "grad_norm": 0.8591864705085754,
      "learning_rate": 2.9464474352050387e-05,
      "loss": 0.2077,
      "num_input_tokens_seen": 15350016,
      "step": 17715
    },
    {
      "epoch": 8.354549740688354,
      "grad_norm": 0.3588745594024658,
      "learning_rate": 2.9454814341251336e-05,
      "loss": 0.1792,
      "num_input_tokens_seen": 15353536,
      "step": 17720
    },
    {
      "epoch": 8.356907119283356,
      "grad_norm": 0.7183419466018677,
      "learning_rate": 2.9445153643463942e-05,
      "loss": 0.1799,
      "num_input_tokens_seen": 15357696,
      "step": 17725
    },
    {
      "epoch": 8.359264497878359,
      "grad_norm": 0.8558833003044128,
      "learning_rate": 2.943549226017798e-05,
      "loss": 0.1652,
      "num_input_tokens_seen": 15361936,
      "step": 17730
    },
    {
      "epoch": 8.361621876473361,
      "grad_norm": 2.5269863605499268,
      "learning_rate": 2.942583019288337e-05,
      "loss": 0.2037,
      "num_input_tokens_seen": 15366000,
      "step": 17735
    },
    {
      "epoch": 8.363979255068363,
      "grad_norm": 0.6199498772621155,
      "learning_rate": 2.9416167443070132e-05,
      "loss": 0.2124,
      "num_input_tokens_seen": 15369792,
      "step": 17740
    },
    {
      "epoch": 8.366336633663366,
      "grad_norm": 0.5282654166221619,
      "learning_rate": 2.9406504012228375e-05,
      "loss": 0.1841,
      "num_input_tokens_seen": 15374800,
      "step": 17745
    },
    {
      "epoch": 8.368694012258368,
      "grad_norm": 0.826917290687561,
      "learning_rate": 2.939683990184832e-05,
      "loss": 0.2018,
      "num_input_tokens_seen": 15378976,
      "step": 17750
    },
    {
      "epoch": 8.37105139085337,
      "grad_norm": 1.4022821187973022,
      "learning_rate": 2.93871751134203e-05,
      "loss": 0.1647,
      "num_input_tokens_seen": 15383968,
      "step": 17755
    },
    {
      "epoch": 8.373408769448373,
      "grad_norm": 0.47147098183631897,
      "learning_rate": 2.9377509648434752e-05,
      "loss": 0.2508,
      "num_input_tokens_seen": 15387776,
      "step": 17760
    },
    {
      "epoch": 8.375766148043375,
      "grad_norm": 0.37032631039619446,
      "learning_rate": 2.9367843508382203e-05,
      "loss": 0.1807,
      "num_input_tokens_seen": 15392624,
      "step": 17765
    },
    {
      "epoch": 8.378123526638378,
      "grad_norm": 1.1369540691375732,
      "learning_rate": 2.9358176694753293e-05,
      "loss": 0.0946,
      "num_input_tokens_seen": 15396768,
      "step": 17770
    },
    {
      "epoch": 8.38048090523338,
      "grad_norm": 0.4903259575366974,
      "learning_rate": 2.9348509209038766e-05,
      "loss": 0.157,
      "num_input_tokens_seen": 15400592,
      "step": 17775
    },
    {
      "epoch": 8.382838283828383,
      "grad_norm": 0.3152092695236206,
      "learning_rate": 2.933884105272947e-05,
      "loss": 0.1704,
      "num_input_tokens_seen": 15405136,
      "step": 17780
    },
    {
      "epoch": 8.385195662423385,
      "grad_norm": 1.9078739881515503,
      "learning_rate": 2.9329172227316366e-05,
      "loss": 0.2016,
      "num_input_tokens_seen": 15409408,
      "step": 17785
    },
    {
      "epoch": 8.387553041018387,
      "grad_norm": 1.0585578680038452,
      "learning_rate": 2.93195027342905e-05,
      "loss": 0.2257,
      "num_input_tokens_seen": 15414112,
      "step": 17790
    },
    {
      "epoch": 8.38991041961339,
      "grad_norm": 1.865581750869751,
      "learning_rate": 2.9309832575143024e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 15418160,
      "step": 17795
    },
    {
      "epoch": 8.392267798208392,
      "grad_norm": 1.626832365989685,
      "learning_rate": 2.930016175136521e-05,
      "loss": 0.1876,
      "num_input_tokens_seen": 15422256,
      "step": 17800
    },
    {
      "epoch": 8.392267798208392,
      "eval_loss": 0.20298327505588531,
      "eval_runtime": 22.1852,
      "eval_samples_per_second": 42.506,
      "eval_steps_per_second": 21.275,
      "num_input_tokens_seen": 15422256,
      "step": 17800
    },
    {
      "epoch": 8.394625176803395,
      "grad_norm": 0.5056160688400269,
      "learning_rate": 2.9290490264448412e-05,
      "loss": 0.1238,
      "num_input_tokens_seen": 15426528,
      "step": 17805
    },
    {
      "epoch": 8.396982555398397,
      "grad_norm": 0.410060316324234,
      "learning_rate": 2.9280818115884094e-05,
      "loss": 0.1866,
      "num_input_tokens_seen": 15430704,
      "step": 17810
    },
    {
      "epoch": 8.3993399339934,
      "grad_norm": 0.6098925471305847,
      "learning_rate": 2.9271145307163828e-05,
      "loss": 0.2332,
      "num_input_tokens_seen": 15434880,
      "step": 17815
    },
    {
      "epoch": 8.401697312588402,
      "grad_norm": 0.636286199092865,
      "learning_rate": 2.9261471839779287e-05,
      "loss": 0.2121,
      "num_input_tokens_seen": 15438992,
      "step": 17820
    },
    {
      "epoch": 8.404054691183404,
      "grad_norm": 0.3373870551586151,
      "learning_rate": 2.925179771522223e-05,
      "loss": 0.1387,
      "num_input_tokens_seen": 15443600,
      "step": 17825
    },
    {
      "epoch": 8.406412069778407,
      "grad_norm": 1.199813961982727,
      "learning_rate": 2.9242122934984535e-05,
      "loss": 0.1923,
      "num_input_tokens_seen": 15448688,
      "step": 17830
    },
    {
      "epoch": 8.408769448373409,
      "grad_norm": 0.7770894169807434,
      "learning_rate": 2.9232447500558176e-05,
      "loss": 0.1419,
      "num_input_tokens_seen": 15452880,
      "step": 17835
    },
    {
      "epoch": 8.411126826968411,
      "grad_norm": 0.9847220182418823,
      "learning_rate": 2.9222771413435225e-05,
      "loss": 0.2567,
      "num_input_tokens_seen": 15457488,
      "step": 17840
    },
    {
      "epoch": 8.413484205563414,
      "grad_norm": 1.3947902917861938,
      "learning_rate": 2.9213094675107848e-05,
      "loss": 0.1608,
      "num_input_tokens_seen": 15461520,
      "step": 17845
    },
    {
      "epoch": 8.415841584158416,
      "grad_norm": 0.8175863027572632,
      "learning_rate": 2.9203417287068335e-05,
      "loss": 0.1425,
      "num_input_tokens_seen": 15465824,
      "step": 17850
    },
    {
      "epoch": 8.418198962753419,
      "grad_norm": 0.6988744735717773,
      "learning_rate": 2.9193739250809042e-05,
      "loss": 0.1366,
      "num_input_tokens_seen": 15470368,
      "step": 17855
    },
    {
      "epoch": 8.420556341348421,
      "grad_norm": 0.6030047535896301,
      "learning_rate": 2.9184060567822463e-05,
      "loss": 0.2297,
      "num_input_tokens_seen": 15475168,
      "step": 17860
    },
    {
      "epoch": 8.422913719943423,
      "grad_norm": 0.5226079821586609,
      "learning_rate": 2.9174381239601166e-05,
      "loss": 0.2464,
      "num_input_tokens_seen": 15479056,
      "step": 17865
    },
    {
      "epoch": 8.425271098538426,
      "grad_norm": 0.5397506952285767,
      "learning_rate": 2.916470126763783e-05,
      "loss": 0.2158,
      "num_input_tokens_seen": 15484176,
      "step": 17870
    },
    {
      "epoch": 8.427628477133428,
      "grad_norm": 1.103083848953247,
      "learning_rate": 2.9155020653425203e-05,
      "loss": 0.2018,
      "num_input_tokens_seen": 15487952,
      "step": 17875
    },
    {
      "epoch": 8.42998585572843,
      "grad_norm": 0.4681907594203949,
      "learning_rate": 2.9145339398456184e-05,
      "loss": 0.1414,
      "num_input_tokens_seen": 15492432,
      "step": 17880
    },
    {
      "epoch": 8.432343234323433,
      "grad_norm": 2.1757595539093018,
      "learning_rate": 2.913565750422374e-05,
      "loss": 0.1721,
      "num_input_tokens_seen": 15497120,
      "step": 17885
    },
    {
      "epoch": 8.434700612918435,
      "grad_norm": 0.43554478883743286,
      "learning_rate": 2.9125974972220938e-05,
      "loss": 0.1761,
      "num_input_tokens_seen": 15502160,
      "step": 17890
    },
    {
      "epoch": 8.437057991513438,
      "grad_norm": 0.654890775680542,
      "learning_rate": 2.9116291803940932e-05,
      "loss": 0.1624,
      "num_input_tokens_seen": 15505920,
      "step": 17895
    },
    {
      "epoch": 8.43941537010844,
      "grad_norm": 1.4472836256027222,
      "learning_rate": 2.910660800087701e-05,
      "loss": 0.1546,
      "num_input_tokens_seen": 15510432,
      "step": 17900
    },
    {
      "epoch": 8.441772748703443,
      "grad_norm": 0.8305056691169739,
      "learning_rate": 2.909692356452254e-05,
      "loss": 0.151,
      "num_input_tokens_seen": 15514688,
      "step": 17905
    },
    {
      "epoch": 8.444130127298443,
      "grad_norm": 0.5347093343734741,
      "learning_rate": 2.9087238496370962e-05,
      "loss": 0.1374,
      "num_input_tokens_seen": 15519424,
      "step": 17910
    },
    {
      "epoch": 8.446487505893446,
      "grad_norm": 0.5683633089065552,
      "learning_rate": 2.907755279791583e-05,
      "loss": 0.1083,
      "num_input_tokens_seen": 15522912,
      "step": 17915
    },
    {
      "epoch": 8.448844884488448,
      "grad_norm": 0.5551364421844482,
      "learning_rate": 2.906786647065083e-05,
      "loss": 0.1879,
      "num_input_tokens_seen": 15526928,
      "step": 17920
    },
    {
      "epoch": 8.45120226308345,
      "grad_norm": 1.5419082641601562,
      "learning_rate": 2.9058179516069695e-05,
      "loss": 0.1177,
      "num_input_tokens_seen": 15530944,
      "step": 17925
    },
    {
      "epoch": 8.453559641678453,
      "grad_norm": 1.6764116287231445,
      "learning_rate": 2.9048491935666282e-05,
      "loss": 0.1541,
      "num_input_tokens_seen": 15535408,
      "step": 17930
    },
    {
      "epoch": 8.455917020273455,
      "grad_norm": 0.6956201791763306,
      "learning_rate": 2.9038803730934534e-05,
      "loss": 0.1094,
      "num_input_tokens_seen": 15540544,
      "step": 17935
    },
    {
      "epoch": 8.458274398868458,
      "grad_norm": 0.9854612946510315,
      "learning_rate": 2.9029114903368503e-05,
      "loss": 0.2374,
      "num_input_tokens_seen": 15544112,
      "step": 17940
    },
    {
      "epoch": 8.46063177746346,
      "grad_norm": 0.5393016934394836,
      "learning_rate": 2.9019425454462318e-05,
      "loss": 0.1047,
      "num_input_tokens_seen": 15548560,
      "step": 17945
    },
    {
      "epoch": 8.462989156058462,
      "grad_norm": 1.0563814640045166,
      "learning_rate": 2.9009735385710212e-05,
      "loss": 0.1059,
      "num_input_tokens_seen": 15553920,
      "step": 17950
    },
    {
      "epoch": 8.465346534653465,
      "grad_norm": 1.6286391019821167,
      "learning_rate": 2.900004469860652e-05,
      "loss": 0.2166,
      "num_input_tokens_seen": 15557712,
      "step": 17955
    },
    {
      "epoch": 8.467703913248467,
      "grad_norm": 1.3814489841461182,
      "learning_rate": 2.8990353394645668e-05,
      "loss": 0.2468,
      "num_input_tokens_seen": 15562544,
      "step": 17960
    },
    {
      "epoch": 8.47006129184347,
      "grad_norm": 1.0230932235717773,
      "learning_rate": 2.8980661475322186e-05,
      "loss": 0.174,
      "num_input_tokens_seen": 15566416,
      "step": 17965
    },
    {
      "epoch": 8.472418670438472,
      "grad_norm": 1.456763744354248,
      "learning_rate": 2.897096894213067e-05,
      "loss": 0.2096,
      "num_input_tokens_seen": 15571792,
      "step": 17970
    },
    {
      "epoch": 8.474776049033474,
      "grad_norm": 2.029188394546509,
      "learning_rate": 2.8961275796565845e-05,
      "loss": 0.1468,
      "num_input_tokens_seen": 15575968,
      "step": 17975
    },
    {
      "epoch": 8.477133427628477,
      "grad_norm": 1.115424394607544,
      "learning_rate": 2.8951582040122517e-05,
      "loss": 0.1832,
      "num_input_tokens_seen": 15579296,
      "step": 17980
    },
    {
      "epoch": 8.47949080622348,
      "grad_norm": 0.9547447562217712,
      "learning_rate": 2.894188767429557e-05,
      "loss": 0.1529,
      "num_input_tokens_seen": 15583408,
      "step": 17985
    },
    {
      "epoch": 8.481848184818482,
      "grad_norm": 1.5262349843978882,
      "learning_rate": 2.8932192700580014e-05,
      "loss": 0.1716,
      "num_input_tokens_seen": 15587680,
      "step": 17990
    },
    {
      "epoch": 8.484205563413484,
      "grad_norm": 0.9757899641990662,
      "learning_rate": 2.8922497120470916e-05,
      "loss": 0.1342,
      "num_input_tokens_seen": 15591344,
      "step": 17995
    },
    {
      "epoch": 8.486562942008486,
      "grad_norm": 1.5437380075454712,
      "learning_rate": 2.891280093546348e-05,
      "loss": 0.181,
      "num_input_tokens_seen": 15595776,
      "step": 18000
    },
    {
      "epoch": 8.486562942008486,
      "eval_loss": 0.20542581379413605,
      "eval_runtime": 22.1627,
      "eval_samples_per_second": 42.549,
      "eval_steps_per_second": 21.297,
      "num_input_tokens_seen": 15595776,
      "step": 18000
    },
    {
      "epoch": 8.488920320603489,
      "grad_norm": 1.0863544940948486,
      "learning_rate": 2.890310414705297e-05,
      "loss": 0.2157,
      "num_input_tokens_seen": 15600048,
      "step": 18005
    },
    {
      "epoch": 8.491277699198491,
      "grad_norm": 1.233013391494751,
      "learning_rate": 2.8893406756734742e-05,
      "loss": 0.1698,
      "num_input_tokens_seen": 15604464,
      "step": 18010
    },
    {
      "epoch": 8.493635077793494,
      "grad_norm": 0.8571586608886719,
      "learning_rate": 2.888370876600427e-05,
      "loss": 0.2777,
      "num_input_tokens_seen": 15608880,
      "step": 18015
    },
    {
      "epoch": 8.495992456388496,
      "grad_norm": 1.2673355340957642,
      "learning_rate": 2.8874010176357104e-05,
      "loss": 0.1616,
      "num_input_tokens_seen": 15613184,
      "step": 18020
    },
    {
      "epoch": 8.498349834983498,
      "grad_norm": 1.017629861831665,
      "learning_rate": 2.886431098928888e-05,
      "loss": 0.168,
      "num_input_tokens_seen": 15616768,
      "step": 18025
    },
    {
      "epoch": 8.500707213578501,
      "grad_norm": 0.9772809147834778,
      "learning_rate": 2.885461120629534e-05,
      "loss": 0.1267,
      "num_input_tokens_seen": 15620656,
      "step": 18030
    },
    {
      "epoch": 8.503064592173503,
      "grad_norm": 0.988387405872345,
      "learning_rate": 2.8844910828872317e-05,
      "loss": 0.1541,
      "num_input_tokens_seen": 15624624,
      "step": 18035
    },
    {
      "epoch": 8.505421970768506,
      "grad_norm": 1.0323506593704224,
      "learning_rate": 2.8835209858515715e-05,
      "loss": 0.1455,
      "num_input_tokens_seen": 15629008,
      "step": 18040
    },
    {
      "epoch": 8.507779349363508,
      "grad_norm": 1.6209731101989746,
      "learning_rate": 2.8825508296721566e-05,
      "loss": 0.2326,
      "num_input_tokens_seen": 15633984,
      "step": 18045
    },
    {
      "epoch": 8.51013672795851,
      "grad_norm": 1.1527318954467773,
      "learning_rate": 2.881580614498596e-05,
      "loss": 0.2552,
      "num_input_tokens_seen": 15638064,
      "step": 18050
    },
    {
      "epoch": 8.512494106553513,
      "grad_norm": 1.1110332012176514,
      "learning_rate": 2.8806103404805103e-05,
      "loss": 0.1887,
      "num_input_tokens_seen": 15642080,
      "step": 18055
    },
    {
      "epoch": 8.514851485148515,
      "grad_norm": 0.9719645977020264,
      "learning_rate": 2.8796400077675257e-05,
      "loss": 0.1546,
      "num_input_tokens_seen": 15647120,
      "step": 18060
    },
    {
      "epoch": 8.517208863743518,
      "grad_norm": 1.7694967985153198,
      "learning_rate": 2.8786696165092812e-05,
      "loss": 0.1726,
      "num_input_tokens_seen": 15651296,
      "step": 18065
    },
    {
      "epoch": 8.51956624233852,
      "grad_norm": 0.6935361623764038,
      "learning_rate": 2.8776991668554236e-05,
      "loss": 0.186,
      "num_input_tokens_seen": 15655088,
      "step": 18070
    },
    {
      "epoch": 8.521923620933523,
      "grad_norm": 0.40298983454704285,
      "learning_rate": 2.876728658955608e-05,
      "loss": 0.2313,
      "num_input_tokens_seen": 15659600,
      "step": 18075
    },
    {
      "epoch": 8.524280999528525,
      "grad_norm": 0.722490131855011,
      "learning_rate": 2.8757580929594986e-05,
      "loss": 0.1866,
      "num_input_tokens_seen": 15663648,
      "step": 18080
    },
    {
      "epoch": 8.526638378123527,
      "grad_norm": 0.9685713648796082,
      "learning_rate": 2.87478746901677e-05,
      "loss": 0.1437,
      "num_input_tokens_seen": 15668224,
      "step": 18085
    },
    {
      "epoch": 8.52899575671853,
      "grad_norm": 0.6059074997901917,
      "learning_rate": 2.873816787277103e-05,
      "loss": 0.1664,
      "num_input_tokens_seen": 15672128,
      "step": 18090
    },
    {
      "epoch": 8.531353135313532,
      "grad_norm": 0.8516188263893127,
      "learning_rate": 2.8728460478901903e-05,
      "loss": 0.1994,
      "num_input_tokens_seen": 15675712,
      "step": 18095
    },
    {
      "epoch": 8.533710513908535,
      "grad_norm": 1.2641311883926392,
      "learning_rate": 2.8718752510057307e-05,
      "loss": 0.153,
      "num_input_tokens_seen": 15680560,
      "step": 18100
    },
    {
      "epoch": 8.536067892503535,
      "grad_norm": 0.6822392344474792,
      "learning_rate": 2.870904396773435e-05,
      "loss": 0.203,
      "num_input_tokens_seen": 15685008,
      "step": 18105
    },
    {
      "epoch": 8.53842527109854,
      "grad_norm": 1.4142227172851562,
      "learning_rate": 2.86993348534302e-05,
      "loss": 0.1191,
      "num_input_tokens_seen": 15689344,
      "step": 18110
    },
    {
      "epoch": 8.54078264969354,
      "grad_norm": 0.43715423345565796,
      "learning_rate": 2.868962516864212e-05,
      "loss": 0.185,
      "num_input_tokens_seen": 15693296,
      "step": 18115
    },
    {
      "epoch": 8.543140028288542,
      "grad_norm": 0.7496107816696167,
      "learning_rate": 2.8679914914867477e-05,
      "loss": 0.1,
      "num_input_tokens_seen": 15697264,
      "step": 18120
    },
    {
      "epoch": 8.545497406883545,
      "grad_norm": 0.8914118409156799,
      "learning_rate": 2.8670204093603713e-05,
      "loss": 0.2018,
      "num_input_tokens_seen": 15702704,
      "step": 18125
    },
    {
      "epoch": 8.547854785478547,
      "grad_norm": 1.2754064798355103,
      "learning_rate": 2.8660492706348357e-05,
      "loss": 0.2549,
      "num_input_tokens_seen": 15707376,
      "step": 18130
    },
    {
      "epoch": 8.55021216407355,
      "grad_norm": 0.8644305467605591,
      "learning_rate": 2.8650780754599022e-05,
      "loss": 0.1262,
      "num_input_tokens_seen": 15711408,
      "step": 18135
    },
    {
      "epoch": 8.552569542668552,
      "grad_norm": 0.8962616324424744,
      "learning_rate": 2.8641068239853407e-05,
      "loss": 0.1692,
      "num_input_tokens_seen": 15716096,
      "step": 18140
    },
    {
      "epoch": 8.554926921263554,
      "grad_norm": 0.644536018371582,
      "learning_rate": 2.863135516360932e-05,
      "loss": 0.2053,
      "num_input_tokens_seen": 15720656,
      "step": 18145
    },
    {
      "epoch": 8.557284299858557,
      "grad_norm": 0.4446660876274109,
      "learning_rate": 2.8621641527364633e-05,
      "loss": 0.1588,
      "num_input_tokens_seen": 15724176,
      "step": 18150
    },
    {
      "epoch": 8.55964167845356,
      "grad_norm": 1.6593291759490967,
      "learning_rate": 2.8611927332617313e-05,
      "loss": 0.2808,
      "num_input_tokens_seen": 15728144,
      "step": 18155
    },
    {
      "epoch": 8.561999057048562,
      "grad_norm": 0.830574095249176,
      "learning_rate": 2.8602212580865405e-05,
      "loss": 0.1568,
      "num_input_tokens_seen": 15733328,
      "step": 18160
    },
    {
      "epoch": 8.564356435643564,
      "grad_norm": 1.0036848783493042,
      "learning_rate": 2.859249727360705e-05,
      "loss": 0.2051,
      "num_input_tokens_seen": 15737712,
      "step": 18165
    },
    {
      "epoch": 8.566713814238566,
      "grad_norm": 1.5440980195999146,
      "learning_rate": 2.8582781412340465e-05,
      "loss": 0.2071,
      "num_input_tokens_seen": 15741952,
      "step": 18170
    },
    {
      "epoch": 8.569071192833569,
      "grad_norm": 1.0131826400756836,
      "learning_rate": 2.857306499856397e-05,
      "loss": 0.1763,
      "num_input_tokens_seen": 15746464,
      "step": 18175
    },
    {
      "epoch": 8.571428571428571,
      "grad_norm": 0.34713292121887207,
      "learning_rate": 2.856334803377594e-05,
      "loss": 0.2114,
      "num_input_tokens_seen": 15750480,
      "step": 18180
    },
    {
      "epoch": 8.573785950023574,
      "grad_norm": 0.49178558588027954,
      "learning_rate": 2.8553630519474867e-05,
      "loss": 0.1911,
      "num_input_tokens_seen": 15754736,
      "step": 18185
    },
    {
      "epoch": 8.576143328618576,
      "grad_norm": 1.1476927995681763,
      "learning_rate": 2.8543912457159317e-05,
      "loss": 0.2489,
      "num_input_tokens_seen": 15759280,
      "step": 18190
    },
    {
      "epoch": 8.578500707213578,
      "grad_norm": 0.5425099730491638,
      "learning_rate": 2.853419384832792e-05,
      "loss": 0.168,
      "num_input_tokens_seen": 15763312,
      "step": 18195
    },
    {
      "epoch": 8.58085808580858,
      "grad_norm": 0.5805662870407104,
      "learning_rate": 2.8524474694479423e-05,
      "loss": 0.1669,
      "num_input_tokens_seen": 15768288,
      "step": 18200
    },
    {
      "epoch": 8.58085808580858,
      "eval_loss": 0.20202966034412384,
      "eval_runtime": 22.1632,
      "eval_samples_per_second": 42.548,
      "eval_steps_per_second": 21.297,
      "num_input_tokens_seen": 15768288,
      "step": 18200
    },
    {
      "epoch": 8.583215464403583,
      "grad_norm": 1.5803784132003784,
      "learning_rate": 2.851475499711264e-05,
      "loss": 0.3037,
      "num_input_tokens_seen": 15772000,
      "step": 18205
    },
    {
      "epoch": 8.585572842998586,
      "grad_norm": 0.7685804963111877,
      "learning_rate": 2.8505034757726468e-05,
      "loss": 0.2215,
      "num_input_tokens_seen": 15775280,
      "step": 18210
    },
    {
      "epoch": 8.587930221593588,
      "grad_norm": 2.5676846504211426,
      "learning_rate": 2.8495313977819886e-05,
      "loss": 0.2454,
      "num_input_tokens_seen": 15780464,
      "step": 18215
    },
    {
      "epoch": 8.59028760018859,
      "grad_norm": 0.8196375370025635,
      "learning_rate": 2.8485592658891956e-05,
      "loss": 0.1762,
      "num_input_tokens_seen": 15785040,
      "step": 18220
    },
    {
      "epoch": 8.592644978783593,
      "grad_norm": 0.3499718904495239,
      "learning_rate": 2.8475870802441844e-05,
      "loss": 0.127,
      "num_input_tokens_seen": 15789072,
      "step": 18225
    },
    {
      "epoch": 8.595002357378595,
      "grad_norm": 0.6821929812431335,
      "learning_rate": 2.8466148409968774e-05,
      "loss": 0.2277,
      "num_input_tokens_seen": 15793744,
      "step": 18230
    },
    {
      "epoch": 8.597359735973598,
      "grad_norm": 0.7979618310928345,
      "learning_rate": 2.8456425482972067e-05,
      "loss": 0.2208,
      "num_input_tokens_seen": 15798080,
      "step": 18235
    },
    {
      "epoch": 8.5997171145686,
      "grad_norm": 0.4553111493587494,
      "learning_rate": 2.84467020229511e-05,
      "loss": 0.1641,
      "num_input_tokens_seen": 15802192,
      "step": 18240
    },
    {
      "epoch": 8.602074493163602,
      "grad_norm": 1.9385855197906494,
      "learning_rate": 2.8436978031405375e-05,
      "loss": 0.1916,
      "num_input_tokens_seen": 15806464,
      "step": 18245
    },
    {
      "epoch": 8.604431871758605,
      "grad_norm": 1.510809063911438,
      "learning_rate": 2.842725350983445e-05,
      "loss": 0.1942,
      "num_input_tokens_seen": 15811232,
      "step": 18250
    },
    {
      "epoch": 8.606789250353607,
      "grad_norm": 0.38915857672691345,
      "learning_rate": 2.8417528459737957e-05,
      "loss": 0.2147,
      "num_input_tokens_seen": 15815600,
      "step": 18255
    },
    {
      "epoch": 8.60914662894861,
      "grad_norm": 1.348555326461792,
      "learning_rate": 2.8407802882615624e-05,
      "loss": 0.2297,
      "num_input_tokens_seen": 15819456,
      "step": 18260
    },
    {
      "epoch": 8.611504007543612,
      "grad_norm": 1.9775934219360352,
      "learning_rate": 2.8398076779967277e-05,
      "loss": 0.2612,
      "num_input_tokens_seen": 15823200,
      "step": 18265
    },
    {
      "epoch": 8.613861386138614,
      "grad_norm": 1.4141714572906494,
      "learning_rate": 2.8388350153292774e-05,
      "loss": 0.1887,
      "num_input_tokens_seen": 15828576,
      "step": 18270
    },
    {
      "epoch": 8.616218764733617,
      "grad_norm": 1.2495176792144775,
      "learning_rate": 2.8378623004092103e-05,
      "loss": 0.1781,
      "num_input_tokens_seen": 15832720,
      "step": 18275
    },
    {
      "epoch": 8.61857614332862,
      "grad_norm": 0.7915297746658325,
      "learning_rate": 2.8368895333865302e-05,
      "loss": 0.1671,
      "num_input_tokens_seen": 15837472,
      "step": 18280
    },
    {
      "epoch": 8.620933521923622,
      "grad_norm": 0.7336122989654541,
      "learning_rate": 2.835916714411251e-05,
      "loss": 0.1631,
      "num_input_tokens_seen": 15841792,
      "step": 18285
    },
    {
      "epoch": 8.623290900518624,
      "grad_norm": 0.6423028111457825,
      "learning_rate": 2.8349438436333926e-05,
      "loss": 0.2432,
      "num_input_tokens_seen": 15846256,
      "step": 18290
    },
    {
      "epoch": 8.625648279113626,
      "grad_norm": 1.2693312168121338,
      "learning_rate": 2.833970921202984e-05,
      "loss": 0.1927,
      "num_input_tokens_seen": 15850416,
      "step": 18295
    },
    {
      "epoch": 8.628005657708629,
      "grad_norm": 0.40899109840393066,
      "learning_rate": 2.8329979472700628e-05,
      "loss": 0.1525,
      "num_input_tokens_seen": 15854912,
      "step": 18300
    },
    {
      "epoch": 8.630363036303631,
      "grad_norm": 1.106114149093628,
      "learning_rate": 2.832024921984674e-05,
      "loss": 0.2087,
      "num_input_tokens_seen": 15859648,
      "step": 18305
    },
    {
      "epoch": 8.632720414898632,
      "grad_norm": 0.8054942488670349,
      "learning_rate": 2.8310518454968693e-05,
      "loss": 0.1837,
      "num_input_tokens_seen": 15863312,
      "step": 18310
    },
    {
      "epoch": 8.635077793493634,
      "grad_norm": 0.9059444069862366,
      "learning_rate": 2.8300787179567095e-05,
      "loss": 0.1837,
      "num_input_tokens_seen": 15868656,
      "step": 18315
    },
    {
      "epoch": 8.637435172088637,
      "grad_norm": 0.8242949843406677,
      "learning_rate": 2.8291055395142636e-05,
      "loss": 0.1932,
      "num_input_tokens_seen": 15872688,
      "step": 18320
    },
    {
      "epoch": 8.639792550683639,
      "grad_norm": 0.7729157209396362,
      "learning_rate": 2.8281323103196073e-05,
      "loss": 0.1935,
      "num_input_tokens_seen": 15876880,
      "step": 18325
    },
    {
      "epoch": 8.642149929278641,
      "grad_norm": 0.9696890115737915,
      "learning_rate": 2.8271590305228256e-05,
      "loss": 0.2309,
      "num_input_tokens_seen": 15881072,
      "step": 18330
    },
    {
      "epoch": 8.644507307873644,
      "grad_norm": 1.9376912117004395,
      "learning_rate": 2.82618570027401e-05,
      "loss": 0.2347,
      "num_input_tokens_seen": 15885120,
      "step": 18335
    },
    {
      "epoch": 8.646864686468646,
      "grad_norm": 1.1762081384658813,
      "learning_rate": 2.8252123197232604e-05,
      "loss": 0.2258,
      "num_input_tokens_seen": 15889584,
      "step": 18340
    },
    {
      "epoch": 8.649222065063649,
      "grad_norm": 0.9932411313056946,
      "learning_rate": 2.8242388890206843e-05,
      "loss": 0.1573,
      "num_input_tokens_seen": 15893568,
      "step": 18345
    },
    {
      "epoch": 8.651579443658651,
      "grad_norm": 0.9040425419807434,
      "learning_rate": 2.8232654083163967e-05,
      "loss": 0.1376,
      "num_input_tokens_seen": 15897952,
      "step": 18350
    },
    {
      "epoch": 8.653936822253653,
      "grad_norm": 1.0374219417572021,
      "learning_rate": 2.822291877760521e-05,
      "loss": 0.1393,
      "num_input_tokens_seen": 15901712,
      "step": 18355
    },
    {
      "epoch": 8.656294200848656,
      "grad_norm": 0.5193894505500793,
      "learning_rate": 2.8213182975031864e-05,
      "loss": 0.1196,
      "num_input_tokens_seen": 15906160,
      "step": 18360
    },
    {
      "epoch": 8.658651579443658,
      "grad_norm": 0.3763711750507355,
      "learning_rate": 2.8203446676945337e-05,
      "loss": 0.1797,
      "num_input_tokens_seen": 15910256,
      "step": 18365
    },
    {
      "epoch": 8.66100895803866,
      "grad_norm": 1.4771957397460938,
      "learning_rate": 2.8193709884847075e-05,
      "loss": 0.1989,
      "num_input_tokens_seen": 15914592,
      "step": 18370
    },
    {
      "epoch": 8.663366336633663,
      "grad_norm": 0.8207273483276367,
      "learning_rate": 2.8183972600238605e-05,
      "loss": 0.1342,
      "num_input_tokens_seen": 15920224,
      "step": 18375
    },
    {
      "epoch": 8.665723715228665,
      "grad_norm": 0.9663313627243042,
      "learning_rate": 2.817423482462156e-05,
      "loss": 0.1575,
      "num_input_tokens_seen": 15924944,
      "step": 18380
    },
    {
      "epoch": 8.668081093823668,
      "grad_norm": 1.7451072931289673,
      "learning_rate": 2.8164496559497605e-05,
      "loss": 0.1545,
      "num_input_tokens_seen": 15929264,
      "step": 18385
    },
    {
      "epoch": 8.67043847241867,
      "grad_norm": 1.1182734966278076,
      "learning_rate": 2.815475780636852e-05,
      "loss": 0.1861,
      "num_input_tokens_seen": 15933264,
      "step": 18390
    },
    {
      "epoch": 8.672795851013673,
      "grad_norm": 1.0788722038269043,
      "learning_rate": 2.814501856673613e-05,
      "loss": 0.1659,
      "num_input_tokens_seen": 15937680,
      "step": 18395
    },
    {
      "epoch": 8.675153229608675,
      "grad_norm": 1.4009209871292114,
      "learning_rate": 2.8135278842102353e-05,
      "loss": 0.2619,
      "num_input_tokens_seen": 15941776,
      "step": 18400
    },
    {
      "epoch": 8.675153229608675,
      "eval_loss": 0.20244957506656647,
      "eval_runtime": 22.1754,
      "eval_samples_per_second": 42.525,
      "eval_steps_per_second": 21.285,
      "num_input_tokens_seen": 15941776,
      "step": 18400
    },
    {
      "epoch": 8.677510608203677,
      "grad_norm": 0.44532373547554016,
      "learning_rate": 2.8125538633969183e-05,
      "loss": 0.1781,
      "num_input_tokens_seen": 15945232,
      "step": 18405
    },
    {
      "epoch": 8.67986798679868,
      "grad_norm": 1.4560130834579468,
      "learning_rate": 2.8115797943838677e-05,
      "loss": 0.1686,
      "num_input_tokens_seen": 15948944,
      "step": 18410
    },
    {
      "epoch": 8.682225365393682,
      "grad_norm": 1.4544957876205444,
      "learning_rate": 2.810605677321298e-05,
      "loss": 0.166,
      "num_input_tokens_seen": 15952848,
      "step": 18415
    },
    {
      "epoch": 8.684582743988685,
      "grad_norm": 0.48117873072624207,
      "learning_rate": 2.809631512359428e-05,
      "loss": 0.1249,
      "num_input_tokens_seen": 15957744,
      "step": 18420
    },
    {
      "epoch": 8.686940122583687,
      "grad_norm": 0.6376723051071167,
      "learning_rate": 2.8086572996484884e-05,
      "loss": 0.1576,
      "num_input_tokens_seen": 15962352,
      "step": 18425
    },
    {
      "epoch": 8.68929750117869,
      "grad_norm": 1.2017980813980103,
      "learning_rate": 2.8076830393387143e-05,
      "loss": 0.2599,
      "num_input_tokens_seen": 15967456,
      "step": 18430
    },
    {
      "epoch": 8.691654879773692,
      "grad_norm": 1.2502646446228027,
      "learning_rate": 2.8067087315803497e-05,
      "loss": 0.1689,
      "num_input_tokens_seen": 15971824,
      "step": 18435
    },
    {
      "epoch": 8.694012258368694,
      "grad_norm": 0.8138018846511841,
      "learning_rate": 2.8057343765236433e-05,
      "loss": 0.1654,
      "num_input_tokens_seen": 15976576,
      "step": 18440
    },
    {
      "epoch": 8.696369636963697,
      "grad_norm": 0.641320526599884,
      "learning_rate": 2.804759974318854e-05,
      "loss": 0.2329,
      "num_input_tokens_seen": 15980272,
      "step": 18445
    },
    {
      "epoch": 8.698727015558699,
      "grad_norm": 0.5749420523643494,
      "learning_rate": 2.8037855251162482e-05,
      "loss": 0.1933,
      "num_input_tokens_seen": 15985248,
      "step": 18450
    },
    {
      "epoch": 8.701084394153701,
      "grad_norm": 0.6934888362884521,
      "learning_rate": 2.802811029066096e-05,
      "loss": 0.1841,
      "num_input_tokens_seen": 15989104,
      "step": 18455
    },
    {
      "epoch": 8.703441772748704,
      "grad_norm": 0.6819186806678772,
      "learning_rate": 2.8018364863186764e-05,
      "loss": 0.1301,
      "num_input_tokens_seen": 15992736,
      "step": 18460
    },
    {
      "epoch": 8.705799151343706,
      "grad_norm": 1.0230882167816162,
      "learning_rate": 2.800861897024279e-05,
      "loss": 0.1812,
      "num_input_tokens_seen": 15996912,
      "step": 18465
    },
    {
      "epoch": 8.708156529938709,
      "grad_norm": 0.564306914806366,
      "learning_rate": 2.799887261333196e-05,
      "loss": 0.1842,
      "num_input_tokens_seen": 16001568,
      "step": 18470
    },
    {
      "epoch": 8.710513908533711,
      "grad_norm": 0.8450314998626709,
      "learning_rate": 2.798912579395728e-05,
      "loss": 0.1218,
      "num_input_tokens_seen": 16006288,
      "step": 18475
    },
    {
      "epoch": 8.712871287128714,
      "grad_norm": 0.3826739490032196,
      "learning_rate": 2.797937851362185e-05,
      "loss": 0.1306,
      "num_input_tokens_seen": 16010976,
      "step": 18480
    },
    {
      "epoch": 8.715228665723716,
      "grad_norm": 1.3557446002960205,
      "learning_rate": 2.7969630773828802e-05,
      "loss": 0.1959,
      "num_input_tokens_seen": 16015840,
      "step": 18485
    },
    {
      "epoch": 8.717586044318718,
      "grad_norm": 0.8700003623962402,
      "learning_rate": 2.7959882576081382e-05,
      "loss": 0.109,
      "num_input_tokens_seen": 16019888,
      "step": 18490
    },
    {
      "epoch": 8.71994342291372,
      "grad_norm": 1.8544145822525024,
      "learning_rate": 2.795013392188286e-05,
      "loss": 0.2081,
      "num_input_tokens_seen": 16023936,
      "step": 18495
    },
    {
      "epoch": 8.722300801508723,
      "grad_norm": 0.4651021957397461,
      "learning_rate": 2.7940384812736614e-05,
      "loss": 0.1965,
      "num_input_tokens_seen": 16028208,
      "step": 18500
    },
    {
      "epoch": 8.724658180103724,
      "grad_norm": 1.7595553398132324,
      "learning_rate": 2.7930635250146087e-05,
      "loss": 0.1445,
      "num_input_tokens_seen": 16032464,
      "step": 18505
    },
    {
      "epoch": 8.727015558698728,
      "grad_norm": 0.5921587944030762,
      "learning_rate": 2.792088523561477e-05,
      "loss": 0.1291,
      "num_input_tokens_seen": 16036960,
      "step": 18510
    },
    {
      "epoch": 8.729372937293729,
      "grad_norm": 1.1814130544662476,
      "learning_rate": 2.7911134770646246e-05,
      "loss": 0.1447,
      "num_input_tokens_seen": 16042112,
      "step": 18515
    },
    {
      "epoch": 8.731730315888731,
      "grad_norm": 0.5058456063270569,
      "learning_rate": 2.7901383856744157e-05,
      "loss": 0.1276,
      "num_input_tokens_seen": 16045904,
      "step": 18520
    },
    {
      "epoch": 8.734087694483733,
      "grad_norm": 0.6077386736869812,
      "learning_rate": 2.7891632495412217e-05,
      "loss": 0.1813,
      "num_input_tokens_seen": 16049824,
      "step": 18525
    },
    {
      "epoch": 8.736445073078736,
      "grad_norm": 1.1759897470474243,
      "learning_rate": 2.7881880688154205e-05,
      "loss": 0.1646,
      "num_input_tokens_seen": 16053728,
      "step": 18530
    },
    {
      "epoch": 8.738802451673738,
      "grad_norm": 1.1833091974258423,
      "learning_rate": 2.7872128436473977e-05,
      "loss": 0.1853,
      "num_input_tokens_seen": 16058032,
      "step": 18535
    },
    {
      "epoch": 8.74115983026874,
      "grad_norm": 0.4470970630645752,
      "learning_rate": 2.7862375741875448e-05,
      "loss": 0.1791,
      "num_input_tokens_seen": 16062384,
      "step": 18540
    },
    {
      "epoch": 8.743517208863743,
      "grad_norm": 1.7646827697753906,
      "learning_rate": 2.785262260586261e-05,
      "loss": 0.3381,
      "num_input_tokens_seen": 16067312,
      "step": 18545
    },
    {
      "epoch": 8.745874587458745,
      "grad_norm": 0.17141233384609222,
      "learning_rate": 2.7842869029939517e-05,
      "loss": 0.1461,
      "num_input_tokens_seen": 16071904,
      "step": 18550
    },
    {
      "epoch": 8.748231966053748,
      "grad_norm": 0.9262630343437195,
      "learning_rate": 2.7833115015610296e-05,
      "loss": 0.1921,
      "num_input_tokens_seen": 16076880,
      "step": 18555
    },
    {
      "epoch": 8.75058934464875,
      "grad_norm": 0.5823333263397217,
      "learning_rate": 2.7823360564379136e-05,
      "loss": 0.2526,
      "num_input_tokens_seen": 16080704,
      "step": 18560
    },
    {
      "epoch": 8.752946723243753,
      "grad_norm": 1.3908771276474,
      "learning_rate": 2.7813605677750297e-05,
      "loss": 0.2282,
      "num_input_tokens_seen": 16084992,
      "step": 18565
    },
    {
      "epoch": 8.755304101838755,
      "grad_norm": 0.6180805563926697,
      "learning_rate": 2.7803850357228102e-05,
      "loss": 0.1541,
      "num_input_tokens_seen": 16090080,
      "step": 18570
    },
    {
      "epoch": 8.757661480433757,
      "grad_norm": 0.9861109852790833,
      "learning_rate": 2.779409460431695e-05,
      "loss": 0.203,
      "num_input_tokens_seen": 16094240,
      "step": 18575
    },
    {
      "epoch": 8.76001885902876,
      "grad_norm": 0.604365885257721,
      "learning_rate": 2.778433842052129e-05,
      "loss": 0.1229,
      "num_input_tokens_seen": 16099184,
      "step": 18580
    },
    {
      "epoch": 8.762376237623762,
      "grad_norm": 0.7979449033737183,
      "learning_rate": 2.7774581807345664e-05,
      "loss": 0.1715,
      "num_input_tokens_seen": 16103680,
      "step": 18585
    },
    {
      "epoch": 8.764733616218765,
      "grad_norm": 1.0595217943191528,
      "learning_rate": 2.776482476629465e-05,
      "loss": 0.2255,
      "num_input_tokens_seen": 16107536,
      "step": 18590
    },
    {
      "epoch": 8.767090994813767,
      "grad_norm": 1.7671165466308594,
      "learning_rate": 2.7755067298872924e-05,
      "loss": 0.2308,
      "num_input_tokens_seen": 16111104,
      "step": 18595
    },
    {
      "epoch": 8.76944837340877,
      "grad_norm": 0.4373203217983246,
      "learning_rate": 2.774530940658518e-05,
      "loss": 0.1593,
      "num_input_tokens_seen": 16115152,
      "step": 18600
    },
    {
      "epoch": 8.76944837340877,
      "eval_loss": 0.20276881754398346,
      "eval_runtime": 22.1663,
      "eval_samples_per_second": 42.542,
      "eval_steps_per_second": 21.294,
      "num_input_tokens_seen": 16115152,
      "step": 18600
    },
    {
      "epoch": 8.771805752003772,
      "grad_norm": 0.6744709014892578,
      "learning_rate": 2.7735551090936236e-05,
      "loss": 0.1855,
      "num_input_tokens_seen": 16119024,
      "step": 18605
    },
    {
      "epoch": 8.774163130598774,
      "grad_norm": 0.7809107899665833,
      "learning_rate": 2.7725792353430934e-05,
      "loss": 0.2792,
      "num_input_tokens_seen": 16123632,
      "step": 18610
    },
    {
      "epoch": 8.776520509193777,
      "grad_norm": 1.3199808597564697,
      "learning_rate": 2.77160331955742e-05,
      "loss": 0.2326,
      "num_input_tokens_seen": 16127584,
      "step": 18615
    },
    {
      "epoch": 8.778877887788779,
      "grad_norm": 1.0647861957550049,
      "learning_rate": 2.7706273618871008e-05,
      "loss": 0.1818,
      "num_input_tokens_seen": 16131536,
      "step": 18620
    },
    {
      "epoch": 8.781235266383781,
      "grad_norm": 1.0798879861831665,
      "learning_rate": 2.769651362482642e-05,
      "loss": 0.1597,
      "num_input_tokens_seen": 16136960,
      "step": 18625
    },
    {
      "epoch": 8.783592644978784,
      "grad_norm": 0.5575692653656006,
      "learning_rate": 2.768675321494555e-05,
      "loss": 0.1386,
      "num_input_tokens_seen": 16140784,
      "step": 18630
    },
    {
      "epoch": 8.785950023573786,
      "grad_norm": 0.7824182510375977,
      "learning_rate": 2.7676992390733565e-05,
      "loss": 0.1252,
      "num_input_tokens_seen": 16144464,
      "step": 18635
    },
    {
      "epoch": 8.788307402168789,
      "grad_norm": 0.6705513596534729,
      "learning_rate": 2.766723115369571e-05,
      "loss": 0.1158,
      "num_input_tokens_seen": 16148048,
      "step": 18640
    },
    {
      "epoch": 8.790664780763791,
      "grad_norm": 0.8000301718711853,
      "learning_rate": 2.765746950533729e-05,
      "loss": 0.259,
      "num_input_tokens_seen": 16152016,
      "step": 18645
    },
    {
      "epoch": 8.793022159358793,
      "grad_norm": 0.5627002120018005,
      "learning_rate": 2.7647707447163684e-05,
      "loss": 0.198,
      "num_input_tokens_seen": 16156848,
      "step": 18650
    },
    {
      "epoch": 8.795379537953796,
      "grad_norm": 1.2451303005218506,
      "learning_rate": 2.7637944980680315e-05,
      "loss": 0.2103,
      "num_input_tokens_seen": 16160960,
      "step": 18655
    },
    {
      "epoch": 8.797736916548798,
      "grad_norm": 0.527006983757019,
      "learning_rate": 2.762818210739268e-05,
      "loss": 0.1665,
      "num_input_tokens_seen": 16164960,
      "step": 18660
    },
    {
      "epoch": 8.8000942951438,
      "grad_norm": 1.112839937210083,
      "learning_rate": 2.7618418828806332e-05,
      "loss": 0.1648,
      "num_input_tokens_seen": 16169504,
      "step": 18665
    },
    {
      "epoch": 8.802451673738803,
      "grad_norm": 0.9314213395118713,
      "learning_rate": 2.76086551464269e-05,
      "loss": 0.1709,
      "num_input_tokens_seen": 16173168,
      "step": 18670
    },
    {
      "epoch": 8.804809052333805,
      "grad_norm": 1.7670636177062988,
      "learning_rate": 2.759889106176006e-05,
      "loss": 0.1813,
      "num_input_tokens_seen": 16177696,
      "step": 18675
    },
    {
      "epoch": 8.807166430928808,
      "grad_norm": 0.8723813891410828,
      "learning_rate": 2.758912657631156e-05,
      "loss": 0.1586,
      "num_input_tokens_seen": 16182000,
      "step": 18680
    },
    {
      "epoch": 8.80952380952381,
      "grad_norm": 0.7372036576271057,
      "learning_rate": 2.7579361691587198e-05,
      "loss": 0.1571,
      "num_input_tokens_seen": 16186544,
      "step": 18685
    },
    {
      "epoch": 8.811881188118813,
      "grad_norm": 0.5762573480606079,
      "learning_rate": 2.756959640909285e-05,
      "loss": 0.1303,
      "num_input_tokens_seen": 16191216,
      "step": 18690
    },
    {
      "epoch": 8.814238566713815,
      "grad_norm": 2.018683433532715,
      "learning_rate": 2.7559830730334452e-05,
      "loss": 0.2433,
      "num_input_tokens_seen": 16196160,
      "step": 18695
    },
    {
      "epoch": 8.816595945308817,
      "grad_norm": 1.0519917011260986,
      "learning_rate": 2.7550064656817988e-05,
      "loss": 0.168,
      "num_input_tokens_seen": 16200160,
      "step": 18700
    },
    {
      "epoch": 8.81895332390382,
      "grad_norm": 0.4609846770763397,
      "learning_rate": 2.7540298190049503e-05,
      "loss": 0.1094,
      "num_input_tokens_seen": 16203376,
      "step": 18705
    },
    {
      "epoch": 8.82131070249882,
      "grad_norm": 0.6858793497085571,
      "learning_rate": 2.7530531331535107e-05,
      "loss": 0.173,
      "num_input_tokens_seen": 16207472,
      "step": 18710
    },
    {
      "epoch": 8.823668081093825,
      "grad_norm": 1.2351278066635132,
      "learning_rate": 2.752076408278099e-05,
      "loss": 0.1619,
      "num_input_tokens_seen": 16212144,
      "step": 18715
    },
    {
      "epoch": 8.826025459688825,
      "grad_norm": 1.0904645919799805,
      "learning_rate": 2.751099644529337e-05,
      "loss": 0.2051,
      "num_input_tokens_seen": 16216432,
      "step": 18720
    },
    {
      "epoch": 8.828382838283828,
      "grad_norm": 1.9487454891204834,
      "learning_rate": 2.7501228420578533e-05,
      "loss": 0.2297,
      "num_input_tokens_seen": 16220848,
      "step": 18725
    },
    {
      "epoch": 8.83074021687883,
      "grad_norm": 1.3574389219284058,
      "learning_rate": 2.7491460010142857e-05,
      "loss": 0.1658,
      "num_input_tokens_seen": 16224768,
      "step": 18730
    },
    {
      "epoch": 8.833097595473832,
      "grad_norm": 0.2735130488872528,
      "learning_rate": 2.7481691215492727e-05,
      "loss": 0.2039,
      "num_input_tokens_seen": 16228896,
      "step": 18735
    },
    {
      "epoch": 8.835454974068835,
      "grad_norm": 1.0494059324264526,
      "learning_rate": 2.747192203813463e-05,
      "loss": 0.1482,
      "num_input_tokens_seen": 16233088,
      "step": 18740
    },
    {
      "epoch": 8.837812352663837,
      "grad_norm": 1.5181617736816406,
      "learning_rate": 2.7462152479575087e-05,
      "loss": 0.1899,
      "num_input_tokens_seen": 16237648,
      "step": 18745
    },
    {
      "epoch": 8.84016973125884,
      "grad_norm": 1.8434092998504639,
      "learning_rate": 2.7452382541320697e-05,
      "loss": 0.1818,
      "num_input_tokens_seen": 16241824,
      "step": 18750
    },
    {
      "epoch": 8.842527109853842,
      "grad_norm": 1.0983213186264038,
      "learning_rate": 2.7442612224878096e-05,
      "loss": 0.1907,
      "num_input_tokens_seen": 16245856,
      "step": 18755
    },
    {
      "epoch": 8.844884488448844,
      "grad_norm": 1.2063379287719727,
      "learning_rate": 2.7432841531753994e-05,
      "loss": 0.2143,
      "num_input_tokens_seen": 16250480,
      "step": 18760
    },
    {
      "epoch": 8.847241867043847,
      "grad_norm": 0.8012604117393494,
      "learning_rate": 2.7423070463455147e-05,
      "loss": 0.1589,
      "num_input_tokens_seen": 16254864,
      "step": 18765
    },
    {
      "epoch": 8.84959924563885,
      "grad_norm": 0.8784675598144531,
      "learning_rate": 2.7413299021488397e-05,
      "loss": 0.2021,
      "num_input_tokens_seen": 16258704,
      "step": 18770
    },
    {
      "epoch": 8.851956624233852,
      "grad_norm": 0.9085725545883179,
      "learning_rate": 2.7403527207360615e-05,
      "loss": 0.2371,
      "num_input_tokens_seen": 16262864,
      "step": 18775
    },
    {
      "epoch": 8.854314002828854,
      "grad_norm": 1.3182965517044067,
      "learning_rate": 2.7393755022578722e-05,
      "loss": 0.2218,
      "num_input_tokens_seen": 16267632,
      "step": 18780
    },
    {
      "epoch": 8.856671381423856,
      "grad_norm": 1.0883091688156128,
      "learning_rate": 2.7383982468649714e-05,
      "loss": 0.163,
      "num_input_tokens_seen": 16271552,
      "step": 18785
    },
    {
      "epoch": 8.859028760018859,
      "grad_norm": 0.6269042491912842,
      "learning_rate": 2.7374209547080665e-05,
      "loss": 0.154,
      "num_input_tokens_seen": 16275680,
      "step": 18790
    },
    {
      "epoch": 8.861386138613861,
      "grad_norm": 0.5916184186935425,
      "learning_rate": 2.7364436259378663e-05,
      "loss": 0.1586,
      "num_input_tokens_seen": 16279744,
      "step": 18795
    },
    {
      "epoch": 8.863743517208864,
      "grad_norm": 1.4053263664245605,
      "learning_rate": 2.735466260705088e-05,
      "loss": 0.19,
      "num_input_tokens_seen": 16284384,
      "step": 18800
    },
    {
      "epoch": 8.863743517208864,
      "eval_loss": 0.20295360684394836,
      "eval_runtime": 22.1432,
      "eval_samples_per_second": 42.586,
      "eval_steps_per_second": 21.316,
      "num_input_tokens_seen": 16284384,
      "step": 18800
    },
    {
      "epoch": 8.866100895803866,
      "grad_norm": 1.3965189456939697,
      "learning_rate": 2.7344888591604524e-05,
      "loss": 0.1722,
      "num_input_tokens_seen": 16288832,
      "step": 18805
    },
    {
      "epoch": 8.868458274398868,
      "grad_norm": 0.8974383473396301,
      "learning_rate": 2.7335114214546893e-05,
      "loss": 0.1572,
      "num_input_tokens_seen": 16293152,
      "step": 18810
    },
    {
      "epoch": 8.87081565299387,
      "grad_norm": 1.356257438659668,
      "learning_rate": 2.7325339477385293e-05,
      "loss": 0.1285,
      "num_input_tokens_seen": 16297616,
      "step": 18815
    },
    {
      "epoch": 8.873173031588873,
      "grad_norm": 0.9156038165092468,
      "learning_rate": 2.7315564381627128e-05,
      "loss": 0.1872,
      "num_input_tokens_seen": 16302224,
      "step": 18820
    },
    {
      "epoch": 8.875530410183876,
      "grad_norm": 0.8896782398223877,
      "learning_rate": 2.7305788928779835e-05,
      "loss": 0.1633,
      "num_input_tokens_seen": 16307392,
      "step": 18825
    },
    {
      "epoch": 8.877887788778878,
      "grad_norm": 1.2200373411178589,
      "learning_rate": 2.729601312035091e-05,
      "loss": 0.3382,
      "num_input_tokens_seen": 16312528,
      "step": 18830
    },
    {
      "epoch": 8.88024516737388,
      "grad_norm": 1.5533092021942139,
      "learning_rate": 2.7286236957847915e-05,
      "loss": 0.2127,
      "num_input_tokens_seen": 16316736,
      "step": 18835
    },
    {
      "epoch": 8.882602545968883,
      "grad_norm": 0.7247341871261597,
      "learning_rate": 2.7276460442778446e-05,
      "loss": 0.1804,
      "num_input_tokens_seen": 16321152,
      "step": 18840
    },
    {
      "epoch": 8.884959924563885,
      "grad_norm": 0.6787034273147583,
      "learning_rate": 2.726668357665017e-05,
      "loss": 0.1464,
      "num_input_tokens_seen": 16325344,
      "step": 18845
    },
    {
      "epoch": 8.887317303158888,
      "grad_norm": 0.4679868519306183,
      "learning_rate": 2.7256906360970808e-05,
      "loss": 0.1332,
      "num_input_tokens_seen": 16328928,
      "step": 18850
    },
    {
      "epoch": 8.88967468175389,
      "grad_norm": 0.4664399325847626,
      "learning_rate": 2.7247128797248117e-05,
      "loss": 0.1987,
      "num_input_tokens_seen": 16333024,
      "step": 18855
    },
    {
      "epoch": 8.892032060348892,
      "grad_norm": 1.1523762941360474,
      "learning_rate": 2.7237350886989925e-05,
      "loss": 0.2026,
      "num_input_tokens_seen": 16337136,
      "step": 18860
    },
    {
      "epoch": 8.894389438943895,
      "grad_norm": 1.1681113243103027,
      "learning_rate": 2.7227572631704107e-05,
      "loss": 0.2004,
      "num_input_tokens_seen": 16341472,
      "step": 18865
    },
    {
      "epoch": 8.896746817538897,
      "grad_norm": 0.7295750379562378,
      "learning_rate": 2.7217794032898596e-05,
      "loss": 0.1274,
      "num_input_tokens_seen": 16346480,
      "step": 18870
    },
    {
      "epoch": 8.8991041961339,
      "grad_norm": 1.1496398448944092,
      "learning_rate": 2.7208015092081384e-05,
      "loss": 0.1759,
      "num_input_tokens_seen": 16350272,
      "step": 18875
    },
    {
      "epoch": 8.901461574728902,
      "grad_norm": 1.3583189249038696,
      "learning_rate": 2.719823581076049e-05,
      "loss": 0.2184,
      "num_input_tokens_seen": 16354160,
      "step": 18880
    },
    {
      "epoch": 8.903818953323904,
      "grad_norm": 1.382176160812378,
      "learning_rate": 2.718845619044401e-05,
      "loss": 0.2055,
      "num_input_tokens_seen": 16359248,
      "step": 18885
    },
    {
      "epoch": 8.906176331918907,
      "grad_norm": 2.560415029525757,
      "learning_rate": 2.7178676232640088e-05,
      "loss": 0.2457,
      "num_input_tokens_seen": 16363664,
      "step": 18890
    },
    {
      "epoch": 8.90853371051391,
      "grad_norm": 1.625309944152832,
      "learning_rate": 2.716889593885691e-05,
      "loss": 0.2069,
      "num_input_tokens_seen": 16367632,
      "step": 18895
    },
    {
      "epoch": 8.910891089108912,
      "grad_norm": 0.6153598427772522,
      "learning_rate": 2.7159115310602716e-05,
      "loss": 0.1899,
      "num_input_tokens_seen": 16371536,
      "step": 18900
    },
    {
      "epoch": 8.913248467703912,
      "grad_norm": 0.8673765063285828,
      "learning_rate": 2.7149334349385814e-05,
      "loss": 0.2369,
      "num_input_tokens_seen": 16375888,
      "step": 18905
    },
    {
      "epoch": 8.915605846298917,
      "grad_norm": 0.658880352973938,
      "learning_rate": 2.713955305671454e-05,
      "loss": 0.1277,
      "num_input_tokens_seen": 16380816,
      "step": 18910
    },
    {
      "epoch": 8.917963224893917,
      "grad_norm": 0.3486378788948059,
      "learning_rate": 2.71297714340973e-05,
      "loss": 0.1794,
      "num_input_tokens_seen": 16384944,
      "step": 18915
    },
    {
      "epoch": 8.92032060348892,
      "grad_norm": 2.17136287689209,
      "learning_rate": 2.7119989483042545e-05,
      "loss": 0.2403,
      "num_input_tokens_seen": 16389296,
      "step": 18920
    },
    {
      "epoch": 8.922677982083922,
      "grad_norm": 1.733189344406128,
      "learning_rate": 2.7110207205058768e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 16393440,
      "step": 18925
    },
    {
      "epoch": 8.925035360678924,
      "grad_norm": 0.8790375590324402,
      "learning_rate": 2.7100424601654517e-05,
      "loss": 0.1976,
      "num_input_tokens_seen": 16397872,
      "step": 18930
    },
    {
      "epoch": 8.927392739273927,
      "grad_norm": 1.0234249830245972,
      "learning_rate": 2.7090641674338403e-05,
      "loss": 0.2328,
      "num_input_tokens_seen": 16401968,
      "step": 18935
    },
    {
      "epoch": 8.92975011786893,
      "grad_norm": 0.9621090292930603,
      "learning_rate": 2.7080858424619072e-05,
      "loss": 0.1211,
      "num_input_tokens_seen": 16406000,
      "step": 18940
    },
    {
      "epoch": 8.932107496463932,
      "grad_norm": 1.2795737981796265,
      "learning_rate": 2.707107485400521e-05,
      "loss": 0.1981,
      "num_input_tokens_seen": 16410336,
      "step": 18945
    },
    {
      "epoch": 8.934464875058934,
      "grad_norm": 1.3813713788986206,
      "learning_rate": 2.7061290964005586e-05,
      "loss": 0.2562,
      "num_input_tokens_seen": 16414896,
      "step": 18950
    },
    {
      "epoch": 8.936822253653936,
      "grad_norm": 0.7733728885650635,
      "learning_rate": 2.7051506756129e-05,
      "loss": 0.1578,
      "num_input_tokens_seen": 16419056,
      "step": 18955
    },
    {
      "epoch": 8.939179632248939,
      "grad_norm": 2.107764482498169,
      "learning_rate": 2.704172223188428e-05,
      "loss": 0.1828,
      "num_input_tokens_seen": 16423088,
      "step": 18960
    },
    {
      "epoch": 8.941537010843941,
      "grad_norm": 0.5536381006240845,
      "learning_rate": 2.7031937392780334e-05,
      "loss": 0.1812,
      "num_input_tokens_seen": 16427152,
      "step": 18965
    },
    {
      "epoch": 8.943894389438944,
      "grad_norm": 0.6241905093193054,
      "learning_rate": 2.702215224032611e-05,
      "loss": 0.1498,
      "num_input_tokens_seen": 16430688,
      "step": 18970
    },
    {
      "epoch": 8.946251768033946,
      "grad_norm": 1.6365330219268799,
      "learning_rate": 2.70123667760306e-05,
      "loss": 0.1787,
      "num_input_tokens_seen": 16435040,
      "step": 18975
    },
    {
      "epoch": 8.948609146628948,
      "grad_norm": 1.288781762123108,
      "learning_rate": 2.7002581001402845e-05,
      "loss": 0.1474,
      "num_input_tokens_seen": 16438960,
      "step": 18980
    },
    {
      "epoch": 8.95096652522395,
      "grad_norm": 0.8721839785575867,
      "learning_rate": 2.6992794917951923e-05,
      "loss": 0.1276,
      "num_input_tokens_seen": 16442656,
      "step": 18985
    },
    {
      "epoch": 8.953323903818953,
      "grad_norm": 0.6087895035743713,
      "learning_rate": 2.6983008527187e-05,
      "loss": 0.2387,
      "num_input_tokens_seen": 16446864,
      "step": 18990
    },
    {
      "epoch": 8.955681282413956,
      "grad_norm": 0.7081970572471619,
      "learning_rate": 2.697322183061723e-05,
      "loss": 0.214,
      "num_input_tokens_seen": 16453104,
      "step": 18995
    },
    {
      "epoch": 8.958038661008958,
      "grad_norm": 0.9068287014961243,
      "learning_rate": 2.696343482975186e-05,
      "loss": 0.1903,
      "num_input_tokens_seen": 16457552,
      "step": 19000
    },
    {
      "epoch": 8.958038661008958,
      "eval_loss": 0.20343394577503204,
      "eval_runtime": 22.1561,
      "eval_samples_per_second": 42.562,
      "eval_steps_per_second": 21.303,
      "num_input_tokens_seen": 16457552,
      "step": 19000
    },
    {
      "epoch": 8.96039603960396,
      "grad_norm": 1.6142451763153076,
      "learning_rate": 2.695364752610016e-05,
      "loss": 0.2235,
      "num_input_tokens_seen": 16461392,
      "step": 19005
    },
    {
      "epoch": 8.962753418198963,
      "grad_norm": 0.962779700756073,
      "learning_rate": 2.6943859921171467e-05,
      "loss": 0.1778,
      "num_input_tokens_seen": 16466320,
      "step": 19010
    },
    {
      "epoch": 8.965110796793965,
      "grad_norm": 0.4311082363128662,
      "learning_rate": 2.6934072016475143e-05,
      "loss": 0.1027,
      "num_input_tokens_seen": 16470816,
      "step": 19015
    },
    {
      "epoch": 8.967468175388968,
      "grad_norm": 0.6653015613555908,
      "learning_rate": 2.6924283813520606e-05,
      "loss": 0.208,
      "num_input_tokens_seen": 16474976,
      "step": 19020
    },
    {
      "epoch": 8.96982555398397,
      "grad_norm": 0.7246946692466736,
      "learning_rate": 2.691449531381733e-05,
      "loss": 0.2269,
      "num_input_tokens_seen": 16479216,
      "step": 19025
    },
    {
      "epoch": 8.972182932578972,
      "grad_norm": 0.8324491381645203,
      "learning_rate": 2.6904706518874816e-05,
      "loss": 0.198,
      "num_input_tokens_seen": 16483536,
      "step": 19030
    },
    {
      "epoch": 8.974540311173975,
      "grad_norm": 0.6189842820167542,
      "learning_rate": 2.6894917430202615e-05,
      "loss": 0.148,
      "num_input_tokens_seen": 16488304,
      "step": 19035
    },
    {
      "epoch": 8.976897689768977,
      "grad_norm": 0.8205829858779907,
      "learning_rate": 2.6885128049310343e-05,
      "loss": 0.2367,
      "num_input_tokens_seen": 16492688,
      "step": 19040
    },
    {
      "epoch": 8.97925506836398,
      "grad_norm": 0.9058135747909546,
      "learning_rate": 2.687533837770762e-05,
      "loss": 0.1897,
      "num_input_tokens_seen": 16496576,
      "step": 19045
    },
    {
      "epoch": 8.981612446958982,
      "grad_norm": 0.7233713865280151,
      "learning_rate": 2.6865548416904162e-05,
      "loss": 0.2163,
      "num_input_tokens_seen": 16501120,
      "step": 19050
    },
    {
      "epoch": 8.983969825553984,
      "grad_norm": 1.0652835369110107,
      "learning_rate": 2.68557581684097e-05,
      "loss": 0.1471,
      "num_input_tokens_seen": 16505520,
      "step": 19055
    },
    {
      "epoch": 8.986327204148987,
      "grad_norm": 0.9564029574394226,
      "learning_rate": 2.6845967633733998e-05,
      "loss": 0.1566,
      "num_input_tokens_seen": 16510032,
      "step": 19060
    },
    {
      "epoch": 8.98868458274399,
      "grad_norm": 0.513271152973175,
      "learning_rate": 2.683617681438689e-05,
      "loss": 0.1843,
      "num_input_tokens_seen": 16514416,
      "step": 19065
    },
    {
      "epoch": 8.991041961338992,
      "grad_norm": 0.539741575717926,
      "learning_rate": 2.682638571187825e-05,
      "loss": 0.1763,
      "num_input_tokens_seen": 16518256,
      "step": 19070
    },
    {
      "epoch": 8.993399339933994,
      "grad_norm": 0.6809326410293579,
      "learning_rate": 2.6816594327717976e-05,
      "loss": 0.1834,
      "num_input_tokens_seen": 16522784,
      "step": 19075
    },
    {
      "epoch": 8.995756718528996,
      "grad_norm": 0.5949167013168335,
      "learning_rate": 2.680680266341603e-05,
      "loss": 0.1835,
      "num_input_tokens_seen": 16527056,
      "step": 19080
    },
    {
      "epoch": 8.998114097123999,
      "grad_norm": 0.7578131556510925,
      "learning_rate": 2.67970107204824e-05,
      "loss": 0.2099,
      "num_input_tokens_seen": 16531328,
      "step": 19085
    },
    {
      "epoch": 9.000471475719001,
      "grad_norm": 0.8708164691925049,
      "learning_rate": 2.6787218500427142e-05,
      "loss": 0.1773,
      "num_input_tokens_seen": 16535808,
      "step": 19090
    },
    {
      "epoch": 9.002828854314004,
      "grad_norm": 0.6089879870414734,
      "learning_rate": 2.6777426004760332e-05,
      "loss": 0.1701,
      "num_input_tokens_seen": 16541664,
      "step": 19095
    },
    {
      "epoch": 9.005186232909006,
      "grad_norm": 0.5763943791389465,
      "learning_rate": 2.6767633234992094e-05,
      "loss": 0.2643,
      "num_input_tokens_seen": 16545808,
      "step": 19100
    },
    {
      "epoch": 9.007543611504008,
      "grad_norm": 0.6569165587425232,
      "learning_rate": 2.6757840192632598e-05,
      "loss": 0.2013,
      "num_input_tokens_seen": 16550400,
      "step": 19105
    },
    {
      "epoch": 9.009900990099009,
      "grad_norm": 2.0740599632263184,
      "learning_rate": 2.6748046879192052e-05,
      "loss": 0.1953,
      "num_input_tokens_seen": 16554240,
      "step": 19110
    },
    {
      "epoch": 9.012258368694011,
      "grad_norm": 0.42417120933532715,
      "learning_rate": 2.673825329618071e-05,
      "loss": 0.1615,
      "num_input_tokens_seen": 16558672,
      "step": 19115
    },
    {
      "epoch": 9.014615747289014,
      "grad_norm": 0.47664111852645874,
      "learning_rate": 2.6728459445108866e-05,
      "loss": 0.1387,
      "num_input_tokens_seen": 16563168,
      "step": 19120
    },
    {
      "epoch": 9.016973125884016,
      "grad_norm": 0.5855072736740112,
      "learning_rate": 2.6718665327486854e-05,
      "loss": 0.1652,
      "num_input_tokens_seen": 16567152,
      "step": 19125
    },
    {
      "epoch": 9.019330504479019,
      "grad_norm": 0.3824607729911804,
      "learning_rate": 2.6708870944825048e-05,
      "loss": 0.1591,
      "num_input_tokens_seen": 16571408,
      "step": 19130
    },
    {
      "epoch": 9.021687883074021,
      "grad_norm": 0.9764345288276672,
      "learning_rate": 2.6699076298633874e-05,
      "loss": 0.114,
      "num_input_tokens_seen": 16575664,
      "step": 19135
    },
    {
      "epoch": 9.024045261669023,
      "grad_norm": 1.2593045234680176,
      "learning_rate": 2.6689281390423788e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 16579744,
      "step": 19140
    },
    {
      "epoch": 9.026402640264026,
      "grad_norm": 0.9506648182868958,
      "learning_rate": 2.667948622170527e-05,
      "loss": 0.1536,
      "num_input_tokens_seen": 16583936,
      "step": 19145
    },
    {
      "epoch": 9.028760018859028,
      "grad_norm": 0.8887495994567871,
      "learning_rate": 2.6669690793988873e-05,
      "loss": 0.1403,
      "num_input_tokens_seen": 16588272,
      "step": 19150
    },
    {
      "epoch": 9.03111739745403,
      "grad_norm": 1.1126550436019897,
      "learning_rate": 2.665989510878518e-05,
      "loss": 0.1567,
      "num_input_tokens_seen": 16591712,
      "step": 19155
    },
    {
      "epoch": 9.033474776049033,
      "grad_norm": 1.5912599563598633,
      "learning_rate": 2.6650099167604793e-05,
      "loss": 0.1993,
      "num_input_tokens_seen": 16595824,
      "step": 19160
    },
    {
      "epoch": 9.035832154644035,
      "grad_norm": 1.8675671815872192,
      "learning_rate": 2.6640302971958376e-05,
      "loss": 0.2596,
      "num_input_tokens_seen": 16600304,
      "step": 19165
    },
    {
      "epoch": 9.038189533239038,
      "grad_norm": 0.43882429599761963,
      "learning_rate": 2.6630506523356635e-05,
      "loss": 0.1508,
      "num_input_tokens_seen": 16604912,
      "step": 19170
    },
    {
      "epoch": 9.04054691183404,
      "grad_norm": 1.6701184511184692,
      "learning_rate": 2.6620709823310297e-05,
      "loss": 0.1807,
      "num_input_tokens_seen": 16609376,
      "step": 19175
    },
    {
      "epoch": 9.042904290429043,
      "grad_norm": 1.0672574043273926,
      "learning_rate": 2.661091287333014e-05,
      "loss": 0.1605,
      "num_input_tokens_seen": 16613296,
      "step": 19180
    },
    {
      "epoch": 9.045261669024045,
      "grad_norm": 0.7580594420433044,
      "learning_rate": 2.660111567492696e-05,
      "loss": 0.201,
      "num_input_tokens_seen": 16617824,
      "step": 19185
    },
    {
      "epoch": 9.047619047619047,
      "grad_norm": 1.476259469985962,
      "learning_rate": 2.6591318229611635e-05,
      "loss": 0.1713,
      "num_input_tokens_seen": 16622768,
      "step": 19190
    },
    {
      "epoch": 9.04997642621405,
      "grad_norm": 1.411295771598816,
      "learning_rate": 2.6581520538895037e-05,
      "loss": 0.2513,
      "num_input_tokens_seen": 16628464,
      "step": 19195
    },
    {
      "epoch": 9.052333804809052,
      "grad_norm": 1.3062070608139038,
      "learning_rate": 2.6571722604288102e-05,
      "loss": 0.1835,
      "num_input_tokens_seen": 16632272,
      "step": 19200
    },
    {
      "epoch": 9.052333804809052,
      "eval_loss": 0.20749124884605408,
      "eval_runtime": 22.1779,
      "eval_samples_per_second": 42.52,
      "eval_steps_per_second": 21.282,
      "num_input_tokens_seen": 16632272,
      "step": 19200
    },
    {
      "epoch": 9.054691183404055,
      "grad_norm": 0.8633865118026733,
      "learning_rate": 2.656192442730179e-05,
      "loss": 0.1669,
      "num_input_tokens_seen": 16635680,
      "step": 19205
    },
    {
      "epoch": 9.057048561999057,
      "grad_norm": 1.3285460472106934,
      "learning_rate": 2.6552126009447098e-05,
      "loss": 0.2643,
      "num_input_tokens_seen": 16640208,
      "step": 19210
    },
    {
      "epoch": 9.05940594059406,
      "grad_norm": 1.3701797723770142,
      "learning_rate": 2.654232735223507e-05,
      "loss": 0.1641,
      "num_input_tokens_seen": 16644784,
      "step": 19215
    },
    {
      "epoch": 9.061763319189062,
      "grad_norm": 0.5281635522842407,
      "learning_rate": 2.6532528457176787e-05,
      "loss": 0.1961,
      "num_input_tokens_seen": 16648912,
      "step": 19220
    },
    {
      "epoch": 9.064120697784064,
      "grad_norm": 0.7861537933349609,
      "learning_rate": 2.6522729325783348e-05,
      "loss": 0.2064,
      "num_input_tokens_seen": 16653216,
      "step": 19225
    },
    {
      "epoch": 9.066478076379067,
      "grad_norm": 1.2307018041610718,
      "learning_rate": 2.6512929959565914e-05,
      "loss": 0.126,
      "num_input_tokens_seen": 16657296,
      "step": 19230
    },
    {
      "epoch": 9.068835454974069,
      "grad_norm": 0.48876798152923584,
      "learning_rate": 2.6503130360035673e-05,
      "loss": 0.158,
      "num_input_tokens_seen": 16661856,
      "step": 19235
    },
    {
      "epoch": 9.071192833569071,
      "grad_norm": 0.658029317855835,
      "learning_rate": 2.6493330528703835e-05,
      "loss": 0.1109,
      "num_input_tokens_seen": 16665792,
      "step": 19240
    },
    {
      "epoch": 9.073550212164074,
      "grad_norm": 0.665662407875061,
      "learning_rate": 2.648353046708167e-05,
      "loss": 0.1666,
      "num_input_tokens_seen": 16670336,
      "step": 19245
    },
    {
      "epoch": 9.075907590759076,
      "grad_norm": 1.290551781654358,
      "learning_rate": 2.647373017668046e-05,
      "loss": 0.1913,
      "num_input_tokens_seen": 16674832,
      "step": 19250
    },
    {
      "epoch": 9.078264969354079,
      "grad_norm": 0.8957157731056213,
      "learning_rate": 2.6463929659011537e-05,
      "loss": 0.2046,
      "num_input_tokens_seen": 16679696,
      "step": 19255
    },
    {
      "epoch": 9.080622347949081,
      "grad_norm": 1.2875727415084839,
      "learning_rate": 2.6454128915586262e-05,
      "loss": 0.1692,
      "num_input_tokens_seen": 16683344,
      "step": 19260
    },
    {
      "epoch": 9.082979726544083,
      "grad_norm": 1.7584184408187866,
      "learning_rate": 2.6444327947916036e-05,
      "loss": 0.1752,
      "num_input_tokens_seen": 16687920,
      "step": 19265
    },
    {
      "epoch": 9.085337105139086,
      "grad_norm": 0.8530916571617126,
      "learning_rate": 2.6434526757512292e-05,
      "loss": 0.1955,
      "num_input_tokens_seen": 16692288,
      "step": 19270
    },
    {
      "epoch": 9.087694483734088,
      "grad_norm": 1.2205127477645874,
      "learning_rate": 2.6424725345886486e-05,
      "loss": 0.1915,
      "num_input_tokens_seen": 16696288,
      "step": 19275
    },
    {
      "epoch": 9.09005186232909,
      "grad_norm": 1.0271824598312378,
      "learning_rate": 2.641492371455014e-05,
      "loss": 0.1713,
      "num_input_tokens_seen": 16700800,
      "step": 19280
    },
    {
      "epoch": 9.092409240924093,
      "grad_norm": 0.689440131187439,
      "learning_rate": 2.640512186501477e-05,
      "loss": 0.2101,
      "num_input_tokens_seen": 16705424,
      "step": 19285
    },
    {
      "epoch": 9.094766619519095,
      "grad_norm": 1.054534912109375,
      "learning_rate": 2.639531979879195e-05,
      "loss": 0.1553,
      "num_input_tokens_seen": 16709648,
      "step": 19290
    },
    {
      "epoch": 9.097123998114098,
      "grad_norm": 0.6756740212440491,
      "learning_rate": 2.638551751739328e-05,
      "loss": 0.2177,
      "num_input_tokens_seen": 16713552,
      "step": 19295
    },
    {
      "epoch": 9.0994813767091,
      "grad_norm": 0.6277413368225098,
      "learning_rate": 2.6375715022330404e-05,
      "loss": 0.1935,
      "num_input_tokens_seen": 16717808,
      "step": 19300
    },
    {
      "epoch": 9.101838755304103,
      "grad_norm": 0.8792551755905151,
      "learning_rate": 2.6365912315114976e-05,
      "loss": 0.1993,
      "num_input_tokens_seen": 16721680,
      "step": 19305
    },
    {
      "epoch": 9.104196133899103,
      "grad_norm": 0.7870965600013733,
      "learning_rate": 2.6356109397258704e-05,
      "loss": 0.2049,
      "num_input_tokens_seen": 16725728,
      "step": 19310
    },
    {
      "epoch": 9.106553512494106,
      "grad_norm": 0.8639470338821411,
      "learning_rate": 2.6346306270273325e-05,
      "loss": 0.1418,
      "num_input_tokens_seen": 16730224,
      "step": 19315
    },
    {
      "epoch": 9.108910891089108,
      "grad_norm": 0.821853756904602,
      "learning_rate": 2.6336502935670608e-05,
      "loss": 0.151,
      "num_input_tokens_seen": 16734128,
      "step": 19320
    },
    {
      "epoch": 9.11126826968411,
      "grad_norm": 1.3475514650344849,
      "learning_rate": 2.6326699394962333e-05,
      "loss": 0.1239,
      "num_input_tokens_seen": 16739024,
      "step": 19325
    },
    {
      "epoch": 9.113625648279113,
      "grad_norm": 0.9155060052871704,
      "learning_rate": 2.6316895649660334e-05,
      "loss": 0.1689,
      "num_input_tokens_seen": 16742800,
      "step": 19330
    },
    {
      "epoch": 9.115983026874115,
      "grad_norm": 1.3159980773925781,
      "learning_rate": 2.6307091701276486e-05,
      "loss": 0.181,
      "num_input_tokens_seen": 16748096,
      "step": 19335
    },
    {
      "epoch": 9.118340405469118,
      "grad_norm": 1.5012078285217285,
      "learning_rate": 2.629728755132267e-05,
      "loss": 0.205,
      "num_input_tokens_seen": 16752224,
      "step": 19340
    },
    {
      "epoch": 9.12069778406412,
      "grad_norm": 0.5697728991508484,
      "learning_rate": 2.628748320131081e-05,
      "loss": 0.1862,
      "num_input_tokens_seen": 16757056,
      "step": 19345
    },
    {
      "epoch": 9.123055162659123,
      "grad_norm": 1.1836498975753784,
      "learning_rate": 2.6277678652752856e-05,
      "loss": 0.1437,
      "num_input_tokens_seen": 16760768,
      "step": 19350
    },
    {
      "epoch": 9.125412541254125,
      "grad_norm": 0.8798470497131348,
      "learning_rate": 2.6267873907160807e-05,
      "loss": 0.1476,
      "num_input_tokens_seen": 16765152,
      "step": 19355
    },
    {
      "epoch": 9.127769919849127,
      "grad_norm": 1.370583176612854,
      "learning_rate": 2.6258068966046668e-05,
      "loss": 0.2457,
      "num_input_tokens_seen": 16769664,
      "step": 19360
    },
    {
      "epoch": 9.13012729844413,
      "grad_norm": 0.7587063908576965,
      "learning_rate": 2.6248263830922475e-05,
      "loss": 0.1271,
      "num_input_tokens_seen": 16774624,
      "step": 19365
    },
    {
      "epoch": 9.132484677039132,
      "grad_norm": 0.5157407522201538,
      "learning_rate": 2.6238458503300318e-05,
      "loss": 0.1515,
      "num_input_tokens_seen": 16778752,
      "step": 19370
    },
    {
      "epoch": 9.134842055634135,
      "grad_norm": 1.2588391304016113,
      "learning_rate": 2.6228652984692292e-05,
      "loss": 0.1408,
      "num_input_tokens_seen": 16783648,
      "step": 19375
    },
    {
      "epoch": 9.137199434229137,
      "grad_norm": 0.46188750863075256,
      "learning_rate": 2.621884727661054e-05,
      "loss": 0.1346,
      "num_input_tokens_seen": 16788640,
      "step": 19380
    },
    {
      "epoch": 9.13955681282414,
      "grad_norm": 0.9767963886260986,
      "learning_rate": 2.6209041380567222e-05,
      "loss": 0.1707,
      "num_input_tokens_seen": 16792560,
      "step": 19385
    },
    {
      "epoch": 9.141914191419142,
      "grad_norm": 0.9096600413322449,
      "learning_rate": 2.6199235298074527e-05,
      "loss": 0.144,
      "num_input_tokens_seen": 16796624,
      "step": 19390
    },
    {
      "epoch": 9.144271570014144,
      "grad_norm": 1.2202038764953613,
      "learning_rate": 2.618942903064468e-05,
      "loss": 0.2172,
      "num_input_tokens_seen": 16801712,
      "step": 19395
    },
    {
      "epoch": 9.146628948609147,
      "grad_norm": 0.6059379577636719,
      "learning_rate": 2.6179622579789932e-05,
      "loss": 0.2002,
      "num_input_tokens_seen": 16806304,
      "step": 19400
    },
    {
      "epoch": 9.146628948609147,
      "eval_loss": 0.2042151242494583,
      "eval_runtime": 22.1423,
      "eval_samples_per_second": 42.588,
      "eval_steps_per_second": 21.317,
      "num_input_tokens_seen": 16806304,
      "step": 19400
    },
    {
      "epoch": 9.148986327204149,
      "grad_norm": 0.8499491214752197,
      "learning_rate": 2.6169815947022553e-05,
      "loss": 0.2698,
      "num_input_tokens_seen": 16811440,
      "step": 19405
    },
    {
      "epoch": 9.151343705799151,
      "grad_norm": 0.572066605091095,
      "learning_rate": 2.6160009133854853e-05,
      "loss": 0.1475,
      "num_input_tokens_seen": 16814816,
      "step": 19410
    },
    {
      "epoch": 9.153701084394154,
      "grad_norm": 0.5535953640937805,
      "learning_rate": 2.6150202141799168e-05,
      "loss": 0.1953,
      "num_input_tokens_seen": 16818672,
      "step": 19415
    },
    {
      "epoch": 9.156058462989156,
      "grad_norm": 0.8425552248954773,
      "learning_rate": 2.614039497236786e-05,
      "loss": 0.1498,
      "num_input_tokens_seen": 16822880,
      "step": 19420
    },
    {
      "epoch": 9.158415841584159,
      "grad_norm": 0.7602952122688293,
      "learning_rate": 2.6130587627073315e-05,
      "loss": 0.1872,
      "num_input_tokens_seen": 16826544,
      "step": 19425
    },
    {
      "epoch": 9.160773220179161,
      "grad_norm": 0.6114056706428528,
      "learning_rate": 2.6120780107427956e-05,
      "loss": 0.1832,
      "num_input_tokens_seen": 16831024,
      "step": 19430
    },
    {
      "epoch": 9.163130598774163,
      "grad_norm": 0.40623730421066284,
      "learning_rate": 2.6110972414944214e-05,
      "loss": 0.1871,
      "num_input_tokens_seen": 16835632,
      "step": 19435
    },
    {
      "epoch": 9.165487977369166,
      "grad_norm": 0.5156939029693604,
      "learning_rate": 2.6101164551134565e-05,
      "loss": 0.205,
      "num_input_tokens_seen": 16839744,
      "step": 19440
    },
    {
      "epoch": 9.167845355964168,
      "grad_norm": 2.0449507236480713,
      "learning_rate": 2.6091356517511505e-05,
      "loss": 0.2021,
      "num_input_tokens_seen": 16844256,
      "step": 19445
    },
    {
      "epoch": 9.17020273455917,
      "grad_norm": 0.4425397217273712,
      "learning_rate": 2.608154831558755e-05,
      "loss": 0.1198,
      "num_input_tokens_seen": 16848400,
      "step": 19450
    },
    {
      "epoch": 9.172560113154173,
      "grad_norm": 0.5370048880577087,
      "learning_rate": 2.607173994687526e-05,
      "loss": 0.1947,
      "num_input_tokens_seen": 16852064,
      "step": 19455
    },
    {
      "epoch": 9.174917491749175,
      "grad_norm": 0.831891655921936,
      "learning_rate": 2.6061931412887196e-05,
      "loss": 0.2159,
      "num_input_tokens_seen": 16856528,
      "step": 19460
    },
    {
      "epoch": 9.177274870344178,
      "grad_norm": 0.902950644493103,
      "learning_rate": 2.6052122715135973e-05,
      "loss": 0.1694,
      "num_input_tokens_seen": 16860736,
      "step": 19465
    },
    {
      "epoch": 9.17963224893918,
      "grad_norm": 1.3158718347549438,
      "learning_rate": 2.60423138551342e-05,
      "loss": 0.2386,
      "num_input_tokens_seen": 16864704,
      "step": 19470
    },
    {
      "epoch": 9.181989627534183,
      "grad_norm": 1.6643859148025513,
      "learning_rate": 2.6032504834394527e-05,
      "loss": 0.1963,
      "num_input_tokens_seen": 16869536,
      "step": 19475
    },
    {
      "epoch": 9.184347006129185,
      "grad_norm": 0.7435295581817627,
      "learning_rate": 2.602269565442964e-05,
      "loss": 0.2095,
      "num_input_tokens_seen": 16874416,
      "step": 19480
    },
    {
      "epoch": 9.186704384724187,
      "grad_norm": 0.9442761540412903,
      "learning_rate": 2.6012886316752227e-05,
      "loss": 0.1562,
      "num_input_tokens_seen": 16878800,
      "step": 19485
    },
    {
      "epoch": 9.18906176331919,
      "grad_norm": 1.2549952268600464,
      "learning_rate": 2.6003076822875018e-05,
      "loss": 0.2672,
      "num_input_tokens_seen": 16883856,
      "step": 19490
    },
    {
      "epoch": 9.191419141914192,
      "grad_norm": 1.3542077541351318,
      "learning_rate": 2.5993267174310755e-05,
      "loss": 0.2006,
      "num_input_tokens_seen": 16887808,
      "step": 19495
    },
    {
      "epoch": 9.193776520509195,
      "grad_norm": 1.6403788328170776,
      "learning_rate": 2.5983457372572218e-05,
      "loss": 0.135,
      "num_input_tokens_seen": 16891536,
      "step": 19500
    },
    {
      "epoch": 9.196133899104197,
      "grad_norm": 0.9198794364929199,
      "learning_rate": 2.597364741917219e-05,
      "loss": 0.1418,
      "num_input_tokens_seen": 16896192,
      "step": 19505
    },
    {
      "epoch": 9.198491277699198,
      "grad_norm": 1.0654475688934326,
      "learning_rate": 2.5963837315623492e-05,
      "loss": 0.181,
      "num_input_tokens_seen": 16900880,
      "step": 19510
    },
    {
      "epoch": 9.2008486562942,
      "grad_norm": 1.6711674928665161,
      "learning_rate": 2.595402706343897e-05,
      "loss": 0.201,
      "num_input_tokens_seen": 16904832,
      "step": 19515
    },
    {
      "epoch": 9.203206034889202,
      "grad_norm": 0.6464036107063293,
      "learning_rate": 2.594421666413148e-05,
      "loss": 0.1944,
      "num_input_tokens_seen": 16909552,
      "step": 19520
    },
    {
      "epoch": 9.205563413484205,
      "grad_norm": 0.4867481291294098,
      "learning_rate": 2.5934406119213928e-05,
      "loss": 0.2437,
      "num_input_tokens_seen": 16913872,
      "step": 19525
    },
    {
      "epoch": 9.207920792079207,
      "grad_norm": 0.5692402720451355,
      "learning_rate": 2.5924595430199193e-05,
      "loss": 0.1947,
      "num_input_tokens_seen": 16918176,
      "step": 19530
    },
    {
      "epoch": 9.21027817067421,
      "grad_norm": 0.7507001757621765,
      "learning_rate": 2.5914784598600238e-05,
      "loss": 0.1784,
      "num_input_tokens_seen": 16922144,
      "step": 19535
    },
    {
      "epoch": 9.212635549269212,
      "grad_norm": 0.6595105528831482,
      "learning_rate": 2.5904973625930002e-05,
      "loss": 0.1897,
      "num_input_tokens_seen": 16926752,
      "step": 19540
    },
    {
      "epoch": 9.214992927864214,
      "grad_norm": 0.6201308965682983,
      "learning_rate": 2.5895162513701456e-05,
      "loss": 0.1949,
      "num_input_tokens_seen": 16931168,
      "step": 19545
    },
    {
      "epoch": 9.217350306459217,
      "grad_norm": 0.5910993218421936,
      "learning_rate": 2.5885351263427593e-05,
      "loss": 0.2301,
      "num_input_tokens_seen": 16935440,
      "step": 19550
    },
    {
      "epoch": 9.21970768505422,
      "grad_norm": 2.130715847015381,
      "learning_rate": 2.5875539876621448e-05,
      "loss": 0.2412,
      "num_input_tokens_seen": 16940512,
      "step": 19555
    },
    {
      "epoch": 9.222065063649222,
      "grad_norm": 0.5977562069892883,
      "learning_rate": 2.586572835479605e-05,
      "loss": 0.1862,
      "num_input_tokens_seen": 16944688,
      "step": 19560
    },
    {
      "epoch": 9.224422442244224,
      "grad_norm": 1.2911708354949951,
      "learning_rate": 2.585591669946446e-05,
      "loss": 0.1685,
      "num_input_tokens_seen": 16948416,
      "step": 19565
    },
    {
      "epoch": 9.226779820839226,
      "grad_norm": 0.7388254404067993,
      "learning_rate": 2.5846104912139756e-05,
      "loss": 0.1663,
      "num_input_tokens_seen": 16952800,
      "step": 19570
    },
    {
      "epoch": 9.229137199434229,
      "grad_norm": 0.7175175547599792,
      "learning_rate": 2.583629299433505e-05,
      "loss": 0.1412,
      "num_input_tokens_seen": 16956032,
      "step": 19575
    },
    {
      "epoch": 9.231494578029231,
      "grad_norm": 0.7116897106170654,
      "learning_rate": 2.582648094756345e-05,
      "loss": 0.1387,
      "num_input_tokens_seen": 16961152,
      "step": 19580
    },
    {
      "epoch": 9.233851956624234,
      "grad_norm": 1.1774417161941528,
      "learning_rate": 2.5816668773338098e-05,
      "loss": 0.21,
      "num_input_tokens_seen": 16965792,
      "step": 19585
    },
    {
      "epoch": 9.236209335219236,
      "grad_norm": 1.386590600013733,
      "learning_rate": 2.580685647317216e-05,
      "loss": 0.1463,
      "num_input_tokens_seen": 16971008,
      "step": 19590
    },
    {
      "epoch": 9.238566713814238,
      "grad_norm": 1.862434983253479,
      "learning_rate": 2.5797044048578818e-05,
      "loss": 0.2331,
      "num_input_tokens_seen": 16974304,
      "step": 19595
    },
    {
      "epoch": 9.24092409240924,
      "grad_norm": 1.481911301612854,
      "learning_rate": 2.5787231501071262e-05,
      "loss": 0.1987,
      "num_input_tokens_seen": 16979072,
      "step": 19600
    },
    {
      "epoch": 9.24092409240924,
      "eval_loss": 0.20148968696594238,
      "eval_runtime": 22.1641,
      "eval_samples_per_second": 42.546,
      "eval_steps_per_second": 21.296,
      "num_input_tokens_seen": 16979072,
      "step": 19600
    },
    {
      "epoch": 9.243281471004243,
      "grad_norm": 1.0415207147598267,
      "learning_rate": 2.577741883216272e-05,
      "loss": 0.2242,
      "num_input_tokens_seen": 16983264,
      "step": 19605
    },
    {
      "epoch": 9.245638849599246,
      "grad_norm": 1.2722415924072266,
      "learning_rate": 2.576760604336642e-05,
      "loss": 0.2971,
      "num_input_tokens_seen": 16987328,
      "step": 19610
    },
    {
      "epoch": 9.247996228194248,
      "grad_norm": 1.4834622144699097,
      "learning_rate": 2.575779313619563e-05,
      "loss": 0.2162,
      "num_input_tokens_seen": 16991472,
      "step": 19615
    },
    {
      "epoch": 9.25035360678925,
      "grad_norm": 0.7930892705917358,
      "learning_rate": 2.5747980112163605e-05,
      "loss": 0.2382,
      "num_input_tokens_seen": 16995536,
      "step": 19620
    },
    {
      "epoch": 9.252710985384253,
      "grad_norm": 0.705223798751831,
      "learning_rate": 2.5738166972783656e-05,
      "loss": 0.2284,
      "num_input_tokens_seen": 16999216,
      "step": 19625
    },
    {
      "epoch": 9.255068363979255,
      "grad_norm": 1.1515226364135742,
      "learning_rate": 2.5728353719569075e-05,
      "loss": 0.2552,
      "num_input_tokens_seen": 17003168,
      "step": 19630
    },
    {
      "epoch": 9.257425742574258,
      "grad_norm": 0.6388946175575256,
      "learning_rate": 2.57185403540332e-05,
      "loss": 0.1692,
      "num_input_tokens_seen": 17007056,
      "step": 19635
    },
    {
      "epoch": 9.25978312116926,
      "grad_norm": 1.890320062637329,
      "learning_rate": 2.5708726877689375e-05,
      "loss": 0.2091,
      "num_input_tokens_seen": 17011008,
      "step": 19640
    },
    {
      "epoch": 9.262140499764262,
      "grad_norm": 0.6054696440696716,
      "learning_rate": 2.5698913292050964e-05,
      "loss": 0.1891,
      "num_input_tokens_seen": 17014976,
      "step": 19645
    },
    {
      "epoch": 9.264497878359265,
      "grad_norm": 0.9542998671531677,
      "learning_rate": 2.568909959863133e-05,
      "loss": 0.1836,
      "num_input_tokens_seen": 17019920,
      "step": 19650
    },
    {
      "epoch": 9.266855256954267,
      "grad_norm": 0.7260638475418091,
      "learning_rate": 2.5679285798943887e-05,
      "loss": 0.198,
      "num_input_tokens_seen": 17024112,
      "step": 19655
    },
    {
      "epoch": 9.26921263554927,
      "grad_norm": 0.7107857465744019,
      "learning_rate": 2.5669471894502035e-05,
      "loss": 0.2362,
      "num_input_tokens_seen": 17028208,
      "step": 19660
    },
    {
      "epoch": 9.271570014144272,
      "grad_norm": 1.0902818441390991,
      "learning_rate": 2.56596578868192e-05,
      "loss": 0.1771,
      "num_input_tokens_seen": 17032288,
      "step": 19665
    },
    {
      "epoch": 9.273927392739274,
      "grad_norm": 0.5030151605606079,
      "learning_rate": 2.564984377740883e-05,
      "loss": 0.1208,
      "num_input_tokens_seen": 17036864,
      "step": 19670
    },
    {
      "epoch": 9.276284771334277,
      "grad_norm": 1.8172271251678467,
      "learning_rate": 2.564002956778438e-05,
      "loss": 0.1541,
      "num_input_tokens_seen": 17040512,
      "step": 19675
    },
    {
      "epoch": 9.27864214992928,
      "grad_norm": 0.6605421304702759,
      "learning_rate": 2.563021525945934e-05,
      "loss": 0.1711,
      "num_input_tokens_seen": 17044544,
      "step": 19680
    },
    {
      "epoch": 9.280999528524282,
      "grad_norm": 0.8599736094474792,
      "learning_rate": 2.562040085394718e-05,
      "loss": 0.1615,
      "num_input_tokens_seen": 17048800,
      "step": 19685
    },
    {
      "epoch": 9.283356907119284,
      "grad_norm": 1.821950912475586,
      "learning_rate": 2.56105863527614e-05,
      "loss": 0.1858,
      "num_input_tokens_seen": 17053232,
      "step": 19690
    },
    {
      "epoch": 9.285714285714286,
      "grad_norm": 0.34317079186439514,
      "learning_rate": 2.5600771757415548e-05,
      "loss": 0.1106,
      "num_input_tokens_seen": 17057328,
      "step": 19695
    },
    {
      "epoch": 9.288071664309289,
      "grad_norm": 0.6618346571922302,
      "learning_rate": 2.5590957069423134e-05,
      "loss": 0.1684,
      "num_input_tokens_seen": 17062208,
      "step": 19700
    },
    {
      "epoch": 9.290429042904291,
      "grad_norm": 1.6132780313491821,
      "learning_rate": 2.5581142290297716e-05,
      "loss": 0.1587,
      "num_input_tokens_seen": 17067200,
      "step": 19705
    },
    {
      "epoch": 9.292786421499294,
      "grad_norm": 1.1150283813476562,
      "learning_rate": 2.557132742155285e-05,
      "loss": 0.1565,
      "num_input_tokens_seen": 17071200,
      "step": 19710
    },
    {
      "epoch": 9.295143800094294,
      "grad_norm": 0.5788206458091736,
      "learning_rate": 2.556151246470212e-05,
      "loss": 0.1449,
      "num_input_tokens_seen": 17075408,
      "step": 19715
    },
    {
      "epoch": 9.297501178689297,
      "grad_norm": 0.4933871924877167,
      "learning_rate": 2.5551697421259114e-05,
      "loss": 0.2239,
      "num_input_tokens_seen": 17079328,
      "step": 19720
    },
    {
      "epoch": 9.299858557284299,
      "grad_norm": 0.8999598622322083,
      "learning_rate": 2.554188229273743e-05,
      "loss": 0.2284,
      "num_input_tokens_seen": 17083808,
      "step": 19725
    },
    {
      "epoch": 9.302215935879302,
      "grad_norm": 0.9333539009094238,
      "learning_rate": 2.5532067080650678e-05,
      "loss": 0.1772,
      "num_input_tokens_seen": 17088944,
      "step": 19730
    },
    {
      "epoch": 9.304573314474304,
      "grad_norm": 2.219184637069702,
      "learning_rate": 2.55222517865125e-05,
      "loss": 0.1442,
      "num_input_tokens_seen": 17093200,
      "step": 19735
    },
    {
      "epoch": 9.306930693069306,
      "grad_norm": 0.7643903493881226,
      "learning_rate": 2.5512436411836538e-05,
      "loss": 0.1892,
      "num_input_tokens_seen": 17097712,
      "step": 19740
    },
    {
      "epoch": 9.309288071664309,
      "grad_norm": 0.5340753197669983,
      "learning_rate": 2.5502620958136443e-05,
      "loss": 0.1917,
      "num_input_tokens_seen": 17102736,
      "step": 19745
    },
    {
      "epoch": 9.311645450259311,
      "grad_norm": 0.312745600938797,
      "learning_rate": 2.5492805426925874e-05,
      "loss": 0.1397,
      "num_input_tokens_seen": 17106240,
      "step": 19750
    },
    {
      "epoch": 9.314002828854314,
      "grad_norm": 1.5106258392333984,
      "learning_rate": 2.5482989819718523e-05,
      "loss": 0.1909,
      "num_input_tokens_seen": 17110816,
      "step": 19755
    },
    {
      "epoch": 9.316360207449316,
      "grad_norm": 0.9103663563728333,
      "learning_rate": 2.5473174138028065e-05,
      "loss": 0.2242,
      "num_input_tokens_seen": 17114848,
      "step": 19760
    },
    {
      "epoch": 9.318717586044318,
      "grad_norm": 0.7551849484443665,
      "learning_rate": 2.5463358383368212e-05,
      "loss": 0.2182,
      "num_input_tokens_seen": 17119488,
      "step": 19765
    },
    {
      "epoch": 9.32107496463932,
      "grad_norm": 1.4828991889953613,
      "learning_rate": 2.545354255725267e-05,
      "loss": 0.2089,
      "num_input_tokens_seen": 17123888,
      "step": 19770
    },
    {
      "epoch": 9.323432343234323,
      "grad_norm": 1.5084336996078491,
      "learning_rate": 2.5443726661195165e-05,
      "loss": 0.1651,
      "num_input_tokens_seen": 17128144,
      "step": 19775
    },
    {
      "epoch": 9.325789721829326,
      "grad_norm": 0.9798581600189209,
      "learning_rate": 2.543391069670944e-05,
      "loss": 0.2211,
      "num_input_tokens_seen": 17131776,
      "step": 19780
    },
    {
      "epoch": 9.328147100424328,
      "grad_norm": 1.516592264175415,
      "learning_rate": 2.5424094665309228e-05,
      "loss": 0.1919,
      "num_input_tokens_seen": 17135792,
      "step": 19785
    },
    {
      "epoch": 9.33050447901933,
      "grad_norm": 0.485762357711792,
      "learning_rate": 2.5414278568508292e-05,
      "loss": 0.1786,
      "num_input_tokens_seen": 17140272,
      "step": 19790
    },
    {
      "epoch": 9.332861857614333,
      "grad_norm": 0.9286555647850037,
      "learning_rate": 2.540446240782039e-05,
      "loss": 0.1611,
      "num_input_tokens_seen": 17145216,
      "step": 19795
    },
    {
      "epoch": 9.335219236209335,
      "grad_norm": 1.7095831632614136,
      "learning_rate": 2.5394646184759307e-05,
      "loss": 0.2362,
      "num_input_tokens_seen": 17150160,
      "step": 19800
    },
    {
      "epoch": 9.335219236209335,
      "eval_loss": 0.2025989443063736,
      "eval_runtime": 22.1542,
      "eval_samples_per_second": 42.565,
      "eval_steps_per_second": 21.305,
      "num_input_tokens_seen": 17150160,
      "step": 19800
    },
    {
      "epoch": 9.337576614804338,
      "grad_norm": 0.523276686668396,
      "learning_rate": 2.538482990083882e-05,
      "loss": 0.1431,
      "num_input_tokens_seen": 17154256,
      "step": 19805
    },
    {
      "epoch": 9.33993399339934,
      "grad_norm": 0.7110109329223633,
      "learning_rate": 2.5375013557572725e-05,
      "loss": 0.1593,
      "num_input_tokens_seen": 17157808,
      "step": 19810
    },
    {
      "epoch": 9.342291371994342,
      "grad_norm": 0.7661017775535583,
      "learning_rate": 2.536519715647483e-05,
      "loss": 0.1638,
      "num_input_tokens_seen": 17162048,
      "step": 19815
    },
    {
      "epoch": 9.344648750589345,
      "grad_norm": 1.0025845766067505,
      "learning_rate": 2.535538069905894e-05,
      "loss": 0.1616,
      "num_input_tokens_seen": 17166272,
      "step": 19820
    },
    {
      "epoch": 9.347006129184347,
      "grad_norm": 1.141798973083496,
      "learning_rate": 2.534556418683888e-05,
      "loss": 0.1609,
      "num_input_tokens_seen": 17170272,
      "step": 19825
    },
    {
      "epoch": 9.34936350777935,
      "grad_norm": 1.5038336515426636,
      "learning_rate": 2.5335747621328486e-05,
      "loss": 0.1905,
      "num_input_tokens_seen": 17174800,
      "step": 19830
    },
    {
      "epoch": 9.351720886374352,
      "grad_norm": 0.940314531326294,
      "learning_rate": 2.5325931004041586e-05,
      "loss": 0.2378,
      "num_input_tokens_seen": 17178688,
      "step": 19835
    },
    {
      "epoch": 9.354078264969354,
      "grad_norm": 0.8697491884231567,
      "learning_rate": 2.5316114336492032e-05,
      "loss": 0.1712,
      "num_input_tokens_seen": 17182720,
      "step": 19840
    },
    {
      "epoch": 9.356435643564357,
      "grad_norm": 0.9646694660186768,
      "learning_rate": 2.530629762019367e-05,
      "loss": 0.2208,
      "num_input_tokens_seen": 17186480,
      "step": 19845
    },
    {
      "epoch": 9.35879302215936,
      "grad_norm": 0.6029452681541443,
      "learning_rate": 2.5296480856660364e-05,
      "loss": 0.2146,
      "num_input_tokens_seen": 17190896,
      "step": 19850
    },
    {
      "epoch": 9.361150400754362,
      "grad_norm": 1.1855802536010742,
      "learning_rate": 2.528666404740599e-05,
      "loss": 0.1968,
      "num_input_tokens_seen": 17195200,
      "step": 19855
    },
    {
      "epoch": 9.363507779349364,
      "grad_norm": 0.8825040459632874,
      "learning_rate": 2.527684719394442e-05,
      "loss": 0.1748,
      "num_input_tokens_seen": 17198880,
      "step": 19860
    },
    {
      "epoch": 9.365865157944366,
      "grad_norm": 0.8096519708633423,
      "learning_rate": 2.526703029778953e-05,
      "loss": 0.1482,
      "num_input_tokens_seen": 17203632,
      "step": 19865
    },
    {
      "epoch": 9.368222536539369,
      "grad_norm": 0.9918960928916931,
      "learning_rate": 2.5257213360455208e-05,
      "loss": 0.2402,
      "num_input_tokens_seen": 17208496,
      "step": 19870
    },
    {
      "epoch": 9.370579915134371,
      "grad_norm": 1.3650832176208496,
      "learning_rate": 2.5247396383455353e-05,
      "loss": 0.1861,
      "num_input_tokens_seen": 17212576,
      "step": 19875
    },
    {
      "epoch": 9.372937293729374,
      "grad_norm": 0.775772750377655,
      "learning_rate": 2.523757936830387e-05,
      "loss": 0.1691,
      "num_input_tokens_seen": 17216288,
      "step": 19880
    },
    {
      "epoch": 9.375294672324376,
      "grad_norm": 1.9551000595092773,
      "learning_rate": 2.5227762316514662e-05,
      "loss": 0.19,
      "num_input_tokens_seen": 17220560,
      "step": 19885
    },
    {
      "epoch": 9.377652050919378,
      "grad_norm": 0.5646829009056091,
      "learning_rate": 2.5217945229601648e-05,
      "loss": 0.1746,
      "num_input_tokens_seen": 17224368,
      "step": 19890
    },
    {
      "epoch": 9.38000942951438,
      "grad_norm": 2.0658979415893555,
      "learning_rate": 2.5208128109078738e-05,
      "loss": 0.1512,
      "num_input_tokens_seen": 17228464,
      "step": 19895
    },
    {
      "epoch": 9.382366808109383,
      "grad_norm": 1.7181193828582764,
      "learning_rate": 2.5198310956459853e-05,
      "loss": 0.1958,
      "num_input_tokens_seen": 17233552,
      "step": 19900
    },
    {
      "epoch": 9.384724186704386,
      "grad_norm": 0.6920815110206604,
      "learning_rate": 2.518849377325893e-05,
      "loss": 0.28,
      "num_input_tokens_seen": 17237536,
      "step": 19905
    },
    {
      "epoch": 9.387081565299386,
      "grad_norm": 0.42866626381874084,
      "learning_rate": 2.51786765609899e-05,
      "loss": 0.1546,
      "num_input_tokens_seen": 17242048,
      "step": 19910
    },
    {
      "epoch": 9.389438943894389,
      "grad_norm": 1.0182790756225586,
      "learning_rate": 2.5168859321166694e-05,
      "loss": 0.2144,
      "num_input_tokens_seen": 17246000,
      "step": 19915
    },
    {
      "epoch": 9.391796322489391,
      "grad_norm": 1.2448638677597046,
      "learning_rate": 2.515904205530326e-05,
      "loss": 0.1675,
      "num_input_tokens_seen": 17250096,
      "step": 19920
    },
    {
      "epoch": 9.394153701084393,
      "grad_norm": 0.8946560621261597,
      "learning_rate": 2.514922476491355e-05,
      "loss": 0.2195,
      "num_input_tokens_seen": 17255152,
      "step": 19925
    },
    {
      "epoch": 9.396511079679396,
      "grad_norm": 0.7409642338752747,
      "learning_rate": 2.51394074515115e-05,
      "loss": 0.2189,
      "num_input_tokens_seen": 17260128,
      "step": 19930
    },
    {
      "epoch": 9.398868458274398,
      "grad_norm": 1.0027409791946411,
      "learning_rate": 2.5129590116611067e-05,
      "loss": 0.2046,
      "num_input_tokens_seen": 17264064,
      "step": 19935
    },
    {
      "epoch": 9.4012258368694,
      "grad_norm": 1.8133609294891357,
      "learning_rate": 2.5119772761726212e-05,
      "loss": 0.1665,
      "num_input_tokens_seen": 17268896,
      "step": 19940
    },
    {
      "epoch": 9.403583215464403,
      "grad_norm": 0.6119510531425476,
      "learning_rate": 2.5109955388370893e-05,
      "loss": 0.1214,
      "num_input_tokens_seen": 17273504,
      "step": 19945
    },
    {
      "epoch": 9.405940594059405,
      "grad_norm": 0.7563053369522095,
      "learning_rate": 2.510013799805907e-05,
      "loss": 0.1425,
      "num_input_tokens_seen": 17278608,
      "step": 19950
    },
    {
      "epoch": 9.408297972654408,
      "grad_norm": 1.5319708585739136,
      "learning_rate": 2.5090320592304706e-05,
      "loss": 0.1943,
      "num_input_tokens_seen": 17282800,
      "step": 19955
    },
    {
      "epoch": 9.41065535124941,
      "grad_norm": 1.0833925008773804,
      "learning_rate": 2.5080503172621777e-05,
      "loss": 0.1724,
      "num_input_tokens_seen": 17286608,
      "step": 19960
    },
    {
      "epoch": 9.413012729844413,
      "grad_norm": 0.2967025935649872,
      "learning_rate": 2.5070685740524246e-05,
      "loss": 0.1502,
      "num_input_tokens_seen": 17290912,
      "step": 19965
    },
    {
      "epoch": 9.415370108439415,
      "grad_norm": 1.635654330253601,
      "learning_rate": 2.5060868297526084e-05,
      "loss": 0.2097,
      "num_input_tokens_seen": 17294976,
      "step": 19970
    },
    {
      "epoch": 9.417727487034417,
      "grad_norm": 0.8789593577384949,
      "learning_rate": 2.5051050845141267e-05,
      "loss": 0.1456,
      "num_input_tokens_seen": 17299744,
      "step": 19975
    },
    {
      "epoch": 9.42008486562942,
      "grad_norm": 0.8931361436843872,
      "learning_rate": 2.5041233384883765e-05,
      "loss": 0.1767,
      "num_input_tokens_seen": 17303296,
      "step": 19980
    },
    {
      "epoch": 9.422442244224422,
      "grad_norm": 1.3024530410766602,
      "learning_rate": 2.5031415918267564e-05,
      "loss": 0.133,
      "num_input_tokens_seen": 17307472,
      "step": 19985
    },
    {
      "epoch": 9.424799622819425,
      "grad_norm": 0.7291197776794434,
      "learning_rate": 2.5021598446806626e-05,
      "loss": 0.1548,
      "num_input_tokens_seen": 17312656,
      "step": 19990
    },
    {
      "epoch": 9.427157001414427,
      "grad_norm": 0.4529426395893097,
      "learning_rate": 2.5011780972014937e-05,
      "loss": 0.1471,
      "num_input_tokens_seen": 17316400,
      "step": 19995
    },
    {
      "epoch": 9.42951438000943,
      "grad_norm": 0.9316900968551636,
      "learning_rate": 2.5001963495406478e-05,
      "loss": 0.2449,
      "num_input_tokens_seen": 17321280,
      "step": 20000
    },
    {
      "epoch": 9.42951438000943,
      "eval_loss": 0.20473915338516235,
      "eval_runtime": 22.1503,
      "eval_samples_per_second": 42.573,
      "eval_steps_per_second": 21.309,
      "num_input_tokens_seen": 17321280,
      "step": 20000
    },
    {
      "epoch": 9.431871758604432,
      "grad_norm": 1.9256911277770996,
      "learning_rate": 2.499214601849522e-05,
      "loss": 0.1757,
      "num_input_tokens_seen": 17325056,
      "step": 20005
    },
    {
      "epoch": 9.434229137199434,
      "grad_norm": 1.7285070419311523,
      "learning_rate": 2.4982328542795148e-05,
      "loss": 0.1384,
      "num_input_tokens_seen": 17329088,
      "step": 20010
    },
    {
      "epoch": 9.436586515794437,
      "grad_norm": 1.2826825380325317,
      "learning_rate": 2.497251106982024e-05,
      "loss": 0.2461,
      "num_input_tokens_seen": 17333536,
      "step": 20015
    },
    {
      "epoch": 9.438943894389439,
      "grad_norm": 0.965855598449707,
      "learning_rate": 2.4962693601084458e-05,
      "loss": 0.2047,
      "num_input_tokens_seen": 17338048,
      "step": 20020
    },
    {
      "epoch": 9.441301272984441,
      "grad_norm": 0.5866786241531372,
      "learning_rate": 2.4952876138101794e-05,
      "loss": 0.117,
      "num_input_tokens_seen": 17341888,
      "step": 20025
    },
    {
      "epoch": 9.443658651579444,
      "grad_norm": 1.3407502174377441,
      "learning_rate": 2.4943058682386233e-05,
      "loss": 0.1846,
      "num_input_tokens_seen": 17346544,
      "step": 20030
    },
    {
      "epoch": 9.446016030174446,
      "grad_norm": 0.8944697380065918,
      "learning_rate": 2.493324123545173e-05,
      "loss": 0.1459,
      "num_input_tokens_seen": 17351072,
      "step": 20035
    },
    {
      "epoch": 9.448373408769449,
      "grad_norm": 1.0880333185195923,
      "learning_rate": 2.4923423798812272e-05,
      "loss": 0.1463,
      "num_input_tokens_seen": 17354960,
      "step": 20040
    },
    {
      "epoch": 9.450730787364451,
      "grad_norm": 0.8322956562042236,
      "learning_rate": 2.4913606373981825e-05,
      "loss": 0.1207,
      "num_input_tokens_seen": 17359376,
      "step": 20045
    },
    {
      "epoch": 9.453088165959453,
      "grad_norm": 0.9240128993988037,
      "learning_rate": 2.4903788962474357e-05,
      "loss": 0.2037,
      "num_input_tokens_seen": 17364336,
      "step": 20050
    },
    {
      "epoch": 9.455445544554456,
      "grad_norm": 0.6645311713218689,
      "learning_rate": 2.489397156580385e-05,
      "loss": 0.1797,
      "num_input_tokens_seen": 17368992,
      "step": 20055
    },
    {
      "epoch": 9.457802923149458,
      "grad_norm": 0.9431477785110474,
      "learning_rate": 2.4884154185484246e-05,
      "loss": 0.1605,
      "num_input_tokens_seen": 17373712,
      "step": 20060
    },
    {
      "epoch": 9.46016030174446,
      "grad_norm": 0.47270095348358154,
      "learning_rate": 2.4874336823029526e-05,
      "loss": 0.1356,
      "num_input_tokens_seen": 17378304,
      "step": 20065
    },
    {
      "epoch": 9.462517680339463,
      "grad_norm": 0.5390760898590088,
      "learning_rate": 2.4864519479953656e-05,
      "loss": 0.141,
      "num_input_tokens_seen": 17382928,
      "step": 20070
    },
    {
      "epoch": 9.464875058934465,
      "grad_norm": 0.9104752540588379,
      "learning_rate": 2.485470215777058e-05,
      "loss": 0.1692,
      "num_input_tokens_seen": 17387024,
      "step": 20075
    },
    {
      "epoch": 9.467232437529468,
      "grad_norm": 1.2182921171188354,
      "learning_rate": 2.4844884857994258e-05,
      "loss": 0.2187,
      "num_input_tokens_seen": 17391056,
      "step": 20080
    },
    {
      "epoch": 9.46958981612447,
      "grad_norm": 1.188805103302002,
      "learning_rate": 2.4835067582138638e-05,
      "loss": 0.1927,
      "num_input_tokens_seen": 17394944,
      "step": 20085
    },
    {
      "epoch": 9.471947194719473,
      "grad_norm": 1.1192973852157593,
      "learning_rate": 2.4825250331717666e-05,
      "loss": 0.1938,
      "num_input_tokens_seen": 17398608,
      "step": 20090
    },
    {
      "epoch": 9.474304573314475,
      "grad_norm": 0.5724523067474365,
      "learning_rate": 2.4815433108245298e-05,
      "loss": 0.2118,
      "num_input_tokens_seen": 17402736,
      "step": 20095
    },
    {
      "epoch": 9.476661951909477,
      "grad_norm": 0.6860583424568176,
      "learning_rate": 2.4805615913235456e-05,
      "loss": 0.1698,
      "num_input_tokens_seen": 17406784,
      "step": 20100
    },
    {
      "epoch": 9.47901933050448,
      "grad_norm": 0.7176621556282043,
      "learning_rate": 2.479579874820208e-05,
      "loss": 0.1427,
      "num_input_tokens_seen": 17411296,
      "step": 20105
    },
    {
      "epoch": 9.481376709099482,
      "grad_norm": 0.7922742962837219,
      "learning_rate": 2.4785981614659115e-05,
      "loss": 0.2296,
      "num_input_tokens_seen": 17415616,
      "step": 20110
    },
    {
      "epoch": 9.483734087694483,
      "grad_norm": 2.715517282485962,
      "learning_rate": 2.477616451412047e-05,
      "loss": 0.2574,
      "num_input_tokens_seen": 17420560,
      "step": 20115
    },
    {
      "epoch": 9.486091466289485,
      "grad_norm": 0.4724370241165161,
      "learning_rate": 2.476634744810007e-05,
      "loss": 0.1843,
      "num_input_tokens_seen": 17424816,
      "step": 20120
    },
    {
      "epoch": 9.488448844884488,
      "grad_norm": 0.541308581829071,
      "learning_rate": 2.475653041811183e-05,
      "loss": 0.1146,
      "num_input_tokens_seen": 17429344,
      "step": 20125
    },
    {
      "epoch": 9.49080622347949,
      "grad_norm": 0.37113627791404724,
      "learning_rate": 2.4746713425669652e-05,
      "loss": 0.2069,
      "num_input_tokens_seen": 17434160,
      "step": 20130
    },
    {
      "epoch": 9.493163602074493,
      "grad_norm": 2.1231484413146973,
      "learning_rate": 2.4736896472287458e-05,
      "loss": 0.2853,
      "num_input_tokens_seen": 17438656,
      "step": 20135
    },
    {
      "epoch": 9.495520980669495,
      "grad_norm": 1.1171936988830566,
      "learning_rate": 2.4727079559479124e-05,
      "loss": 0.2528,
      "num_input_tokens_seen": 17443888,
      "step": 20140
    },
    {
      "epoch": 9.497878359264497,
      "grad_norm": 0.8014938831329346,
      "learning_rate": 2.4717262688758557e-05,
      "loss": 0.2198,
      "num_input_tokens_seen": 17448336,
      "step": 20145
    },
    {
      "epoch": 9.5002357378595,
      "grad_norm": 0.5937250256538391,
      "learning_rate": 2.4707445861639637e-05,
      "loss": 0.1599,
      "num_input_tokens_seen": 17452928,
      "step": 20150
    },
    {
      "epoch": 9.502593116454502,
      "grad_norm": 0.7667165398597717,
      "learning_rate": 2.4697629079636244e-05,
      "loss": 0.2035,
      "num_input_tokens_seen": 17456848,
      "step": 20155
    },
    {
      "epoch": 9.504950495049505,
      "grad_norm": 1.6761484146118164,
      "learning_rate": 2.4687812344262244e-05,
      "loss": 0.2122,
      "num_input_tokens_seen": 17460656,
      "step": 20160
    },
    {
      "epoch": 9.507307873644507,
      "grad_norm": 1.1838117837905884,
      "learning_rate": 2.46779956570315e-05,
      "loss": 0.2287,
      "num_input_tokens_seen": 17464528,
      "step": 20165
    },
    {
      "epoch": 9.50966525223951,
      "grad_norm": 0.5876703262329102,
      "learning_rate": 2.466817901945787e-05,
      "loss": 0.1552,
      "num_input_tokens_seen": 17468544,
      "step": 20170
    },
    {
      "epoch": 9.512022630834512,
      "grad_norm": 1.0037246942520142,
      "learning_rate": 2.4658362433055217e-05,
      "loss": 0.1519,
      "num_input_tokens_seen": 17472576,
      "step": 20175
    },
    {
      "epoch": 9.514380009429514,
      "grad_norm": 1.0766208171844482,
      "learning_rate": 2.4648545899337356e-05,
      "loss": 0.2403,
      "num_input_tokens_seen": 17476688,
      "step": 20180
    },
    {
      "epoch": 9.516737388024517,
      "grad_norm": 0.5811607837677002,
      "learning_rate": 2.4638729419818143e-05,
      "loss": 0.2104,
      "num_input_tokens_seen": 17481872,
      "step": 20185
    },
    {
      "epoch": 9.519094766619519,
      "grad_norm": 1.322239875793457,
      "learning_rate": 2.46289129960114e-05,
      "loss": 0.1722,
      "num_input_tokens_seen": 17486224,
      "step": 20190
    },
    {
      "epoch": 9.521452145214521,
      "grad_norm": 1.1064900159835815,
      "learning_rate": 2.4619096629430924e-05,
      "loss": 0.1948,
      "num_input_tokens_seen": 17491120,
      "step": 20195
    },
    {
      "epoch": 9.523809523809524,
      "grad_norm": 2.381747245788574,
      "learning_rate": 2.4609280321590543e-05,
      "loss": 0.1978,
      "num_input_tokens_seen": 17495488,
      "step": 20200
    },
    {
      "epoch": 9.523809523809524,
      "eval_loss": 0.20058976113796234,
      "eval_runtime": 22.1711,
      "eval_samples_per_second": 42.533,
      "eval_steps_per_second": 21.289,
      "num_input_tokens_seen": 17495488,
      "step": 20200
    },
    {
      "epoch": 9.526166902404526,
      "grad_norm": 0.8902534246444702,
      "learning_rate": 2.4599464074004037e-05,
      "loss": 0.1507,
      "num_input_tokens_seen": 17499936,
      "step": 20205
    },
    {
      "epoch": 9.528524280999529,
      "grad_norm": 2.0385043621063232,
      "learning_rate": 2.4589647888185204e-05,
      "loss": 0.1771,
      "num_input_tokens_seen": 17504512,
      "step": 20210
    },
    {
      "epoch": 9.530881659594531,
      "grad_norm": 1.2385271787643433,
      "learning_rate": 2.4579831765647836e-05,
      "loss": 0.1846,
      "num_input_tokens_seen": 17508800,
      "step": 20215
    },
    {
      "epoch": 9.533239038189533,
      "grad_norm": 0.9354729652404785,
      "learning_rate": 2.4570015707905676e-05,
      "loss": 0.1686,
      "num_input_tokens_seen": 17512608,
      "step": 20220
    },
    {
      "epoch": 9.535596416784536,
      "grad_norm": 0.9687459468841553,
      "learning_rate": 2.4560199716472508e-05,
      "loss": 0.1644,
      "num_input_tokens_seen": 17517376,
      "step": 20225
    },
    {
      "epoch": 9.537953795379538,
      "grad_norm": 0.8689837455749512,
      "learning_rate": 2.455038379286207e-05,
      "loss": 0.1428,
      "num_input_tokens_seen": 17521360,
      "step": 20230
    },
    {
      "epoch": 9.54031117397454,
      "grad_norm": 1.3306269645690918,
      "learning_rate": 2.4540567938588095e-05,
      "loss": 0.2694,
      "num_input_tokens_seen": 17525440,
      "step": 20235
    },
    {
      "epoch": 9.542668552569543,
      "grad_norm": 1.3681919574737549,
      "learning_rate": 2.4530752155164328e-05,
      "loss": 0.1894,
      "num_input_tokens_seen": 17530208,
      "step": 20240
    },
    {
      "epoch": 9.545025931164545,
      "grad_norm": 0.8396546244621277,
      "learning_rate": 2.4520936444104463e-05,
      "loss": 0.1607,
      "num_input_tokens_seen": 17534224,
      "step": 20245
    },
    {
      "epoch": 9.547383309759548,
      "grad_norm": 0.9357109069824219,
      "learning_rate": 2.4511120806922218e-05,
      "loss": 0.1314,
      "num_input_tokens_seen": 17538768,
      "step": 20250
    },
    {
      "epoch": 9.54974068835455,
      "grad_norm": 0.4683522880077362,
      "learning_rate": 2.45013052451313e-05,
      "loss": 0.1418,
      "num_input_tokens_seen": 17543088,
      "step": 20255
    },
    {
      "epoch": 9.552098066949553,
      "grad_norm": 0.4790043532848358,
      "learning_rate": 2.4491489760245376e-05,
      "loss": 0.1299,
      "num_input_tokens_seen": 17547984,
      "step": 20260
    },
    {
      "epoch": 9.554455445544555,
      "grad_norm": 1.0399186611175537,
      "learning_rate": 2.4481674353778115e-05,
      "loss": 0.1915,
      "num_input_tokens_seen": 17551808,
      "step": 20265
    },
    {
      "epoch": 9.556812824139557,
      "grad_norm": 0.9062173366546631,
      "learning_rate": 2.447185902724319e-05,
      "loss": 0.1952,
      "num_input_tokens_seen": 17555456,
      "step": 20270
    },
    {
      "epoch": 9.55917020273456,
      "grad_norm": 0.883118212223053,
      "learning_rate": 2.4462043782154233e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 17559840,
      "step": 20275
    },
    {
      "epoch": 9.561527581329562,
      "grad_norm": 0.8659515976905823,
      "learning_rate": 2.4452228620024895e-05,
      "loss": 0.1543,
      "num_input_tokens_seen": 17564016,
      "step": 20280
    },
    {
      "epoch": 9.563884959924565,
      "grad_norm": 0.6248417496681213,
      "learning_rate": 2.4442413542368776e-05,
      "loss": 0.203,
      "num_input_tokens_seen": 17568320,
      "step": 20285
    },
    {
      "epoch": 9.566242338519567,
      "grad_norm": 1.8223490715026855,
      "learning_rate": 2.4432598550699502e-05,
      "loss": 0.1982,
      "num_input_tokens_seen": 17572544,
      "step": 20290
    },
    {
      "epoch": 9.56859971711457,
      "grad_norm": 1.8315179347991943,
      "learning_rate": 2.4422783646530663e-05,
      "loss": 0.1696,
      "num_input_tokens_seen": 17576704,
      "step": 20295
    },
    {
      "epoch": 9.570957095709572,
      "grad_norm": 0.8754879236221313,
      "learning_rate": 2.441296883137584e-05,
      "loss": 0.1233,
      "num_input_tokens_seen": 17581056,
      "step": 20300
    },
    {
      "epoch": 9.573314474304574,
      "grad_norm": 0.9532091021537781,
      "learning_rate": 2.4403154106748592e-05,
      "loss": 0.1245,
      "num_input_tokens_seen": 17585168,
      "step": 20305
    },
    {
      "epoch": 9.575671852899575,
      "grad_norm": 0.7099881172180176,
      "learning_rate": 2.4393339474162494e-05,
      "loss": 0.1426,
      "num_input_tokens_seen": 17589472,
      "step": 20310
    },
    {
      "epoch": 9.578029231494579,
      "grad_norm": 0.4683946371078491,
      "learning_rate": 2.4383524935131062e-05,
      "loss": 0.2038,
      "num_input_tokens_seen": 17593456,
      "step": 20315
    },
    {
      "epoch": 9.58038661008958,
      "grad_norm": 1.5456105470657349,
      "learning_rate": 2.437371049116784e-05,
      "loss": 0.1519,
      "num_input_tokens_seen": 17598576,
      "step": 20320
    },
    {
      "epoch": 9.582743988684582,
      "grad_norm": 0.6612688899040222,
      "learning_rate": 2.436389614378632e-05,
      "loss": 0.1556,
      "num_input_tokens_seen": 17602704,
      "step": 20325
    },
    {
      "epoch": 9.585101367279584,
      "grad_norm": 1.8300961256027222,
      "learning_rate": 2.435408189450002e-05,
      "loss": 0.2208,
      "num_input_tokens_seen": 17607328,
      "step": 20330
    },
    {
      "epoch": 9.587458745874587,
      "grad_norm": 0.7152625918388367,
      "learning_rate": 2.4344267744822406e-05,
      "loss": 0.1698,
      "num_input_tokens_seen": 17611696,
      "step": 20335
    },
    {
      "epoch": 9.58981612446959,
      "grad_norm": 0.8517228960990906,
      "learning_rate": 2.4334453696266944e-05,
      "loss": 0.1414,
      "num_input_tokens_seen": 17616928,
      "step": 20340
    },
    {
      "epoch": 9.592173503064592,
      "grad_norm": 0.9325321316719055,
      "learning_rate": 2.432463975034708e-05,
      "loss": 0.1882,
      "num_input_tokens_seen": 17621360,
      "step": 20345
    },
    {
      "epoch": 9.594530881659594,
      "grad_norm": 2.1142749786376953,
      "learning_rate": 2.4314825908576265e-05,
      "loss": 0.1562,
      "num_input_tokens_seen": 17625952,
      "step": 20350
    },
    {
      "epoch": 9.596888260254596,
      "grad_norm": 1.110050916671753,
      "learning_rate": 2.4305012172467897e-05,
      "loss": 0.1747,
      "num_input_tokens_seen": 17630480,
      "step": 20355
    },
    {
      "epoch": 9.599245638849599,
      "grad_norm": 0.5854708552360535,
      "learning_rate": 2.4295198543535393e-05,
      "loss": 0.1643,
      "num_input_tokens_seen": 17634944,
      "step": 20360
    },
    {
      "epoch": 9.601603017444601,
      "grad_norm": 0.9896730780601501,
      "learning_rate": 2.4285385023292124e-05,
      "loss": 0.1712,
      "num_input_tokens_seen": 17638992,
      "step": 20365
    },
    {
      "epoch": 9.603960396039604,
      "grad_norm": 0.7264599204063416,
      "learning_rate": 2.427557161325147e-05,
      "loss": 0.1101,
      "num_input_tokens_seen": 17643328,
      "step": 20370
    },
    {
      "epoch": 9.606317774634606,
      "grad_norm": 0.5680750012397766,
      "learning_rate": 2.4265758314926778e-05,
      "loss": 0.1594,
      "num_input_tokens_seen": 17648112,
      "step": 20375
    },
    {
      "epoch": 9.608675153229608,
      "grad_norm": 1.627468466758728,
      "learning_rate": 2.4255945129831373e-05,
      "loss": 0.2158,
      "num_input_tokens_seen": 17653600,
      "step": 20380
    },
    {
      "epoch": 9.61103253182461,
      "grad_norm": 0.5440817475318909,
      "learning_rate": 2.4246132059478578e-05,
      "loss": 0.1323,
      "num_input_tokens_seen": 17658096,
      "step": 20385
    },
    {
      "epoch": 9.613389910419613,
      "grad_norm": 1.2428438663482666,
      "learning_rate": 2.4236319105381706e-05,
      "loss": 0.1616,
      "num_input_tokens_seen": 17661776,
      "step": 20390
    },
    {
      "epoch": 9.615747289014616,
      "grad_norm": 1.374666690826416,
      "learning_rate": 2.422650626905401e-05,
      "loss": 0.2054,
      "num_input_tokens_seen": 17665664,
      "step": 20395
    },
    {
      "epoch": 9.618104667609618,
      "grad_norm": 0.9928940534591675,
      "learning_rate": 2.4216693552008785e-05,
      "loss": 0.2193,
      "num_input_tokens_seen": 17670576,
      "step": 20400
    },
    {
      "epoch": 9.618104667609618,
      "eval_loss": 0.20429562032222748,
      "eval_runtime": 22.2003,
      "eval_samples_per_second": 42.477,
      "eval_steps_per_second": 21.261,
      "num_input_tokens_seen": 17670576,
      "step": 20400
    },
    {
      "epoch": 9.62046204620462,
      "grad_norm": 2.005547523498535,
      "learning_rate": 2.4206880955759247e-05,
      "loss": 0.1962,
      "num_input_tokens_seen": 17674976,
      "step": 20405
    },
    {
      "epoch": 9.622819424799623,
      "grad_norm": 0.7622870802879333,
      "learning_rate": 2.419706848181863e-05,
      "loss": 0.2029,
      "num_input_tokens_seen": 17679840,
      "step": 20410
    },
    {
      "epoch": 9.625176803394625,
      "grad_norm": 0.6464874148368835,
      "learning_rate": 2.4187256131700153e-05,
      "loss": 0.1448,
      "num_input_tokens_seen": 17684336,
      "step": 20415
    },
    {
      "epoch": 9.627534181989628,
      "grad_norm": 0.6685307025909424,
      "learning_rate": 2.4177443906916985e-05,
      "loss": 0.2745,
      "num_input_tokens_seen": 17688432,
      "step": 20420
    },
    {
      "epoch": 9.62989156058463,
      "grad_norm": 0.8400667905807495,
      "learning_rate": 2.4167631808982303e-05,
      "loss": 0.1864,
      "num_input_tokens_seen": 17692896,
      "step": 20425
    },
    {
      "epoch": 9.632248939179632,
      "grad_norm": 1.2925243377685547,
      "learning_rate": 2.4157819839409264e-05,
      "loss": 0.2003,
      "num_input_tokens_seen": 17696560,
      "step": 20430
    },
    {
      "epoch": 9.634606317774635,
      "grad_norm": 0.9034615159034729,
      "learning_rate": 2.414800799971098e-05,
      "loss": 0.1547,
      "num_input_tokens_seen": 17700432,
      "step": 20435
    },
    {
      "epoch": 9.636963696369637,
      "grad_norm": 0.9897530674934387,
      "learning_rate": 2.4138196291400582e-05,
      "loss": 0.2241,
      "num_input_tokens_seen": 17704800,
      "step": 20440
    },
    {
      "epoch": 9.63932107496464,
      "grad_norm": 0.9164755940437317,
      "learning_rate": 2.412838471599114e-05,
      "loss": 0.1218,
      "num_input_tokens_seen": 17708368,
      "step": 20445
    },
    {
      "epoch": 9.641678453559642,
      "grad_norm": 0.8621382713317871,
      "learning_rate": 2.411857327499572e-05,
      "loss": 0.2504,
      "num_input_tokens_seen": 17712976,
      "step": 20450
    },
    {
      "epoch": 9.644035832154644,
      "grad_norm": 0.9328822493553162,
      "learning_rate": 2.410876196992739e-05,
      "loss": 0.1931,
      "num_input_tokens_seen": 17716992,
      "step": 20455
    },
    {
      "epoch": 9.646393210749647,
      "grad_norm": 1.161285161972046,
      "learning_rate": 2.4098950802299156e-05,
      "loss": 0.2158,
      "num_input_tokens_seen": 17721200,
      "step": 20460
    },
    {
      "epoch": 9.64875058934465,
      "grad_norm": 2.0411407947540283,
      "learning_rate": 2.4089139773624027e-05,
      "loss": 0.1538,
      "num_input_tokens_seen": 17724432,
      "step": 20465
    },
    {
      "epoch": 9.651107967939652,
      "grad_norm": 0.6787992715835571,
      "learning_rate": 2.4079328885415007e-05,
      "loss": 0.2269,
      "num_input_tokens_seen": 17728928,
      "step": 20470
    },
    {
      "epoch": 9.653465346534654,
      "grad_norm": 0.4570913314819336,
      "learning_rate": 2.4069518139185036e-05,
      "loss": 0.2069,
      "num_input_tokens_seen": 17732768,
      "step": 20475
    },
    {
      "epoch": 9.655822725129656,
      "grad_norm": 0.7063820362091064,
      "learning_rate": 2.405970753644706e-05,
      "loss": 0.2486,
      "num_input_tokens_seen": 17737584,
      "step": 20480
    },
    {
      "epoch": 9.658180103724659,
      "grad_norm": 1.8720930814743042,
      "learning_rate": 2.4049897078714e-05,
      "loss": 0.1748,
      "num_input_tokens_seen": 17742608,
      "step": 20485
    },
    {
      "epoch": 9.660537482319661,
      "grad_norm": 1.4161783456802368,
      "learning_rate": 2.404008676749874e-05,
      "loss": 0.2322,
      "num_input_tokens_seen": 17747696,
      "step": 20490
    },
    {
      "epoch": 9.662894860914664,
      "grad_norm": 1.9414031505584717,
      "learning_rate": 2.403027660431418e-05,
      "loss": 0.2252,
      "num_input_tokens_seen": 17752144,
      "step": 20495
    },
    {
      "epoch": 9.665252239509666,
      "grad_norm": 1.1775689125061035,
      "learning_rate": 2.402046659067314e-05,
      "loss": 0.1966,
      "num_input_tokens_seen": 17756320,
      "step": 20500
    },
    {
      "epoch": 9.667609618104667,
      "grad_norm": 0.5885893702507019,
      "learning_rate": 2.401065672808847e-05,
      "loss": 0.2144,
      "num_input_tokens_seen": 17760304,
      "step": 20505
    },
    {
      "epoch": 9.66996699669967,
      "grad_norm": 0.785905122756958,
      "learning_rate": 2.400084701807296e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 17765152,
      "step": 20510
    },
    {
      "epoch": 9.672324375294671,
      "grad_norm": 0.37853074073791504,
      "learning_rate": 2.39910374621394e-05,
      "loss": 0.116,
      "num_input_tokens_seen": 17769280,
      "step": 20515
    },
    {
      "epoch": 9.674681753889674,
      "grad_norm": 0.45019423961639404,
      "learning_rate": 2.3981228061800544e-05,
      "loss": 0.2242,
      "num_input_tokens_seen": 17774432,
      "step": 20520
    },
    {
      "epoch": 9.677039132484676,
      "grad_norm": 0.6516244411468506,
      "learning_rate": 2.3971418818569115e-05,
      "loss": 0.2647,
      "num_input_tokens_seen": 17778400,
      "step": 20525
    },
    {
      "epoch": 9.679396511079679,
      "grad_norm": 1.6869912147521973,
      "learning_rate": 2.3961609733957832e-05,
      "loss": 0.2567,
      "num_input_tokens_seen": 17782720,
      "step": 20530
    },
    {
      "epoch": 9.681753889674681,
      "grad_norm": 1.9141415357589722,
      "learning_rate": 2.395180080947939e-05,
      "loss": 0.2533,
      "num_input_tokens_seen": 17786832,
      "step": 20535
    },
    {
      "epoch": 9.684111268269683,
      "grad_norm": 0.6576740145683289,
      "learning_rate": 2.394199204664642e-05,
      "loss": 0.1325,
      "num_input_tokens_seen": 17790912,
      "step": 20540
    },
    {
      "epoch": 9.686468646864686,
      "grad_norm": 0.8346157073974609,
      "learning_rate": 2.3932183446971583e-05,
      "loss": 0.1591,
      "num_input_tokens_seen": 17795424,
      "step": 20545
    },
    {
      "epoch": 9.688826025459688,
      "grad_norm": 1.3148810863494873,
      "learning_rate": 2.3922375011967473e-05,
      "loss": 0.1238,
      "num_input_tokens_seen": 17799744,
      "step": 20550
    },
    {
      "epoch": 9.69118340405469,
      "grad_norm": 0.656443178653717,
      "learning_rate": 2.3912566743146676e-05,
      "loss": 0.1532,
      "num_input_tokens_seen": 17803872,
      "step": 20555
    },
    {
      "epoch": 9.693540782649693,
      "grad_norm": 0.6662218570709229,
      "learning_rate": 2.390275864202176e-05,
      "loss": 0.1444,
      "num_input_tokens_seen": 17808880,
      "step": 20560
    },
    {
      "epoch": 9.695898161244696,
      "grad_norm": 1.535007357597351,
      "learning_rate": 2.3892950710105243e-05,
      "loss": 0.1947,
      "num_input_tokens_seen": 17813328,
      "step": 20565
    },
    {
      "epoch": 9.698255539839698,
      "grad_norm": 0.7873338460922241,
      "learning_rate": 2.3883142948909635e-05,
      "loss": 0.1415,
      "num_input_tokens_seen": 17817760,
      "step": 20570
    },
    {
      "epoch": 9.7006129184347,
      "grad_norm": 0.8061013221740723,
      "learning_rate": 2.3873335359947433e-05,
      "loss": 0.1765,
      "num_input_tokens_seen": 17822080,
      "step": 20575
    },
    {
      "epoch": 9.702970297029703,
      "grad_norm": 0.952102541923523,
      "learning_rate": 2.3863527944731066e-05,
      "loss": 0.2267,
      "num_input_tokens_seen": 17825984,
      "step": 20580
    },
    {
      "epoch": 9.705327675624705,
      "grad_norm": 2.9026076793670654,
      "learning_rate": 2.385372070477298e-05,
      "loss": 0.1852,
      "num_input_tokens_seen": 17830240,
      "step": 20585
    },
    {
      "epoch": 9.707685054219708,
      "grad_norm": 1.287312626838684,
      "learning_rate": 2.384391364158556e-05,
      "loss": 0.1594,
      "num_input_tokens_seen": 17834912,
      "step": 20590
    },
    {
      "epoch": 9.71004243281471,
      "grad_norm": 0.8653079867362976,
      "learning_rate": 2.3834106756681185e-05,
      "loss": 0.2319,
      "num_input_tokens_seen": 17838576,
      "step": 20595
    },
    {
      "epoch": 9.712399811409712,
      "grad_norm": 0.7867031097412109,
      "learning_rate": 2.3824300051572206e-05,
      "loss": 0.1625,
      "num_input_tokens_seen": 17843440,
      "step": 20600
    },
    {
      "epoch": 9.712399811409712,
      "eval_loss": 0.20307381451129913,
      "eval_runtime": 22.2068,
      "eval_samples_per_second": 42.465,
      "eval_steps_per_second": 21.255,
      "num_input_tokens_seen": 17843440,
      "step": 20600
    },
    {
      "epoch": 9.714757190004715,
      "grad_norm": 1.1564326286315918,
      "learning_rate": 2.3814493527770923e-05,
      "loss": 0.2215,
      "num_input_tokens_seen": 17848352,
      "step": 20605
    },
    {
      "epoch": 9.717114568599717,
      "grad_norm": 1.52826988697052,
      "learning_rate": 2.3804687186789637e-05,
      "loss": 0.1546,
      "num_input_tokens_seen": 17852848,
      "step": 20610
    },
    {
      "epoch": 9.71947194719472,
      "grad_norm": 1.1573710441589355,
      "learning_rate": 2.379488103014062e-05,
      "loss": 0.1808,
      "num_input_tokens_seen": 17856944,
      "step": 20615
    },
    {
      "epoch": 9.721829325789722,
      "grad_norm": 1.4598201513290405,
      "learning_rate": 2.3785075059336086e-05,
      "loss": 0.2826,
      "num_input_tokens_seen": 17861616,
      "step": 20620
    },
    {
      "epoch": 9.724186704384724,
      "grad_norm": 0.3597964942455292,
      "learning_rate": 2.3775269275888248e-05,
      "loss": 0.1578,
      "num_input_tokens_seen": 17865472,
      "step": 20625
    },
    {
      "epoch": 9.726544082979727,
      "grad_norm": 0.9565565586090088,
      "learning_rate": 2.3765463681309274e-05,
      "loss": 0.1853,
      "num_input_tokens_seen": 17870048,
      "step": 20630
    },
    {
      "epoch": 9.72890146157473,
      "grad_norm": 0.6527553200721741,
      "learning_rate": 2.3755658277111313e-05,
      "loss": 0.222,
      "num_input_tokens_seen": 17874512,
      "step": 20635
    },
    {
      "epoch": 9.731258840169732,
      "grad_norm": 1.1748254299163818,
      "learning_rate": 2.374585306480649e-05,
      "loss": 0.1842,
      "num_input_tokens_seen": 17879136,
      "step": 20640
    },
    {
      "epoch": 9.733616218764734,
      "grad_norm": 1.2322499752044678,
      "learning_rate": 2.3736048045906877e-05,
      "loss": 0.2132,
      "num_input_tokens_seen": 17882528,
      "step": 20645
    },
    {
      "epoch": 9.735973597359736,
      "grad_norm": 0.7359440922737122,
      "learning_rate": 2.372624322192454e-05,
      "loss": 0.1159,
      "num_input_tokens_seen": 17886176,
      "step": 20650
    },
    {
      "epoch": 9.738330975954739,
      "grad_norm": 1.1472468376159668,
      "learning_rate": 2.3716438594371516e-05,
      "loss": 0.1854,
      "num_input_tokens_seen": 17890720,
      "step": 20655
    },
    {
      "epoch": 9.740688354549741,
      "grad_norm": 0.6550520658493042,
      "learning_rate": 2.3706634164759784e-05,
      "loss": 0.2251,
      "num_input_tokens_seen": 17894688,
      "step": 20660
    },
    {
      "epoch": 9.743045733144744,
      "grad_norm": 0.4854090213775635,
      "learning_rate": 2.3696829934601323e-05,
      "loss": 0.2235,
      "num_input_tokens_seen": 17898480,
      "step": 20665
    },
    {
      "epoch": 9.745403111739746,
      "grad_norm": 0.6692149639129639,
      "learning_rate": 2.3687025905408053e-05,
      "loss": 0.2191,
      "num_input_tokens_seen": 17902384,
      "step": 20670
    },
    {
      "epoch": 9.747760490334748,
      "grad_norm": 0.5031152963638306,
      "learning_rate": 2.3677222078691886e-05,
      "loss": 0.1467,
      "num_input_tokens_seen": 17906288,
      "step": 20675
    },
    {
      "epoch": 9.75011786892975,
      "grad_norm": 0.8850587606430054,
      "learning_rate": 2.366741845596471e-05,
      "loss": 0.239,
      "num_input_tokens_seen": 17910800,
      "step": 20680
    },
    {
      "epoch": 9.752475247524753,
      "grad_norm": 1.3658900260925293,
      "learning_rate": 2.3657615038738343e-05,
      "loss": 0.2374,
      "num_input_tokens_seen": 17915488,
      "step": 20685
    },
    {
      "epoch": 9.754832626119756,
      "grad_norm": 1.795453429222107,
      "learning_rate": 2.3647811828524614e-05,
      "loss": 0.1754,
      "num_input_tokens_seen": 17919136,
      "step": 20690
    },
    {
      "epoch": 9.757190004714758,
      "grad_norm": 0.7152372002601624,
      "learning_rate": 2.363800882683529e-05,
      "loss": 0.2335,
      "num_input_tokens_seen": 17922928,
      "step": 20695
    },
    {
      "epoch": 9.75954738330976,
      "grad_norm": 1.694046974182129,
      "learning_rate": 2.3628206035182125e-05,
      "loss": 0.2263,
      "num_input_tokens_seen": 17927984,
      "step": 20700
    },
    {
      "epoch": 9.761904761904763,
      "grad_norm": 0.8257821798324585,
      "learning_rate": 2.361840345507683e-05,
      "loss": 0.2182,
      "num_input_tokens_seen": 17932144,
      "step": 20705
    },
    {
      "epoch": 9.764262140499763,
      "grad_norm": 2.119673252105713,
      "learning_rate": 2.3608601088031073e-05,
      "loss": 0.17,
      "num_input_tokens_seen": 17936032,
      "step": 20710
    },
    {
      "epoch": 9.766619519094768,
      "grad_norm": 1.6267352104187012,
      "learning_rate": 2.3598798935556516e-05,
      "loss": 0.1613,
      "num_input_tokens_seen": 17940464,
      "step": 20715
    },
    {
      "epoch": 9.768976897689768,
      "grad_norm": 1.2176357507705688,
      "learning_rate": 2.3588996999164784e-05,
      "loss": 0.1671,
      "num_input_tokens_seen": 17944352,
      "step": 20720
    },
    {
      "epoch": 9.77133427628477,
      "grad_norm": 1.0038894414901733,
      "learning_rate": 2.3579195280367434e-05,
      "loss": 0.2347,
      "num_input_tokens_seen": 17948528,
      "step": 20725
    },
    {
      "epoch": 9.773691654879773,
      "grad_norm": 1.0231958627700806,
      "learning_rate": 2.356939378067603e-05,
      "loss": 0.1719,
      "num_input_tokens_seen": 17952496,
      "step": 20730
    },
    {
      "epoch": 9.776049033474775,
      "grad_norm": 0.9431200623512268,
      "learning_rate": 2.3559592501602092e-05,
      "loss": 0.1499,
      "num_input_tokens_seen": 17956880,
      "step": 20735
    },
    {
      "epoch": 9.778406412069778,
      "grad_norm": 1.6993792057037354,
      "learning_rate": 2.3549791444657076e-05,
      "loss": 0.168,
      "num_input_tokens_seen": 17961392,
      "step": 20740
    },
    {
      "epoch": 9.78076379066478,
      "grad_norm": 1.1649925708770752,
      "learning_rate": 2.353999061135246e-05,
      "loss": 0.1796,
      "num_input_tokens_seen": 17965568,
      "step": 20745
    },
    {
      "epoch": 9.783121169259783,
      "grad_norm": 1.5510437488555908,
      "learning_rate": 2.3530190003199626e-05,
      "loss": 0.1775,
      "num_input_tokens_seen": 17969648,
      "step": 20750
    },
    {
      "epoch": 9.785478547854785,
      "grad_norm": 1.34163498878479,
      "learning_rate": 2.3520389621709965e-05,
      "loss": 0.1419,
      "num_input_tokens_seen": 17974256,
      "step": 20755
    },
    {
      "epoch": 9.787835926449787,
      "grad_norm": 0.7477098703384399,
      "learning_rate": 2.351058946839483e-05,
      "loss": 0.2065,
      "num_input_tokens_seen": 17978192,
      "step": 20760
    },
    {
      "epoch": 9.79019330504479,
      "grad_norm": 1.4506824016571045,
      "learning_rate": 2.350078954476551e-05,
      "loss": 0.2271,
      "num_input_tokens_seen": 17981904,
      "step": 20765
    },
    {
      "epoch": 9.792550683639792,
      "grad_norm": 1.2792637348175049,
      "learning_rate": 2.3490989852333272e-05,
      "loss": 0.1089,
      "num_input_tokens_seen": 17986512,
      "step": 20770
    },
    {
      "epoch": 9.794908062234795,
      "grad_norm": 1.475193738937378,
      "learning_rate": 2.3481190392609377e-05,
      "loss": 0.2148,
      "num_input_tokens_seen": 17990464,
      "step": 20775
    },
    {
      "epoch": 9.797265440829797,
      "grad_norm": 1.0509395599365234,
      "learning_rate": 2.3471391167105e-05,
      "loss": 0.1817,
      "num_input_tokens_seen": 17995040,
      "step": 20780
    },
    {
      "epoch": 9.7996228194248,
      "grad_norm": 0.73553466796875,
      "learning_rate": 2.3461592177331325e-05,
      "loss": 0.2321,
      "num_input_tokens_seen": 17999104,
      "step": 20785
    },
    {
      "epoch": 9.801980198019802,
      "grad_norm": 1.6309797763824463,
      "learning_rate": 2.345179342479946e-05,
      "loss": 0.2287,
      "num_input_tokens_seen": 18003776,
      "step": 20790
    },
    {
      "epoch": 9.804337576614804,
      "grad_norm": 0.7442489266395569,
      "learning_rate": 2.3441994911020503e-05,
      "loss": 0.1744,
      "num_input_tokens_seen": 18008064,
      "step": 20795
    },
    {
      "epoch": 9.806694955209807,
      "grad_norm": 1.198710560798645,
      "learning_rate": 2.3432196637505522e-05,
      "loss": 0.1655,
      "num_input_tokens_seen": 18012496,
      "step": 20800
    },
    {
      "epoch": 9.806694955209807,
      "eval_loss": 0.20453613996505737,
      "eval_runtime": 22.1576,
      "eval_samples_per_second": 42.559,
      "eval_steps_per_second": 21.302,
      "num_input_tokens_seen": 18012496,
      "step": 20800
    },
    {
      "epoch": 9.809052333804809,
      "grad_norm": 1.0309267044067383,
      "learning_rate": 2.3422398605765515e-05,
      "loss": 0.1524,
      "num_input_tokens_seen": 18016016,
      "step": 20805
    },
    {
      "epoch": 9.811409712399811,
      "grad_norm": 1.2048777341842651,
      "learning_rate": 2.3412600817311462e-05,
      "loss": 0.1807,
      "num_input_tokens_seen": 18019792,
      "step": 20810
    },
    {
      "epoch": 9.813767090994814,
      "grad_norm": 2.171645164489746,
      "learning_rate": 2.3402803273654326e-05,
      "loss": 0.2118,
      "num_input_tokens_seen": 18023792,
      "step": 20815
    },
    {
      "epoch": 9.816124469589816,
      "grad_norm": 1.5088634490966797,
      "learning_rate": 2.3393005976304983e-05,
      "loss": 0.2296,
      "num_input_tokens_seen": 18027696,
      "step": 20820
    },
    {
      "epoch": 9.818481848184819,
      "grad_norm": 1.1147236824035645,
      "learning_rate": 2.338320892677432e-05,
      "loss": 0.1889,
      "num_input_tokens_seen": 18031760,
      "step": 20825
    },
    {
      "epoch": 9.820839226779821,
      "grad_norm": 1.1950846910476685,
      "learning_rate": 2.3373412126573155e-05,
      "loss": 0.1995,
      "num_input_tokens_seen": 18036416,
      "step": 20830
    },
    {
      "epoch": 9.823196605374823,
      "grad_norm": 1.1504544019699097,
      "learning_rate": 2.3363615577212285e-05,
      "loss": 0.1743,
      "num_input_tokens_seen": 18041248,
      "step": 20835
    },
    {
      "epoch": 9.825553983969826,
      "grad_norm": 0.530267059803009,
      "learning_rate": 2.3353819280202455e-05,
      "loss": 0.1535,
      "num_input_tokens_seen": 18045680,
      "step": 20840
    },
    {
      "epoch": 9.827911362564828,
      "grad_norm": 0.8467192053794861,
      "learning_rate": 2.334402323705438e-05,
      "loss": 0.1769,
      "num_input_tokens_seen": 18049680,
      "step": 20845
    },
    {
      "epoch": 9.83026874115983,
      "grad_norm": 1.039473056793213,
      "learning_rate": 2.3334227449278725e-05,
      "loss": 0.1751,
      "num_input_tokens_seen": 18054688,
      "step": 20850
    },
    {
      "epoch": 9.832626119754833,
      "grad_norm": 2.228668689727783,
      "learning_rate": 2.3324431918386143e-05,
      "loss": 0.1974,
      "num_input_tokens_seen": 18059600,
      "step": 20855
    },
    {
      "epoch": 9.834983498349835,
      "grad_norm": 0.3985774517059326,
      "learning_rate": 2.3314636645887207e-05,
      "loss": 0.1333,
      "num_input_tokens_seen": 18064272,
      "step": 20860
    },
    {
      "epoch": 9.837340876944838,
      "grad_norm": 0.8061315417289734,
      "learning_rate": 2.3304841633292487e-05,
      "loss": 0.1656,
      "num_input_tokens_seen": 18068448,
      "step": 20865
    },
    {
      "epoch": 9.83969825553984,
      "grad_norm": 1.8073471784591675,
      "learning_rate": 2.329504688211248e-05,
      "loss": 0.2581,
      "num_input_tokens_seen": 18072768,
      "step": 20870
    },
    {
      "epoch": 9.842055634134843,
      "grad_norm": 0.9724053740501404,
      "learning_rate": 2.3285252393857677e-05,
      "loss": 0.1471,
      "num_input_tokens_seen": 18077264,
      "step": 20875
    },
    {
      "epoch": 9.844413012729845,
      "grad_norm": 1.4995518922805786,
      "learning_rate": 2.327545817003851e-05,
      "loss": 0.221,
      "num_input_tokens_seen": 18082560,
      "step": 20880
    },
    {
      "epoch": 9.846770391324847,
      "grad_norm": 0.5003247261047363,
      "learning_rate": 2.326566421216535e-05,
      "loss": 0.1693,
      "num_input_tokens_seen": 18086432,
      "step": 20885
    },
    {
      "epoch": 9.84912776991985,
      "grad_norm": 1.621264100074768,
      "learning_rate": 2.3255870521748565e-05,
      "loss": 0.3269,
      "num_input_tokens_seen": 18090832,
      "step": 20890
    },
    {
      "epoch": 9.851485148514852,
      "grad_norm": 0.721210777759552,
      "learning_rate": 2.3246077100298474e-05,
      "loss": 0.1413,
      "num_input_tokens_seen": 18094240,
      "step": 20895
    },
    {
      "epoch": 9.853842527109855,
      "grad_norm": 0.8761459589004517,
      "learning_rate": 2.3236283949325328e-05,
      "loss": 0.1404,
      "num_input_tokens_seen": 18099104,
      "step": 20900
    },
    {
      "epoch": 9.856199905704855,
      "grad_norm": 0.6147610545158386,
      "learning_rate": 2.3226491070339368e-05,
      "loss": 0.1677,
      "num_input_tokens_seen": 18103296,
      "step": 20905
    },
    {
      "epoch": 9.85855728429986,
      "grad_norm": 0.9113844037055969,
      "learning_rate": 2.3216698464850762e-05,
      "loss": 0.212,
      "num_input_tokens_seen": 18107856,
      "step": 20910
    },
    {
      "epoch": 9.86091466289486,
      "grad_norm": 1.2499562501907349,
      "learning_rate": 2.320690613436967e-05,
      "loss": 0.1479,
      "num_input_tokens_seen": 18112064,
      "step": 20915
    },
    {
      "epoch": 9.863272041489862,
      "grad_norm": 1.5394182205200195,
      "learning_rate": 2.3197114080406192e-05,
      "loss": 0.1944,
      "num_input_tokens_seen": 18116464,
      "step": 20920
    },
    {
      "epoch": 9.865629420084865,
      "grad_norm": 1.1993523836135864,
      "learning_rate": 2.3187322304470365e-05,
      "loss": 0.1905,
      "num_input_tokens_seen": 18120896,
      "step": 20925
    },
    {
      "epoch": 9.867986798679867,
      "grad_norm": 1.105039119720459,
      "learning_rate": 2.3177530808072222e-05,
      "loss": 0.2005,
      "num_input_tokens_seen": 18125712,
      "step": 20930
    },
    {
      "epoch": 9.87034417727487,
      "grad_norm": 0.5303094983100891,
      "learning_rate": 2.316773959272174e-05,
      "loss": 0.1661,
      "num_input_tokens_seen": 18130592,
      "step": 20935
    },
    {
      "epoch": 9.872701555869872,
      "grad_norm": 1.5139769315719604,
      "learning_rate": 2.3157948659928823e-05,
      "loss": 0.1974,
      "num_input_tokens_seen": 18135248,
      "step": 20940
    },
    {
      "epoch": 9.875058934464874,
      "grad_norm": 0.5501163601875305,
      "learning_rate": 2.3148158011203388e-05,
      "loss": 0.1111,
      "num_input_tokens_seen": 18139216,
      "step": 20945
    },
    {
      "epoch": 9.877416313059877,
      "grad_norm": 0.5725337862968445,
      "learning_rate": 2.3138367648055253e-05,
      "loss": 0.1819,
      "num_input_tokens_seen": 18144016,
      "step": 20950
    },
    {
      "epoch": 9.87977369165488,
      "grad_norm": 0.7359203696250916,
      "learning_rate": 2.312857757199422e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 18147984,
      "step": 20955
    },
    {
      "epoch": 9.882131070249882,
      "grad_norm": 0.6070775985717773,
      "learning_rate": 2.3118787784530048e-05,
      "loss": 0.171,
      "num_input_tokens_seen": 18152080,
      "step": 20960
    },
    {
      "epoch": 9.884488448844884,
      "grad_norm": 0.8185896277427673,
      "learning_rate": 2.310899828717243e-05,
      "loss": 0.1697,
      "num_input_tokens_seen": 18156128,
      "step": 20965
    },
    {
      "epoch": 9.886845827439886,
      "grad_norm": 0.3092438578605652,
      "learning_rate": 2.309920908143104e-05,
      "loss": 0.1447,
      "num_input_tokens_seen": 18160224,
      "step": 20970
    },
    {
      "epoch": 9.889203206034889,
      "grad_norm": 0.7481637597084045,
      "learning_rate": 2.308942016881551e-05,
      "loss": 0.1486,
      "num_input_tokens_seen": 18164128,
      "step": 20975
    },
    {
      "epoch": 9.891560584629891,
      "grad_norm": 1.0289719104766846,
      "learning_rate": 2.307963155083539e-05,
      "loss": 0.1201,
      "num_input_tokens_seen": 18168800,
      "step": 20980
    },
    {
      "epoch": 9.893917963224894,
      "grad_norm": 0.7300512790679932,
      "learning_rate": 2.306984322900022e-05,
      "loss": 0.1263,
      "num_input_tokens_seen": 18173200,
      "step": 20985
    },
    {
      "epoch": 9.896275341819896,
      "grad_norm": 0.4032100737094879,
      "learning_rate": 2.3060055204819482e-05,
      "loss": 0.1509,
      "num_input_tokens_seen": 18177344,
      "step": 20990
    },
    {
      "epoch": 9.898632720414899,
      "grad_norm": 1.4448925256729126,
      "learning_rate": 2.3050267479802604e-05,
      "loss": 0.1071,
      "num_input_tokens_seen": 18181968,
      "step": 20995
    },
    {
      "epoch": 9.900990099009901,
      "grad_norm": 1.2664117813110352,
      "learning_rate": 2.304048005545899e-05,
      "loss": 0.161,
      "num_input_tokens_seen": 18186480,
      "step": 21000
    },
    {
      "epoch": 9.900990099009901,
      "eval_loss": 0.20169991254806519,
      "eval_runtime": 22.1392,
      "eval_samples_per_second": 42.594,
      "eval_steps_per_second": 21.32,
      "num_input_tokens_seen": 18186480,
      "step": 21000
    },
    {
      "epoch": 9.903347477604903,
      "grad_norm": 1.3777490854263306,
      "learning_rate": 2.3030692933297972e-05,
      "loss": 0.1591,
      "num_input_tokens_seen": 18189808,
      "step": 21005
    },
    {
      "epoch": 9.905704856199906,
      "grad_norm": 1.3947962522506714,
      "learning_rate": 2.3020906114828843e-05,
      "loss": 0.1611,
      "num_input_tokens_seen": 18195152,
      "step": 21010
    },
    {
      "epoch": 9.908062234794908,
      "grad_norm": 0.8093017935752869,
      "learning_rate": 2.301111960156088e-05,
      "loss": 0.1546,
      "num_input_tokens_seen": 18199424,
      "step": 21015
    },
    {
      "epoch": 9.91041961338991,
      "grad_norm": 0.6018768548965454,
      "learning_rate": 2.300133339500326e-05,
      "loss": 0.129,
      "num_input_tokens_seen": 18204592,
      "step": 21020
    },
    {
      "epoch": 9.912776991984913,
      "grad_norm": 1.094236969947815,
      "learning_rate": 2.2991547496665148e-05,
      "loss": 0.1411,
      "num_input_tokens_seen": 18209456,
      "step": 21025
    },
    {
      "epoch": 9.915134370579915,
      "grad_norm": 1.084227204322815,
      "learning_rate": 2.298176190805565e-05,
      "loss": 0.2123,
      "num_input_tokens_seen": 18213952,
      "step": 21030
    },
    {
      "epoch": 9.917491749174918,
      "grad_norm": 1.2703847885131836,
      "learning_rate": 2.2971976630683826e-05,
      "loss": 0.2138,
      "num_input_tokens_seen": 18217600,
      "step": 21035
    },
    {
      "epoch": 9.91984912776992,
      "grad_norm": 0.48065608739852905,
      "learning_rate": 2.29621916660587e-05,
      "loss": 0.1749,
      "num_input_tokens_seen": 18221904,
      "step": 21040
    },
    {
      "epoch": 9.922206506364923,
      "grad_norm": 1.0223937034606934,
      "learning_rate": 2.295240701568922e-05,
      "loss": 0.2197,
      "num_input_tokens_seen": 18226464,
      "step": 21045
    },
    {
      "epoch": 9.924563884959925,
      "grad_norm": 0.7734251618385315,
      "learning_rate": 2.2942622681084312e-05,
      "loss": 0.1445,
      "num_input_tokens_seen": 18230192,
      "step": 21050
    },
    {
      "epoch": 9.926921263554927,
      "grad_norm": 1.2513164281845093,
      "learning_rate": 2.293283866375284e-05,
      "loss": 0.2759,
      "num_input_tokens_seen": 18234624,
      "step": 21055
    },
    {
      "epoch": 9.92927864214993,
      "grad_norm": 1.0240954160690308,
      "learning_rate": 2.2923054965203627e-05,
      "loss": 0.2055,
      "num_input_tokens_seen": 18238992,
      "step": 21060
    },
    {
      "epoch": 9.931636020744932,
      "grad_norm": 0.5824181437492371,
      "learning_rate": 2.2913271586945443e-05,
      "loss": 0.181,
      "num_input_tokens_seen": 18242864,
      "step": 21065
    },
    {
      "epoch": 9.933993399339935,
      "grad_norm": 0.8751112222671509,
      "learning_rate": 2.290348853048699e-05,
      "loss": 0.2269,
      "num_input_tokens_seen": 18247504,
      "step": 21070
    },
    {
      "epoch": 9.936350777934937,
      "grad_norm": 0.4771482050418854,
      "learning_rate": 2.2893705797336956e-05,
      "loss": 0.133,
      "num_input_tokens_seen": 18251024,
      "step": 21075
    },
    {
      "epoch": 9.93870815652994,
      "grad_norm": 0.47226259112358093,
      "learning_rate": 2.288392338900397e-05,
      "loss": 0.169,
      "num_input_tokens_seen": 18256160,
      "step": 21080
    },
    {
      "epoch": 9.941065535124942,
      "grad_norm": 1.6674535274505615,
      "learning_rate": 2.2874141306996576e-05,
      "loss": 0.1345,
      "num_input_tokens_seen": 18260336,
      "step": 21085
    },
    {
      "epoch": 9.943422913719944,
      "grad_norm": 1.9459502696990967,
      "learning_rate": 2.2864359552823312e-05,
      "loss": 0.2093,
      "num_input_tokens_seen": 18265536,
      "step": 21090
    },
    {
      "epoch": 9.945780292314947,
      "grad_norm": 1.6950231790542603,
      "learning_rate": 2.2854578127992648e-05,
      "loss": 0.1821,
      "num_input_tokens_seen": 18269280,
      "step": 21095
    },
    {
      "epoch": 9.948137670909949,
      "grad_norm": 0.7188493609428406,
      "learning_rate": 2.2844797034012988e-05,
      "loss": 0.1531,
      "num_input_tokens_seen": 18273552,
      "step": 21100
    },
    {
      "epoch": 9.950495049504951,
      "grad_norm": 0.8296502828598022,
      "learning_rate": 2.2835016272392722e-05,
      "loss": 0.1975,
      "num_input_tokens_seen": 18277632,
      "step": 21105
    },
    {
      "epoch": 9.952852428099952,
      "grad_norm": 1.9895614385604858,
      "learning_rate": 2.2825235844640142e-05,
      "loss": 0.1573,
      "num_input_tokens_seen": 18282576,
      "step": 21110
    },
    {
      "epoch": 9.955209806694956,
      "grad_norm": 0.6596464514732361,
      "learning_rate": 2.2815455752263522e-05,
      "loss": 0.209,
      "num_input_tokens_seen": 18286064,
      "step": 21115
    },
    {
      "epoch": 9.957567185289957,
      "grad_norm": 1.02486252784729,
      "learning_rate": 2.2805675996771092e-05,
      "loss": 0.1857,
      "num_input_tokens_seen": 18291392,
      "step": 21120
    },
    {
      "epoch": 9.95992456388496,
      "grad_norm": 1.8707079887390137,
      "learning_rate": 2.2795896579670987e-05,
      "loss": 0.1653,
      "num_input_tokens_seen": 18295888,
      "step": 21125
    },
    {
      "epoch": 9.962281942479962,
      "grad_norm": 1.87435781955719,
      "learning_rate": 2.2786117502471337e-05,
      "loss": 0.1575,
      "num_input_tokens_seen": 18299856,
      "step": 21130
    },
    {
      "epoch": 9.964639321074964,
      "grad_norm": 1.354357123374939,
      "learning_rate": 2.2776338766680185e-05,
      "loss": 0.1769,
      "num_input_tokens_seen": 18303584,
      "step": 21135
    },
    {
      "epoch": 9.966996699669966,
      "grad_norm": 1.564168930053711,
      "learning_rate": 2.2766560373805533e-05,
      "loss": 0.1762,
      "num_input_tokens_seen": 18307792,
      "step": 21140
    },
    {
      "epoch": 9.969354078264969,
      "grad_norm": 0.707224428653717,
      "learning_rate": 2.2756782325355353e-05,
      "loss": 0.1952,
      "num_input_tokens_seen": 18312416,
      "step": 21145
    },
    {
      "epoch": 9.971711456859971,
      "grad_norm": 0.48307979106903076,
      "learning_rate": 2.2747004622837514e-05,
      "loss": 0.213,
      "num_input_tokens_seen": 18316560,
      "step": 21150
    },
    {
      "epoch": 9.974068835454974,
      "grad_norm": 1.4874557256698608,
      "learning_rate": 2.2737227267759878e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 18321568,
      "step": 21155
    },
    {
      "epoch": 9.976426214049976,
      "grad_norm": 1.3696211576461792,
      "learning_rate": 2.272745026163024e-05,
      "loss": 0.2039,
      "num_input_tokens_seen": 18326016,
      "step": 21160
    },
    {
      "epoch": 9.978783592644978,
      "grad_norm": 0.4911412298679352,
      "learning_rate": 2.271767360595633e-05,
      "loss": 0.1717,
      "num_input_tokens_seen": 18330560,
      "step": 21165
    },
    {
      "epoch": 9.98114097123998,
      "grad_norm": 1.5893751382827759,
      "learning_rate": 2.270789730224583e-05,
      "loss": 0.1921,
      "num_input_tokens_seen": 18335040,
      "step": 21170
    },
    {
      "epoch": 9.983498349834983,
      "grad_norm": 1.2809480428695679,
      "learning_rate": 2.2698121352006367e-05,
      "loss": 0.1859,
      "num_input_tokens_seen": 18339024,
      "step": 21175
    },
    {
      "epoch": 9.985855728429986,
      "grad_norm": 0.8817849159240723,
      "learning_rate": 2.2688345756745517e-05,
      "loss": 0.1194,
      "num_input_tokens_seen": 18343520,
      "step": 21180
    },
    {
      "epoch": 9.988213107024988,
      "grad_norm": 0.5098994374275208,
      "learning_rate": 2.267857051797081e-05,
      "loss": 0.1913,
      "num_input_tokens_seen": 18348304,
      "step": 21185
    },
    {
      "epoch": 9.99057048561999,
      "grad_norm": 0.6444498896598816,
      "learning_rate": 2.2668795637189695e-05,
      "loss": 0.126,
      "num_input_tokens_seen": 18352272,
      "step": 21190
    },
    {
      "epoch": 9.992927864214993,
      "grad_norm": 0.785392701625824,
      "learning_rate": 2.2659021115909586e-05,
      "loss": 0.1374,
      "num_input_tokens_seen": 18356640,
      "step": 21195
    },
    {
      "epoch": 9.995285242809995,
      "grad_norm": 0.9938814640045166,
      "learning_rate": 2.2649246955637847e-05,
      "loss": 0.1796,
      "num_input_tokens_seen": 18360368,
      "step": 21200
    },
    {
      "epoch": 9.995285242809995,
      "eval_loss": 0.20162230730056763,
      "eval_runtime": 22.1561,
      "eval_samples_per_second": 42.562,
      "eval_steps_per_second": 21.303,
      "num_input_tokens_seen": 18360368,
      "step": 21200
    },
    {
      "epoch": 9.997642621404998,
      "grad_norm": 0.49685239791870117,
      "learning_rate": 2.2639473157881766e-05,
      "loss": 0.23,
      "num_input_tokens_seen": 18364368,
      "step": 21205
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.6676758527755737,
      "learning_rate": 2.2629699724148594e-05,
      "loss": 0.1453,
      "num_input_tokens_seen": 18368192,
      "step": 21210
    },
    {
      "epoch": 10.002357378595002,
      "grad_norm": 1.6939502954483032,
      "learning_rate": 2.26199266559455e-05,
      "loss": 0.1829,
      "num_input_tokens_seen": 18372992,
      "step": 21215
    },
    {
      "epoch": 10.004714757190005,
      "grad_norm": 0.47055986523628235,
      "learning_rate": 2.2610153954779625e-05,
      "loss": 0.1648,
      "num_input_tokens_seen": 18377920,
      "step": 21220
    },
    {
      "epoch": 10.007072135785007,
      "grad_norm": 1.8418359756469727,
      "learning_rate": 2.2600381622158056e-05,
      "loss": 0.2167,
      "num_input_tokens_seen": 18383072,
      "step": 21225
    },
    {
      "epoch": 10.00942951438001,
      "grad_norm": 0.9832539558410645,
      "learning_rate": 2.2590609659587783e-05,
      "loss": 0.1705,
      "num_input_tokens_seen": 18387824,
      "step": 21230
    },
    {
      "epoch": 10.011786892975012,
      "grad_norm": 1.116883635520935,
      "learning_rate": 2.2580838068575787e-05,
      "loss": 0.1647,
      "num_input_tokens_seen": 18391760,
      "step": 21235
    },
    {
      "epoch": 10.014144271570014,
      "grad_norm": 0.5500665903091431,
      "learning_rate": 2.257106685062896e-05,
      "loss": 0.3012,
      "num_input_tokens_seen": 18396720,
      "step": 21240
    },
    {
      "epoch": 10.016501650165017,
      "grad_norm": 1.0829808712005615,
      "learning_rate": 2.256129600725415e-05,
      "loss": 0.1283,
      "num_input_tokens_seen": 18400704,
      "step": 21245
    },
    {
      "epoch": 10.01885902876002,
      "grad_norm": 1.0851233005523682,
      "learning_rate": 2.2551525539958145e-05,
      "loss": 0.1628,
      "num_input_tokens_seen": 18404624,
      "step": 21250
    },
    {
      "epoch": 10.021216407355022,
      "grad_norm": 0.5026887059211731,
      "learning_rate": 2.2541755450247663e-05,
      "loss": 0.1502,
      "num_input_tokens_seen": 18409104,
      "step": 21255
    },
    {
      "epoch": 10.023573785950024,
      "grad_norm": 1.9038326740264893,
      "learning_rate": 2.2531985739629382e-05,
      "loss": 0.1563,
      "num_input_tokens_seen": 18413168,
      "step": 21260
    },
    {
      "epoch": 10.025931164545026,
      "grad_norm": 0.49238401651382446,
      "learning_rate": 2.2522216409609924e-05,
      "loss": 0.1742,
      "num_input_tokens_seen": 18417456,
      "step": 21265
    },
    {
      "epoch": 10.028288543140029,
      "grad_norm": 0.5881630182266235,
      "learning_rate": 2.2512447461695826e-05,
      "loss": 0.2222,
      "num_input_tokens_seen": 18422256,
      "step": 21270
    },
    {
      "epoch": 10.030645921735031,
      "grad_norm": 1.2602406740188599,
      "learning_rate": 2.2502678897393593e-05,
      "loss": 0.1329,
      "num_input_tokens_seen": 18428192,
      "step": 21275
    },
    {
      "epoch": 10.033003300330034,
      "grad_norm": 0.6455239057540894,
      "learning_rate": 2.2492910718209665e-05,
      "loss": 0.1607,
      "num_input_tokens_seen": 18432384,
      "step": 21280
    },
    {
      "epoch": 10.035360678925036,
      "grad_norm": 2.4800543785095215,
      "learning_rate": 2.2483142925650398e-05,
      "loss": 0.2132,
      "num_input_tokens_seen": 18436784,
      "step": 21285
    },
    {
      "epoch": 10.037718057520038,
      "grad_norm": 0.4995984733104706,
      "learning_rate": 2.247337552122213e-05,
      "loss": 0.2767,
      "num_input_tokens_seen": 18441312,
      "step": 21290
    },
    {
      "epoch": 10.04007543611504,
      "grad_norm": 0.8571516871452332,
      "learning_rate": 2.24636085064311e-05,
      "loss": 0.1456,
      "num_input_tokens_seen": 18445728,
      "step": 21295
    },
    {
      "epoch": 10.042432814710043,
      "grad_norm": 2.0830626487731934,
      "learning_rate": 2.245384188278351e-05,
      "loss": 0.2219,
      "num_input_tokens_seen": 18450096,
      "step": 21300
    },
    {
      "epoch": 10.044790193305046,
      "grad_norm": 2.2404253482818604,
      "learning_rate": 2.2444075651785513e-05,
      "loss": 0.2211,
      "num_input_tokens_seen": 18454560,
      "step": 21305
    },
    {
      "epoch": 10.047147571900048,
      "grad_norm": 1.2765084505081177,
      "learning_rate": 2.243430981494316e-05,
      "loss": 0.192,
      "num_input_tokens_seen": 18458656,
      "step": 21310
    },
    {
      "epoch": 10.049504950495049,
      "grad_norm": 1.913975715637207,
      "learning_rate": 2.2424544373762475e-05,
      "loss": 0.2059,
      "num_input_tokens_seen": 18463072,
      "step": 21315
    },
    {
      "epoch": 10.051862329090051,
      "grad_norm": 1.743324875831604,
      "learning_rate": 2.2414779329749418e-05,
      "loss": 0.1751,
      "num_input_tokens_seen": 18467424,
      "step": 21320
    },
    {
      "epoch": 10.054219707685053,
      "grad_norm": 0.7775964140892029,
      "learning_rate": 2.2405014684409873e-05,
      "loss": 0.1603,
      "num_input_tokens_seen": 18471952,
      "step": 21325
    },
    {
      "epoch": 10.056577086280056,
      "grad_norm": 0.5320525169372559,
      "learning_rate": 2.239525043924968e-05,
      "loss": 0.2067,
      "num_input_tokens_seen": 18476416,
      "step": 21330
    },
    {
      "epoch": 10.058934464875058,
      "grad_norm": 1.169677734375,
      "learning_rate": 2.2385486595774592e-05,
      "loss": 0.1523,
      "num_input_tokens_seen": 18480768,
      "step": 21335
    },
    {
      "epoch": 10.06129184347006,
      "grad_norm": 0.4576520621776581,
      "learning_rate": 2.237572315549033e-05,
      "loss": 0.1337,
      "num_input_tokens_seen": 18485472,
      "step": 21340
    },
    {
      "epoch": 10.063649222065063,
      "grad_norm": 0.9152936935424805,
      "learning_rate": 2.2365960119902545e-05,
      "loss": 0.178,
      "num_input_tokens_seen": 18490336,
      "step": 21345
    },
    {
      "epoch": 10.066006600660065,
      "grad_norm": 1.3537465333938599,
      "learning_rate": 2.2356197490516806e-05,
      "loss": 0.2016,
      "num_input_tokens_seen": 18494864,
      "step": 21350
    },
    {
      "epoch": 10.068363979255068,
      "grad_norm": 1.332167625427246,
      "learning_rate": 2.234643526883863e-05,
      "loss": 0.2035,
      "num_input_tokens_seen": 18498960,
      "step": 21355
    },
    {
      "epoch": 10.07072135785007,
      "grad_norm": 1.3567755222320557,
      "learning_rate": 2.2336673456373497e-05,
      "loss": 0.2408,
      "num_input_tokens_seen": 18503136,
      "step": 21360
    },
    {
      "epoch": 10.073078736445073,
      "grad_norm": 0.6286781430244446,
      "learning_rate": 2.2326912054626772e-05,
      "loss": 0.2149,
      "num_input_tokens_seen": 18507808,
      "step": 21365
    },
    {
      "epoch": 10.075436115040075,
      "grad_norm": 1.0773073434829712,
      "learning_rate": 2.2317151065103813e-05,
      "loss": 0.1369,
      "num_input_tokens_seen": 18512496,
      "step": 21370
    },
    {
      "epoch": 10.077793493635077,
      "grad_norm": 0.8847166895866394,
      "learning_rate": 2.2307390489309865e-05,
      "loss": 0.1485,
      "num_input_tokens_seen": 18516976,
      "step": 21375
    },
    {
      "epoch": 10.08015087223008,
      "grad_norm": 1.2782049179077148,
      "learning_rate": 2.2297630328750146e-05,
      "loss": 0.1571,
      "num_input_tokens_seen": 18521216,
      "step": 21380
    },
    {
      "epoch": 10.082508250825082,
      "grad_norm": 1.2739183902740479,
      "learning_rate": 2.228787058492979e-05,
      "loss": 0.2265,
      "num_input_tokens_seen": 18526768,
      "step": 21385
    },
    {
      "epoch": 10.084865629420085,
      "grad_norm": 0.7549823522567749,
      "learning_rate": 2.2278111259353875e-05,
      "loss": 0.1471,
      "num_input_tokens_seen": 18530624,
      "step": 21390
    },
    {
      "epoch": 10.087223008015087,
      "grad_norm": 0.4522794187068939,
      "learning_rate": 2.2268352353527395e-05,
      "loss": 0.2015,
      "num_input_tokens_seen": 18535696,
      "step": 21395
    },
    {
      "epoch": 10.08958038661009,
      "grad_norm": 0.563372015953064,
      "learning_rate": 2.225859386895533e-05,
      "loss": 0.1749,
      "num_input_tokens_seen": 18539664,
      "step": 21400
    },
    {
      "epoch": 10.08958038661009,
      "eval_loss": 0.20263011753559113,
      "eval_runtime": 22.1393,
      "eval_samples_per_second": 42.594,
      "eval_steps_per_second": 21.32,
      "num_input_tokens_seen": 18539664,
      "step": 21400
    },
    {
      "epoch": 10.091937765205092,
      "grad_norm": 1.5722119808197021,
      "learning_rate": 2.2248835807142525e-05,
      "loss": 0.2831,
      "num_input_tokens_seen": 18544992,
      "step": 21405
    },
    {
      "epoch": 10.094295143800094,
      "grad_norm": 0.9289899468421936,
      "learning_rate": 2.2239078169593826e-05,
      "loss": 0.2049,
      "num_input_tokens_seen": 18549200,
      "step": 21410
    },
    {
      "epoch": 10.096652522395097,
      "grad_norm": 1.0996088981628418,
      "learning_rate": 2.222932095781396e-05,
      "loss": 0.2176,
      "num_input_tokens_seen": 18553632,
      "step": 21415
    },
    {
      "epoch": 10.099009900990099,
      "grad_norm": 1.1866761445999146,
      "learning_rate": 2.221956417330762e-05,
      "loss": 0.1582,
      "num_input_tokens_seen": 18557632,
      "step": 21420
    },
    {
      "epoch": 10.101367279585101,
      "grad_norm": 1.3773212432861328,
      "learning_rate": 2.2209807817579438e-05,
      "loss": 0.1747,
      "num_input_tokens_seen": 18562704,
      "step": 21425
    },
    {
      "epoch": 10.103724658180104,
      "grad_norm": 0.6537896394729614,
      "learning_rate": 2.220005189213394e-05,
      "loss": 0.1607,
      "num_input_tokens_seen": 18567216,
      "step": 21430
    },
    {
      "epoch": 10.106082036775106,
      "grad_norm": 0.8447272777557373,
      "learning_rate": 2.2190296398475624e-05,
      "loss": 0.1754,
      "num_input_tokens_seen": 18571856,
      "step": 21435
    },
    {
      "epoch": 10.108439415370109,
      "grad_norm": 0.48076391220092773,
      "learning_rate": 2.2180541338108926e-05,
      "loss": 0.1948,
      "num_input_tokens_seen": 18575936,
      "step": 21440
    },
    {
      "epoch": 10.110796793965111,
      "grad_norm": 0.8844450116157532,
      "learning_rate": 2.2170786712538176e-05,
      "loss": 0.1436,
      "num_input_tokens_seen": 18580064,
      "step": 21445
    },
    {
      "epoch": 10.113154172560114,
      "grad_norm": 0.906254231929779,
      "learning_rate": 2.216103252326768e-05,
      "loss": 0.241,
      "num_input_tokens_seen": 18584336,
      "step": 21450
    },
    {
      "epoch": 10.115511551155116,
      "grad_norm": 1.2791780233383179,
      "learning_rate": 2.2151278771801635e-05,
      "loss": 0.1714,
      "num_input_tokens_seen": 18588656,
      "step": 21455
    },
    {
      "epoch": 10.117868929750118,
      "grad_norm": 0.5213234424591064,
      "learning_rate": 2.21415254596442e-05,
      "loss": 0.1066,
      "num_input_tokens_seen": 18592096,
      "step": 21460
    },
    {
      "epoch": 10.12022630834512,
      "grad_norm": 0.6419104933738708,
      "learning_rate": 2.213177258829947e-05,
      "loss": 0.1511,
      "num_input_tokens_seen": 18596944,
      "step": 21465
    },
    {
      "epoch": 10.122583686940123,
      "grad_norm": 1.1777397394180298,
      "learning_rate": 2.2122020159271445e-05,
      "loss": 0.2641,
      "num_input_tokens_seen": 18601856,
      "step": 21470
    },
    {
      "epoch": 10.124941065535126,
      "grad_norm": 0.7773908376693726,
      "learning_rate": 2.2112268174064075e-05,
      "loss": 0.2045,
      "num_input_tokens_seen": 18607088,
      "step": 21475
    },
    {
      "epoch": 10.127298444130128,
      "grad_norm": 1.3176002502441406,
      "learning_rate": 2.2102516634181253e-05,
      "loss": 0.1986,
      "num_input_tokens_seen": 18611888,
      "step": 21480
    },
    {
      "epoch": 10.12965582272513,
      "grad_norm": 0.5798171758651733,
      "learning_rate": 2.209276554112677e-05,
      "loss": 0.1332,
      "num_input_tokens_seen": 18615904,
      "step": 21485
    },
    {
      "epoch": 10.132013201320133,
      "grad_norm": 1.2305279970169067,
      "learning_rate": 2.2083014896404384e-05,
      "loss": 0.1731,
      "num_input_tokens_seen": 18620224,
      "step": 21490
    },
    {
      "epoch": 10.134370579915135,
      "grad_norm": 0.49054762721061707,
      "learning_rate": 2.207326470151775e-05,
      "loss": 0.1705,
      "num_input_tokens_seen": 18624160,
      "step": 21495
    },
    {
      "epoch": 10.136727958510138,
      "grad_norm": 0.39749541878700256,
      "learning_rate": 2.2063514957970477e-05,
      "loss": 0.1557,
      "num_input_tokens_seen": 18628496,
      "step": 21500
    },
    {
      "epoch": 10.13908533710514,
      "grad_norm": 0.90678471326828,
      "learning_rate": 2.205376566726611e-05,
      "loss": 0.1729,
      "num_input_tokens_seen": 18632304,
      "step": 21505
    },
    {
      "epoch": 10.14144271570014,
      "grad_norm": 1.1650543212890625,
      "learning_rate": 2.204401683090809e-05,
      "loss": 0.111,
      "num_input_tokens_seen": 18637008,
      "step": 21510
    },
    {
      "epoch": 10.143800094295143,
      "grad_norm": 1.2698523998260498,
      "learning_rate": 2.203426845039982e-05,
      "loss": 0.2899,
      "num_input_tokens_seen": 18641392,
      "step": 21515
    },
    {
      "epoch": 10.146157472890145,
      "grad_norm": 1.0070009231567383,
      "learning_rate": 2.202452052724464e-05,
      "loss": 0.1683,
      "num_input_tokens_seen": 18646368,
      "step": 21520
    },
    {
      "epoch": 10.148514851485148,
      "grad_norm": 0.9778646230697632,
      "learning_rate": 2.2014773062945777e-05,
      "loss": 0.1871,
      "num_input_tokens_seen": 18650240,
      "step": 21525
    },
    {
      "epoch": 10.15087223008015,
      "grad_norm": 1.531511664390564,
      "learning_rate": 2.2005026059006427e-05,
      "loss": 0.1905,
      "num_input_tokens_seen": 18655136,
      "step": 21530
    },
    {
      "epoch": 10.153229608675153,
      "grad_norm": 1.5466982126235962,
      "learning_rate": 2.1995279516929695e-05,
      "loss": 0.1694,
      "num_input_tokens_seen": 18659632,
      "step": 21535
    },
    {
      "epoch": 10.155586987270155,
      "grad_norm": 0.8164313435554504,
      "learning_rate": 2.1985533438218613e-05,
      "loss": 0.1124,
      "num_input_tokens_seen": 18664080,
      "step": 21540
    },
    {
      "epoch": 10.157944365865157,
      "grad_norm": 0.5665033459663391,
      "learning_rate": 2.197578782437617e-05,
      "loss": 0.172,
      "num_input_tokens_seen": 18668272,
      "step": 21545
    },
    {
      "epoch": 10.16030174446016,
      "grad_norm": 0.5709864497184753,
      "learning_rate": 2.196604267690524e-05,
      "loss": 0.1371,
      "num_input_tokens_seen": 18672352,
      "step": 21550
    },
    {
      "epoch": 10.162659123055162,
      "grad_norm": 1.0967538356781006,
      "learning_rate": 2.195629799730865e-05,
      "loss": 0.3057,
      "num_input_tokens_seen": 18678400,
      "step": 21555
    },
    {
      "epoch": 10.165016501650165,
      "grad_norm": 0.6776449084281921,
      "learning_rate": 2.1946553787089173e-05,
      "loss": 0.1474,
      "num_input_tokens_seen": 18682608,
      "step": 21560
    },
    {
      "epoch": 10.167373880245167,
      "grad_norm": 0.7764477729797363,
      "learning_rate": 2.193681004774947e-05,
      "loss": 0.2337,
      "num_input_tokens_seen": 18686832,
      "step": 21565
    },
    {
      "epoch": 10.16973125884017,
      "grad_norm": 1.7530282735824585,
      "learning_rate": 2.1927066780792154e-05,
      "loss": 0.179,
      "num_input_tokens_seen": 18691808,
      "step": 21570
    },
    {
      "epoch": 10.172088637435172,
      "grad_norm": 1.1429414749145508,
      "learning_rate": 2.191732398771975e-05,
      "loss": 0.1651,
      "num_input_tokens_seen": 18697088,
      "step": 21575
    },
    {
      "epoch": 10.174446016030174,
      "grad_norm": 1.61149263381958,
      "learning_rate": 2.1907581670034725e-05,
      "loss": 0.1513,
      "num_input_tokens_seen": 18701440,
      "step": 21580
    },
    {
      "epoch": 10.176803394625177,
      "grad_norm": 0.9761525988578796,
      "learning_rate": 2.189783982923948e-05,
      "loss": 0.1992,
      "num_input_tokens_seen": 18705472,
      "step": 21585
    },
    {
      "epoch": 10.179160773220179,
      "grad_norm": 1.0609108209609985,
      "learning_rate": 2.1888098466836303e-05,
      "loss": 0.153,
      "num_input_tokens_seen": 18709616,
      "step": 21590
    },
    {
      "epoch": 10.181518151815181,
      "grad_norm": 0.47730517387390137,
      "learning_rate": 2.1878357584327457e-05,
      "loss": 0.1633,
      "num_input_tokens_seen": 18714176,
      "step": 21595
    },
    {
      "epoch": 10.183875530410184,
      "grad_norm": 0.3511694371700287,
      "learning_rate": 2.1868617183215103e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 18718016,
      "step": 21600
    },
    {
      "epoch": 10.183875530410184,
      "eval_loss": 0.20161639153957367,
      "eval_runtime": 22.1434,
      "eval_samples_per_second": 42.586,
      "eval_steps_per_second": 21.316,
      "num_input_tokens_seen": 18718016,
      "step": 21600
    },
    {
      "epoch": 10.186232909005186,
      "grad_norm": 2.109114408493042,
      "learning_rate": 2.1858877265001327e-05,
      "loss": 0.2437,
      "num_input_tokens_seen": 18723008,
      "step": 21605
    },
    {
      "epoch": 10.188590287600189,
      "grad_norm": 0.7895712852478027,
      "learning_rate": 2.184913783118816e-05,
      "loss": 0.1984,
      "num_input_tokens_seen": 18727184,
      "step": 21610
    },
    {
      "epoch": 10.190947666195191,
      "grad_norm": 0.49853771924972534,
      "learning_rate": 2.1839398883277522e-05,
      "loss": 0.1824,
      "num_input_tokens_seen": 18731072,
      "step": 21615
    },
    {
      "epoch": 10.193305044790193,
      "grad_norm": 0.7030458450317383,
      "learning_rate": 2.182966042277129e-05,
      "loss": 0.2067,
      "num_input_tokens_seen": 18735424,
      "step": 21620
    },
    {
      "epoch": 10.195662423385196,
      "grad_norm": 1.3064957857131958,
      "learning_rate": 2.181992245117128e-05,
      "loss": 0.1377,
      "num_input_tokens_seen": 18739856,
      "step": 21625
    },
    {
      "epoch": 10.198019801980198,
      "grad_norm": 1.785194993019104,
      "learning_rate": 2.181018496997918e-05,
      "loss": 0.1556,
      "num_input_tokens_seen": 18744720,
      "step": 21630
    },
    {
      "epoch": 10.2003771805752,
      "grad_norm": 0.975426435470581,
      "learning_rate": 2.1800447980696648e-05,
      "loss": 0.1806,
      "num_input_tokens_seen": 18748656,
      "step": 21635
    },
    {
      "epoch": 10.202734559170203,
      "grad_norm": 0.6193059682846069,
      "learning_rate": 2.1790711484825248e-05,
      "loss": 0.1862,
      "num_input_tokens_seen": 18751968,
      "step": 21640
    },
    {
      "epoch": 10.205091937765205,
      "grad_norm": 0.8036287426948547,
      "learning_rate": 2.178097548386646e-05,
      "loss": 0.1977,
      "num_input_tokens_seen": 18756032,
      "step": 21645
    },
    {
      "epoch": 10.207449316360208,
      "grad_norm": 1.4222910404205322,
      "learning_rate": 2.1771239979321712e-05,
      "loss": 0.1709,
      "num_input_tokens_seen": 18759856,
      "step": 21650
    },
    {
      "epoch": 10.20980669495521,
      "grad_norm": 0.5200932621955872,
      "learning_rate": 2.1761504972692327e-05,
      "loss": 0.1254,
      "num_input_tokens_seen": 18764064,
      "step": 21655
    },
    {
      "epoch": 10.212164073550213,
      "grad_norm": 0.7285264134407043,
      "learning_rate": 2.1751770465479572e-05,
      "loss": 0.1922,
      "num_input_tokens_seen": 18768512,
      "step": 21660
    },
    {
      "epoch": 10.214521452145215,
      "grad_norm": 1.4230315685272217,
      "learning_rate": 2.174203645918464e-05,
      "loss": 0.172,
      "num_input_tokens_seen": 18772544,
      "step": 21665
    },
    {
      "epoch": 10.216878830740217,
      "grad_norm": 0.9207444787025452,
      "learning_rate": 2.1732302955308624e-05,
      "loss": 0.1189,
      "num_input_tokens_seen": 18776656,
      "step": 21670
    },
    {
      "epoch": 10.21923620933522,
      "grad_norm": 0.4633946418762207,
      "learning_rate": 2.172256995535255e-05,
      "loss": 0.1763,
      "num_input_tokens_seen": 18781200,
      "step": 21675
    },
    {
      "epoch": 10.221593587930222,
      "grad_norm": 1.0236998796463013,
      "learning_rate": 2.171283746081739e-05,
      "loss": 0.1553,
      "num_input_tokens_seen": 18785024,
      "step": 21680
    },
    {
      "epoch": 10.223950966525225,
      "grad_norm": 1.1673080921173096,
      "learning_rate": 2.1703105473203988e-05,
      "loss": 0.1656,
      "num_input_tokens_seen": 18789616,
      "step": 21685
    },
    {
      "epoch": 10.226308345120227,
      "grad_norm": 0.4742825925350189,
      "learning_rate": 2.1693373994013168e-05,
      "loss": 0.13,
      "num_input_tokens_seen": 18793872,
      "step": 21690
    },
    {
      "epoch": 10.22866572371523,
      "grad_norm": 1.2681329250335693,
      "learning_rate": 2.168364302474562e-05,
      "loss": 0.1336,
      "num_input_tokens_seen": 18797920,
      "step": 21695
    },
    {
      "epoch": 10.231023102310232,
      "grad_norm": 1.1898382902145386,
      "learning_rate": 2.167391256690199e-05,
      "loss": 0.1522,
      "num_input_tokens_seen": 18802784,
      "step": 21700
    },
    {
      "epoch": 10.233380480905234,
      "grad_norm": 0.6437375545501709,
      "learning_rate": 2.1664182621982855e-05,
      "loss": 0.1371,
      "num_input_tokens_seen": 18807200,
      "step": 21705
    },
    {
      "epoch": 10.235737859500237,
      "grad_norm": 1.2961708307266235,
      "learning_rate": 2.1654453191488673e-05,
      "loss": 0.1759,
      "num_input_tokens_seen": 18812544,
      "step": 21710
    },
    {
      "epoch": 10.238095238095237,
      "grad_norm": 0.5781790614128113,
      "learning_rate": 2.1644724276919846e-05,
      "loss": 0.2132,
      "num_input_tokens_seen": 18816752,
      "step": 21715
    },
    {
      "epoch": 10.24045261669024,
      "grad_norm": 1.3901443481445312,
      "learning_rate": 2.1634995879776715e-05,
      "loss": 0.2407,
      "num_input_tokens_seen": 18821456,
      "step": 21720
    },
    {
      "epoch": 10.242809995285242,
      "grad_norm": 1.144775629043579,
      "learning_rate": 2.162526800155949e-05,
      "loss": 0.2118,
      "num_input_tokens_seen": 18826032,
      "step": 21725
    },
    {
      "epoch": 10.245167373880244,
      "grad_norm": 0.8728311061859131,
      "learning_rate": 2.1615540643768363e-05,
      "loss": 0.1057,
      "num_input_tokens_seen": 18830304,
      "step": 21730
    },
    {
      "epoch": 10.247524752475247,
      "grad_norm": 1.1209158897399902,
      "learning_rate": 2.160581380790339e-05,
      "loss": 0.162,
      "num_input_tokens_seen": 18834512,
      "step": 21735
    },
    {
      "epoch": 10.24988213107025,
      "grad_norm": 1.379789113998413,
      "learning_rate": 2.1596087495464586e-05,
      "loss": 0.1672,
      "num_input_tokens_seen": 18838416,
      "step": 21740
    },
    {
      "epoch": 10.252239509665252,
      "grad_norm": 1.3350800275802612,
      "learning_rate": 2.1586361707951866e-05,
      "loss": 0.1691,
      "num_input_tokens_seen": 18842912,
      "step": 21745
    },
    {
      "epoch": 10.254596888260254,
      "grad_norm": 0.4627259373664856,
      "learning_rate": 2.157663644686507e-05,
      "loss": 0.156,
      "num_input_tokens_seen": 18846944,
      "step": 21750
    },
    {
      "epoch": 10.256954266855256,
      "grad_norm": 0.2802494764328003,
      "learning_rate": 2.156691171370396e-05,
      "loss": 0.1441,
      "num_input_tokens_seen": 18851040,
      "step": 21755
    },
    {
      "epoch": 10.259311645450259,
      "grad_norm": 0.65956711769104,
      "learning_rate": 2.1557187509968195e-05,
      "loss": 0.226,
      "num_input_tokens_seen": 18855200,
      "step": 21760
    },
    {
      "epoch": 10.261669024045261,
      "grad_norm": 1.3859692811965942,
      "learning_rate": 2.1547463837157382e-05,
      "loss": 0.1821,
      "num_input_tokens_seen": 18859408,
      "step": 21765
    },
    {
      "epoch": 10.264026402640264,
      "grad_norm": 1.0332417488098145,
      "learning_rate": 2.1537740696771045e-05,
      "loss": 0.1722,
      "num_input_tokens_seen": 18863216,
      "step": 21770
    },
    {
      "epoch": 10.266383781235266,
      "grad_norm": 1.270840048789978,
      "learning_rate": 2.1528018090308587e-05,
      "loss": 0.1766,
      "num_input_tokens_seen": 18867184,
      "step": 21775
    },
    {
      "epoch": 10.268741159830268,
      "grad_norm": 2.054513692855835,
      "learning_rate": 2.151829601926938e-05,
      "loss": 0.2073,
      "num_input_tokens_seen": 18871856,
      "step": 21780
    },
    {
      "epoch": 10.27109853842527,
      "grad_norm": 0.5351089239120483,
      "learning_rate": 2.1508574485152684e-05,
      "loss": 0.1593,
      "num_input_tokens_seen": 18876384,
      "step": 21785
    },
    {
      "epoch": 10.273455917020273,
      "grad_norm": 0.659218430519104,
      "learning_rate": 2.1498853489457667e-05,
      "loss": 0.1797,
      "num_input_tokens_seen": 18880672,
      "step": 21790
    },
    {
      "epoch": 10.275813295615276,
      "grad_norm": 0.9596639275550842,
      "learning_rate": 2.1489133033683455e-05,
      "loss": 0.2303,
      "num_input_tokens_seen": 18884240,
      "step": 21795
    },
    {
      "epoch": 10.278170674210278,
      "grad_norm": 1.1014419794082642,
      "learning_rate": 2.1479413119329038e-05,
      "loss": 0.16,
      "num_input_tokens_seen": 18888560,
      "step": 21800
    },
    {
      "epoch": 10.278170674210278,
      "eval_loss": 0.20253358781337738,
      "eval_runtime": 22.1692,
      "eval_samples_per_second": 42.536,
      "eval_steps_per_second": 21.291,
      "num_input_tokens_seen": 18888560,
      "step": 21800
    },
    {
      "epoch": 10.28052805280528,
      "grad_norm": 0.7935968637466431,
      "learning_rate": 2.1469693747893355e-05,
      "loss": 0.1712,
      "num_input_tokens_seen": 18893248,
      "step": 21805
    },
    {
      "epoch": 10.282885431400283,
      "grad_norm": 0.5561950206756592,
      "learning_rate": 2.1459974920875274e-05,
      "loss": 0.1741,
      "num_input_tokens_seen": 18898320,
      "step": 21810
    },
    {
      "epoch": 10.285242809995285,
      "grad_norm": 0.3435556888580322,
      "learning_rate": 2.145025663977354e-05,
      "loss": 0.1272,
      "num_input_tokens_seen": 18902720,
      "step": 21815
    },
    {
      "epoch": 10.287600188590288,
      "grad_norm": 0.7926916480064392,
      "learning_rate": 2.1440538906086844e-05,
      "loss": 0.1326,
      "num_input_tokens_seen": 18906912,
      "step": 21820
    },
    {
      "epoch": 10.28995756718529,
      "grad_norm": 0.7664534449577332,
      "learning_rate": 2.1430821721313782e-05,
      "loss": 0.1702,
      "num_input_tokens_seen": 18911728,
      "step": 21825
    },
    {
      "epoch": 10.292314945780292,
      "grad_norm": 0.9693820476531982,
      "learning_rate": 2.142110508695286e-05,
      "loss": 0.1488,
      "num_input_tokens_seen": 18916000,
      "step": 21830
    },
    {
      "epoch": 10.294672324375295,
      "grad_norm": 1.313506007194519,
      "learning_rate": 2.1411389004502515e-05,
      "loss": 0.2082,
      "num_input_tokens_seen": 18920320,
      "step": 21835
    },
    {
      "epoch": 10.297029702970297,
      "grad_norm": 1.0475506782531738,
      "learning_rate": 2.140167347546107e-05,
      "loss": 0.2049,
      "num_input_tokens_seen": 18924272,
      "step": 21840
    },
    {
      "epoch": 10.2993870815653,
      "grad_norm": 0.6000980734825134,
      "learning_rate": 2.1391958501326793e-05,
      "loss": 0.2203,
      "num_input_tokens_seen": 18929360,
      "step": 21845
    },
    {
      "epoch": 10.301744460160302,
      "grad_norm": 0.9124854803085327,
      "learning_rate": 2.1382244083597873e-05,
      "loss": 0.1842,
      "num_input_tokens_seen": 18934096,
      "step": 21850
    },
    {
      "epoch": 10.304101838755304,
      "grad_norm": 0.6693963408470154,
      "learning_rate": 2.137253022377237e-05,
      "loss": 0.2035,
      "num_input_tokens_seen": 18938336,
      "step": 21855
    },
    {
      "epoch": 10.306459217350307,
      "grad_norm": 1.2840644121170044,
      "learning_rate": 2.136281692334829e-05,
      "loss": 0.1648,
      "num_input_tokens_seen": 18942704,
      "step": 21860
    },
    {
      "epoch": 10.30881659594531,
      "grad_norm": 1.139608383178711,
      "learning_rate": 2.135310418382356e-05,
      "loss": 0.1707,
      "num_input_tokens_seen": 18946672,
      "step": 21865
    },
    {
      "epoch": 10.311173974540312,
      "grad_norm": 0.5694892406463623,
      "learning_rate": 2.134339200669598e-05,
      "loss": 0.1436,
      "num_input_tokens_seen": 18950800,
      "step": 21870
    },
    {
      "epoch": 10.313531353135314,
      "grad_norm": 1.7392451763153076,
      "learning_rate": 2.133368039346331e-05,
      "loss": 0.2342,
      "num_input_tokens_seen": 18955888,
      "step": 21875
    },
    {
      "epoch": 10.315888731730317,
      "grad_norm": 0.7927244305610657,
      "learning_rate": 2.1323969345623195e-05,
      "loss": 0.1868,
      "num_input_tokens_seen": 18959792,
      "step": 21880
    },
    {
      "epoch": 10.318246110325319,
      "grad_norm": 0.5544301271438599,
      "learning_rate": 2.1314258864673207e-05,
      "loss": 0.1854,
      "num_input_tokens_seen": 18964224,
      "step": 21885
    },
    {
      "epoch": 10.320603488920321,
      "grad_norm": 2.0040829181671143,
      "learning_rate": 2.130454895211082e-05,
      "loss": 0.2139,
      "num_input_tokens_seen": 18968368,
      "step": 21890
    },
    {
      "epoch": 10.322960867515324,
      "grad_norm": 1.0951937437057495,
      "learning_rate": 2.129483960943342e-05,
      "loss": 0.15,
      "num_input_tokens_seen": 18973152,
      "step": 21895
    },
    {
      "epoch": 10.325318246110326,
      "grad_norm": 2.018706798553467,
      "learning_rate": 2.128513083813831e-05,
      "loss": 0.1824,
      "num_input_tokens_seen": 18976816,
      "step": 21900
    },
    {
      "epoch": 10.327675624705329,
      "grad_norm": 0.6421876549720764,
      "learning_rate": 2.1275422639722724e-05,
      "loss": 0.0747,
      "num_input_tokens_seen": 18980352,
      "step": 21905
    },
    {
      "epoch": 10.33003300330033,
      "grad_norm": 0.9252476692199707,
      "learning_rate": 2.126571501568376e-05,
      "loss": 0.1356,
      "num_input_tokens_seen": 18984288,
      "step": 21910
    },
    {
      "epoch": 10.332390381895332,
      "grad_norm": 1.3887465000152588,
      "learning_rate": 2.1256007967518478e-05,
      "loss": 0.1526,
      "num_input_tokens_seen": 18988544,
      "step": 21915
    },
    {
      "epoch": 10.334747760490334,
      "grad_norm": 0.7868514060974121,
      "learning_rate": 2.124630149672381e-05,
      "loss": 0.157,
      "num_input_tokens_seen": 18992448,
      "step": 21920
    },
    {
      "epoch": 10.337105139085336,
      "grad_norm": 0.5851542353630066,
      "learning_rate": 2.1236595604796624e-05,
      "loss": 0.1998,
      "num_input_tokens_seen": 18997696,
      "step": 21925
    },
    {
      "epoch": 10.339462517680339,
      "grad_norm": 0.5008936524391174,
      "learning_rate": 2.1226890293233693e-05,
      "loss": 0.1849,
      "num_input_tokens_seen": 19001856,
      "step": 21930
    },
    {
      "epoch": 10.341819896275341,
      "grad_norm": 0.9699210524559021,
      "learning_rate": 2.1217185563531694e-05,
      "loss": 0.2031,
      "num_input_tokens_seen": 19006672,
      "step": 21935
    },
    {
      "epoch": 10.344177274870344,
      "grad_norm": 0.7290472388267517,
      "learning_rate": 2.120748141718721e-05,
      "loss": 0.2124,
      "num_input_tokens_seen": 19010432,
      "step": 21940
    },
    {
      "epoch": 10.346534653465346,
      "grad_norm": 1.373564600944519,
      "learning_rate": 2.1197777855696765e-05,
      "loss": 0.1841,
      "num_input_tokens_seen": 19014720,
      "step": 21945
    },
    {
      "epoch": 10.348892032060348,
      "grad_norm": 0.8721588850021362,
      "learning_rate": 2.1188074880556746e-05,
      "loss": 0.1375,
      "num_input_tokens_seen": 19018624,
      "step": 21950
    },
    {
      "epoch": 10.35124941065535,
      "grad_norm": 2.1574771404266357,
      "learning_rate": 2.1178372493263495e-05,
      "loss": 0.1612,
      "num_input_tokens_seen": 19022880,
      "step": 21955
    },
    {
      "epoch": 10.353606789250353,
      "grad_norm": 1.0745296478271484,
      "learning_rate": 2.116867069531322e-05,
      "loss": 0.186,
      "num_input_tokens_seen": 19026288,
      "step": 21960
    },
    {
      "epoch": 10.355964167845356,
      "grad_norm": 1.140344500541687,
      "learning_rate": 2.1158969488202073e-05,
      "loss": 0.1633,
      "num_input_tokens_seen": 19030720,
      "step": 21965
    },
    {
      "epoch": 10.358321546440358,
      "grad_norm": 0.9148508310317993,
      "learning_rate": 2.114926887342611e-05,
      "loss": 0.1509,
      "num_input_tokens_seen": 19035856,
      "step": 21970
    },
    {
      "epoch": 10.36067892503536,
      "grad_norm": 1.655653476715088,
      "learning_rate": 2.113956885248127e-05,
      "loss": 0.2162,
      "num_input_tokens_seen": 19040464,
      "step": 21975
    },
    {
      "epoch": 10.363036303630363,
      "grad_norm": 1.0980713367462158,
      "learning_rate": 2.112986942686342e-05,
      "loss": 0.167,
      "num_input_tokens_seen": 19044592,
      "step": 21980
    },
    {
      "epoch": 10.365393682225365,
      "grad_norm": 0.4396168887615204,
      "learning_rate": 2.112017059806835e-05,
      "loss": 0.1523,
      "num_input_tokens_seen": 19048640,
      "step": 21985
    },
    {
      "epoch": 10.367751060820368,
      "grad_norm": 1.0368527173995972,
      "learning_rate": 2.1110472367591724e-05,
      "loss": 0.2252,
      "num_input_tokens_seen": 19053008,
      "step": 21990
    },
    {
      "epoch": 10.37010843941537,
      "grad_norm": 1.4365087747573853,
      "learning_rate": 2.1100774736929145e-05,
      "loss": 0.1535,
      "num_input_tokens_seen": 19057808,
      "step": 21995
    },
    {
      "epoch": 10.372465818010372,
      "grad_norm": 1.5051605701446533,
      "learning_rate": 2.10910777075761e-05,
      "loss": 0.1502,
      "num_input_tokens_seen": 19061328,
      "step": 22000
    },
    {
      "epoch": 10.372465818010372,
      "eval_loss": 0.20590490102767944,
      "eval_runtime": 22.1934,
      "eval_samples_per_second": 42.49,
      "eval_steps_per_second": 21.268,
      "num_input_tokens_seen": 19061328,
      "step": 22000
    },
    {
      "epoch": 10.374823196605375,
      "grad_norm": 1.7246954441070557,
      "learning_rate": 2.108138128102799e-05,
      "loss": 0.1351,
      "num_input_tokens_seen": 19065376,
      "step": 22005
    },
    {
      "epoch": 10.377180575200377,
      "grad_norm": 0.4703231453895569,
      "learning_rate": 2.107168545878014e-05,
      "loss": 0.1951,
      "num_input_tokens_seen": 19069360,
      "step": 22010
    },
    {
      "epoch": 10.37953795379538,
      "grad_norm": 0.5695770978927612,
      "learning_rate": 2.106199024232775e-05,
      "loss": 0.1814,
      "num_input_tokens_seen": 19073296,
      "step": 22015
    },
    {
      "epoch": 10.381895332390382,
      "grad_norm": 0.9028769731521606,
      "learning_rate": 2.105229563316595e-05,
      "loss": 0.1451,
      "num_input_tokens_seen": 19077008,
      "step": 22020
    },
    {
      "epoch": 10.384252710985384,
      "grad_norm": 1.227699875831604,
      "learning_rate": 2.1042601632789784e-05,
      "loss": 0.1943,
      "num_input_tokens_seen": 19081856,
      "step": 22025
    },
    {
      "epoch": 10.386610089580387,
      "grad_norm": 1.2260466814041138,
      "learning_rate": 2.103290824269417e-05,
      "loss": 0.1601,
      "num_input_tokens_seen": 19086576,
      "step": 22030
    },
    {
      "epoch": 10.38896746817539,
      "grad_norm": 0.7542808055877686,
      "learning_rate": 2.1023215464373965e-05,
      "loss": 0.1338,
      "num_input_tokens_seen": 19091408,
      "step": 22035
    },
    {
      "epoch": 10.391324846770392,
      "grad_norm": 0.910786509513855,
      "learning_rate": 2.1013523299323908e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 19096608,
      "step": 22040
    },
    {
      "epoch": 10.393682225365394,
      "grad_norm": 0.6138972640037537,
      "learning_rate": 2.1003831749038654e-05,
      "loss": 0.1432,
      "num_input_tokens_seen": 19100768,
      "step": 22045
    },
    {
      "epoch": 10.396039603960396,
      "grad_norm": 1.2817987203598022,
      "learning_rate": 2.099414081501277e-05,
      "loss": 0.1949,
      "num_input_tokens_seen": 19105440,
      "step": 22050
    },
    {
      "epoch": 10.398396982555399,
      "grad_norm": 0.6749522089958191,
      "learning_rate": 2.09844504987407e-05,
      "loss": 0.1626,
      "num_input_tokens_seen": 19109488,
      "step": 22055
    },
    {
      "epoch": 10.400754361150401,
      "grad_norm": 0.661352276802063,
      "learning_rate": 2.097476080171683e-05,
      "loss": 0.1255,
      "num_input_tokens_seen": 19113808,
      "step": 22060
    },
    {
      "epoch": 10.403111739745404,
      "grad_norm": 0.4634983241558075,
      "learning_rate": 2.0965071725435436e-05,
      "loss": 0.243,
      "num_input_tokens_seen": 19117488,
      "step": 22065
    },
    {
      "epoch": 10.405469118340406,
      "grad_norm": 0.898535966873169,
      "learning_rate": 2.0955383271390684e-05,
      "loss": 0.2369,
      "num_input_tokens_seen": 19122448,
      "step": 22070
    },
    {
      "epoch": 10.407826496935408,
      "grad_norm": 1.3856817483901978,
      "learning_rate": 2.094569544107666e-05,
      "loss": 0.1657,
      "num_input_tokens_seen": 19126384,
      "step": 22075
    },
    {
      "epoch": 10.41018387553041,
      "grad_norm": 0.987543523311615,
      "learning_rate": 2.093600823598735e-05,
      "loss": 0.2778,
      "num_input_tokens_seen": 19130944,
      "step": 22080
    },
    {
      "epoch": 10.412541254125413,
      "grad_norm": 0.8714144825935364,
      "learning_rate": 2.092632165761663e-05,
      "loss": 0.1873,
      "num_input_tokens_seen": 19135232,
      "step": 22085
    },
    {
      "epoch": 10.414898632720416,
      "grad_norm": 0.5656450390815735,
      "learning_rate": 2.091663570745832e-05,
      "loss": 0.2046,
      "num_input_tokens_seen": 19139664,
      "step": 22090
    },
    {
      "epoch": 10.417256011315418,
      "grad_norm": 0.824065625667572,
      "learning_rate": 2.0906950387006086e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 19144336,
      "step": 22095
    },
    {
      "epoch": 10.41961338991042,
      "grad_norm": 1.1975051164627075,
      "learning_rate": 2.0897265697753543e-05,
      "loss": 0.1696,
      "num_input_tokens_seen": 19150320,
      "step": 22100
    },
    {
      "epoch": 10.421970768505423,
      "grad_norm": 0.5114638805389404,
      "learning_rate": 2.088758164119419e-05,
      "loss": 0.1858,
      "num_input_tokens_seen": 19154768,
      "step": 22105
    },
    {
      "epoch": 10.424328147100425,
      "grad_norm": 1.5094640254974365,
      "learning_rate": 2.0877898218821428e-05,
      "loss": 0.1675,
      "num_input_tokens_seen": 19158976,
      "step": 22110
    },
    {
      "epoch": 10.426685525695426,
      "grad_norm": 1.756321907043457,
      "learning_rate": 2.0868215432128565e-05,
      "loss": 0.2513,
      "num_input_tokens_seen": 19162960,
      "step": 22115
    },
    {
      "epoch": 10.429042904290428,
      "grad_norm": 2.286142110824585,
      "learning_rate": 2.0858533282608796e-05,
      "loss": 0.303,
      "num_input_tokens_seen": 19167296,
      "step": 22120
    },
    {
      "epoch": 10.43140028288543,
      "grad_norm": 1.196448802947998,
      "learning_rate": 2.084885177175524e-05,
      "loss": 0.2234,
      "num_input_tokens_seen": 19171104,
      "step": 22125
    },
    {
      "epoch": 10.433757661480433,
      "grad_norm": 1.2884224653244019,
      "learning_rate": 2.0839170901060917e-05,
      "loss": 0.1833,
      "num_input_tokens_seen": 19174880,
      "step": 22130
    },
    {
      "epoch": 10.436115040075435,
      "grad_norm": 0.6025993824005127,
      "learning_rate": 2.082949067201872e-05,
      "loss": 0.1488,
      "num_input_tokens_seen": 19179296,
      "step": 22135
    },
    {
      "epoch": 10.438472418670438,
      "grad_norm": 0.720244288444519,
      "learning_rate": 2.0819811086121475e-05,
      "loss": 0.1907,
      "num_input_tokens_seen": 19183504,
      "step": 22140
    },
    {
      "epoch": 10.44082979726544,
      "grad_norm": 1.3040833473205566,
      "learning_rate": 2.08101321448619e-05,
      "loss": 0.2111,
      "num_input_tokens_seen": 19188096,
      "step": 22145
    },
    {
      "epoch": 10.443187175860443,
      "grad_norm": 0.6896134614944458,
      "learning_rate": 2.080045384973259e-05,
      "loss": 0.1799,
      "num_input_tokens_seen": 19192464,
      "step": 22150
    },
    {
      "epoch": 10.445544554455445,
      "grad_norm": 0.7208622097969055,
      "learning_rate": 2.0790776202226082e-05,
      "loss": 0.1775,
      "num_input_tokens_seen": 19196576,
      "step": 22155
    },
    {
      "epoch": 10.447901933050447,
      "grad_norm": 0.6371626853942871,
      "learning_rate": 2.078109920383477e-05,
      "loss": 0.2008,
      "num_input_tokens_seen": 19200784,
      "step": 22160
    },
    {
      "epoch": 10.45025931164545,
      "grad_norm": 0.5537049174308777,
      "learning_rate": 2.0771422856050978e-05,
      "loss": 0.2177,
      "num_input_tokens_seen": 19204864,
      "step": 22165
    },
    {
      "epoch": 10.452616690240452,
      "grad_norm": 0.4197816848754883,
      "learning_rate": 2.076174716036693e-05,
      "loss": 0.1929,
      "num_input_tokens_seen": 19209232,
      "step": 22170
    },
    {
      "epoch": 10.454974068835455,
      "grad_norm": 1.042786955833435,
      "learning_rate": 2.075207211827472e-05,
      "loss": 0.1647,
      "num_input_tokens_seen": 19213552,
      "step": 22175
    },
    {
      "epoch": 10.457331447430457,
      "grad_norm": 0.4007701277732849,
      "learning_rate": 2.074239773126638e-05,
      "loss": 0.1635,
      "num_input_tokens_seen": 19218224,
      "step": 22180
    },
    {
      "epoch": 10.45968882602546,
      "grad_norm": 0.40177804231643677,
      "learning_rate": 2.073272400083382e-05,
      "loss": 0.1067,
      "num_input_tokens_seen": 19222880,
      "step": 22185
    },
    {
      "epoch": 10.462046204620462,
      "grad_norm": 1.0317213535308838,
      "learning_rate": 2.072305092846883e-05,
      "loss": 0.2032,
      "num_input_tokens_seen": 19227040,
      "step": 22190
    },
    {
      "epoch": 10.464403583215464,
      "grad_norm": 1.9812241792678833,
      "learning_rate": 2.0713378515663152e-05,
      "loss": 0.162,
      "num_input_tokens_seen": 19231008,
      "step": 22195
    },
    {
      "epoch": 10.466760961810467,
      "grad_norm": 0.6799085140228271,
      "learning_rate": 2.070370676390836e-05,
      "loss": 0.174,
      "num_input_tokens_seen": 19236176,
      "step": 22200
    },
    {
      "epoch": 10.466760961810467,
      "eval_loss": 0.2043340653181076,
      "eval_runtime": 22.1581,
      "eval_samples_per_second": 42.558,
      "eval_steps_per_second": 21.302,
      "num_input_tokens_seen": 19236176,
      "step": 22200
    },
    {
      "epoch": 10.469118340405469,
      "grad_norm": 1.0224955081939697,
      "learning_rate": 2.0694035674695974e-05,
      "loss": 0.2041,
      "num_input_tokens_seen": 19239840,
      "step": 22205
    },
    {
      "epoch": 10.471475719000471,
      "grad_norm": 2.133862018585205,
      "learning_rate": 2.0684365249517416e-05,
      "loss": 0.2388,
      "num_input_tokens_seen": 19244080,
      "step": 22210
    },
    {
      "epoch": 10.473833097595474,
      "grad_norm": 1.1526755094528198,
      "learning_rate": 2.067469548986396e-05,
      "loss": 0.1863,
      "num_input_tokens_seen": 19247856,
      "step": 22215
    },
    {
      "epoch": 10.476190476190476,
      "grad_norm": 1.1017919778823853,
      "learning_rate": 2.066502639722681e-05,
      "loss": 0.1999,
      "num_input_tokens_seen": 19251760,
      "step": 22220
    },
    {
      "epoch": 10.478547854785479,
      "grad_norm": 2.0613203048706055,
      "learning_rate": 2.065535797309708e-05,
      "loss": 0.1663,
      "num_input_tokens_seen": 19256480,
      "step": 22225
    },
    {
      "epoch": 10.480905233380481,
      "grad_norm": 1.519866943359375,
      "learning_rate": 2.0645690218965736e-05,
      "loss": 0.1612,
      "num_input_tokens_seen": 19260720,
      "step": 22230
    },
    {
      "epoch": 10.483262611975483,
      "grad_norm": 0.8844563961029053,
      "learning_rate": 2.063602313632369e-05,
      "loss": 0.1798,
      "num_input_tokens_seen": 19264672,
      "step": 22235
    },
    {
      "epoch": 10.485619990570486,
      "grad_norm": 0.7617820501327515,
      "learning_rate": 2.0626356726661704e-05,
      "loss": 0.1227,
      "num_input_tokens_seen": 19269312,
      "step": 22240
    },
    {
      "epoch": 10.487977369165488,
      "grad_norm": 0.5408512353897095,
      "learning_rate": 2.0616690991470477e-05,
      "loss": 0.2495,
      "num_input_tokens_seen": 19273920,
      "step": 22245
    },
    {
      "epoch": 10.49033474776049,
      "grad_norm": 0.7934156060218811,
      "learning_rate": 2.0607025932240595e-05,
      "loss": 0.1423,
      "num_input_tokens_seen": 19278112,
      "step": 22250
    },
    {
      "epoch": 10.492692126355493,
      "grad_norm": 1.8315324783325195,
      "learning_rate": 2.059736155046251e-05,
      "loss": 0.162,
      "num_input_tokens_seen": 19282784,
      "step": 22255
    },
    {
      "epoch": 10.495049504950495,
      "grad_norm": 1.2061538696289062,
      "learning_rate": 2.0587697847626603e-05,
      "loss": 0.2183,
      "num_input_tokens_seen": 19287840,
      "step": 22260
    },
    {
      "epoch": 10.497406883545498,
      "grad_norm": 0.827060878276825,
      "learning_rate": 2.057803482522314e-05,
      "loss": 0.1202,
      "num_input_tokens_seen": 19291664,
      "step": 22265
    },
    {
      "epoch": 10.4997642621405,
      "grad_norm": 0.8380371332168579,
      "learning_rate": 2.056837248474227e-05,
      "loss": 0.1369,
      "num_input_tokens_seen": 19295872,
      "step": 22270
    },
    {
      "epoch": 10.502121640735503,
      "grad_norm": 0.4866775572299957,
      "learning_rate": 2.0558710827674064e-05,
      "loss": 0.1655,
      "num_input_tokens_seen": 19300320,
      "step": 22275
    },
    {
      "epoch": 10.504479019330505,
      "grad_norm": 0.7377066016197205,
      "learning_rate": 2.054904985550845e-05,
      "loss": 0.1658,
      "num_input_tokens_seen": 19303872,
      "step": 22280
    },
    {
      "epoch": 10.506836397925507,
      "grad_norm": 1.4753119945526123,
      "learning_rate": 2.0539389569735287e-05,
      "loss": 0.1969,
      "num_input_tokens_seen": 19308656,
      "step": 22285
    },
    {
      "epoch": 10.50919377652051,
      "grad_norm": 1.9238662719726562,
      "learning_rate": 2.052972997184431e-05,
      "loss": 0.2223,
      "num_input_tokens_seen": 19312928,
      "step": 22290
    },
    {
      "epoch": 10.511551155115512,
      "grad_norm": 0.38836613297462463,
      "learning_rate": 2.0520071063325146e-05,
      "loss": 0.1259,
      "num_input_tokens_seen": 19316912,
      "step": 22295
    },
    {
      "epoch": 10.513908533710515,
      "grad_norm": 1.1170001029968262,
      "learning_rate": 2.051041284566732e-05,
      "loss": 0.211,
      "num_input_tokens_seen": 19321040,
      "step": 22300
    },
    {
      "epoch": 10.516265912305517,
      "grad_norm": 0.5383704900741577,
      "learning_rate": 2.050075532036026e-05,
      "loss": 0.1825,
      "num_input_tokens_seen": 19324960,
      "step": 22305
    },
    {
      "epoch": 10.518623290900518,
      "grad_norm": 0.6872249841690063,
      "learning_rate": 2.0491098488893264e-05,
      "loss": 0.1956,
      "num_input_tokens_seen": 19329648,
      "step": 22310
    },
    {
      "epoch": 10.520980669495522,
      "grad_norm": 0.9813538193702698,
      "learning_rate": 2.0481442352755546e-05,
      "loss": 0.1941,
      "num_input_tokens_seen": 19333296,
      "step": 22315
    },
    {
      "epoch": 10.523338048090523,
      "grad_norm": 0.7226388454437256,
      "learning_rate": 2.0471786913436198e-05,
      "loss": 0.1719,
      "num_input_tokens_seen": 19337952,
      "step": 22320
    },
    {
      "epoch": 10.525695426685525,
      "grad_norm": 1.16147780418396,
      "learning_rate": 2.0462132172424218e-05,
      "loss": 0.1878,
      "num_input_tokens_seen": 19341616,
      "step": 22325
    },
    {
      "epoch": 10.528052805280527,
      "grad_norm": 0.9222947359085083,
      "learning_rate": 2.0452478131208484e-05,
      "loss": 0.1684,
      "num_input_tokens_seen": 19345568,
      "step": 22330
    },
    {
      "epoch": 10.53041018387553,
      "grad_norm": 0.40792909264564514,
      "learning_rate": 2.0442824791277765e-05,
      "loss": 0.151,
      "num_input_tokens_seen": 19349568,
      "step": 22335
    },
    {
      "epoch": 10.532767562470532,
      "grad_norm": 0.5532940626144409,
      "learning_rate": 2.0433172154120727e-05,
      "loss": 0.1599,
      "num_input_tokens_seen": 19352816,
      "step": 22340
    },
    {
      "epoch": 10.535124941065535,
      "grad_norm": 0.5954071283340454,
      "learning_rate": 2.0423520221225947e-05,
      "loss": 0.2038,
      "num_input_tokens_seen": 19357392,
      "step": 22345
    },
    {
      "epoch": 10.537482319660537,
      "grad_norm": 0.9024198651313782,
      "learning_rate": 2.0413868994081848e-05,
      "loss": 0.2179,
      "num_input_tokens_seen": 19361376,
      "step": 22350
    },
    {
      "epoch": 10.53983969825554,
      "grad_norm": 1.0160454511642456,
      "learning_rate": 2.0404218474176795e-05,
      "loss": 0.1981,
      "num_input_tokens_seen": 19365680,
      "step": 22355
    },
    {
      "epoch": 10.542197076850542,
      "grad_norm": 0.6217955350875854,
      "learning_rate": 2.0394568662999002e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 19370256,
      "step": 22360
    },
    {
      "epoch": 10.544554455445544,
      "grad_norm": 0.7665778398513794,
      "learning_rate": 2.0384919562036593e-05,
      "loss": 0.1915,
      "num_input_tokens_seen": 19373936,
      "step": 22365
    },
    {
      "epoch": 10.546911834040547,
      "grad_norm": 0.5094192028045654,
      "learning_rate": 2.0375271172777593e-05,
      "loss": 0.2165,
      "num_input_tokens_seen": 19378096,
      "step": 22370
    },
    {
      "epoch": 10.549269212635549,
      "grad_norm": 0.7503447532653809,
      "learning_rate": 2.0365623496709885e-05,
      "loss": 0.1857,
      "num_input_tokens_seen": 19383152,
      "step": 22375
    },
    {
      "epoch": 10.551626591230551,
      "grad_norm": 1.185051441192627,
      "learning_rate": 2.0355976535321283e-05,
      "loss": 0.1396,
      "num_input_tokens_seen": 19388176,
      "step": 22380
    },
    {
      "epoch": 10.553983969825554,
      "grad_norm": 1.5355230569839478,
      "learning_rate": 2.034633029009945e-05,
      "loss": 0.2065,
      "num_input_tokens_seen": 19392720,
      "step": 22385
    },
    {
      "epoch": 10.556341348420556,
      "grad_norm": 2.447253465652466,
      "learning_rate": 2.0336684762531972e-05,
      "loss": 0.217,
      "num_input_tokens_seen": 19396944,
      "step": 22390
    },
    {
      "epoch": 10.558698727015559,
      "grad_norm": 0.5449923276901245,
      "learning_rate": 2.032703995410631e-05,
      "loss": 0.3148,
      "num_input_tokens_seen": 19400576,
      "step": 22395
    },
    {
      "epoch": 10.561056105610561,
      "grad_norm": 0.7239357233047485,
      "learning_rate": 2.031739586630981e-05,
      "loss": 0.1802,
      "num_input_tokens_seen": 19404288,
      "step": 22400
    },
    {
      "epoch": 10.561056105610561,
      "eval_loss": 0.20716425776481628,
      "eval_runtime": 22.1393,
      "eval_samples_per_second": 42.594,
      "eval_steps_per_second": 21.32,
      "num_input_tokens_seen": 19404288,
      "step": 22400
    },
    {
      "epoch": 10.563413484205563,
      "grad_norm": 0.9643945693969727,
      "learning_rate": 2.0307752500629707e-05,
      "loss": 0.2077,
      "num_input_tokens_seen": 19407504,
      "step": 22405
    },
    {
      "epoch": 10.565770862800566,
      "grad_norm": 0.7820362448692322,
      "learning_rate": 2.0298109858553144e-05,
      "loss": 0.1396,
      "num_input_tokens_seen": 19411504,
      "step": 22410
    },
    {
      "epoch": 10.568128241395568,
      "grad_norm": 1.6691079139709473,
      "learning_rate": 2.028846794156712e-05,
      "loss": 0.212,
      "num_input_tokens_seen": 19415856,
      "step": 22415
    },
    {
      "epoch": 10.57048561999057,
      "grad_norm": 0.7030676603317261,
      "learning_rate": 2.027882675115856e-05,
      "loss": 0.2056,
      "num_input_tokens_seen": 19420304,
      "step": 22420
    },
    {
      "epoch": 10.572842998585573,
      "grad_norm": 0.7645986080169678,
      "learning_rate": 2.026918628881423e-05,
      "loss": 0.2047,
      "num_input_tokens_seen": 19424640,
      "step": 22425
    },
    {
      "epoch": 10.575200377180575,
      "grad_norm": 0.35321855545043945,
      "learning_rate": 2.0259546556020833e-05,
      "loss": 0.1618,
      "num_input_tokens_seen": 19429072,
      "step": 22430
    },
    {
      "epoch": 10.577557755775578,
      "grad_norm": 0.5455840229988098,
      "learning_rate": 2.024990755426493e-05,
      "loss": 0.2047,
      "num_input_tokens_seen": 19433232,
      "step": 22435
    },
    {
      "epoch": 10.57991513437058,
      "grad_norm": 1.6569145917892456,
      "learning_rate": 2.0240269285032975e-05,
      "loss": 0.2062,
      "num_input_tokens_seen": 19438240,
      "step": 22440
    },
    {
      "epoch": 10.582272512965583,
      "grad_norm": 0.5187669396400452,
      "learning_rate": 2.0230631749811306e-05,
      "loss": 0.1489,
      "num_input_tokens_seen": 19442832,
      "step": 22445
    },
    {
      "epoch": 10.584629891560585,
      "grad_norm": 0.6775956153869629,
      "learning_rate": 2.0220994950086162e-05,
      "loss": 0.2134,
      "num_input_tokens_seen": 19447152,
      "step": 22450
    },
    {
      "epoch": 10.586987270155587,
      "grad_norm": 1.447052001953125,
      "learning_rate": 2.021135888734365e-05,
      "loss": 0.1252,
      "num_input_tokens_seen": 19451872,
      "step": 22455
    },
    {
      "epoch": 10.58934464875059,
      "grad_norm": 0.43254542350769043,
      "learning_rate": 2.0201723563069783e-05,
      "loss": 0.2084,
      "num_input_tokens_seen": 19456448,
      "step": 22460
    },
    {
      "epoch": 10.591702027345592,
      "grad_norm": 0.6985853314399719,
      "learning_rate": 2.0192088978750433e-05,
      "loss": 0.1353,
      "num_input_tokens_seen": 19460896,
      "step": 22465
    },
    {
      "epoch": 10.594059405940595,
      "grad_norm": 0.9510381817817688,
      "learning_rate": 2.0182455135871385e-05,
      "loss": 0.1499,
      "num_input_tokens_seen": 19465088,
      "step": 22470
    },
    {
      "epoch": 10.596416784535597,
      "grad_norm": 0.5516671538352966,
      "learning_rate": 2.0172822035918305e-05,
      "loss": 0.1798,
      "num_input_tokens_seen": 19469232,
      "step": 22475
    },
    {
      "epoch": 10.5987741631306,
      "grad_norm": 1.239419937133789,
      "learning_rate": 2.016318968037671e-05,
      "loss": 0.2393,
      "num_input_tokens_seen": 19473424,
      "step": 22480
    },
    {
      "epoch": 10.601131541725602,
      "grad_norm": 1.3588494062423706,
      "learning_rate": 2.015355807073206e-05,
      "loss": 0.1872,
      "num_input_tokens_seen": 19477728,
      "step": 22485
    },
    {
      "epoch": 10.603488920320604,
      "grad_norm": 0.4976968765258789,
      "learning_rate": 2.0143927208469664e-05,
      "loss": 0.113,
      "num_input_tokens_seen": 19481888,
      "step": 22490
    },
    {
      "epoch": 10.605846298915607,
      "grad_norm": 0.8832781314849854,
      "learning_rate": 2.0134297095074708e-05,
      "loss": 0.1411,
      "num_input_tokens_seen": 19485872,
      "step": 22495
    },
    {
      "epoch": 10.608203677510609,
      "grad_norm": 0.46328532695770264,
      "learning_rate": 2.0124667732032297e-05,
      "loss": 0.2165,
      "num_input_tokens_seen": 19489952,
      "step": 22500
    },
    {
      "epoch": 10.61056105610561,
      "grad_norm": 1.2089694738388062,
      "learning_rate": 2.011503912082738e-05,
      "loss": 0.1333,
      "num_input_tokens_seen": 19493568,
      "step": 22505
    },
    {
      "epoch": 10.612918434700614,
      "grad_norm": 0.40841442346572876,
      "learning_rate": 2.0105411262944823e-05,
      "loss": 0.1846,
      "num_input_tokens_seen": 19497216,
      "step": 22510
    },
    {
      "epoch": 10.615275813295614,
      "grad_norm": 1.437760353088379,
      "learning_rate": 2.0095784159869366e-05,
      "loss": 0.1776,
      "num_input_tokens_seen": 19501984,
      "step": 22515
    },
    {
      "epoch": 10.617633191890617,
      "grad_norm": 0.719245970249176,
      "learning_rate": 2.0086157813085608e-05,
      "loss": 0.2566,
      "num_input_tokens_seen": 19506528,
      "step": 22520
    },
    {
      "epoch": 10.61999057048562,
      "grad_norm": 0.5190545320510864,
      "learning_rate": 2.0076532224078068e-05,
      "loss": 0.2345,
      "num_input_tokens_seen": 19510784,
      "step": 22525
    },
    {
      "epoch": 10.622347949080622,
      "grad_norm": 1.113452434539795,
      "learning_rate": 2.0066907394331142e-05,
      "loss": 0.2004,
      "num_input_tokens_seen": 19515392,
      "step": 22530
    },
    {
      "epoch": 10.624705327675624,
      "grad_norm": 0.4568403363227844,
      "learning_rate": 2.0057283325329077e-05,
      "loss": 0.1929,
      "num_input_tokens_seen": 19519424,
      "step": 22535
    },
    {
      "epoch": 10.627062706270626,
      "grad_norm": 0.5897207856178284,
      "learning_rate": 2.0047660018556047e-05,
      "loss": 0.2492,
      "num_input_tokens_seen": 19523424,
      "step": 22540
    },
    {
      "epoch": 10.629420084865629,
      "grad_norm": 0.713524580001831,
      "learning_rate": 2.0038037475496075e-05,
      "loss": 0.135,
      "num_input_tokens_seen": 19527904,
      "step": 22545
    },
    {
      "epoch": 10.631777463460631,
      "grad_norm": 1.2570109367370605,
      "learning_rate": 2.0028415697633073e-05,
      "loss": 0.1185,
      "num_input_tokens_seen": 19532192,
      "step": 22550
    },
    {
      "epoch": 10.634134842055634,
      "grad_norm": 1.8772122859954834,
      "learning_rate": 2.0018794686450858e-05,
      "loss": 0.1723,
      "num_input_tokens_seen": 19536992,
      "step": 22555
    },
    {
      "epoch": 10.636492220650636,
      "grad_norm": 0.5378044843673706,
      "learning_rate": 2.0009174443433088e-05,
      "loss": 0.1028,
      "num_input_tokens_seen": 19541168,
      "step": 22560
    },
    {
      "epoch": 10.638849599245638,
      "grad_norm": 0.865265429019928,
      "learning_rate": 1.999955497006334e-05,
      "loss": 0.2351,
      "num_input_tokens_seen": 19545024,
      "step": 22565
    },
    {
      "epoch": 10.64120697784064,
      "grad_norm": 0.705531895160675,
      "learning_rate": 1.9989936267825067e-05,
      "loss": 0.2242,
      "num_input_tokens_seen": 19549120,
      "step": 22570
    },
    {
      "epoch": 10.643564356435643,
      "grad_norm": 0.8891657590866089,
      "learning_rate": 1.9980318338201572e-05,
      "loss": 0.2358,
      "num_input_tokens_seen": 19553520,
      "step": 22575
    },
    {
      "epoch": 10.645921735030646,
      "grad_norm": 1.8386160135269165,
      "learning_rate": 1.997070118267607e-05,
      "loss": 0.1944,
      "num_input_tokens_seen": 19558096,
      "step": 22580
    },
    {
      "epoch": 10.648279113625648,
      "grad_norm": 1.7947449684143066,
      "learning_rate": 1.9961084802731654e-05,
      "loss": 0.1754,
      "num_input_tokens_seen": 19562224,
      "step": 22585
    },
    {
      "epoch": 10.65063649222065,
      "grad_norm": 0.8837350010871887,
      "learning_rate": 1.9951469199851273e-05,
      "loss": 0.2217,
      "num_input_tokens_seen": 19566000,
      "step": 22590
    },
    {
      "epoch": 10.652993870815653,
      "grad_norm": 0.6190546751022339,
      "learning_rate": 1.99418543755178e-05,
      "loss": 0.2114,
      "num_input_tokens_seen": 19569680,
      "step": 22595
    },
    {
      "epoch": 10.655351249410655,
      "grad_norm": 0.5867920517921448,
      "learning_rate": 1.9932240331213936e-05,
      "loss": 0.1499,
      "num_input_tokens_seen": 19574224,
      "step": 22600
    },
    {
      "epoch": 10.655351249410655,
      "eval_loss": 0.20554836094379425,
      "eval_runtime": 22.1566,
      "eval_samples_per_second": 42.561,
      "eval_steps_per_second": 21.303,
      "num_input_tokens_seen": 19574224,
      "step": 22600
    },
    {
      "epoch": 10.657708628005658,
      "grad_norm": 0.5591682195663452,
      "learning_rate": 1.9922627068422297e-05,
      "loss": 0.0968,
      "num_input_tokens_seen": 19578000,
      "step": 22605
    },
    {
      "epoch": 10.66006600660066,
      "grad_norm": 1.1232548952102661,
      "learning_rate": 1.991301458862538e-05,
      "loss": 0.1498,
      "num_input_tokens_seen": 19581552,
      "step": 22610
    },
    {
      "epoch": 10.662423385195662,
      "grad_norm": 1.472653865814209,
      "learning_rate": 1.9903402893305536e-05,
      "loss": 0.1855,
      "num_input_tokens_seen": 19585184,
      "step": 22615
    },
    {
      "epoch": 10.664780763790665,
      "grad_norm": 0.9487419724464417,
      "learning_rate": 1.9893791983945016e-05,
      "loss": 0.1733,
      "num_input_tokens_seen": 19589840,
      "step": 22620
    },
    {
      "epoch": 10.667138142385667,
      "grad_norm": 0.7308679819107056,
      "learning_rate": 1.988418186202594e-05,
      "loss": 0.2456,
      "num_input_tokens_seen": 19593744,
      "step": 22625
    },
    {
      "epoch": 10.66949552098067,
      "grad_norm": 1.317770004272461,
      "learning_rate": 1.98745725290303e-05,
      "loss": 0.1714,
      "num_input_tokens_seen": 19598112,
      "step": 22630
    },
    {
      "epoch": 10.671852899575672,
      "grad_norm": 0.6333796977996826,
      "learning_rate": 1.986496398644e-05,
      "loss": 0.1242,
      "num_input_tokens_seen": 19602320,
      "step": 22635
    },
    {
      "epoch": 10.674210278170674,
      "grad_norm": 0.6361583471298218,
      "learning_rate": 1.9855356235736777e-05,
      "loss": 0.1517,
      "num_input_tokens_seen": 19606416,
      "step": 22640
    },
    {
      "epoch": 10.676567656765677,
      "grad_norm": 0.6287312507629395,
      "learning_rate": 1.9845749278402277e-05,
      "loss": 0.1932,
      "num_input_tokens_seen": 19610896,
      "step": 22645
    },
    {
      "epoch": 10.67892503536068,
      "grad_norm": 1.8276386260986328,
      "learning_rate": 1.9836143115918006e-05,
      "loss": 0.2995,
      "num_input_tokens_seen": 19615904,
      "step": 22650
    },
    {
      "epoch": 10.681282413955682,
      "grad_norm": 0.7043393850326538,
      "learning_rate": 1.9826537749765367e-05,
      "loss": 0.1878,
      "num_input_tokens_seen": 19620096,
      "step": 22655
    },
    {
      "epoch": 10.683639792550684,
      "grad_norm": 0.9505472779273987,
      "learning_rate": 1.9816933181425625e-05,
      "loss": 0.1929,
      "num_input_tokens_seen": 19623712,
      "step": 22660
    },
    {
      "epoch": 10.685997171145686,
      "grad_norm": 0.5923447012901306,
      "learning_rate": 1.9807329412379903e-05,
      "loss": 0.1101,
      "num_input_tokens_seen": 19628432,
      "step": 22665
    },
    {
      "epoch": 10.688354549740689,
      "grad_norm": 2.0359549522399902,
      "learning_rate": 1.9797726444109247e-05,
      "loss": 0.2951,
      "num_input_tokens_seen": 19631952,
      "step": 22670
    },
    {
      "epoch": 10.690711928335691,
      "grad_norm": 1.5066155195236206,
      "learning_rate": 1.9788124278094557e-05,
      "loss": 0.182,
      "num_input_tokens_seen": 19636432,
      "step": 22675
    },
    {
      "epoch": 10.693069306930694,
      "grad_norm": 1.9225910902023315,
      "learning_rate": 1.9778522915816594e-05,
      "loss": 0.1456,
      "num_input_tokens_seen": 19641120,
      "step": 22680
    },
    {
      "epoch": 10.695426685525696,
      "grad_norm": 1.8876867294311523,
      "learning_rate": 1.9768922358756014e-05,
      "loss": 0.2484,
      "num_input_tokens_seen": 19645216,
      "step": 22685
    },
    {
      "epoch": 10.697784064120698,
      "grad_norm": 0.8824220895767212,
      "learning_rate": 1.9759322608393353e-05,
      "loss": 0.209,
      "num_input_tokens_seen": 19649232,
      "step": 22690
    },
    {
      "epoch": 10.700141442715701,
      "grad_norm": 0.560635507106781,
      "learning_rate": 1.9749723666208992e-05,
      "loss": 0.2212,
      "num_input_tokens_seen": 19653344,
      "step": 22695
    },
    {
      "epoch": 10.702498821310703,
      "grad_norm": 0.4042558968067169,
      "learning_rate": 1.9740125533683235e-05,
      "loss": 0.2141,
      "num_input_tokens_seen": 19657904,
      "step": 22700
    },
    {
      "epoch": 10.704856199905706,
      "grad_norm": 0.6656500101089478,
      "learning_rate": 1.9730528212296208e-05,
      "loss": 0.1655,
      "num_input_tokens_seen": 19662368,
      "step": 22705
    },
    {
      "epoch": 10.707213578500706,
      "grad_norm": 0.9934701919555664,
      "learning_rate": 1.9720931703527945e-05,
      "loss": 0.1403,
      "num_input_tokens_seen": 19666608,
      "step": 22710
    },
    {
      "epoch": 10.70957095709571,
      "grad_norm": 1.117411494255066,
      "learning_rate": 1.9711336008858373e-05,
      "loss": 0.1663,
      "num_input_tokens_seen": 19671088,
      "step": 22715
    },
    {
      "epoch": 10.711928335690711,
      "grad_norm": 0.7325334548950195,
      "learning_rate": 1.9701741129767233e-05,
      "loss": 0.1916,
      "num_input_tokens_seen": 19675568,
      "step": 22720
    },
    {
      "epoch": 10.714285714285714,
      "grad_norm": 1.292484164237976,
      "learning_rate": 1.9692147067734202e-05,
      "loss": 0.2003,
      "num_input_tokens_seen": 19680336,
      "step": 22725
    },
    {
      "epoch": 10.716643092880716,
      "grad_norm": 0.9909115433692932,
      "learning_rate": 1.96825538242388e-05,
      "loss": 0.175,
      "num_input_tokens_seen": 19683920,
      "step": 22730
    },
    {
      "epoch": 10.719000471475718,
      "grad_norm": 0.48690298199653625,
      "learning_rate": 1.967296140076041e-05,
      "loss": 0.1497,
      "num_input_tokens_seen": 19689360,
      "step": 22735
    },
    {
      "epoch": 10.72135785007072,
      "grad_norm": 0.7666659355163574,
      "learning_rate": 1.966336979877833e-05,
      "loss": 0.2112,
      "num_input_tokens_seen": 19693200,
      "step": 22740
    },
    {
      "epoch": 10.723715228665723,
      "grad_norm": 1.4671519994735718,
      "learning_rate": 1.9653779019771678e-05,
      "loss": 0.1355,
      "num_input_tokens_seen": 19697216,
      "step": 22745
    },
    {
      "epoch": 10.726072607260726,
      "grad_norm": 0.708812415599823,
      "learning_rate": 1.9644189065219488e-05,
      "loss": 0.1567,
      "num_input_tokens_seen": 19701952,
      "step": 22750
    },
    {
      "epoch": 10.728429985855728,
      "grad_norm": 2.143373727798462,
      "learning_rate": 1.9634599936600655e-05,
      "loss": 0.2247,
      "num_input_tokens_seen": 19706176,
      "step": 22755
    },
    {
      "epoch": 10.73078736445073,
      "grad_norm": 0.5122739672660828,
      "learning_rate": 1.9625011635393935e-05,
      "loss": 0.1684,
      "num_input_tokens_seen": 19710400,
      "step": 22760
    },
    {
      "epoch": 10.733144743045733,
      "grad_norm": 0.8969156742095947,
      "learning_rate": 1.9615424163077963e-05,
      "loss": 0.1563,
      "num_input_tokens_seen": 19714832,
      "step": 22765
    },
    {
      "epoch": 10.735502121640735,
      "grad_norm": 0.5187214612960815,
      "learning_rate": 1.9605837521131263e-05,
      "loss": 0.1882,
      "num_input_tokens_seen": 19718592,
      "step": 22770
    },
    {
      "epoch": 10.737859500235738,
      "grad_norm": 0.6569729447364807,
      "learning_rate": 1.9596251711032192e-05,
      "loss": 0.1498,
      "num_input_tokens_seen": 19722464,
      "step": 22775
    },
    {
      "epoch": 10.74021687883074,
      "grad_norm": 1.7494621276855469,
      "learning_rate": 1.958666673425903e-05,
      "loss": 0.1945,
      "num_input_tokens_seen": 19727200,
      "step": 22780
    },
    {
      "epoch": 10.742574257425742,
      "grad_norm": 1.1989972591400146,
      "learning_rate": 1.957708259228987e-05,
      "loss": 0.1901,
      "num_input_tokens_seen": 19732000,
      "step": 22785
    },
    {
      "epoch": 10.744931636020745,
      "grad_norm": 0.6658492088317871,
      "learning_rate": 1.956749928660273e-05,
      "loss": 0.1882,
      "num_input_tokens_seen": 19735808,
      "step": 22790
    },
    {
      "epoch": 10.747289014615747,
      "grad_norm": 1.8149869441986084,
      "learning_rate": 1.955791681867547e-05,
      "loss": 0.2375,
      "num_input_tokens_seen": 19740624,
      "step": 22795
    },
    {
      "epoch": 10.74964639321075,
      "grad_norm": 1.0303785800933838,
      "learning_rate": 1.9548335189985824e-05,
      "loss": 0.2155,
      "num_input_tokens_seen": 19744496,
      "step": 22800
    },
    {
      "epoch": 10.74964639321075,
      "eval_loss": 0.20229695737361908,
      "eval_runtime": 22.1964,
      "eval_samples_per_second": 42.484,
      "eval_steps_per_second": 21.265,
      "num_input_tokens_seen": 19744496,
      "step": 22800
    },
    {
      "epoch": 10.752003771805752,
      "grad_norm": 1.1561673879623413,
      "learning_rate": 1.9538754402011396e-05,
      "loss": 0.211,
      "num_input_tokens_seen": 19748960,
      "step": 22805
    },
    {
      "epoch": 10.754361150400754,
      "grad_norm": 0.9164857864379883,
      "learning_rate": 1.952917445622968e-05,
      "loss": 0.1771,
      "num_input_tokens_seen": 19752928,
      "step": 22810
    },
    {
      "epoch": 10.756718528995757,
      "grad_norm": 0.5369678139686584,
      "learning_rate": 1.9519595354118005e-05,
      "loss": 0.1873,
      "num_input_tokens_seen": 19756784,
      "step": 22815
    },
    {
      "epoch": 10.75907590759076,
      "grad_norm": 0.5459575653076172,
      "learning_rate": 1.951001709715361e-05,
      "loss": 0.1574,
      "num_input_tokens_seen": 19761248,
      "step": 22820
    },
    {
      "epoch": 10.761433286185762,
      "grad_norm": 1.0002570152282715,
      "learning_rate": 1.9500439686813556e-05,
      "loss": 0.1817,
      "num_input_tokens_seen": 19764736,
      "step": 22825
    },
    {
      "epoch": 10.763790664780764,
      "grad_norm": 0.8621525168418884,
      "learning_rate": 1.949086312457482e-05,
      "loss": 0.143,
      "num_input_tokens_seen": 19768704,
      "step": 22830
    },
    {
      "epoch": 10.766148043375766,
      "grad_norm": 1.1940280199050903,
      "learning_rate": 1.9481287411914223e-05,
      "loss": 0.1514,
      "num_input_tokens_seen": 19773488,
      "step": 22835
    },
    {
      "epoch": 10.768505421970769,
      "grad_norm": 0.7995002865791321,
      "learning_rate": 1.9471712550308457e-05,
      "loss": 0.1469,
      "num_input_tokens_seen": 19777360,
      "step": 22840
    },
    {
      "epoch": 10.770862800565771,
      "grad_norm": 1.2105700969696045,
      "learning_rate": 1.946213854123409e-05,
      "loss": 0.1191,
      "num_input_tokens_seen": 19781904,
      "step": 22845
    },
    {
      "epoch": 10.773220179160774,
      "grad_norm": 0.6164396405220032,
      "learning_rate": 1.9452565386167554e-05,
      "loss": 0.1975,
      "num_input_tokens_seen": 19785472,
      "step": 22850
    },
    {
      "epoch": 10.775577557755776,
      "grad_norm": 0.9290472865104675,
      "learning_rate": 1.9442993086585142e-05,
      "loss": 0.1099,
      "num_input_tokens_seen": 19790160,
      "step": 22855
    },
    {
      "epoch": 10.777934936350778,
      "grad_norm": 1.2259601354599,
      "learning_rate": 1.9433421643963043e-05,
      "loss": 0.198,
      "num_input_tokens_seen": 19794448,
      "step": 22860
    },
    {
      "epoch": 10.78029231494578,
      "grad_norm": 0.6725147366523743,
      "learning_rate": 1.942385105977727e-05,
      "loss": 0.2185,
      "num_input_tokens_seen": 19798560,
      "step": 22865
    },
    {
      "epoch": 10.782649693540783,
      "grad_norm": 1.750649333000183,
      "learning_rate": 1.9414281335503743e-05,
      "loss": 0.2172,
      "num_input_tokens_seen": 19803056,
      "step": 22870
    },
    {
      "epoch": 10.785007072135786,
      "grad_norm": 1.1837334632873535,
      "learning_rate": 1.9404712472618232e-05,
      "loss": 0.1847,
      "num_input_tokens_seen": 19807056,
      "step": 22875
    },
    {
      "epoch": 10.787364450730788,
      "grad_norm": 0.578210711479187,
      "learning_rate": 1.939514447259636e-05,
      "loss": 0.1717,
      "num_input_tokens_seen": 19811808,
      "step": 22880
    },
    {
      "epoch": 10.78972182932579,
      "grad_norm": 0.3969174027442932,
      "learning_rate": 1.938557733691365e-05,
      "loss": 0.1445,
      "num_input_tokens_seen": 19816032,
      "step": 22885
    },
    {
      "epoch": 10.792079207920793,
      "grad_norm": 1.1107057332992554,
      "learning_rate": 1.9376011067045476e-05,
      "loss": 0.1919,
      "num_input_tokens_seen": 19820112,
      "step": 22890
    },
    {
      "epoch": 10.794436586515795,
      "grad_norm": 0.9708033204078674,
      "learning_rate": 1.9366445664467065e-05,
      "loss": 0.2219,
      "num_input_tokens_seen": 19824656,
      "step": 22895
    },
    {
      "epoch": 10.796793965110798,
      "grad_norm": 0.7467743754386902,
      "learning_rate": 1.9356881130653533e-05,
      "loss": 0.1988,
      "num_input_tokens_seen": 19828832,
      "step": 22900
    },
    {
      "epoch": 10.799151343705798,
      "grad_norm": 0.36505869030952454,
      "learning_rate": 1.9347317467079846e-05,
      "loss": 0.1585,
      "num_input_tokens_seen": 19833072,
      "step": 22905
    },
    {
      "epoch": 10.801508722300802,
      "grad_norm": 0.5260916352272034,
      "learning_rate": 1.9337754675220836e-05,
      "loss": 0.1906,
      "num_input_tokens_seen": 19836976,
      "step": 22910
    },
    {
      "epoch": 10.803866100895803,
      "grad_norm": 0.9214611053466797,
      "learning_rate": 1.9328192756551218e-05,
      "loss": 0.1507,
      "num_input_tokens_seen": 19841248,
      "step": 22915
    },
    {
      "epoch": 10.806223479490805,
      "grad_norm": 1.0148183107376099,
      "learning_rate": 1.931863171254555e-05,
      "loss": 0.1144,
      "num_input_tokens_seen": 19845728,
      "step": 22920
    },
    {
      "epoch": 10.808580858085808,
      "grad_norm": 0.964490532875061,
      "learning_rate": 1.930907154467826e-05,
      "loss": 0.1949,
      "num_input_tokens_seen": 19850640,
      "step": 22925
    },
    {
      "epoch": 10.81093823668081,
      "grad_norm": 1.3165743350982666,
      "learning_rate": 1.9299512254423673e-05,
      "loss": 0.2031,
      "num_input_tokens_seen": 19857152,
      "step": 22930
    },
    {
      "epoch": 10.813295615275813,
      "grad_norm": 0.6486037969589233,
      "learning_rate": 1.9289953843255914e-05,
      "loss": 0.1489,
      "num_input_tokens_seen": 19860704,
      "step": 22935
    },
    {
      "epoch": 10.815652993870815,
      "grad_norm": 0.6107425093650818,
      "learning_rate": 1.9280396312649048e-05,
      "loss": 0.1225,
      "num_input_tokens_seen": 19865360,
      "step": 22940
    },
    {
      "epoch": 10.818010372465817,
      "grad_norm": 0.8191775679588318,
      "learning_rate": 1.9270839664076936e-05,
      "loss": 0.1996,
      "num_input_tokens_seen": 19870048,
      "step": 22945
    },
    {
      "epoch": 10.82036775106082,
      "grad_norm": 0.3175641894340515,
      "learning_rate": 1.9261283899013345e-05,
      "loss": 0.121,
      "num_input_tokens_seen": 19874336,
      "step": 22950
    },
    {
      "epoch": 10.822725129655822,
      "grad_norm": 0.6449944972991943,
      "learning_rate": 1.92517290189319e-05,
      "loss": 0.1836,
      "num_input_tokens_seen": 19877984,
      "step": 22955
    },
    {
      "epoch": 10.825082508250825,
      "grad_norm": 2.8019964694976807,
      "learning_rate": 1.924217502530607e-05,
      "loss": 0.2069,
      "num_input_tokens_seen": 19881744,
      "step": 22960
    },
    {
      "epoch": 10.827439886845827,
      "grad_norm": 0.9423580169677734,
      "learning_rate": 1.9232621919609207e-05,
      "loss": 0.2124,
      "num_input_tokens_seen": 19885920,
      "step": 22965
    },
    {
      "epoch": 10.82979726544083,
      "grad_norm": 0.7991328239440918,
      "learning_rate": 1.9223069703314534e-05,
      "loss": 0.1422,
      "num_input_tokens_seen": 19889456,
      "step": 22970
    },
    {
      "epoch": 10.832154644035832,
      "grad_norm": 0.7629505395889282,
      "learning_rate": 1.92135183778951e-05,
      "loss": 0.13,
      "num_input_tokens_seen": 19893696,
      "step": 22975
    },
    {
      "epoch": 10.834512022630834,
      "grad_norm": 1.0796560049057007,
      "learning_rate": 1.9203967944823857e-05,
      "loss": 0.2027,
      "num_input_tokens_seen": 19898064,
      "step": 22980
    },
    {
      "epoch": 10.836869401225837,
      "grad_norm": 0.6407715082168579,
      "learning_rate": 1.9194418405573588e-05,
      "loss": 0.1619,
      "num_input_tokens_seen": 19902480,
      "step": 22985
    },
    {
      "epoch": 10.839226779820839,
      "grad_norm": 0.9090464115142822,
      "learning_rate": 1.9184869761616954e-05,
      "loss": 0.1162,
      "num_input_tokens_seen": 19907008,
      "step": 22990
    },
    {
      "epoch": 10.841584158415841,
      "grad_norm": 1.2613552808761597,
      "learning_rate": 1.9175322014426495e-05,
      "loss": 0.2522,
      "num_input_tokens_seen": 19911696,
      "step": 22995
    },
    {
      "epoch": 10.843941537010844,
      "grad_norm": 0.873407781124115,
      "learning_rate": 1.9165775165474565e-05,
      "loss": 0.1557,
      "num_input_tokens_seen": 19915984,
      "step": 23000
    },
    {
      "epoch": 10.843941537010844,
      "eval_loss": 0.2035198211669922,
      "eval_runtime": 22.1623,
      "eval_samples_per_second": 42.55,
      "eval_steps_per_second": 21.297,
      "num_input_tokens_seen": 19915984,
      "step": 23000
    },
    {
      "epoch": 10.846298915605846,
      "grad_norm": 1.2523187398910522,
      "learning_rate": 1.9156229216233434e-05,
      "loss": 0.2502,
      "num_input_tokens_seen": 19920032,
      "step": 23005
    },
    {
      "epoch": 10.848656294200849,
      "grad_norm": 2.0151004791259766,
      "learning_rate": 1.9146684168175184e-05,
      "loss": 0.1855,
      "num_input_tokens_seen": 19925024,
      "step": 23010
    },
    {
      "epoch": 10.851013672795851,
      "grad_norm": 1.4835516214370728,
      "learning_rate": 1.9137140022771796e-05,
      "loss": 0.2512,
      "num_input_tokens_seen": 19929424,
      "step": 23015
    },
    {
      "epoch": 10.853371051390853,
      "grad_norm": 0.5072636604309082,
      "learning_rate": 1.9127596781495103e-05,
      "loss": 0.1537,
      "num_input_tokens_seen": 19934048,
      "step": 23020
    },
    {
      "epoch": 10.855728429985856,
      "grad_norm": 0.8425182104110718,
      "learning_rate": 1.9118054445816767e-05,
      "loss": 0.2175,
      "num_input_tokens_seen": 19937968,
      "step": 23025
    },
    {
      "epoch": 10.858085808580858,
      "grad_norm": 1.6567858457565308,
      "learning_rate": 1.9108513017208356e-05,
      "loss": 0.2317,
      "num_input_tokens_seen": 19942416,
      "step": 23030
    },
    {
      "epoch": 10.86044318717586,
      "grad_norm": 1.3479156494140625,
      "learning_rate": 1.9098972497141287e-05,
      "loss": 0.1776,
      "num_input_tokens_seen": 19946688,
      "step": 23035
    },
    {
      "epoch": 10.862800565770863,
      "grad_norm": 0.7532241344451904,
      "learning_rate": 1.9089432887086806e-05,
      "loss": 0.1853,
      "num_input_tokens_seen": 19950592,
      "step": 23040
    },
    {
      "epoch": 10.865157944365865,
      "grad_norm": 1.2807254791259766,
      "learning_rate": 1.9079894188516056e-05,
      "loss": 0.2082,
      "num_input_tokens_seen": 19955296,
      "step": 23045
    },
    {
      "epoch": 10.867515322960868,
      "grad_norm": 0.6504724621772766,
      "learning_rate": 1.907035640290002e-05,
      "loss": 0.1924,
      "num_input_tokens_seen": 19959328,
      "step": 23050
    },
    {
      "epoch": 10.86987270155587,
      "grad_norm": 0.9379006624221802,
      "learning_rate": 1.9060819531709534e-05,
      "loss": 0.1815,
      "num_input_tokens_seen": 19963296,
      "step": 23055
    },
    {
      "epoch": 10.872230080150873,
      "grad_norm": 1.038901925086975,
      "learning_rate": 1.9051283576415325e-05,
      "loss": 0.2099,
      "num_input_tokens_seen": 19967520,
      "step": 23060
    },
    {
      "epoch": 10.874587458745875,
      "grad_norm": 0.6258488893508911,
      "learning_rate": 1.904174853848793e-05,
      "loss": 0.1753,
      "num_input_tokens_seen": 19972416,
      "step": 23065
    },
    {
      "epoch": 10.876944837340877,
      "grad_norm": 0.7510301470756531,
      "learning_rate": 1.903221441939779e-05,
      "loss": 0.1883,
      "num_input_tokens_seen": 19977296,
      "step": 23070
    },
    {
      "epoch": 10.87930221593588,
      "grad_norm": 1.1560548543930054,
      "learning_rate": 1.9022681220615194e-05,
      "loss": 0.1481,
      "num_input_tokens_seen": 19981968,
      "step": 23075
    },
    {
      "epoch": 10.881659594530882,
      "grad_norm": 1.193294882774353,
      "learning_rate": 1.9013148943610255e-05,
      "loss": 0.2076,
      "num_input_tokens_seen": 19986864,
      "step": 23080
    },
    {
      "epoch": 10.884016973125885,
      "grad_norm": 0.75892174243927,
      "learning_rate": 1.9003617589852998e-05,
      "loss": 0.1883,
      "num_input_tokens_seen": 19990992,
      "step": 23085
    },
    {
      "epoch": 10.886374351720887,
      "grad_norm": 0.710504412651062,
      "learning_rate": 1.899408716081326e-05,
      "loss": 0.1502,
      "num_input_tokens_seen": 19995280,
      "step": 23090
    },
    {
      "epoch": 10.88873173031589,
      "grad_norm": 0.8990337252616882,
      "learning_rate": 1.898455765796075e-05,
      "loss": 0.2043,
      "num_input_tokens_seen": 19999376,
      "step": 23095
    },
    {
      "epoch": 10.891089108910892,
      "grad_norm": 0.5493698716163635,
      "learning_rate": 1.8975029082765053e-05,
      "loss": 0.1902,
      "num_input_tokens_seen": 20003072,
      "step": 23100
    },
    {
      "epoch": 10.893446487505894,
      "grad_norm": 0.4692363142967224,
      "learning_rate": 1.8965501436695577e-05,
      "loss": 0.1842,
      "num_input_tokens_seen": 20007936,
      "step": 23105
    },
    {
      "epoch": 10.895803866100895,
      "grad_norm": 1.1662003993988037,
      "learning_rate": 1.895597472122161e-05,
      "loss": 0.2193,
      "num_input_tokens_seen": 20012464,
      "step": 23110
    },
    {
      "epoch": 10.898161244695899,
      "grad_norm": 2.0201079845428467,
      "learning_rate": 1.894644893781231e-05,
      "loss": 0.292,
      "num_input_tokens_seen": 20016064,
      "step": 23115
    },
    {
      "epoch": 10.9005186232909,
      "grad_norm": 0.9879921078681946,
      "learning_rate": 1.893692408793665e-05,
      "loss": 0.1601,
      "num_input_tokens_seen": 20020288,
      "step": 23120
    },
    {
      "epoch": 10.902876001885902,
      "grad_norm": 1.8056976795196533,
      "learning_rate": 1.8927400173063493e-05,
      "loss": 0.2042,
      "num_input_tokens_seen": 20024512,
      "step": 23125
    },
    {
      "epoch": 10.905233380480905,
      "grad_norm": 0.5429677963256836,
      "learning_rate": 1.891787719466154e-05,
      "loss": 0.1846,
      "num_input_tokens_seen": 20028768,
      "step": 23130
    },
    {
      "epoch": 10.907590759075907,
      "grad_norm": 0.5849343538284302,
      "learning_rate": 1.8908355154199346e-05,
      "loss": 0.1484,
      "num_input_tokens_seen": 20033248,
      "step": 23135
    },
    {
      "epoch": 10.90994813767091,
      "grad_norm": 1.1636022329330444,
      "learning_rate": 1.8898834053145357e-05,
      "loss": 0.2027,
      "num_input_tokens_seen": 20037920,
      "step": 23140
    },
    {
      "epoch": 10.912305516265912,
      "grad_norm": 0.8616999387741089,
      "learning_rate": 1.8889313892967813e-05,
      "loss": 0.1632,
      "num_input_tokens_seen": 20044048,
      "step": 23145
    },
    {
      "epoch": 10.914662894860914,
      "grad_norm": 0.7840429544448853,
      "learning_rate": 1.8879794675134863e-05,
      "loss": 0.1929,
      "num_input_tokens_seen": 20048096,
      "step": 23150
    },
    {
      "epoch": 10.917020273455917,
      "grad_norm": 0.7121015787124634,
      "learning_rate": 1.8870276401114494e-05,
      "loss": 0.1736,
      "num_input_tokens_seen": 20052576,
      "step": 23155
    },
    {
      "epoch": 10.919377652050919,
      "grad_norm": 0.6205703616142273,
      "learning_rate": 1.886075907237453e-05,
      "loss": 0.1444,
      "num_input_tokens_seen": 20056496,
      "step": 23160
    },
    {
      "epoch": 10.921735030645921,
      "grad_norm": 1.5607036352157593,
      "learning_rate": 1.8851242690382672e-05,
      "loss": 0.2933,
      "num_input_tokens_seen": 20060992,
      "step": 23165
    },
    {
      "epoch": 10.924092409240924,
      "grad_norm": 1.0331438779830933,
      "learning_rate": 1.884172725660645e-05,
      "loss": 0.24,
      "num_input_tokens_seen": 20065104,
      "step": 23170
    },
    {
      "epoch": 10.926449787835926,
      "grad_norm": 1.3392308950424194,
      "learning_rate": 1.8832212772513277e-05,
      "loss": 0.1233,
      "num_input_tokens_seen": 20069200,
      "step": 23175
    },
    {
      "epoch": 10.928807166430929,
      "grad_norm": 1.2608448266983032,
      "learning_rate": 1.8822699239570414e-05,
      "loss": 0.2219,
      "num_input_tokens_seen": 20073232,
      "step": 23180
    },
    {
      "epoch": 10.931164545025931,
      "grad_norm": 0.6568114161491394,
      "learning_rate": 1.8813186659244943e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 20077392,
      "step": 23185
    },
    {
      "epoch": 10.933521923620933,
      "grad_norm": 1.3157501220703125,
      "learning_rate": 1.880367503300385e-05,
      "loss": 0.1853,
      "num_input_tokens_seen": 20082448,
      "step": 23190
    },
    {
      "epoch": 10.935879302215936,
      "grad_norm": 1.278550624847412,
      "learning_rate": 1.8794164362313927e-05,
      "loss": 0.1851,
      "num_input_tokens_seen": 20086544,
      "step": 23195
    },
    {
      "epoch": 10.938236680810938,
      "grad_norm": 0.777962327003479,
      "learning_rate": 1.878465464864185e-05,
      "loss": 0.2032,
      "num_input_tokens_seen": 20090944,
      "step": 23200
    },
    {
      "epoch": 10.938236680810938,
      "eval_loss": 0.2003687173128128,
      "eval_runtime": 22.1468,
      "eval_samples_per_second": 42.579,
      "eval_steps_per_second": 21.312,
      "num_input_tokens_seen": 20090944,
      "step": 23200
    },
    {
      "epoch": 10.94059405940594,
      "grad_norm": 0.5318717956542969,
      "learning_rate": 1.877514589345414e-05,
      "loss": 0.1566,
      "num_input_tokens_seen": 20095760,
      "step": 23205
    },
    {
      "epoch": 10.942951438000943,
      "grad_norm": 0.5314643979072571,
      "learning_rate": 1.876563809821715e-05,
      "loss": 0.1713,
      "num_input_tokens_seen": 20100384,
      "step": 23210
    },
    {
      "epoch": 10.945308816595945,
      "grad_norm": 1.670151710510254,
      "learning_rate": 1.8756131264397106e-05,
      "loss": 0.2148,
      "num_input_tokens_seen": 20104880,
      "step": 23215
    },
    {
      "epoch": 10.947666195190948,
      "grad_norm": 1.1141077280044556,
      "learning_rate": 1.87466253934601e-05,
      "loss": 0.163,
      "num_input_tokens_seen": 20109648,
      "step": 23220
    },
    {
      "epoch": 10.95002357378595,
      "grad_norm": 0.4778267443180084,
      "learning_rate": 1.8737120486872033e-05,
      "loss": 0.2336,
      "num_input_tokens_seen": 20113456,
      "step": 23225
    },
    {
      "epoch": 10.952380952380953,
      "grad_norm": 0.4812975525856018,
      "learning_rate": 1.8727616546098696e-05,
      "loss": 0.2148,
      "num_input_tokens_seen": 20118160,
      "step": 23230
    },
    {
      "epoch": 10.954738330975955,
      "grad_norm": 1.4809023141860962,
      "learning_rate": 1.8718113572605716e-05,
      "loss": 0.2145,
      "num_input_tokens_seen": 20122384,
      "step": 23235
    },
    {
      "epoch": 10.957095709570957,
      "grad_norm": 1.2756576538085938,
      "learning_rate": 1.8708611567858554e-05,
      "loss": 0.2153,
      "num_input_tokens_seen": 20125760,
      "step": 23240
    },
    {
      "epoch": 10.95945308816596,
      "grad_norm": 0.5867846012115479,
      "learning_rate": 1.8699110533322565e-05,
      "loss": 0.182,
      "num_input_tokens_seen": 20130048,
      "step": 23245
    },
    {
      "epoch": 10.961810466760962,
      "grad_norm": 1.1801459789276123,
      "learning_rate": 1.8689610470462897e-05,
      "loss": 0.1842,
      "num_input_tokens_seen": 20134480,
      "step": 23250
    },
    {
      "epoch": 10.964167845355965,
      "grad_norm": 0.9843763709068298,
      "learning_rate": 1.8680111380744604e-05,
      "loss": 0.1486,
      "num_input_tokens_seen": 20139264,
      "step": 23255
    },
    {
      "epoch": 10.966525223950967,
      "grad_norm": 1.1478140354156494,
      "learning_rate": 1.8670613265632564e-05,
      "loss": 0.1783,
      "num_input_tokens_seen": 20143312,
      "step": 23260
    },
    {
      "epoch": 10.96888260254597,
      "grad_norm": 0.6410378217697144,
      "learning_rate": 1.866111612659149e-05,
      "loss": 0.142,
      "num_input_tokens_seen": 20147840,
      "step": 23265
    },
    {
      "epoch": 10.971239981140972,
      "grad_norm": 0.6143513321876526,
      "learning_rate": 1.8651619965085967e-05,
      "loss": 0.162,
      "num_input_tokens_seen": 20151856,
      "step": 23270
    },
    {
      "epoch": 10.973597359735974,
      "grad_norm": 0.6124769449234009,
      "learning_rate": 1.8642124782580433e-05,
      "loss": 0.2563,
      "num_input_tokens_seen": 20155424,
      "step": 23275
    },
    {
      "epoch": 10.975954738330977,
      "grad_norm": 0.37505173683166504,
      "learning_rate": 1.8632630580539144e-05,
      "loss": 0.1559,
      "num_input_tokens_seen": 20159472,
      "step": 23280
    },
    {
      "epoch": 10.978312116925979,
      "grad_norm": 0.7280723452568054,
      "learning_rate": 1.862313736042625e-05,
      "loss": 0.1445,
      "num_input_tokens_seen": 20163488,
      "step": 23285
    },
    {
      "epoch": 10.980669495520981,
      "grad_norm": 0.9270986318588257,
      "learning_rate": 1.8613645123705703e-05,
      "loss": 0.1646,
      "num_input_tokens_seen": 20168320,
      "step": 23290
    },
    {
      "epoch": 10.983026874115984,
      "grad_norm": 0.8147082924842834,
      "learning_rate": 1.8604153871841328e-05,
      "loss": 0.1373,
      "num_input_tokens_seen": 20172992,
      "step": 23295
    },
    {
      "epoch": 10.985384252710986,
      "grad_norm": 2.0444483757019043,
      "learning_rate": 1.859466360629682e-05,
      "loss": 0.1731,
      "num_input_tokens_seen": 20177136,
      "step": 23300
    },
    {
      "epoch": 10.987741631305987,
      "grad_norm": 1.0688029527664185,
      "learning_rate": 1.8585174328535666e-05,
      "loss": 0.1675,
      "num_input_tokens_seen": 20181984,
      "step": 23305
    },
    {
      "epoch": 10.990099009900991,
      "grad_norm": 1.619010329246521,
      "learning_rate": 1.857568604002124e-05,
      "loss": 0.1699,
      "num_input_tokens_seen": 20185840,
      "step": 23310
    },
    {
      "epoch": 10.992456388495992,
      "grad_norm": 0.9562480449676514,
      "learning_rate": 1.8566198742216774e-05,
      "loss": 0.1255,
      "num_input_tokens_seen": 20191200,
      "step": 23315
    },
    {
      "epoch": 10.994813767090994,
      "grad_norm": 0.9740175604820251,
      "learning_rate": 1.85567124365853e-05,
      "loss": 0.1711,
      "num_input_tokens_seen": 20195264,
      "step": 23320
    },
    {
      "epoch": 10.997171145685996,
      "grad_norm": 1.249362826347351,
      "learning_rate": 1.854722712458975e-05,
      "loss": 0.1597,
      "num_input_tokens_seen": 20199808,
      "step": 23325
    },
    {
      "epoch": 10.999528524280999,
      "grad_norm": 1.2894744873046875,
      "learning_rate": 1.853774280769286e-05,
      "loss": 0.1806,
      "num_input_tokens_seen": 20203920,
      "step": 23330
    },
    {
      "epoch": 11.001885902876001,
      "grad_norm": 0.41901692748069763,
      "learning_rate": 1.852825948735724e-05,
      "loss": 0.1567,
      "num_input_tokens_seen": 20207728,
      "step": 23335
    },
    {
      "epoch": 11.004243281471004,
      "grad_norm": 0.9626701474189758,
      "learning_rate": 1.851877716504534e-05,
      "loss": 0.1833,
      "num_input_tokens_seen": 20212592,
      "step": 23340
    },
    {
      "epoch": 11.006600660066006,
      "grad_norm": 1.4874767065048218,
      "learning_rate": 1.8509295842219448e-05,
      "loss": 0.1773,
      "num_input_tokens_seen": 20217056,
      "step": 23345
    },
    {
      "epoch": 11.008958038661008,
      "grad_norm": 0.8862606883049011,
      "learning_rate": 1.8499815520341697e-05,
      "loss": 0.1331,
      "num_input_tokens_seen": 20220848,
      "step": 23350
    },
    {
      "epoch": 11.01131541725601,
      "grad_norm": 1.0894031524658203,
      "learning_rate": 1.8490336200874094e-05,
      "loss": 0.2124,
      "num_input_tokens_seen": 20225360,
      "step": 23355
    },
    {
      "epoch": 11.013672795851013,
      "grad_norm": 1.9435157775878906,
      "learning_rate": 1.848085788527844e-05,
      "loss": 0.2138,
      "num_input_tokens_seen": 20230480,
      "step": 23360
    },
    {
      "epoch": 11.016030174446016,
      "grad_norm": 1.3749189376831055,
      "learning_rate": 1.847138057501644e-05,
      "loss": 0.1952,
      "num_input_tokens_seen": 20234112,
      "step": 23365
    },
    {
      "epoch": 11.018387553041018,
      "grad_norm": 0.37779319286346436,
      "learning_rate": 1.8461904271549582e-05,
      "loss": 0.2457,
      "num_input_tokens_seen": 20238320,
      "step": 23370
    },
    {
      "epoch": 11.02074493163602,
      "grad_norm": 0.7895139455795288,
      "learning_rate": 1.845242897633926e-05,
      "loss": 0.1002,
      "num_input_tokens_seen": 20242592,
      "step": 23375
    },
    {
      "epoch": 11.023102310231023,
      "grad_norm": 0.6950656771659851,
      "learning_rate": 1.844295469084667e-05,
      "loss": 0.1599,
      "num_input_tokens_seen": 20246912,
      "step": 23380
    },
    {
      "epoch": 11.025459688826025,
      "grad_norm": 1.2453829050064087,
      "learning_rate": 1.843348141653286e-05,
      "loss": 0.2584,
      "num_input_tokens_seen": 20251216,
      "step": 23385
    },
    {
      "epoch": 11.027817067421028,
      "grad_norm": 0.7509268522262573,
      "learning_rate": 1.842400915485874e-05,
      "loss": 0.1881,
      "num_input_tokens_seen": 20256000,
      "step": 23390
    },
    {
      "epoch": 11.03017444601603,
      "grad_norm": 1.9787847995758057,
      "learning_rate": 1.8414537907285053e-05,
      "loss": 0.1935,
      "num_input_tokens_seen": 20259760,
      "step": 23395
    },
    {
      "epoch": 11.032531824611032,
      "grad_norm": 1.043272852897644,
      "learning_rate": 1.840506767527237e-05,
      "loss": 0.1394,
      "num_input_tokens_seen": 20264992,
      "step": 23400
    },
    {
      "epoch": 11.032531824611032,
      "eval_loss": 0.20058400928974152,
      "eval_runtime": 22.153,
      "eval_samples_per_second": 42.568,
      "eval_steps_per_second": 21.306,
      "num_input_tokens_seen": 20264992,
      "step": 23400
    },
    {
      "epoch": 11.034889203206035,
      "grad_norm": 1.0992294549942017,
      "learning_rate": 1.8395598460281137e-05,
      "loss": 0.2171,
      "num_input_tokens_seen": 20268720,
      "step": 23405
    },
    {
      "epoch": 11.037246581801037,
      "grad_norm": 1.0712177753448486,
      "learning_rate": 1.838613026377161e-05,
      "loss": 0.2176,
      "num_input_tokens_seen": 20273072,
      "step": 23410
    },
    {
      "epoch": 11.03960396039604,
      "grad_norm": 0.9319260120391846,
      "learning_rate": 1.8376663087203917e-05,
      "loss": 0.1389,
      "num_input_tokens_seen": 20276864,
      "step": 23415
    },
    {
      "epoch": 11.041961338991042,
      "grad_norm": 0.9032689929008484,
      "learning_rate": 1.8367196932038014e-05,
      "loss": 0.1752,
      "num_input_tokens_seen": 20281168,
      "step": 23420
    },
    {
      "epoch": 11.044318717586044,
      "grad_norm": 1.0435967445373535,
      "learning_rate": 1.8357731799733686e-05,
      "loss": 0.1549,
      "num_input_tokens_seen": 20285664,
      "step": 23425
    },
    {
      "epoch": 11.046676096181047,
      "grad_norm": 0.9122189879417419,
      "learning_rate": 1.8348267691750586e-05,
      "loss": 0.1789,
      "num_input_tokens_seen": 20289680,
      "step": 23430
    },
    {
      "epoch": 11.04903347477605,
      "grad_norm": 1.205739140510559,
      "learning_rate": 1.833880460954821e-05,
      "loss": 0.2048,
      "num_input_tokens_seen": 20293824,
      "step": 23435
    },
    {
      "epoch": 11.051390853371052,
      "grad_norm": 0.7988952994346619,
      "learning_rate": 1.8329342554585866e-05,
      "loss": 0.1488,
      "num_input_tokens_seen": 20298160,
      "step": 23440
    },
    {
      "epoch": 11.053748231966054,
      "grad_norm": 1.5231536626815796,
      "learning_rate": 1.8319881528322735e-05,
      "loss": 0.1634,
      "num_input_tokens_seen": 20302096,
      "step": 23445
    },
    {
      "epoch": 11.056105610561056,
      "grad_norm": 1.2973573207855225,
      "learning_rate": 1.8310421532217815e-05,
      "loss": 0.2094,
      "num_input_tokens_seen": 20306432,
      "step": 23450
    },
    {
      "epoch": 11.058462989156059,
      "grad_norm": 1.0154715776443481,
      "learning_rate": 1.8300962567729958e-05,
      "loss": 0.196,
      "num_input_tokens_seen": 20310736,
      "step": 23455
    },
    {
      "epoch": 11.060820367751061,
      "grad_norm": 0.9050886631011963,
      "learning_rate": 1.8291504636317866e-05,
      "loss": 0.1568,
      "num_input_tokens_seen": 20315904,
      "step": 23460
    },
    {
      "epoch": 11.063177746346064,
      "grad_norm": 0.5486920475959778,
      "learning_rate": 1.8282047739440055e-05,
      "loss": 0.1779,
      "num_input_tokens_seen": 20319680,
      "step": 23465
    },
    {
      "epoch": 11.065535124941066,
      "grad_norm": 1.572740912437439,
      "learning_rate": 1.8272591878554903e-05,
      "loss": 0.1621,
      "num_input_tokens_seen": 20323760,
      "step": 23470
    },
    {
      "epoch": 11.067892503536068,
      "grad_norm": 0.4525289833545685,
      "learning_rate": 1.8263137055120638e-05,
      "loss": 0.1476,
      "num_input_tokens_seen": 20328608,
      "step": 23475
    },
    {
      "epoch": 11.07024988213107,
      "grad_norm": 1.415173888206482,
      "learning_rate": 1.8253683270595295e-05,
      "loss": 0.2234,
      "num_input_tokens_seen": 20333280,
      "step": 23480
    },
    {
      "epoch": 11.072607260726073,
      "grad_norm": 0.9445692896842957,
      "learning_rate": 1.824423052643677e-05,
      "loss": 0.2219,
      "num_input_tokens_seen": 20337568,
      "step": 23485
    },
    {
      "epoch": 11.074964639321076,
      "grad_norm": 0.9063090682029724,
      "learning_rate": 1.82347788241028e-05,
      "loss": 0.1432,
      "num_input_tokens_seen": 20341600,
      "step": 23490
    },
    {
      "epoch": 11.077322017916078,
      "grad_norm": 0.9626246094703674,
      "learning_rate": 1.8225328165050942e-05,
      "loss": 0.1781,
      "num_input_tokens_seen": 20345696,
      "step": 23495
    },
    {
      "epoch": 11.07967939651108,
      "grad_norm": 0.9050880074501038,
      "learning_rate": 1.821587855073863e-05,
      "loss": 0.1875,
      "num_input_tokens_seen": 20350384,
      "step": 23500
    },
    {
      "epoch": 11.082036775106083,
      "grad_norm": 1.8278837203979492,
      "learning_rate": 1.8206429982623086e-05,
      "loss": 0.216,
      "num_input_tokens_seen": 20354224,
      "step": 23505
    },
    {
      "epoch": 11.084394153701085,
      "grad_norm": 1.217753291130066,
      "learning_rate": 1.8196982462161416e-05,
      "loss": 0.1682,
      "num_input_tokens_seen": 20358384,
      "step": 23510
    },
    {
      "epoch": 11.086751532296086,
      "grad_norm": 1.7207518815994263,
      "learning_rate": 1.818753599081055e-05,
      "loss": 0.1668,
      "num_input_tokens_seen": 20362944,
      "step": 23515
    },
    {
      "epoch": 11.089108910891088,
      "grad_norm": 1.0049203634262085,
      "learning_rate": 1.817809057002724e-05,
      "loss": 0.1252,
      "num_input_tokens_seen": 20367584,
      "step": 23520
    },
    {
      "epoch": 11.09146628948609,
      "grad_norm": 0.6265076994895935,
      "learning_rate": 1.8168646201268096e-05,
      "loss": 0.1644,
      "num_input_tokens_seen": 20371488,
      "step": 23525
    },
    {
      "epoch": 11.093823668081093,
      "grad_norm": 0.9472841620445251,
      "learning_rate": 1.8159202885989557e-05,
      "loss": 0.1838,
      "num_input_tokens_seen": 20375920,
      "step": 23530
    },
    {
      "epoch": 11.096181046676096,
      "grad_norm": 0.7813743352890015,
      "learning_rate": 1.814976062564789e-05,
      "loss": 0.1786,
      "num_input_tokens_seen": 20380320,
      "step": 23535
    },
    {
      "epoch": 11.098538425271098,
      "grad_norm": 0.4338204562664032,
      "learning_rate": 1.8140319421699234e-05,
      "loss": 0.1074,
      "num_input_tokens_seen": 20384400,
      "step": 23540
    },
    {
      "epoch": 11.1008958038661,
      "grad_norm": 0.6166663765907288,
      "learning_rate": 1.8130879275599515e-05,
      "loss": 0.1957,
      "num_input_tokens_seen": 20389120,
      "step": 23545
    },
    {
      "epoch": 11.103253182461103,
      "grad_norm": 1.1002477407455444,
      "learning_rate": 1.8121440188804544e-05,
      "loss": 0.1537,
      "num_input_tokens_seen": 20393408,
      "step": 23550
    },
    {
      "epoch": 11.105610561056105,
      "grad_norm": 0.8837639093399048,
      "learning_rate": 1.811200216276993e-05,
      "loss": 0.1588,
      "num_input_tokens_seen": 20398784,
      "step": 23555
    },
    {
      "epoch": 11.107967939651108,
      "grad_norm": 1.5598434209823608,
      "learning_rate": 1.810256519895115e-05,
      "loss": 0.2192,
      "num_input_tokens_seen": 20404048,
      "step": 23560
    },
    {
      "epoch": 11.11032531824611,
      "grad_norm": 2.1109495162963867,
      "learning_rate": 1.8093129298803494e-05,
      "loss": 0.1311,
      "num_input_tokens_seen": 20408448,
      "step": 23565
    },
    {
      "epoch": 11.112682696841112,
      "grad_norm": 1.7323195934295654,
      "learning_rate": 1.808369446378209e-05,
      "loss": 0.187,
      "num_input_tokens_seen": 20412864,
      "step": 23570
    },
    {
      "epoch": 11.115040075436115,
      "grad_norm": 0.6088112592697144,
      "learning_rate": 1.8074260695341914e-05,
      "loss": 0.2001,
      "num_input_tokens_seen": 20416800,
      "step": 23575
    },
    {
      "epoch": 11.117397454031117,
      "grad_norm": 1.2201943397521973,
      "learning_rate": 1.8064827994937782e-05,
      "loss": 0.2503,
      "num_input_tokens_seen": 20421296,
      "step": 23580
    },
    {
      "epoch": 11.11975483262612,
      "grad_norm": 1.453325629234314,
      "learning_rate": 1.8055396364024317e-05,
      "loss": 0.1759,
      "num_input_tokens_seen": 20424912,
      "step": 23585
    },
    {
      "epoch": 11.122112211221122,
      "grad_norm": 1.1563482284545898,
      "learning_rate": 1.804596580405601e-05,
      "loss": 0.2142,
      "num_input_tokens_seen": 20428992,
      "step": 23590
    },
    {
      "epoch": 11.124469589816124,
      "grad_norm": 0.9706671833992004,
      "learning_rate": 1.8036536316487174e-05,
      "loss": 0.1568,
      "num_input_tokens_seen": 20433168,
      "step": 23595
    },
    {
      "epoch": 11.126826968411127,
      "grad_norm": 1.1820544004440308,
      "learning_rate": 1.802710790277193e-05,
      "loss": 0.1736,
      "num_input_tokens_seen": 20437952,
      "step": 23600
    },
    {
      "epoch": 11.126826968411127,
      "eval_loss": 0.20097039639949799,
      "eval_runtime": 22.1632,
      "eval_samples_per_second": 42.548,
      "eval_steps_per_second": 21.297,
      "num_input_tokens_seen": 20437952,
      "step": 23600
    },
    {
      "epoch": 11.12918434700613,
      "grad_norm": 0.93578040599823,
      "learning_rate": 1.801768056436429e-05,
      "loss": 0.186,
      "num_input_tokens_seen": 20442672,
      "step": 23605
    },
    {
      "epoch": 11.131541725601132,
      "grad_norm": 0.5691214203834534,
      "learning_rate": 1.8008254302718035e-05,
      "loss": 0.139,
      "num_input_tokens_seen": 20447072,
      "step": 23610
    },
    {
      "epoch": 11.133899104196134,
      "grad_norm": 0.7213246822357178,
      "learning_rate": 1.7998829119286837e-05,
      "loss": 0.1673,
      "num_input_tokens_seen": 20451088,
      "step": 23615
    },
    {
      "epoch": 11.136256482791136,
      "grad_norm": 0.297884464263916,
      "learning_rate": 1.798940501552418e-05,
      "loss": 0.2062,
      "num_input_tokens_seen": 20455472,
      "step": 23620
    },
    {
      "epoch": 11.138613861386139,
      "grad_norm": 0.5216708779335022,
      "learning_rate": 1.797998199288336e-05,
      "loss": 0.129,
      "num_input_tokens_seen": 20459728,
      "step": 23625
    },
    {
      "epoch": 11.140971239981141,
      "grad_norm": 1.514301061630249,
      "learning_rate": 1.7970560052817543e-05,
      "loss": 0.1949,
      "num_input_tokens_seen": 20464112,
      "step": 23630
    },
    {
      "epoch": 11.143328618576144,
      "grad_norm": 2.2298293113708496,
      "learning_rate": 1.7961139196779702e-05,
      "loss": 0.2565,
      "num_input_tokens_seen": 20468112,
      "step": 23635
    },
    {
      "epoch": 11.145685997171146,
      "grad_norm": 0.6893033981323242,
      "learning_rate": 1.7951719426222647e-05,
      "loss": 0.1332,
      "num_input_tokens_seen": 20473696,
      "step": 23640
    },
    {
      "epoch": 11.148043375766148,
      "grad_norm": 1.1699436902999878,
      "learning_rate": 1.794230074259904e-05,
      "loss": 0.2089,
      "num_input_tokens_seen": 20477616,
      "step": 23645
    },
    {
      "epoch": 11.15040075436115,
      "grad_norm": 0.7008761167526245,
      "learning_rate": 1.7932883147361336e-05,
      "loss": 0.2197,
      "num_input_tokens_seen": 20481472,
      "step": 23650
    },
    {
      "epoch": 11.152758132956153,
      "grad_norm": 0.9161515235900879,
      "learning_rate": 1.7923466641961865e-05,
      "loss": 0.1672,
      "num_input_tokens_seen": 20485984,
      "step": 23655
    },
    {
      "epoch": 11.155115511551156,
      "grad_norm": 1.080470085144043,
      "learning_rate": 1.791405122785278e-05,
      "loss": 0.1619,
      "num_input_tokens_seen": 20490512,
      "step": 23660
    },
    {
      "epoch": 11.157472890146158,
      "grad_norm": 0.7579237222671509,
      "learning_rate": 1.7904636906486037e-05,
      "loss": 0.2514,
      "num_input_tokens_seen": 20494288,
      "step": 23665
    },
    {
      "epoch": 11.15983026874116,
      "grad_norm": 1.9611696004867554,
      "learning_rate": 1.7895223679313448e-05,
      "loss": 0.1932,
      "num_input_tokens_seen": 20498368,
      "step": 23670
    },
    {
      "epoch": 11.162187647336163,
      "grad_norm": 1.1137804985046387,
      "learning_rate": 1.7885811547786653e-05,
      "loss": 0.2026,
      "num_input_tokens_seen": 20502176,
      "step": 23675
    },
    {
      "epoch": 11.164545025931165,
      "grad_norm": 1.2361595630645752,
      "learning_rate": 1.7876400513357115e-05,
      "loss": 0.1294,
      "num_input_tokens_seen": 20506560,
      "step": 23680
    },
    {
      "epoch": 11.166902404526168,
      "grad_norm": 1.0506572723388672,
      "learning_rate": 1.7866990577476146e-05,
      "loss": 0.1496,
      "num_input_tokens_seen": 20510544,
      "step": 23685
    },
    {
      "epoch": 11.16925978312117,
      "grad_norm": 1.3978064060211182,
      "learning_rate": 1.7857581741594863e-05,
      "loss": 0.2102,
      "num_input_tokens_seen": 20515072,
      "step": 23690
    },
    {
      "epoch": 11.171617161716172,
      "grad_norm": 0.6387233138084412,
      "learning_rate": 1.7848174007164237e-05,
      "loss": 0.1546,
      "num_input_tokens_seen": 20519472,
      "step": 23695
    },
    {
      "epoch": 11.173974540311175,
      "grad_norm": 0.840644896030426,
      "learning_rate": 1.7838767375635052e-05,
      "loss": 0.1769,
      "num_input_tokens_seen": 20523744,
      "step": 23700
    },
    {
      "epoch": 11.176331918906177,
      "grad_norm": 1.5258506536483765,
      "learning_rate": 1.782936184845793e-05,
      "loss": 0.1384,
      "num_input_tokens_seen": 20527968,
      "step": 23705
    },
    {
      "epoch": 11.17868929750118,
      "grad_norm": 1.0172696113586426,
      "learning_rate": 1.7819957427083334e-05,
      "loss": 0.1043,
      "num_input_tokens_seen": 20531616,
      "step": 23710
    },
    {
      "epoch": 11.18104667609618,
      "grad_norm": 1.7568607330322266,
      "learning_rate": 1.7810554112961516e-05,
      "loss": 0.3167,
      "num_input_tokens_seen": 20535920,
      "step": 23715
    },
    {
      "epoch": 11.183404054691183,
      "grad_norm": 0.8125528693199158,
      "learning_rate": 1.7801151907542607e-05,
      "loss": 0.1163,
      "num_input_tokens_seen": 20541616,
      "step": 23720
    },
    {
      "epoch": 11.185761433286185,
      "grad_norm": 0.8266952633857727,
      "learning_rate": 1.7791750812276547e-05,
      "loss": 0.1413,
      "num_input_tokens_seen": 20545200,
      "step": 23725
    },
    {
      "epoch": 11.188118811881187,
      "grad_norm": 0.7324470281600952,
      "learning_rate": 1.778235082861309e-05,
      "loss": 0.147,
      "num_input_tokens_seen": 20549392,
      "step": 23730
    },
    {
      "epoch": 11.19047619047619,
      "grad_norm": 0.6136787533760071,
      "learning_rate": 1.777295195800184e-05,
      "loss": 0.1321,
      "num_input_tokens_seen": 20553008,
      "step": 23735
    },
    {
      "epoch": 11.192833569071192,
      "grad_norm": 0.802829384803772,
      "learning_rate": 1.7763554201892215e-05,
      "loss": 0.1669,
      "num_input_tokens_seen": 20557264,
      "step": 23740
    },
    {
      "epoch": 11.195190947666195,
      "grad_norm": 0.6631004810333252,
      "learning_rate": 1.7754157561733476e-05,
      "loss": 0.1316,
      "num_input_tokens_seen": 20562144,
      "step": 23745
    },
    {
      "epoch": 11.197548326261197,
      "grad_norm": 0.6824273467063904,
      "learning_rate": 1.7744762038974702e-05,
      "loss": 0.2017,
      "num_input_tokens_seen": 20565952,
      "step": 23750
    },
    {
      "epoch": 11.1999057048562,
      "grad_norm": 1.1487044095993042,
      "learning_rate": 1.7735367635064788e-05,
      "loss": 0.1921,
      "num_input_tokens_seen": 20570544,
      "step": 23755
    },
    {
      "epoch": 11.202263083451202,
      "grad_norm": 1.1107553243637085,
      "learning_rate": 1.7725974351452474e-05,
      "loss": 0.1546,
      "num_input_tokens_seen": 20576096,
      "step": 23760
    },
    {
      "epoch": 11.204620462046204,
      "grad_norm": 0.7298727035522461,
      "learning_rate": 1.771658218958634e-05,
      "loss": 0.2037,
      "num_input_tokens_seen": 20579984,
      "step": 23765
    },
    {
      "epoch": 11.206977840641207,
      "grad_norm": 0.8818562626838684,
      "learning_rate": 1.770719115091475e-05,
      "loss": 0.2225,
      "num_input_tokens_seen": 20584528,
      "step": 23770
    },
    {
      "epoch": 11.209335219236209,
      "grad_norm": 1.134826898574829,
      "learning_rate": 1.7697801236885935e-05,
      "loss": 0.2211,
      "num_input_tokens_seen": 20588384,
      "step": 23775
    },
    {
      "epoch": 11.211692597831211,
      "grad_norm": 1.5345913171768188,
      "learning_rate": 1.7688412448947944e-05,
      "loss": 0.2305,
      "num_input_tokens_seen": 20592992,
      "step": 23780
    },
    {
      "epoch": 11.214049976426214,
      "grad_norm": 0.9212716221809387,
      "learning_rate": 1.767902478854862e-05,
      "loss": 0.2105,
      "num_input_tokens_seen": 20597712,
      "step": 23785
    },
    {
      "epoch": 11.216407355021216,
      "grad_norm": 0.9696792960166931,
      "learning_rate": 1.766963825713569e-05,
      "loss": 0.2649,
      "num_input_tokens_seen": 20603216,
      "step": 23790
    },
    {
      "epoch": 11.218764733616219,
      "grad_norm": 1.1826742887496948,
      "learning_rate": 1.766025285615665e-05,
      "loss": 0.1672,
      "num_input_tokens_seen": 20607344,
      "step": 23795
    },
    {
      "epoch": 11.221122112211221,
      "grad_norm": 1.9719023704528809,
      "learning_rate": 1.7650868587058854e-05,
      "loss": 0.2597,
      "num_input_tokens_seen": 20611040,
      "step": 23800
    },
    {
      "epoch": 11.221122112211221,
      "eval_loss": 0.20324134826660156,
      "eval_runtime": 22.1889,
      "eval_samples_per_second": 42.499,
      "eval_steps_per_second": 21.272,
      "num_input_tokens_seen": 20611040,
      "step": 23800
    },
    {
      "epoch": 11.223479490806223,
      "grad_norm": 1.4798387289047241,
      "learning_rate": 1.7641485451289484e-05,
      "loss": 0.2149,
      "num_input_tokens_seen": 20614848,
      "step": 23805
    },
    {
      "epoch": 11.225836869401226,
      "grad_norm": 0.6813480854034424,
      "learning_rate": 1.7632103450295534e-05,
      "loss": 0.1638,
      "num_input_tokens_seen": 20619584,
      "step": 23810
    },
    {
      "epoch": 11.228194247996228,
      "grad_norm": 0.6684139370918274,
      "learning_rate": 1.762272258552381e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 20623408,
      "step": 23815
    },
    {
      "epoch": 11.23055162659123,
      "grad_norm": 0.7519953846931458,
      "learning_rate": 1.7613342858420988e-05,
      "loss": 0.1712,
      "num_input_tokens_seen": 20628128,
      "step": 23820
    },
    {
      "epoch": 11.232909005186233,
      "grad_norm": 1.4859540462493896,
      "learning_rate": 1.760396427043351e-05,
      "loss": 0.2261,
      "num_input_tokens_seen": 20632848,
      "step": 23825
    },
    {
      "epoch": 11.235266383781235,
      "grad_norm": 2.0364527702331543,
      "learning_rate": 1.7594586823007696e-05,
      "loss": 0.1463,
      "num_input_tokens_seen": 20637168,
      "step": 23830
    },
    {
      "epoch": 11.237623762376238,
      "grad_norm": 1.1809053421020508,
      "learning_rate": 1.7585210517589646e-05,
      "loss": 0.2093,
      "num_input_tokens_seen": 20641744,
      "step": 23835
    },
    {
      "epoch": 11.23998114097124,
      "grad_norm": 0.6880074739456177,
      "learning_rate": 1.7575835355625314e-05,
      "loss": 0.1687,
      "num_input_tokens_seen": 20646256,
      "step": 23840
    },
    {
      "epoch": 11.242338519566243,
      "grad_norm": 0.895853579044342,
      "learning_rate": 1.756646133856048e-05,
      "loss": 0.1327,
      "num_input_tokens_seen": 20650688,
      "step": 23845
    },
    {
      "epoch": 11.244695898161245,
      "grad_norm": 0.7876710295677185,
      "learning_rate": 1.7557088467840714e-05,
      "loss": 0.1996,
      "num_input_tokens_seen": 20655152,
      "step": 23850
    },
    {
      "epoch": 11.247053276756247,
      "grad_norm": 0.8934595584869385,
      "learning_rate": 1.7547716744911438e-05,
      "loss": 0.241,
      "num_input_tokens_seen": 20659712,
      "step": 23855
    },
    {
      "epoch": 11.24941065535125,
      "grad_norm": 1.860917568206787,
      "learning_rate": 1.7538346171217902e-05,
      "loss": 0.1464,
      "num_input_tokens_seen": 20664048,
      "step": 23860
    },
    {
      "epoch": 11.251768033946252,
      "grad_norm": 1.1997103691101074,
      "learning_rate": 1.7528976748205146e-05,
      "loss": 0.1282,
      "num_input_tokens_seen": 20667888,
      "step": 23865
    },
    {
      "epoch": 11.254125412541255,
      "grad_norm": 2.585282325744629,
      "learning_rate": 1.751960847731807e-05,
      "loss": 0.2296,
      "num_input_tokens_seen": 20672256,
      "step": 23870
    },
    {
      "epoch": 11.256482791136257,
      "grad_norm": 1.131248116493225,
      "learning_rate": 1.7510241360001362e-05,
      "loss": 0.2271,
      "num_input_tokens_seen": 20676400,
      "step": 23875
    },
    {
      "epoch": 11.25884016973126,
      "grad_norm": 0.9956717491149902,
      "learning_rate": 1.7500875397699562e-05,
      "loss": 0.1806,
      "num_input_tokens_seen": 20680352,
      "step": 23880
    },
    {
      "epoch": 11.261197548326262,
      "grad_norm": 1.5442192554473877,
      "learning_rate": 1.7491510591857015e-05,
      "loss": 0.2074,
      "num_input_tokens_seen": 20686368,
      "step": 23885
    },
    {
      "epoch": 11.263554926921264,
      "grad_norm": 0.7640234231948853,
      "learning_rate": 1.7482146943917896e-05,
      "loss": 0.2189,
      "num_input_tokens_seen": 20692064,
      "step": 23890
    },
    {
      "epoch": 11.265912305516267,
      "grad_norm": 0.523444414138794,
      "learning_rate": 1.7472784455326185e-05,
      "loss": 0.1569,
      "num_input_tokens_seen": 20697072,
      "step": 23895
    },
    {
      "epoch": 11.268269684111269,
      "grad_norm": 0.7585139870643616,
      "learning_rate": 1.746342312752572e-05,
      "loss": 0.1667,
      "num_input_tokens_seen": 20701232,
      "step": 23900
    },
    {
      "epoch": 11.270627062706271,
      "grad_norm": 1.5660617351531982,
      "learning_rate": 1.74540629619601e-05,
      "loss": 0.1836,
      "num_input_tokens_seen": 20705104,
      "step": 23905
    },
    {
      "epoch": 11.272984441301272,
      "grad_norm": 1.169471025466919,
      "learning_rate": 1.7444703960072815e-05,
      "loss": 0.1645,
      "num_input_tokens_seen": 20709616,
      "step": 23910
    },
    {
      "epoch": 11.275341819896274,
      "grad_norm": 1.5657061338424683,
      "learning_rate": 1.7435346123307118e-05,
      "loss": 0.1558,
      "num_input_tokens_seen": 20714368,
      "step": 23915
    },
    {
      "epoch": 11.277699198491277,
      "grad_norm": 1.0641130208969116,
      "learning_rate": 1.742598945310611e-05,
      "loss": 0.1808,
      "num_input_tokens_seen": 20719072,
      "step": 23920
    },
    {
      "epoch": 11.28005657708628,
      "grad_norm": 0.8541812300682068,
      "learning_rate": 1.741663395091272e-05,
      "loss": 0.1425,
      "num_input_tokens_seen": 20722896,
      "step": 23925
    },
    {
      "epoch": 11.282413955681282,
      "grad_norm": 0.6175968647003174,
      "learning_rate": 1.7407279618169657e-05,
      "loss": 0.161,
      "num_input_tokens_seen": 20727664,
      "step": 23930
    },
    {
      "epoch": 11.284771334276284,
      "grad_norm": 0.7991706132888794,
      "learning_rate": 1.73979264563195e-05,
      "loss": 0.1744,
      "num_input_tokens_seen": 20731520,
      "step": 23935
    },
    {
      "epoch": 11.287128712871286,
      "grad_norm": 0.6196434497833252,
      "learning_rate": 1.7388574466804625e-05,
      "loss": 0.2354,
      "num_input_tokens_seen": 20736208,
      "step": 23940
    },
    {
      "epoch": 11.289486091466289,
      "grad_norm": 1.276373028755188,
      "learning_rate": 1.7379223651067207e-05,
      "loss": 0.1986,
      "num_input_tokens_seen": 20740448,
      "step": 23945
    },
    {
      "epoch": 11.291843470061291,
      "grad_norm": 0.52667236328125,
      "learning_rate": 1.736987401054928e-05,
      "loss": 0.206,
      "num_input_tokens_seen": 20744640,
      "step": 23950
    },
    {
      "epoch": 11.294200848656294,
      "grad_norm": 0.7418292760848999,
      "learning_rate": 1.736052554669266e-05,
      "loss": 0.1947,
      "num_input_tokens_seen": 20749296,
      "step": 23955
    },
    {
      "epoch": 11.296558227251296,
      "grad_norm": 0.9661464095115662,
      "learning_rate": 1.7351178260939007e-05,
      "loss": 0.2155,
      "num_input_tokens_seen": 20753712,
      "step": 23960
    },
    {
      "epoch": 11.298915605846299,
      "grad_norm": 0.3700721263885498,
      "learning_rate": 1.7341832154729794e-05,
      "loss": 0.1622,
      "num_input_tokens_seen": 20757968,
      "step": 23965
    },
    {
      "epoch": 11.301272984441301,
      "grad_norm": 1.1551063060760498,
      "learning_rate": 1.7332487229506286e-05,
      "loss": 0.1362,
      "num_input_tokens_seen": 20762432,
      "step": 23970
    },
    {
      "epoch": 11.303630363036303,
      "grad_norm": 0.5663784742355347,
      "learning_rate": 1.732314348670961e-05,
      "loss": 0.2574,
      "num_input_tokens_seen": 20766320,
      "step": 23975
    },
    {
      "epoch": 11.305987741631306,
      "grad_norm": 1.8078259229660034,
      "learning_rate": 1.7313800927780686e-05,
      "loss": 0.1712,
      "num_input_tokens_seen": 20769872,
      "step": 23980
    },
    {
      "epoch": 11.308345120226308,
      "grad_norm": 1.6710171699523926,
      "learning_rate": 1.7304459554160245e-05,
      "loss": 0.2375,
      "num_input_tokens_seen": 20774912,
      "step": 23985
    },
    {
      "epoch": 11.31070249882131,
      "grad_norm": 0.5415478348731995,
      "learning_rate": 1.7295119367288853e-05,
      "loss": 0.1326,
      "num_input_tokens_seen": 20779072,
      "step": 23990
    },
    {
      "epoch": 11.313059877416313,
      "grad_norm": 1.1098743677139282,
      "learning_rate": 1.728578036860688e-05,
      "loss": 0.1556,
      "num_input_tokens_seen": 20783584,
      "step": 23995
    },
    {
      "epoch": 11.315417256011315,
      "grad_norm": 1.8470053672790527,
      "learning_rate": 1.7276442559554513e-05,
      "loss": 0.2103,
      "num_input_tokens_seen": 20787488,
      "step": 24000
    },
    {
      "epoch": 11.315417256011315,
      "eval_loss": 0.2023724466562271,
      "eval_runtime": 22.1127,
      "eval_samples_per_second": 42.645,
      "eval_steps_per_second": 21.345,
      "num_input_tokens_seen": 20787488,
      "step": 24000
    },
    {
      "epoch": 11.317774634606318,
      "grad_norm": 0.7628883123397827,
      "learning_rate": 1.726710594157177e-05,
      "loss": 0.156,
      "num_input_tokens_seen": 20791376,
      "step": 24005
    },
    {
      "epoch": 11.32013201320132,
      "grad_norm": 0.8095505237579346,
      "learning_rate": 1.725777051609846e-05,
      "loss": 0.199,
      "num_input_tokens_seen": 20795712,
      "step": 24010
    },
    {
      "epoch": 11.322489391796323,
      "grad_norm": 1.1596503257751465,
      "learning_rate": 1.7248436284574228e-05,
      "loss": 0.213,
      "num_input_tokens_seen": 20799520,
      "step": 24015
    },
    {
      "epoch": 11.324846770391325,
      "grad_norm": 0.39536741375923157,
      "learning_rate": 1.723910324843855e-05,
      "loss": 0.1544,
      "num_input_tokens_seen": 20804240,
      "step": 24020
    },
    {
      "epoch": 11.327204148986327,
      "grad_norm": 0.7076999545097351,
      "learning_rate": 1.722977140913067e-05,
      "loss": 0.1951,
      "num_input_tokens_seen": 20807936,
      "step": 24025
    },
    {
      "epoch": 11.32956152758133,
      "grad_norm": 0.6010247468948364,
      "learning_rate": 1.7220440768089688e-05,
      "loss": 0.2187,
      "num_input_tokens_seen": 20812464,
      "step": 24030
    },
    {
      "epoch": 11.331918906176332,
      "grad_norm": 0.7248252034187317,
      "learning_rate": 1.7211111326754505e-05,
      "loss": 0.1583,
      "num_input_tokens_seen": 20816640,
      "step": 24035
    },
    {
      "epoch": 11.334276284771335,
      "grad_norm": 1.8479503393173218,
      "learning_rate": 1.720178308656383e-05,
      "loss": 0.2294,
      "num_input_tokens_seen": 20820736,
      "step": 24040
    },
    {
      "epoch": 11.336633663366337,
      "grad_norm": 0.5402957201004028,
      "learning_rate": 1.719245604895621e-05,
      "loss": 0.1659,
      "num_input_tokens_seen": 20824448,
      "step": 24045
    },
    {
      "epoch": 11.33899104196134,
      "grad_norm": 0.7614288330078125,
      "learning_rate": 1.7183130215369972e-05,
      "loss": 0.2024,
      "num_input_tokens_seen": 20830240,
      "step": 24050
    },
    {
      "epoch": 11.341348420556342,
      "grad_norm": 0.9830533266067505,
      "learning_rate": 1.7173805587243292e-05,
      "loss": 0.1651,
      "num_input_tokens_seen": 20833584,
      "step": 24055
    },
    {
      "epoch": 11.343705799151344,
      "grad_norm": 1.3931732177734375,
      "learning_rate": 1.7164482166014147e-05,
      "loss": 0.2019,
      "num_input_tokens_seen": 20837424,
      "step": 24060
    },
    {
      "epoch": 11.346063177746347,
      "grad_norm": 1.0912232398986816,
      "learning_rate": 1.7155159953120313e-05,
      "loss": 0.1492,
      "num_input_tokens_seen": 20842736,
      "step": 24065
    },
    {
      "epoch": 11.348420556341349,
      "grad_norm": 0.5583867430686951,
      "learning_rate": 1.714583894999941e-05,
      "loss": 0.1519,
      "num_input_tokens_seen": 20846624,
      "step": 24070
    },
    {
      "epoch": 11.350777934936351,
      "grad_norm": 1.768693447113037,
      "learning_rate": 1.7136519158088826e-05,
      "loss": 0.205,
      "num_input_tokens_seen": 20850816,
      "step": 24075
    },
    {
      "epoch": 11.353135313531354,
      "grad_norm": 1.3845851421356201,
      "learning_rate": 1.712720057882581e-05,
      "loss": 0.1738,
      "num_input_tokens_seen": 20855344,
      "step": 24080
    },
    {
      "epoch": 11.355492692126356,
      "grad_norm": 0.9459949135780334,
      "learning_rate": 1.7117883213647413e-05,
      "loss": 0.144,
      "num_input_tokens_seen": 20859968,
      "step": 24085
    },
    {
      "epoch": 11.357850070721359,
      "grad_norm": 1.0335298776626587,
      "learning_rate": 1.710856706399046e-05,
      "loss": 0.1683,
      "num_input_tokens_seen": 20864912,
      "step": 24090
    },
    {
      "epoch": 11.360207449316361,
      "grad_norm": 0.8219898343086243,
      "learning_rate": 1.7099252131291648e-05,
      "loss": 0.2498,
      "num_input_tokens_seen": 20870352,
      "step": 24095
    },
    {
      "epoch": 11.362564827911363,
      "grad_norm": 1.6781047582626343,
      "learning_rate": 1.708993841698744e-05,
      "loss": 0.1872,
      "num_input_tokens_seen": 20874976,
      "step": 24100
    },
    {
      "epoch": 11.364922206506366,
      "grad_norm": 1.0913100242614746,
      "learning_rate": 1.7080625922514132e-05,
      "loss": 0.1184,
      "num_input_tokens_seen": 20878848,
      "step": 24105
    },
    {
      "epoch": 11.367279585101368,
      "grad_norm": 0.6796855330467224,
      "learning_rate": 1.7071314649307836e-05,
      "loss": 0.1689,
      "num_input_tokens_seen": 20883072,
      "step": 24110
    },
    {
      "epoch": 11.369636963696369,
      "grad_norm": 0.891778290271759,
      "learning_rate": 1.7062004598804448e-05,
      "loss": 0.2271,
      "num_input_tokens_seen": 20888032,
      "step": 24115
    },
    {
      "epoch": 11.371994342291371,
      "grad_norm": 0.9896553754806519,
      "learning_rate": 1.7052695772439702e-05,
      "loss": 0.2352,
      "num_input_tokens_seen": 20892432,
      "step": 24120
    },
    {
      "epoch": 11.374351720886374,
      "grad_norm": 0.7729418277740479,
      "learning_rate": 1.7043388171649154e-05,
      "loss": 0.1187,
      "num_input_tokens_seen": 20896432,
      "step": 24125
    },
    {
      "epoch": 11.376709099481376,
      "grad_norm": 1.320183277130127,
      "learning_rate": 1.7034081797868127e-05,
      "loss": 0.167,
      "num_input_tokens_seen": 20900656,
      "step": 24130
    },
    {
      "epoch": 11.379066478076378,
      "grad_norm": 0.9703401327133179,
      "learning_rate": 1.70247766525318e-05,
      "loss": 0.2191,
      "num_input_tokens_seen": 20904128,
      "step": 24135
    },
    {
      "epoch": 11.38142385667138,
      "grad_norm": 0.7657676339149475,
      "learning_rate": 1.701547273707514e-05,
      "loss": 0.1967,
      "num_input_tokens_seen": 20907728,
      "step": 24140
    },
    {
      "epoch": 11.383781235266383,
      "grad_norm": 0.8404051661491394,
      "learning_rate": 1.7006170052932916e-05,
      "loss": 0.2071,
      "num_input_tokens_seen": 20912224,
      "step": 24145
    },
    {
      "epoch": 11.386138613861386,
      "grad_norm": 0.8600794672966003,
      "learning_rate": 1.6996868601539735e-05,
      "loss": 0.1647,
      "num_input_tokens_seen": 20916576,
      "step": 24150
    },
    {
      "epoch": 11.388495992456388,
      "grad_norm": 1.0698598623275757,
      "learning_rate": 1.6987568384329977e-05,
      "loss": 0.1777,
      "num_input_tokens_seen": 20920464,
      "step": 24155
    },
    {
      "epoch": 11.39085337105139,
      "grad_norm": 1.1427483558654785,
      "learning_rate": 1.6978269402737866e-05,
      "loss": 0.1485,
      "num_input_tokens_seen": 20925232,
      "step": 24160
    },
    {
      "epoch": 11.393210749646393,
      "grad_norm": 0.8440130352973938,
      "learning_rate": 1.696897165819743e-05,
      "loss": 0.2057,
      "num_input_tokens_seen": 20929056,
      "step": 24165
    },
    {
      "epoch": 11.395568128241395,
      "grad_norm": 1.6081115007400513,
      "learning_rate": 1.6959675152142487e-05,
      "loss": 0.1707,
      "num_input_tokens_seen": 20933264,
      "step": 24170
    },
    {
      "epoch": 11.397925506836398,
      "grad_norm": 0.5988689661026001,
      "learning_rate": 1.6950379886006667e-05,
      "loss": 0.2065,
      "num_input_tokens_seen": 20936944,
      "step": 24175
    },
    {
      "epoch": 11.4002828854314,
      "grad_norm": 1.5990351438522339,
      "learning_rate": 1.6941085861223438e-05,
      "loss": 0.1876,
      "num_input_tokens_seen": 20940768,
      "step": 24180
    },
    {
      "epoch": 11.402640264026402,
      "grad_norm": 0.7345959544181824,
      "learning_rate": 1.6931793079226034e-05,
      "loss": 0.2103,
      "num_input_tokens_seen": 20944736,
      "step": 24185
    },
    {
      "epoch": 11.404997642621405,
      "grad_norm": 2.296865463256836,
      "learning_rate": 1.692250154144754e-05,
      "loss": 0.1826,
      "num_input_tokens_seen": 20948960,
      "step": 24190
    },
    {
      "epoch": 11.407355021216407,
      "grad_norm": 1.4703984260559082,
      "learning_rate": 1.6913211249320807e-05,
      "loss": 0.1284,
      "num_input_tokens_seen": 20953440,
      "step": 24195
    },
    {
      "epoch": 11.40971239981141,
      "grad_norm": 0.820636510848999,
      "learning_rate": 1.6903922204278522e-05,
      "loss": 0.2394,
      "num_input_tokens_seen": 20958240,
      "step": 24200
    },
    {
      "epoch": 11.40971239981141,
      "eval_loss": 0.20195287466049194,
      "eval_runtime": 22.1171,
      "eval_samples_per_second": 42.637,
      "eval_steps_per_second": 21.341,
      "num_input_tokens_seen": 20958240,
      "step": 24200
    },
    {
      "epoch": 11.412069778406412,
      "grad_norm": 0.8148500323295593,
      "learning_rate": 1.6894634407753186e-05,
      "loss": 0.1694,
      "num_input_tokens_seen": 20962608,
      "step": 24205
    },
    {
      "epoch": 11.414427157001414,
      "grad_norm": 1.7950761318206787,
      "learning_rate": 1.6885347861177077e-05,
      "loss": 0.1555,
      "num_input_tokens_seen": 20966848,
      "step": 24210
    },
    {
      "epoch": 11.416784535596417,
      "grad_norm": 0.5965088605880737,
      "learning_rate": 1.6876062565982298e-05,
      "loss": 0.1846,
      "num_input_tokens_seen": 20971184,
      "step": 24215
    },
    {
      "epoch": 11.41914191419142,
      "grad_norm": 0.6758869290351868,
      "learning_rate": 1.6866778523600774e-05,
      "loss": 0.1983,
      "num_input_tokens_seen": 20975568,
      "step": 24220
    },
    {
      "epoch": 11.421499292786422,
      "grad_norm": 0.4647976756095886,
      "learning_rate": 1.6857495735464195e-05,
      "loss": 0.1508,
      "num_input_tokens_seen": 20979264,
      "step": 24225
    },
    {
      "epoch": 11.423856671381424,
      "grad_norm": 1.6470284461975098,
      "learning_rate": 1.6848214203004115e-05,
      "loss": 0.1687,
      "num_input_tokens_seen": 20983312,
      "step": 24230
    },
    {
      "epoch": 11.426214049976426,
      "grad_norm": 1.1801023483276367,
      "learning_rate": 1.6838933927651835e-05,
      "loss": 0.1183,
      "num_input_tokens_seen": 20988480,
      "step": 24235
    },
    {
      "epoch": 11.428571428571429,
      "grad_norm": 0.33375445008277893,
      "learning_rate": 1.6829654910838506e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 20994288,
      "step": 24240
    },
    {
      "epoch": 11.430928807166431,
      "grad_norm": 0.7563997507095337,
      "learning_rate": 1.6820377153995065e-05,
      "loss": 0.141,
      "num_input_tokens_seen": 20999264,
      "step": 24245
    },
    {
      "epoch": 11.433286185761434,
      "grad_norm": 1.212962031364441,
      "learning_rate": 1.681110065855226e-05,
      "loss": 0.1544,
      "num_input_tokens_seen": 21003504,
      "step": 24250
    },
    {
      "epoch": 11.435643564356436,
      "grad_norm": 0.5085329413414001,
      "learning_rate": 1.6801825425940642e-05,
      "loss": 0.1437,
      "num_input_tokens_seen": 21007328,
      "step": 24255
    },
    {
      "epoch": 11.438000942951438,
      "grad_norm": 0.41125592589378357,
      "learning_rate": 1.679255145759056e-05,
      "loss": 0.2588,
      "num_input_tokens_seen": 21011200,
      "step": 24260
    },
    {
      "epoch": 11.44035832154644,
      "grad_norm": 0.7692433595657349,
      "learning_rate": 1.6783278754932187e-05,
      "loss": 0.1794,
      "num_input_tokens_seen": 21015712,
      "step": 24265
    },
    {
      "epoch": 11.442715700141443,
      "grad_norm": 0.6771591901779175,
      "learning_rate": 1.6774007319395496e-05,
      "loss": 0.1278,
      "num_input_tokens_seen": 21019792,
      "step": 24270
    },
    {
      "epoch": 11.445073078736446,
      "grad_norm": 0.6152214407920837,
      "learning_rate": 1.6764737152410243e-05,
      "loss": 0.2496,
      "num_input_tokens_seen": 21024128,
      "step": 24275
    },
    {
      "epoch": 11.447430457331448,
      "grad_norm": 0.9169282913208008,
      "learning_rate": 1.6755468255406016e-05,
      "loss": 0.1541,
      "num_input_tokens_seen": 21028816,
      "step": 24280
    },
    {
      "epoch": 11.44978783592645,
      "grad_norm": 1.2966947555541992,
      "learning_rate": 1.674620062981219e-05,
      "loss": 0.2315,
      "num_input_tokens_seen": 21033824,
      "step": 24285
    },
    {
      "epoch": 11.452145214521453,
      "grad_norm": 1.441890835762024,
      "learning_rate": 1.6736934277057947e-05,
      "loss": 0.1953,
      "num_input_tokens_seen": 21037248,
      "step": 24290
    },
    {
      "epoch": 11.454502593116455,
      "grad_norm": 1.1056917905807495,
      "learning_rate": 1.6727669198572286e-05,
      "loss": 0.1249,
      "num_input_tokens_seen": 21041168,
      "step": 24295
    },
    {
      "epoch": 11.456859971711458,
      "grad_norm": 0.8865184187889099,
      "learning_rate": 1.6718405395783984e-05,
      "loss": 0.1661,
      "num_input_tokens_seen": 21044640,
      "step": 24300
    },
    {
      "epoch": 11.45921735030646,
      "grad_norm": 0.7160131931304932,
      "learning_rate": 1.6709142870121643e-05,
      "loss": 0.1563,
      "num_input_tokens_seen": 21048944,
      "step": 24305
    },
    {
      "epoch": 11.46157472890146,
      "grad_norm": 0.744384229183197,
      "learning_rate": 1.669988162301367e-05,
      "loss": 0.2012,
      "num_input_tokens_seen": 21054048,
      "step": 24310
    },
    {
      "epoch": 11.463932107496463,
      "grad_norm": 0.4283408522605896,
      "learning_rate": 1.6690621655888243e-05,
      "loss": 0.1944,
      "num_input_tokens_seen": 21058400,
      "step": 24315
    },
    {
      "epoch": 11.466289486091465,
      "grad_norm": 2.040719509124756,
      "learning_rate": 1.6681362970173386e-05,
      "loss": 0.3242,
      "num_input_tokens_seen": 21063056,
      "step": 24320
    },
    {
      "epoch": 11.468646864686468,
      "grad_norm": 0.7168094515800476,
      "learning_rate": 1.6672105567296904e-05,
      "loss": 0.1646,
      "num_input_tokens_seen": 21067328,
      "step": 24325
    },
    {
      "epoch": 11.47100424328147,
      "grad_norm": 0.7833216190338135,
      "learning_rate": 1.666284944868639e-05,
      "loss": 0.1907,
      "num_input_tokens_seen": 21072160,
      "step": 24330
    },
    {
      "epoch": 11.473361621876473,
      "grad_norm": 0.867788553237915,
      "learning_rate": 1.665359461576927e-05,
      "loss": 0.2287,
      "num_input_tokens_seen": 21076128,
      "step": 24335
    },
    {
      "epoch": 11.475719000471475,
      "grad_norm": 0.9347960948944092,
      "learning_rate": 1.6644341069972736e-05,
      "loss": 0.1852,
      "num_input_tokens_seen": 21080208,
      "step": 24340
    },
    {
      "epoch": 11.478076379066477,
      "grad_norm": 1.2007813453674316,
      "learning_rate": 1.6635088812723813e-05,
      "loss": 0.2137,
      "num_input_tokens_seen": 21085040,
      "step": 24345
    },
    {
      "epoch": 11.48043375766148,
      "grad_norm": 0.5310762524604797,
      "learning_rate": 1.6625837845449328e-05,
      "loss": 0.1704,
      "num_input_tokens_seen": 21089200,
      "step": 24350
    },
    {
      "epoch": 11.482791136256482,
      "grad_norm": 0.3461885154247284,
      "learning_rate": 1.6616588169575874e-05,
      "loss": 0.1374,
      "num_input_tokens_seen": 21092944,
      "step": 24355
    },
    {
      "epoch": 11.485148514851485,
      "grad_norm": 1.5480281114578247,
      "learning_rate": 1.6607339786529878e-05,
      "loss": 0.1559,
      "num_input_tokens_seen": 21097136,
      "step": 24360
    },
    {
      "epoch": 11.487505893446487,
      "grad_norm": 0.777969479560852,
      "learning_rate": 1.659809269773756e-05,
      "loss": 0.1441,
      "num_input_tokens_seen": 21102496,
      "step": 24365
    },
    {
      "epoch": 11.48986327204149,
      "grad_norm": 2.2137537002563477,
      "learning_rate": 1.658884690462493e-05,
      "loss": 0.2449,
      "num_input_tokens_seen": 21106624,
      "step": 24370
    },
    {
      "epoch": 11.492220650636492,
      "grad_norm": 1.0362876653671265,
      "learning_rate": 1.6579602408617813e-05,
      "loss": 0.1998,
      "num_input_tokens_seen": 21111632,
      "step": 24375
    },
    {
      "epoch": 11.494578029231494,
      "grad_norm": 0.77972012758255,
      "learning_rate": 1.657035921114181e-05,
      "loss": 0.1361,
      "num_input_tokens_seen": 21115680,
      "step": 24380
    },
    {
      "epoch": 11.496935407826497,
      "grad_norm": 0.9175909161567688,
      "learning_rate": 1.656111731362236e-05,
      "loss": 0.1782,
      "num_input_tokens_seen": 21120192,
      "step": 24385
    },
    {
      "epoch": 11.499292786421499,
      "grad_norm": 0.693439245223999,
      "learning_rate": 1.6551876717484666e-05,
      "loss": 0.2038,
      "num_input_tokens_seen": 21124144,
      "step": 24390
    },
    {
      "epoch": 11.501650165016502,
      "grad_norm": 1.1665743589401245,
      "learning_rate": 1.6542637424153752e-05,
      "loss": 0.1851,
      "num_input_tokens_seen": 21129024,
      "step": 24395
    },
    {
      "epoch": 11.504007543611504,
      "grad_norm": 1.7367985248565674,
      "learning_rate": 1.6533399435054418e-05,
      "loss": 0.2223,
      "num_input_tokens_seen": 21133392,
      "step": 24400
    },
    {
      "epoch": 11.504007543611504,
      "eval_loss": 0.20144544541835785,
      "eval_runtime": 22.1465,
      "eval_samples_per_second": 42.58,
      "eval_steps_per_second": 21.313,
      "num_input_tokens_seen": 21133392,
      "step": 24400
    },
    {
      "epoch": 11.506364922206506,
      "grad_norm": 1.2246366739273071,
      "learning_rate": 1.6524162751611304e-05,
      "loss": 0.1521,
      "num_input_tokens_seen": 21137664,
      "step": 24405
    },
    {
      "epoch": 11.508722300801509,
      "grad_norm": 0.6022456288337708,
      "learning_rate": 1.6514927375248796e-05,
      "loss": 0.1639,
      "num_input_tokens_seen": 21141136,
      "step": 24410
    },
    {
      "epoch": 11.511079679396511,
      "grad_norm": 0.3412388265132904,
      "learning_rate": 1.6505693307391127e-05,
      "loss": 0.137,
      "num_input_tokens_seen": 21146080,
      "step": 24415
    },
    {
      "epoch": 11.513437057991514,
      "grad_norm": 0.758820652961731,
      "learning_rate": 1.6496460549462288e-05,
      "loss": 0.1867,
      "num_input_tokens_seen": 21149920,
      "step": 24420
    },
    {
      "epoch": 11.515794436586516,
      "grad_norm": 1.4116379022598267,
      "learning_rate": 1.6487229102886097e-05,
      "loss": 0.2196,
      "num_input_tokens_seen": 21153952,
      "step": 24425
    },
    {
      "epoch": 11.518151815181518,
      "grad_norm": 0.6922993063926697,
      "learning_rate": 1.6477998969086155e-05,
      "loss": 0.1933,
      "num_input_tokens_seen": 21158768,
      "step": 24430
    },
    {
      "epoch": 11.52050919377652,
      "grad_norm": 1.1000475883483887,
      "learning_rate": 1.646877014948587e-05,
      "loss": 0.184,
      "num_input_tokens_seen": 21162720,
      "step": 24435
    },
    {
      "epoch": 11.522866572371523,
      "grad_norm": 0.6671265959739685,
      "learning_rate": 1.6459542645508433e-05,
      "loss": 0.2025,
      "num_input_tokens_seen": 21166944,
      "step": 24440
    },
    {
      "epoch": 11.525223950966526,
      "grad_norm": 1.0156469345092773,
      "learning_rate": 1.6450316458576852e-05,
      "loss": 0.1702,
      "num_input_tokens_seen": 21171904,
      "step": 24445
    },
    {
      "epoch": 11.527581329561528,
      "grad_norm": 1.0871717929840088,
      "learning_rate": 1.6441091590113912e-05,
      "loss": 0.2027,
      "num_input_tokens_seen": 21175728,
      "step": 24450
    },
    {
      "epoch": 11.52993870815653,
      "grad_norm": 1.2718652486801147,
      "learning_rate": 1.6431868041542213e-05,
      "loss": 0.1579,
      "num_input_tokens_seen": 21180192,
      "step": 24455
    },
    {
      "epoch": 11.532296086751533,
      "grad_norm": 1.2084397077560425,
      "learning_rate": 1.6422645814284123e-05,
      "loss": 0.1367,
      "num_input_tokens_seen": 21184224,
      "step": 24460
    },
    {
      "epoch": 11.534653465346535,
      "grad_norm": 1.0261218547821045,
      "learning_rate": 1.6413424909761846e-05,
      "loss": 0.2337,
      "num_input_tokens_seen": 21188656,
      "step": 24465
    },
    {
      "epoch": 11.537010843941538,
      "grad_norm": 1.5129669904708862,
      "learning_rate": 1.640420532939736e-05,
      "loss": 0.1447,
      "num_input_tokens_seen": 21192912,
      "step": 24470
    },
    {
      "epoch": 11.53936822253654,
      "grad_norm": 0.9586169719696045,
      "learning_rate": 1.639498707461242e-05,
      "loss": 0.1717,
      "num_input_tokens_seen": 21197680,
      "step": 24475
    },
    {
      "epoch": 11.541725601131542,
      "grad_norm": 0.49878379702568054,
      "learning_rate": 1.6385770146828614e-05,
      "loss": 0.1605,
      "num_input_tokens_seen": 21201408,
      "step": 24480
    },
    {
      "epoch": 11.544082979726545,
      "grad_norm": 0.9673119783401489,
      "learning_rate": 1.637655454746731e-05,
      "loss": 0.1939,
      "num_input_tokens_seen": 21205392,
      "step": 24485
    },
    {
      "epoch": 11.546440358321547,
      "grad_norm": 1.2510573863983154,
      "learning_rate": 1.6367340277949658e-05,
      "loss": 0.2233,
      "num_input_tokens_seen": 21210048,
      "step": 24490
    },
    {
      "epoch": 11.54879773691655,
      "grad_norm": 0.563922643661499,
      "learning_rate": 1.635812733969663e-05,
      "loss": 0.1704,
      "num_input_tokens_seen": 21213840,
      "step": 24495
    },
    {
      "epoch": 11.551155115511552,
      "grad_norm": 1.0569757223129272,
      "learning_rate": 1.634891573412896e-05,
      "loss": 0.205,
      "num_input_tokens_seen": 21218160,
      "step": 24500
    },
    {
      "epoch": 11.553512494106554,
      "grad_norm": 1.0396267175674438,
      "learning_rate": 1.6339705462667196e-05,
      "loss": 0.2286,
      "num_input_tokens_seen": 21222688,
      "step": 24505
    },
    {
      "epoch": 11.555869872701557,
      "grad_norm": 0.6256810426712036,
      "learning_rate": 1.633049652673169e-05,
      "loss": 0.1337,
      "num_input_tokens_seen": 21226464,
      "step": 24510
    },
    {
      "epoch": 11.558227251296557,
      "grad_norm": 1.118095874786377,
      "learning_rate": 1.632128892774256e-05,
      "loss": 0.2409,
      "num_input_tokens_seen": 21230544,
      "step": 24515
    },
    {
      "epoch": 11.56058462989156,
      "grad_norm": 1.8801697492599487,
      "learning_rate": 1.6312082667119737e-05,
      "loss": 0.2035,
      "num_input_tokens_seen": 21235088,
      "step": 24520
    },
    {
      "epoch": 11.562942008486562,
      "grad_norm": 1.3692238330841064,
      "learning_rate": 1.630287774628296e-05,
      "loss": 0.1774,
      "num_input_tokens_seen": 21239504,
      "step": 24525
    },
    {
      "epoch": 11.565299387081565,
      "grad_norm": 1.408896803855896,
      "learning_rate": 1.6293674166651718e-05,
      "loss": 0.1815,
      "num_input_tokens_seen": 21243392,
      "step": 24530
    },
    {
      "epoch": 11.567656765676567,
      "grad_norm": 1.8748209476470947,
      "learning_rate": 1.6284471929645338e-05,
      "loss": 0.2087,
      "num_input_tokens_seen": 21246816,
      "step": 24535
    },
    {
      "epoch": 11.57001414427157,
      "grad_norm": 0.9871891736984253,
      "learning_rate": 1.627527103668291e-05,
      "loss": 0.1032,
      "num_input_tokens_seen": 21251920,
      "step": 24540
    },
    {
      "epoch": 11.572371522866572,
      "grad_norm": 1.5118013620376587,
      "learning_rate": 1.6266071489183327e-05,
      "loss": 0.1974,
      "num_input_tokens_seen": 21256640,
      "step": 24545
    },
    {
      "epoch": 11.574728901461574,
      "grad_norm": 0.5161932110786438,
      "learning_rate": 1.6256873288565283e-05,
      "loss": 0.1915,
      "num_input_tokens_seen": 21260784,
      "step": 24550
    },
    {
      "epoch": 11.577086280056577,
      "grad_norm": 1.0635205507278442,
      "learning_rate": 1.6247676436247245e-05,
      "loss": 0.1003,
      "num_input_tokens_seen": 21265712,
      "step": 24555
    },
    {
      "epoch": 11.579443658651579,
      "grad_norm": 0.8676229119300842,
      "learning_rate": 1.6238480933647486e-05,
      "loss": 0.1806,
      "num_input_tokens_seen": 21269360,
      "step": 24560
    },
    {
      "epoch": 11.581801037246581,
      "grad_norm": 0.6803581714630127,
      "learning_rate": 1.6229286782184083e-05,
      "loss": 0.123,
      "num_input_tokens_seen": 21273440,
      "step": 24565
    },
    {
      "epoch": 11.584158415841584,
      "grad_norm": 3.0590884685516357,
      "learning_rate": 1.622009398327487e-05,
      "loss": 0.2466,
      "num_input_tokens_seen": 21278064,
      "step": 24570
    },
    {
      "epoch": 11.586515794436586,
      "grad_norm": 1.058477759361267,
      "learning_rate": 1.6210902538337502e-05,
      "loss": 0.2478,
      "num_input_tokens_seen": 21282336,
      "step": 24575
    },
    {
      "epoch": 11.588873173031589,
      "grad_norm": 0.9900107383728027,
      "learning_rate": 1.6201712448789413e-05,
      "loss": 0.2601,
      "num_input_tokens_seen": 21286544,
      "step": 24580
    },
    {
      "epoch": 11.591230551626591,
      "grad_norm": 0.41711127758026123,
      "learning_rate": 1.6192523716047827e-05,
      "loss": 0.1571,
      "num_input_tokens_seen": 21290352,
      "step": 24585
    },
    {
      "epoch": 11.593587930221593,
      "grad_norm": 0.5574778914451599,
      "learning_rate": 1.6183336341529776e-05,
      "loss": 0.1714,
      "num_input_tokens_seen": 21295072,
      "step": 24590
    },
    {
      "epoch": 11.595945308816596,
      "grad_norm": 1.2004488706588745,
      "learning_rate": 1.6174150326652047e-05,
      "loss": 0.1777,
      "num_input_tokens_seen": 21299440,
      "step": 24595
    },
    {
      "epoch": 11.598302687411598,
      "grad_norm": 1.2507059574127197,
      "learning_rate": 1.6164965672831256e-05,
      "loss": 0.1571,
      "num_input_tokens_seen": 21303360,
      "step": 24600
    },
    {
      "epoch": 11.598302687411598,
      "eval_loss": 0.20018865168094635,
      "eval_runtime": 22.1116,
      "eval_samples_per_second": 42.647,
      "eval_steps_per_second": 21.346,
      "num_input_tokens_seen": 21303360,
      "step": 24600
    },
    {
      "epoch": 11.6006600660066,
      "grad_norm": 0.7882770895957947,
      "learning_rate": 1.6155782381483784e-05,
      "loss": 0.1613,
      "num_input_tokens_seen": 21307776,
      "step": 24605
    },
    {
      "epoch": 11.603017444601603,
      "grad_norm": 1.5404224395751953,
      "learning_rate": 1.6146600454025813e-05,
      "loss": 0.2298,
      "num_input_tokens_seen": 21311696,
      "step": 24610
    },
    {
      "epoch": 11.605374823196605,
      "grad_norm": 1.0231261253356934,
      "learning_rate": 1.6137419891873317e-05,
      "loss": 0.1821,
      "num_input_tokens_seen": 21315808,
      "step": 24615
    },
    {
      "epoch": 11.607732201791608,
      "grad_norm": 0.9479862451553345,
      "learning_rate": 1.6128240696442038e-05,
      "loss": 0.1839,
      "num_input_tokens_seen": 21319968,
      "step": 24620
    },
    {
      "epoch": 11.61008958038661,
      "grad_norm": 1.1673953533172607,
      "learning_rate": 1.611906286914753e-05,
      "loss": 0.1447,
      "num_input_tokens_seen": 21324416,
      "step": 24625
    },
    {
      "epoch": 11.612446958981613,
      "grad_norm": 0.8835504651069641,
      "learning_rate": 1.6109886411405144e-05,
      "loss": 0.1471,
      "num_input_tokens_seen": 21328784,
      "step": 24630
    },
    {
      "epoch": 11.614804337576615,
      "grad_norm": 1.933476209640503,
      "learning_rate": 1.6100711324629985e-05,
      "loss": 0.2017,
      "num_input_tokens_seen": 21333744,
      "step": 24635
    },
    {
      "epoch": 11.617161716171617,
      "grad_norm": 0.47356149554252625,
      "learning_rate": 1.609153761023698e-05,
      "loss": 0.1762,
      "num_input_tokens_seen": 21337696,
      "step": 24640
    },
    {
      "epoch": 11.61951909476662,
      "grad_norm": 1.2119898796081543,
      "learning_rate": 1.608236526964083e-05,
      "loss": 0.226,
      "num_input_tokens_seen": 21342272,
      "step": 24645
    },
    {
      "epoch": 11.621876473361622,
      "grad_norm": 0.3636210560798645,
      "learning_rate": 1.607319430425601e-05,
      "loss": 0.124,
      "num_input_tokens_seen": 21345936,
      "step": 24650
    },
    {
      "epoch": 11.624233851956625,
      "grad_norm": 1.5921895503997803,
      "learning_rate": 1.606402471549682e-05,
      "loss": 0.1823,
      "num_input_tokens_seen": 21350672,
      "step": 24655
    },
    {
      "epoch": 11.626591230551627,
      "grad_norm": 0.889611542224884,
      "learning_rate": 1.6054856504777312e-05,
      "loss": 0.1733,
      "num_input_tokens_seen": 21355232,
      "step": 24660
    },
    {
      "epoch": 11.62894860914663,
      "grad_norm": 0.8986194729804993,
      "learning_rate": 1.6045689673511334e-05,
      "loss": 0.1901,
      "num_input_tokens_seen": 21359744,
      "step": 24665
    },
    {
      "epoch": 11.631305987741632,
      "grad_norm": 2.212350845336914,
      "learning_rate": 1.6036524223112548e-05,
      "loss": 0.2272,
      "num_input_tokens_seen": 21363104,
      "step": 24670
    },
    {
      "epoch": 11.633663366336634,
      "grad_norm": 1.0784621238708496,
      "learning_rate": 1.602736015499436e-05,
      "loss": 0.1617,
      "num_input_tokens_seen": 21367232,
      "step": 24675
    },
    {
      "epoch": 11.636020744931637,
      "grad_norm": 0.7521235346794128,
      "learning_rate": 1.601819747057e-05,
      "loss": 0.1347,
      "num_input_tokens_seen": 21371808,
      "step": 24680
    },
    {
      "epoch": 11.638378123526639,
      "grad_norm": 0.6575235724449158,
      "learning_rate": 1.6009036171252465e-05,
      "loss": 0.1209,
      "num_input_tokens_seen": 21375840,
      "step": 24685
    },
    {
      "epoch": 11.640735502121641,
      "grad_norm": 0.7435993552207947,
      "learning_rate": 1.599987625845453e-05,
      "loss": 0.2281,
      "num_input_tokens_seen": 21380112,
      "step": 24690
    },
    {
      "epoch": 11.643092880716644,
      "grad_norm": 0.9461956024169922,
      "learning_rate": 1.599071773358879e-05,
      "loss": 0.1983,
      "num_input_tokens_seen": 21384096,
      "step": 24695
    },
    {
      "epoch": 11.645450259311646,
      "grad_norm": 2.0221636295318604,
      "learning_rate": 1.598156059806758e-05,
      "loss": 0.134,
      "num_input_tokens_seen": 21387728,
      "step": 24700
    },
    {
      "epoch": 11.647807637906649,
      "grad_norm": 1.3191345930099487,
      "learning_rate": 1.5972404853303062e-05,
      "loss": 0.2548,
      "num_input_tokens_seen": 21391808,
      "step": 24705
    },
    {
      "epoch": 11.65016501650165,
      "grad_norm": 1.1094589233398438,
      "learning_rate": 1.5963250500707172e-05,
      "loss": 0.1667,
      "num_input_tokens_seen": 21396832,
      "step": 24710
    },
    {
      "epoch": 11.652522395096653,
      "grad_norm": 1.606717824935913,
      "learning_rate": 1.5954097541691612e-05,
      "loss": 0.169,
      "num_input_tokens_seen": 21401184,
      "step": 24715
    },
    {
      "epoch": 11.654879773691654,
      "grad_norm": 1.0892565250396729,
      "learning_rate": 1.5944945977667884e-05,
      "loss": 0.1377,
      "num_input_tokens_seen": 21406176,
      "step": 24720
    },
    {
      "epoch": 11.657237152286656,
      "grad_norm": 0.972034215927124,
      "learning_rate": 1.593579581004729e-05,
      "loss": 0.1691,
      "num_input_tokens_seen": 21411312,
      "step": 24725
    },
    {
      "epoch": 11.659594530881659,
      "grad_norm": 0.6625174880027771,
      "learning_rate": 1.592664704024088e-05,
      "loss": 0.1179,
      "num_input_tokens_seen": 21415616,
      "step": 24730
    },
    {
      "epoch": 11.661951909476661,
      "grad_norm": 0.4909258782863617,
      "learning_rate": 1.591749966965953e-05,
      "loss": 0.1966,
      "num_input_tokens_seen": 21419904,
      "step": 24735
    },
    {
      "epoch": 11.664309288071664,
      "grad_norm": 0.6824101805686951,
      "learning_rate": 1.5908353699713856e-05,
      "loss": 0.1527,
      "num_input_tokens_seen": 21424704,
      "step": 24740
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 0.32003912329673767,
      "learning_rate": 1.5899209131814298e-05,
      "loss": 0.0936,
      "num_input_tokens_seen": 21429472,
      "step": 24745
    },
    {
      "epoch": 11.669024045261668,
      "grad_norm": 1.402883768081665,
      "learning_rate": 1.5890065967371067e-05,
      "loss": 0.2216,
      "num_input_tokens_seen": 21433952,
      "step": 24750
    },
    {
      "epoch": 11.67138142385667,
      "grad_norm": 0.5915811061859131,
      "learning_rate": 1.5880924207794144e-05,
      "loss": 0.1642,
      "num_input_tokens_seen": 21438128,
      "step": 24755
    },
    {
      "epoch": 11.673738802451673,
      "grad_norm": 0.654824435710907,
      "learning_rate": 1.5871783854493298e-05,
      "loss": 0.1555,
      "num_input_tokens_seen": 21441504,
      "step": 24760
    },
    {
      "epoch": 11.676096181046676,
      "grad_norm": 0.45139044523239136,
      "learning_rate": 1.5862644908878106e-05,
      "loss": 0.2048,
      "num_input_tokens_seen": 21445232,
      "step": 24765
    },
    {
      "epoch": 11.678453559641678,
      "grad_norm": 0.6976479291915894,
      "learning_rate": 1.5853507372357885e-05,
      "loss": 0.1793,
      "num_input_tokens_seen": 21449280,
      "step": 24770
    },
    {
      "epoch": 11.68081093823668,
      "grad_norm": 0.39913415908813477,
      "learning_rate": 1.5844371246341776e-05,
      "loss": 0.1612,
      "num_input_tokens_seen": 21453200,
      "step": 24775
    },
    {
      "epoch": 11.683168316831683,
      "grad_norm": 1.437680721282959,
      "learning_rate": 1.5835236532238674e-05,
      "loss": 0.2268,
      "num_input_tokens_seen": 21457888,
      "step": 24780
    },
    {
      "epoch": 11.685525695426685,
      "grad_norm": 0.682899534702301,
      "learning_rate": 1.582610323145727e-05,
      "loss": 0.1482,
      "num_input_tokens_seen": 21462176,
      "step": 24785
    },
    {
      "epoch": 11.687883074021688,
      "grad_norm": 0.8126052021980286,
      "learning_rate": 1.5816971345406035e-05,
      "loss": 0.1581,
      "num_input_tokens_seen": 21466336,
      "step": 24790
    },
    {
      "epoch": 11.69024045261669,
      "grad_norm": 1.13790762424469,
      "learning_rate": 1.5807840875493225e-05,
      "loss": 0.2042,
      "num_input_tokens_seen": 21471296,
      "step": 24795
    },
    {
      "epoch": 11.692597831211692,
      "grad_norm": 0.8505696058273315,
      "learning_rate": 1.5798711823126854e-05,
      "loss": 0.1842,
      "num_input_tokens_seen": 21475184,
      "step": 24800
    },
    {
      "epoch": 11.692597831211692,
      "eval_loss": 0.20011581480503082,
      "eval_runtime": 22.1542,
      "eval_samples_per_second": 42.565,
      "eval_steps_per_second": 21.305,
      "num_input_tokens_seen": 21475184,
      "step": 24800
    },
    {
      "epoch": 11.694955209806695,
      "grad_norm": 0.7676085829734802,
      "learning_rate": 1.578958418971477e-05,
      "loss": 0.1835,
      "num_input_tokens_seen": 21480320,
      "step": 24805
    },
    {
      "epoch": 11.697312588401697,
      "grad_norm": 1.241850733757019,
      "learning_rate": 1.578045797666453e-05,
      "loss": 0.3085,
      "num_input_tokens_seen": 21484800,
      "step": 24810
    },
    {
      "epoch": 11.6996699669967,
      "grad_norm": 3.3275585174560547,
      "learning_rate": 1.5771333185383548e-05,
      "loss": 0.2761,
      "num_input_tokens_seen": 21489104,
      "step": 24815
    },
    {
      "epoch": 11.702027345591702,
      "grad_norm": 0.6258659362792969,
      "learning_rate": 1.576220981727895e-05,
      "loss": 0.124,
      "num_input_tokens_seen": 21492928,
      "step": 24820
    },
    {
      "epoch": 11.704384724186705,
      "grad_norm": 0.5655291080474854,
      "learning_rate": 1.575308787375769e-05,
      "loss": 0.1405,
      "num_input_tokens_seen": 21497456,
      "step": 24825
    },
    {
      "epoch": 11.706742102781707,
      "grad_norm": 1.053939938545227,
      "learning_rate": 1.5743967356226492e-05,
      "loss": 0.1694,
      "num_input_tokens_seen": 21501440,
      "step": 24830
    },
    {
      "epoch": 11.70909948137671,
      "grad_norm": 0.9301292300224304,
      "learning_rate": 1.5734848266091835e-05,
      "loss": 0.1842,
      "num_input_tokens_seen": 21506640,
      "step": 24835
    },
    {
      "epoch": 11.711456859971712,
      "grad_norm": 0.7091415524482727,
      "learning_rate": 1.572573060476001e-05,
      "loss": 0.1511,
      "num_input_tokens_seen": 21511536,
      "step": 24840
    },
    {
      "epoch": 11.713814238566714,
      "grad_norm": 1.8150726556777954,
      "learning_rate": 1.5716614373637085e-05,
      "loss": 0.2479,
      "num_input_tokens_seen": 21516192,
      "step": 24845
    },
    {
      "epoch": 11.716171617161717,
      "grad_norm": 0.8772522211074829,
      "learning_rate": 1.570749957412887e-05,
      "loss": 0.1826,
      "num_input_tokens_seen": 21520032,
      "step": 24850
    },
    {
      "epoch": 11.718528995756719,
      "grad_norm": 0.35085195302963257,
      "learning_rate": 1.5698386207641013e-05,
      "loss": 0.111,
      "num_input_tokens_seen": 21524368,
      "step": 24855
    },
    {
      "epoch": 11.720886374351721,
      "grad_norm": 1.5129435062408447,
      "learning_rate": 1.5689274275578884e-05,
      "loss": 0.2036,
      "num_input_tokens_seen": 21528832,
      "step": 24860
    },
    {
      "epoch": 11.723243752946724,
      "grad_norm": 1.3805400133132935,
      "learning_rate": 1.5680163779347667e-05,
      "loss": 0.1667,
      "num_input_tokens_seen": 21533328,
      "step": 24865
    },
    {
      "epoch": 11.725601131541726,
      "grad_norm": 0.593344509601593,
      "learning_rate": 1.5671054720352327e-05,
      "loss": 0.1106,
      "num_input_tokens_seen": 21537072,
      "step": 24870
    },
    {
      "epoch": 11.727958510136729,
      "grad_norm": 0.6876255869865417,
      "learning_rate": 1.566194709999757e-05,
      "loss": 0.1469,
      "num_input_tokens_seen": 21541584,
      "step": 24875
    },
    {
      "epoch": 11.730315888731731,
      "grad_norm": 0.6965705752372742,
      "learning_rate": 1.5652840919687933e-05,
      "loss": 0.204,
      "num_input_tokens_seen": 21545728,
      "step": 24880
    },
    {
      "epoch": 11.732673267326733,
      "grad_norm": 0.4701138436794281,
      "learning_rate": 1.5643736180827676e-05,
      "loss": 0.2257,
      "num_input_tokens_seen": 21549536,
      "step": 24885
    },
    {
      "epoch": 11.735030645921736,
      "grad_norm": 1.7041505575180054,
      "learning_rate": 1.5634632884820878e-05,
      "loss": 0.1487,
      "num_input_tokens_seen": 21554736,
      "step": 24890
    },
    {
      "epoch": 11.737388024516738,
      "grad_norm": 0.5171929597854614,
      "learning_rate": 1.5625531033071395e-05,
      "loss": 0.1599,
      "num_input_tokens_seen": 21558768,
      "step": 24895
    },
    {
      "epoch": 11.73974540311174,
      "grad_norm": 0.6392348408699036,
      "learning_rate": 1.5616430626982828e-05,
      "loss": 0.1995,
      "num_input_tokens_seen": 21563232,
      "step": 24900
    },
    {
      "epoch": 11.742102781706743,
      "grad_norm": 0.3952465057373047,
      "learning_rate": 1.5607331667958575e-05,
      "loss": 0.1572,
      "num_input_tokens_seen": 21567040,
      "step": 24905
    },
    {
      "epoch": 11.744460160301745,
      "grad_norm": 0.6617560386657715,
      "learning_rate": 1.5598234157401824e-05,
      "loss": 0.1886,
      "num_input_tokens_seen": 21571392,
      "step": 24910
    },
    {
      "epoch": 11.746817538896746,
      "grad_norm": 0.4392480254173279,
      "learning_rate": 1.5589138096715503e-05,
      "loss": 0.1837,
      "num_input_tokens_seen": 21576016,
      "step": 24915
    },
    {
      "epoch": 11.749174917491748,
      "grad_norm": 0.9779781103134155,
      "learning_rate": 1.5580043487302365e-05,
      "loss": 0.2038,
      "num_input_tokens_seen": 21579920,
      "step": 24920
    },
    {
      "epoch": 11.75153229608675,
      "grad_norm": 1.2359203100204468,
      "learning_rate": 1.5570950330564888e-05,
      "loss": 0.2117,
      "num_input_tokens_seen": 21584592,
      "step": 24925
    },
    {
      "epoch": 11.753889674681753,
      "grad_norm": 1.5149723291397095,
      "learning_rate": 1.5561858627905367e-05,
      "loss": 0.1985,
      "num_input_tokens_seen": 21588944,
      "step": 24930
    },
    {
      "epoch": 11.756247053276756,
      "grad_norm": 1.5447208881378174,
      "learning_rate": 1.5552768380725857e-05,
      "loss": 0.1504,
      "num_input_tokens_seen": 21593808,
      "step": 24935
    },
    {
      "epoch": 11.758604431871758,
      "grad_norm": 0.6872416138648987,
      "learning_rate": 1.5543679590428183e-05,
      "loss": 0.1577,
      "num_input_tokens_seen": 21598560,
      "step": 24940
    },
    {
      "epoch": 11.76096181046676,
      "grad_norm": 1.077425479888916,
      "learning_rate": 1.5534592258413943e-05,
      "loss": 0.2007,
      "num_input_tokens_seen": 21602944,
      "step": 24945
    },
    {
      "epoch": 11.763319189061763,
      "grad_norm": 0.5222394466400146,
      "learning_rate": 1.5525506386084538e-05,
      "loss": 0.2456,
      "num_input_tokens_seen": 21607392,
      "step": 24950
    },
    {
      "epoch": 11.765676567656765,
      "grad_norm": 1.328148603439331,
      "learning_rate": 1.55164219748411e-05,
      "loss": 0.1595,
      "num_input_tokens_seen": 21611680,
      "step": 24955
    },
    {
      "epoch": 11.768033946251768,
      "grad_norm": 0.7522818446159363,
      "learning_rate": 1.550733902608459e-05,
      "loss": 0.1285,
      "num_input_tokens_seen": 21615632,
      "step": 24960
    },
    {
      "epoch": 11.77039132484677,
      "grad_norm": 0.7852062582969666,
      "learning_rate": 1.549825754121568e-05,
      "loss": 0.2719,
      "num_input_tokens_seen": 21619216,
      "step": 24965
    },
    {
      "epoch": 11.772748703441772,
      "grad_norm": 1.12738037109375,
      "learning_rate": 1.5489177521634864e-05,
      "loss": 0.2264,
      "num_input_tokens_seen": 21623808,
      "step": 24970
    },
    {
      "epoch": 11.775106082036775,
      "grad_norm": 0.7437605261802673,
      "learning_rate": 1.5480098968742402e-05,
      "loss": 0.1647,
      "num_input_tokens_seen": 21627888,
      "step": 24975
    },
    {
      "epoch": 11.777463460631777,
      "grad_norm": 0.4029112160205841,
      "learning_rate": 1.5471021883938304e-05,
      "loss": 0.1265,
      "num_input_tokens_seen": 21632256,
      "step": 24980
    },
    {
      "epoch": 11.77982083922678,
      "grad_norm": 0.7890406847000122,
      "learning_rate": 1.546194626862238e-05,
      "loss": 0.2317,
      "num_input_tokens_seen": 21636576,
      "step": 24985
    },
    {
      "epoch": 11.782178217821782,
      "grad_norm": 1.3006335496902466,
      "learning_rate": 1.5452872124194216e-05,
      "loss": 0.2626,
      "num_input_tokens_seen": 21640560,
      "step": 24990
    },
    {
      "epoch": 11.784535596416784,
      "grad_norm": 0.862337052822113,
      "learning_rate": 1.5443799452053136e-05,
      "loss": 0.0966,
      "num_input_tokens_seen": 21644864,
      "step": 24995
    },
    {
      "epoch": 11.786892975011787,
      "grad_norm": 1.1908963918685913,
      "learning_rate": 1.543472825359828e-05,
      "loss": 0.206,
      "num_input_tokens_seen": 21649744,
      "step": 25000
    },
    {
      "epoch": 11.786892975011787,
      "eval_loss": 0.19930247962474823,
      "eval_runtime": 22.1574,
      "eval_samples_per_second": 42.559,
      "eval_steps_per_second": 21.302,
      "num_input_tokens_seen": 21649744,
      "step": 25000
    },
    {
      "epoch": 11.78925035360679,
      "grad_norm": 1.7211741209030151,
      "learning_rate": 1.5425658530228522e-05,
      "loss": 0.2155,
      "num_input_tokens_seen": 21653408,
      "step": 25005
    },
    {
      "epoch": 11.791607732201792,
      "grad_norm": 1.169339895248413,
      "learning_rate": 1.5416590283342546e-05,
      "loss": 0.167,
      "num_input_tokens_seen": 21657904,
      "step": 25010
    },
    {
      "epoch": 11.793965110796794,
      "grad_norm": 1.2280045747756958,
      "learning_rate": 1.5407523514338783e-05,
      "loss": 0.1567,
      "num_input_tokens_seen": 21662016,
      "step": 25015
    },
    {
      "epoch": 11.796322489391796,
      "grad_norm": 0.22865420579910278,
      "learning_rate": 1.539845822461543e-05,
      "loss": 0.2099,
      "num_input_tokens_seen": 21665904,
      "step": 25020
    },
    {
      "epoch": 11.798679867986799,
      "grad_norm": 1.3040903806686401,
      "learning_rate": 1.538939441557048e-05,
      "loss": 0.2344,
      "num_input_tokens_seen": 21670480,
      "step": 25025
    },
    {
      "epoch": 11.801037246581801,
      "grad_norm": 1.281349778175354,
      "learning_rate": 1.5380332088601696e-05,
      "loss": 0.147,
      "num_input_tokens_seen": 21674928,
      "step": 25030
    },
    {
      "epoch": 11.803394625176804,
      "grad_norm": 0.642000138759613,
      "learning_rate": 1.537127124510658e-05,
      "loss": 0.2238,
      "num_input_tokens_seen": 21678784,
      "step": 25035
    },
    {
      "epoch": 11.805752003771806,
      "grad_norm": 0.5658220052719116,
      "learning_rate": 1.5362211886482457e-05,
      "loss": 0.1681,
      "num_input_tokens_seen": 21682832,
      "step": 25040
    },
    {
      "epoch": 11.808109382366808,
      "grad_norm": 1.0287429094314575,
      "learning_rate": 1.5353154014126363e-05,
      "loss": 0.146,
      "num_input_tokens_seen": 21687184,
      "step": 25045
    },
    {
      "epoch": 11.81046676096181,
      "grad_norm": 1.3547941446304321,
      "learning_rate": 1.534409762943515e-05,
      "loss": 0.1781,
      "num_input_tokens_seen": 21691424,
      "step": 25050
    },
    {
      "epoch": 11.812824139556813,
      "grad_norm": 1.3289297819137573,
      "learning_rate": 1.5335042733805438e-05,
      "loss": 0.2374,
      "num_input_tokens_seen": 21695584,
      "step": 25055
    },
    {
      "epoch": 11.815181518151816,
      "grad_norm": 0.5710821151733398,
      "learning_rate": 1.532598932863358e-05,
      "loss": 0.1427,
      "num_input_tokens_seen": 21699520,
      "step": 25060
    },
    {
      "epoch": 11.817538896746818,
      "grad_norm": 0.7949731945991516,
      "learning_rate": 1.531693741531574e-05,
      "loss": 0.2435,
      "num_input_tokens_seen": 21703520,
      "step": 25065
    },
    {
      "epoch": 11.81989627534182,
      "grad_norm": 1.8420014381408691,
      "learning_rate": 1.5307886995247844e-05,
      "loss": 0.1422,
      "num_input_tokens_seen": 21708512,
      "step": 25070
    },
    {
      "epoch": 11.822253653936823,
      "grad_norm": 0.5612462162971497,
      "learning_rate": 1.529883806982557e-05,
      "loss": 0.1903,
      "num_input_tokens_seen": 21712752,
      "step": 25075
    },
    {
      "epoch": 11.824611032531825,
      "grad_norm": 0.5872528553009033,
      "learning_rate": 1.5289790640444376e-05,
      "loss": 0.1211,
      "num_input_tokens_seen": 21717680,
      "step": 25080
    },
    {
      "epoch": 11.826968411126828,
      "grad_norm": 0.4322367012500763,
      "learning_rate": 1.5280744708499494e-05,
      "loss": 0.1512,
      "num_input_tokens_seen": 21721776,
      "step": 25085
    },
    {
      "epoch": 11.82932578972183,
      "grad_norm": 0.5251156091690063,
      "learning_rate": 1.527170027538591e-05,
      "loss": 0.1624,
      "num_input_tokens_seen": 21726112,
      "step": 25090
    },
    {
      "epoch": 11.831683168316832,
      "grad_norm": 0.8603505492210388,
      "learning_rate": 1.5262657342498407e-05,
      "loss": 0.1327,
      "num_input_tokens_seen": 21730416,
      "step": 25095
    },
    {
      "epoch": 11.834040546911835,
      "grad_norm": 0.7046492099761963,
      "learning_rate": 1.52536159112315e-05,
      "loss": 0.1633,
      "num_input_tokens_seen": 21734608,
      "step": 25100
    },
    {
      "epoch": 11.836397925506837,
      "grad_norm": 0.837164580821991,
      "learning_rate": 1.5244575982979497e-05,
      "loss": 0.2462,
      "num_input_tokens_seen": 21738752,
      "step": 25105
    },
    {
      "epoch": 11.838755304101838,
      "grad_norm": 1.2229236364364624,
      "learning_rate": 1.5235537559136487e-05,
      "loss": 0.1498,
      "num_input_tokens_seen": 21743168,
      "step": 25110
    },
    {
      "epoch": 11.841112682696842,
      "grad_norm": 1.3226056098937988,
      "learning_rate": 1.5226500641096286e-05,
      "loss": 0.1574,
      "num_input_tokens_seen": 21747824,
      "step": 25115
    },
    {
      "epoch": 11.843470061291843,
      "grad_norm": 1.1316486597061157,
      "learning_rate": 1.5217465230252509e-05,
      "loss": 0.1321,
      "num_input_tokens_seen": 21751952,
      "step": 25120
    },
    {
      "epoch": 11.845827439886845,
      "grad_norm": 0.6369245052337646,
      "learning_rate": 1.5208431327998523e-05,
      "loss": 0.0925,
      "num_input_tokens_seen": 21756544,
      "step": 25125
    },
    {
      "epoch": 11.848184818481847,
      "grad_norm": 1.3633733987808228,
      "learning_rate": 1.5199398935727477e-05,
      "loss": 0.2051,
      "num_input_tokens_seen": 21760368,
      "step": 25130
    },
    {
      "epoch": 11.85054219707685,
      "grad_norm": 1.666072964668274,
      "learning_rate": 1.5190368054832282e-05,
      "loss": 0.1322,
      "num_input_tokens_seen": 21765008,
      "step": 25135
    },
    {
      "epoch": 11.852899575671852,
      "grad_norm": 2.1883509159088135,
      "learning_rate": 1.5181338686705601e-05,
      "loss": 0.1942,
      "num_input_tokens_seen": 21769824,
      "step": 25140
    },
    {
      "epoch": 11.855256954266855,
      "grad_norm": 0.7355301976203918,
      "learning_rate": 1.5172310832739889e-05,
      "loss": 0.2247,
      "num_input_tokens_seen": 21773728,
      "step": 25145
    },
    {
      "epoch": 11.857614332861857,
      "grad_norm": 0.5590580105781555,
      "learning_rate": 1.5163284494327346e-05,
      "loss": 0.1074,
      "num_input_tokens_seen": 21778016,
      "step": 25150
    },
    {
      "epoch": 11.85997171145686,
      "grad_norm": 0.8964572548866272,
      "learning_rate": 1.5154259672859952e-05,
      "loss": 0.2911,
      "num_input_tokens_seen": 21782000,
      "step": 25155
    },
    {
      "epoch": 11.862329090051862,
      "grad_norm": 0.730681300163269,
      "learning_rate": 1.5145236369729452e-05,
      "loss": 0.153,
      "num_input_tokens_seen": 21785840,
      "step": 25160
    },
    {
      "epoch": 11.864686468646864,
      "grad_norm": 0.5873236656188965,
      "learning_rate": 1.5136214586327335e-05,
      "loss": 0.1585,
      "num_input_tokens_seen": 21789680,
      "step": 25165
    },
    {
      "epoch": 11.867043847241867,
      "grad_norm": 1.2751281261444092,
      "learning_rate": 1.5127194324044885e-05,
      "loss": 0.2228,
      "num_input_tokens_seen": 21794352,
      "step": 25170
    },
    {
      "epoch": 11.869401225836869,
      "grad_norm": 0.4132826328277588,
      "learning_rate": 1.5118175584273148e-05,
      "loss": 0.2098,
      "num_input_tokens_seen": 21798640,
      "step": 25175
    },
    {
      "epoch": 11.871758604431871,
      "grad_norm": 1.210775375366211,
      "learning_rate": 1.5109158368402909e-05,
      "loss": 0.1904,
      "num_input_tokens_seen": 21803024,
      "step": 25180
    },
    {
      "epoch": 11.874115983026874,
      "grad_norm": 0.8232323527336121,
      "learning_rate": 1.5100142677824753e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 21806640,
      "step": 25185
    },
    {
      "epoch": 11.876473361621876,
      "grad_norm": 1.1342964172363281,
      "learning_rate": 1.509112851392901e-05,
      "loss": 0.1893,
      "num_input_tokens_seen": 21811232,
      "step": 25190
    },
    {
      "epoch": 11.878830740216879,
      "grad_norm": 1.9056810140609741,
      "learning_rate": 1.5082115878105763e-05,
      "loss": 0.1745,
      "num_input_tokens_seen": 21815152,
      "step": 25195
    },
    {
      "epoch": 11.881188118811881,
      "grad_norm": 2.129314422607422,
      "learning_rate": 1.5073104771744892e-05,
      "loss": 0.2084,
      "num_input_tokens_seen": 21819728,
      "step": 25200
    },
    {
      "epoch": 11.881188118811881,
      "eval_loss": 0.20066215097904205,
      "eval_runtime": 22.1608,
      "eval_samples_per_second": 42.553,
      "eval_steps_per_second": 21.299,
      "num_input_tokens_seen": 21819728,
      "step": 25200
    },
    {
      "epoch": 11.883545497406883,
      "grad_norm": 0.88011634349823,
      "learning_rate": 1.5064095196236006e-05,
      "loss": 0.1442,
      "num_input_tokens_seen": 21823952,
      "step": 25205
    },
    {
      "epoch": 11.885902876001886,
      "grad_norm": 0.952643096446991,
      "learning_rate": 1.50550871529685e-05,
      "loss": 0.1791,
      "num_input_tokens_seen": 21827984,
      "step": 25210
    },
    {
      "epoch": 11.888260254596888,
      "grad_norm": 1.3876850605010986,
      "learning_rate": 1.5046080643331546e-05,
      "loss": 0.2026,
      "num_input_tokens_seen": 21832000,
      "step": 25215
    },
    {
      "epoch": 11.89061763319189,
      "grad_norm": 1.3601927757263184,
      "learning_rate": 1.5037075668714028e-05,
      "loss": 0.1509,
      "num_input_tokens_seen": 21836176,
      "step": 25220
    },
    {
      "epoch": 11.892975011786893,
      "grad_norm": 0.6418851017951965,
      "learning_rate": 1.5028072230504656e-05,
      "loss": 0.1801,
      "num_input_tokens_seen": 21840848,
      "step": 25225
    },
    {
      "epoch": 11.895332390381895,
      "grad_norm": 1.05406653881073,
      "learning_rate": 1.5019070330091861e-05,
      "loss": 0.1583,
      "num_input_tokens_seen": 21845344,
      "step": 25230
    },
    {
      "epoch": 11.897689768976898,
      "grad_norm": 0.6188300251960754,
      "learning_rate": 1.5010069968863843e-05,
      "loss": 0.1832,
      "num_input_tokens_seen": 21850208,
      "step": 25235
    },
    {
      "epoch": 11.9000471475719,
      "grad_norm": 0.6619958877563477,
      "learning_rate": 1.5001071148208584e-05,
      "loss": 0.167,
      "num_input_tokens_seen": 21854240,
      "step": 25240
    },
    {
      "epoch": 11.902404526166903,
      "grad_norm": 2.5078165531158447,
      "learning_rate": 1.49920738695138e-05,
      "loss": 0.1896,
      "num_input_tokens_seen": 21858960,
      "step": 25245
    },
    {
      "epoch": 11.904761904761905,
      "grad_norm": 0.6358357071876526,
      "learning_rate": 1.4983078134166995e-05,
      "loss": 0.2035,
      "num_input_tokens_seen": 21863472,
      "step": 25250
    },
    {
      "epoch": 11.907119283356908,
      "grad_norm": 0.6233124732971191,
      "learning_rate": 1.4974083943555428e-05,
      "loss": 0.1657,
      "num_input_tokens_seen": 21867632,
      "step": 25255
    },
    {
      "epoch": 11.90947666195191,
      "grad_norm": 1.105772614479065,
      "learning_rate": 1.496509129906611e-05,
      "loss": 0.1784,
      "num_input_tokens_seen": 21871760,
      "step": 25260
    },
    {
      "epoch": 11.911834040546912,
      "grad_norm": 1.8652076721191406,
      "learning_rate": 1.4956100202085809e-05,
      "loss": 0.1767,
      "num_input_tokens_seen": 21876096,
      "step": 25265
    },
    {
      "epoch": 11.914191419141915,
      "grad_norm": 0.44858813285827637,
      "learning_rate": 1.4947110654001093e-05,
      "loss": 0.1456,
      "num_input_tokens_seen": 21880048,
      "step": 25270
    },
    {
      "epoch": 11.916548797736917,
      "grad_norm": 0.5181200504302979,
      "learning_rate": 1.4938122656198234e-05,
      "loss": 0.1655,
      "num_input_tokens_seen": 21883696,
      "step": 25275
    },
    {
      "epoch": 11.91890617633192,
      "grad_norm": 1.4642839431762695,
      "learning_rate": 1.4929136210063316e-05,
      "loss": 0.1767,
      "num_input_tokens_seen": 21888368,
      "step": 25280
    },
    {
      "epoch": 11.921263554926922,
      "grad_norm": 0.653458833694458,
      "learning_rate": 1.4920151316982146e-05,
      "loss": 0.1603,
      "num_input_tokens_seen": 21892976,
      "step": 25285
    },
    {
      "epoch": 11.923620933521924,
      "grad_norm": 1.0529979467391968,
      "learning_rate": 1.4911167978340312e-05,
      "loss": 0.2155,
      "num_input_tokens_seen": 21896976,
      "step": 25290
    },
    {
      "epoch": 11.925978312116927,
      "grad_norm": 0.14023452997207642,
      "learning_rate": 1.4902186195523166e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 21900672,
      "step": 25295
    },
    {
      "epoch": 11.92833569071193,
      "grad_norm": 1.0239835977554321,
      "learning_rate": 1.4893205969915805e-05,
      "loss": 0.1967,
      "num_input_tokens_seen": 21904576,
      "step": 25300
    },
    {
      "epoch": 11.930693069306932,
      "grad_norm": 1.211143136024475,
      "learning_rate": 1.4884227302903086e-05,
      "loss": 0.21,
      "num_input_tokens_seen": 21908784,
      "step": 25305
    },
    {
      "epoch": 11.933050447901934,
      "grad_norm": 1.2037616968154907,
      "learning_rate": 1.4875250195869653e-05,
      "loss": 0.2238,
      "num_input_tokens_seen": 21913808,
      "step": 25310
    },
    {
      "epoch": 11.935407826496935,
      "grad_norm": 1.5099109411239624,
      "learning_rate": 1.4866274650199862e-05,
      "loss": 0.2188,
      "num_input_tokens_seen": 21917824,
      "step": 25315
    },
    {
      "epoch": 11.937765205091937,
      "grad_norm": 0.6399657130241394,
      "learning_rate": 1.485730066727788e-05,
      "loss": 0.2251,
      "num_input_tokens_seen": 21922464,
      "step": 25320
    },
    {
      "epoch": 11.94012258368694,
      "grad_norm": 1.6764863729476929,
      "learning_rate": 1.4848328248487586e-05,
      "loss": 0.2145,
      "num_input_tokens_seen": 21928352,
      "step": 25325
    },
    {
      "epoch": 11.942479962281942,
      "grad_norm": 0.6716168522834778,
      "learning_rate": 1.4839357395212656e-05,
      "loss": 0.1648,
      "num_input_tokens_seen": 21932704,
      "step": 25330
    },
    {
      "epoch": 11.944837340876944,
      "grad_norm": 0.6768725514411926,
      "learning_rate": 1.4830388108836502e-05,
      "loss": 0.1916,
      "num_input_tokens_seen": 21937104,
      "step": 25335
    },
    {
      "epoch": 11.947194719471947,
      "grad_norm": 0.551946222782135,
      "learning_rate": 1.4821420390742299e-05,
      "loss": 0.1505,
      "num_input_tokens_seen": 21940704,
      "step": 25340
    },
    {
      "epoch": 11.949552098066949,
      "grad_norm": 0.6754173040390015,
      "learning_rate": 1.4812454242312979e-05,
      "loss": 0.0988,
      "num_input_tokens_seen": 21944544,
      "step": 25345
    },
    {
      "epoch": 11.951909476661951,
      "grad_norm": 0.5255511999130249,
      "learning_rate": 1.4803489664931253e-05,
      "loss": 0.1884,
      "num_input_tokens_seen": 21948720,
      "step": 25350
    },
    {
      "epoch": 11.954266855256954,
      "grad_norm": 0.6352415084838867,
      "learning_rate": 1.4794526659979544e-05,
      "loss": 0.1611,
      "num_input_tokens_seen": 21953712,
      "step": 25355
    },
    {
      "epoch": 11.956624233851956,
      "grad_norm": 1.153688669204712,
      "learning_rate": 1.4785565228840086e-05,
      "loss": 0.229,
      "num_input_tokens_seen": 21957552,
      "step": 25360
    },
    {
      "epoch": 11.958981612446959,
      "grad_norm": 1.5078575611114502,
      "learning_rate": 1.4776605372894819e-05,
      "loss": 0.2968,
      "num_input_tokens_seen": 21962592,
      "step": 25365
    },
    {
      "epoch": 11.961338991041961,
      "grad_norm": 0.8268031477928162,
      "learning_rate": 1.4767647093525488e-05,
      "loss": 0.1531,
      "num_input_tokens_seen": 21966752,
      "step": 25370
    },
    {
      "epoch": 11.963696369636963,
      "grad_norm": 0.9706369042396545,
      "learning_rate": 1.4758690392113566e-05,
      "loss": 0.1548,
      "num_input_tokens_seen": 21971136,
      "step": 25375
    },
    {
      "epoch": 11.966053748231966,
      "grad_norm": 1.8309377431869507,
      "learning_rate": 1.4749735270040276e-05,
      "loss": 0.2113,
      "num_input_tokens_seen": 21976112,
      "step": 25380
    },
    {
      "epoch": 11.968411126826968,
      "grad_norm": 0.6367422342300415,
      "learning_rate": 1.4740781728686623e-05,
      "loss": 0.2537,
      "num_input_tokens_seen": 21980672,
      "step": 25385
    },
    {
      "epoch": 11.97076850542197,
      "grad_norm": 0.5494503974914551,
      "learning_rate": 1.4731829769433358e-05,
      "loss": 0.1527,
      "num_input_tokens_seen": 21985824,
      "step": 25390
    },
    {
      "epoch": 11.973125884016973,
      "grad_norm": 0.9752506017684937,
      "learning_rate": 1.4722879393660976e-05,
      "loss": 0.1914,
      "num_input_tokens_seen": 21989232,
      "step": 25395
    },
    {
      "epoch": 11.975483262611975,
      "grad_norm": 2.2120490074157715,
      "learning_rate": 1.4713930602749748e-05,
      "loss": 0.168,
      "num_input_tokens_seen": 21993120,
      "step": 25400
    },
    {
      "epoch": 11.975483262611975,
      "eval_loss": 0.2004028707742691,
      "eval_runtime": 22.1416,
      "eval_samples_per_second": 42.589,
      "eval_steps_per_second": 21.317,
      "num_input_tokens_seen": 21993120,
      "step": 25400
    },
    {
      "epoch": 11.977840641206978,
      "grad_norm": 0.7426704168319702,
      "learning_rate": 1.470498339807968e-05,
      "loss": 0.1643,
      "num_input_tokens_seen": 21997584,
      "step": 25405
    },
    {
      "epoch": 11.98019801980198,
      "grad_norm": 1.3647993803024292,
      "learning_rate": 1.4696037781030542e-05,
      "loss": 0.1346,
      "num_input_tokens_seen": 22001488,
      "step": 25410
    },
    {
      "epoch": 11.982555398396983,
      "grad_norm": 1.4564167261123657,
      "learning_rate": 1.4687093752981876e-05,
      "loss": 0.2013,
      "num_input_tokens_seen": 22006192,
      "step": 25415
    },
    {
      "epoch": 11.984912776991985,
      "grad_norm": 0.744270920753479,
      "learning_rate": 1.4678151315312943e-05,
      "loss": 0.1517,
      "num_input_tokens_seen": 22010176,
      "step": 25420
    },
    {
      "epoch": 11.987270155586987,
      "grad_norm": 0.7691128849983215,
      "learning_rate": 1.4669210469402789e-05,
      "loss": 0.2063,
      "num_input_tokens_seen": 22013968,
      "step": 25425
    },
    {
      "epoch": 11.98962753418199,
      "grad_norm": 1.6371033191680908,
      "learning_rate": 1.4660271216630218e-05,
      "loss": 0.1757,
      "num_input_tokens_seen": 22017920,
      "step": 25430
    },
    {
      "epoch": 11.991984912776992,
      "grad_norm": 1.2933037281036377,
      "learning_rate": 1.4651333558373748e-05,
      "loss": 0.2564,
      "num_input_tokens_seen": 22021968,
      "step": 25435
    },
    {
      "epoch": 11.994342291371995,
      "grad_norm": 0.8212460279464722,
      "learning_rate": 1.4642397496011707e-05,
      "loss": 0.1085,
      "num_input_tokens_seen": 22026832,
      "step": 25440
    },
    {
      "epoch": 11.996699669966997,
      "grad_norm": 1.5355819463729858,
      "learning_rate": 1.4633463030922129e-05,
      "loss": 0.1567,
      "num_input_tokens_seen": 22031008,
      "step": 25445
    },
    {
      "epoch": 11.999057048562,
      "grad_norm": 0.33401724696159363,
      "learning_rate": 1.462453016448282e-05,
      "loss": 0.0774,
      "num_input_tokens_seen": 22034768,
      "step": 25450
    },
    {
      "epoch": 12.001414427157002,
      "grad_norm": 0.7942261099815369,
      "learning_rate": 1.4615598898071354e-05,
      "loss": 0.1718,
      "num_input_tokens_seen": 22039216,
      "step": 25455
    },
    {
      "epoch": 12.003771805752004,
      "grad_norm": 1.0640591382980347,
      "learning_rate": 1.4606669233065026e-05,
      "loss": 0.2144,
      "num_input_tokens_seen": 22043216,
      "step": 25460
    },
    {
      "epoch": 12.006129184347007,
      "grad_norm": 0.24409857392311096,
      "learning_rate": 1.4597741170840914e-05,
      "loss": 0.1291,
      "num_input_tokens_seen": 22047968,
      "step": 25465
    },
    {
      "epoch": 12.008486562942009,
      "grad_norm": 1.3333473205566406,
      "learning_rate": 1.4588814712775853e-05,
      "loss": 0.1389,
      "num_input_tokens_seen": 22052000,
      "step": 25470
    },
    {
      "epoch": 12.010843941537011,
      "grad_norm": 0.8512163758277893,
      "learning_rate": 1.4579889860246382e-05,
      "loss": 0.1523,
      "num_input_tokens_seen": 22055824,
      "step": 25475
    },
    {
      "epoch": 12.013201320132014,
      "grad_norm": 0.6941679120063782,
      "learning_rate": 1.457096661462885e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 22060128,
      "step": 25480
    },
    {
      "epoch": 12.015558698727016,
      "grad_norm": 0.5479028820991516,
      "learning_rate": 1.4562044977299322e-05,
      "loss": 0.1951,
      "num_input_tokens_seen": 22065184,
      "step": 25485
    },
    {
      "epoch": 12.017916077322019,
      "grad_norm": 0.6695169806480408,
      "learning_rate": 1.4553124949633623e-05,
      "loss": 0.1756,
      "num_input_tokens_seen": 22068912,
      "step": 25490
    },
    {
      "epoch": 12.020273455917021,
      "grad_norm": 0.552466094493866,
      "learning_rate": 1.4544206533007354e-05,
      "loss": 0.2676,
      "num_input_tokens_seen": 22072992,
      "step": 25495
    },
    {
      "epoch": 12.022630834512023,
      "grad_norm": 0.7879061102867126,
      "learning_rate": 1.4535289728795821e-05,
      "loss": 0.2236,
      "num_input_tokens_seen": 22077312,
      "step": 25500
    },
    {
      "epoch": 12.024988213107026,
      "grad_norm": 1.6733018159866333,
      "learning_rate": 1.4526374538374132e-05,
      "loss": 0.1987,
      "num_input_tokens_seen": 22081488,
      "step": 25505
    },
    {
      "epoch": 12.027345591702028,
      "grad_norm": 0.552544891834259,
      "learning_rate": 1.4517460963117097e-05,
      "loss": 0.1805,
      "num_input_tokens_seen": 22085136,
      "step": 25510
    },
    {
      "epoch": 12.029702970297029,
      "grad_norm": 0.3365324139595032,
      "learning_rate": 1.4508549004399314e-05,
      "loss": 0.1066,
      "num_input_tokens_seen": 22089520,
      "step": 25515
    },
    {
      "epoch": 12.032060348892031,
      "grad_norm": 0.8499159216880798,
      "learning_rate": 1.449963866359513e-05,
      "loss": 0.203,
      "num_input_tokens_seen": 22094560,
      "step": 25520
    },
    {
      "epoch": 12.034417727487034,
      "grad_norm": 0.7022693753242493,
      "learning_rate": 1.4490729942078607e-05,
      "loss": 0.1867,
      "num_input_tokens_seen": 22098560,
      "step": 25525
    },
    {
      "epoch": 12.036775106082036,
      "grad_norm": 1.0279390811920166,
      "learning_rate": 1.4481822841223608e-05,
      "loss": 0.2001,
      "num_input_tokens_seen": 22103328,
      "step": 25530
    },
    {
      "epoch": 12.039132484677038,
      "grad_norm": 1.215786099433899,
      "learning_rate": 1.4472917362403704e-05,
      "loss": 0.1646,
      "num_input_tokens_seen": 22106880,
      "step": 25535
    },
    {
      "epoch": 12.04148986327204,
      "grad_norm": 0.8220831155776978,
      "learning_rate": 1.4464013506992224e-05,
      "loss": 0.1927,
      "num_input_tokens_seen": 22111536,
      "step": 25540
    },
    {
      "epoch": 12.043847241867043,
      "grad_norm": 0.7002930045127869,
      "learning_rate": 1.4455111276362277e-05,
      "loss": 0.1556,
      "num_input_tokens_seen": 22115968,
      "step": 25545
    },
    {
      "epoch": 12.046204620462046,
      "grad_norm": 0.49308857321739197,
      "learning_rate": 1.4446210671886676e-05,
      "loss": 0.1676,
      "num_input_tokens_seen": 22119488,
      "step": 25550
    },
    {
      "epoch": 12.048561999057048,
      "grad_norm": 0.7541100382804871,
      "learning_rate": 1.4437311694938015e-05,
      "loss": 0.2118,
      "num_input_tokens_seen": 22123088,
      "step": 25555
    },
    {
      "epoch": 12.05091937765205,
      "grad_norm": 0.5564988255500793,
      "learning_rate": 1.442841434688864e-05,
      "loss": 0.1938,
      "num_input_tokens_seen": 22128080,
      "step": 25560
    },
    {
      "epoch": 12.053276756247053,
      "grad_norm": 1.1664706468582153,
      "learning_rate": 1.4419518629110615e-05,
      "loss": 0.1465,
      "num_input_tokens_seen": 22132560,
      "step": 25565
    },
    {
      "epoch": 12.055634134842055,
      "grad_norm": 0.9569830298423767,
      "learning_rate": 1.4410624542975778e-05,
      "loss": 0.1891,
      "num_input_tokens_seen": 22137648,
      "step": 25570
    },
    {
      "epoch": 12.057991513437058,
      "grad_norm": 0.7839100956916809,
      "learning_rate": 1.4401732089855724e-05,
      "loss": 0.2766,
      "num_input_tokens_seen": 22141504,
      "step": 25575
    },
    {
      "epoch": 12.06034889203206,
      "grad_norm": 2.305680274963379,
      "learning_rate": 1.4392841271121754e-05,
      "loss": 0.2318,
      "num_input_tokens_seen": 22145664,
      "step": 25580
    },
    {
      "epoch": 12.062706270627062,
      "grad_norm": 1.0086543560028076,
      "learning_rate": 1.438395208814497e-05,
      "loss": 0.1437,
      "num_input_tokens_seen": 22150448,
      "step": 25585
    },
    {
      "epoch": 12.065063649222065,
      "grad_norm": 0.5419625043869019,
      "learning_rate": 1.4375064542296174e-05,
      "loss": 0.1662,
      "num_input_tokens_seen": 22156480,
      "step": 25590
    },
    {
      "epoch": 12.067421027817067,
      "grad_norm": 1.6759178638458252,
      "learning_rate": 1.4366178634945946e-05,
      "loss": 0.1954,
      "num_input_tokens_seen": 22160016,
      "step": 25595
    },
    {
      "epoch": 12.06977840641207,
      "grad_norm": 0.9937723875045776,
      "learning_rate": 1.4357294367464616e-05,
      "loss": 0.1637,
      "num_input_tokens_seen": 22164624,
      "step": 25600
    },
    {
      "epoch": 12.06977840641207,
      "eval_loss": 0.2008320540189743,
      "eval_runtime": 22.1466,
      "eval_samples_per_second": 42.58,
      "eval_steps_per_second": 21.313,
      "num_input_tokens_seen": 22164624,
      "step": 25600
    },
    {
      "epoch": 12.072135785007072,
      "grad_norm": 0.59818035364151,
      "learning_rate": 1.434841174122224e-05,
      "loss": 0.1075,
      "num_input_tokens_seen": 22168592,
      "step": 25605
    },
    {
      "epoch": 12.074493163602074,
      "grad_norm": 1.119001865386963,
      "learning_rate": 1.4339530757588615e-05,
      "loss": 0.1235,
      "num_input_tokens_seen": 22172672,
      "step": 25610
    },
    {
      "epoch": 12.076850542197077,
      "grad_norm": 1.0913091897964478,
      "learning_rate": 1.433065141793333e-05,
      "loss": 0.2485,
      "num_input_tokens_seen": 22177136,
      "step": 25615
    },
    {
      "epoch": 12.07920792079208,
      "grad_norm": 1.0881770849227905,
      "learning_rate": 1.4321773723625665e-05,
      "loss": 0.1438,
      "num_input_tokens_seen": 22181696,
      "step": 25620
    },
    {
      "epoch": 12.081565299387082,
      "grad_norm": 0.5263729095458984,
      "learning_rate": 1.4312897676034693e-05,
      "loss": 0.1088,
      "num_input_tokens_seen": 22186032,
      "step": 25625
    },
    {
      "epoch": 12.083922677982084,
      "grad_norm": 0.40870076417922974,
      "learning_rate": 1.4304023276529188e-05,
      "loss": 0.2389,
      "num_input_tokens_seen": 22190736,
      "step": 25630
    },
    {
      "epoch": 12.086280056577086,
      "grad_norm": 0.7050721645355225,
      "learning_rate": 1.4295150526477712e-05,
      "loss": 0.2199,
      "num_input_tokens_seen": 22195280,
      "step": 25635
    },
    {
      "epoch": 12.088637435172089,
      "grad_norm": 0.3449245095252991,
      "learning_rate": 1.4286279427248562e-05,
      "loss": 0.1336,
      "num_input_tokens_seen": 22198944,
      "step": 25640
    },
    {
      "epoch": 12.090994813767091,
      "grad_norm": 1.1664551496505737,
      "learning_rate": 1.4277409980209747e-05,
      "loss": 0.17,
      "num_input_tokens_seen": 22203264,
      "step": 25645
    },
    {
      "epoch": 12.093352192362094,
      "grad_norm": 1.07377290725708,
      "learning_rate": 1.4268542186729061e-05,
      "loss": 0.146,
      "num_input_tokens_seen": 22208016,
      "step": 25650
    },
    {
      "epoch": 12.095709570957096,
      "grad_norm": 0.5111498832702637,
      "learning_rate": 1.4259676048174043e-05,
      "loss": 0.1437,
      "num_input_tokens_seen": 22212304,
      "step": 25655
    },
    {
      "epoch": 12.098066949552098,
      "grad_norm": 1.1881178617477417,
      "learning_rate": 1.4250811565911937e-05,
      "loss": 0.1543,
      "num_input_tokens_seen": 22216672,
      "step": 25660
    },
    {
      "epoch": 12.100424328147101,
      "grad_norm": 0.3342926800251007,
      "learning_rate": 1.4241948741309782e-05,
      "loss": 0.1736,
      "num_input_tokens_seen": 22221920,
      "step": 25665
    },
    {
      "epoch": 12.102781706742103,
      "grad_norm": 0.59037846326828,
      "learning_rate": 1.4233087575734317e-05,
      "loss": 0.1772,
      "num_input_tokens_seen": 22226640,
      "step": 25670
    },
    {
      "epoch": 12.105139085337106,
      "grad_norm": 0.8623335957527161,
      "learning_rate": 1.422422807055206e-05,
      "loss": 0.1512,
      "num_input_tokens_seen": 22230672,
      "step": 25675
    },
    {
      "epoch": 12.107496463932108,
      "grad_norm": 0.8561792373657227,
      "learning_rate": 1.4215370227129243e-05,
      "loss": 0.2397,
      "num_input_tokens_seen": 22235568,
      "step": 25680
    },
    {
      "epoch": 12.10985384252711,
      "grad_norm": 0.8982670307159424,
      "learning_rate": 1.4206514046831876e-05,
      "loss": 0.1812,
      "num_input_tokens_seen": 22239888,
      "step": 25685
    },
    {
      "epoch": 12.112211221122113,
      "grad_norm": 0.5272141695022583,
      "learning_rate": 1.419765953102567e-05,
      "loss": 0.1191,
      "num_input_tokens_seen": 22243696,
      "step": 25690
    },
    {
      "epoch": 12.114568599717115,
      "grad_norm": 1.5940958261489868,
      "learning_rate": 1.4188806681076125e-05,
      "loss": 0.1336,
      "num_input_tokens_seen": 22248128,
      "step": 25695
    },
    {
      "epoch": 12.116925978312118,
      "grad_norm": 0.3026042580604553,
      "learning_rate": 1.4179955498348443e-05,
      "loss": 0.184,
      "num_input_tokens_seen": 22251664,
      "step": 25700
    },
    {
      "epoch": 12.11928335690712,
      "grad_norm": 0.674284040927887,
      "learning_rate": 1.4171105984207605e-05,
      "loss": 0.1228,
      "num_input_tokens_seen": 22256064,
      "step": 25705
    },
    {
      "epoch": 12.121640735502123,
      "grad_norm": 0.47048479318618774,
      "learning_rate": 1.4162258140018304e-05,
      "loss": 0.2425,
      "num_input_tokens_seen": 22261136,
      "step": 25710
    },
    {
      "epoch": 12.123998114097123,
      "grad_norm": 1.1523524522781372,
      "learning_rate": 1.4153411967144986e-05,
      "loss": 0.1388,
      "num_input_tokens_seen": 22265840,
      "step": 25715
    },
    {
      "epoch": 12.126355492692126,
      "grad_norm": 1.2074421644210815,
      "learning_rate": 1.4144567466951864e-05,
      "loss": 0.1913,
      "num_input_tokens_seen": 22270192,
      "step": 25720
    },
    {
      "epoch": 12.128712871287128,
      "grad_norm": 1.2017170190811157,
      "learning_rate": 1.4135724640802844e-05,
      "loss": 0.1463,
      "num_input_tokens_seen": 22274272,
      "step": 25725
    },
    {
      "epoch": 12.13107024988213,
      "grad_norm": 0.6006807088851929,
      "learning_rate": 1.4126883490061615e-05,
      "loss": 0.2013,
      "num_input_tokens_seen": 22278976,
      "step": 25730
    },
    {
      "epoch": 12.133427628477133,
      "grad_norm": 0.9135948419570923,
      "learning_rate": 1.4118044016091603e-05,
      "loss": 0.2395,
      "num_input_tokens_seen": 22283728,
      "step": 25735
    },
    {
      "epoch": 12.135785007072135,
      "grad_norm": 0.46686574816703796,
      "learning_rate": 1.410920622025594e-05,
      "loss": 0.1466,
      "num_input_tokens_seen": 22287760,
      "step": 25740
    },
    {
      "epoch": 12.138142385667138,
      "grad_norm": 0.8879020810127258,
      "learning_rate": 1.4100370103917554e-05,
      "loss": 0.1864,
      "num_input_tokens_seen": 22291472,
      "step": 25745
    },
    {
      "epoch": 12.14049976426214,
      "grad_norm": 0.44238701462745667,
      "learning_rate": 1.409153566843907e-05,
      "loss": 0.256,
      "num_input_tokens_seen": 22295792,
      "step": 25750
    },
    {
      "epoch": 12.142857142857142,
      "grad_norm": 1.2935130596160889,
      "learning_rate": 1.408270291518286e-05,
      "loss": 0.2253,
      "num_input_tokens_seen": 22300256,
      "step": 25755
    },
    {
      "epoch": 12.145214521452145,
      "grad_norm": 0.9666107296943665,
      "learning_rate": 1.407387184551107e-05,
      "loss": 0.2243,
      "num_input_tokens_seen": 22304400,
      "step": 25760
    },
    {
      "epoch": 12.147571900047147,
      "grad_norm": 1.0419071912765503,
      "learning_rate": 1.4065042460785532e-05,
      "loss": 0.1509,
      "num_input_tokens_seen": 22308864,
      "step": 25765
    },
    {
      "epoch": 12.14992927864215,
      "grad_norm": 2.5626211166381836,
      "learning_rate": 1.405621476236787e-05,
      "loss": 0.1319,
      "num_input_tokens_seen": 22313200,
      "step": 25770
    },
    {
      "epoch": 12.152286657237152,
      "grad_norm": 0.76249098777771,
      "learning_rate": 1.4047388751619423e-05,
      "loss": 0.1982,
      "num_input_tokens_seen": 22317536,
      "step": 25775
    },
    {
      "epoch": 12.154644035832154,
      "grad_norm": 1.144867181777954,
      "learning_rate": 1.4038564429901264e-05,
      "loss": 0.2774,
      "num_input_tokens_seen": 22322208,
      "step": 25780
    },
    {
      "epoch": 12.157001414427157,
      "grad_norm": 1.7563810348510742,
      "learning_rate": 1.4029741798574227e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 22327072,
      "step": 25785
    },
    {
      "epoch": 12.15935879302216,
      "grad_norm": 0.5894182920455933,
      "learning_rate": 1.402092085899886e-05,
      "loss": 0.1714,
      "num_input_tokens_seen": 22331024,
      "step": 25790
    },
    {
      "epoch": 12.161716171617162,
      "grad_norm": 0.9997113347053528,
      "learning_rate": 1.4012101612535464e-05,
      "loss": 0.2224,
      "num_input_tokens_seen": 22335088,
      "step": 25795
    },
    {
      "epoch": 12.164073550212164,
      "grad_norm": 1.1725783348083496,
      "learning_rate": 1.4003284060544092e-05,
      "loss": 0.2579,
      "num_input_tokens_seen": 22340064,
      "step": 25800
    },
    {
      "epoch": 12.164073550212164,
      "eval_loss": 0.20141737163066864,
      "eval_runtime": 22.1483,
      "eval_samples_per_second": 42.577,
      "eval_steps_per_second": 21.311,
      "num_input_tokens_seen": 22340064,
      "step": 25800
    },
    {
      "epoch": 12.166430928807166,
      "grad_norm": 0.7874097228050232,
      "learning_rate": 1.3994468204384504e-05,
      "loss": 0.1573,
      "num_input_tokens_seen": 22344432,
      "step": 25805
    },
    {
      "epoch": 12.168788307402169,
      "grad_norm": 0.7498232126235962,
      "learning_rate": 1.398565404541622e-05,
      "loss": 0.1714,
      "num_input_tokens_seen": 22349040,
      "step": 25810
    },
    {
      "epoch": 12.171145685997171,
      "grad_norm": 0.9429686069488525,
      "learning_rate": 1.3976841584998513e-05,
      "loss": 0.1442,
      "num_input_tokens_seen": 22352928,
      "step": 25815
    },
    {
      "epoch": 12.173503064592174,
      "grad_norm": 0.7538814544677734,
      "learning_rate": 1.3968030824490352e-05,
      "loss": 0.1324,
      "num_input_tokens_seen": 22357312,
      "step": 25820
    },
    {
      "epoch": 12.175860443187176,
      "grad_norm": 0.9142833948135376,
      "learning_rate": 1.3959221765250469e-05,
      "loss": 0.1795,
      "num_input_tokens_seen": 22361824,
      "step": 25825
    },
    {
      "epoch": 12.178217821782178,
      "grad_norm": 1.3573241233825684,
      "learning_rate": 1.3950414408637343e-05,
      "loss": 0.2408,
      "num_input_tokens_seen": 22365696,
      "step": 25830
    },
    {
      "epoch": 12.18057520037718,
      "grad_norm": 0.5236617922782898,
      "learning_rate": 1.3941608756009166e-05,
      "loss": 0.1199,
      "num_input_tokens_seen": 22370816,
      "step": 25835
    },
    {
      "epoch": 12.182932578972183,
      "grad_norm": 0.6949323415756226,
      "learning_rate": 1.3932804808723898e-05,
      "loss": 0.1443,
      "num_input_tokens_seen": 22375936,
      "step": 25840
    },
    {
      "epoch": 12.185289957567186,
      "grad_norm": 1.9223191738128662,
      "learning_rate": 1.3924002568139194e-05,
      "loss": 0.1646,
      "num_input_tokens_seen": 22379936,
      "step": 25845
    },
    {
      "epoch": 12.187647336162188,
      "grad_norm": 1.172307014465332,
      "learning_rate": 1.3915202035612485e-05,
      "loss": 0.2115,
      "num_input_tokens_seen": 22385072,
      "step": 25850
    },
    {
      "epoch": 12.19000471475719,
      "grad_norm": 0.9116866588592529,
      "learning_rate": 1.3906403212500935e-05,
      "loss": 0.2023,
      "num_input_tokens_seen": 22389072,
      "step": 25855
    },
    {
      "epoch": 12.192362093352193,
      "grad_norm": 1.1890931129455566,
      "learning_rate": 1.3897606100161409e-05,
      "loss": 0.1782,
      "num_input_tokens_seen": 22393424,
      "step": 25860
    },
    {
      "epoch": 12.194719471947195,
      "grad_norm": 0.9093512296676636,
      "learning_rate": 1.388881069995055e-05,
      "loss": 0.1851,
      "num_input_tokens_seen": 22398112,
      "step": 25865
    },
    {
      "epoch": 12.197076850542198,
      "grad_norm": 0.5931593775749207,
      "learning_rate": 1.3880017013224708e-05,
      "loss": 0.1373,
      "num_input_tokens_seen": 22403456,
      "step": 25870
    },
    {
      "epoch": 12.1994342291372,
      "grad_norm": 0.8039612174034119,
      "learning_rate": 1.3871225041339984e-05,
      "loss": 0.112,
      "num_input_tokens_seen": 22406992,
      "step": 25875
    },
    {
      "epoch": 12.201791607732202,
      "grad_norm": 0.5766472220420837,
      "learning_rate": 1.386243478565222e-05,
      "loss": 0.1606,
      "num_input_tokens_seen": 22411184,
      "step": 25880
    },
    {
      "epoch": 12.204148986327205,
      "grad_norm": 0.7567035555839539,
      "learning_rate": 1.3853646247516966e-05,
      "loss": 0.2034,
      "num_input_tokens_seen": 22416688,
      "step": 25885
    },
    {
      "epoch": 12.206506364922207,
      "grad_norm": 1.7855224609375,
      "learning_rate": 1.3844859428289545e-05,
      "loss": 0.2348,
      "num_input_tokens_seen": 22421536,
      "step": 25890
    },
    {
      "epoch": 12.20886374351721,
      "grad_norm": 0.9825354814529419,
      "learning_rate": 1.3836074329324984e-05,
      "loss": 0.1426,
      "num_input_tokens_seen": 22426624,
      "step": 25895
    },
    {
      "epoch": 12.211221122112212,
      "grad_norm": 1.099150538444519,
      "learning_rate": 1.3827290951978044e-05,
      "loss": 0.1892,
      "num_input_tokens_seen": 22430448,
      "step": 25900
    },
    {
      "epoch": 12.213578500707214,
      "grad_norm": 0.5664853453636169,
      "learning_rate": 1.381850929760326e-05,
      "loss": 0.1503,
      "num_input_tokens_seen": 22434592,
      "step": 25905
    },
    {
      "epoch": 12.215935879302217,
      "grad_norm": 0.5254608392715454,
      "learning_rate": 1.3809729367554842e-05,
      "loss": 0.1741,
      "num_input_tokens_seen": 22438816,
      "step": 25910
    },
    {
      "epoch": 12.218293257897217,
      "grad_norm": 0.6634894609451294,
      "learning_rate": 1.3800951163186784e-05,
      "loss": 0.1231,
      "num_input_tokens_seen": 22442848,
      "step": 25915
    },
    {
      "epoch": 12.22065063649222,
      "grad_norm": 0.7434160709381104,
      "learning_rate": 1.3792174685852801e-05,
      "loss": 0.2582,
      "num_input_tokens_seen": 22446768,
      "step": 25920
    },
    {
      "epoch": 12.223008015087222,
      "grad_norm": 0.638166606426239,
      "learning_rate": 1.378339993690632e-05,
      "loss": 0.169,
      "num_input_tokens_seen": 22451600,
      "step": 25925
    },
    {
      "epoch": 12.225365393682225,
      "grad_norm": 1.500511646270752,
      "learning_rate": 1.3774626917700523e-05,
      "loss": 0.178,
      "num_input_tokens_seen": 22455504,
      "step": 25930
    },
    {
      "epoch": 12.227722772277227,
      "grad_norm": 1.0265341997146606,
      "learning_rate": 1.3765855629588334e-05,
      "loss": 0.139,
      "num_input_tokens_seen": 22458944,
      "step": 25935
    },
    {
      "epoch": 12.23008015087223,
      "grad_norm": 0.8997228145599365,
      "learning_rate": 1.3757086073922374e-05,
      "loss": 0.1952,
      "num_input_tokens_seen": 22462560,
      "step": 25940
    },
    {
      "epoch": 12.232437529467232,
      "grad_norm": 1.184867262840271,
      "learning_rate": 1.3748318252055038e-05,
      "loss": 0.2383,
      "num_input_tokens_seen": 22466976,
      "step": 25945
    },
    {
      "epoch": 12.234794908062234,
      "grad_norm": 0.8383160829544067,
      "learning_rate": 1.3739552165338416e-05,
      "loss": 0.1694,
      "num_input_tokens_seen": 22471664,
      "step": 25950
    },
    {
      "epoch": 12.237152286657237,
      "grad_norm": 1.5361825227737427,
      "learning_rate": 1.3730787815124354e-05,
      "loss": 0.1733,
      "num_input_tokens_seen": 22476448,
      "step": 25955
    },
    {
      "epoch": 12.239509665252239,
      "grad_norm": 1.016868233680725,
      "learning_rate": 1.3722025202764443e-05,
      "loss": 0.1893,
      "num_input_tokens_seen": 22480960,
      "step": 25960
    },
    {
      "epoch": 12.241867043847241,
      "grad_norm": 1.277968168258667,
      "learning_rate": 1.371326432960997e-05,
      "loss": 0.1277,
      "num_input_tokens_seen": 22484784,
      "step": 25965
    },
    {
      "epoch": 12.244224422442244,
      "grad_norm": 0.5368862152099609,
      "learning_rate": 1.3704505197011969e-05,
      "loss": 0.2058,
      "num_input_tokens_seen": 22489088,
      "step": 25970
    },
    {
      "epoch": 12.246581801037246,
      "grad_norm": 1.0961153507232666,
      "learning_rate": 1.3695747806321224e-05,
      "loss": 0.1728,
      "num_input_tokens_seen": 22493520,
      "step": 25975
    },
    {
      "epoch": 12.248939179632249,
      "grad_norm": 1.2784613370895386,
      "learning_rate": 1.3686992158888212e-05,
      "loss": 0.1181,
      "num_input_tokens_seen": 22497248,
      "step": 25980
    },
    {
      "epoch": 12.251296558227251,
      "grad_norm": 1.8432320356369019,
      "learning_rate": 1.367823825606319e-05,
      "loss": 0.223,
      "num_input_tokens_seen": 22502880,
      "step": 25985
    },
    {
      "epoch": 12.253653936822253,
      "grad_norm": 1.549390196800232,
      "learning_rate": 1.36694860991961e-05,
      "loss": 0.1722,
      "num_input_tokens_seen": 22507184,
      "step": 25990
    },
    {
      "epoch": 12.256011315417256,
      "grad_norm": 1.0215636491775513,
      "learning_rate": 1.3660735689636636e-05,
      "loss": 0.2023,
      "num_input_tokens_seen": 22511520,
      "step": 25995
    },
    {
      "epoch": 12.258368694012258,
      "grad_norm": 0.6906536817550659,
      "learning_rate": 1.365198702873424e-05,
      "loss": 0.1575,
      "num_input_tokens_seen": 22515088,
      "step": 26000
    },
    {
      "epoch": 12.258368694012258,
      "eval_loss": 0.20272080600261688,
      "eval_runtime": 22.1678,
      "eval_samples_per_second": 42.539,
      "eval_steps_per_second": 21.292,
      "num_input_tokens_seen": 22515088,
      "step": 26000
    },
    {
      "epoch": 12.26072607260726,
      "grad_norm": 0.4573255777359009,
      "learning_rate": 1.364324011783804e-05,
      "loss": 0.144,
      "num_input_tokens_seen": 22519088,
      "step": 26005
    },
    {
      "epoch": 12.263083451202263,
      "grad_norm": 0.7444007396697998,
      "learning_rate": 1.3634494958296934e-05,
      "loss": 0.1929,
      "num_input_tokens_seen": 22524144,
      "step": 26010
    },
    {
      "epoch": 12.265440829797265,
      "grad_norm": 1.1705166101455688,
      "learning_rate": 1.3625751551459542e-05,
      "loss": 0.2253,
      "num_input_tokens_seen": 22528816,
      "step": 26015
    },
    {
      "epoch": 12.267798208392268,
      "grad_norm": 1.8330532312393188,
      "learning_rate": 1.3617009898674188e-05,
      "loss": 0.2019,
      "num_input_tokens_seen": 22532704,
      "step": 26020
    },
    {
      "epoch": 12.27015558698727,
      "grad_norm": 1.0180710554122925,
      "learning_rate": 1.3608270001288967e-05,
      "loss": 0.1866,
      "num_input_tokens_seen": 22537456,
      "step": 26025
    },
    {
      "epoch": 12.272512965582273,
      "grad_norm": 0.9964539408683777,
      "learning_rate": 1.359953186065166e-05,
      "loss": 0.153,
      "num_input_tokens_seen": 22542176,
      "step": 26030
    },
    {
      "epoch": 12.274870344177275,
      "grad_norm": 0.6135226488113403,
      "learning_rate": 1.3590795478109814e-05,
      "loss": 0.2088,
      "num_input_tokens_seen": 22546560,
      "step": 26035
    },
    {
      "epoch": 12.277227722772277,
      "grad_norm": 1.23873770236969,
      "learning_rate": 1.3582060855010675e-05,
      "loss": 0.1882,
      "num_input_tokens_seen": 22550848,
      "step": 26040
    },
    {
      "epoch": 12.27958510136728,
      "grad_norm": 2.318521738052368,
      "learning_rate": 1.3573327992701245e-05,
      "loss": 0.2398,
      "num_input_tokens_seen": 22555696,
      "step": 26045
    },
    {
      "epoch": 12.281942479962282,
      "grad_norm": 1.8039488792419434,
      "learning_rate": 1.356459689252823e-05,
      "loss": 0.1624,
      "num_input_tokens_seen": 22559232,
      "step": 26050
    },
    {
      "epoch": 12.284299858557285,
      "grad_norm": 0.5314809083938599,
      "learning_rate": 1.3555867555838087e-05,
      "loss": 0.1928,
      "num_input_tokens_seen": 22563728,
      "step": 26055
    },
    {
      "epoch": 12.286657237152287,
      "grad_norm": 1.3925952911376953,
      "learning_rate": 1.3547139983976975e-05,
      "loss": 0.1242,
      "num_input_tokens_seen": 22569664,
      "step": 26060
    },
    {
      "epoch": 12.28901461574729,
      "grad_norm": 0.3630935847759247,
      "learning_rate": 1.3538414178290815e-05,
      "loss": 0.1514,
      "num_input_tokens_seen": 22574096,
      "step": 26065
    },
    {
      "epoch": 12.291371994342292,
      "grad_norm": 0.5023082494735718,
      "learning_rate": 1.3529690140125209e-05,
      "loss": 0.1587,
      "num_input_tokens_seen": 22578384,
      "step": 26070
    },
    {
      "epoch": 12.293729372937294,
      "grad_norm": 1.6518100500106812,
      "learning_rate": 1.352096787082553e-05,
      "loss": 0.1624,
      "num_input_tokens_seen": 22582400,
      "step": 26075
    },
    {
      "epoch": 12.296086751532297,
      "grad_norm": 0.7791001796722412,
      "learning_rate": 1.3512247371736871e-05,
      "loss": 0.1186,
      "num_input_tokens_seen": 22587168,
      "step": 26080
    },
    {
      "epoch": 12.298444130127299,
      "grad_norm": 1.9985321760177612,
      "learning_rate": 1.3503528644204022e-05,
      "loss": 0.2059,
      "num_input_tokens_seen": 22591952,
      "step": 26085
    },
    {
      "epoch": 12.300801508722301,
      "grad_norm": 1.2299672365188599,
      "learning_rate": 1.349481168957153e-05,
      "loss": 0.1708,
      "num_input_tokens_seen": 22596368,
      "step": 26090
    },
    {
      "epoch": 12.303158887317304,
      "grad_norm": 0.7738209366798401,
      "learning_rate": 1.3486096509183665e-05,
      "loss": 0.1684,
      "num_input_tokens_seen": 22600192,
      "step": 26095
    },
    {
      "epoch": 12.305516265912306,
      "grad_norm": 1.7829961776733398,
      "learning_rate": 1.3477383104384406e-05,
      "loss": 0.2015,
      "num_input_tokens_seen": 22604768,
      "step": 26100
    },
    {
      "epoch": 12.307873644507309,
      "grad_norm": 1.2748359441757202,
      "learning_rate": 1.3468671476517481e-05,
      "loss": 0.1914,
      "num_input_tokens_seen": 22608992,
      "step": 26105
    },
    {
      "epoch": 12.310231023102311,
      "grad_norm": 1.6642669439315796,
      "learning_rate": 1.3459961626926326e-05,
      "loss": 0.3007,
      "num_input_tokens_seen": 22614000,
      "step": 26110
    },
    {
      "epoch": 12.312588401697312,
      "grad_norm": 1.3558396100997925,
      "learning_rate": 1.3451253556954101e-05,
      "loss": 0.1402,
      "num_input_tokens_seen": 22618224,
      "step": 26115
    },
    {
      "epoch": 12.314945780292314,
      "grad_norm": 0.5252816081047058,
      "learning_rate": 1.3442547267943717e-05,
      "loss": 0.1659,
      "num_input_tokens_seen": 22622768,
      "step": 26120
    },
    {
      "epoch": 12.317303158887317,
      "grad_norm": 0.5595914125442505,
      "learning_rate": 1.3433842761237774e-05,
      "loss": 0.2074,
      "num_input_tokens_seen": 22627456,
      "step": 26125
    },
    {
      "epoch": 12.319660537482319,
      "grad_norm": 1.1636271476745605,
      "learning_rate": 1.3425140038178639e-05,
      "loss": 0.1452,
      "num_input_tokens_seen": 22631616,
      "step": 26130
    },
    {
      "epoch": 12.322017916077321,
      "grad_norm": 0.4914737045764923,
      "learning_rate": 1.3416439100108358e-05,
      "loss": 0.2412,
      "num_input_tokens_seen": 22636048,
      "step": 26135
    },
    {
      "epoch": 12.324375294672324,
      "grad_norm": 0.9338358044624329,
      "learning_rate": 1.3407739948368734e-05,
      "loss": 0.1608,
      "num_input_tokens_seen": 22640640,
      "step": 26140
    },
    {
      "epoch": 12.326732673267326,
      "grad_norm": 1.1276267766952515,
      "learning_rate": 1.3399042584301298e-05,
      "loss": 0.1805,
      "num_input_tokens_seen": 22645728,
      "step": 26145
    },
    {
      "epoch": 12.329090051862329,
      "grad_norm": 1.2042728662490845,
      "learning_rate": 1.3390347009247272e-05,
      "loss": 0.2392,
      "num_input_tokens_seen": 22649968,
      "step": 26150
    },
    {
      "epoch": 12.331447430457331,
      "grad_norm": 0.5502938032150269,
      "learning_rate": 1.3381653224547635e-05,
      "loss": 0.152,
      "num_input_tokens_seen": 22654016,
      "step": 26155
    },
    {
      "epoch": 12.333804809052333,
      "grad_norm": 2.590339422225952,
      "learning_rate": 1.3372961231543086e-05,
      "loss": 0.2111,
      "num_input_tokens_seen": 22657888,
      "step": 26160
    },
    {
      "epoch": 12.336162187647336,
      "grad_norm": 2.253267526626587,
      "learning_rate": 1.3364271031574016e-05,
      "loss": 0.235,
      "num_input_tokens_seen": 22661968,
      "step": 26165
    },
    {
      "epoch": 12.338519566242338,
      "grad_norm": 0.9157306551933289,
      "learning_rate": 1.335558262598059e-05,
      "loss": 0.2079,
      "num_input_tokens_seen": 22665648,
      "step": 26170
    },
    {
      "epoch": 12.34087694483734,
      "grad_norm": 1.4312829971313477,
      "learning_rate": 1.3346896016102645e-05,
      "loss": 0.1967,
      "num_input_tokens_seen": 22669760,
      "step": 26175
    },
    {
      "epoch": 12.343234323432343,
      "grad_norm": 1.067774772644043,
      "learning_rate": 1.3338211203279788e-05,
      "loss": 0.1721,
      "num_input_tokens_seen": 22674240,
      "step": 26180
    },
    {
      "epoch": 12.345591702027345,
      "grad_norm": 0.9520806074142456,
      "learning_rate": 1.3329528188851303e-05,
      "loss": 0.1535,
      "num_input_tokens_seen": 22678464,
      "step": 26185
    },
    {
      "epoch": 12.347949080622348,
      "grad_norm": 1.6584498882293701,
      "learning_rate": 1.3320846974156242e-05,
      "loss": 0.1973,
      "num_input_tokens_seen": 22683152,
      "step": 26190
    },
    {
      "epoch": 12.35030645921735,
      "grad_norm": 1.4351707696914673,
      "learning_rate": 1.3312167560533337e-05,
      "loss": 0.2083,
      "num_input_tokens_seen": 22687664,
      "step": 26195
    },
    {
      "epoch": 12.352663837812353,
      "grad_norm": 0.7301865220069885,
      "learning_rate": 1.3303489949321082e-05,
      "loss": 0.1485,
      "num_input_tokens_seen": 22692240,
      "step": 26200
    },
    {
      "epoch": 12.352663837812353,
      "eval_loss": 0.20149853825569153,
      "eval_runtime": 22.1297,
      "eval_samples_per_second": 42.612,
      "eval_steps_per_second": 21.329,
      "num_input_tokens_seen": 22692240,
      "step": 26200
    },
    {
      "epoch": 12.355021216407355,
      "grad_norm": 0.896356463432312,
      "learning_rate": 1.3294814141857653e-05,
      "loss": 0.2001,
      "num_input_tokens_seen": 22697136,
      "step": 26205
    },
    {
      "epoch": 12.357378595002357,
      "grad_norm": 1.051675796508789,
      "learning_rate": 1.3286140139480992e-05,
      "loss": 0.212,
      "num_input_tokens_seen": 22701184,
      "step": 26210
    },
    {
      "epoch": 12.35973597359736,
      "grad_norm": 1.6972858905792236,
      "learning_rate": 1.3277467943528719e-05,
      "loss": 0.1778,
      "num_input_tokens_seen": 22704944,
      "step": 26215
    },
    {
      "epoch": 12.362093352192362,
      "grad_norm": 1.7822511196136475,
      "learning_rate": 1.3268797555338203e-05,
      "loss": 0.1791,
      "num_input_tokens_seen": 22709488,
      "step": 26220
    },
    {
      "epoch": 12.364450730787365,
      "grad_norm": 0.704693078994751,
      "learning_rate": 1.3260128976246533e-05,
      "loss": 0.1436,
      "num_input_tokens_seen": 22713616,
      "step": 26225
    },
    {
      "epoch": 12.366808109382367,
      "grad_norm": 1.6847076416015625,
      "learning_rate": 1.32514622075905e-05,
      "loss": 0.158,
      "num_input_tokens_seen": 22717856,
      "step": 26230
    },
    {
      "epoch": 12.36916548797737,
      "grad_norm": 0.8796930909156799,
      "learning_rate": 1.3242797250706638e-05,
      "loss": 0.1898,
      "num_input_tokens_seen": 22721712,
      "step": 26235
    },
    {
      "epoch": 12.371522866572372,
      "grad_norm": 2.000711441040039,
      "learning_rate": 1.3234134106931195e-05,
      "loss": 0.1745,
      "num_input_tokens_seen": 22726736,
      "step": 26240
    },
    {
      "epoch": 12.373880245167374,
      "grad_norm": 1.4062272310256958,
      "learning_rate": 1.322547277760013e-05,
      "loss": 0.157,
      "num_input_tokens_seen": 22730848,
      "step": 26245
    },
    {
      "epoch": 12.376237623762377,
      "grad_norm": 2.6683101654052734,
      "learning_rate": 1.3216813264049132e-05,
      "loss": 0.1519,
      "num_input_tokens_seen": 22735056,
      "step": 26250
    },
    {
      "epoch": 12.378595002357379,
      "grad_norm": 0.6081883907318115,
      "learning_rate": 1.32081555676136e-05,
      "loss": 0.1938,
      "num_input_tokens_seen": 22739872,
      "step": 26255
    },
    {
      "epoch": 12.380952380952381,
      "grad_norm": 0.8276988863945007,
      "learning_rate": 1.3199499689628674e-05,
      "loss": 0.1372,
      "num_input_tokens_seen": 22744224,
      "step": 26260
    },
    {
      "epoch": 12.383309759547384,
      "grad_norm": 0.542558491230011,
      "learning_rate": 1.3190845631429192e-05,
      "loss": 0.1792,
      "num_input_tokens_seen": 22748720,
      "step": 26265
    },
    {
      "epoch": 12.385667138142386,
      "grad_norm": 0.715173065662384,
      "learning_rate": 1.3182193394349704e-05,
      "loss": 0.2098,
      "num_input_tokens_seen": 22753152,
      "step": 26270
    },
    {
      "epoch": 12.388024516737389,
      "grad_norm": 0.4564567506313324,
      "learning_rate": 1.3173542979724507e-05,
      "loss": 0.1305,
      "num_input_tokens_seen": 22756848,
      "step": 26275
    },
    {
      "epoch": 12.390381895332391,
      "grad_norm": 0.7724074125289917,
      "learning_rate": 1.3164894388887617e-05,
      "loss": 0.1732,
      "num_input_tokens_seen": 22760256,
      "step": 26280
    },
    {
      "epoch": 12.392739273927393,
      "grad_norm": 0.9416939616203308,
      "learning_rate": 1.3156247623172727e-05,
      "loss": 0.1473,
      "num_input_tokens_seen": 22765280,
      "step": 26285
    },
    {
      "epoch": 12.395096652522396,
      "grad_norm": 1.331099510192871,
      "learning_rate": 1.3147602683913302e-05,
      "loss": 0.1401,
      "num_input_tokens_seen": 22769536,
      "step": 26290
    },
    {
      "epoch": 12.397454031117398,
      "grad_norm": 0.4594915509223938,
      "learning_rate": 1.3138959572442481e-05,
      "loss": 0.1508,
      "num_input_tokens_seen": 22773792,
      "step": 26295
    },
    {
      "epoch": 12.3998114097124,
      "grad_norm": 1.8457502126693726,
      "learning_rate": 1.3130318290093146e-05,
      "loss": 0.1902,
      "num_input_tokens_seen": 22778112,
      "step": 26300
    },
    {
      "epoch": 12.402168788307403,
      "grad_norm": 0.36446094512939453,
      "learning_rate": 1.3121678838197909e-05,
      "loss": 0.1463,
      "num_input_tokens_seen": 22782720,
      "step": 26305
    },
    {
      "epoch": 12.404526166902404,
      "grad_norm": 1.635817050933838,
      "learning_rate": 1.3113041218089056e-05,
      "loss": 0.2528,
      "num_input_tokens_seen": 22787296,
      "step": 26310
    },
    {
      "epoch": 12.406883545497408,
      "grad_norm": 0.6035531163215637,
      "learning_rate": 1.3104405431098626e-05,
      "loss": 0.1241,
      "num_input_tokens_seen": 22791312,
      "step": 26315
    },
    {
      "epoch": 12.409240924092408,
      "grad_norm": 1.0808236598968506,
      "learning_rate": 1.3095771478558377e-05,
      "loss": 0.1665,
      "num_input_tokens_seen": 22795920,
      "step": 26320
    },
    {
      "epoch": 12.41159830268741,
      "grad_norm": 0.41122597455978394,
      "learning_rate": 1.3087139361799766e-05,
      "loss": 0.1619,
      "num_input_tokens_seen": 22800512,
      "step": 26325
    },
    {
      "epoch": 12.413955681282413,
      "grad_norm": 1.5447074174880981,
      "learning_rate": 1.3078509082153964e-05,
      "loss": 0.1725,
      "num_input_tokens_seen": 22804544,
      "step": 26330
    },
    {
      "epoch": 12.416313059877416,
      "grad_norm": 0.49688899517059326,
      "learning_rate": 1.3069880640951885e-05,
      "loss": 0.1295,
      "num_input_tokens_seen": 22809456,
      "step": 26335
    },
    {
      "epoch": 12.418670438472418,
      "grad_norm": 0.7409977912902832,
      "learning_rate": 1.3061254039524123e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 22813648,
      "step": 26340
    },
    {
      "epoch": 12.42102781706742,
      "grad_norm": 1.0388052463531494,
      "learning_rate": 1.3052629279201028e-05,
      "loss": 0.2217,
      "num_input_tokens_seen": 22817808,
      "step": 26345
    },
    {
      "epoch": 12.423385195662423,
      "grad_norm": 0.7629923820495605,
      "learning_rate": 1.3044006361312633e-05,
      "loss": 0.1545,
      "num_input_tokens_seen": 22821648,
      "step": 26350
    },
    {
      "epoch": 12.425742574257425,
      "grad_norm": 1.5166523456573486,
      "learning_rate": 1.30353852871887e-05,
      "loss": 0.1771,
      "num_input_tokens_seen": 22826560,
      "step": 26355
    },
    {
      "epoch": 12.428099952852428,
      "grad_norm": 1.1027796268463135,
      "learning_rate": 1.302676605815873e-05,
      "loss": 0.2168,
      "num_input_tokens_seen": 22830096,
      "step": 26360
    },
    {
      "epoch": 12.43045733144743,
      "grad_norm": 2.700120687484741,
      "learning_rate": 1.3018148675551884e-05,
      "loss": 0.1192,
      "num_input_tokens_seen": 22834560,
      "step": 26365
    },
    {
      "epoch": 12.432814710042432,
      "grad_norm": 1.3693257570266724,
      "learning_rate": 1.3009533140697094e-05,
      "loss": 0.2286,
      "num_input_tokens_seen": 22839264,
      "step": 26370
    },
    {
      "epoch": 12.435172088637435,
      "grad_norm": 1.1453804969787598,
      "learning_rate": 1.3000919454922966e-05,
      "loss": 0.1893,
      "num_input_tokens_seen": 22843296,
      "step": 26375
    },
    {
      "epoch": 12.437529467232437,
      "grad_norm": 1.2220064401626587,
      "learning_rate": 1.299230761955785e-05,
      "loss": 0.2198,
      "num_input_tokens_seen": 22847712,
      "step": 26380
    },
    {
      "epoch": 12.43988684582744,
      "grad_norm": 0.955959677696228,
      "learning_rate": 1.2983697635929807e-05,
      "loss": 0.2159,
      "num_input_tokens_seen": 22852080,
      "step": 26385
    },
    {
      "epoch": 12.442244224422442,
      "grad_norm": 1.0669766664505005,
      "learning_rate": 1.2975089505366584e-05,
      "loss": 0.1556,
      "num_input_tokens_seen": 22856368,
      "step": 26390
    },
    {
      "epoch": 12.444601603017444,
      "grad_norm": 1.053932785987854,
      "learning_rate": 1.2966483229195683e-05,
      "loss": 0.2158,
      "num_input_tokens_seen": 22860624,
      "step": 26395
    },
    {
      "epoch": 12.446958981612447,
      "grad_norm": 0.3135656714439392,
      "learning_rate": 1.2957878808744283e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 22864512,
      "step": 26400
    },
    {
      "epoch": 12.446958981612447,
      "eval_loss": 0.2017444670200348,
      "eval_runtime": 22.2035,
      "eval_samples_per_second": 42.471,
      "eval_steps_per_second": 21.258,
      "num_input_tokens_seen": 22864512,
      "step": 26400
    },
    {
      "epoch": 12.44931636020745,
      "grad_norm": 1.26803457736969,
      "learning_rate": 1.294927624533931e-05,
      "loss": 0.1683,
      "num_input_tokens_seen": 22869232,
      "step": 26405
    },
    {
      "epoch": 12.451673738802452,
      "grad_norm": 1.5558652877807617,
      "learning_rate": 1.2940675540307378e-05,
      "loss": 0.203,
      "num_input_tokens_seen": 22873360,
      "step": 26410
    },
    {
      "epoch": 12.454031117397454,
      "grad_norm": 1.1064486503601074,
      "learning_rate": 1.2932076694974814e-05,
      "loss": 0.2501,
      "num_input_tokens_seen": 22877184,
      "step": 26415
    },
    {
      "epoch": 12.456388495992456,
      "grad_norm": 0.8569207191467285,
      "learning_rate": 1.2923479710667682e-05,
      "loss": 0.1472,
      "num_input_tokens_seen": 22881600,
      "step": 26420
    },
    {
      "epoch": 12.458745874587459,
      "grad_norm": 0.6522246599197388,
      "learning_rate": 1.2914884588711751e-05,
      "loss": 0.1879,
      "num_input_tokens_seen": 22885904,
      "step": 26425
    },
    {
      "epoch": 12.461103253182461,
      "grad_norm": 1.5130289793014526,
      "learning_rate": 1.2906291330432475e-05,
      "loss": 0.1267,
      "num_input_tokens_seen": 22890672,
      "step": 26430
    },
    {
      "epoch": 12.463460631777464,
      "grad_norm": 0.7450770139694214,
      "learning_rate": 1.2897699937155055e-05,
      "loss": 0.1912,
      "num_input_tokens_seen": 22895216,
      "step": 26435
    },
    {
      "epoch": 12.465818010372466,
      "grad_norm": 0.7547131776809692,
      "learning_rate": 1.2889110410204403e-05,
      "loss": 0.1744,
      "num_input_tokens_seen": 22900000,
      "step": 26440
    },
    {
      "epoch": 12.468175388967468,
      "grad_norm": 1.3905330896377563,
      "learning_rate": 1.2880522750905111e-05,
      "loss": 0.1886,
      "num_input_tokens_seen": 22904336,
      "step": 26445
    },
    {
      "epoch": 12.47053276756247,
      "grad_norm": 0.6403145790100098,
      "learning_rate": 1.2871936960581523e-05,
      "loss": 0.1508,
      "num_input_tokens_seen": 22909472,
      "step": 26450
    },
    {
      "epoch": 12.472890146157473,
      "grad_norm": 1.0649291276931763,
      "learning_rate": 1.2863353040557658e-05,
      "loss": 0.2134,
      "num_input_tokens_seen": 22913632,
      "step": 26455
    },
    {
      "epoch": 12.475247524752476,
      "grad_norm": 1.274760127067566,
      "learning_rate": 1.2854770992157273e-05,
      "loss": 0.1665,
      "num_input_tokens_seen": 22918016,
      "step": 26460
    },
    {
      "epoch": 12.477604903347478,
      "grad_norm": 1.4032137393951416,
      "learning_rate": 1.2846190816703835e-05,
      "loss": 0.1561,
      "num_input_tokens_seen": 22922192,
      "step": 26465
    },
    {
      "epoch": 12.47996228194248,
      "grad_norm": 1.109297513961792,
      "learning_rate": 1.2837612515520498e-05,
      "loss": 0.1544,
      "num_input_tokens_seen": 22927072,
      "step": 26470
    },
    {
      "epoch": 12.482319660537483,
      "grad_norm": 0.5252145528793335,
      "learning_rate": 1.2829036089930163e-05,
      "loss": 0.1852,
      "num_input_tokens_seen": 22931376,
      "step": 26475
    },
    {
      "epoch": 12.484677039132485,
      "grad_norm": 0.9359779357910156,
      "learning_rate": 1.2820461541255412e-05,
      "loss": 0.158,
      "num_input_tokens_seen": 22935296,
      "step": 26480
    },
    {
      "epoch": 12.487034417727488,
      "grad_norm": 0.6496657133102417,
      "learning_rate": 1.2811888870818543e-05,
      "loss": 0.1673,
      "num_input_tokens_seen": 22939280,
      "step": 26485
    },
    {
      "epoch": 12.48939179632249,
      "grad_norm": 0.5221185088157654,
      "learning_rate": 1.2803318079941581e-05,
      "loss": 0.1773,
      "num_input_tokens_seen": 22943680,
      "step": 26490
    },
    {
      "epoch": 12.491749174917492,
      "grad_norm": 0.7148948311805725,
      "learning_rate": 1.2794749169946235e-05,
      "loss": 0.1889,
      "num_input_tokens_seen": 22948912,
      "step": 26495
    },
    {
      "epoch": 12.494106553512495,
      "grad_norm": 0.5809556841850281,
      "learning_rate": 1.2786182142153952e-05,
      "loss": 0.1996,
      "num_input_tokens_seen": 22953376,
      "step": 26500
    },
    {
      "epoch": 12.496463932107497,
      "grad_norm": 0.8628658056259155,
      "learning_rate": 1.2777616997885878e-05,
      "loss": 0.1464,
      "num_input_tokens_seen": 22957952,
      "step": 26505
    },
    {
      "epoch": 12.4988213107025,
      "grad_norm": 1.6913864612579346,
      "learning_rate": 1.2769053738462847e-05,
      "loss": 0.189,
      "num_input_tokens_seen": 22961632,
      "step": 26510
    },
    {
      "epoch": 12.5011786892975,
      "grad_norm": 0.6568983197212219,
      "learning_rate": 1.2760492365205434e-05,
      "loss": 0.1693,
      "num_input_tokens_seen": 22966080,
      "step": 26515
    },
    {
      "epoch": 12.503536067892503,
      "grad_norm": 0.6508215069770813,
      "learning_rate": 1.2751932879433919e-05,
      "loss": 0.1687,
      "num_input_tokens_seen": 22970000,
      "step": 26520
    },
    {
      "epoch": 12.505893446487505,
      "grad_norm": 1.122745394706726,
      "learning_rate": 1.2743375282468267e-05,
      "loss": 0.2345,
      "num_input_tokens_seen": 22974560,
      "step": 26525
    },
    {
      "epoch": 12.508250825082508,
      "grad_norm": 0.5459104776382446,
      "learning_rate": 1.2734819575628182e-05,
      "loss": 0.1321,
      "num_input_tokens_seen": 22978912,
      "step": 26530
    },
    {
      "epoch": 12.51060820367751,
      "grad_norm": 0.6970847845077515,
      "learning_rate": 1.2726265760233039e-05,
      "loss": 0.1708,
      "num_input_tokens_seen": 22982944,
      "step": 26535
    },
    {
      "epoch": 12.512965582272512,
      "grad_norm": 1.204940676689148,
      "learning_rate": 1.271771383760197e-05,
      "loss": 0.1834,
      "num_input_tokens_seen": 22987088,
      "step": 26540
    },
    {
      "epoch": 12.515322960867515,
      "grad_norm": 0.9333414435386658,
      "learning_rate": 1.2709163809053764e-05,
      "loss": 0.1392,
      "num_input_tokens_seen": 22991168,
      "step": 26545
    },
    {
      "epoch": 12.517680339462517,
      "grad_norm": 0.9269161820411682,
      "learning_rate": 1.2700615675906963e-05,
      "loss": 0.1511,
      "num_input_tokens_seen": 22995760,
      "step": 26550
    },
    {
      "epoch": 12.52003771805752,
      "grad_norm": 1.567359447479248,
      "learning_rate": 1.269206943947978e-05,
      "loss": 0.2311,
      "num_input_tokens_seen": 22999744,
      "step": 26555
    },
    {
      "epoch": 12.522395096652522,
      "grad_norm": 1.3268587589263916,
      "learning_rate": 1.2683525101090177e-05,
      "loss": 0.2157,
      "num_input_tokens_seen": 23004016,
      "step": 26560
    },
    {
      "epoch": 12.524752475247524,
      "grad_norm": 1.0637506246566772,
      "learning_rate": 1.2674982662055765e-05,
      "loss": 0.2678,
      "num_input_tokens_seen": 23008080,
      "step": 26565
    },
    {
      "epoch": 12.527109853842527,
      "grad_norm": 0.8785350918769836,
      "learning_rate": 1.2666442123693922e-05,
      "loss": 0.1241,
      "num_input_tokens_seen": 23012832,
      "step": 26570
    },
    {
      "epoch": 12.52946723243753,
      "grad_norm": 1.3909573554992676,
      "learning_rate": 1.265790348732169e-05,
      "loss": 0.1952,
      "num_input_tokens_seen": 23016672,
      "step": 26575
    },
    {
      "epoch": 12.531824611032532,
      "grad_norm": 0.6041833758354187,
      "learning_rate": 1.264936675425584e-05,
      "loss": 0.1253,
      "num_input_tokens_seen": 23020640,
      "step": 26580
    },
    {
      "epoch": 12.534181989627534,
      "grad_norm": 1.1506879329681396,
      "learning_rate": 1.2640831925812852e-05,
      "loss": 0.1652,
      "num_input_tokens_seen": 23025248,
      "step": 26585
    },
    {
      "epoch": 12.536539368222536,
      "grad_norm": 1.2509690523147583,
      "learning_rate": 1.263229900330889e-05,
      "loss": 0.1627,
      "num_input_tokens_seen": 23029424,
      "step": 26590
    },
    {
      "epoch": 12.538896746817539,
      "grad_norm": 0.6931034922599792,
      "learning_rate": 1.2623767988059843e-05,
      "loss": 0.2279,
      "num_input_tokens_seen": 23033120,
      "step": 26595
    },
    {
      "epoch": 12.541254125412541,
      "grad_norm": 1.2713524103164673,
      "learning_rate": 1.2615238881381309e-05,
      "loss": 0.1747,
      "num_input_tokens_seen": 23037568,
      "step": 26600
    },
    {
      "epoch": 12.541254125412541,
      "eval_loss": 0.20116207003593445,
      "eval_runtime": 22.1654,
      "eval_samples_per_second": 42.544,
      "eval_steps_per_second": 21.294,
      "num_input_tokens_seen": 23037568,
      "step": 26600
    },
    {
      "epoch": 12.543611504007544,
      "grad_norm": 0.8482495546340942,
      "learning_rate": 1.2606711684588568e-05,
      "loss": 0.206,
      "num_input_tokens_seen": 23041920,
      "step": 26605
    },
    {
      "epoch": 12.545968882602546,
      "grad_norm": 0.5668378472328186,
      "learning_rate": 1.2598186398996636e-05,
      "loss": 0.1979,
      "num_input_tokens_seen": 23046816,
      "step": 26610
    },
    {
      "epoch": 12.548326261197548,
      "grad_norm": 1.4612935781478882,
      "learning_rate": 1.2589663025920207e-05,
      "loss": 0.2069,
      "num_input_tokens_seen": 23051152,
      "step": 26615
    },
    {
      "epoch": 12.55068363979255,
      "grad_norm": 1.0398457050323486,
      "learning_rate": 1.2581141566673705e-05,
      "loss": 0.1549,
      "num_input_tokens_seen": 23055440,
      "step": 26620
    },
    {
      "epoch": 12.553041018387553,
      "grad_norm": 0.6727153658866882,
      "learning_rate": 1.257262202257124e-05,
      "loss": 0.2402,
      "num_input_tokens_seen": 23060224,
      "step": 26625
    },
    {
      "epoch": 12.555398396982556,
      "grad_norm": 0.7632288336753845,
      "learning_rate": 1.2564104394926618e-05,
      "loss": 0.1743,
      "num_input_tokens_seen": 23064128,
      "step": 26630
    },
    {
      "epoch": 12.557755775577558,
      "grad_norm": 1.3730077743530273,
      "learning_rate": 1.2555588685053383e-05,
      "loss": 0.1243,
      "num_input_tokens_seen": 23068336,
      "step": 26635
    },
    {
      "epoch": 12.56011315417256,
      "grad_norm": 0.5094214081764221,
      "learning_rate": 1.2547074894264762e-05,
      "loss": 0.2092,
      "num_input_tokens_seen": 23072048,
      "step": 26640
    },
    {
      "epoch": 12.562470532767563,
      "grad_norm": 0.6249430179595947,
      "learning_rate": 1.2538563023873679e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 23076064,
      "step": 26645
    },
    {
      "epoch": 12.564827911362565,
      "grad_norm": 1.4391802549362183,
      "learning_rate": 1.2530053075192789e-05,
      "loss": 0.2205,
      "num_input_tokens_seen": 23080496,
      "step": 26650
    },
    {
      "epoch": 12.567185289957568,
      "grad_norm": 1.2938227653503418,
      "learning_rate": 1.252154504953441e-05,
      "loss": 0.1529,
      "num_input_tokens_seen": 23084560,
      "step": 26655
    },
    {
      "epoch": 12.56954266855257,
      "grad_norm": 0.7283514738082886,
      "learning_rate": 1.25130389482106e-05,
      "loss": 0.24,
      "num_input_tokens_seen": 23089008,
      "step": 26660
    },
    {
      "epoch": 12.571900047147572,
      "grad_norm": 1.3061002492904663,
      "learning_rate": 1.2504534772533116e-05,
      "loss": 0.1634,
      "num_input_tokens_seen": 23093248,
      "step": 26665
    },
    {
      "epoch": 12.574257425742575,
      "grad_norm": 0.7806869745254517,
      "learning_rate": 1.2496032523813387e-05,
      "loss": 0.166,
      "num_input_tokens_seen": 23097120,
      "step": 26670
    },
    {
      "epoch": 12.576614804337577,
      "grad_norm": 2.0281660556793213,
      "learning_rate": 1.2487532203362576e-05,
      "loss": 0.2733,
      "num_input_tokens_seen": 23101600,
      "step": 26675
    },
    {
      "epoch": 12.57897218293258,
      "grad_norm": 0.31804656982421875,
      "learning_rate": 1.247903381249155e-05,
      "loss": 0.1304,
      "num_input_tokens_seen": 23105680,
      "step": 26680
    },
    {
      "epoch": 12.581329561527582,
      "grad_norm": 0.7548238039016724,
      "learning_rate": 1.2470537352510853e-05,
      "loss": 0.1587,
      "num_input_tokens_seen": 23110032,
      "step": 26685
    },
    {
      "epoch": 12.583686940122584,
      "grad_norm": 0.8995374441146851,
      "learning_rate": 1.2462042824730758e-05,
      "loss": 0.1972,
      "num_input_tokens_seen": 23114640,
      "step": 26690
    },
    {
      "epoch": 12.586044318717587,
      "grad_norm": 2.459531545639038,
      "learning_rate": 1.245355023046122e-05,
      "loss": 0.2474,
      "num_input_tokens_seen": 23118864,
      "step": 26695
    },
    {
      "epoch": 12.58840169731259,
      "grad_norm": 0.9273573160171509,
      "learning_rate": 1.2445059571011896e-05,
      "loss": 0.17,
      "num_input_tokens_seen": 23123488,
      "step": 26700
    },
    {
      "epoch": 12.590759075907592,
      "grad_norm": 0.5023062825202942,
      "learning_rate": 1.2436570847692173e-05,
      "loss": 0.1735,
      "num_input_tokens_seen": 23129072,
      "step": 26705
    },
    {
      "epoch": 12.593116454502592,
      "grad_norm": 0.6258382201194763,
      "learning_rate": 1.2428084061811096e-05,
      "loss": 0.2098,
      "num_input_tokens_seen": 23133040,
      "step": 26710
    },
    {
      "epoch": 12.595473833097596,
      "grad_norm": 0.45832589268684387,
      "learning_rate": 1.2419599214677447e-05,
      "loss": 0.1347,
      "num_input_tokens_seen": 23136720,
      "step": 26715
    },
    {
      "epoch": 12.597831211692597,
      "grad_norm": 1.2757738828659058,
      "learning_rate": 1.2411116307599702e-05,
      "loss": 0.1567,
      "num_input_tokens_seen": 23141696,
      "step": 26720
    },
    {
      "epoch": 12.6001885902876,
      "grad_norm": 1.0443130731582642,
      "learning_rate": 1.2402635341886016e-05,
      "loss": 0.1697,
      "num_input_tokens_seen": 23146192,
      "step": 26725
    },
    {
      "epoch": 12.602545968882602,
      "grad_norm": 1.6601953506469727,
      "learning_rate": 1.2394156318844278e-05,
      "loss": 0.2163,
      "num_input_tokens_seen": 23150560,
      "step": 26730
    },
    {
      "epoch": 12.604903347477604,
      "grad_norm": 1.7042012214660645,
      "learning_rate": 1.2385679239782039e-05,
      "loss": 0.1595,
      "num_input_tokens_seen": 23154944,
      "step": 26735
    },
    {
      "epoch": 12.607260726072607,
      "grad_norm": 0.48037710785865784,
      "learning_rate": 1.2377204106006585e-05,
      "loss": 0.1602,
      "num_input_tokens_seen": 23159616,
      "step": 26740
    },
    {
      "epoch": 12.609618104667609,
      "grad_norm": 0.44572076201438904,
      "learning_rate": 1.2368730918824891e-05,
      "loss": 0.0987,
      "num_input_tokens_seen": 23163760,
      "step": 26745
    },
    {
      "epoch": 12.611975483262611,
      "grad_norm": 0.5814863443374634,
      "learning_rate": 1.236025967954362e-05,
      "loss": 0.092,
      "num_input_tokens_seen": 23167824,
      "step": 26750
    },
    {
      "epoch": 12.614332861857614,
      "grad_norm": 1.5072500705718994,
      "learning_rate": 1.2351790389469153e-05,
      "loss": 0.1818,
      "num_input_tokens_seen": 23172384,
      "step": 26755
    },
    {
      "epoch": 12.616690240452616,
      "grad_norm": 1.0410326719284058,
      "learning_rate": 1.234332304990755e-05,
      "loss": 0.1961,
      "num_input_tokens_seen": 23177040,
      "step": 26760
    },
    {
      "epoch": 12.619047619047619,
      "grad_norm": 0.5788867473602295,
      "learning_rate": 1.2334857662164593e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 23181424,
      "step": 26765
    },
    {
      "epoch": 12.621404997642621,
      "grad_norm": 0.8182691335678101,
      "learning_rate": 1.2326394227545743e-05,
      "loss": 0.2085,
      "num_input_tokens_seen": 23185456,
      "step": 26770
    },
    {
      "epoch": 12.623762376237623,
      "grad_norm": 0.6473545432090759,
      "learning_rate": 1.2317932747356162e-05,
      "loss": 0.2795,
      "num_input_tokens_seen": 23188736,
      "step": 26775
    },
    {
      "epoch": 12.626119754832626,
      "grad_norm": 1.3619524240493774,
      "learning_rate": 1.2309473222900726e-05,
      "loss": 0.2185,
      "num_input_tokens_seen": 23193088,
      "step": 26780
    },
    {
      "epoch": 12.628477133427628,
      "grad_norm": 2.482442617416382,
      "learning_rate": 1.2301015655484006e-05,
      "loss": 0.193,
      "num_input_tokens_seen": 23196800,
      "step": 26785
    },
    {
      "epoch": 12.63083451202263,
      "grad_norm": 0.25286638736724854,
      "learning_rate": 1.2292560046410245e-05,
      "loss": 0.0996,
      "num_input_tokens_seen": 23200784,
      "step": 26790
    },
    {
      "epoch": 12.633191890617633,
      "grad_norm": 1.3431557416915894,
      "learning_rate": 1.228410639698343e-05,
      "loss": 0.1983,
      "num_input_tokens_seen": 23204192,
      "step": 26795
    },
    {
      "epoch": 12.635549269212635,
      "grad_norm": 0.6801823973655701,
      "learning_rate": 1.2275654708507195e-05,
      "loss": 0.2277,
      "num_input_tokens_seen": 23207936,
      "step": 26800
    },
    {
      "epoch": 12.635549269212635,
      "eval_loss": 0.20015037059783936,
      "eval_runtime": 22.1646,
      "eval_samples_per_second": 42.545,
      "eval_steps_per_second": 21.295,
      "num_input_tokens_seen": 23207936,
      "step": 26800
    },
    {
      "epoch": 12.637906647807638,
      "grad_norm": 0.5632161498069763,
      "learning_rate": 1.2267204982284908e-05,
      "loss": 0.1456,
      "num_input_tokens_seen": 23212432,
      "step": 26805
    },
    {
      "epoch": 12.64026402640264,
      "grad_norm": 0.7745083570480347,
      "learning_rate": 1.2258757219619635e-05,
      "loss": 0.1114,
      "num_input_tokens_seen": 23217584,
      "step": 26810
    },
    {
      "epoch": 12.642621404997643,
      "grad_norm": 0.6176864504814148,
      "learning_rate": 1.2250311421814104e-05,
      "loss": 0.2296,
      "num_input_tokens_seen": 23221536,
      "step": 26815
    },
    {
      "epoch": 12.644978783592645,
      "grad_norm": 2.745605230331421,
      "learning_rate": 1.2241867590170772e-05,
      "loss": 0.199,
      "num_input_tokens_seen": 23225360,
      "step": 26820
    },
    {
      "epoch": 12.647336162187647,
      "grad_norm": 0.9966856837272644,
      "learning_rate": 1.2233425725991799e-05,
      "loss": 0.1858,
      "num_input_tokens_seen": 23230320,
      "step": 26825
    },
    {
      "epoch": 12.64969354078265,
      "grad_norm": 0.7256116271018982,
      "learning_rate": 1.2224985830579003e-05,
      "loss": 0.1718,
      "num_input_tokens_seen": 23235296,
      "step": 26830
    },
    {
      "epoch": 12.652050919377652,
      "grad_norm": 0.8336031436920166,
      "learning_rate": 1.2216547905233944e-05,
      "loss": 0.172,
      "num_input_tokens_seen": 23240240,
      "step": 26835
    },
    {
      "epoch": 12.654408297972655,
      "grad_norm": 0.5923427939414978,
      "learning_rate": 1.2208111951257842e-05,
      "loss": 0.1789,
      "num_input_tokens_seen": 23245088,
      "step": 26840
    },
    {
      "epoch": 12.656765676567657,
      "grad_norm": 1.1019341945648193,
      "learning_rate": 1.2199677969951622e-05,
      "loss": 0.2328,
      "num_input_tokens_seen": 23249744,
      "step": 26845
    },
    {
      "epoch": 12.65912305516266,
      "grad_norm": 0.8233985900878906,
      "learning_rate": 1.2191245962615927e-05,
      "loss": 0.2207,
      "num_input_tokens_seen": 23253984,
      "step": 26850
    },
    {
      "epoch": 12.661480433757662,
      "grad_norm": 0.9422771334648132,
      "learning_rate": 1.218281593055106e-05,
      "loss": 0.1745,
      "num_input_tokens_seen": 23258464,
      "step": 26855
    },
    {
      "epoch": 12.663837812352664,
      "grad_norm": 1.5905580520629883,
      "learning_rate": 1.217438787505705e-05,
      "loss": 0.2086,
      "num_input_tokens_seen": 23263792,
      "step": 26860
    },
    {
      "epoch": 12.666195190947667,
      "grad_norm": 0.7098751664161682,
      "learning_rate": 1.2165961797433615e-05,
      "loss": 0.1791,
      "num_input_tokens_seen": 23267936,
      "step": 26865
    },
    {
      "epoch": 12.668552569542669,
      "grad_norm": 1.9687312841415405,
      "learning_rate": 1.215753769898014e-05,
      "loss": 0.1722,
      "num_input_tokens_seen": 23272624,
      "step": 26870
    },
    {
      "epoch": 12.670909948137671,
      "grad_norm": 1.1256380081176758,
      "learning_rate": 1.2149115580995755e-05,
      "loss": 0.2348,
      "num_input_tokens_seen": 23277280,
      "step": 26875
    },
    {
      "epoch": 12.673267326732674,
      "grad_norm": 1.5331578254699707,
      "learning_rate": 1.2140695444779227e-05,
      "loss": 0.2467,
      "num_input_tokens_seen": 23281984,
      "step": 26880
    },
    {
      "epoch": 12.675624705327676,
      "grad_norm": 1.669206142425537,
      "learning_rate": 1.2132277291629066e-05,
      "loss": 0.2286,
      "num_input_tokens_seen": 23286160,
      "step": 26885
    },
    {
      "epoch": 12.677982083922679,
      "grad_norm": 1.2158801555633545,
      "learning_rate": 1.2123861122843458e-05,
      "loss": 0.1829,
      "num_input_tokens_seen": 23290800,
      "step": 26890
    },
    {
      "epoch": 12.680339462517681,
      "grad_norm": 1.0219777822494507,
      "learning_rate": 1.2115446939720271e-05,
      "loss": 0.1902,
      "num_input_tokens_seen": 23295328,
      "step": 26895
    },
    {
      "epoch": 12.682696841112683,
      "grad_norm": 1.1113651990890503,
      "learning_rate": 1.210703474355708e-05,
      "loss": 0.1957,
      "num_input_tokens_seen": 23299088,
      "step": 26900
    },
    {
      "epoch": 12.685054219707686,
      "grad_norm": 0.4246830344200134,
      "learning_rate": 1.2098624535651164e-05,
      "loss": 0.1693,
      "num_input_tokens_seen": 23302800,
      "step": 26905
    },
    {
      "epoch": 12.687411598302688,
      "grad_norm": 0.5680114030838013,
      "learning_rate": 1.2090216317299477e-05,
      "loss": 0.1681,
      "num_input_tokens_seen": 23306864,
      "step": 26910
    },
    {
      "epoch": 12.689768976897689,
      "grad_norm": 1.1502349376678467,
      "learning_rate": 1.2081810089798668e-05,
      "loss": 0.1936,
      "num_input_tokens_seen": 23311168,
      "step": 26915
    },
    {
      "epoch": 12.692126355492691,
      "grad_norm": 1.264572024345398,
      "learning_rate": 1.2073405854445072e-05,
      "loss": 0.1903,
      "num_input_tokens_seen": 23315200,
      "step": 26920
    },
    {
      "epoch": 12.694483734087694,
      "grad_norm": 0.6027985215187073,
      "learning_rate": 1.206500361253474e-05,
      "loss": 0.1581,
      "num_input_tokens_seen": 23320064,
      "step": 26925
    },
    {
      "epoch": 12.696841112682696,
      "grad_norm": 1.1156502962112427,
      "learning_rate": 1.2056603365363409e-05,
      "loss": 0.256,
      "num_input_tokens_seen": 23324496,
      "step": 26930
    },
    {
      "epoch": 12.699198491277699,
      "grad_norm": 1.2866572141647339,
      "learning_rate": 1.2048205114226487e-05,
      "loss": 0.1817,
      "num_input_tokens_seen": 23328528,
      "step": 26935
    },
    {
      "epoch": 12.701555869872701,
      "grad_norm": 1.5663986206054688,
      "learning_rate": 1.2039808860419102e-05,
      "loss": 0.1706,
      "num_input_tokens_seen": 23333472,
      "step": 26940
    },
    {
      "epoch": 12.703913248467703,
      "grad_norm": 1.5022034645080566,
      "learning_rate": 1.2031414605236066e-05,
      "loss": 0.1477,
      "num_input_tokens_seen": 23337152,
      "step": 26945
    },
    {
      "epoch": 12.706270627062706,
      "grad_norm": 0.9041879773139954,
      "learning_rate": 1.2023022349971862e-05,
      "loss": 0.1291,
      "num_input_tokens_seen": 23340848,
      "step": 26950
    },
    {
      "epoch": 12.708628005657708,
      "grad_norm": 1.2056769132614136,
      "learning_rate": 1.20146320959207e-05,
      "loss": 0.1404,
      "num_input_tokens_seen": 23344752,
      "step": 26955
    },
    {
      "epoch": 12.71098538425271,
      "grad_norm": 1.0587142705917358,
      "learning_rate": 1.2006243844376445e-05,
      "loss": 0.1929,
      "num_input_tokens_seen": 23348912,
      "step": 26960
    },
    {
      "epoch": 12.713342762847713,
      "grad_norm": 1.8293602466583252,
      "learning_rate": 1.1997857596632678e-05,
      "loss": 0.1789,
      "num_input_tokens_seen": 23353008,
      "step": 26965
    },
    {
      "epoch": 12.715700141442715,
      "grad_norm": 1.3377724885940552,
      "learning_rate": 1.1989473353982672e-05,
      "loss": 0.1811,
      "num_input_tokens_seen": 23356736,
      "step": 26970
    },
    {
      "epoch": 12.718057520037718,
      "grad_norm": 0.8579534888267517,
      "learning_rate": 1.198109111771937e-05,
      "loss": 0.1515,
      "num_input_tokens_seen": 23360784,
      "step": 26975
    },
    {
      "epoch": 12.72041489863272,
      "grad_norm": 1.997638463973999,
      "learning_rate": 1.197271088913543e-05,
      "loss": 0.188,
      "num_input_tokens_seen": 23364848,
      "step": 26980
    },
    {
      "epoch": 12.722772277227723,
      "grad_norm": 0.47318434715270996,
      "learning_rate": 1.1964332669523182e-05,
      "loss": 0.1626,
      "num_input_tokens_seen": 23368880,
      "step": 26985
    },
    {
      "epoch": 12.725129655822725,
      "grad_norm": 0.7750253081321716,
      "learning_rate": 1.1955956460174645e-05,
      "loss": 0.1666,
      "num_input_tokens_seen": 23372928,
      "step": 26990
    },
    {
      "epoch": 12.727487034417727,
      "grad_norm": 0.8088942766189575,
      "learning_rate": 1.1947582262381552e-05,
      "loss": 0.1595,
      "num_input_tokens_seen": 23377088,
      "step": 26995
    },
    {
      "epoch": 12.72984441301273,
      "grad_norm": 0.6914220452308655,
      "learning_rate": 1.1939210077435293e-05,
      "loss": 0.1293,
      "num_input_tokens_seen": 23381376,
      "step": 27000
    },
    {
      "epoch": 12.72984441301273,
      "eval_loss": 0.2000862956047058,
      "eval_runtime": 22.1749,
      "eval_samples_per_second": 42.526,
      "eval_steps_per_second": 21.285,
      "num_input_tokens_seen": 23381376,
      "step": 27000
    },
    {
      "epoch": 12.732201791607732,
      "grad_norm": 1.224565029144287,
      "learning_rate": 1.193083990662697e-05,
      "loss": 0.1738,
      "num_input_tokens_seen": 23385232,
      "step": 27005
    },
    {
      "epoch": 12.734559170202735,
      "grad_norm": 0.8871161937713623,
      "learning_rate": 1.192247175124738e-05,
      "loss": 0.2039,
      "num_input_tokens_seen": 23389280,
      "step": 27010
    },
    {
      "epoch": 12.736916548797737,
      "grad_norm": 2.813981056213379,
      "learning_rate": 1.191410561258698e-05,
      "loss": 0.3207,
      "num_input_tokens_seen": 23393296,
      "step": 27015
    },
    {
      "epoch": 12.73927392739274,
      "grad_norm": 1.2664904594421387,
      "learning_rate": 1.1905741491935944e-05,
      "loss": 0.1704,
      "num_input_tokens_seen": 23397456,
      "step": 27020
    },
    {
      "epoch": 12.741631305987742,
      "grad_norm": 0.8425100445747375,
      "learning_rate": 1.1897379390584129e-05,
      "loss": 0.2252,
      "num_input_tokens_seen": 23402000,
      "step": 27025
    },
    {
      "epoch": 12.743988684582744,
      "grad_norm": 1.0504114627838135,
      "learning_rate": 1.1889019309821062e-05,
      "loss": 0.2209,
      "num_input_tokens_seen": 23406592,
      "step": 27030
    },
    {
      "epoch": 12.746346063177747,
      "grad_norm": 0.45774027705192566,
      "learning_rate": 1.188066125093599e-05,
      "loss": 0.1347,
      "num_input_tokens_seen": 23410864,
      "step": 27035
    },
    {
      "epoch": 12.748703441772749,
      "grad_norm": 1.104559063911438,
      "learning_rate": 1.1872305215217811e-05,
      "loss": 0.1771,
      "num_input_tokens_seen": 23415120,
      "step": 27040
    },
    {
      "epoch": 12.751060820367751,
      "grad_norm": 0.9648807048797607,
      "learning_rate": 1.186395120395514e-05,
      "loss": 0.1574,
      "num_input_tokens_seen": 23419696,
      "step": 27045
    },
    {
      "epoch": 12.753418198962754,
      "grad_norm": 0.9219727516174316,
      "learning_rate": 1.1855599218436283e-05,
      "loss": 0.2647,
      "num_input_tokens_seen": 23423648,
      "step": 27050
    },
    {
      "epoch": 12.755775577557756,
      "grad_norm": 0.542502760887146,
      "learning_rate": 1.1847249259949209e-05,
      "loss": 0.19,
      "num_input_tokens_seen": 23428000,
      "step": 27055
    },
    {
      "epoch": 12.758132956152759,
      "grad_norm": 0.8641098737716675,
      "learning_rate": 1.1838901329781574e-05,
      "loss": 0.1368,
      "num_input_tokens_seen": 23432368,
      "step": 27060
    },
    {
      "epoch": 12.760490334747761,
      "grad_norm": 0.6633122563362122,
      "learning_rate": 1.1830555429220758e-05,
      "loss": 0.1304,
      "num_input_tokens_seen": 23436848,
      "step": 27065
    },
    {
      "epoch": 12.762847713342763,
      "grad_norm": 1.7373515367507935,
      "learning_rate": 1.1822211559553784e-05,
      "loss": 0.1955,
      "num_input_tokens_seen": 23441152,
      "step": 27070
    },
    {
      "epoch": 12.765205091937766,
      "grad_norm": 0.5249028205871582,
      "learning_rate": 1.18138697220674e-05,
      "loss": 0.1614,
      "num_input_tokens_seen": 23445696,
      "step": 27075
    },
    {
      "epoch": 12.767562470532768,
      "grad_norm": 1.1174407005310059,
      "learning_rate": 1.1805529918048e-05,
      "loss": 0.2053,
      "num_input_tokens_seen": 23449888,
      "step": 27080
    },
    {
      "epoch": 12.76991984912777,
      "grad_norm": 1.460530161857605,
      "learning_rate": 1.1797192148781702e-05,
      "loss": 0.1466,
      "num_input_tokens_seen": 23455232,
      "step": 27085
    },
    {
      "epoch": 12.772277227722773,
      "grad_norm": 0.82078617811203,
      "learning_rate": 1.1788856415554297e-05,
      "loss": 0.1479,
      "num_input_tokens_seen": 23460448,
      "step": 27090
    },
    {
      "epoch": 12.774634606317775,
      "grad_norm": 0.4641667902469635,
      "learning_rate": 1.1780522719651249e-05,
      "loss": 0.184,
      "num_input_tokens_seen": 23465152,
      "step": 27095
    },
    {
      "epoch": 12.776991984912778,
      "grad_norm": 0.7296306490898132,
      "learning_rate": 1.1772191062357721e-05,
      "loss": 0.1397,
      "num_input_tokens_seen": 23470240,
      "step": 27100
    },
    {
      "epoch": 12.77934936350778,
      "grad_norm": 1.323307752609253,
      "learning_rate": 1.1763861444958573e-05,
      "loss": 0.1879,
      "num_input_tokens_seen": 23474368,
      "step": 27105
    },
    {
      "epoch": 12.78170674210278,
      "grad_norm": 0.5877262353897095,
      "learning_rate": 1.1755533868738317e-05,
      "loss": 0.1341,
      "num_input_tokens_seen": 23478880,
      "step": 27110
    },
    {
      "epoch": 12.784064120697785,
      "grad_norm": 0.8572908043861389,
      "learning_rate": 1.1747208334981185e-05,
      "loss": 0.1742,
      "num_input_tokens_seen": 23483744,
      "step": 27115
    },
    {
      "epoch": 12.786421499292786,
      "grad_norm": 1.024013876914978,
      "learning_rate": 1.1738884844971067e-05,
      "loss": 0.1313,
      "num_input_tokens_seen": 23487248,
      "step": 27120
    },
    {
      "epoch": 12.788778877887788,
      "grad_norm": 0.9016984701156616,
      "learning_rate": 1.1730563399991563e-05,
      "loss": 0.1815,
      "num_input_tokens_seen": 23490976,
      "step": 27125
    },
    {
      "epoch": 12.79113625648279,
      "grad_norm": 0.8138012290000916,
      "learning_rate": 1.1722244001325938e-05,
      "loss": 0.1229,
      "num_input_tokens_seen": 23495472,
      "step": 27130
    },
    {
      "epoch": 12.793493635077793,
      "grad_norm": 0.7577952742576599,
      "learning_rate": 1.1713926650257137e-05,
      "loss": 0.2428,
      "num_input_tokens_seen": 23498832,
      "step": 27135
    },
    {
      "epoch": 12.795851013672795,
      "grad_norm": 0.5881874561309814,
      "learning_rate": 1.170561134806781e-05,
      "loss": 0.1811,
      "num_input_tokens_seen": 23503216,
      "step": 27140
    },
    {
      "epoch": 12.798208392267798,
      "grad_norm": 2.475132942199707,
      "learning_rate": 1.1697298096040287e-05,
      "loss": 0.2949,
      "num_input_tokens_seen": 23507504,
      "step": 27145
    },
    {
      "epoch": 12.8005657708628,
      "grad_norm": 1.0155909061431885,
      "learning_rate": 1.1688986895456567e-05,
      "loss": 0.2173,
      "num_input_tokens_seen": 23511968,
      "step": 27150
    },
    {
      "epoch": 12.802923149457802,
      "grad_norm": 0.9310283660888672,
      "learning_rate": 1.1680677747598349e-05,
      "loss": 0.1669,
      "num_input_tokens_seen": 23516080,
      "step": 27155
    },
    {
      "epoch": 12.805280528052805,
      "grad_norm": 2.2100937366485596,
      "learning_rate": 1.1672370653746995e-05,
      "loss": 0.2452,
      "num_input_tokens_seen": 23520768,
      "step": 27160
    },
    {
      "epoch": 12.807637906647807,
      "grad_norm": 1.6786342859268188,
      "learning_rate": 1.166406561518357e-05,
      "loss": 0.2005,
      "num_input_tokens_seen": 23524736,
      "step": 27165
    },
    {
      "epoch": 12.80999528524281,
      "grad_norm": 0.45267799496650696,
      "learning_rate": 1.1655762633188826e-05,
      "loss": 0.2341,
      "num_input_tokens_seen": 23529264,
      "step": 27170
    },
    {
      "epoch": 12.812352663837812,
      "grad_norm": 0.5198317766189575,
      "learning_rate": 1.1647461709043172e-05,
      "loss": 0.0827,
      "num_input_tokens_seen": 23533248,
      "step": 27175
    },
    {
      "epoch": 12.814710042432814,
      "grad_norm": 1.5310728549957275,
      "learning_rate": 1.1639162844026722e-05,
      "loss": 0.1758,
      "num_input_tokens_seen": 23537536,
      "step": 27180
    },
    {
      "epoch": 12.817067421027817,
      "grad_norm": 0.9619582295417786,
      "learning_rate": 1.163086603941927e-05,
      "loss": 0.1965,
      "num_input_tokens_seen": 23541824,
      "step": 27185
    },
    {
      "epoch": 12.81942479962282,
      "grad_norm": 0.9594809412956238,
      "learning_rate": 1.1622571296500273e-05,
      "loss": 0.1938,
      "num_input_tokens_seen": 23545600,
      "step": 27190
    },
    {
      "epoch": 12.821782178217822,
      "grad_norm": 0.6648644804954529,
      "learning_rate": 1.1614278616548904e-05,
      "loss": 0.1595,
      "num_input_tokens_seen": 23549280,
      "step": 27195
    },
    {
      "epoch": 12.824139556812824,
      "grad_norm": 0.9350987076759338,
      "learning_rate": 1.1605988000843986e-05,
      "loss": 0.165,
      "num_input_tokens_seen": 23553008,
      "step": 27200
    },
    {
      "epoch": 12.824139556812824,
      "eval_loss": 0.20093616843223572,
      "eval_runtime": 22.1282,
      "eval_samples_per_second": 42.615,
      "eval_steps_per_second": 21.33,
      "num_input_tokens_seen": 23553008,
      "step": 27200
    },
    {
      "epoch": 12.826496935407826,
      "grad_norm": 1.683773398399353,
      "learning_rate": 1.1597699450664028e-05,
      "loss": 0.1351,
      "num_input_tokens_seen": 23556704,
      "step": 27205
    },
    {
      "epoch": 12.828854314002829,
      "grad_norm": 0.6610498428344727,
      "learning_rate": 1.1589412967287252e-05,
      "loss": 0.1305,
      "num_input_tokens_seen": 23561360,
      "step": 27210
    },
    {
      "epoch": 12.831211692597831,
      "grad_norm": 2.741046905517578,
      "learning_rate": 1.1581128551991514e-05,
      "loss": 0.2126,
      "num_input_tokens_seen": 23565888,
      "step": 27215
    },
    {
      "epoch": 12.833569071192834,
      "grad_norm": 0.6933498978614807,
      "learning_rate": 1.1572846206054383e-05,
      "loss": 0.2432,
      "num_input_tokens_seen": 23570400,
      "step": 27220
    },
    {
      "epoch": 12.835926449787836,
      "grad_norm": 1.6352790594100952,
      "learning_rate": 1.1564565930753113e-05,
      "loss": 0.2417,
      "num_input_tokens_seen": 23574432,
      "step": 27225
    },
    {
      "epoch": 12.838283828382838,
      "grad_norm": 0.8534215092658997,
      "learning_rate": 1.1556287727364606e-05,
      "loss": 0.1436,
      "num_input_tokens_seen": 23578592,
      "step": 27230
    },
    {
      "epoch": 12.84064120697784,
      "grad_norm": 0.7130349278450012,
      "learning_rate": 1.1548011597165489e-05,
      "loss": 0.1937,
      "num_input_tokens_seen": 23582784,
      "step": 27235
    },
    {
      "epoch": 12.842998585572843,
      "grad_norm": 1.036177396774292,
      "learning_rate": 1.1539737541432019e-05,
      "loss": 0.2006,
      "num_input_tokens_seen": 23586944,
      "step": 27240
    },
    {
      "epoch": 12.845355964167846,
      "grad_norm": 0.7024903893470764,
      "learning_rate": 1.1531465561440174e-05,
      "loss": 0.1361,
      "num_input_tokens_seen": 23591328,
      "step": 27245
    },
    {
      "epoch": 12.847713342762848,
      "grad_norm": 1.420285701751709,
      "learning_rate": 1.1523195658465605e-05,
      "loss": 0.1904,
      "num_input_tokens_seen": 23595264,
      "step": 27250
    },
    {
      "epoch": 12.85007072135785,
      "grad_norm": 0.36987823247909546,
      "learning_rate": 1.1514927833783618e-05,
      "loss": 0.1809,
      "num_input_tokens_seen": 23598960,
      "step": 27255
    },
    {
      "epoch": 12.852428099952853,
      "grad_norm": 0.6678282618522644,
      "learning_rate": 1.150666208866922e-05,
      "loss": 0.2137,
      "num_input_tokens_seen": 23603184,
      "step": 27260
    },
    {
      "epoch": 12.854785478547855,
      "grad_norm": 1.1315950155258179,
      "learning_rate": 1.1498398424397106e-05,
      "loss": 0.1337,
      "num_input_tokens_seen": 23606944,
      "step": 27265
    },
    {
      "epoch": 12.857142857142858,
      "grad_norm": 0.9566529393196106,
      "learning_rate": 1.1490136842241628e-05,
      "loss": 0.2171,
      "num_input_tokens_seen": 23611808,
      "step": 27270
    },
    {
      "epoch": 12.85950023573786,
      "grad_norm": 0.6239786148071289,
      "learning_rate": 1.1481877343476813e-05,
      "loss": 0.2051,
      "num_input_tokens_seen": 23616368,
      "step": 27275
    },
    {
      "epoch": 12.861857614332862,
      "grad_norm": 1.7231799364089966,
      "learning_rate": 1.14736199293764e-05,
      "loss": 0.2005,
      "num_input_tokens_seen": 23620800,
      "step": 27280
    },
    {
      "epoch": 12.864214992927865,
      "grad_norm": 1.8313159942626953,
      "learning_rate": 1.1465364601213771e-05,
      "loss": 0.1663,
      "num_input_tokens_seen": 23625136,
      "step": 27285
    },
    {
      "epoch": 12.866572371522867,
      "grad_norm": 0.8890312910079956,
      "learning_rate": 1.1457111360262012e-05,
      "loss": 0.1478,
      "num_input_tokens_seen": 23628720,
      "step": 27290
    },
    {
      "epoch": 12.86892975011787,
      "grad_norm": 0.6682963371276855,
      "learning_rate": 1.1448860207793869e-05,
      "loss": 0.1224,
      "num_input_tokens_seen": 23632992,
      "step": 27295
    },
    {
      "epoch": 12.871287128712872,
      "grad_norm": 1.2853806018829346,
      "learning_rate": 1.144061114508177e-05,
      "loss": 0.2283,
      "num_input_tokens_seen": 23636832,
      "step": 27300
    },
    {
      "epoch": 12.873644507307874,
      "grad_norm": 1.0352184772491455,
      "learning_rate": 1.1432364173397842e-05,
      "loss": 0.1563,
      "num_input_tokens_seen": 23640752,
      "step": 27305
    },
    {
      "epoch": 12.876001885902877,
      "grad_norm": 0.6844066977500916,
      "learning_rate": 1.1424119294013852e-05,
      "loss": 0.1664,
      "num_input_tokens_seen": 23645184,
      "step": 27310
    },
    {
      "epoch": 12.878359264497877,
      "grad_norm": 1.7234524488449097,
      "learning_rate": 1.1415876508201279e-05,
      "loss": 0.1771,
      "num_input_tokens_seen": 23649392,
      "step": 27315
    },
    {
      "epoch": 12.88071664309288,
      "grad_norm": 1.2679685354232788,
      "learning_rate": 1.140763581723125e-05,
      "loss": 0.249,
      "num_input_tokens_seen": 23653520,
      "step": 27320
    },
    {
      "epoch": 12.883074021687882,
      "grad_norm": 0.7641607522964478,
      "learning_rate": 1.1399397222374588e-05,
      "loss": 0.1389,
      "num_input_tokens_seen": 23657696,
      "step": 27325
    },
    {
      "epoch": 12.885431400282885,
      "grad_norm": 0.8093212246894836,
      "learning_rate": 1.1391160724901804e-05,
      "loss": 0.1831,
      "num_input_tokens_seen": 23661856,
      "step": 27330
    },
    {
      "epoch": 12.887788778877887,
      "grad_norm": 1.7562134265899658,
      "learning_rate": 1.138292632608304e-05,
      "loss": 0.1646,
      "num_input_tokens_seen": 23665904,
      "step": 27335
    },
    {
      "epoch": 12.89014615747289,
      "grad_norm": 0.312921404838562,
      "learning_rate": 1.1374694027188174e-05,
      "loss": 0.154,
      "num_input_tokens_seen": 23670528,
      "step": 27340
    },
    {
      "epoch": 12.892503536067892,
      "grad_norm": 0.4898776710033417,
      "learning_rate": 1.1366463829486711e-05,
      "loss": 0.1186,
      "num_input_tokens_seen": 23674544,
      "step": 27345
    },
    {
      "epoch": 12.894860914662894,
      "grad_norm": 1.607582688331604,
      "learning_rate": 1.1358235734247849e-05,
      "loss": 0.2276,
      "num_input_tokens_seen": 23679680,
      "step": 27350
    },
    {
      "epoch": 12.897218293257897,
      "grad_norm": 0.5275335311889648,
      "learning_rate": 1.1350009742740478e-05,
      "loss": 0.1849,
      "num_input_tokens_seen": 23683472,
      "step": 27355
    },
    {
      "epoch": 12.899575671852899,
      "grad_norm": 1.422922134399414,
      "learning_rate": 1.134178585623313e-05,
      "loss": 0.1684,
      "num_input_tokens_seen": 23687216,
      "step": 27360
    },
    {
      "epoch": 12.901933050447902,
      "grad_norm": 1.0694712400436401,
      "learning_rate": 1.1333564075994047e-05,
      "loss": 0.2232,
      "num_input_tokens_seen": 23691392,
      "step": 27365
    },
    {
      "epoch": 12.904290429042904,
      "grad_norm": 0.607294499874115,
      "learning_rate": 1.1325344403291133e-05,
      "loss": 0.1747,
      "num_input_tokens_seen": 23695776,
      "step": 27370
    },
    {
      "epoch": 12.906647807637906,
      "grad_norm": 0.7849457263946533,
      "learning_rate": 1.1317126839391951e-05,
      "loss": 0.1982,
      "num_input_tokens_seen": 23700640,
      "step": 27375
    },
    {
      "epoch": 12.909005186232909,
      "grad_norm": 1.5891014337539673,
      "learning_rate": 1.1308911385563766e-05,
      "loss": 0.1674,
      "num_input_tokens_seen": 23705168,
      "step": 27380
    },
    {
      "epoch": 12.911362564827911,
      "grad_norm": 0.9237022995948792,
      "learning_rate": 1.1300698043073494e-05,
      "loss": 0.0999,
      "num_input_tokens_seen": 23709104,
      "step": 27385
    },
    {
      "epoch": 12.913719943422914,
      "grad_norm": 0.43907618522644043,
      "learning_rate": 1.1292486813187736e-05,
      "loss": 0.1653,
      "num_input_tokens_seen": 23713520,
      "step": 27390
    },
    {
      "epoch": 12.916077322017916,
      "grad_norm": 1.0114340782165527,
      "learning_rate": 1.1284277697172782e-05,
      "loss": 0.2175,
      "num_input_tokens_seen": 23718160,
      "step": 27395
    },
    {
      "epoch": 12.918434700612918,
      "grad_norm": 1.4826167821884155,
      "learning_rate": 1.127607069629456e-05,
      "loss": 0.2417,
      "num_input_tokens_seen": 23722608,
      "step": 27400
    },
    {
      "epoch": 12.918434700612918,
      "eval_loss": 0.20061133801937103,
      "eval_runtime": 22.1077,
      "eval_samples_per_second": 42.655,
      "eval_steps_per_second": 21.35,
      "num_input_tokens_seen": 23722608,
      "step": 27400
    },
    {
      "epoch": 12.92079207920792,
      "grad_norm": 1.6134570837020874,
      "learning_rate": 1.1267865811818701e-05,
      "loss": 0.141,
      "num_input_tokens_seen": 23727392,
      "step": 27405
    },
    {
      "epoch": 12.923149457802923,
      "grad_norm": 0.7828065752983093,
      "learning_rate": 1.1259663045010513e-05,
      "loss": 0.2074,
      "num_input_tokens_seen": 23733056,
      "step": 27410
    },
    {
      "epoch": 12.925506836397926,
      "grad_norm": 0.8193246722221375,
      "learning_rate": 1.1251462397134957e-05,
      "loss": 0.1015,
      "num_input_tokens_seen": 23737488,
      "step": 27415
    },
    {
      "epoch": 12.927864214992928,
      "grad_norm": 0.5876080393791199,
      "learning_rate": 1.1243263869456664e-05,
      "loss": 0.1794,
      "num_input_tokens_seen": 23741792,
      "step": 27420
    },
    {
      "epoch": 12.93022159358793,
      "grad_norm": 1.146461844444275,
      "learning_rate": 1.1235067463239967e-05,
      "loss": 0.1532,
      "num_input_tokens_seen": 23746512,
      "step": 27425
    },
    {
      "epoch": 12.932578972182933,
      "grad_norm": 0.8696014285087585,
      "learning_rate": 1.122687317974884e-05,
      "loss": 0.2061,
      "num_input_tokens_seen": 23751040,
      "step": 27430
    },
    {
      "epoch": 12.934936350777935,
      "grad_norm": 0.5576220154762268,
      "learning_rate": 1.1218681020246963e-05,
      "loss": 0.2544,
      "num_input_tokens_seen": 23755440,
      "step": 27435
    },
    {
      "epoch": 12.937293729372938,
      "grad_norm": 1.3345868587493896,
      "learning_rate": 1.1210490985997652e-05,
      "loss": 0.1802,
      "num_input_tokens_seen": 23759200,
      "step": 27440
    },
    {
      "epoch": 12.93965110796794,
      "grad_norm": 1.5760780572891235,
      "learning_rate": 1.1202303078263917e-05,
      "loss": 0.1746,
      "num_input_tokens_seen": 23763392,
      "step": 27445
    },
    {
      "epoch": 12.942008486562942,
      "grad_norm": 0.6845310926437378,
      "learning_rate": 1.1194117298308451e-05,
      "loss": 0.112,
      "num_input_tokens_seen": 23767696,
      "step": 27450
    },
    {
      "epoch": 12.944365865157945,
      "grad_norm": 1.1021873950958252,
      "learning_rate": 1.1185933647393585e-05,
      "loss": 0.2685,
      "num_input_tokens_seen": 23772096,
      "step": 27455
    },
    {
      "epoch": 12.946723243752947,
      "grad_norm": 1.0094454288482666,
      "learning_rate": 1.1177752126781354e-05,
      "loss": 0.1545,
      "num_input_tokens_seen": 23775344,
      "step": 27460
    },
    {
      "epoch": 12.94908062234795,
      "grad_norm": 1.9455740451812744,
      "learning_rate": 1.1169572737733441e-05,
      "loss": 0.1776,
      "num_input_tokens_seen": 23779792,
      "step": 27465
    },
    {
      "epoch": 12.951438000942952,
      "grad_norm": 0.5283916592597961,
      "learning_rate": 1.1161395481511216e-05,
      "loss": 0.1511,
      "num_input_tokens_seen": 23783744,
      "step": 27470
    },
    {
      "epoch": 12.953795379537954,
      "grad_norm": 1.7466456890106201,
      "learning_rate": 1.1153220359375722e-05,
      "loss": 0.2216,
      "num_input_tokens_seen": 23788208,
      "step": 27475
    },
    {
      "epoch": 12.956152758132957,
      "grad_norm": 1.7033188343048096,
      "learning_rate": 1.114504737258765e-05,
      "loss": 0.1508,
      "num_input_tokens_seen": 23792496,
      "step": 27480
    },
    {
      "epoch": 12.95851013672796,
      "grad_norm": 0.6537677049636841,
      "learning_rate": 1.1136876522407393e-05,
      "loss": 0.1929,
      "num_input_tokens_seen": 23797072,
      "step": 27485
    },
    {
      "epoch": 12.960867515322962,
      "grad_norm": 2.449319362640381,
      "learning_rate": 1.1128707810094985e-05,
      "loss": 0.2233,
      "num_input_tokens_seen": 23800976,
      "step": 27490
    },
    {
      "epoch": 12.963224893917964,
      "grad_norm": 1.3610725402832031,
      "learning_rate": 1.1120541236910157e-05,
      "loss": 0.2265,
      "num_input_tokens_seen": 23805072,
      "step": 27495
    },
    {
      "epoch": 12.965582272512966,
      "grad_norm": 1.2540276050567627,
      "learning_rate": 1.111237680411229e-05,
      "loss": 0.1734,
      "num_input_tokens_seen": 23809424,
      "step": 27500
    },
    {
      "epoch": 12.967939651107969,
      "grad_norm": 0.7843773365020752,
      "learning_rate": 1.1104214512960433e-05,
      "loss": 0.1332,
      "num_input_tokens_seen": 23813184,
      "step": 27505
    },
    {
      "epoch": 12.97029702970297,
      "grad_norm": 1.3611537218093872,
      "learning_rate": 1.1096054364713327e-05,
      "loss": 0.1478,
      "num_input_tokens_seen": 23817008,
      "step": 27510
    },
    {
      "epoch": 12.972654408297974,
      "grad_norm": 0.4272540211677551,
      "learning_rate": 1.1087896360629371e-05,
      "loss": 0.1979,
      "num_input_tokens_seen": 23821856,
      "step": 27515
    },
    {
      "epoch": 12.975011786892974,
      "grad_norm": 1.7515548467636108,
      "learning_rate": 1.107974050196662e-05,
      "loss": 0.2613,
      "num_input_tokens_seen": 23826368,
      "step": 27520
    },
    {
      "epoch": 12.977369165487977,
      "grad_norm": 1.3281610012054443,
      "learning_rate": 1.1071586789982816e-05,
      "loss": 0.1312,
      "num_input_tokens_seen": 23830928,
      "step": 27525
    },
    {
      "epoch": 12.979726544082979,
      "grad_norm": 1.5098129510879517,
      "learning_rate": 1.1063435225935373e-05,
      "loss": 0.2322,
      "num_input_tokens_seen": 23835248,
      "step": 27530
    },
    {
      "epoch": 12.982083922677981,
      "grad_norm": 0.7442829608917236,
      "learning_rate": 1.1055285811081348e-05,
      "loss": 0.1612,
      "num_input_tokens_seen": 23838992,
      "step": 27535
    },
    {
      "epoch": 12.984441301272984,
      "grad_norm": 0.7673740386962891,
      "learning_rate": 1.1047138546677499e-05,
      "loss": 0.1416,
      "num_input_tokens_seen": 23843072,
      "step": 27540
    },
    {
      "epoch": 12.986798679867986,
      "grad_norm": 0.8848484754562378,
      "learning_rate": 1.1038993433980219e-05,
      "loss": 0.1817,
      "num_input_tokens_seen": 23847296,
      "step": 27545
    },
    {
      "epoch": 12.989156058462989,
      "grad_norm": 0.8497651815414429,
      "learning_rate": 1.1030850474245597e-05,
      "loss": 0.188,
      "num_input_tokens_seen": 23851488,
      "step": 27550
    },
    {
      "epoch": 12.991513437057991,
      "grad_norm": 1.0012836456298828,
      "learning_rate": 1.102270966872939e-05,
      "loss": 0.1821,
      "num_input_tokens_seen": 23855248,
      "step": 27555
    },
    {
      "epoch": 12.993870815652993,
      "grad_norm": 2.6551806926727295,
      "learning_rate": 1.1014571018687e-05,
      "loss": 0.1738,
      "num_input_tokens_seen": 23860064,
      "step": 27560
    },
    {
      "epoch": 12.996228194247996,
      "grad_norm": 0.5560300946235657,
      "learning_rate": 1.1006434525373502e-05,
      "loss": 0.2091,
      "num_input_tokens_seen": 23864064,
      "step": 27565
    },
    {
      "epoch": 12.998585572842998,
      "grad_norm": 0.6764346957206726,
      "learning_rate": 1.0998300190043664e-05,
      "loss": 0.1173,
      "num_input_tokens_seen": 23867856,
      "step": 27570
    },
    {
      "epoch": 13.000942951438,
      "grad_norm": 0.9117735028266907,
      "learning_rate": 1.0990168013951882e-05,
      "loss": 0.2401,
      "num_input_tokens_seen": 23872560,
      "step": 27575
    },
    {
      "epoch": 13.003300330033003,
      "grad_norm": 0.8273669481277466,
      "learning_rate": 1.0982037998352263e-05,
      "loss": 0.1592,
      "num_input_tokens_seen": 23877344,
      "step": 27580
    },
    {
      "epoch": 13.005657708628005,
      "grad_norm": 0.7092344164848328,
      "learning_rate": 1.0973910144498534e-05,
      "loss": 0.1883,
      "num_input_tokens_seen": 23881312,
      "step": 27585
    },
    {
      "epoch": 13.008015087223008,
      "grad_norm": 0.5444493293762207,
      "learning_rate": 1.0965784453644123e-05,
      "loss": 0.1117,
      "num_input_tokens_seen": 23885248,
      "step": 27590
    },
    {
      "epoch": 13.01037246581801,
      "grad_norm": 0.8501327037811279,
      "learning_rate": 1.0957660927042127e-05,
      "loss": 0.1865,
      "num_input_tokens_seen": 23889312,
      "step": 27595
    },
    {
      "epoch": 13.012729844413013,
      "grad_norm": 0.8015831112861633,
      "learning_rate": 1.094953956594527e-05,
      "loss": 0.188,
      "num_input_tokens_seen": 23892928,
      "step": 27600
    },
    {
      "epoch": 13.012729844413013,
      "eval_loss": 0.20131340622901917,
      "eval_runtime": 22.1962,
      "eval_samples_per_second": 42.485,
      "eval_steps_per_second": 21.265,
      "num_input_tokens_seen": 23892928,
      "step": 27600
    },
    {
      "epoch": 13.015087223008015,
      "grad_norm": 1.3529324531555176,
      "learning_rate": 1.0941420371605981e-05,
      "loss": 0.189,
      "num_input_tokens_seen": 23897472,
      "step": 27605
    },
    {
      "epoch": 13.017444601603017,
      "grad_norm": 1.5583782196044922,
      "learning_rate": 1.0933303345276354e-05,
      "loss": 0.1331,
      "num_input_tokens_seen": 23901392,
      "step": 27610
    },
    {
      "epoch": 13.01980198019802,
      "grad_norm": 0.6920907497406006,
      "learning_rate": 1.0925188488208112e-05,
      "loss": 0.1689,
      "num_input_tokens_seen": 23905616,
      "step": 27615
    },
    {
      "epoch": 13.022159358793022,
      "grad_norm": 0.6389724612236023,
      "learning_rate": 1.0917075801652694e-05,
      "loss": 0.1445,
      "num_input_tokens_seen": 23909008,
      "step": 27620
    },
    {
      "epoch": 13.024516737388025,
      "grad_norm": 0.5595930814743042,
      "learning_rate": 1.0908965286861151e-05,
      "loss": 0.1888,
      "num_input_tokens_seen": 23912496,
      "step": 27625
    },
    {
      "epoch": 13.026874115983027,
      "grad_norm": 3.4142115116119385,
      "learning_rate": 1.090085694508425e-05,
      "loss": 0.2749,
      "num_input_tokens_seen": 23916320,
      "step": 27630
    },
    {
      "epoch": 13.02923149457803,
      "grad_norm": 0.5258936882019043,
      "learning_rate": 1.089275077757238e-05,
      "loss": 0.2525,
      "num_input_tokens_seen": 23920992,
      "step": 27635
    },
    {
      "epoch": 13.031588873173032,
      "grad_norm": 1.0078493356704712,
      "learning_rate": 1.0884646785575633e-05,
      "loss": 0.1779,
      "num_input_tokens_seen": 23925792,
      "step": 27640
    },
    {
      "epoch": 13.033946251768034,
      "grad_norm": 1.0769203901290894,
      "learning_rate": 1.0876544970343728e-05,
      "loss": 0.231,
      "num_input_tokens_seen": 23930672,
      "step": 27645
    },
    {
      "epoch": 13.036303630363037,
      "grad_norm": 0.9273089170455933,
      "learning_rate": 1.0868445333126082e-05,
      "loss": 0.1985,
      "num_input_tokens_seen": 23934848,
      "step": 27650
    },
    {
      "epoch": 13.038661008958039,
      "grad_norm": 0.5302481055259705,
      "learning_rate": 1.0860347875171745e-05,
      "loss": 0.2038,
      "num_input_tokens_seen": 23939488,
      "step": 27655
    },
    {
      "epoch": 13.041018387553041,
      "grad_norm": 1.1377607583999634,
      "learning_rate": 1.0852252597729465e-05,
      "loss": 0.2082,
      "num_input_tokens_seen": 23943120,
      "step": 27660
    },
    {
      "epoch": 13.043375766148044,
      "grad_norm": 0.9543406963348389,
      "learning_rate": 1.0844159502047615e-05,
      "loss": 0.1894,
      "num_input_tokens_seen": 23947728,
      "step": 27665
    },
    {
      "epoch": 13.045733144743046,
      "grad_norm": 0.6796595454216003,
      "learning_rate": 1.0836068589374265e-05,
      "loss": 0.2152,
      "num_input_tokens_seen": 23952336,
      "step": 27670
    },
    {
      "epoch": 13.048090523338049,
      "grad_norm": 2.058438777923584,
      "learning_rate": 1.0827979860957144e-05,
      "loss": 0.1846,
      "num_input_tokens_seen": 23956560,
      "step": 27675
    },
    {
      "epoch": 13.050447901933051,
      "grad_norm": 0.6698653697967529,
      "learning_rate": 1.0819893318043615e-05,
      "loss": 0.2295,
      "num_input_tokens_seen": 23961408,
      "step": 27680
    },
    {
      "epoch": 13.052805280528053,
      "grad_norm": 0.8930062055587769,
      "learning_rate": 1.0811808961880734e-05,
      "loss": 0.2175,
      "num_input_tokens_seen": 23965504,
      "step": 27685
    },
    {
      "epoch": 13.055162659123056,
      "grad_norm": 1.7735358476638794,
      "learning_rate": 1.080372679371522e-05,
      "loss": 0.1798,
      "num_input_tokens_seen": 23969472,
      "step": 27690
    },
    {
      "epoch": 13.057520037718058,
      "grad_norm": 0.5494157671928406,
      "learning_rate": 1.0795646814793428e-05,
      "loss": 0.212,
      "num_input_tokens_seen": 23973680,
      "step": 27695
    },
    {
      "epoch": 13.05987741631306,
      "grad_norm": 0.40645527839660645,
      "learning_rate": 1.078756902636141e-05,
      "loss": 0.1505,
      "num_input_tokens_seen": 23978592,
      "step": 27700
    },
    {
      "epoch": 13.062234794908063,
      "grad_norm": 0.48930442333221436,
      "learning_rate": 1.077949342966485e-05,
      "loss": 0.2156,
      "num_input_tokens_seen": 23983520,
      "step": 27705
    },
    {
      "epoch": 13.064592173503065,
      "grad_norm": 1.4690663814544678,
      "learning_rate": 1.0771420025949103e-05,
      "loss": 0.2026,
      "num_input_tokens_seen": 23987232,
      "step": 27710
    },
    {
      "epoch": 13.066949552098066,
      "grad_norm": 0.9860767126083374,
      "learning_rate": 1.0763348816459204e-05,
      "loss": 0.2731,
      "num_input_tokens_seen": 23991632,
      "step": 27715
    },
    {
      "epoch": 13.069306930693068,
      "grad_norm": 1.4303737878799438,
      "learning_rate": 1.0755279802439816e-05,
      "loss": 0.1698,
      "num_input_tokens_seen": 23996000,
      "step": 27720
    },
    {
      "epoch": 13.07166430928807,
      "grad_norm": 1.244262933731079,
      "learning_rate": 1.0747212985135293e-05,
      "loss": 0.1764,
      "num_input_tokens_seen": 23999632,
      "step": 27725
    },
    {
      "epoch": 13.074021687883073,
      "grad_norm": 1.7880924940109253,
      "learning_rate": 1.073914836578965e-05,
      "loss": 0.2088,
      "num_input_tokens_seen": 24004368,
      "step": 27730
    },
    {
      "epoch": 13.076379066478076,
      "grad_norm": 1.2996134757995605,
      "learning_rate": 1.0731085945646529e-05,
      "loss": 0.0965,
      "num_input_tokens_seen": 24008608,
      "step": 27735
    },
    {
      "epoch": 13.078736445073078,
      "grad_norm": 0.7290377020835876,
      "learning_rate": 1.0723025725949285e-05,
      "loss": 0.1449,
      "num_input_tokens_seen": 24013280,
      "step": 27740
    },
    {
      "epoch": 13.08109382366808,
      "grad_norm": 1.3243656158447266,
      "learning_rate": 1.0714967707940875e-05,
      "loss": 0.1761,
      "num_input_tokens_seen": 24017664,
      "step": 27745
    },
    {
      "epoch": 13.083451202263083,
      "grad_norm": 0.5582170486450195,
      "learning_rate": 1.0706911892863963e-05,
      "loss": 0.16,
      "num_input_tokens_seen": 24022112,
      "step": 27750
    },
    {
      "epoch": 13.085808580858085,
      "grad_norm": 0.6216561794281006,
      "learning_rate": 1.0698858281960866e-05,
      "loss": 0.2203,
      "num_input_tokens_seen": 24026032,
      "step": 27755
    },
    {
      "epoch": 13.088165959453088,
      "grad_norm": 1.5715514421463013,
      "learning_rate": 1.069080687647353e-05,
      "loss": 0.1904,
      "num_input_tokens_seen": 24030272,
      "step": 27760
    },
    {
      "epoch": 13.09052333804809,
      "grad_norm": 0.5989189743995667,
      "learning_rate": 1.0682757677643596e-05,
      "loss": 0.1527,
      "num_input_tokens_seen": 24034352,
      "step": 27765
    },
    {
      "epoch": 13.092880716643092,
      "grad_norm": 1.1891658306121826,
      "learning_rate": 1.0674710686712359e-05,
      "loss": 0.1776,
      "num_input_tokens_seen": 24038336,
      "step": 27770
    },
    {
      "epoch": 13.095238095238095,
      "grad_norm": 2.000013589859009,
      "learning_rate": 1.0666665904920756e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 24042240,
      "step": 27775
    },
    {
      "epoch": 13.097595473833097,
      "grad_norm": 0.6898412108421326,
      "learning_rate": 1.0658623333509385e-05,
      "loss": 0.1827,
      "num_input_tokens_seen": 24046496,
      "step": 27780
    },
    {
      "epoch": 13.0999528524281,
      "grad_norm": 0.7050005793571472,
      "learning_rate": 1.0650582973718532e-05,
      "loss": 0.148,
      "num_input_tokens_seen": 24050800,
      "step": 27785
    },
    {
      "epoch": 13.102310231023102,
      "grad_norm": 1.5260318517684937,
      "learning_rate": 1.0642544826788098e-05,
      "loss": 0.1935,
      "num_input_tokens_seen": 24055104,
      "step": 27790
    },
    {
      "epoch": 13.104667609618105,
      "grad_norm": 0.558136522769928,
      "learning_rate": 1.063450889395769e-05,
      "loss": 0.174,
      "num_input_tokens_seen": 24059456,
      "step": 27795
    },
    {
      "epoch": 13.107024988213107,
      "grad_norm": 0.5067992210388184,
      "learning_rate": 1.062647517646653e-05,
      "loss": 0.1443,
      "num_input_tokens_seen": 24063632,
      "step": 27800
    },
    {
      "epoch": 13.107024988213107,
      "eval_loss": 0.2010423243045807,
      "eval_runtime": 22.1445,
      "eval_samples_per_second": 42.584,
      "eval_steps_per_second": 21.315,
      "num_input_tokens_seen": 24063632,
      "step": 27800
    },
    {
      "epoch": 13.10938236680811,
      "grad_norm": 0.7050521373748779,
      "learning_rate": 1.0618443675553527e-05,
      "loss": 0.1671,
      "num_input_tokens_seen": 24068320,
      "step": 27805
    },
    {
      "epoch": 13.111739745403112,
      "grad_norm": 2.6938657760620117,
      "learning_rate": 1.0610414392457247e-05,
      "loss": 0.1749,
      "num_input_tokens_seen": 24073200,
      "step": 27810
    },
    {
      "epoch": 13.114097123998114,
      "grad_norm": 0.5324445366859436,
      "learning_rate": 1.0602387328415888e-05,
      "loss": 0.1447,
      "num_input_tokens_seen": 24077280,
      "step": 27815
    },
    {
      "epoch": 13.116454502593117,
      "grad_norm": 1.4892191886901855,
      "learning_rate": 1.0594362484667347e-05,
      "loss": 0.173,
      "num_input_tokens_seen": 24080848,
      "step": 27820
    },
    {
      "epoch": 13.118811881188119,
      "grad_norm": 1.1288845539093018,
      "learning_rate": 1.0586339862449132e-05,
      "loss": 0.2013,
      "num_input_tokens_seen": 24085520,
      "step": 27825
    },
    {
      "epoch": 13.121169259783121,
      "grad_norm": 1.3886168003082275,
      "learning_rate": 1.0578319462998445e-05,
      "loss": 0.2012,
      "num_input_tokens_seen": 24090064,
      "step": 27830
    },
    {
      "epoch": 13.123526638378124,
      "grad_norm": 1.2895125150680542,
      "learning_rate": 1.057030128755214e-05,
      "loss": 0.1573,
      "num_input_tokens_seen": 24093952,
      "step": 27835
    },
    {
      "epoch": 13.125884016973126,
      "grad_norm": 1.2502192258834839,
      "learning_rate": 1.0562285337346703e-05,
      "loss": 0.1309,
      "num_input_tokens_seen": 24098176,
      "step": 27840
    },
    {
      "epoch": 13.128241395568129,
      "grad_norm": 0.9424232244491577,
      "learning_rate": 1.0554271613618308e-05,
      "loss": 0.1727,
      "num_input_tokens_seen": 24102176,
      "step": 27845
    },
    {
      "epoch": 13.130598774163131,
      "grad_norm": 0.7299245595932007,
      "learning_rate": 1.054626011760276e-05,
      "loss": 0.1786,
      "num_input_tokens_seen": 24106960,
      "step": 27850
    },
    {
      "epoch": 13.132956152758133,
      "grad_norm": 0.8267210721969604,
      "learning_rate": 1.0538250850535549e-05,
      "loss": 0.1812,
      "num_input_tokens_seen": 24110720,
      "step": 27855
    },
    {
      "epoch": 13.135313531353136,
      "grad_norm": 0.6941401958465576,
      "learning_rate": 1.0530243813651794e-05,
      "loss": 0.2147,
      "num_input_tokens_seen": 24115232,
      "step": 27860
    },
    {
      "epoch": 13.137670909948138,
      "grad_norm": 2.0757734775543213,
      "learning_rate": 1.0522239008186271e-05,
      "loss": 0.2026,
      "num_input_tokens_seen": 24119712,
      "step": 27865
    },
    {
      "epoch": 13.14002828854314,
      "grad_norm": 0.9672611355781555,
      "learning_rate": 1.0514236435373434e-05,
      "loss": 0.2172,
      "num_input_tokens_seen": 24123504,
      "step": 27870
    },
    {
      "epoch": 13.142385667138143,
      "grad_norm": 0.538604736328125,
      "learning_rate": 1.0506236096447386e-05,
      "loss": 0.2187,
      "num_input_tokens_seen": 24127856,
      "step": 27875
    },
    {
      "epoch": 13.144743045733145,
      "grad_norm": 1.0479291677474976,
      "learning_rate": 1.049823799264186e-05,
      "loss": 0.1483,
      "num_input_tokens_seen": 24132704,
      "step": 27880
    },
    {
      "epoch": 13.147100424328148,
      "grad_norm": 0.6708202362060547,
      "learning_rate": 1.049024212519028e-05,
      "loss": 0.2848,
      "num_input_tokens_seen": 24137584,
      "step": 27885
    },
    {
      "epoch": 13.14945780292315,
      "grad_norm": 1.001341700553894,
      "learning_rate": 1.0482248495325713e-05,
      "loss": 0.1426,
      "num_input_tokens_seen": 24141840,
      "step": 27890
    },
    {
      "epoch": 13.151815181518153,
      "grad_norm": 1.285774827003479,
      "learning_rate": 1.047425710428086e-05,
      "loss": 0.2351,
      "num_input_tokens_seen": 24146512,
      "step": 27895
    },
    {
      "epoch": 13.154172560113155,
      "grad_norm": 0.5294156670570374,
      "learning_rate": 1.0466267953288114e-05,
      "loss": 0.1962,
      "num_input_tokens_seen": 24150976,
      "step": 27900
    },
    {
      "epoch": 13.156529938708157,
      "grad_norm": 1.1781935691833496,
      "learning_rate": 1.0458281043579482e-05,
      "loss": 0.1698,
      "num_input_tokens_seen": 24156128,
      "step": 27905
    },
    {
      "epoch": 13.15888731730316,
      "grad_norm": 0.8731048703193665,
      "learning_rate": 1.0450296376386657e-05,
      "loss": 0.1616,
      "num_input_tokens_seen": 24160096,
      "step": 27910
    },
    {
      "epoch": 13.16124469589816,
      "grad_norm": 1.0222082138061523,
      "learning_rate": 1.044231395294098e-05,
      "loss": 0.21,
      "num_input_tokens_seen": 24164624,
      "step": 27915
    },
    {
      "epoch": 13.163602074493163,
      "grad_norm": 1.0227346420288086,
      "learning_rate": 1.0434333774473435e-05,
      "loss": 0.1231,
      "num_input_tokens_seen": 24168416,
      "step": 27920
    },
    {
      "epoch": 13.165959453088165,
      "grad_norm": 0.9909266233444214,
      "learning_rate": 1.0426355842214657e-05,
      "loss": 0.1938,
      "num_input_tokens_seen": 24173216,
      "step": 27925
    },
    {
      "epoch": 13.168316831683168,
      "grad_norm": 0.7328718900680542,
      "learning_rate": 1.0418380157394963e-05,
      "loss": 0.2079,
      "num_input_tokens_seen": 24177040,
      "step": 27930
    },
    {
      "epoch": 13.17067421027817,
      "grad_norm": 0.5709068775177002,
      "learning_rate": 1.0410406721244281e-05,
      "loss": 0.1114,
      "num_input_tokens_seen": 24180880,
      "step": 27935
    },
    {
      "epoch": 13.173031588873172,
      "grad_norm": 0.8443426489830017,
      "learning_rate": 1.0402435534992238e-05,
      "loss": 0.1015,
      "num_input_tokens_seen": 24185136,
      "step": 27940
    },
    {
      "epoch": 13.175388967468175,
      "grad_norm": 1.3682795763015747,
      "learning_rate": 1.0394466599868071e-05,
      "loss": 0.2006,
      "num_input_tokens_seen": 24188800,
      "step": 27945
    },
    {
      "epoch": 13.177746346063177,
      "grad_norm": 1.040582537651062,
      "learning_rate": 1.0386499917100697e-05,
      "loss": 0.2333,
      "num_input_tokens_seen": 24192832,
      "step": 27950
    },
    {
      "epoch": 13.18010372465818,
      "grad_norm": 1.8141934871673584,
      "learning_rate": 1.0378535487918692e-05,
      "loss": 0.1744,
      "num_input_tokens_seen": 24197440,
      "step": 27955
    },
    {
      "epoch": 13.182461103253182,
      "grad_norm": 0.6265061497688293,
      "learning_rate": 1.037057331355025e-05,
      "loss": 0.1253,
      "num_input_tokens_seen": 24202592,
      "step": 27960
    },
    {
      "epoch": 13.184818481848184,
      "grad_norm": 0.5211203694343567,
      "learning_rate": 1.0362613395223247e-05,
      "loss": 0.1481,
      "num_input_tokens_seen": 24207568,
      "step": 27965
    },
    {
      "epoch": 13.187175860443187,
      "grad_norm": 0.8485364317893982,
      "learning_rate": 1.0354655734165212e-05,
      "loss": 0.1833,
      "num_input_tokens_seen": 24211664,
      "step": 27970
    },
    {
      "epoch": 13.18953323903819,
      "grad_norm": 1.5394909381866455,
      "learning_rate": 1.03467003316033e-05,
      "loss": 0.1459,
      "num_input_tokens_seen": 24215904,
      "step": 27975
    },
    {
      "epoch": 13.191890617633192,
      "grad_norm": 1.7944555282592773,
      "learning_rate": 1.033874718876435e-05,
      "loss": 0.185,
      "num_input_tokens_seen": 24220016,
      "step": 27980
    },
    {
      "epoch": 13.194247996228194,
      "grad_norm": 0.9833961129188538,
      "learning_rate": 1.0330796306874818e-05,
      "loss": 0.2015,
      "num_input_tokens_seen": 24224016,
      "step": 27985
    },
    {
      "epoch": 13.196605374823196,
      "grad_norm": 0.6602535247802734,
      "learning_rate": 1.032284768716085e-05,
      "loss": 0.184,
      "num_input_tokens_seen": 24228736,
      "step": 27990
    },
    {
      "epoch": 13.198962753418199,
      "grad_norm": 0.5274674296379089,
      "learning_rate": 1.0314901330848206e-05,
      "loss": 0.2055,
      "num_input_tokens_seen": 24232400,
      "step": 27995
    },
    {
      "epoch": 13.201320132013201,
      "grad_norm": 0.659539520740509,
      "learning_rate": 1.030695723916233e-05,
      "loss": 0.1565,
      "num_input_tokens_seen": 24237248,
      "step": 28000
    },
    {
      "epoch": 13.201320132013201,
      "eval_loss": 0.20160354673862457,
      "eval_runtime": 22.1661,
      "eval_samples_per_second": 42.543,
      "eval_steps_per_second": 21.294,
      "num_input_tokens_seen": 24237248,
      "step": 28000
    },
    {
      "epoch": 13.203677510608204,
      "grad_norm": 0.9328359365463257,
      "learning_rate": 1.0299015413328289e-05,
      "loss": 0.165,
      "num_input_tokens_seen": 24241456,
      "step": 28005
    },
    {
      "epoch": 13.206034889203206,
      "grad_norm": 0.6862813830375671,
      "learning_rate": 1.0291075854570809e-05,
      "loss": 0.2195,
      "num_input_tokens_seen": 24245712,
      "step": 28010
    },
    {
      "epoch": 13.208392267798208,
      "grad_norm": 2.192819118499756,
      "learning_rate": 1.0283138564114275e-05,
      "loss": 0.2262,
      "num_input_tokens_seen": 24249904,
      "step": 28015
    },
    {
      "epoch": 13.21074964639321,
      "grad_norm": 1.5891003608703613,
      "learning_rate": 1.027520354318273e-05,
      "loss": 0.1381,
      "num_input_tokens_seen": 24253808,
      "step": 28020
    },
    {
      "epoch": 13.213107024988213,
      "grad_norm": 2.297419786453247,
      "learning_rate": 1.0267270792999828e-05,
      "loss": 0.1607,
      "num_input_tokens_seen": 24257472,
      "step": 28025
    },
    {
      "epoch": 13.215464403583216,
      "grad_norm": 0.6509454250335693,
      "learning_rate": 1.0259340314788919e-05,
      "loss": 0.21,
      "num_input_tokens_seen": 24261040,
      "step": 28030
    },
    {
      "epoch": 13.217821782178218,
      "grad_norm": 2.609475612640381,
      "learning_rate": 1.0251412109772979e-05,
      "loss": 0.2777,
      "num_input_tokens_seen": 24264976,
      "step": 28035
    },
    {
      "epoch": 13.22017916077322,
      "grad_norm": 0.7409912943840027,
      "learning_rate": 1.0243486179174627e-05,
      "loss": 0.112,
      "num_input_tokens_seen": 24269920,
      "step": 28040
    },
    {
      "epoch": 13.222536539368223,
      "grad_norm": 0.6090894341468811,
      "learning_rate": 1.0235562524216158e-05,
      "loss": 0.1435,
      "num_input_tokens_seen": 24276144,
      "step": 28045
    },
    {
      "epoch": 13.224893917963225,
      "grad_norm": 1.9710640907287598,
      "learning_rate": 1.022764114611948e-05,
      "loss": 0.1911,
      "num_input_tokens_seen": 24280688,
      "step": 28050
    },
    {
      "epoch": 13.227251296558228,
      "grad_norm": 2.3551762104034424,
      "learning_rate": 1.0219722046106178e-05,
      "loss": 0.1698,
      "num_input_tokens_seen": 24285424,
      "step": 28055
    },
    {
      "epoch": 13.22960867515323,
      "grad_norm": 0.6129821538925171,
      "learning_rate": 1.0211805225397486e-05,
      "loss": 0.2057,
      "num_input_tokens_seen": 24289136,
      "step": 28060
    },
    {
      "epoch": 13.231966053748232,
      "grad_norm": 0.7593371868133545,
      "learning_rate": 1.020389068521426e-05,
      "loss": 0.1797,
      "num_input_tokens_seen": 24293072,
      "step": 28065
    },
    {
      "epoch": 13.234323432343235,
      "grad_norm": 0.81557697057724,
      "learning_rate": 1.0195978426777039e-05,
      "loss": 0.24,
      "num_input_tokens_seen": 24296752,
      "step": 28070
    },
    {
      "epoch": 13.236680810938237,
      "grad_norm": 0.8638442754745483,
      "learning_rate": 1.0188068451305982e-05,
      "loss": 0.1561,
      "num_input_tokens_seen": 24301920,
      "step": 28075
    },
    {
      "epoch": 13.23903818953324,
      "grad_norm": 1.0550647974014282,
      "learning_rate": 1.0180160760020902e-05,
      "loss": 0.1962,
      "num_input_tokens_seen": 24305952,
      "step": 28080
    },
    {
      "epoch": 13.241395568128242,
      "grad_norm": 1.1216741800308228,
      "learning_rate": 1.0172255354141278e-05,
      "loss": 0.1743,
      "num_input_tokens_seen": 24310016,
      "step": 28085
    },
    {
      "epoch": 13.243752946723244,
      "grad_norm": 0.6741437911987305,
      "learning_rate": 1.0164352234886205e-05,
      "loss": 0.1613,
      "num_input_tokens_seen": 24314544,
      "step": 28090
    },
    {
      "epoch": 13.246110325318247,
      "grad_norm": 1.2048219442367554,
      "learning_rate": 1.0156451403474454e-05,
      "loss": 0.2088,
      "num_input_tokens_seen": 24319616,
      "step": 28095
    },
    {
      "epoch": 13.24846770391325,
      "grad_norm": 1.0500798225402832,
      "learning_rate": 1.0148552861124443e-05,
      "loss": 0.1575,
      "num_input_tokens_seen": 24323568,
      "step": 28100
    },
    {
      "epoch": 13.250825082508252,
      "grad_norm": 0.9099064469337463,
      "learning_rate": 1.0140656609054205e-05,
      "loss": 0.1587,
      "num_input_tokens_seen": 24328016,
      "step": 28105
    },
    {
      "epoch": 13.253182461103254,
      "grad_norm": 0.5885050296783447,
      "learning_rate": 1.0132762648481455e-05,
      "loss": 0.1373,
      "num_input_tokens_seen": 24331904,
      "step": 28110
    },
    {
      "epoch": 13.255539839698255,
      "grad_norm": 0.37027114629745483,
      "learning_rate": 1.0124870980623543e-05,
      "loss": 0.1828,
      "num_input_tokens_seen": 24336368,
      "step": 28115
    },
    {
      "epoch": 13.257897218293257,
      "grad_norm": 0.8098682165145874,
      "learning_rate": 1.0116981606697453e-05,
      "loss": 0.1563,
      "num_input_tokens_seen": 24341824,
      "step": 28120
    },
    {
      "epoch": 13.26025459688826,
      "grad_norm": 1.197999119758606,
      "learning_rate": 1.0109094527919838e-05,
      "loss": 0.218,
      "num_input_tokens_seen": 24346336,
      "step": 28125
    },
    {
      "epoch": 13.262611975483262,
      "grad_norm": 1.6807754039764404,
      "learning_rate": 1.010120974550697e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 24350960,
      "step": 28130
    },
    {
      "epoch": 13.264969354078264,
      "grad_norm": 0.2597126066684723,
      "learning_rate": 1.0093327260674795e-05,
      "loss": 0.1539,
      "num_input_tokens_seen": 24355344,
      "step": 28135
    },
    {
      "epoch": 13.267326732673267,
      "grad_norm": 0.7466571927070618,
      "learning_rate": 1.0085447074638878e-05,
      "loss": 0.2667,
      "num_input_tokens_seen": 24359360,
      "step": 28140
    },
    {
      "epoch": 13.269684111268269,
      "grad_norm": 1.8485618829727173,
      "learning_rate": 1.0077569188614461e-05,
      "loss": 0.2221,
      "num_input_tokens_seen": 24363360,
      "step": 28145
    },
    {
      "epoch": 13.272041489863271,
      "grad_norm": 0.5231190323829651,
      "learning_rate": 1.0069693603816393e-05,
      "loss": 0.2201,
      "num_input_tokens_seen": 24367792,
      "step": 28150
    },
    {
      "epoch": 13.274398868458274,
      "grad_norm": 1.664719820022583,
      "learning_rate": 1.0061820321459204e-05,
      "loss": 0.2591,
      "num_input_tokens_seen": 24372528,
      "step": 28155
    },
    {
      "epoch": 13.276756247053276,
      "grad_norm": 1.3062454462051392,
      "learning_rate": 1.0053949342757038e-05,
      "loss": 0.1451,
      "num_input_tokens_seen": 24376768,
      "step": 28160
    },
    {
      "epoch": 13.279113625648279,
      "grad_norm": 0.5414755344390869,
      "learning_rate": 1.0046080668923717e-05,
      "loss": 0.1656,
      "num_input_tokens_seen": 24380896,
      "step": 28165
    },
    {
      "epoch": 13.281471004243281,
      "grad_norm": 1.4035749435424805,
      "learning_rate": 1.003821430117267e-05,
      "loss": 0.201,
      "num_input_tokens_seen": 24384912,
      "step": 28170
    },
    {
      "epoch": 13.283828382838283,
      "grad_norm": 0.8873651623725891,
      "learning_rate": 1.0030350240716999e-05,
      "loss": 0.1403,
      "num_input_tokens_seen": 24389264,
      "step": 28175
    },
    {
      "epoch": 13.286185761433286,
      "grad_norm": 0.9271333813667297,
      "learning_rate": 1.0022488488769449e-05,
      "loss": 0.191,
      "num_input_tokens_seen": 24393584,
      "step": 28180
    },
    {
      "epoch": 13.288543140028288,
      "grad_norm": 1.1627895832061768,
      "learning_rate": 1.0014629046542387e-05,
      "loss": 0.183,
      "num_input_tokens_seen": 24398896,
      "step": 28185
    },
    {
      "epoch": 13.29090051862329,
      "grad_norm": 0.9785323739051819,
      "learning_rate": 1.0006771915247842e-05,
      "loss": 0.1348,
      "num_input_tokens_seen": 24403440,
      "step": 28190
    },
    {
      "epoch": 13.293257897218293,
      "grad_norm": 0.6676890850067139,
      "learning_rate": 9.998917096097495e-06,
      "loss": 0.1796,
      "num_input_tokens_seen": 24408000,
      "step": 28195
    },
    {
      "epoch": 13.295615275813295,
      "grad_norm": 0.7978713512420654,
      "learning_rate": 9.991064590302638e-06,
      "loss": 0.2199,
      "num_input_tokens_seen": 24411712,
      "step": 28200
    },
    {
      "epoch": 13.295615275813295,
      "eval_loss": 0.20028379559516907,
      "eval_runtime": 22.1731,
      "eval_samples_per_second": 42.529,
      "eval_steps_per_second": 21.287,
      "num_input_tokens_seen": 24411712,
      "step": 28200
    },
    {
      "epoch": 13.297972654408298,
      "grad_norm": 0.9058188199996948,
      "learning_rate": 9.983214399074241e-06,
      "loss": 0.1372,
      "num_input_tokens_seen": 24415824,
      "step": 28205
    },
    {
      "epoch": 13.3003300330033,
      "grad_norm": 0.7457078099250793,
      "learning_rate": 9.975366523622893e-06,
      "loss": 0.177,
      "num_input_tokens_seen": 24420096,
      "step": 28210
    },
    {
      "epoch": 13.302687411598303,
      "grad_norm": 1.4469361305236816,
      "learning_rate": 9.967520965158841e-06,
      "loss": 0.161,
      "num_input_tokens_seen": 24424336,
      "step": 28215
    },
    {
      "epoch": 13.305044790193305,
      "grad_norm": 1.1418706178665161,
      "learning_rate": 9.95967772489197e-06,
      "loss": 0.1338,
      "num_input_tokens_seen": 24427904,
      "step": 28220
    },
    {
      "epoch": 13.307402168788308,
      "grad_norm": 1.281305193901062,
      "learning_rate": 9.951836804031794e-06,
      "loss": 0.2108,
      "num_input_tokens_seen": 24432480,
      "step": 28225
    },
    {
      "epoch": 13.30975954738331,
      "grad_norm": 1.2007555961608887,
      "learning_rate": 9.943998203787489e-06,
      "loss": 0.1432,
      "num_input_tokens_seen": 24436688,
      "step": 28230
    },
    {
      "epoch": 13.312116925978312,
      "grad_norm": 0.4229377210140228,
      "learning_rate": 9.936161925367874e-06,
      "loss": 0.1458,
      "num_input_tokens_seen": 24440032,
      "step": 28235
    },
    {
      "epoch": 13.314474304573315,
      "grad_norm": 0.3066205084323883,
      "learning_rate": 9.928327969981386e-06,
      "loss": 0.183,
      "num_input_tokens_seen": 24444480,
      "step": 28240
    },
    {
      "epoch": 13.316831683168317,
      "grad_norm": 1.0683478116989136,
      "learning_rate": 9.920496338836135e-06,
      "loss": 0.1508,
      "num_input_tokens_seen": 24449104,
      "step": 28245
    },
    {
      "epoch": 13.31918906176332,
      "grad_norm": 1.5246034860610962,
      "learning_rate": 9.912667033139844e-06,
      "loss": 0.1968,
      "num_input_tokens_seen": 24452800,
      "step": 28250
    },
    {
      "epoch": 13.321546440358322,
      "grad_norm": 0.3334369659423828,
      "learning_rate": 9.904840054099893e-06,
      "loss": 0.0879,
      "num_input_tokens_seen": 24457200,
      "step": 28255
    },
    {
      "epoch": 13.323903818953324,
      "grad_norm": 2.276012897491455,
      "learning_rate": 9.897015402923312e-06,
      "loss": 0.2195,
      "num_input_tokens_seen": 24461776,
      "step": 28260
    },
    {
      "epoch": 13.326261197548327,
      "grad_norm": 1.2561160326004028,
      "learning_rate": 9.889193080816744e-06,
      "loss": 0.2329,
      "num_input_tokens_seen": 24465504,
      "step": 28265
    },
    {
      "epoch": 13.32861857614333,
      "grad_norm": 0.898218035697937,
      "learning_rate": 9.881373088986498e-06,
      "loss": 0.1566,
      "num_input_tokens_seen": 24470240,
      "step": 28270
    },
    {
      "epoch": 13.330975954738332,
      "grad_norm": 2.652449131011963,
      "learning_rate": 9.873555428638523e-06,
      "loss": 0.178,
      "num_input_tokens_seen": 24474624,
      "step": 28275
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 0.5519058704376221,
      "learning_rate": 9.865740100978383e-06,
      "loss": 0.2288,
      "num_input_tokens_seen": 24479024,
      "step": 28280
    },
    {
      "epoch": 13.335690711928336,
      "grad_norm": 0.7567720413208008,
      "learning_rate": 9.857927107211315e-06,
      "loss": 0.2149,
      "num_input_tokens_seen": 24483584,
      "step": 28285
    },
    {
      "epoch": 13.338048090523339,
      "grad_norm": 0.6141287088394165,
      "learning_rate": 9.850116448542177e-06,
      "loss": 0.1774,
      "num_input_tokens_seen": 24488208,
      "step": 28290
    },
    {
      "epoch": 13.340405469118341,
      "grad_norm": 0.9825791716575623,
      "learning_rate": 9.842308126175457e-06,
      "loss": 0.2242,
      "num_input_tokens_seen": 24492240,
      "step": 28295
    },
    {
      "epoch": 13.342762847713344,
      "grad_norm": 0.7281460165977478,
      "learning_rate": 9.834502141315315e-06,
      "loss": 0.1614,
      "num_input_tokens_seen": 24496848,
      "step": 28300
    },
    {
      "epoch": 13.345120226308346,
      "grad_norm": 2.8220176696777344,
      "learning_rate": 9.82669849516552e-06,
      "loss": 0.144,
      "num_input_tokens_seen": 24502480,
      "step": 28305
    },
    {
      "epoch": 13.347477604903348,
      "grad_norm": 0.7826404571533203,
      "learning_rate": 9.818897188929493e-06,
      "loss": 0.1957,
      "num_input_tokens_seen": 24506736,
      "step": 28310
    },
    {
      "epoch": 13.34983498349835,
      "grad_norm": 0.8763823509216309,
      "learning_rate": 9.811098223810309e-06,
      "loss": 0.1402,
      "num_input_tokens_seen": 24511040,
      "step": 28315
    },
    {
      "epoch": 13.352192362093351,
      "grad_norm": 0.721804678440094,
      "learning_rate": 9.803301601010641e-06,
      "loss": 0.237,
      "num_input_tokens_seen": 24514608,
      "step": 28320
    },
    {
      "epoch": 13.354549740688354,
      "grad_norm": 0.9232965707778931,
      "learning_rate": 9.795507321732853e-06,
      "loss": 0.1377,
      "num_input_tokens_seen": 24518368,
      "step": 28325
    },
    {
      "epoch": 13.356907119283356,
      "grad_norm": 1.4158849716186523,
      "learning_rate": 9.787715387178898e-06,
      "loss": 0.2387,
      "num_input_tokens_seen": 24522752,
      "step": 28330
    },
    {
      "epoch": 13.359264497878359,
      "grad_norm": 0.911292314529419,
      "learning_rate": 9.779925798550399e-06,
      "loss": 0.1506,
      "num_input_tokens_seen": 24526688,
      "step": 28335
    },
    {
      "epoch": 13.361621876473361,
      "grad_norm": 1.0203046798706055,
      "learning_rate": 9.772138557048619e-06,
      "loss": 0.176,
      "num_input_tokens_seen": 24531984,
      "step": 28340
    },
    {
      "epoch": 13.363979255068363,
      "grad_norm": 0.626774251461029,
      "learning_rate": 9.764353663874426e-06,
      "loss": 0.1937,
      "num_input_tokens_seen": 24536320,
      "step": 28345
    },
    {
      "epoch": 13.366336633663366,
      "grad_norm": 0.9966804385185242,
      "learning_rate": 9.756571120228375e-06,
      "loss": 0.1852,
      "num_input_tokens_seen": 24541600,
      "step": 28350
    },
    {
      "epoch": 13.368694012258368,
      "grad_norm": 0.473570317029953,
      "learning_rate": 9.748790927310605e-06,
      "loss": 0.2332,
      "num_input_tokens_seen": 24545600,
      "step": 28355
    },
    {
      "epoch": 13.37105139085337,
      "grad_norm": 0.7786096334457397,
      "learning_rate": 9.741013086320946e-06,
      "loss": 0.1121,
      "num_input_tokens_seen": 24549952,
      "step": 28360
    },
    {
      "epoch": 13.373408769448373,
      "grad_norm": 0.6334929466247559,
      "learning_rate": 9.733237598458821e-06,
      "loss": 0.1554,
      "num_input_tokens_seen": 24554560,
      "step": 28365
    },
    {
      "epoch": 13.375766148043375,
      "grad_norm": 0.597587525844574,
      "learning_rate": 9.725464464923308e-06,
      "loss": 0.1229,
      "num_input_tokens_seen": 24559040,
      "step": 28370
    },
    {
      "epoch": 13.378123526638378,
      "grad_norm": 1.2059310674667358,
      "learning_rate": 9.717693686913123e-06,
      "loss": 0.1416,
      "num_input_tokens_seen": 24563408,
      "step": 28375
    },
    {
      "epoch": 13.38048090523338,
      "grad_norm": 1.3422329425811768,
      "learning_rate": 9.709925265626632e-06,
      "loss": 0.226,
      "num_input_tokens_seen": 24567072,
      "step": 28380
    },
    {
      "epoch": 13.382838283828383,
      "grad_norm": 0.5382468104362488,
      "learning_rate": 9.702159202261801e-06,
      "loss": 0.1306,
      "num_input_tokens_seen": 24571824,
      "step": 28385
    },
    {
      "epoch": 13.385195662423385,
      "grad_norm": 1.0960047245025635,
      "learning_rate": 9.694395498016268e-06,
      "loss": 0.1879,
      "num_input_tokens_seen": 24576272,
      "step": 28390
    },
    {
      "epoch": 13.387553041018387,
      "grad_norm": 0.9997598528862,
      "learning_rate": 9.686634154087298e-06,
      "loss": 0.1668,
      "num_input_tokens_seen": 24580304,
      "step": 28395
    },
    {
      "epoch": 13.38991041961339,
      "grad_norm": 0.6945285201072693,
      "learning_rate": 9.678875171671776e-06,
      "loss": 0.2351,
      "num_input_tokens_seen": 24584800,
      "step": 28400
    },
    {
      "epoch": 13.38991041961339,
      "eval_loss": 0.2001599818468094,
      "eval_runtime": 22.1823,
      "eval_samples_per_second": 42.511,
      "eval_steps_per_second": 21.278,
      "num_input_tokens_seen": 24584800,
      "step": 28400
    },
    {
      "epoch": 13.392267798208392,
      "grad_norm": 2.5932514667510986,
      "learning_rate": 9.671118551966246e-06,
      "loss": 0.1503,
      "num_input_tokens_seen": 24589376,
      "step": 28405
    },
    {
      "epoch": 13.394625176803395,
      "grad_norm": 1.937033772468567,
      "learning_rate": 9.66336429616686e-06,
      "loss": 0.1688,
      "num_input_tokens_seen": 24593840,
      "step": 28410
    },
    {
      "epoch": 13.396982555398397,
      "grad_norm": 1.1788455247879028,
      "learning_rate": 9.655612405469436e-06,
      "loss": 0.2664,
      "num_input_tokens_seen": 24598752,
      "step": 28415
    },
    {
      "epoch": 13.3993399339934,
      "grad_norm": 1.4013874530792236,
      "learning_rate": 9.647862881069413e-06,
      "loss": 0.2201,
      "num_input_tokens_seen": 24602368,
      "step": 28420
    },
    {
      "epoch": 13.401697312588402,
      "grad_norm": 1.7941181659698486,
      "learning_rate": 9.640115724161855e-06,
      "loss": 0.1738,
      "num_input_tokens_seen": 24606480,
      "step": 28425
    },
    {
      "epoch": 13.404054691183404,
      "grad_norm": 0.523609459400177,
      "learning_rate": 9.632370935941483e-06,
      "loss": 0.1688,
      "num_input_tokens_seen": 24610544,
      "step": 28430
    },
    {
      "epoch": 13.406412069778407,
      "grad_norm": 0.9897738695144653,
      "learning_rate": 9.624628517602634e-06,
      "loss": 0.185,
      "num_input_tokens_seen": 24615680,
      "step": 28435
    },
    {
      "epoch": 13.408769448373409,
      "grad_norm": 1.4451619386672974,
      "learning_rate": 9.61688847033928e-06,
      "loss": 0.2594,
      "num_input_tokens_seen": 24620128,
      "step": 28440
    },
    {
      "epoch": 13.411126826968411,
      "grad_norm": 0.8476455211639404,
      "learning_rate": 9.609150795345051e-06,
      "loss": 0.1571,
      "num_input_tokens_seen": 24625072,
      "step": 28445
    },
    {
      "epoch": 13.413484205563414,
      "grad_norm": 0.7961642742156982,
      "learning_rate": 9.601415493813171e-06,
      "loss": 0.1579,
      "num_input_tokens_seen": 24628704,
      "step": 28450
    },
    {
      "epoch": 13.415841584158416,
      "grad_norm": 0.785173237323761,
      "learning_rate": 9.593682566936533e-06,
      "loss": 0.1953,
      "num_input_tokens_seen": 24633424,
      "step": 28455
    },
    {
      "epoch": 13.418198962753419,
      "grad_norm": 0.8481107950210571,
      "learning_rate": 9.58595201590766e-06,
      "loss": 0.1208,
      "num_input_tokens_seen": 24637136,
      "step": 28460
    },
    {
      "epoch": 13.420556341348421,
      "grad_norm": 0.534109354019165,
      "learning_rate": 9.578223841918681e-06,
      "loss": 0.1041,
      "num_input_tokens_seen": 24641072,
      "step": 28465
    },
    {
      "epoch": 13.422913719943423,
      "grad_norm": 1.4636683464050293,
      "learning_rate": 9.570498046161389e-06,
      "loss": 0.1524,
      "num_input_tokens_seen": 24644912,
      "step": 28470
    },
    {
      "epoch": 13.425271098538426,
      "grad_norm": 0.7956243753433228,
      "learning_rate": 9.562774629827206e-06,
      "loss": 0.192,
      "num_input_tokens_seen": 24649344,
      "step": 28475
    },
    {
      "epoch": 13.427628477133428,
      "grad_norm": 0.45447489619255066,
      "learning_rate": 9.555053594107163e-06,
      "loss": 0.1875,
      "num_input_tokens_seen": 24653808,
      "step": 28480
    },
    {
      "epoch": 13.42998585572843,
      "grad_norm": 0.6456600427627563,
      "learning_rate": 9.547334940191957e-06,
      "loss": 0.2205,
      "num_input_tokens_seen": 24658528,
      "step": 28485
    },
    {
      "epoch": 13.432343234323433,
      "grad_norm": 1.0157994031906128,
      "learning_rate": 9.539618669271886e-06,
      "loss": 0.1434,
      "num_input_tokens_seen": 24662848,
      "step": 28490
    },
    {
      "epoch": 13.434700612918435,
      "grad_norm": 1.1604782342910767,
      "learning_rate": 9.531904782536904e-06,
      "loss": 0.1351,
      "num_input_tokens_seen": 24667056,
      "step": 28495
    },
    {
      "epoch": 13.437057991513438,
      "grad_norm": 0.9643362164497375,
      "learning_rate": 9.524193281176597e-06,
      "loss": 0.2127,
      "num_input_tokens_seen": 24671360,
      "step": 28500
    },
    {
      "epoch": 13.43941537010844,
      "grad_norm": 0.8700309991836548,
      "learning_rate": 9.516484166380165e-06,
      "loss": 0.1915,
      "num_input_tokens_seen": 24676480,
      "step": 28505
    },
    {
      "epoch": 13.441772748703443,
      "grad_norm": 0.575862467288971,
      "learning_rate": 9.508777439336447e-06,
      "loss": 0.2038,
      "num_input_tokens_seen": 24680864,
      "step": 28510
    },
    {
      "epoch": 13.444130127298443,
      "grad_norm": 1.9263954162597656,
      "learning_rate": 9.50107310123393e-06,
      "loss": 0.1963,
      "num_input_tokens_seen": 24684944,
      "step": 28515
    },
    {
      "epoch": 13.446487505893446,
      "grad_norm": 0.9392013549804688,
      "learning_rate": 9.493371153260702e-06,
      "loss": 0.1681,
      "num_input_tokens_seen": 24688864,
      "step": 28520
    },
    {
      "epoch": 13.448844884488448,
      "grad_norm": 1.9231384992599487,
      "learning_rate": 9.485671596604523e-06,
      "loss": 0.1806,
      "num_input_tokens_seen": 24693632,
      "step": 28525
    },
    {
      "epoch": 13.45120226308345,
      "grad_norm": 0.8970178961753845,
      "learning_rate": 9.477974432452738e-06,
      "loss": 0.214,
      "num_input_tokens_seen": 24697456,
      "step": 28530
    },
    {
      "epoch": 13.453559641678453,
      "grad_norm": 0.6606221795082092,
      "learning_rate": 9.470279661992356e-06,
      "loss": 0.1996,
      "num_input_tokens_seen": 24701920,
      "step": 28535
    },
    {
      "epoch": 13.455917020273455,
      "grad_norm": 0.5701212882995605,
      "learning_rate": 9.462587286410021e-06,
      "loss": 0.1409,
      "num_input_tokens_seen": 24706080,
      "step": 28540
    },
    {
      "epoch": 13.458274398868458,
      "grad_norm": 0.7903903126716614,
      "learning_rate": 9.454897306891972e-06,
      "loss": 0.2021,
      "num_input_tokens_seen": 24710160,
      "step": 28545
    },
    {
      "epoch": 13.46063177746346,
      "grad_norm": 1.7097406387329102,
      "learning_rate": 9.44720972462411e-06,
      "loss": 0.1632,
      "num_input_tokens_seen": 24714256,
      "step": 28550
    },
    {
      "epoch": 13.462989156058462,
      "grad_norm": 1.5123569965362549,
      "learning_rate": 9.439524540791964e-06,
      "loss": 0.1237,
      "num_input_tokens_seen": 24717872,
      "step": 28555
    },
    {
      "epoch": 13.465346534653465,
      "grad_norm": 1.1556874513626099,
      "learning_rate": 9.431841756580673e-06,
      "loss": 0.1311,
      "num_input_tokens_seen": 24723472,
      "step": 28560
    },
    {
      "epoch": 13.467703913248467,
      "grad_norm": 1.0641556978225708,
      "learning_rate": 9.42416137317503e-06,
      "loss": 0.2642,
      "num_input_tokens_seen": 24729120,
      "step": 28565
    },
    {
      "epoch": 13.47006129184347,
      "grad_norm": 0.6616072654724121,
      "learning_rate": 9.416483391759437e-06,
      "loss": 0.2737,
      "num_input_tokens_seen": 24733552,
      "step": 28570
    },
    {
      "epoch": 13.472418670438472,
      "grad_norm": 1.3707598447799683,
      "learning_rate": 9.408807813517945e-06,
      "loss": 0.1037,
      "num_input_tokens_seen": 24738496,
      "step": 28575
    },
    {
      "epoch": 13.474776049033474,
      "grad_norm": 0.7447033524513245,
      "learning_rate": 9.401134639634221e-06,
      "loss": 0.1608,
      "num_input_tokens_seen": 24742256,
      "step": 28580
    },
    {
      "epoch": 13.477133427628477,
      "grad_norm": 1.0030606985092163,
      "learning_rate": 9.393463871291555e-06,
      "loss": 0.2367,
      "num_input_tokens_seen": 24746768,
      "step": 28585
    },
    {
      "epoch": 13.47949080622348,
      "grad_norm": 0.4985327124595642,
      "learning_rate": 9.385795509672881e-06,
      "loss": 0.1679,
      "num_input_tokens_seen": 24750960,
      "step": 28590
    },
    {
      "epoch": 13.481848184818482,
      "grad_norm": 1.0845283269882202,
      "learning_rate": 9.378129555960771e-06,
      "loss": 0.1466,
      "num_input_tokens_seen": 24755120,
      "step": 28595
    },
    {
      "epoch": 13.484205563413484,
      "grad_norm": 1.1213029623031616,
      "learning_rate": 9.370466011337392e-06,
      "loss": 0.1459,
      "num_input_tokens_seen": 24759888,
      "step": 28600
    },
    {
      "epoch": 13.484205563413484,
      "eval_loss": 0.2006627768278122,
      "eval_runtime": 22.1499,
      "eval_samples_per_second": 42.574,
      "eval_steps_per_second": 21.309,
      "num_input_tokens_seen": 24759888,
      "step": 28600
    },
    {
      "epoch": 13.486562942008486,
      "grad_norm": 1.5305202007293701,
      "learning_rate": 9.362804876984573e-06,
      "loss": 0.144,
      "num_input_tokens_seen": 24764560,
      "step": 28605
    },
    {
      "epoch": 13.488920320603489,
      "grad_norm": 0.7055389285087585,
      "learning_rate": 9.355146154083747e-06,
      "loss": 0.1519,
      "num_input_tokens_seen": 24769184,
      "step": 28610
    },
    {
      "epoch": 13.491277699198491,
      "grad_norm": 1.0532338619232178,
      "learning_rate": 9.347489843815987e-06,
      "loss": 0.1674,
      "num_input_tokens_seen": 24773088,
      "step": 28615
    },
    {
      "epoch": 13.493635077793494,
      "grad_norm": 1.411482334136963,
      "learning_rate": 9.339835947362002e-06,
      "loss": 0.1677,
      "num_input_tokens_seen": 24777520,
      "step": 28620
    },
    {
      "epoch": 13.495992456388496,
      "grad_norm": 0.8198084831237793,
      "learning_rate": 9.332184465902105e-06,
      "loss": 0.2344,
      "num_input_tokens_seen": 24781360,
      "step": 28625
    },
    {
      "epoch": 13.498349834983498,
      "grad_norm": 1.132393479347229,
      "learning_rate": 9.324535400616266e-06,
      "loss": 0.116,
      "num_input_tokens_seen": 24785648,
      "step": 28630
    },
    {
      "epoch": 13.500707213578501,
      "grad_norm": 1.5680779218673706,
      "learning_rate": 9.31688875268405e-06,
      "loss": 0.1925,
      "num_input_tokens_seen": 24790304,
      "step": 28635
    },
    {
      "epoch": 13.503064592173503,
      "grad_norm": 0.8831533789634705,
      "learning_rate": 9.309244523284674e-06,
      "loss": 0.1589,
      "num_input_tokens_seen": 24794048,
      "step": 28640
    },
    {
      "epoch": 13.505421970768506,
      "grad_norm": 0.5050777196884155,
      "learning_rate": 9.301602713596982e-06,
      "loss": 0.1947,
      "num_input_tokens_seen": 24799040,
      "step": 28645
    },
    {
      "epoch": 13.507779349363508,
      "grad_norm": 1.886544108390808,
      "learning_rate": 9.293963324799432e-06,
      "loss": 0.2226,
      "num_input_tokens_seen": 24802992,
      "step": 28650
    },
    {
      "epoch": 13.51013672795851,
      "grad_norm": 0.7525604963302612,
      "learning_rate": 9.286326358070104e-06,
      "loss": 0.1612,
      "num_input_tokens_seen": 24807248,
      "step": 28655
    },
    {
      "epoch": 13.512494106553513,
      "grad_norm": 0.7044686675071716,
      "learning_rate": 9.278691814586729e-06,
      "loss": 0.1733,
      "num_input_tokens_seen": 24811696,
      "step": 28660
    },
    {
      "epoch": 13.514851485148515,
      "grad_norm": 1.9728740453720093,
      "learning_rate": 9.271059695526635e-06,
      "loss": 0.2536,
      "num_input_tokens_seen": 24815376,
      "step": 28665
    },
    {
      "epoch": 13.517208863743518,
      "grad_norm": 0.6718676090240479,
      "learning_rate": 9.263430002066805e-06,
      "loss": 0.1453,
      "num_input_tokens_seen": 24819712,
      "step": 28670
    },
    {
      "epoch": 13.51956624233852,
      "grad_norm": 1.1795222759246826,
      "learning_rate": 9.25580273538382e-06,
      "loss": 0.1569,
      "num_input_tokens_seen": 24824768,
      "step": 28675
    },
    {
      "epoch": 13.521923620933523,
      "grad_norm": 1.480603575706482,
      "learning_rate": 9.248177896653907e-06,
      "loss": 0.1745,
      "num_input_tokens_seen": 24829104,
      "step": 28680
    },
    {
      "epoch": 13.524280999528525,
      "grad_norm": 0.7519298791885376,
      "learning_rate": 9.240555487052918e-06,
      "loss": 0.0968,
      "num_input_tokens_seen": 24832864,
      "step": 28685
    },
    {
      "epoch": 13.526638378123527,
      "grad_norm": 0.8370193839073181,
      "learning_rate": 9.232935507756313e-06,
      "loss": 0.1338,
      "num_input_tokens_seen": 24836656,
      "step": 28690
    },
    {
      "epoch": 13.52899575671853,
      "grad_norm": 2.95166015625,
      "learning_rate": 9.225317959939193e-06,
      "loss": 0.2199,
      "num_input_tokens_seen": 24841120,
      "step": 28695
    },
    {
      "epoch": 13.531353135313532,
      "grad_norm": 0.8225114345550537,
      "learning_rate": 9.217702844776287e-06,
      "loss": 0.1677,
      "num_input_tokens_seen": 24845792,
      "step": 28700
    },
    {
      "epoch": 13.533710513908535,
      "grad_norm": 1.7566485404968262,
      "learning_rate": 9.210090163441929e-06,
      "loss": 0.1336,
      "num_input_tokens_seen": 24850704,
      "step": 28705
    },
    {
      "epoch": 13.536067892503535,
      "grad_norm": 0.7777831554412842,
      "learning_rate": 9.202479917110105e-06,
      "loss": 0.1465,
      "num_input_tokens_seen": 24855200,
      "step": 28710
    },
    {
      "epoch": 13.53842527109854,
      "grad_norm": 0.7829527854919434,
      "learning_rate": 9.194872106954392e-06,
      "loss": 0.2367,
      "num_input_tokens_seen": 24859696,
      "step": 28715
    },
    {
      "epoch": 13.54078264969354,
      "grad_norm": 0.8530531525611877,
      "learning_rate": 9.187266734148029e-06,
      "loss": 0.1636,
      "num_input_tokens_seen": 24864000,
      "step": 28720
    },
    {
      "epoch": 13.543140028288542,
      "grad_norm": 1.469146966934204,
      "learning_rate": 9.179663799863849e-06,
      "loss": 0.1963,
      "num_input_tokens_seen": 24868720,
      "step": 28725
    },
    {
      "epoch": 13.545497406883545,
      "grad_norm": 0.9400542378425598,
      "learning_rate": 9.172063305274317e-06,
      "loss": 0.1354,
      "num_input_tokens_seen": 24873056,
      "step": 28730
    },
    {
      "epoch": 13.547854785478547,
      "grad_norm": 0.7705539464950562,
      "learning_rate": 9.164465251551527e-06,
      "loss": 0.1977,
      "num_input_tokens_seen": 24877200,
      "step": 28735
    },
    {
      "epoch": 13.55021216407355,
      "grad_norm": 1.063805341720581,
      "learning_rate": 9.156869639867205e-06,
      "loss": 0.2665,
      "num_input_tokens_seen": 24881584,
      "step": 28740
    },
    {
      "epoch": 13.552569542668552,
      "grad_norm": 0.6615839600563049,
      "learning_rate": 9.149276471392677e-06,
      "loss": 0.2139,
      "num_input_tokens_seen": 24886544,
      "step": 28745
    },
    {
      "epoch": 13.554926921263554,
      "grad_norm": 0.6083362698554993,
      "learning_rate": 9.141685747298914e-06,
      "loss": 0.2337,
      "num_input_tokens_seen": 24891664,
      "step": 28750
    },
    {
      "epoch": 13.557284299858557,
      "grad_norm": 1.1660289764404297,
      "learning_rate": 9.13409746875649e-06,
      "loss": 0.2571,
      "num_input_tokens_seen": 24895920,
      "step": 28755
    },
    {
      "epoch": 13.55964167845356,
      "grad_norm": 1.390304684638977,
      "learning_rate": 9.12651163693562e-06,
      "loss": 0.1368,
      "num_input_tokens_seen": 24900304,
      "step": 28760
    },
    {
      "epoch": 13.561999057048562,
      "grad_norm": 1.3609718084335327,
      "learning_rate": 9.11892825300614e-06,
      "loss": 0.1803,
      "num_input_tokens_seen": 24904432,
      "step": 28765
    },
    {
      "epoch": 13.564356435643564,
      "grad_norm": 1.5454614162445068,
      "learning_rate": 9.111347318137491e-06,
      "loss": 0.2499,
      "num_input_tokens_seen": 24909568,
      "step": 28770
    },
    {
      "epoch": 13.566713814238566,
      "grad_norm": 0.5177340507507324,
      "learning_rate": 9.103768833498755e-06,
      "loss": 0.1738,
      "num_input_tokens_seen": 24913616,
      "step": 28775
    },
    {
      "epoch": 13.569071192833569,
      "grad_norm": 0.7810370326042175,
      "learning_rate": 9.096192800258639e-06,
      "loss": 0.1565,
      "num_input_tokens_seen": 24918048,
      "step": 28780
    },
    {
      "epoch": 13.571428571428571,
      "grad_norm": 1.5711239576339722,
      "learning_rate": 9.088619219585443e-06,
      "loss": 0.2706,
      "num_input_tokens_seen": 24922464,
      "step": 28785
    },
    {
      "epoch": 13.573785950023574,
      "grad_norm": 0.6845369935035706,
      "learning_rate": 9.081048092647127e-06,
      "loss": 0.1867,
      "num_input_tokens_seen": 24926800,
      "step": 28790
    },
    {
      "epoch": 13.576143328618576,
      "grad_norm": 1.01654851436615,
      "learning_rate": 9.073479420611245e-06,
      "loss": 0.1719,
      "num_input_tokens_seen": 24932192,
      "step": 28795
    },
    {
      "epoch": 13.578500707213578,
      "grad_norm": 1.3059186935424805,
      "learning_rate": 9.065913204644974e-06,
      "loss": 0.2332,
      "num_input_tokens_seen": 24936720,
      "step": 28800
    },
    {
      "epoch": 13.578500707213578,
      "eval_loss": 0.19963344931602478,
      "eval_runtime": 22.1543,
      "eval_samples_per_second": 42.565,
      "eval_steps_per_second": 21.305,
      "num_input_tokens_seen": 24936720,
      "step": 28800
    },
    {
      "epoch": 13.58085808580858,
      "grad_norm": 1.1139683723449707,
      "learning_rate": 9.058349445915135e-06,
      "loss": 0.1578,
      "num_input_tokens_seen": 24941280,
      "step": 28805
    },
    {
      "epoch": 13.583215464403583,
      "grad_norm": 0.6704456210136414,
      "learning_rate": 9.050788145588138e-06,
      "loss": 0.1276,
      "num_input_tokens_seen": 24945088,
      "step": 28810
    },
    {
      "epoch": 13.585572842998586,
      "grad_norm": 0.6163750886917114,
      "learning_rate": 9.043229304830039e-06,
      "loss": 0.1561,
      "num_input_tokens_seen": 24949456,
      "step": 28815
    },
    {
      "epoch": 13.587930221593588,
      "grad_norm": 1.6076147556304932,
      "learning_rate": 9.035672924806515e-06,
      "loss": 0.19,
      "num_input_tokens_seen": 24954336,
      "step": 28820
    },
    {
      "epoch": 13.59028760018859,
      "grad_norm": 0.8879321217536926,
      "learning_rate": 9.028119006682839e-06,
      "loss": 0.2379,
      "num_input_tokens_seen": 24958752,
      "step": 28825
    },
    {
      "epoch": 13.592644978783593,
      "grad_norm": 1.1618834733963013,
      "learning_rate": 9.020567551623935e-06,
      "loss": 0.1217,
      "num_input_tokens_seen": 24963296,
      "step": 28830
    },
    {
      "epoch": 13.595002357378595,
      "grad_norm": 0.7077571153640747,
      "learning_rate": 9.013018560794318e-06,
      "loss": 0.2107,
      "num_input_tokens_seen": 24967728,
      "step": 28835
    },
    {
      "epoch": 13.597359735973598,
      "grad_norm": 2.198136329650879,
      "learning_rate": 9.005472035358139e-06,
      "loss": 0.2726,
      "num_input_tokens_seen": 24972544,
      "step": 28840
    },
    {
      "epoch": 13.5997171145686,
      "grad_norm": 0.6542454957962036,
      "learning_rate": 8.997927976479185e-06,
      "loss": 0.1515,
      "num_input_tokens_seen": 24977168,
      "step": 28845
    },
    {
      "epoch": 13.602074493163602,
      "grad_norm": 0.5838233232498169,
      "learning_rate": 8.99038638532082e-06,
      "loss": 0.1951,
      "num_input_tokens_seen": 24981632,
      "step": 28850
    },
    {
      "epoch": 13.604431871758605,
      "grad_norm": 0.8539848327636719,
      "learning_rate": 8.982847263046065e-06,
      "loss": 0.1964,
      "num_input_tokens_seen": 24985552,
      "step": 28855
    },
    {
      "epoch": 13.606789250353607,
      "grad_norm": 0.41996169090270996,
      "learning_rate": 8.975310610817555e-06,
      "loss": 0.2005,
      "num_input_tokens_seen": 24989712,
      "step": 28860
    },
    {
      "epoch": 13.60914662894861,
      "grad_norm": 0.5272237658500671,
      "learning_rate": 8.967776429797528e-06,
      "loss": 0.2156,
      "num_input_tokens_seen": 24994224,
      "step": 28865
    },
    {
      "epoch": 13.611504007543612,
      "grad_norm": 1.203905463218689,
      "learning_rate": 8.960244721147842e-06,
      "loss": 0.1603,
      "num_input_tokens_seen": 24998928,
      "step": 28870
    },
    {
      "epoch": 13.613861386138614,
      "grad_norm": 0.6777842044830322,
      "learning_rate": 8.952715486029995e-06,
      "loss": 0.1693,
      "num_input_tokens_seen": 25003504,
      "step": 28875
    },
    {
      "epoch": 13.616218764733617,
      "grad_norm": 0.6615727543830872,
      "learning_rate": 8.945188725605075e-06,
      "loss": 0.1271,
      "num_input_tokens_seen": 25008064,
      "step": 28880
    },
    {
      "epoch": 13.61857614332862,
      "grad_norm": 1.5631139278411865,
      "learning_rate": 8.937664441033817e-06,
      "loss": 0.2365,
      "num_input_tokens_seen": 25011840,
      "step": 28885
    },
    {
      "epoch": 13.620933521923622,
      "grad_norm": 0.4748738408088684,
      "learning_rate": 8.930142633476549e-06,
      "loss": 0.1113,
      "num_input_tokens_seen": 25015680,
      "step": 28890
    },
    {
      "epoch": 13.623290900518624,
      "grad_norm": 0.3566029369831085,
      "learning_rate": 8.92262330409323e-06,
      "loss": 0.168,
      "num_input_tokens_seen": 25020464,
      "step": 28895
    },
    {
      "epoch": 13.625648279113626,
      "grad_norm": 1.1840143203735352,
      "learning_rate": 8.915106454043448e-06,
      "loss": 0.1297,
      "num_input_tokens_seen": 25025296,
      "step": 28900
    },
    {
      "epoch": 13.628005657708629,
      "grad_norm": 0.7404469847679138,
      "learning_rate": 8.90759208448638e-06,
      "loss": 0.2051,
      "num_input_tokens_seen": 25029696,
      "step": 28905
    },
    {
      "epoch": 13.630363036303631,
      "grad_norm": 0.5053998231887817,
      "learning_rate": 8.900080196580848e-06,
      "loss": 0.1444,
      "num_input_tokens_seen": 25034592,
      "step": 28910
    },
    {
      "epoch": 13.632720414898632,
      "grad_norm": 1.1763859987258911,
      "learning_rate": 8.892570791485267e-06,
      "loss": 0.1773,
      "num_input_tokens_seen": 25039120,
      "step": 28915
    },
    {
      "epoch": 13.635077793493634,
      "grad_norm": 2.0058786869049072,
      "learning_rate": 8.885063870357688e-06,
      "loss": 0.1577,
      "num_input_tokens_seen": 25043648,
      "step": 28920
    },
    {
      "epoch": 13.637435172088637,
      "grad_norm": 0.46481141448020935,
      "learning_rate": 8.87755943435578e-06,
      "loss": 0.2245,
      "num_input_tokens_seen": 25047728,
      "step": 28925
    },
    {
      "epoch": 13.639792550683639,
      "grad_norm": 0.935066819190979,
      "learning_rate": 8.87005748463681e-06,
      "loss": 0.13,
      "num_input_tokens_seen": 25052112,
      "step": 28930
    },
    {
      "epoch": 13.642149929278641,
      "grad_norm": 1.355580449104309,
      "learning_rate": 8.862558022357681e-06,
      "loss": 0.1969,
      "num_input_tokens_seen": 25056416,
      "step": 28935
    },
    {
      "epoch": 13.644507307873644,
      "grad_norm": 0.6559155583381653,
      "learning_rate": 8.855061048674903e-06,
      "loss": 0.1597,
      "num_input_tokens_seen": 25060096,
      "step": 28940
    },
    {
      "epoch": 13.646864686468646,
      "grad_norm": 1.4893410205841064,
      "learning_rate": 8.847566564744595e-06,
      "loss": 0.1794,
      "num_input_tokens_seen": 25064640,
      "step": 28945
    },
    {
      "epoch": 13.649222065063649,
      "grad_norm": 2.3314645290374756,
      "learning_rate": 8.840074571722512e-06,
      "loss": 0.2063,
      "num_input_tokens_seen": 25068736,
      "step": 28950
    },
    {
      "epoch": 13.651579443658651,
      "grad_norm": 0.3892892599105835,
      "learning_rate": 8.832585070764002e-06,
      "loss": 0.1144,
      "num_input_tokens_seen": 25072736,
      "step": 28955
    },
    {
      "epoch": 13.653936822253653,
      "grad_norm": 2.680431842803955,
      "learning_rate": 8.825098063024045e-06,
      "loss": 0.3083,
      "num_input_tokens_seen": 25077712,
      "step": 28960
    },
    {
      "epoch": 13.656294200848656,
      "grad_norm": 0.748672604560852,
      "learning_rate": 8.817613549657244e-06,
      "loss": 0.1628,
      "num_input_tokens_seen": 25082208,
      "step": 28965
    },
    {
      "epoch": 13.658651579443658,
      "grad_norm": 1.583905577659607,
      "learning_rate": 8.810131531817783e-06,
      "loss": 0.2407,
      "num_input_tokens_seen": 25086864,
      "step": 28970
    },
    {
      "epoch": 13.66100895803866,
      "grad_norm": 0.8147802352905273,
      "learning_rate": 8.802652010659496e-06,
      "loss": 0.1937,
      "num_input_tokens_seen": 25091376,
      "step": 28975
    },
    {
      "epoch": 13.663366336633663,
      "grad_norm": 0.5900242924690247,
      "learning_rate": 8.795174987335827e-06,
      "loss": 0.1108,
      "num_input_tokens_seen": 25095424,
      "step": 28980
    },
    {
      "epoch": 13.665723715228665,
      "grad_norm": 1.3729535341262817,
      "learning_rate": 8.787700462999807e-06,
      "loss": 0.1862,
      "num_input_tokens_seen": 25099360,
      "step": 28985
    },
    {
      "epoch": 13.668081093823668,
      "grad_norm": 1.0200573205947876,
      "learning_rate": 8.780228438804122e-06,
      "loss": 0.1833,
      "num_input_tokens_seen": 25103376,
      "step": 28990
    },
    {
      "epoch": 13.67043847241867,
      "grad_norm": 1.483525037765503,
      "learning_rate": 8.772758915901032e-06,
      "loss": 0.2172,
      "num_input_tokens_seen": 25107440,
      "step": 28995
    },
    {
      "epoch": 13.672795851013673,
      "grad_norm": 0.3989827036857605,
      "learning_rate": 8.765291895442443e-06,
      "loss": 0.1106,
      "num_input_tokens_seen": 25110864,
      "step": 29000
    },
    {
      "epoch": 13.672795851013673,
      "eval_loss": 0.20014061033725739,
      "eval_runtime": 22.1407,
      "eval_samples_per_second": 42.591,
      "eval_steps_per_second": 21.318,
      "num_input_tokens_seen": 25110864,
      "step": 29000
    },
    {
      "epoch": 13.675153229608675,
      "grad_norm": 2.1324241161346436,
      "learning_rate": 8.75782737857987e-06,
      "loss": 0.1908,
      "num_input_tokens_seen": 25115472,
      "step": 29005
    },
    {
      "epoch": 13.677510608203677,
      "grad_norm": 0.6638823747634888,
      "learning_rate": 8.750365366464425e-06,
      "loss": 0.2196,
      "num_input_tokens_seen": 25119696,
      "step": 29010
    },
    {
      "epoch": 13.67986798679868,
      "grad_norm": 0.576482892036438,
      "learning_rate": 8.742905860246838e-06,
      "loss": 0.1396,
      "num_input_tokens_seen": 25123824,
      "step": 29015
    },
    {
      "epoch": 13.682225365393682,
      "grad_norm": 1.0907065868377686,
      "learning_rate": 8.735448861077478e-06,
      "loss": 0.1963,
      "num_input_tokens_seen": 25127536,
      "step": 29020
    },
    {
      "epoch": 13.684582743988685,
      "grad_norm": 0.4524258077144623,
      "learning_rate": 8.727994370106288e-06,
      "loss": 0.1404,
      "num_input_tokens_seen": 25131536,
      "step": 29025
    },
    {
      "epoch": 13.686940122583687,
      "grad_norm": 0.6537094712257385,
      "learning_rate": 8.720542388482861e-06,
      "loss": 0.1168,
      "num_input_tokens_seen": 25136128,
      "step": 29030
    },
    {
      "epoch": 13.68929750117869,
      "grad_norm": 1.2591654062271118,
      "learning_rate": 8.71309291735637e-06,
      "loss": 0.1817,
      "num_input_tokens_seen": 25140768,
      "step": 29035
    },
    {
      "epoch": 13.691654879773692,
      "grad_norm": 0.9281764030456543,
      "learning_rate": 8.705645957875621e-06,
      "loss": 0.1452,
      "num_input_tokens_seen": 25144608,
      "step": 29040
    },
    {
      "epoch": 13.694012258368694,
      "grad_norm": 1.1537511348724365,
      "learning_rate": 8.698201511189048e-06,
      "loss": 0.1994,
      "num_input_tokens_seen": 25148688,
      "step": 29045
    },
    {
      "epoch": 13.696369636963697,
      "grad_norm": 1.3043670654296875,
      "learning_rate": 8.690759578444649e-06,
      "loss": 0.2,
      "num_input_tokens_seen": 25154336,
      "step": 29050
    },
    {
      "epoch": 13.698727015558699,
      "grad_norm": 0.3770429790019989,
      "learning_rate": 8.68332016079008e-06,
      "loss": 0.1476,
      "num_input_tokens_seen": 25159040,
      "step": 29055
    },
    {
      "epoch": 13.701084394153701,
      "grad_norm": 0.3652510643005371,
      "learning_rate": 8.6758832593726e-06,
      "loss": 0.1488,
      "num_input_tokens_seen": 25162768,
      "step": 29060
    },
    {
      "epoch": 13.703441772748704,
      "grad_norm": 1.0926995277404785,
      "learning_rate": 8.668448875339053e-06,
      "loss": 0.1429,
      "num_input_tokens_seen": 25166560,
      "step": 29065
    },
    {
      "epoch": 13.705799151343706,
      "grad_norm": 1.0356690883636475,
      "learning_rate": 8.661017009835933e-06,
      "loss": 0.198,
      "num_input_tokens_seen": 25170272,
      "step": 29070
    },
    {
      "epoch": 13.708156529938709,
      "grad_norm": 0.6715324521064758,
      "learning_rate": 8.653587664009311e-06,
      "loss": 0.1591,
      "num_input_tokens_seen": 25174272,
      "step": 29075
    },
    {
      "epoch": 13.710513908533711,
      "grad_norm": 0.46329161524772644,
      "learning_rate": 8.646160839004902e-06,
      "loss": 0.1423,
      "num_input_tokens_seen": 25178688,
      "step": 29080
    },
    {
      "epoch": 13.712871287128714,
      "grad_norm": 0.9814568758010864,
      "learning_rate": 8.638736535967998e-06,
      "loss": 0.1825,
      "num_input_tokens_seen": 25182752,
      "step": 29085
    },
    {
      "epoch": 13.715228665723716,
      "grad_norm": 0.6068205237388611,
      "learning_rate": 8.631314756043535e-06,
      "loss": 0.1145,
      "num_input_tokens_seen": 25187344,
      "step": 29090
    },
    {
      "epoch": 13.717586044318718,
      "grad_norm": 1.6490503549575806,
      "learning_rate": 8.62389550037603e-06,
      "loss": 0.1627,
      "num_input_tokens_seen": 25191920,
      "step": 29095
    },
    {
      "epoch": 13.71994342291372,
      "grad_norm": 0.6026099324226379,
      "learning_rate": 8.616478770109646e-06,
      "loss": 0.1325,
      "num_input_tokens_seen": 25196368,
      "step": 29100
    },
    {
      "epoch": 13.722300801508723,
      "grad_norm": 1.0266069173812866,
      "learning_rate": 8.609064566388111e-06,
      "loss": 0.1712,
      "num_input_tokens_seen": 25201264,
      "step": 29105
    },
    {
      "epoch": 13.724658180103724,
      "grad_norm": 1.644388198852539,
      "learning_rate": 8.601652890354815e-06,
      "loss": 0.1448,
      "num_input_tokens_seen": 25205952,
      "step": 29110
    },
    {
      "epoch": 13.727015558698728,
      "grad_norm": 0.6656533479690552,
      "learning_rate": 8.594243743152705e-06,
      "loss": 0.1842,
      "num_input_tokens_seen": 25210208,
      "step": 29115
    },
    {
      "epoch": 13.729372937293729,
      "grad_norm": 0.670244038105011,
      "learning_rate": 8.58683712592438e-06,
      "loss": 0.1699,
      "num_input_tokens_seen": 25214720,
      "step": 29120
    },
    {
      "epoch": 13.731730315888731,
      "grad_norm": 1.1069554090499878,
      "learning_rate": 8.579433039812037e-06,
      "loss": 0.2215,
      "num_input_tokens_seen": 25219040,
      "step": 29125
    },
    {
      "epoch": 13.734087694483733,
      "grad_norm": 1.856331467628479,
      "learning_rate": 8.572031485957466e-06,
      "loss": 0.2081,
      "num_input_tokens_seen": 25222992,
      "step": 29130
    },
    {
      "epoch": 13.736445073078736,
      "grad_norm": 1.8751800060272217,
      "learning_rate": 8.564632465502084e-06,
      "loss": 0.2112,
      "num_input_tokens_seen": 25227472,
      "step": 29135
    },
    {
      "epoch": 13.738802451673738,
      "grad_norm": 0.7135925889015198,
      "learning_rate": 8.557235979586928e-06,
      "loss": 0.1632,
      "num_input_tokens_seen": 25232400,
      "step": 29140
    },
    {
      "epoch": 13.74115983026874,
      "grad_norm": 1.051513910293579,
      "learning_rate": 8.549842029352606e-06,
      "loss": 0.1674,
      "num_input_tokens_seen": 25236192,
      "step": 29145
    },
    {
      "epoch": 13.743517208863743,
      "grad_norm": 0.9222533106803894,
      "learning_rate": 8.542450615939376e-06,
      "loss": 0.2094,
      "num_input_tokens_seen": 25240640,
      "step": 29150
    },
    {
      "epoch": 13.745874587458745,
      "grad_norm": 1.503047227859497,
      "learning_rate": 8.535061740487082e-06,
      "loss": 0.1514,
      "num_input_tokens_seen": 25244752,
      "step": 29155
    },
    {
      "epoch": 13.748231966053748,
      "grad_norm": 0.6449356079101562,
      "learning_rate": 8.527675404135168e-06,
      "loss": 0.2703,
      "num_input_tokens_seen": 25249632,
      "step": 29160
    },
    {
      "epoch": 13.75058934464875,
      "grad_norm": 1.7762521505355835,
      "learning_rate": 8.520291608022724e-06,
      "loss": 0.2154,
      "num_input_tokens_seen": 25254384,
      "step": 29165
    },
    {
      "epoch": 13.752946723243753,
      "grad_norm": 1.0126310586929321,
      "learning_rate": 8.512910353288398e-06,
      "loss": 0.2117,
      "num_input_tokens_seen": 25258208,
      "step": 29170
    },
    {
      "epoch": 13.755304101838755,
      "grad_norm": 0.9630042314529419,
      "learning_rate": 8.505531641070486e-06,
      "loss": 0.1445,
      "num_input_tokens_seen": 25261840,
      "step": 29175
    },
    {
      "epoch": 13.757661480433757,
      "grad_norm": 1.1456046104431152,
      "learning_rate": 8.498155472506885e-06,
      "loss": 0.1032,
      "num_input_tokens_seen": 25266096,
      "step": 29180
    },
    {
      "epoch": 13.76001885902876,
      "grad_norm": 1.1342321634292603,
      "learning_rate": 8.49078184873508e-06,
      "loss": 0.2229,
      "num_input_tokens_seen": 25270720,
      "step": 29185
    },
    {
      "epoch": 13.762376237623762,
      "grad_norm": 1.016515851020813,
      "learning_rate": 8.483410770892188e-06,
      "loss": 0.1567,
      "num_input_tokens_seen": 25275344,
      "step": 29190
    },
    {
      "epoch": 13.764733616218765,
      "grad_norm": 1.5277231931686401,
      "learning_rate": 8.476042240114909e-06,
      "loss": 0.2395,
      "num_input_tokens_seen": 25279920,
      "step": 29195
    },
    {
      "epoch": 13.767090994813767,
      "grad_norm": 2.587883710861206,
      "learning_rate": 8.468676257539568e-06,
      "loss": 0.2217,
      "num_input_tokens_seen": 25284944,
      "step": 29200
    },
    {
      "epoch": 13.767090994813767,
      "eval_loss": 0.20020969212055206,
      "eval_runtime": 22.1624,
      "eval_samples_per_second": 42.55,
      "eval_steps_per_second": 21.297,
      "num_input_tokens_seen": 25284944,
      "step": 29200
    },
    {
      "epoch": 13.76944837340877,
      "grad_norm": 1.5468751192092896,
      "learning_rate": 8.4613128243021e-06,
      "loss": 0.1856,
      "num_input_tokens_seen": 25288480,
      "step": 29205
    },
    {
      "epoch": 13.771805752003772,
      "grad_norm": 0.6193552017211914,
      "learning_rate": 8.453951941538028e-06,
      "loss": 0.1548,
      "num_input_tokens_seen": 25292736,
      "step": 29210
    },
    {
      "epoch": 13.774163130598774,
      "grad_norm": 0.45216819643974304,
      "learning_rate": 8.446593610382495e-06,
      "loss": 0.1207,
      "num_input_tokens_seen": 25297280,
      "step": 29215
    },
    {
      "epoch": 13.776520509193777,
      "grad_norm": 0.8432392477989197,
      "learning_rate": 8.439237831970259e-06,
      "loss": 0.1493,
      "num_input_tokens_seen": 25302320,
      "step": 29220
    },
    {
      "epoch": 13.778877887788779,
      "grad_norm": 0.5791231393814087,
      "learning_rate": 8.431884607435667e-06,
      "loss": 0.1265,
      "num_input_tokens_seen": 25306416,
      "step": 29225
    },
    {
      "epoch": 13.781235266383781,
      "grad_norm": 0.9642370343208313,
      "learning_rate": 8.424533937912665e-06,
      "loss": 0.2222,
      "num_input_tokens_seen": 25310768,
      "step": 29230
    },
    {
      "epoch": 13.783592644978784,
      "grad_norm": 1.0482410192489624,
      "learning_rate": 8.41718582453484e-06,
      "loss": 0.196,
      "num_input_tokens_seen": 25315056,
      "step": 29235
    },
    {
      "epoch": 13.785950023573786,
      "grad_norm": 0.808412492275238,
      "learning_rate": 8.409840268435346e-06,
      "loss": 0.1284,
      "num_input_tokens_seen": 25319584,
      "step": 29240
    },
    {
      "epoch": 13.788307402168789,
      "grad_norm": 0.7934390902519226,
      "learning_rate": 8.402497270746976e-06,
      "loss": 0.1965,
      "num_input_tokens_seen": 25323376,
      "step": 29245
    },
    {
      "epoch": 13.790664780763791,
      "grad_norm": 0.664463460445404,
      "learning_rate": 8.395156832602095e-06,
      "loss": 0.1891,
      "num_input_tokens_seen": 25328160,
      "step": 29250
    },
    {
      "epoch": 13.793022159358793,
      "grad_norm": 0.3799958825111389,
      "learning_rate": 8.387818955132707e-06,
      "loss": 0.1934,
      "num_input_tokens_seen": 25332464,
      "step": 29255
    },
    {
      "epoch": 13.795379537953796,
      "grad_norm": 1.172903299331665,
      "learning_rate": 8.38048363947039e-06,
      "loss": 0.1511,
      "num_input_tokens_seen": 25336896,
      "step": 29260
    },
    {
      "epoch": 13.797736916548798,
      "grad_norm": 0.9860852956771851,
      "learning_rate": 8.373150886746351e-06,
      "loss": 0.166,
      "num_input_tokens_seen": 25341312,
      "step": 29265
    },
    {
      "epoch": 13.8000942951438,
      "grad_norm": 0.9530287981033325,
      "learning_rate": 8.365820698091397e-06,
      "loss": 0.1126,
      "num_input_tokens_seen": 25345968,
      "step": 29270
    },
    {
      "epoch": 13.802451673738803,
      "grad_norm": 0.9595927596092224,
      "learning_rate": 8.358493074635922e-06,
      "loss": 0.2124,
      "num_input_tokens_seen": 25349728,
      "step": 29275
    },
    {
      "epoch": 13.804809052333805,
      "grad_norm": 2.391026735305786,
      "learning_rate": 8.351168017509948e-06,
      "loss": 0.2443,
      "num_input_tokens_seen": 25354032,
      "step": 29280
    },
    {
      "epoch": 13.807166430928808,
      "grad_norm": 2.19706392288208,
      "learning_rate": 8.343845527843094e-06,
      "loss": 0.2044,
      "num_input_tokens_seen": 25357792,
      "step": 29285
    },
    {
      "epoch": 13.80952380952381,
      "grad_norm": 0.8304182291030884,
      "learning_rate": 8.336525606764566e-06,
      "loss": 0.2162,
      "num_input_tokens_seen": 25362128,
      "step": 29290
    },
    {
      "epoch": 13.811881188118813,
      "grad_norm": 0.4365807771682739,
      "learning_rate": 8.329208255403204e-06,
      "loss": 0.1771,
      "num_input_tokens_seen": 25365600,
      "step": 29295
    },
    {
      "epoch": 13.814238566713815,
      "grad_norm": 0.5926506519317627,
      "learning_rate": 8.321893474887426e-06,
      "loss": 0.1255,
      "num_input_tokens_seen": 25369952,
      "step": 29300
    },
    {
      "epoch": 13.816595945308817,
      "grad_norm": 0.5912235379219055,
      "learning_rate": 8.31458126634526e-06,
      "loss": 0.1248,
      "num_input_tokens_seen": 25374816,
      "step": 29305
    },
    {
      "epoch": 13.81895332390382,
      "grad_norm": 1.0718337297439575,
      "learning_rate": 8.30727163090435e-06,
      "loss": 0.116,
      "num_input_tokens_seen": 25380176,
      "step": 29310
    },
    {
      "epoch": 13.82131070249882,
      "grad_norm": 1.1459100246429443,
      "learning_rate": 8.29996456969192e-06,
      "loss": 0.1376,
      "num_input_tokens_seen": 25383856,
      "step": 29315
    },
    {
      "epoch": 13.823668081093825,
      "grad_norm": 0.9791961312294006,
      "learning_rate": 8.292660083834818e-06,
      "loss": 0.1801,
      "num_input_tokens_seen": 25387312,
      "step": 29320
    },
    {
      "epoch": 13.826025459688825,
      "grad_norm": 1.1345396041870117,
      "learning_rate": 8.2853581744595e-06,
      "loss": 0.1201,
      "num_input_tokens_seen": 25392192,
      "step": 29325
    },
    {
      "epoch": 13.828382838283828,
      "grad_norm": 1.3272814750671387,
      "learning_rate": 8.278058842691991e-06,
      "loss": 0.1858,
      "num_input_tokens_seen": 25396240,
      "step": 29330
    },
    {
      "epoch": 13.83074021687883,
      "grad_norm": 1.4364196062088013,
      "learning_rate": 8.27076208965796e-06,
      "loss": 0.235,
      "num_input_tokens_seen": 25400768,
      "step": 29335
    },
    {
      "epoch": 13.833097595473832,
      "grad_norm": 0.8997614979743958,
      "learning_rate": 8.263467916482637e-06,
      "loss": 0.1494,
      "num_input_tokens_seen": 25404960,
      "step": 29340
    },
    {
      "epoch": 13.835454974068835,
      "grad_norm": 1.4908571243286133,
      "learning_rate": 8.256176324290885e-06,
      "loss": 0.2097,
      "num_input_tokens_seen": 25408768,
      "step": 29345
    },
    {
      "epoch": 13.837812352663837,
      "grad_norm": 2.8723957538604736,
      "learning_rate": 8.248887314207168e-06,
      "loss": 0.1977,
      "num_input_tokens_seen": 25413712,
      "step": 29350
    },
    {
      "epoch": 13.84016973125884,
      "grad_norm": 0.6796513795852661,
      "learning_rate": 8.24160088735553e-06,
      "loss": 0.1254,
      "num_input_tokens_seen": 25417888,
      "step": 29355
    },
    {
      "epoch": 13.842527109853842,
      "grad_norm": 0.43760067224502563,
      "learning_rate": 8.234317044859629e-06,
      "loss": 0.2591,
      "num_input_tokens_seen": 25422176,
      "step": 29360
    },
    {
      "epoch": 13.844884488448844,
      "grad_norm": 0.4097949266433716,
      "learning_rate": 8.227035787842744e-06,
      "loss": 0.2682,
      "num_input_tokens_seen": 25427216,
      "step": 29365
    },
    {
      "epoch": 13.847241867043847,
      "grad_norm": 0.9256018996238708,
      "learning_rate": 8.219757117427721e-06,
      "loss": 0.1105,
      "num_input_tokens_seen": 25431376,
      "step": 29370
    },
    {
      "epoch": 13.84959924563885,
      "grad_norm": 0.7641186118125916,
      "learning_rate": 8.212481034737014e-06,
      "loss": 0.1855,
      "num_input_tokens_seen": 25435680,
      "step": 29375
    },
    {
      "epoch": 13.851956624233852,
      "grad_norm": 0.9965980052947998,
      "learning_rate": 8.205207540892707e-06,
      "loss": 0.1992,
      "num_input_tokens_seen": 25439936,
      "step": 29380
    },
    {
      "epoch": 13.854314002828854,
      "grad_norm": 0.5187287926673889,
      "learning_rate": 8.197936637016442e-06,
      "loss": 0.1532,
      "num_input_tokens_seen": 25444720,
      "step": 29385
    },
    {
      "epoch": 13.856671381423856,
      "grad_norm": 1.2991678714752197,
      "learning_rate": 8.190668324229508e-06,
      "loss": 0.1834,
      "num_input_tokens_seen": 25448480,
      "step": 29390
    },
    {
      "epoch": 13.859028760018859,
      "grad_norm": 1.321941614151001,
      "learning_rate": 8.183402603652749e-06,
      "loss": 0.1345,
      "num_input_tokens_seen": 25452464,
      "step": 29395
    },
    {
      "epoch": 13.861386138613861,
      "grad_norm": 0.9365843534469604,
      "learning_rate": 8.176139476406635e-06,
      "loss": 0.2221,
      "num_input_tokens_seen": 25456816,
      "step": 29400
    },
    {
      "epoch": 13.861386138613861,
      "eval_loss": 0.20104682445526123,
      "eval_runtime": 22.1883,
      "eval_samples_per_second": 42.5,
      "eval_steps_per_second": 21.273,
      "num_input_tokens_seen": 25456816,
      "step": 29400
    },
    {
      "epoch": 13.863743517208864,
      "grad_norm": 1.208789587020874,
      "learning_rate": 8.16887894361125e-06,
      "loss": 0.2366,
      "num_input_tokens_seen": 25461280,
      "step": 29405
    },
    {
      "epoch": 13.866100895803866,
      "grad_norm": 1.1936861276626587,
      "learning_rate": 8.161621006386233e-06,
      "loss": 0.1123,
      "num_input_tokens_seen": 25465664,
      "step": 29410
    },
    {
      "epoch": 13.868458274398868,
      "grad_norm": 1.0492544174194336,
      "learning_rate": 8.154365665850869e-06,
      "loss": 0.1915,
      "num_input_tokens_seen": 25469504,
      "step": 29415
    },
    {
      "epoch": 13.87081565299387,
      "grad_norm": 0.6605067253112793,
      "learning_rate": 8.147112923124005e-06,
      "loss": 0.1782,
      "num_input_tokens_seen": 25473072,
      "step": 29420
    },
    {
      "epoch": 13.873173031588873,
      "grad_norm": 0.8070706129074097,
      "learning_rate": 8.13986277932412e-06,
      "loss": 0.1644,
      "num_input_tokens_seen": 25476624,
      "step": 29425
    },
    {
      "epoch": 13.875530410183876,
      "grad_norm": 0.35368937253952026,
      "learning_rate": 8.132615235569277e-06,
      "loss": 0.1564,
      "num_input_tokens_seen": 25482160,
      "step": 29430
    },
    {
      "epoch": 13.877887788778878,
      "grad_norm": 1.3069266080856323,
      "learning_rate": 8.125370292977124e-06,
      "loss": 0.1919,
      "num_input_tokens_seen": 25487296,
      "step": 29435
    },
    {
      "epoch": 13.88024516737388,
      "grad_norm": 1.1480604410171509,
      "learning_rate": 8.118127952664944e-06,
      "loss": 0.1864,
      "num_input_tokens_seen": 25491056,
      "step": 29440
    },
    {
      "epoch": 13.882602545968883,
      "grad_norm": 0.8057440519332886,
      "learning_rate": 8.110888215749574e-06,
      "loss": 0.1613,
      "num_input_tokens_seen": 25495520,
      "step": 29445
    },
    {
      "epoch": 13.884959924563885,
      "grad_norm": 0.9897560477256775,
      "learning_rate": 8.10365108334749e-06,
      "loss": 0.1247,
      "num_input_tokens_seen": 25500352,
      "step": 29450
    },
    {
      "epoch": 13.887317303158888,
      "grad_norm": 0.5120867490768433,
      "learning_rate": 8.096416556574743e-06,
      "loss": 0.1378,
      "num_input_tokens_seen": 25505200,
      "step": 29455
    },
    {
      "epoch": 13.88967468175389,
      "grad_norm": 1.5268380641937256,
      "learning_rate": 8.08918463654698e-06,
      "loss": 0.2487,
      "num_input_tokens_seen": 25509552,
      "step": 29460
    },
    {
      "epoch": 13.892032060348892,
      "grad_norm": 0.349714070558548,
      "learning_rate": 8.081955324379458e-06,
      "loss": 0.1308,
      "num_input_tokens_seen": 25513840,
      "step": 29465
    },
    {
      "epoch": 13.894389438943895,
      "grad_norm": 0.5686280131340027,
      "learning_rate": 8.074728621187039e-06,
      "loss": 0.1216,
      "num_input_tokens_seen": 25518016,
      "step": 29470
    },
    {
      "epoch": 13.896746817538897,
      "grad_norm": 0.7988592386245728,
      "learning_rate": 8.067504528084158e-06,
      "loss": 0.2263,
      "num_input_tokens_seen": 25522688,
      "step": 29475
    },
    {
      "epoch": 13.8991041961339,
      "grad_norm": 1.649152398109436,
      "learning_rate": 8.060283046184861e-06,
      "loss": 0.1396,
      "num_input_tokens_seen": 25528400,
      "step": 29480
    },
    {
      "epoch": 13.901461574728902,
      "grad_norm": 1.4255144596099854,
      "learning_rate": 8.053064176602806e-06,
      "loss": 0.155,
      "num_input_tokens_seen": 25532496,
      "step": 29485
    },
    {
      "epoch": 13.903818953323904,
      "grad_norm": 1.4468388557434082,
      "learning_rate": 8.045847920451216e-06,
      "loss": 0.1846,
      "num_input_tokens_seen": 25536736,
      "step": 29490
    },
    {
      "epoch": 13.906176331918907,
      "grad_norm": 0.46303653717041016,
      "learning_rate": 8.038634278842944e-06,
      "loss": 0.1167,
      "num_input_tokens_seen": 25540912,
      "step": 29495
    },
    {
      "epoch": 13.90853371051391,
      "grad_norm": 0.7752310037612915,
      "learning_rate": 8.031423252890408e-06,
      "loss": 0.1648,
      "num_input_tokens_seen": 25544768,
      "step": 29500
    },
    {
      "epoch": 13.910891089108912,
      "grad_norm": 0.6092898845672607,
      "learning_rate": 8.024214843705646e-06,
      "loss": 0.1762,
      "num_input_tokens_seen": 25549200,
      "step": 29505
    },
    {
      "epoch": 13.913248467703912,
      "grad_norm": 1.5254464149475098,
      "learning_rate": 8.017009052400295e-06,
      "loss": 0.1588,
      "num_input_tokens_seen": 25552816,
      "step": 29510
    },
    {
      "epoch": 13.915605846298917,
      "grad_norm": 0.728187084197998,
      "learning_rate": 8.00980588008557e-06,
      "loss": 0.1451,
      "num_input_tokens_seen": 25557744,
      "step": 29515
    },
    {
      "epoch": 13.917963224893917,
      "grad_norm": 1.3026561737060547,
      "learning_rate": 8.002605327872282e-06,
      "loss": 0.1195,
      "num_input_tokens_seen": 25561344,
      "step": 29520
    },
    {
      "epoch": 13.92032060348892,
      "grad_norm": 0.6342382431030273,
      "learning_rate": 7.995407396870862e-06,
      "loss": 0.1877,
      "num_input_tokens_seen": 25566496,
      "step": 29525
    },
    {
      "epoch": 13.922677982083922,
      "grad_norm": 0.6497116088867188,
      "learning_rate": 7.988212088191307e-06,
      "loss": 0.1649,
      "num_input_tokens_seen": 25570800,
      "step": 29530
    },
    {
      "epoch": 13.925035360678924,
      "grad_norm": 0.7867460250854492,
      "learning_rate": 7.98101940294324e-06,
      "loss": 0.1544,
      "num_input_tokens_seen": 25574816,
      "step": 29535
    },
    {
      "epoch": 13.927392739273927,
      "grad_norm": 0.5876978039741516,
      "learning_rate": 7.973829342235847e-06,
      "loss": 0.1307,
      "num_input_tokens_seen": 25579936,
      "step": 29540
    },
    {
      "epoch": 13.92975011786893,
      "grad_norm": 1.607958436012268,
      "learning_rate": 7.966641907177936e-06,
      "loss": 0.1914,
      "num_input_tokens_seen": 25584032,
      "step": 29545
    },
    {
      "epoch": 13.932107496463932,
      "grad_norm": 1.5141175985336304,
      "learning_rate": 7.959457098877901e-06,
      "loss": 0.1529,
      "num_input_tokens_seen": 25588160,
      "step": 29550
    },
    {
      "epoch": 13.934464875058934,
      "grad_norm": 1.1163748502731323,
      "learning_rate": 7.952274918443719e-06,
      "loss": 0.216,
      "num_input_tokens_seen": 25592272,
      "step": 29555
    },
    {
      "epoch": 13.936822253653936,
      "grad_norm": 2.531888961791992,
      "learning_rate": 7.945095366982983e-06,
      "loss": 0.2934,
      "num_input_tokens_seen": 25596544,
      "step": 29560
    },
    {
      "epoch": 13.939179632248939,
      "grad_norm": 0.815471351146698,
      "learning_rate": 7.937918445602871e-06,
      "loss": 0.2813,
      "num_input_tokens_seen": 25600512,
      "step": 29565
    },
    {
      "epoch": 13.941537010843941,
      "grad_norm": 1.063787817955017,
      "learning_rate": 7.930744155410145e-06,
      "loss": 0.0882,
      "num_input_tokens_seen": 25604240,
      "step": 29570
    },
    {
      "epoch": 13.943894389438944,
      "grad_norm": 0.5543541312217712,
      "learning_rate": 7.923572497511181e-06,
      "loss": 0.1741,
      "num_input_tokens_seen": 25609344,
      "step": 29575
    },
    {
      "epoch": 13.946251768033946,
      "grad_norm": 1.0965420007705688,
      "learning_rate": 7.916403473011927e-06,
      "loss": 0.1672,
      "num_input_tokens_seen": 25613728,
      "step": 29580
    },
    {
      "epoch": 13.948609146628948,
      "grad_norm": 1.7476110458374023,
      "learning_rate": 7.909237083017953e-06,
      "loss": 0.2111,
      "num_input_tokens_seen": 25618240,
      "step": 29585
    },
    {
      "epoch": 13.95096652522395,
      "grad_norm": 0.9891989827156067,
      "learning_rate": 7.902073328634389e-06,
      "loss": 0.2546,
      "num_input_tokens_seen": 25622240,
      "step": 29590
    },
    {
      "epoch": 13.953323903818953,
      "grad_norm": 0.3950904309749603,
      "learning_rate": 7.894912210965987e-06,
      "loss": 0.1573,
      "num_input_tokens_seen": 25627232,
      "step": 29595
    },
    {
      "epoch": 13.955681282413956,
      "grad_norm": 0.9837313890457153,
      "learning_rate": 7.887753731117075e-06,
      "loss": 0.1589,
      "num_input_tokens_seen": 25631728,
      "step": 29600
    },
    {
      "epoch": 13.955681282413956,
      "eval_loss": 0.2009042352437973,
      "eval_runtime": 22.1687,
      "eval_samples_per_second": 42.538,
      "eval_steps_per_second": 21.291,
      "num_input_tokens_seen": 25631728,
      "step": 29600
    },
    {
      "epoch": 13.958038661008958,
      "grad_norm": 0.8046411275863647,
      "learning_rate": 7.880597890191587e-06,
      "loss": 0.1425,
      "num_input_tokens_seen": 25635200,
      "step": 29605
    },
    {
      "epoch": 13.96039603960396,
      "grad_norm": 2.2143871784210205,
      "learning_rate": 7.873444689293036e-06,
      "loss": 0.2275,
      "num_input_tokens_seen": 25639456,
      "step": 29610
    },
    {
      "epoch": 13.962753418198963,
      "grad_norm": 1.7195502519607544,
      "learning_rate": 7.866294129524548e-06,
      "loss": 0.1725,
      "num_input_tokens_seen": 25643824,
      "step": 29615
    },
    {
      "epoch": 13.965110796793965,
      "grad_norm": 0.9429389834403992,
      "learning_rate": 7.859146211988811e-06,
      "loss": 0.195,
      "num_input_tokens_seen": 25648688,
      "step": 29620
    },
    {
      "epoch": 13.967468175388968,
      "grad_norm": 0.9099541306495667,
      "learning_rate": 7.852000937788134e-06,
      "loss": 0.1186,
      "num_input_tokens_seen": 25652464,
      "step": 29625
    },
    {
      "epoch": 13.96982555398397,
      "grad_norm": 2.2382678985595703,
      "learning_rate": 7.844858308024416e-06,
      "loss": 0.1878,
      "num_input_tokens_seen": 25657248,
      "step": 29630
    },
    {
      "epoch": 13.972182932578972,
      "grad_norm": 1.8988755941390991,
      "learning_rate": 7.837718323799122e-06,
      "loss": 0.2701,
      "num_input_tokens_seen": 25661472,
      "step": 29635
    },
    {
      "epoch": 13.974540311173975,
      "grad_norm": 1.0609549283981323,
      "learning_rate": 7.83058098621334e-06,
      "loss": 0.2101,
      "num_input_tokens_seen": 25665584,
      "step": 29640
    },
    {
      "epoch": 13.976897689768977,
      "grad_norm": 1.0633403062820435,
      "learning_rate": 7.823446296367739e-06,
      "loss": 0.1111,
      "num_input_tokens_seen": 25669536,
      "step": 29645
    },
    {
      "epoch": 13.97925506836398,
      "grad_norm": 2.2669599056243896,
      "learning_rate": 7.81631425536257e-06,
      "loss": 0.175,
      "num_input_tokens_seen": 25673584,
      "step": 29650
    },
    {
      "epoch": 13.981612446958982,
      "grad_norm": 0.7863162755966187,
      "learning_rate": 7.809184864297689e-06,
      "loss": 0.1366,
      "num_input_tokens_seen": 25678592,
      "step": 29655
    },
    {
      "epoch": 13.983969825553984,
      "grad_norm": 0.7770519852638245,
      "learning_rate": 7.802058124272532e-06,
      "loss": 0.1683,
      "num_input_tokens_seen": 25682640,
      "step": 29660
    },
    {
      "epoch": 13.986327204148987,
      "grad_norm": 0.39013558626174927,
      "learning_rate": 7.79493403638614e-06,
      "loss": 0.1942,
      "num_input_tokens_seen": 25686704,
      "step": 29665
    },
    {
      "epoch": 13.98868458274399,
      "grad_norm": 0.6215249300003052,
      "learning_rate": 7.787812601737132e-06,
      "loss": 0.1638,
      "num_input_tokens_seen": 25690832,
      "step": 29670
    },
    {
      "epoch": 13.991041961338992,
      "grad_norm": 2.4564390182495117,
      "learning_rate": 7.780693821423715e-06,
      "loss": 0.222,
      "num_input_tokens_seen": 25694864,
      "step": 29675
    },
    {
      "epoch": 13.993399339933994,
      "grad_norm": 0.6743127107620239,
      "learning_rate": 7.773577696543705e-06,
      "loss": 0.1401,
      "num_input_tokens_seen": 25699120,
      "step": 29680
    },
    {
      "epoch": 13.995756718528996,
      "grad_norm": 0.6726715564727783,
      "learning_rate": 7.7664642281945e-06,
      "loss": 0.2303,
      "num_input_tokens_seen": 25704080,
      "step": 29685
    },
    {
      "epoch": 13.998114097123999,
      "grad_norm": 1.2013952732086182,
      "learning_rate": 7.759353417473072e-06,
      "loss": 0.1494,
      "num_input_tokens_seen": 25708384,
      "step": 29690
    },
    {
      "epoch": 14.000471475719001,
      "grad_norm": 0.5386176705360413,
      "learning_rate": 7.752245265476016e-06,
      "loss": 0.2275,
      "num_input_tokens_seen": 25712288,
      "step": 29695
    },
    {
      "epoch": 14.002828854314004,
      "grad_norm": 0.6788761615753174,
      "learning_rate": 7.745139773299481e-06,
      "loss": 0.2004,
      "num_input_tokens_seen": 25716256,
      "step": 29700
    },
    {
      "epoch": 14.005186232909006,
      "grad_norm": 1.700595736503601,
      "learning_rate": 7.738036942039232e-06,
      "loss": 0.2446,
      "num_input_tokens_seen": 25720176,
      "step": 29705
    },
    {
      "epoch": 14.007543611504008,
      "grad_norm": 0.43755847215652466,
      "learning_rate": 7.73093677279062e-06,
      "loss": 0.1622,
      "num_input_tokens_seen": 25724432,
      "step": 29710
    },
    {
      "epoch": 14.009900990099009,
      "grad_norm": 0.8794018626213074,
      "learning_rate": 7.72383926664857e-06,
      "loss": 0.1904,
      "num_input_tokens_seen": 25729664,
      "step": 29715
    },
    {
      "epoch": 14.012258368694011,
      "grad_norm": 0.49887579679489136,
      "learning_rate": 7.716744424707606e-06,
      "loss": 0.1805,
      "num_input_tokens_seen": 25734256,
      "step": 29720
    },
    {
      "epoch": 14.014615747289014,
      "grad_norm": 0.8308819532394409,
      "learning_rate": 7.709652248061858e-06,
      "loss": 0.1535,
      "num_input_tokens_seen": 25738176,
      "step": 29725
    },
    {
      "epoch": 14.016973125884016,
      "grad_norm": 0.3200397491455078,
      "learning_rate": 7.702562737805017e-06,
      "loss": 0.1239,
      "num_input_tokens_seen": 25741680,
      "step": 29730
    },
    {
      "epoch": 14.019330504479019,
      "grad_norm": 2.3887429237365723,
      "learning_rate": 7.695475895030365e-06,
      "loss": 0.2739,
      "num_input_tokens_seen": 25745440,
      "step": 29735
    },
    {
      "epoch": 14.021687883074021,
      "grad_norm": 1.5061547756195068,
      "learning_rate": 7.6883917208308e-06,
      "loss": 0.2569,
      "num_input_tokens_seen": 25750128,
      "step": 29740
    },
    {
      "epoch": 14.024045261669023,
      "grad_norm": 0.44077685475349426,
      "learning_rate": 7.681310216298778e-06,
      "loss": 0.1417,
      "num_input_tokens_seen": 25755088,
      "step": 29745
    },
    {
      "epoch": 14.026402640264026,
      "grad_norm": 1.295090913772583,
      "learning_rate": 7.674231382526367e-06,
      "loss": 0.216,
      "num_input_tokens_seen": 25759072,
      "step": 29750
    },
    {
      "epoch": 14.028760018859028,
      "grad_norm": 1.7084485292434692,
      "learning_rate": 7.667155220605198e-06,
      "loss": 0.1552,
      "num_input_tokens_seen": 25763312,
      "step": 29755
    },
    {
      "epoch": 14.03111739745403,
      "grad_norm": 1.602980613708496,
      "learning_rate": 7.660081731626515e-06,
      "loss": 0.1773,
      "num_input_tokens_seen": 25767984,
      "step": 29760
    },
    {
      "epoch": 14.033474776049033,
      "grad_norm": 0.8234833478927612,
      "learning_rate": 7.653010916681141e-06,
      "loss": 0.1762,
      "num_input_tokens_seen": 25771824,
      "step": 29765
    },
    {
      "epoch": 14.035832154644035,
      "grad_norm": 0.940879762172699,
      "learning_rate": 7.645942776859472e-06,
      "loss": 0.1702,
      "num_input_tokens_seen": 25776048,
      "step": 29770
    },
    {
      "epoch": 14.038189533239038,
      "grad_norm": 0.9053022265434265,
      "learning_rate": 7.63887731325152e-06,
      "loss": 0.2502,
      "num_input_tokens_seen": 25780816,
      "step": 29775
    },
    {
      "epoch": 14.04054691183404,
      "grad_norm": 0.7610349059104919,
      "learning_rate": 7.63181452694685e-06,
      "loss": 0.1285,
      "num_input_tokens_seen": 25784992,
      "step": 29780
    },
    {
      "epoch": 14.042904290429043,
      "grad_norm": 1.3761073350906372,
      "learning_rate": 7.624754419034644e-06,
      "loss": 0.1401,
      "num_input_tokens_seen": 25788352,
      "step": 29785
    },
    {
      "epoch": 14.045261669024045,
      "grad_norm": 2.5533833503723145,
      "learning_rate": 7.6176969906036645e-06,
      "loss": 0.1707,
      "num_input_tokens_seen": 25792400,
      "step": 29790
    },
    {
      "epoch": 14.047619047619047,
      "grad_norm": 0.6368961930274963,
      "learning_rate": 7.610642242742242e-06,
      "loss": 0.1209,
      "num_input_tokens_seen": 25796800,
      "step": 29795
    },
    {
      "epoch": 14.04997642621405,
      "grad_norm": 1.404702067375183,
      "learning_rate": 7.603590176538322e-06,
      "loss": 0.2079,
      "num_input_tokens_seen": 25801056,
      "step": 29800
    },
    {
      "epoch": 14.04997642621405,
      "eval_loss": 0.20244766771793365,
      "eval_runtime": 22.2041,
      "eval_samples_per_second": 42.47,
      "eval_steps_per_second": 21.257,
      "num_input_tokens_seen": 25801056,
      "step": 29800
    },
    {
      "epoch": 14.052333804809052,
      "grad_norm": 0.859574019908905,
      "learning_rate": 7.596540793079404e-06,
      "loss": 0.143,
      "num_input_tokens_seen": 25805520,
      "step": 29805
    },
    {
      "epoch": 14.054691183404055,
      "grad_norm": 0.8265509009361267,
      "learning_rate": 7.5894940934526125e-06,
      "loss": 0.2932,
      "num_input_tokens_seen": 25810032,
      "step": 29810
    },
    {
      "epoch": 14.057048561999057,
      "grad_norm": 0.8525673151016235,
      "learning_rate": 7.582450078744621e-06,
      "loss": 0.1662,
      "num_input_tokens_seen": 25814080,
      "step": 29815
    },
    {
      "epoch": 14.05940594059406,
      "grad_norm": 1.5994199514389038,
      "learning_rate": 7.575408750041707e-06,
      "loss": 0.2055,
      "num_input_tokens_seen": 25818144,
      "step": 29820
    },
    {
      "epoch": 14.061763319189062,
      "grad_norm": 0.4389601945877075,
      "learning_rate": 7.568370108429732e-06,
      "loss": 0.1517,
      "num_input_tokens_seen": 25822496,
      "step": 29825
    },
    {
      "epoch": 14.064120697784064,
      "grad_norm": 1.1231781244277954,
      "learning_rate": 7.561334154994154e-06,
      "loss": 0.2084,
      "num_input_tokens_seen": 25826768,
      "step": 29830
    },
    {
      "epoch": 14.066478076379067,
      "grad_norm": 0.8466101884841919,
      "learning_rate": 7.55430089081999e-06,
      "loss": 0.1386,
      "num_input_tokens_seen": 25831408,
      "step": 29835
    },
    {
      "epoch": 14.068835454974069,
      "grad_norm": 0.38674160838127136,
      "learning_rate": 7.547270316991864e-06,
      "loss": 0.2673,
      "num_input_tokens_seen": 25835744,
      "step": 29840
    },
    {
      "epoch": 14.071192833569071,
      "grad_norm": 0.8618287444114685,
      "learning_rate": 7.5402424345939884e-06,
      "loss": 0.1176,
      "num_input_tokens_seen": 25840144,
      "step": 29845
    },
    {
      "epoch": 14.073550212164074,
      "grad_norm": 1.1051273345947266,
      "learning_rate": 7.533217244710133e-06,
      "loss": 0.1772,
      "num_input_tokens_seen": 25844736,
      "step": 29850
    },
    {
      "epoch": 14.075907590759076,
      "grad_norm": 0.6949992179870605,
      "learning_rate": 7.52619474842369e-06,
      "loss": 0.105,
      "num_input_tokens_seen": 25849056,
      "step": 29855
    },
    {
      "epoch": 14.078264969354079,
      "grad_norm": 0.5015005469322205,
      "learning_rate": 7.519174946817597e-06,
      "loss": 0.1784,
      "num_input_tokens_seen": 25852992,
      "step": 29860
    },
    {
      "epoch": 14.080622347949081,
      "grad_norm": 0.7386866807937622,
      "learning_rate": 7.512157840974407e-06,
      "loss": 0.2134,
      "num_input_tokens_seen": 25856880,
      "step": 29865
    },
    {
      "epoch": 14.082979726544083,
      "grad_norm": 1.3026291131973267,
      "learning_rate": 7.5051434319762496e-06,
      "loss": 0.1326,
      "num_input_tokens_seen": 25860864,
      "step": 29870
    },
    {
      "epoch": 14.085337105139086,
      "grad_norm": 0.7796523571014404,
      "learning_rate": 7.498131720904822e-06,
      "loss": 0.2357,
      "num_input_tokens_seen": 25865616,
      "step": 29875
    },
    {
      "epoch": 14.087694483734088,
      "grad_norm": 1.7463781833648682,
      "learning_rate": 7.491122708841433e-06,
      "loss": 0.2027,
      "num_input_tokens_seen": 25869520,
      "step": 29880
    },
    {
      "epoch": 14.09005186232909,
      "grad_norm": 0.7175498008728027,
      "learning_rate": 7.4841163968669524e-06,
      "loss": 0.2345,
      "num_input_tokens_seen": 25873776,
      "step": 29885
    },
    {
      "epoch": 14.092409240924093,
      "grad_norm": 0.8169807195663452,
      "learning_rate": 7.4771127860618355e-06,
      "loss": 0.1424,
      "num_input_tokens_seen": 25878528,
      "step": 29890
    },
    {
      "epoch": 14.094766619519095,
      "grad_norm": 0.9757396578788757,
      "learning_rate": 7.470111877506139e-06,
      "loss": 0.1311,
      "num_input_tokens_seen": 25883376,
      "step": 29895
    },
    {
      "epoch": 14.097123998114098,
      "grad_norm": 2.1657168865203857,
      "learning_rate": 7.463113672279479e-06,
      "loss": 0.1574,
      "num_input_tokens_seen": 25887552,
      "step": 29900
    },
    {
      "epoch": 14.0994813767091,
      "grad_norm": 0.9285408854484558,
      "learning_rate": 7.456118171461071e-06,
      "loss": 0.1807,
      "num_input_tokens_seen": 25891664,
      "step": 29905
    },
    {
      "epoch": 14.101838755304103,
      "grad_norm": 0.869805097579956,
      "learning_rate": 7.449125376129721e-06,
      "loss": 0.1752,
      "num_input_tokens_seen": 25895584,
      "step": 29910
    },
    {
      "epoch": 14.104196133899103,
      "grad_norm": 0.510375440120697,
      "learning_rate": 7.442135287363788e-06,
      "loss": 0.2283,
      "num_input_tokens_seen": 25899584,
      "step": 29915
    },
    {
      "epoch": 14.106553512494106,
      "grad_norm": 0.5291690230369568,
      "learning_rate": 7.435147906241247e-06,
      "loss": 0.1693,
      "num_input_tokens_seen": 25904576,
      "step": 29920
    },
    {
      "epoch": 14.108910891089108,
      "grad_norm": 0.8979179859161377,
      "learning_rate": 7.428163233839624e-06,
      "loss": 0.2011,
      "num_input_tokens_seen": 25909168,
      "step": 29925
    },
    {
      "epoch": 14.11126826968411,
      "grad_norm": 1.176875114440918,
      "learning_rate": 7.4211812712360525e-06,
      "loss": 0.1617,
      "num_input_tokens_seen": 25914672,
      "step": 29930
    },
    {
      "epoch": 14.113625648279113,
      "grad_norm": 1.0800586938858032,
      "learning_rate": 7.4142020195072464e-06,
      "loss": 0.1675,
      "num_input_tokens_seen": 25918960,
      "step": 29935
    },
    {
      "epoch": 14.115983026874115,
      "grad_norm": 0.9604811668395996,
      "learning_rate": 7.407225479729479e-06,
      "loss": 0.2146,
      "num_input_tokens_seen": 25923808,
      "step": 29940
    },
    {
      "epoch": 14.118340405469118,
      "grad_norm": 0.7625590562820435,
      "learning_rate": 7.400251652978632e-06,
      "loss": 0.1441,
      "num_input_tokens_seen": 25928528,
      "step": 29945
    },
    {
      "epoch": 14.12069778406412,
      "grad_norm": 0.686346709728241,
      "learning_rate": 7.393280540330147e-06,
      "loss": 0.1355,
      "num_input_tokens_seen": 25933408,
      "step": 29950
    },
    {
      "epoch": 14.123055162659123,
      "grad_norm": 1.3367431163787842,
      "learning_rate": 7.386312142859069e-06,
      "loss": 0.1815,
      "num_input_tokens_seen": 25938464,
      "step": 29955
    },
    {
      "epoch": 14.125412541254125,
      "grad_norm": 0.5973284244537354,
      "learning_rate": 7.379346461640008e-06,
      "loss": 0.1944,
      "num_input_tokens_seen": 25943168,
      "step": 29960
    },
    {
      "epoch": 14.127769919849127,
      "grad_norm": 1.3081449270248413,
      "learning_rate": 7.372383497747149e-06,
      "loss": 0.1868,
      "num_input_tokens_seen": 25947872,
      "step": 29965
    },
    {
      "epoch": 14.13012729844413,
      "grad_norm": 0.9822768568992615,
      "learning_rate": 7.3654232522542775e-06,
      "loss": 0.1219,
      "num_input_tokens_seen": 25952608,
      "step": 29970
    },
    {
      "epoch": 14.132484677039132,
      "grad_norm": 1.675977349281311,
      "learning_rate": 7.358465726234756e-06,
      "loss": 0.1611,
      "num_input_tokens_seen": 25956704,
      "step": 29975
    },
    {
      "epoch": 14.134842055634135,
      "grad_norm": 1.2574306726455688,
      "learning_rate": 7.351510920761512e-06,
      "loss": 0.165,
      "num_input_tokens_seen": 25962080,
      "step": 29980
    },
    {
      "epoch": 14.137199434229137,
      "grad_norm": 0.6249713897705078,
      "learning_rate": 7.344558836907067e-06,
      "loss": 0.1081,
      "num_input_tokens_seen": 25966352,
      "step": 29985
    },
    {
      "epoch": 14.13955681282414,
      "grad_norm": 0.9872996211051941,
      "learning_rate": 7.3376094757435285e-06,
      "loss": 0.2125,
      "num_input_tokens_seen": 25970224,
      "step": 29990
    },
    {
      "epoch": 14.141914191419142,
      "grad_norm": 1.1222649812698364,
      "learning_rate": 7.330662838342561e-06,
      "loss": 0.1893,
      "num_input_tokens_seen": 25973968,
      "step": 29995
    },
    {
      "epoch": 14.144271570014144,
      "grad_norm": 0.6633251309394836,
      "learning_rate": 7.323718925775438e-06,
      "loss": 0.2091,
      "num_input_tokens_seen": 25978896,
      "step": 30000
    },
    {
      "epoch": 14.144271570014144,
      "eval_loss": 0.20048509538173676,
      "eval_runtime": 22.2223,
      "eval_samples_per_second": 42.435,
      "eval_steps_per_second": 21.24,
      "num_input_tokens_seen": 25978896,
      "step": 30000
    },
    {
      "epoch": 14.146628948609147,
      "grad_norm": 0.8074001669883728,
      "learning_rate": 7.316777739112985e-06,
      "loss": 0.2132,
      "num_input_tokens_seen": 25983040,
      "step": 30005
    },
    {
      "epoch": 14.148986327204149,
      "grad_norm": 0.9050359725952148,
      "learning_rate": 7.309839279425626e-06,
      "loss": 0.219,
      "num_input_tokens_seen": 25987488,
      "step": 30010
    },
    {
      "epoch": 14.151343705799151,
      "grad_norm": 0.6945421099662781,
      "learning_rate": 7.302903547783366e-06,
      "loss": 0.1267,
      "num_input_tokens_seen": 25992016,
      "step": 30015
    },
    {
      "epoch": 14.153701084394154,
      "grad_norm": 0.5621294975280762,
      "learning_rate": 7.2959705452557644e-06,
      "loss": 0.1336,
      "num_input_tokens_seen": 25996160,
      "step": 30020
    },
    {
      "epoch": 14.156058462989156,
      "grad_norm": 0.46363794803619385,
      "learning_rate": 7.289040272911996e-06,
      "loss": 0.1416,
      "num_input_tokens_seen": 26000816,
      "step": 30025
    },
    {
      "epoch": 14.158415841584159,
      "grad_norm": 0.8889262676239014,
      "learning_rate": 7.282112731820789e-06,
      "loss": 0.1593,
      "num_input_tokens_seen": 26005232,
      "step": 30030
    },
    {
      "epoch": 14.160773220179161,
      "grad_norm": 0.2986850142478943,
      "learning_rate": 7.275187923050447e-06,
      "loss": 0.1303,
      "num_input_tokens_seen": 26009088,
      "step": 30035
    },
    {
      "epoch": 14.163130598774163,
      "grad_norm": 0.7800512909889221,
      "learning_rate": 7.268265847668879e-06,
      "loss": 0.1199,
      "num_input_tokens_seen": 26013232,
      "step": 30040
    },
    {
      "epoch": 14.165487977369166,
      "grad_norm": 0.6441956162452698,
      "learning_rate": 7.261346506743538e-06,
      "loss": 0.1908,
      "num_input_tokens_seen": 26016976,
      "step": 30045
    },
    {
      "epoch": 14.167845355964168,
      "grad_norm": 0.9332231283187866,
      "learning_rate": 7.254429901341486e-06,
      "loss": 0.1407,
      "num_input_tokens_seen": 26021872,
      "step": 30050
    },
    {
      "epoch": 14.17020273455917,
      "grad_norm": 1.6628925800323486,
      "learning_rate": 7.247516032529356e-06,
      "loss": 0.1752,
      "num_input_tokens_seen": 26026352,
      "step": 30055
    },
    {
      "epoch": 14.172560113154173,
      "grad_norm": 1.3779560327529907,
      "learning_rate": 7.240604901373338e-06,
      "loss": 0.2621,
      "num_input_tokens_seen": 26031728,
      "step": 30060
    },
    {
      "epoch": 14.174917491749175,
      "grad_norm": 1.0996778011322021,
      "learning_rate": 7.233696508939223e-06,
      "loss": 0.1338,
      "num_input_tokens_seen": 26035840,
      "step": 30065
    },
    {
      "epoch": 14.177274870344178,
      "grad_norm": 1.3214828968048096,
      "learning_rate": 7.226790856292376e-06,
      "loss": 0.1302,
      "num_input_tokens_seen": 26039360,
      "step": 30070
    },
    {
      "epoch": 14.17963224893918,
      "grad_norm": 0.30222681164741516,
      "learning_rate": 7.219887944497727e-06,
      "loss": 0.1312,
      "num_input_tokens_seen": 26043696,
      "step": 30075
    },
    {
      "epoch": 14.181989627534183,
      "grad_norm": 0.684980034828186,
      "learning_rate": 7.2129877746198e-06,
      "loss": 0.2186,
      "num_input_tokens_seen": 26047680,
      "step": 30080
    },
    {
      "epoch": 14.184347006129185,
      "grad_norm": 0.856623649597168,
      "learning_rate": 7.20609034772268e-06,
      "loss": 0.2082,
      "num_input_tokens_seen": 26052208,
      "step": 30085
    },
    {
      "epoch": 14.186704384724187,
      "grad_norm": 0.8984589576721191,
      "learning_rate": 7.19919566487004e-06,
      "loss": 0.2259,
      "num_input_tokens_seen": 26057408,
      "step": 30090
    },
    {
      "epoch": 14.18906176331919,
      "grad_norm": 0.5946028828620911,
      "learning_rate": 7.192303727125132e-06,
      "loss": 0.1164,
      "num_input_tokens_seen": 26062224,
      "step": 30095
    },
    {
      "epoch": 14.191419141914192,
      "grad_norm": 0.978752613067627,
      "learning_rate": 7.185414535550777e-06,
      "loss": 0.174,
      "num_input_tokens_seen": 26067664,
      "step": 30100
    },
    {
      "epoch": 14.193776520509195,
      "grad_norm": 1.2369714975357056,
      "learning_rate": 7.178528091209363e-06,
      "loss": 0.1555,
      "num_input_tokens_seen": 26071520,
      "step": 30105
    },
    {
      "epoch": 14.196133899104197,
      "grad_norm": 0.8613169193267822,
      "learning_rate": 7.171644395162888e-06,
      "loss": 0.2152,
      "num_input_tokens_seen": 26076064,
      "step": 30110
    },
    {
      "epoch": 14.198491277699198,
      "grad_norm": 2.018399715423584,
      "learning_rate": 7.164763448472881e-06,
      "loss": 0.1905,
      "num_input_tokens_seen": 26080656,
      "step": 30115
    },
    {
      "epoch": 14.2008486562942,
      "grad_norm": 1.3949519395828247,
      "learning_rate": 7.157885252200491e-06,
      "loss": 0.2011,
      "num_input_tokens_seen": 26085328,
      "step": 30120
    },
    {
      "epoch": 14.203206034889202,
      "grad_norm": 0.7739353179931641,
      "learning_rate": 7.151009807406403e-06,
      "loss": 0.1434,
      "num_input_tokens_seen": 26089568,
      "step": 30125
    },
    {
      "epoch": 14.205563413484205,
      "grad_norm": 0.979504406452179,
      "learning_rate": 7.144137115150909e-06,
      "loss": 0.1497,
      "num_input_tokens_seen": 26094128,
      "step": 30130
    },
    {
      "epoch": 14.207920792079207,
      "grad_norm": 1.2214607000350952,
      "learning_rate": 7.1372671764938725e-06,
      "loss": 0.2316,
      "num_input_tokens_seen": 26098976,
      "step": 30135
    },
    {
      "epoch": 14.21027817067421,
      "grad_norm": 1.5208758115768433,
      "learning_rate": 7.130399992494705e-06,
      "loss": 0.1757,
      "num_input_tokens_seen": 26103744,
      "step": 30140
    },
    {
      "epoch": 14.212635549269212,
      "grad_norm": 1.5494253635406494,
      "learning_rate": 7.123535564212419e-06,
      "loss": 0.1868,
      "num_input_tokens_seen": 26107776,
      "step": 30145
    },
    {
      "epoch": 14.214992927864214,
      "grad_norm": 0.7926974296569824,
      "learning_rate": 7.116673892705611e-06,
      "loss": 0.1628,
      "num_input_tokens_seen": 26112032,
      "step": 30150
    },
    {
      "epoch": 14.217350306459217,
      "grad_norm": 0.5255433320999146,
      "learning_rate": 7.109814979032415e-06,
      "loss": 0.1824,
      "num_input_tokens_seen": 26116384,
      "step": 30155
    },
    {
      "epoch": 14.21970768505422,
      "grad_norm": 1.8813599348068237,
      "learning_rate": 7.102958824250577e-06,
      "loss": 0.112,
      "num_input_tokens_seen": 26120640,
      "step": 30160
    },
    {
      "epoch": 14.222065063649222,
      "grad_norm": 1.1951700448989868,
      "learning_rate": 7.096105429417393e-06,
      "loss": 0.1996,
      "num_input_tokens_seen": 26125216,
      "step": 30165
    },
    {
      "epoch": 14.224422442244224,
      "grad_norm": 0.4826122522354126,
      "learning_rate": 7.0892547955897506e-06,
      "loss": 0.1422,
      "num_input_tokens_seen": 26129376,
      "step": 30170
    },
    {
      "epoch": 14.226779820839226,
      "grad_norm": 1.3954354524612427,
      "learning_rate": 7.0824069238241e-06,
      "loss": 0.225,
      "num_input_tokens_seen": 26133616,
      "step": 30175
    },
    {
      "epoch": 14.229137199434229,
      "grad_norm": 0.88782799243927,
      "learning_rate": 7.075561815176462e-06,
      "loss": 0.2088,
      "num_input_tokens_seen": 26137392,
      "step": 30180
    },
    {
      "epoch": 14.231494578029231,
      "grad_norm": 0.6825718283653259,
      "learning_rate": 7.068719470702445e-06,
      "loss": 0.1509,
      "num_input_tokens_seen": 26141616,
      "step": 30185
    },
    {
      "epoch": 14.233851956624234,
      "grad_norm": 0.8214496970176697,
      "learning_rate": 7.061879891457229e-06,
      "loss": 0.2403,
      "num_input_tokens_seen": 26146592,
      "step": 30190
    },
    {
      "epoch": 14.236209335219236,
      "grad_norm": 1.02255380153656,
      "learning_rate": 7.0550430784955515e-06,
      "loss": 0.2113,
      "num_input_tokens_seen": 26150960,
      "step": 30195
    },
    {
      "epoch": 14.238566713814238,
      "grad_norm": 0.737054169178009,
      "learning_rate": 7.048209032871752e-06,
      "loss": 0.1508,
      "num_input_tokens_seen": 26156672,
      "step": 30200
    },
    {
      "epoch": 14.238566713814238,
      "eval_loss": 0.20047873258590698,
      "eval_runtime": 22.1452,
      "eval_samples_per_second": 42.583,
      "eval_steps_per_second": 21.314,
      "num_input_tokens_seen": 26156672,
      "step": 30200
    },
    {
      "epoch": 14.24092409240924,
      "grad_norm": 0.8306350708007812,
      "learning_rate": 7.0413777556397055e-06,
      "loss": 0.1668,
      "num_input_tokens_seen": 26161312,
      "step": 30205
    },
    {
      "epoch": 14.243281471004243,
      "grad_norm": 0.47472935914993286,
      "learning_rate": 7.0345492478528925e-06,
      "loss": 0.1746,
      "num_input_tokens_seen": 26165568,
      "step": 30210
    },
    {
      "epoch": 14.245638849599246,
      "grad_norm": 1.3130947351455688,
      "learning_rate": 7.02772351056436e-06,
      "loss": 0.133,
      "num_input_tokens_seen": 26169584,
      "step": 30215
    },
    {
      "epoch": 14.247996228194248,
      "grad_norm": 0.6980262398719788,
      "learning_rate": 7.020900544826709e-06,
      "loss": 0.2059,
      "num_input_tokens_seen": 26174208,
      "step": 30220
    },
    {
      "epoch": 14.25035360678925,
      "grad_norm": 1.2901676893234253,
      "learning_rate": 7.014080351692134e-06,
      "loss": 0.2448,
      "num_input_tokens_seen": 26179088,
      "step": 30225
    },
    {
      "epoch": 14.252710985384253,
      "grad_norm": 1.4662622213363647,
      "learning_rate": 7.0072629322124024e-06,
      "loss": 0.2108,
      "num_input_tokens_seen": 26182768,
      "step": 30230
    },
    {
      "epoch": 14.255068363979255,
      "grad_norm": 0.4025469422340393,
      "learning_rate": 7.000448287438827e-06,
      "loss": 0.2179,
      "num_input_tokens_seen": 26186880,
      "step": 30235
    },
    {
      "epoch": 14.257425742574258,
      "grad_norm": 1.1736482381820679,
      "learning_rate": 6.993636418422331e-06,
      "loss": 0.2479,
      "num_input_tokens_seen": 26191472,
      "step": 30240
    },
    {
      "epoch": 14.25978312116926,
      "grad_norm": 1.339173674583435,
      "learning_rate": 6.986827326213383e-06,
      "loss": 0.2485,
      "num_input_tokens_seen": 26197504,
      "step": 30245
    },
    {
      "epoch": 14.262140499764262,
      "grad_norm": 0.6706222891807556,
      "learning_rate": 6.9800210118620205e-06,
      "loss": 0.1289,
      "num_input_tokens_seen": 26201056,
      "step": 30250
    },
    {
      "epoch": 14.264497878359265,
      "grad_norm": 2.1006383895874023,
      "learning_rate": 6.973217476417876e-06,
      "loss": 0.2609,
      "num_input_tokens_seen": 26205952,
      "step": 30255
    },
    {
      "epoch": 14.266855256954267,
      "grad_norm": 1.078908920288086,
      "learning_rate": 6.96641672093013e-06,
      "loss": 0.162,
      "num_input_tokens_seen": 26209904,
      "step": 30260
    },
    {
      "epoch": 14.26921263554927,
      "grad_norm": 0.743901789188385,
      "learning_rate": 6.95961874644755e-06,
      "loss": 0.1885,
      "num_input_tokens_seen": 26214720,
      "step": 30265
    },
    {
      "epoch": 14.271570014144272,
      "grad_norm": 2.3768727779388428,
      "learning_rate": 6.952823554018476e-06,
      "loss": 0.2122,
      "num_input_tokens_seen": 26219120,
      "step": 30270
    },
    {
      "epoch": 14.273927392739274,
      "grad_norm": 0.6880722045898438,
      "learning_rate": 6.946031144690798e-06,
      "loss": 0.1635,
      "num_input_tokens_seen": 26223776,
      "step": 30275
    },
    {
      "epoch": 14.276284771334277,
      "grad_norm": 1.159741997718811,
      "learning_rate": 6.939241519512005e-06,
      "loss": 0.1893,
      "num_input_tokens_seen": 26228448,
      "step": 30280
    },
    {
      "epoch": 14.27864214992928,
      "grad_norm": 1.331396222114563,
      "learning_rate": 6.932454679529129e-06,
      "loss": 0.146,
      "num_input_tokens_seen": 26233024,
      "step": 30285
    },
    {
      "epoch": 14.280999528524282,
      "grad_norm": 1.0323938131332397,
      "learning_rate": 6.925670625788791e-06,
      "loss": 0.21,
      "num_input_tokens_seen": 26238400,
      "step": 30290
    },
    {
      "epoch": 14.283356907119284,
      "grad_norm": 0.7331550121307373,
      "learning_rate": 6.918889359337186e-06,
      "loss": 0.2119,
      "num_input_tokens_seen": 26242304,
      "step": 30295
    },
    {
      "epoch": 14.285714285714286,
      "grad_norm": 1.6289966106414795,
      "learning_rate": 6.912110881220058e-06,
      "loss": 0.2108,
      "num_input_tokens_seen": 26246752,
      "step": 30300
    },
    {
      "epoch": 14.288071664309289,
      "grad_norm": 0.6080501675605774,
      "learning_rate": 6.905335192482735e-06,
      "loss": 0.118,
      "num_input_tokens_seen": 26251440,
      "step": 30305
    },
    {
      "epoch": 14.290429042904291,
      "grad_norm": 1.4211194515228271,
      "learning_rate": 6.8985622941701275e-06,
      "loss": 0.1525,
      "num_input_tokens_seen": 26255776,
      "step": 30310
    },
    {
      "epoch": 14.292786421499294,
      "grad_norm": 0.682325005531311,
      "learning_rate": 6.89179218732669e-06,
      "loss": 0.143,
      "num_input_tokens_seen": 26259888,
      "step": 30315
    },
    {
      "epoch": 14.295143800094294,
      "grad_norm": 0.40044018626213074,
      "learning_rate": 6.8850248729964595e-06,
      "loss": 0.1988,
      "num_input_tokens_seen": 26263712,
      "step": 30320
    },
    {
      "epoch": 14.297501178689297,
      "grad_norm": 1.892099142074585,
      "learning_rate": 6.8782603522230314e-06,
      "loss": 0.2232,
      "num_input_tokens_seen": 26267760,
      "step": 30325
    },
    {
      "epoch": 14.299858557284299,
      "grad_norm": 1.2333959341049194,
      "learning_rate": 6.871498626049591e-06,
      "loss": 0.2108,
      "num_input_tokens_seen": 26272304,
      "step": 30330
    },
    {
      "epoch": 14.302215935879302,
      "grad_norm": 2.291926383972168,
      "learning_rate": 6.8647396955188875e-06,
      "loss": 0.2233,
      "num_input_tokens_seen": 26276784,
      "step": 30335
    },
    {
      "epoch": 14.304573314474304,
      "grad_norm": 1.19161057472229,
      "learning_rate": 6.857983561673218e-06,
      "loss": 0.1681,
      "num_input_tokens_seen": 26281104,
      "step": 30340
    },
    {
      "epoch": 14.306930693069306,
      "grad_norm": 1.1627495288848877,
      "learning_rate": 6.851230225554467e-06,
      "loss": 0.1228,
      "num_input_tokens_seen": 26285536,
      "step": 30345
    },
    {
      "epoch": 14.309288071664309,
      "grad_norm": 1.4332118034362793,
      "learning_rate": 6.8444796882040946e-06,
      "loss": 0.231,
      "num_input_tokens_seen": 26289616,
      "step": 30350
    },
    {
      "epoch": 14.311645450259311,
      "grad_norm": 1.1448805332183838,
      "learning_rate": 6.837731950663106e-06,
      "loss": 0.1767,
      "num_input_tokens_seen": 26294000,
      "step": 30355
    },
    {
      "epoch": 14.314002828854314,
      "grad_norm": 0.984918475151062,
      "learning_rate": 6.830987013972098e-06,
      "loss": 0.1966,
      "num_input_tokens_seen": 26298496,
      "step": 30360
    },
    {
      "epoch": 14.316360207449316,
      "grad_norm": 0.5350186824798584,
      "learning_rate": 6.82424487917121e-06,
      "loss": 0.1369,
      "num_input_tokens_seen": 26302880,
      "step": 30365
    },
    {
      "epoch": 14.318717586044318,
      "grad_norm": 0.48747876286506653,
      "learning_rate": 6.8175055473001735e-06,
      "loss": 0.2095,
      "num_input_tokens_seen": 26307008,
      "step": 30370
    },
    {
      "epoch": 14.32107496463932,
      "grad_norm": 2.1879384517669678,
      "learning_rate": 6.8107690193982855e-06,
      "loss": 0.2283,
      "num_input_tokens_seen": 26310720,
      "step": 30375
    },
    {
      "epoch": 14.323432343234323,
      "grad_norm": 1.2297968864440918,
      "learning_rate": 6.804035296504385e-06,
      "loss": 0.2309,
      "num_input_tokens_seen": 26314832,
      "step": 30380
    },
    {
      "epoch": 14.325789721829326,
      "grad_norm": 0.6013527512550354,
      "learning_rate": 6.797304379656916e-06,
      "loss": 0.1583,
      "num_input_tokens_seen": 26319152,
      "step": 30385
    },
    {
      "epoch": 14.328147100424328,
      "grad_norm": 0.7956259846687317,
      "learning_rate": 6.790576269893861e-06,
      "loss": 0.1855,
      "num_input_tokens_seen": 26322992,
      "step": 30390
    },
    {
      "epoch": 14.33050447901933,
      "grad_norm": 1.452505111694336,
      "learning_rate": 6.783850968252772e-06,
      "loss": 0.2107,
      "num_input_tokens_seen": 26326352,
      "step": 30395
    },
    {
      "epoch": 14.332861857614333,
      "grad_norm": 0.9270124435424805,
      "learning_rate": 6.777128475770789e-06,
      "loss": 0.1096,
      "num_input_tokens_seen": 26330592,
      "step": 30400
    },
    {
      "epoch": 14.332861857614333,
      "eval_loss": 0.1997813880443573,
      "eval_runtime": 22.1498,
      "eval_samples_per_second": 42.574,
      "eval_steps_per_second": 21.309,
      "num_input_tokens_seen": 26330592,
      "step": 30400
    },
    {
      "epoch": 14.335219236209335,
      "grad_norm": 1.086334466934204,
      "learning_rate": 6.77040879348459e-06,
      "loss": 0.1557,
      "num_input_tokens_seen": 26334704,
      "step": 30405
    },
    {
      "epoch": 14.337576614804338,
      "grad_norm": 0.6392337679862976,
      "learning_rate": 6.763691922430443e-06,
      "loss": 0.1508,
      "num_input_tokens_seen": 26340400,
      "step": 30410
    },
    {
      "epoch": 14.33993399339934,
      "grad_norm": 2.553370475769043,
      "learning_rate": 6.756977863644178e-06,
      "loss": 0.1958,
      "num_input_tokens_seen": 26344192,
      "step": 30415
    },
    {
      "epoch": 14.342291371994342,
      "grad_norm": 0.6296906471252441,
      "learning_rate": 6.7502666181611804e-06,
      "loss": 0.1877,
      "num_input_tokens_seen": 26348128,
      "step": 30420
    },
    {
      "epoch": 14.344648750589345,
      "grad_norm": 0.9252628087997437,
      "learning_rate": 6.743558187016405e-06,
      "loss": 0.1903,
      "num_input_tokens_seen": 26352320,
      "step": 30425
    },
    {
      "epoch": 14.347006129184347,
      "grad_norm": 1.9359976053237915,
      "learning_rate": 6.7368525712443925e-06,
      "loss": 0.1583,
      "num_input_tokens_seen": 26356384,
      "step": 30430
    },
    {
      "epoch": 14.34936350777935,
      "grad_norm": 0.22676345705986023,
      "learning_rate": 6.7301497718792155e-06,
      "loss": 0.1314,
      "num_input_tokens_seen": 26361152,
      "step": 30435
    },
    {
      "epoch": 14.351720886374352,
      "grad_norm": 0.6473042964935303,
      "learning_rate": 6.723449789954544e-06,
      "loss": 0.1122,
      "num_input_tokens_seen": 26366256,
      "step": 30440
    },
    {
      "epoch": 14.354078264969354,
      "grad_norm": 0.6040889024734497,
      "learning_rate": 6.716752626503586e-06,
      "loss": 0.1445,
      "num_input_tokens_seen": 26370368,
      "step": 30445
    },
    {
      "epoch": 14.356435643564357,
      "grad_norm": 1.7259962558746338,
      "learning_rate": 6.710058282559131e-06,
      "loss": 0.249,
      "num_input_tokens_seen": 26374864,
      "step": 30450
    },
    {
      "epoch": 14.35879302215936,
      "grad_norm": 1.6760108470916748,
      "learning_rate": 6.703366759153545e-06,
      "loss": 0.21,
      "num_input_tokens_seen": 26378960,
      "step": 30455
    },
    {
      "epoch": 14.361150400754362,
      "grad_norm": 0.6960738897323608,
      "learning_rate": 6.6966780573187335e-06,
      "loss": 0.1979,
      "num_input_tokens_seen": 26383616,
      "step": 30460
    },
    {
      "epoch": 14.363507779349364,
      "grad_norm": 0.7321069836616516,
      "learning_rate": 6.689992178086174e-06,
      "loss": 0.1708,
      "num_input_tokens_seen": 26387696,
      "step": 30465
    },
    {
      "epoch": 14.365865157944366,
      "grad_norm": 1.0523029565811157,
      "learning_rate": 6.683309122486925e-06,
      "loss": 0.1909,
      "num_input_tokens_seen": 26391312,
      "step": 30470
    },
    {
      "epoch": 14.368222536539369,
      "grad_norm": 1.0545380115509033,
      "learning_rate": 6.676628891551584e-06,
      "loss": 0.1602,
      "num_input_tokens_seen": 26395440,
      "step": 30475
    },
    {
      "epoch": 14.370579915134371,
      "grad_norm": 1.336134672164917,
      "learning_rate": 6.6699514863103385e-06,
      "loss": 0.1533,
      "num_input_tokens_seen": 26399616,
      "step": 30480
    },
    {
      "epoch": 14.372937293729374,
      "grad_norm": 0.7106510996818542,
      "learning_rate": 6.663276907792921e-06,
      "loss": 0.1671,
      "num_input_tokens_seen": 26404800,
      "step": 30485
    },
    {
      "epoch": 14.375294672324376,
      "grad_norm": 1.5315874814987183,
      "learning_rate": 6.656605157028634e-06,
      "loss": 0.2594,
      "num_input_tokens_seen": 26408848,
      "step": 30490
    },
    {
      "epoch": 14.377652050919378,
      "grad_norm": 0.7086440324783325,
      "learning_rate": 6.649936235046358e-06,
      "loss": 0.177,
      "num_input_tokens_seen": 26413328,
      "step": 30495
    },
    {
      "epoch": 14.38000942951438,
      "grad_norm": 0.3944576382637024,
      "learning_rate": 6.643270142874508e-06,
      "loss": 0.2467,
      "num_input_tokens_seen": 26417072,
      "step": 30500
    },
    {
      "epoch": 14.382366808109383,
      "grad_norm": 0.37743058800697327,
      "learning_rate": 6.636606881541094e-06,
      "loss": 0.1496,
      "num_input_tokens_seen": 26420544,
      "step": 30505
    },
    {
      "epoch": 14.384724186704386,
      "grad_norm": 0.42643922567367554,
      "learning_rate": 6.629946452073662e-06,
      "loss": 0.1546,
      "num_input_tokens_seen": 26424464,
      "step": 30510
    },
    {
      "epoch": 14.387081565299386,
      "grad_norm": 0.5701088309288025,
      "learning_rate": 6.6232888554993375e-06,
      "loss": 0.1936,
      "num_input_tokens_seen": 26428464,
      "step": 30515
    },
    {
      "epoch": 14.389438943894389,
      "grad_norm": 0.9266867637634277,
      "learning_rate": 6.616634092844817e-06,
      "loss": 0.1915,
      "num_input_tokens_seen": 26432704,
      "step": 30520
    },
    {
      "epoch": 14.391796322489391,
      "grad_norm": 1.101103663444519,
      "learning_rate": 6.609982165136331e-06,
      "loss": 0.1634,
      "num_input_tokens_seen": 26437472,
      "step": 30525
    },
    {
      "epoch": 14.394153701084393,
      "grad_norm": 0.7767329812049866,
      "learning_rate": 6.603333073399706e-06,
      "loss": 0.1384,
      "num_input_tokens_seen": 26441680,
      "step": 30530
    },
    {
      "epoch": 14.396511079679396,
      "grad_norm": 2.3046669960021973,
      "learning_rate": 6.596686818660308e-06,
      "loss": 0.2406,
      "num_input_tokens_seen": 26446608,
      "step": 30535
    },
    {
      "epoch": 14.398868458274398,
      "grad_norm": 0.658165454864502,
      "learning_rate": 6.590043401943066e-06,
      "loss": 0.1499,
      "num_input_tokens_seen": 26451152,
      "step": 30540
    },
    {
      "epoch": 14.4012258368694,
      "grad_norm": 1.0118002891540527,
      "learning_rate": 6.583402824272494e-06,
      "loss": 0.2279,
      "num_input_tokens_seen": 26455328,
      "step": 30545
    },
    {
      "epoch": 14.403583215464403,
      "grad_norm": 0.751324474811554,
      "learning_rate": 6.576765086672634e-06,
      "loss": 0.1515,
      "num_input_tokens_seen": 26459616,
      "step": 30550
    },
    {
      "epoch": 14.405940594059405,
      "grad_norm": 0.8313192129135132,
      "learning_rate": 6.57013019016712e-06,
      "loss": 0.1159,
      "num_input_tokens_seen": 26464336,
      "step": 30555
    },
    {
      "epoch": 14.408297972654408,
      "grad_norm": 1.265446662902832,
      "learning_rate": 6.563498135779142e-06,
      "loss": 0.1635,
      "num_input_tokens_seen": 26468160,
      "step": 30560
    },
    {
      "epoch": 14.41065535124941,
      "grad_norm": 1.41526460647583,
      "learning_rate": 6.556868924531431e-06,
      "loss": 0.2389,
      "num_input_tokens_seen": 26471936,
      "step": 30565
    },
    {
      "epoch": 14.413012729844413,
      "grad_norm": 0.8921352624893188,
      "learning_rate": 6.550242557446304e-06,
      "loss": 0.1613,
      "num_input_tokens_seen": 26476480,
      "step": 30570
    },
    {
      "epoch": 14.415370108439415,
      "grad_norm": 0.4726189374923706,
      "learning_rate": 6.543619035545634e-06,
      "loss": 0.1123,
      "num_input_tokens_seen": 26481536,
      "step": 30575
    },
    {
      "epoch": 14.417727487034417,
      "grad_norm": 1.984998106956482,
      "learning_rate": 6.53699835985084e-06,
      "loss": 0.1947,
      "num_input_tokens_seen": 26485952,
      "step": 30580
    },
    {
      "epoch": 14.42008486562942,
      "grad_norm": 0.7393662333488464,
      "learning_rate": 6.530380531382927e-06,
      "loss": 0.2456,
      "num_input_tokens_seen": 26490160,
      "step": 30585
    },
    {
      "epoch": 14.422442244224422,
      "grad_norm": 0.786800742149353,
      "learning_rate": 6.523765551162433e-06,
      "loss": 0.1413,
      "num_input_tokens_seen": 26494624,
      "step": 30590
    },
    {
      "epoch": 14.424799622819425,
      "grad_norm": 1.9932409524917603,
      "learning_rate": 6.517153420209476e-06,
      "loss": 0.1937,
      "num_input_tokens_seen": 26498400,
      "step": 30595
    },
    {
      "epoch": 14.427157001414427,
      "grad_norm": 0.9143979549407959,
      "learning_rate": 6.510544139543739e-06,
      "loss": 0.1928,
      "num_input_tokens_seen": 26502800,
      "step": 30600
    },
    {
      "epoch": 14.427157001414427,
      "eval_loss": 0.199966698884964,
      "eval_runtime": 22.1128,
      "eval_samples_per_second": 42.645,
      "eval_steps_per_second": 21.345,
      "num_input_tokens_seen": 26502800,
      "step": 30600
    },
    {
      "epoch": 14.42951438000943,
      "grad_norm": 0.5090118646621704,
      "learning_rate": 6.503937710184452e-06,
      "loss": 0.0993,
      "num_input_tokens_seen": 26507264,
      "step": 30605
    },
    {
      "epoch": 14.431871758604432,
      "grad_norm": 1.0224484205245972,
      "learning_rate": 6.4973341331503954e-06,
      "loss": 0.2379,
      "num_input_tokens_seen": 26511008,
      "step": 30610
    },
    {
      "epoch": 14.434229137199434,
      "grad_norm": 1.263173222541809,
      "learning_rate": 6.490733409459942e-06,
      "loss": 0.1624,
      "num_input_tokens_seen": 26515712,
      "step": 30615
    },
    {
      "epoch": 14.436586515794437,
      "grad_norm": 1.148697853088379,
      "learning_rate": 6.484135540130995e-06,
      "loss": 0.2036,
      "num_input_tokens_seen": 26519616,
      "step": 30620
    },
    {
      "epoch": 14.438943894389439,
      "grad_norm": 0.8857609033584595,
      "learning_rate": 6.4775405261810364e-06,
      "loss": 0.1878,
      "num_input_tokens_seen": 26523728,
      "step": 30625
    },
    {
      "epoch": 14.441301272984441,
      "grad_norm": 1.3281141519546509,
      "learning_rate": 6.470948368627092e-06,
      "loss": 0.1967,
      "num_input_tokens_seen": 26527664,
      "step": 30630
    },
    {
      "epoch": 14.443658651579444,
      "grad_norm": 0.9108049869537354,
      "learning_rate": 6.464359068485756e-06,
      "loss": 0.1525,
      "num_input_tokens_seen": 26532528,
      "step": 30635
    },
    {
      "epoch": 14.446016030174446,
      "grad_norm": 0.4056035876274109,
      "learning_rate": 6.457772626773195e-06,
      "loss": 0.1056,
      "num_input_tokens_seen": 26536304,
      "step": 30640
    },
    {
      "epoch": 14.448373408769449,
      "grad_norm": 0.551838755607605,
      "learning_rate": 6.451189044505104e-06,
      "loss": 0.1554,
      "num_input_tokens_seen": 26540192,
      "step": 30645
    },
    {
      "epoch": 14.450730787364451,
      "grad_norm": 0.6434744000434875,
      "learning_rate": 6.44460832269676e-06,
      "loss": 0.1244,
      "num_input_tokens_seen": 26544720,
      "step": 30650
    },
    {
      "epoch": 14.453088165959453,
      "grad_norm": 0.8795821070671082,
      "learning_rate": 6.438030462363001e-06,
      "loss": 0.1912,
      "num_input_tokens_seen": 26549248,
      "step": 30655
    },
    {
      "epoch": 14.455445544554456,
      "grad_norm": 0.4625733196735382,
      "learning_rate": 6.431455464518205e-06,
      "loss": 0.1557,
      "num_input_tokens_seen": 26553152,
      "step": 30660
    },
    {
      "epoch": 14.457802923149458,
      "grad_norm": 0.5957898497581482,
      "learning_rate": 6.424883330176326e-06,
      "loss": 0.1479,
      "num_input_tokens_seen": 26557520,
      "step": 30665
    },
    {
      "epoch": 14.46016030174446,
      "grad_norm": 0.31989648938179016,
      "learning_rate": 6.418314060350864e-06,
      "loss": 0.1744,
      "num_input_tokens_seen": 26561616,
      "step": 30670
    },
    {
      "epoch": 14.462517680339463,
      "grad_norm": 0.794940710067749,
      "learning_rate": 6.4117476560548895e-06,
      "loss": 0.1973,
      "num_input_tokens_seen": 26566496,
      "step": 30675
    },
    {
      "epoch": 14.464875058934465,
      "grad_norm": 0.7385790944099426,
      "learning_rate": 6.405184118301016e-06,
      "loss": 0.2072,
      "num_input_tokens_seen": 26570784,
      "step": 30680
    },
    {
      "epoch": 14.467232437529468,
      "grad_norm": 1.6794862747192383,
      "learning_rate": 6.398623448101434e-06,
      "loss": 0.1981,
      "num_input_tokens_seen": 26574800,
      "step": 30685
    },
    {
      "epoch": 14.46958981612447,
      "grad_norm": 1.3758013248443604,
      "learning_rate": 6.392065646467871e-06,
      "loss": 0.2487,
      "num_input_tokens_seen": 26579152,
      "step": 30690
    },
    {
      "epoch": 14.471947194719473,
      "grad_norm": 1.4351584911346436,
      "learning_rate": 6.385510714411632e-06,
      "loss": 0.1861,
      "num_input_tokens_seen": 26583664,
      "step": 30695
    },
    {
      "epoch": 14.474304573314475,
      "grad_norm": 0.6624383926391602,
      "learning_rate": 6.378958652943559e-06,
      "loss": 0.1634,
      "num_input_tokens_seen": 26587744,
      "step": 30700
    },
    {
      "epoch": 14.476661951909477,
      "grad_norm": 1.215569019317627,
      "learning_rate": 6.3724094630740776e-06,
      "loss": 0.1915,
      "num_input_tokens_seen": 26592720,
      "step": 30705
    },
    {
      "epoch": 14.47901933050448,
      "grad_norm": 0.7815809845924377,
      "learning_rate": 6.365863145813136e-06,
      "loss": 0.2486,
      "num_input_tokens_seen": 26596736,
      "step": 30710
    },
    {
      "epoch": 14.481376709099482,
      "grad_norm": 0.641643226146698,
      "learning_rate": 6.359319702170269e-06,
      "loss": 0.157,
      "num_input_tokens_seen": 26600592,
      "step": 30715
    },
    {
      "epoch": 14.483734087694483,
      "grad_norm": 2.017515182495117,
      "learning_rate": 6.352779133154566e-06,
      "loss": 0.1814,
      "num_input_tokens_seen": 26604896,
      "step": 30720
    },
    {
      "epoch": 14.486091466289485,
      "grad_norm": 0.5823014378547668,
      "learning_rate": 6.346241439774648e-06,
      "loss": 0.2077,
      "num_input_tokens_seen": 26608512,
      "step": 30725
    },
    {
      "epoch": 14.488448844884488,
      "grad_norm": 0.8009173274040222,
      "learning_rate": 6.339706623038716e-06,
      "loss": 0.1617,
      "num_input_tokens_seen": 26612624,
      "step": 30730
    },
    {
      "epoch": 14.49080622347949,
      "grad_norm": 2.038393020629883,
      "learning_rate": 6.333174683954532e-06,
      "loss": 0.1648,
      "num_input_tokens_seen": 26617072,
      "step": 30735
    },
    {
      "epoch": 14.493163602074493,
      "grad_norm": 1.5934677124023438,
      "learning_rate": 6.326645623529387e-06,
      "loss": 0.1565,
      "num_input_tokens_seen": 26621312,
      "step": 30740
    },
    {
      "epoch": 14.495520980669495,
      "grad_norm": 2.174532175064087,
      "learning_rate": 6.320119442770156e-06,
      "loss": 0.148,
      "num_input_tokens_seen": 26625456,
      "step": 30745
    },
    {
      "epoch": 14.497878359264497,
      "grad_norm": 2.035590410232544,
      "learning_rate": 6.313596142683254e-06,
      "loss": 0.2286,
      "num_input_tokens_seen": 26629456,
      "step": 30750
    },
    {
      "epoch": 14.5002357378595,
      "grad_norm": 0.9718795418739319,
      "learning_rate": 6.307075724274647e-06,
      "loss": 0.1328,
      "num_input_tokens_seen": 26634224,
      "step": 30755
    },
    {
      "epoch": 14.502593116454502,
      "grad_norm": 0.9062981009483337,
      "learning_rate": 6.300558188549882e-06,
      "loss": 0.2221,
      "num_input_tokens_seen": 26638624,
      "step": 30760
    },
    {
      "epoch": 14.504950495049505,
      "grad_norm": 1.4814249277114868,
      "learning_rate": 6.29404353651403e-06,
      "loss": 0.2046,
      "num_input_tokens_seen": 26642368,
      "step": 30765
    },
    {
      "epoch": 14.507307873644507,
      "grad_norm": 1.1805036067962646,
      "learning_rate": 6.287531769171737e-06,
      "loss": 0.1968,
      "num_input_tokens_seen": 26646336,
      "step": 30770
    },
    {
      "epoch": 14.50966525223951,
      "grad_norm": 0.5895925760269165,
      "learning_rate": 6.2810228875272045e-06,
      "loss": 0.1255,
      "num_input_tokens_seen": 26650016,
      "step": 30775
    },
    {
      "epoch": 14.512022630834512,
      "grad_norm": 2.115931510925293,
      "learning_rate": 6.274516892584179e-06,
      "loss": 0.1848,
      "num_input_tokens_seen": 26653664,
      "step": 30780
    },
    {
      "epoch": 14.514380009429514,
      "grad_norm": 0.8750956058502197,
      "learning_rate": 6.268013785345969e-06,
      "loss": 0.1319,
      "num_input_tokens_seen": 26658112,
      "step": 30785
    },
    {
      "epoch": 14.516737388024517,
      "grad_norm": 2.163763999938965,
      "learning_rate": 6.26151356681543e-06,
      "loss": 0.19,
      "num_input_tokens_seen": 26663104,
      "step": 30790
    },
    {
      "epoch": 14.519094766619519,
      "grad_norm": 0.7121869325637817,
      "learning_rate": 6.255016237994981e-06,
      "loss": 0.1942,
      "num_input_tokens_seen": 26667280,
      "step": 30795
    },
    {
      "epoch": 14.521452145214521,
      "grad_norm": 0.5537320375442505,
      "learning_rate": 6.248521799886603e-06,
      "loss": 0.1728,
      "num_input_tokens_seen": 26671584,
      "step": 30800
    },
    {
      "epoch": 14.521452145214521,
      "eval_loss": 0.20086118578910828,
      "eval_runtime": 22.1978,
      "eval_samples_per_second": 42.482,
      "eval_steps_per_second": 21.263,
      "num_input_tokens_seen": 26671584,
      "step": 30800
    },
    {
      "epoch": 14.523809523809524,
      "grad_norm": 0.3166695833206177,
      "learning_rate": 6.242030253491798e-06,
      "loss": 0.124,
      "num_input_tokens_seen": 26675920,
      "step": 30805
    },
    {
      "epoch": 14.526166902404526,
      "grad_norm": 1.2636396884918213,
      "learning_rate": 6.235541599811656e-06,
      "loss": 0.127,
      "num_input_tokens_seen": 26680016,
      "step": 30810
    },
    {
      "epoch": 14.528524280999529,
      "grad_norm": 0.3276123106479645,
      "learning_rate": 6.229055839846814e-06,
      "loss": 0.2089,
      "num_input_tokens_seen": 26684416,
      "step": 30815
    },
    {
      "epoch": 14.530881659594531,
      "grad_norm": 1.0461136102676392,
      "learning_rate": 6.222572974597455e-06,
      "loss": 0.173,
      "num_input_tokens_seen": 26688736,
      "step": 30820
    },
    {
      "epoch": 14.533239038189533,
      "grad_norm": 0.2644376754760742,
      "learning_rate": 6.216093005063306e-06,
      "loss": 0.1743,
      "num_input_tokens_seen": 26692848,
      "step": 30825
    },
    {
      "epoch": 14.535596416784536,
      "grad_norm": 0.6361153721809387,
      "learning_rate": 6.209615932243678e-06,
      "loss": 0.1758,
      "num_input_tokens_seen": 26697888,
      "step": 30830
    },
    {
      "epoch": 14.537953795379538,
      "grad_norm": 0.8497425317764282,
      "learning_rate": 6.203141757137399e-06,
      "loss": 0.155,
      "num_input_tokens_seen": 26701648,
      "step": 30835
    },
    {
      "epoch": 14.54031117397454,
      "grad_norm": 0.9327563643455505,
      "learning_rate": 6.196670480742886e-06,
      "loss": 0.185,
      "num_input_tokens_seen": 26706240,
      "step": 30840
    },
    {
      "epoch": 14.542668552569543,
      "grad_norm": 1.3633148670196533,
      "learning_rate": 6.190202104058074e-06,
      "loss": 0.2459,
      "num_input_tokens_seen": 26710656,
      "step": 30845
    },
    {
      "epoch": 14.545025931164545,
      "grad_norm": 1.1815180778503418,
      "learning_rate": 6.183736628080475e-06,
      "loss": 0.1548,
      "num_input_tokens_seen": 26715264,
      "step": 30850
    },
    {
      "epoch": 14.547383309759548,
      "grad_norm": 0.9094225764274597,
      "learning_rate": 6.177274053807155e-06,
      "loss": 0.1633,
      "num_input_tokens_seen": 26720176,
      "step": 30855
    },
    {
      "epoch": 14.54974068835455,
      "grad_norm": 0.6358590722084045,
      "learning_rate": 6.170814382234713e-06,
      "loss": 0.1419,
      "num_input_tokens_seen": 26724448,
      "step": 30860
    },
    {
      "epoch": 14.552098066949553,
      "grad_norm": 0.9552934765815735,
      "learning_rate": 6.16435761435932e-06,
      "loss": 0.1397,
      "num_input_tokens_seen": 26728944,
      "step": 30865
    },
    {
      "epoch": 14.554455445544555,
      "grad_norm": 0.6239291429519653,
      "learning_rate": 6.157903751176681e-06,
      "loss": 0.1027,
      "num_input_tokens_seen": 26732432,
      "step": 30870
    },
    {
      "epoch": 14.556812824139557,
      "grad_norm": 0.8255319595336914,
      "learning_rate": 6.151452793682066e-06,
      "loss": 0.1606,
      "num_input_tokens_seen": 26736368,
      "step": 30875
    },
    {
      "epoch": 14.55917020273456,
      "grad_norm": 0.5642181038856506,
      "learning_rate": 6.145004742870305e-06,
      "loss": 0.2191,
      "num_input_tokens_seen": 26740752,
      "step": 30880
    },
    {
      "epoch": 14.561527581329562,
      "grad_norm": 1.164698600769043,
      "learning_rate": 6.138559599735752e-06,
      "loss": 0.181,
      "num_input_tokens_seen": 26744560,
      "step": 30885
    },
    {
      "epoch": 14.563884959924565,
      "grad_norm": 0.559991717338562,
      "learning_rate": 6.132117365272344e-06,
      "loss": 0.1861,
      "num_input_tokens_seen": 26748480,
      "step": 30890
    },
    {
      "epoch": 14.566242338519567,
      "grad_norm": 1.3656384944915771,
      "learning_rate": 6.125678040473545e-06,
      "loss": 0.1671,
      "num_input_tokens_seen": 26752448,
      "step": 30895
    },
    {
      "epoch": 14.56859971711457,
      "grad_norm": 1.6823407411575317,
      "learning_rate": 6.1192416263323755e-06,
      "loss": 0.189,
      "num_input_tokens_seen": 26756816,
      "step": 30900
    },
    {
      "epoch": 14.570957095709572,
      "grad_norm": 0.4311002194881439,
      "learning_rate": 6.112808123841424e-06,
      "loss": 0.1338,
      "num_input_tokens_seen": 26760992,
      "step": 30905
    },
    {
      "epoch": 14.573314474304574,
      "grad_norm": 0.9990582466125488,
      "learning_rate": 6.106377533992805e-06,
      "loss": 0.1388,
      "num_input_tokens_seen": 26765184,
      "step": 30910
    },
    {
      "epoch": 14.575671852899575,
      "grad_norm": 0.5706522464752197,
      "learning_rate": 6.099949857778204e-06,
      "loss": 0.1538,
      "num_input_tokens_seen": 26769472,
      "step": 30915
    },
    {
      "epoch": 14.578029231494579,
      "grad_norm": 0.7714611887931824,
      "learning_rate": 6.093525096188852e-06,
      "loss": 0.1552,
      "num_input_tokens_seen": 26774112,
      "step": 30920
    },
    {
      "epoch": 14.58038661008958,
      "grad_norm": 0.9116371273994446,
      "learning_rate": 6.087103250215518e-06,
      "loss": 0.1438,
      "num_input_tokens_seen": 26778480,
      "step": 30925
    },
    {
      "epoch": 14.582743988684582,
      "grad_norm": 2.107661008834839,
      "learning_rate": 6.080684320848537e-06,
      "loss": 0.2321,
      "num_input_tokens_seen": 26782720,
      "step": 30930
    },
    {
      "epoch": 14.585101367279584,
      "grad_norm": 0.5783594846725464,
      "learning_rate": 6.074268309077794e-06,
      "loss": 0.2358,
      "num_input_tokens_seen": 26786624,
      "step": 30935
    },
    {
      "epoch": 14.587458745874587,
      "grad_norm": 1.122457504272461,
      "learning_rate": 6.067855215892709e-06,
      "loss": 0.2245,
      "num_input_tokens_seen": 26790512,
      "step": 30940
    },
    {
      "epoch": 14.58981612446959,
      "grad_norm": 0.6131851077079773,
      "learning_rate": 6.061445042282271e-06,
      "loss": 0.1207,
      "num_input_tokens_seen": 26795648,
      "step": 30945
    },
    {
      "epoch": 14.592173503064592,
      "grad_norm": 1.7153302431106567,
      "learning_rate": 6.055037789234999e-06,
      "loss": 0.2094,
      "num_input_tokens_seen": 26800896,
      "step": 30950
    },
    {
      "epoch": 14.594530881659594,
      "grad_norm": 0.9682632088661194,
      "learning_rate": 6.048633457738975e-06,
      "loss": 0.1453,
      "num_input_tokens_seen": 26805776,
      "step": 30955
    },
    {
      "epoch": 14.596888260254596,
      "grad_norm": 1.2999259233474731,
      "learning_rate": 6.042232048781837e-06,
      "loss": 0.181,
      "num_input_tokens_seen": 26810432,
      "step": 30960
    },
    {
      "epoch": 14.599245638849599,
      "grad_norm": 1.00568425655365,
      "learning_rate": 6.035833563350757e-06,
      "loss": 0.1391,
      "num_input_tokens_seen": 26814848,
      "step": 30965
    },
    {
      "epoch": 14.601603017444601,
      "grad_norm": 1.7009243965148926,
      "learning_rate": 6.0294380024324525e-06,
      "loss": 0.2404,
      "num_input_tokens_seen": 26820000,
      "step": 30970
    },
    {
      "epoch": 14.603960396039604,
      "grad_norm": 1.1109930276870728,
      "learning_rate": 6.023045367013213e-06,
      "loss": 0.2521,
      "num_input_tokens_seen": 26824288,
      "step": 30975
    },
    {
      "epoch": 14.606317774634606,
      "grad_norm": 1.6193872690200806,
      "learning_rate": 6.016655658078851e-06,
      "loss": 0.2119,
      "num_input_tokens_seen": 26829280,
      "step": 30980
    },
    {
      "epoch": 14.608675153229608,
      "grad_norm": 0.6702356934547424,
      "learning_rate": 6.010268876614753e-06,
      "loss": 0.1507,
      "num_input_tokens_seen": 26833456,
      "step": 30985
    },
    {
      "epoch": 14.61103253182461,
      "grad_norm": 0.5430036187171936,
      "learning_rate": 6.0038850236058266e-06,
      "loss": 0.1559,
      "num_input_tokens_seen": 26837472,
      "step": 30990
    },
    {
      "epoch": 14.613389910419613,
      "grad_norm": 1.562927484512329,
      "learning_rate": 5.997504100036549e-06,
      "loss": 0.1426,
      "num_input_tokens_seen": 26841904,
      "step": 30995
    },
    {
      "epoch": 14.615747289014616,
      "grad_norm": 1.0400490760803223,
      "learning_rate": 5.991126106890949e-06,
      "loss": 0.2114,
      "num_input_tokens_seen": 26845568,
      "step": 31000
    },
    {
      "epoch": 14.615747289014616,
      "eval_loss": 0.19982114434242249,
      "eval_runtime": 22.1929,
      "eval_samples_per_second": 42.491,
      "eval_steps_per_second": 21.268,
      "num_input_tokens_seen": 26845568,
      "step": 31000
    },
    {
      "epoch": 14.618104667609618,
      "grad_norm": 1.0151954889297485,
      "learning_rate": 5.984751045152576e-06,
      "loss": 0.1232,
      "num_input_tokens_seen": 26849472,
      "step": 31005
    },
    {
      "epoch": 14.62046204620462,
      "grad_norm": 0.5335571765899658,
      "learning_rate": 5.978378915804553e-06,
      "loss": 0.1709,
      "num_input_tokens_seen": 26854160,
      "step": 31010
    },
    {
      "epoch": 14.622819424799623,
      "grad_norm": 0.7182631492614746,
      "learning_rate": 5.972009719829547e-06,
      "loss": 0.1482,
      "num_input_tokens_seen": 26859184,
      "step": 31015
    },
    {
      "epoch": 14.625176803394625,
      "grad_norm": 1.024728775024414,
      "learning_rate": 5.965643458209755e-06,
      "loss": 0.1694,
      "num_input_tokens_seen": 26862912,
      "step": 31020
    },
    {
      "epoch": 14.627534181989628,
      "grad_norm": 1.64027738571167,
      "learning_rate": 5.95928013192695e-06,
      "loss": 0.2194,
      "num_input_tokens_seen": 26868416,
      "step": 31025
    },
    {
      "epoch": 14.62989156058463,
      "grad_norm": 1.2400785684585571,
      "learning_rate": 5.952919741962423e-06,
      "loss": 0.1437,
      "num_input_tokens_seen": 26871872,
      "step": 31030
    },
    {
      "epoch": 14.632248939179632,
      "grad_norm": 0.6479269862174988,
      "learning_rate": 5.946562289297042e-06,
      "loss": 0.233,
      "num_input_tokens_seen": 26875488,
      "step": 31035
    },
    {
      "epoch": 14.634606317774635,
      "grad_norm": 1.2425928115844727,
      "learning_rate": 5.9402077749111855e-06,
      "loss": 0.1579,
      "num_input_tokens_seen": 26879584,
      "step": 31040
    },
    {
      "epoch": 14.636963696369637,
      "grad_norm": 0.5810297727584839,
      "learning_rate": 5.933856199784821e-06,
      "loss": 0.1632,
      "num_input_tokens_seen": 26883344,
      "step": 31045
    },
    {
      "epoch": 14.63932107496464,
      "grad_norm": 0.42007720470428467,
      "learning_rate": 5.927507564897419e-06,
      "loss": 0.1887,
      "num_input_tokens_seen": 26887840,
      "step": 31050
    },
    {
      "epoch": 14.641678453559642,
      "grad_norm": 1.3947467803955078,
      "learning_rate": 5.9211618712280395e-06,
      "loss": 0.1578,
      "num_input_tokens_seen": 26891648,
      "step": 31055
    },
    {
      "epoch": 14.644035832154644,
      "grad_norm": 1.412793517112732,
      "learning_rate": 5.914819119755255e-06,
      "loss": 0.2706,
      "num_input_tokens_seen": 26896416,
      "step": 31060
    },
    {
      "epoch": 14.646393210749647,
      "grad_norm": 0.8327175378799438,
      "learning_rate": 5.908479311457205e-06,
      "loss": 0.2082,
      "num_input_tokens_seen": 26901696,
      "step": 31065
    },
    {
      "epoch": 14.64875058934465,
      "grad_norm": 0.4940634071826935,
      "learning_rate": 5.902142447311559e-06,
      "loss": 0.1563,
      "num_input_tokens_seen": 26905696,
      "step": 31070
    },
    {
      "epoch": 14.651107967939652,
      "grad_norm": 1.174597144126892,
      "learning_rate": 5.895808528295546e-06,
      "loss": 0.1772,
      "num_input_tokens_seen": 26909600,
      "step": 31075
    },
    {
      "epoch": 14.653465346534654,
      "grad_norm": 2.381572723388672,
      "learning_rate": 5.889477555385941e-06,
      "loss": 0.2009,
      "num_input_tokens_seen": 26914304,
      "step": 31080
    },
    {
      "epoch": 14.655822725129656,
      "grad_norm": 0.9188838601112366,
      "learning_rate": 5.883149529559051e-06,
      "loss": 0.152,
      "num_input_tokens_seen": 26918400,
      "step": 31085
    },
    {
      "epoch": 14.658180103724659,
      "grad_norm": 0.8575523495674133,
      "learning_rate": 5.876824451790738e-06,
      "loss": 0.1215,
      "num_input_tokens_seen": 26922720,
      "step": 31090
    },
    {
      "epoch": 14.660537482319661,
      "grad_norm": 0.5407374501228333,
      "learning_rate": 5.87050232305642e-06,
      "loss": 0.1486,
      "num_input_tokens_seen": 26927504,
      "step": 31095
    },
    {
      "epoch": 14.662894860914664,
      "grad_norm": 0.815608024597168,
      "learning_rate": 5.864183144331034e-06,
      "loss": 0.1644,
      "num_input_tokens_seen": 26932000,
      "step": 31100
    },
    {
      "epoch": 14.665252239509666,
      "grad_norm": 3.4015650749206543,
      "learning_rate": 5.857866916589089e-06,
      "loss": 0.2292,
      "num_input_tokens_seen": 26936704,
      "step": 31105
    },
    {
      "epoch": 14.667609618104667,
      "grad_norm": 0.9307727813720703,
      "learning_rate": 5.8515536408046216e-06,
      "loss": 0.199,
      "num_input_tokens_seen": 26941296,
      "step": 31110
    },
    {
      "epoch": 14.66996699669967,
      "grad_norm": 0.3161645233631134,
      "learning_rate": 5.845243317951208e-06,
      "loss": 0.1319,
      "num_input_tokens_seen": 26946336,
      "step": 31115
    },
    {
      "epoch": 14.672324375294671,
      "grad_norm": 1.0342774391174316,
      "learning_rate": 5.838935949001997e-06,
      "loss": 0.1057,
      "num_input_tokens_seen": 26950544,
      "step": 31120
    },
    {
      "epoch": 14.674681753889674,
      "grad_norm": 0.8384506106376648,
      "learning_rate": 5.8326315349296476e-06,
      "loss": 0.1553,
      "num_input_tokens_seen": 26955088,
      "step": 31125
    },
    {
      "epoch": 14.677039132484676,
      "grad_norm": 0.9690625071525574,
      "learning_rate": 5.826330076706396e-06,
      "loss": 0.1953,
      "num_input_tokens_seen": 26959920,
      "step": 31130
    },
    {
      "epoch": 14.679396511079679,
      "grad_norm": 2.0974676609039307,
      "learning_rate": 5.820031575303988e-06,
      "loss": 0.1753,
      "num_input_tokens_seen": 26964160,
      "step": 31135
    },
    {
      "epoch": 14.681753889674681,
      "grad_norm": 0.787668764591217,
      "learning_rate": 5.813736031693745e-06,
      "loss": 0.1783,
      "num_input_tokens_seen": 26968160,
      "step": 31140
    },
    {
      "epoch": 14.684111268269683,
      "grad_norm": 1.0989868640899658,
      "learning_rate": 5.807443446846522e-06,
      "loss": 0.2565,
      "num_input_tokens_seen": 26972560,
      "step": 31145
    },
    {
      "epoch": 14.686468646864686,
      "grad_norm": 0.9555286765098572,
      "learning_rate": 5.801153821732699e-06,
      "loss": 0.1486,
      "num_input_tokens_seen": 26976464,
      "step": 31150
    },
    {
      "epoch": 14.688826025459688,
      "grad_norm": 0.9896498322486877,
      "learning_rate": 5.794867157322229e-06,
      "loss": 0.2106,
      "num_input_tokens_seen": 26981040,
      "step": 31155
    },
    {
      "epoch": 14.69118340405469,
      "grad_norm": 1.7708349227905273,
      "learning_rate": 5.788583454584593e-06,
      "loss": 0.1692,
      "num_input_tokens_seen": 26985152,
      "step": 31160
    },
    {
      "epoch": 14.693540782649693,
      "grad_norm": 0.9629276990890503,
      "learning_rate": 5.7823027144888075e-06,
      "loss": 0.1936,
      "num_input_tokens_seen": 26989168,
      "step": 31165
    },
    {
      "epoch": 14.695898161244696,
      "grad_norm": 1.9240213632583618,
      "learning_rate": 5.776024938003455e-06,
      "loss": 0.1746,
      "num_input_tokens_seen": 26994112,
      "step": 31170
    },
    {
      "epoch": 14.698255539839698,
      "grad_norm": 0.6358458995819092,
      "learning_rate": 5.7697501260966345e-06,
      "loss": 0.1617,
      "num_input_tokens_seen": 26998272,
      "step": 31175
    },
    {
      "epoch": 14.7006129184347,
      "grad_norm": 1.8318554162979126,
      "learning_rate": 5.7634782797360145e-06,
      "loss": 0.1547,
      "num_input_tokens_seen": 27002240,
      "step": 31180
    },
    {
      "epoch": 14.702970297029703,
      "grad_norm": 0.9257902503013611,
      "learning_rate": 5.757209399888777e-06,
      "loss": 0.1906,
      "num_input_tokens_seen": 27006240,
      "step": 31185
    },
    {
      "epoch": 14.705327675624705,
      "grad_norm": 1.8944975137710571,
      "learning_rate": 5.750943487521679e-06,
      "loss": 0.1622,
      "num_input_tokens_seen": 27010288,
      "step": 31190
    },
    {
      "epoch": 14.707685054219708,
      "grad_norm": 0.34261059761047363,
      "learning_rate": 5.744680543600986e-06,
      "loss": 0.1564,
      "num_input_tokens_seen": 27014144,
      "step": 31195
    },
    {
      "epoch": 14.71004243281471,
      "grad_norm": 0.45116370916366577,
      "learning_rate": 5.738420569092537e-06,
      "loss": 0.265,
      "num_input_tokens_seen": 27017952,
      "step": 31200
    },
    {
      "epoch": 14.71004243281471,
      "eval_loss": 0.20018665492534637,
      "eval_runtime": 22.18,
      "eval_samples_per_second": 42.516,
      "eval_steps_per_second": 21.28,
      "num_input_tokens_seen": 27017952,
      "step": 31200
    },
    {
      "epoch": 14.712399811409712,
      "grad_norm": 0.4339020848274231,
      "learning_rate": 5.732163564961684e-06,
      "loss": 0.2022,
      "num_input_tokens_seen": 27021600,
      "step": 31205
    },
    {
      "epoch": 14.714757190004715,
      "grad_norm": 1.4643691778182983,
      "learning_rate": 5.725909532173354e-06,
      "loss": 0.1052,
      "num_input_tokens_seen": 27026400,
      "step": 31210
    },
    {
      "epoch": 14.717114568599717,
      "grad_norm": 1.7387672662734985,
      "learning_rate": 5.719658471691977e-06,
      "loss": 0.1935,
      "num_input_tokens_seen": 27030352,
      "step": 31215
    },
    {
      "epoch": 14.71947194719472,
      "grad_norm": 1.1935838460922241,
      "learning_rate": 5.71341038448156e-06,
      "loss": 0.1504,
      "num_input_tokens_seen": 27034528,
      "step": 31220
    },
    {
      "epoch": 14.721829325789722,
      "grad_norm": 0.6285324692726135,
      "learning_rate": 5.707165271505635e-06,
      "loss": 0.169,
      "num_input_tokens_seen": 27038640,
      "step": 31225
    },
    {
      "epoch": 14.724186704384724,
      "grad_norm": 1.8163694143295288,
      "learning_rate": 5.700923133727271e-06,
      "loss": 0.1181,
      "num_input_tokens_seen": 27042960,
      "step": 31230
    },
    {
      "epoch": 14.726544082979727,
      "grad_norm": 1.8436402082443237,
      "learning_rate": 5.694683972109083e-06,
      "loss": 0.1991,
      "num_input_tokens_seen": 27048592,
      "step": 31235
    },
    {
      "epoch": 14.72890146157473,
      "grad_norm": 1.5893877744674683,
      "learning_rate": 5.688447787613241e-06,
      "loss": 0.218,
      "num_input_tokens_seen": 27053072,
      "step": 31240
    },
    {
      "epoch": 14.731258840169732,
      "grad_norm": 1.6774982213974,
      "learning_rate": 5.6822145812014285e-06,
      "loss": 0.1541,
      "num_input_tokens_seen": 27056784,
      "step": 31245
    },
    {
      "epoch": 14.733616218764734,
      "grad_norm": 1.9204015731811523,
      "learning_rate": 5.675984353834896e-06,
      "loss": 0.1838,
      "num_input_tokens_seen": 27060288,
      "step": 31250
    },
    {
      "epoch": 14.735973597359736,
      "grad_norm": 0.6600548028945923,
      "learning_rate": 5.66975710647441e-06,
      "loss": 0.1234,
      "num_input_tokens_seen": 27065344,
      "step": 31255
    },
    {
      "epoch": 14.738330975954739,
      "grad_norm": 2.373466968536377,
      "learning_rate": 5.663532840080304e-06,
      "loss": 0.2059,
      "num_input_tokens_seen": 27069184,
      "step": 31260
    },
    {
      "epoch": 14.740688354549741,
      "grad_norm": 0.6714272499084473,
      "learning_rate": 5.6573115556124325e-06,
      "loss": 0.1384,
      "num_input_tokens_seen": 27074496,
      "step": 31265
    },
    {
      "epoch": 14.743045733144744,
      "grad_norm": 0.7611920833587646,
      "learning_rate": 5.651093254030185e-06,
      "loss": 0.1926,
      "num_input_tokens_seen": 27078304,
      "step": 31270
    },
    {
      "epoch": 14.745403111739746,
      "grad_norm": 0.9248074293136597,
      "learning_rate": 5.644877936292514e-06,
      "loss": 0.1865,
      "num_input_tokens_seen": 27082848,
      "step": 31275
    },
    {
      "epoch": 14.747760490334748,
      "grad_norm": 1.2385300397872925,
      "learning_rate": 5.638665603357901e-06,
      "loss": 0.2114,
      "num_input_tokens_seen": 27087184,
      "step": 31280
    },
    {
      "epoch": 14.75011786892975,
      "grad_norm": 2.1967015266418457,
      "learning_rate": 5.632456256184357e-06,
      "loss": 0.1772,
      "num_input_tokens_seen": 27091072,
      "step": 31285
    },
    {
      "epoch": 14.752475247524753,
      "grad_norm": 0.6693161129951477,
      "learning_rate": 5.626249895729452e-06,
      "loss": 0.1448,
      "num_input_tokens_seen": 27095680,
      "step": 31290
    },
    {
      "epoch": 14.754832626119756,
      "grad_norm": 0.9427210092544556,
      "learning_rate": 5.620046522950273e-06,
      "loss": 0.203,
      "num_input_tokens_seen": 27100160,
      "step": 31295
    },
    {
      "epoch": 14.757190004714758,
      "grad_norm": 0.8190235495567322,
      "learning_rate": 5.613846138803464e-06,
      "loss": 0.1744,
      "num_input_tokens_seen": 27104384,
      "step": 31300
    },
    {
      "epoch": 14.75954738330976,
      "grad_norm": 1.0713279247283936,
      "learning_rate": 5.607648744245206e-06,
      "loss": 0.2552,
      "num_input_tokens_seen": 27108208,
      "step": 31305
    },
    {
      "epoch": 14.761904761904763,
      "grad_norm": 1.1377657651901245,
      "learning_rate": 5.601454340231207e-06,
      "loss": 0.1654,
      "num_input_tokens_seen": 27112384,
      "step": 31310
    },
    {
      "epoch": 14.764262140499763,
      "grad_norm": 0.6442176103591919,
      "learning_rate": 5.595262927716724e-06,
      "loss": 0.2172,
      "num_input_tokens_seen": 27116704,
      "step": 31315
    },
    {
      "epoch": 14.766619519094768,
      "grad_norm": 0.9814587831497192,
      "learning_rate": 5.589074507656561e-06,
      "loss": 0.1864,
      "num_input_tokens_seen": 27121776,
      "step": 31320
    },
    {
      "epoch": 14.768976897689768,
      "grad_norm": 0.6668563485145569,
      "learning_rate": 5.582889081005044e-06,
      "loss": 0.1825,
      "num_input_tokens_seen": 27127392,
      "step": 31325
    },
    {
      "epoch": 14.77133427628477,
      "grad_norm": 0.8615580201148987,
      "learning_rate": 5.5767066487160316e-06,
      "loss": 0.1952,
      "num_input_tokens_seen": 27131360,
      "step": 31330
    },
    {
      "epoch": 14.773691654879773,
      "grad_norm": 1.2765545845031738,
      "learning_rate": 5.570527211742949e-06,
      "loss": 0.1211,
      "num_input_tokens_seen": 27135536,
      "step": 31335
    },
    {
      "epoch": 14.776049033474775,
      "grad_norm": 0.8872185349464417,
      "learning_rate": 5.564350771038731e-06,
      "loss": 0.2098,
      "num_input_tokens_seen": 27140528,
      "step": 31340
    },
    {
      "epoch": 14.778406412069778,
      "grad_norm": 1.4375312328338623,
      "learning_rate": 5.558177327555875e-06,
      "loss": 0.2429,
      "num_input_tokens_seen": 27145216,
      "step": 31345
    },
    {
      "epoch": 14.78076379066478,
      "grad_norm": 1.5670757293701172,
      "learning_rate": 5.552006882246388e-06,
      "loss": 0.2735,
      "num_input_tokens_seen": 27149792,
      "step": 31350
    },
    {
      "epoch": 14.783121169259783,
      "grad_norm": 0.7693326473236084,
      "learning_rate": 5.545839436061839e-06,
      "loss": 0.1446,
      "num_input_tokens_seen": 27154832,
      "step": 31355
    },
    {
      "epoch": 14.785478547854785,
      "grad_norm": 1.631891131401062,
      "learning_rate": 5.539674989953331e-06,
      "loss": 0.235,
      "num_input_tokens_seen": 27158352,
      "step": 31360
    },
    {
      "epoch": 14.787835926449787,
      "grad_norm": 0.7116912603378296,
      "learning_rate": 5.533513544871488e-06,
      "loss": 0.2204,
      "num_input_tokens_seen": 27162208,
      "step": 31365
    },
    {
      "epoch": 14.79019330504479,
      "grad_norm": 2.3226335048675537,
      "learning_rate": 5.527355101766493e-06,
      "loss": 0.161,
      "num_input_tokens_seen": 27166272,
      "step": 31370
    },
    {
      "epoch": 14.792550683639792,
      "grad_norm": 1.1094145774841309,
      "learning_rate": 5.521199661588044e-06,
      "loss": 0.1443,
      "num_input_tokens_seen": 27170208,
      "step": 31375
    },
    {
      "epoch": 14.794908062234795,
      "grad_norm": 1.9432411193847656,
      "learning_rate": 5.5150472252853944e-06,
      "loss": 0.151,
      "num_input_tokens_seen": 27174336,
      "step": 31380
    },
    {
      "epoch": 14.797265440829797,
      "grad_norm": 0.6646655201911926,
      "learning_rate": 5.50889779380733e-06,
      "loss": 0.1494,
      "num_input_tokens_seen": 27179008,
      "step": 31385
    },
    {
      "epoch": 14.7996228194248,
      "grad_norm": 0.7381029725074768,
      "learning_rate": 5.5027513681021605e-06,
      "loss": 0.2318,
      "num_input_tokens_seen": 27183536,
      "step": 31390
    },
    {
      "epoch": 14.801980198019802,
      "grad_norm": 2.4307546615600586,
      "learning_rate": 5.4966079491177545e-06,
      "loss": 0.224,
      "num_input_tokens_seen": 27187728,
      "step": 31395
    },
    {
      "epoch": 14.804337576614804,
      "grad_norm": 0.5869829058647156,
      "learning_rate": 5.490467537801491e-06,
      "loss": 0.2021,
      "num_input_tokens_seen": 27191600,
      "step": 31400
    },
    {
      "epoch": 14.804337576614804,
      "eval_loss": 0.19886288046836853,
      "eval_runtime": 22.1912,
      "eval_samples_per_second": 42.494,
      "eval_steps_per_second": 21.27,
      "num_input_tokens_seen": 27191600,
      "step": 31400
    },
    {
      "epoch": 14.806694955209807,
      "grad_norm": 0.6150519251823425,
      "learning_rate": 5.484330135100313e-06,
      "loss": 0.2693,
      "num_input_tokens_seen": 27195712,
      "step": 31405
    },
    {
      "epoch": 14.809052333804809,
      "grad_norm": 1.3764522075653076,
      "learning_rate": 5.4781957419606785e-06,
      "loss": 0.237,
      "num_input_tokens_seen": 27199472,
      "step": 31410
    },
    {
      "epoch": 14.811409712399811,
      "grad_norm": 0.46802350878715515,
      "learning_rate": 5.472064359328577e-06,
      "loss": 0.1658,
      "num_input_tokens_seen": 27204000,
      "step": 31415
    },
    {
      "epoch": 14.813767090994814,
      "grad_norm": 2.349331855773926,
      "learning_rate": 5.4659359881495565e-06,
      "loss": 0.2332,
      "num_input_tokens_seen": 27208208,
      "step": 31420
    },
    {
      "epoch": 14.816124469589816,
      "grad_norm": 0.5176834464073181,
      "learning_rate": 5.4598106293686916e-06,
      "loss": 0.1278,
      "num_input_tokens_seen": 27212256,
      "step": 31425
    },
    {
      "epoch": 14.818481848184819,
      "grad_norm": 0.5546648502349854,
      "learning_rate": 5.45368828393058e-06,
      "loss": 0.1476,
      "num_input_tokens_seen": 27216736,
      "step": 31430
    },
    {
      "epoch": 14.820839226779821,
      "grad_norm": 0.5306808352470398,
      "learning_rate": 5.44756895277937e-06,
      "loss": 0.1354,
      "num_input_tokens_seen": 27221072,
      "step": 31435
    },
    {
      "epoch": 14.823196605374823,
      "grad_norm": 1.7383337020874023,
      "learning_rate": 5.441452636858746e-06,
      "loss": 0.2032,
      "num_input_tokens_seen": 27224688,
      "step": 31440
    },
    {
      "epoch": 14.825553983969826,
      "grad_norm": 0.9426677823066711,
      "learning_rate": 5.435339337111905e-06,
      "loss": 0.2481,
      "num_input_tokens_seen": 27228656,
      "step": 31445
    },
    {
      "epoch": 14.827911362564828,
      "grad_norm": 0.5876560211181641,
      "learning_rate": 5.42922905448161e-06,
      "loss": 0.1866,
      "num_input_tokens_seen": 27232704,
      "step": 31450
    },
    {
      "epoch": 14.83026874115983,
      "grad_norm": 0.3849015235900879,
      "learning_rate": 5.423121789910129e-06,
      "loss": 0.122,
      "num_input_tokens_seen": 27236640,
      "step": 31455
    },
    {
      "epoch": 14.832626119754833,
      "grad_norm": 0.5169526934623718,
      "learning_rate": 5.417017544339287e-06,
      "loss": 0.1869,
      "num_input_tokens_seen": 27240704,
      "step": 31460
    },
    {
      "epoch": 14.834983498349835,
      "grad_norm": 1.3148298263549805,
      "learning_rate": 5.410916318710443e-06,
      "loss": 0.2241,
      "num_input_tokens_seen": 27244944,
      "step": 31465
    },
    {
      "epoch": 14.837340876944838,
      "grad_norm": 0.8316904902458191,
      "learning_rate": 5.404818113964466e-06,
      "loss": 0.1896,
      "num_input_tokens_seen": 27249680,
      "step": 31470
    },
    {
      "epoch": 14.83969825553984,
      "grad_norm": 0.7465587854385376,
      "learning_rate": 5.398722931041792e-06,
      "loss": 0.1626,
      "num_input_tokens_seen": 27254688,
      "step": 31475
    },
    {
      "epoch": 14.842055634134843,
      "grad_norm": 0.6789965629577637,
      "learning_rate": 5.392630770882367e-06,
      "loss": 0.1494,
      "num_input_tokens_seen": 27258448,
      "step": 31480
    },
    {
      "epoch": 14.844413012729845,
      "grad_norm": 2.0071861743927,
      "learning_rate": 5.3865416344256705e-06,
      "loss": 0.1594,
      "num_input_tokens_seen": 27262848,
      "step": 31485
    },
    {
      "epoch": 14.846770391324847,
      "grad_norm": 0.5177015662193298,
      "learning_rate": 5.380455522610742e-06,
      "loss": 0.1228,
      "num_input_tokens_seen": 27267520,
      "step": 31490
    },
    {
      "epoch": 14.84912776991985,
      "grad_norm": 0.8210535645484924,
      "learning_rate": 5.374372436376116e-06,
      "loss": 0.1686,
      "num_input_tokens_seen": 27271456,
      "step": 31495
    },
    {
      "epoch": 14.851485148514852,
      "grad_norm": 0.6928185820579529,
      "learning_rate": 5.368292376659895e-06,
      "loss": 0.1993,
      "num_input_tokens_seen": 27275360,
      "step": 31500
    },
    {
      "epoch": 14.853842527109855,
      "grad_norm": 0.800110936164856,
      "learning_rate": 5.362215344399701e-06,
      "loss": 0.1988,
      "num_input_tokens_seen": 27279312,
      "step": 31505
    },
    {
      "epoch": 14.856199905704855,
      "grad_norm": 1.7080858945846558,
      "learning_rate": 5.356141340532678e-06,
      "loss": 0.1133,
      "num_input_tokens_seen": 27284432,
      "step": 31510
    },
    {
      "epoch": 14.85855728429986,
      "grad_norm": 0.6058204174041748,
      "learning_rate": 5.350070365995522e-06,
      "loss": 0.2398,
      "num_input_tokens_seen": 27288432,
      "step": 31515
    },
    {
      "epoch": 14.86091466289486,
      "grad_norm": 0.5349116325378418,
      "learning_rate": 5.344002421724459e-06,
      "loss": 0.1787,
      "num_input_tokens_seen": 27293024,
      "step": 31520
    },
    {
      "epoch": 14.863272041489862,
      "grad_norm": 0.7981399297714233,
      "learning_rate": 5.337937508655228e-06,
      "loss": 0.1683,
      "num_input_tokens_seen": 27297424,
      "step": 31525
    },
    {
      "epoch": 14.865629420084865,
      "grad_norm": 1.6096453666687012,
      "learning_rate": 5.331875627723126e-06,
      "loss": 0.1858,
      "num_input_tokens_seen": 27301696,
      "step": 31530
    },
    {
      "epoch": 14.867986798679867,
      "grad_norm": 0.7652379274368286,
      "learning_rate": 5.325816779862963e-06,
      "loss": 0.2438,
      "num_input_tokens_seen": 27306608,
      "step": 31535
    },
    {
      "epoch": 14.87034417727487,
      "grad_norm": 0.9705696105957031,
      "learning_rate": 5.319760966009102e-06,
      "loss": 0.1934,
      "num_input_tokens_seen": 27310464,
      "step": 31540
    },
    {
      "epoch": 14.872701555869872,
      "grad_norm": 1.1192429065704346,
      "learning_rate": 5.3137081870954096e-06,
      "loss": 0.1529,
      "num_input_tokens_seen": 27314992,
      "step": 31545
    },
    {
      "epoch": 14.875058934464874,
      "grad_norm": 0.5996915102005005,
      "learning_rate": 5.307658444055313e-06,
      "loss": 0.1262,
      "num_input_tokens_seen": 27319104,
      "step": 31550
    },
    {
      "epoch": 14.877416313059877,
      "grad_norm": 1.4352387189865112,
      "learning_rate": 5.301611737821749e-06,
      "loss": 0.2361,
      "num_input_tokens_seen": 27323168,
      "step": 31555
    },
    {
      "epoch": 14.87977369165488,
      "grad_norm": 0.7910253405570984,
      "learning_rate": 5.295568069327206e-06,
      "loss": 0.2467,
      "num_input_tokens_seen": 27327408,
      "step": 31560
    },
    {
      "epoch": 14.882131070249882,
      "grad_norm": 0.772239089012146,
      "learning_rate": 5.289527439503683e-06,
      "loss": 0.1529,
      "num_input_tokens_seen": 27331456,
      "step": 31565
    },
    {
      "epoch": 14.884488448844884,
      "grad_norm": 0.4797782003879547,
      "learning_rate": 5.28348984928273e-06,
      "loss": 0.1561,
      "num_input_tokens_seen": 27335168,
      "step": 31570
    },
    {
      "epoch": 14.886845827439886,
      "grad_norm": 1.772045612335205,
      "learning_rate": 5.27745529959541e-06,
      "loss": 0.2103,
      "num_input_tokens_seen": 27339392,
      "step": 31575
    },
    {
      "epoch": 14.889203206034889,
      "grad_norm": 0.6632040739059448,
      "learning_rate": 5.271423791372335e-06,
      "loss": 0.129,
      "num_input_tokens_seen": 27343776,
      "step": 31580
    },
    {
      "epoch": 14.891560584629891,
      "grad_norm": 1.3320975303649902,
      "learning_rate": 5.26539532554364e-06,
      "loss": 0.1903,
      "num_input_tokens_seen": 27348016,
      "step": 31585
    },
    {
      "epoch": 14.893917963224894,
      "grad_norm": 1.3982086181640625,
      "learning_rate": 5.25936990303898e-06,
      "loss": 0.1905,
      "num_input_tokens_seen": 27352528,
      "step": 31590
    },
    {
      "epoch": 14.896275341819896,
      "grad_norm": 1.2718743085861206,
      "learning_rate": 5.253347524787555e-06,
      "loss": 0.1714,
      "num_input_tokens_seen": 27357856,
      "step": 31595
    },
    {
      "epoch": 14.898632720414899,
      "grad_norm": 1.7755401134490967,
      "learning_rate": 5.2473281917181035e-06,
      "loss": 0.1362,
      "num_input_tokens_seen": 27362144,
      "step": 31600
    },
    {
      "epoch": 14.898632720414899,
      "eval_loss": 0.19971530139446259,
      "eval_runtime": 22.2044,
      "eval_samples_per_second": 42.469,
      "eval_steps_per_second": 21.257,
      "num_input_tokens_seen": 27362144,
      "step": 31600
    },
    {
      "epoch": 14.900990099009901,
      "grad_norm": 1.23027765750885,
      "learning_rate": 5.241311904758864e-06,
      "loss": 0.1365,
      "num_input_tokens_seen": 27366016,
      "step": 31605
    },
    {
      "epoch": 14.903347477604903,
      "grad_norm": 0.8002904653549194,
      "learning_rate": 5.23529866483764e-06,
      "loss": 0.12,
      "num_input_tokens_seen": 27370320,
      "step": 31610
    },
    {
      "epoch": 14.905704856199906,
      "grad_norm": 0.8601003289222717,
      "learning_rate": 5.229288472881732e-06,
      "loss": 0.2269,
      "num_input_tokens_seen": 27374096,
      "step": 31615
    },
    {
      "epoch": 14.908062234794908,
      "grad_norm": 0.9681161046028137,
      "learning_rate": 5.2232813298180025e-06,
      "loss": 0.1588,
      "num_input_tokens_seen": 27377984,
      "step": 31620
    },
    {
      "epoch": 14.91041961338991,
      "grad_norm": 0.9392572045326233,
      "learning_rate": 5.217277236572824e-06,
      "loss": 0.2626,
      "num_input_tokens_seen": 27382368,
      "step": 31625
    },
    {
      "epoch": 14.912776991984913,
      "grad_norm": 0.94877028465271,
      "learning_rate": 5.211276194072093e-06,
      "loss": 0.1606,
      "num_input_tokens_seen": 27386832,
      "step": 31630
    },
    {
      "epoch": 14.915134370579915,
      "grad_norm": 1.3165462017059326,
      "learning_rate": 5.205278203241254e-06,
      "loss": 0.168,
      "num_input_tokens_seen": 27390736,
      "step": 31635
    },
    {
      "epoch": 14.917491749174918,
      "grad_norm": 0.5450656414031982,
      "learning_rate": 5.199283265005278e-06,
      "loss": 0.1747,
      "num_input_tokens_seen": 27394976,
      "step": 31640
    },
    {
      "epoch": 14.91984912776992,
      "grad_norm": 0.5240219235420227,
      "learning_rate": 5.193291380288648e-06,
      "loss": 0.1852,
      "num_input_tokens_seen": 27399840,
      "step": 31645
    },
    {
      "epoch": 14.922206506364923,
      "grad_norm": 1.2544487714767456,
      "learning_rate": 5.1873025500153995e-06,
      "loss": 0.2398,
      "num_input_tokens_seen": 27404400,
      "step": 31650
    },
    {
      "epoch": 14.924563884959925,
      "grad_norm": 1.3490544557571411,
      "learning_rate": 5.181316775109071e-06,
      "loss": 0.1858,
      "num_input_tokens_seen": 27408176,
      "step": 31655
    },
    {
      "epoch": 14.926921263554927,
      "grad_norm": 1.1886217594146729,
      "learning_rate": 5.1753340564927564e-06,
      "loss": 0.1605,
      "num_input_tokens_seen": 27412720,
      "step": 31660
    },
    {
      "epoch": 14.92927864214993,
      "grad_norm": 0.9869866967201233,
      "learning_rate": 5.169354395089068e-06,
      "loss": 0.1791,
      "num_input_tokens_seen": 27416688,
      "step": 31665
    },
    {
      "epoch": 14.931636020744932,
      "grad_norm": 0.8090296387672424,
      "learning_rate": 5.1633777918201346e-06,
      "loss": 0.1475,
      "num_input_tokens_seen": 27420976,
      "step": 31670
    },
    {
      "epoch": 14.933993399339935,
      "grad_norm": 1.0117406845092773,
      "learning_rate": 5.157404247607625e-06,
      "loss": 0.174,
      "num_input_tokens_seen": 27425088,
      "step": 31675
    },
    {
      "epoch": 14.936350777934937,
      "grad_norm": 1.2712606191635132,
      "learning_rate": 5.1514337633727454e-06,
      "loss": 0.1562,
      "num_input_tokens_seen": 27429952,
      "step": 31680
    },
    {
      "epoch": 14.93870815652994,
      "grad_norm": 2.015594244003296,
      "learning_rate": 5.145466340036206e-06,
      "loss": 0.1325,
      "num_input_tokens_seen": 27434144,
      "step": 31685
    },
    {
      "epoch": 14.941065535124942,
      "grad_norm": 1.4962679147720337,
      "learning_rate": 5.139501978518274e-06,
      "loss": 0.1383,
      "num_input_tokens_seen": 27438640,
      "step": 31690
    },
    {
      "epoch": 14.943422913719944,
      "grad_norm": 0.640068531036377,
      "learning_rate": 5.133540679738716e-06,
      "loss": 0.1375,
      "num_input_tokens_seen": 27444000,
      "step": 31695
    },
    {
      "epoch": 14.945780292314947,
      "grad_norm": 0.8307481408119202,
      "learning_rate": 5.127582444616838e-06,
      "loss": 0.2128,
      "num_input_tokens_seen": 27448160,
      "step": 31700
    },
    {
      "epoch": 14.948137670909949,
      "grad_norm": 0.9668331742286682,
      "learning_rate": 5.121627274071486e-06,
      "loss": 0.1605,
      "num_input_tokens_seen": 27453040,
      "step": 31705
    },
    {
      "epoch": 14.950495049504951,
      "grad_norm": 2.2707111835479736,
      "learning_rate": 5.115675169021009e-06,
      "loss": 0.2148,
      "num_input_tokens_seen": 27457280,
      "step": 31710
    },
    {
      "epoch": 14.952852428099952,
      "grad_norm": 0.43970152735710144,
      "learning_rate": 5.1097261303832994e-06,
      "loss": 0.1445,
      "num_input_tokens_seen": 27461136,
      "step": 31715
    },
    {
      "epoch": 14.955209806694956,
      "grad_norm": 0.817765474319458,
      "learning_rate": 5.103780159075788e-06,
      "loss": 0.1826,
      "num_input_tokens_seen": 27465216,
      "step": 31720
    },
    {
      "epoch": 14.957567185289957,
      "grad_norm": 0.7543616890907288,
      "learning_rate": 5.0978372560154e-06,
      "loss": 0.244,
      "num_input_tokens_seen": 27470448,
      "step": 31725
    },
    {
      "epoch": 14.95992456388496,
      "grad_norm": 0.6913953423500061,
      "learning_rate": 5.091897422118619e-06,
      "loss": 0.1594,
      "num_input_tokens_seen": 27474912,
      "step": 31730
    },
    {
      "epoch": 14.962281942479962,
      "grad_norm": 1.5779149532318115,
      "learning_rate": 5.0859606583014305e-06,
      "loss": 0.1446,
      "num_input_tokens_seen": 27479776,
      "step": 31735
    },
    {
      "epoch": 14.964639321074964,
      "grad_norm": 1.5882296562194824,
      "learning_rate": 5.080026965479365e-06,
      "loss": 0.1698,
      "num_input_tokens_seen": 27484400,
      "step": 31740
    },
    {
      "epoch": 14.966996699669966,
      "grad_norm": 0.8786389231681824,
      "learning_rate": 5.074096344567475e-06,
      "loss": 0.2414,
      "num_input_tokens_seen": 27488912,
      "step": 31745
    },
    {
      "epoch": 14.969354078264969,
      "grad_norm": 1.3222180604934692,
      "learning_rate": 5.0681687964803294e-06,
      "loss": 0.1789,
      "num_input_tokens_seen": 27493024,
      "step": 31750
    },
    {
      "epoch": 14.971711456859971,
      "grad_norm": 0.6075050830841064,
      "learning_rate": 5.06224432213204e-06,
      "loss": 0.1519,
      "num_input_tokens_seen": 27497456,
      "step": 31755
    },
    {
      "epoch": 14.974068835454974,
      "grad_norm": 0.6562567949295044,
      "learning_rate": 5.056322922436224e-06,
      "loss": 0.2148,
      "num_input_tokens_seen": 27501968,
      "step": 31760
    },
    {
      "epoch": 14.976426214049976,
      "grad_norm": 0.7401714324951172,
      "learning_rate": 5.0504045983060465e-06,
      "loss": 0.1684,
      "num_input_tokens_seen": 27505936,
      "step": 31765
    },
    {
      "epoch": 14.978783592644978,
      "grad_norm": 0.8153507113456726,
      "learning_rate": 5.044489350654183e-06,
      "loss": 0.1537,
      "num_input_tokens_seen": 27510208,
      "step": 31770
    },
    {
      "epoch": 14.98114097123998,
      "grad_norm": 1.700142502784729,
      "learning_rate": 5.038577180392831e-06,
      "loss": 0.144,
      "num_input_tokens_seen": 27514640,
      "step": 31775
    },
    {
      "epoch": 14.983498349834983,
      "grad_norm": 1.0308871269226074,
      "learning_rate": 5.032668088433729e-06,
      "loss": 0.1259,
      "num_input_tokens_seen": 27519344,
      "step": 31780
    },
    {
      "epoch": 14.985855728429986,
      "grad_norm": 0.4835364520549774,
      "learning_rate": 5.02676207568814e-06,
      "loss": 0.1374,
      "num_input_tokens_seen": 27523408,
      "step": 31785
    },
    {
      "epoch": 14.988213107024988,
      "grad_norm": 0.8745425939559937,
      "learning_rate": 5.02085914306683e-06,
      "loss": 0.2025,
      "num_input_tokens_seen": 27528176,
      "step": 31790
    },
    {
      "epoch": 14.99057048561999,
      "grad_norm": 0.7273910641670227,
      "learning_rate": 5.014959291480123e-06,
      "loss": 0.2199,
      "num_input_tokens_seen": 27532640,
      "step": 31795
    },
    {
      "epoch": 14.992927864214993,
      "grad_norm": 0.6610400080680847,
      "learning_rate": 5.009062521837835e-06,
      "loss": 0.1321,
      "num_input_tokens_seen": 27536992,
      "step": 31800
    },
    {
      "epoch": 14.992927864214993,
      "eval_loss": 0.1998426914215088,
      "eval_runtime": 22.1499,
      "eval_samples_per_second": 42.574,
      "eval_steps_per_second": 21.309,
      "num_input_tokens_seen": 27536992,
      "step": 31800
    },
    {
      "epoch": 14.995285242809995,
      "grad_norm": 0.6765905618667603,
      "learning_rate": 5.003168835049324e-06,
      "loss": 0.167,
      "num_input_tokens_seen": 27541840,
      "step": 31805
    },
    {
      "epoch": 14.997642621404998,
      "grad_norm": 1.556758165359497,
      "learning_rate": 4.997278232023483e-06,
      "loss": 0.1946,
      "num_input_tokens_seen": 27546128,
      "step": 31810
    },
    {
      "epoch": 15.0,
      "grad_norm": 1.751418948173523,
      "learning_rate": 4.9913907136687036e-06,
      "loss": 0.1657,
      "num_input_tokens_seen": 27550432,
      "step": 31815
    },
    {
      "epoch": 15.002357378595002,
      "grad_norm": 0.4616970121860504,
      "learning_rate": 4.985506280892918e-06,
      "loss": 0.1237,
      "num_input_tokens_seen": 27554720,
      "step": 31820
    },
    {
      "epoch": 15.004714757190005,
      "grad_norm": 0.908376932144165,
      "learning_rate": 4.979624934603589e-06,
      "loss": 0.2783,
      "num_input_tokens_seen": 27559920,
      "step": 31825
    },
    {
      "epoch": 15.007072135785007,
      "grad_norm": 0.5921745300292969,
      "learning_rate": 4.97374667570768e-06,
      "loss": 0.1275,
      "num_input_tokens_seen": 27563696,
      "step": 31830
    },
    {
      "epoch": 15.00942951438001,
      "grad_norm": 1.78781259059906,
      "learning_rate": 4.967871505111704e-06,
      "loss": 0.1912,
      "num_input_tokens_seen": 27567888,
      "step": 31835
    },
    {
      "epoch": 15.011786892975012,
      "grad_norm": 0.48379117250442505,
      "learning_rate": 4.961999423721686e-06,
      "loss": 0.1766,
      "num_input_tokens_seen": 27572704,
      "step": 31840
    },
    {
      "epoch": 15.014144271570014,
      "grad_norm": 0.30580267310142517,
      "learning_rate": 4.956130432443159e-06,
      "loss": 0.1065,
      "num_input_tokens_seen": 27576832,
      "step": 31845
    },
    {
      "epoch": 15.016501650165017,
      "grad_norm": 0.5878511071205139,
      "learning_rate": 4.950264532181215e-06,
      "loss": 0.1886,
      "num_input_tokens_seen": 27581008,
      "step": 31850
    },
    {
      "epoch": 15.01885902876002,
      "grad_norm": 0.5127586126327515,
      "learning_rate": 4.944401723840433e-06,
      "loss": 0.1743,
      "num_input_tokens_seen": 27585152,
      "step": 31855
    },
    {
      "epoch": 15.021216407355022,
      "grad_norm": 1.1348265409469604,
      "learning_rate": 4.938542008324942e-06,
      "loss": 0.1626,
      "num_input_tokens_seen": 27590336,
      "step": 31860
    },
    {
      "epoch": 15.023573785950024,
      "grad_norm": 1.1217530965805054,
      "learning_rate": 4.9326853865383855e-06,
      "loss": 0.2316,
      "num_input_tokens_seen": 27594208,
      "step": 31865
    },
    {
      "epoch": 15.025931164545026,
      "grad_norm": 1.0617600679397583,
      "learning_rate": 4.926831859383918e-06,
      "loss": 0.1263,
      "num_input_tokens_seen": 27598272,
      "step": 31870
    },
    {
      "epoch": 15.028288543140029,
      "grad_norm": 1.2764047384262085,
      "learning_rate": 4.92098142776424e-06,
      "loss": 0.1961,
      "num_input_tokens_seen": 27602064,
      "step": 31875
    },
    {
      "epoch": 15.030645921735031,
      "grad_norm": 0.3491075932979584,
      "learning_rate": 4.91513409258155e-06,
      "loss": 0.1298,
      "num_input_tokens_seen": 27605952,
      "step": 31880
    },
    {
      "epoch": 15.033003300330034,
      "grad_norm": 2.5802321434020996,
      "learning_rate": 4.909289854737581e-06,
      "loss": 0.2342,
      "num_input_tokens_seen": 27611088,
      "step": 31885
    },
    {
      "epoch": 15.035360678925036,
      "grad_norm": 0.7549448013305664,
      "learning_rate": 4.903448715133602e-06,
      "loss": 0.1429,
      "num_input_tokens_seen": 27615952,
      "step": 31890
    },
    {
      "epoch": 15.037718057520038,
      "grad_norm": 1.0013357400894165,
      "learning_rate": 4.897610674670372e-06,
      "loss": 0.1575,
      "num_input_tokens_seen": 27620432,
      "step": 31895
    },
    {
      "epoch": 15.04007543611504,
      "grad_norm": 0.4260270297527313,
      "learning_rate": 4.8917757342482e-06,
      "loss": 0.2479,
      "num_input_tokens_seen": 27624064,
      "step": 31900
    },
    {
      "epoch": 15.042432814710043,
      "grad_norm": 0.6980153322219849,
      "learning_rate": 4.885943894766909e-06,
      "loss": 0.133,
      "num_input_tokens_seen": 27628128,
      "step": 31905
    },
    {
      "epoch": 15.044790193305046,
      "grad_norm": 1.106559157371521,
      "learning_rate": 4.880115157125842e-06,
      "loss": 0.2198,
      "num_input_tokens_seen": 27631920,
      "step": 31910
    },
    {
      "epoch": 15.047147571900048,
      "grad_norm": 1.283617615699768,
      "learning_rate": 4.874289522223857e-06,
      "loss": 0.1487,
      "num_input_tokens_seen": 27636032,
      "step": 31915
    },
    {
      "epoch": 15.049504950495049,
      "grad_norm": 0.6268705129623413,
      "learning_rate": 4.868466990959339e-06,
      "loss": 0.2527,
      "num_input_tokens_seen": 27639744,
      "step": 31920
    },
    {
      "epoch": 15.051862329090051,
      "grad_norm": 2.20430326461792,
      "learning_rate": 4.8626475642301964e-06,
      "loss": 0.1688,
      "num_input_tokens_seen": 27644064,
      "step": 31925
    },
    {
      "epoch": 15.054219707685053,
      "grad_norm": 0.9148632287979126,
      "learning_rate": 4.856831242933871e-06,
      "loss": 0.2218,
      "num_input_tokens_seen": 27648528,
      "step": 31930
    },
    {
      "epoch": 15.056577086280056,
      "grad_norm": 0.5378922820091248,
      "learning_rate": 4.851018027967294e-06,
      "loss": 0.1661,
      "num_input_tokens_seen": 27653072,
      "step": 31935
    },
    {
      "epoch": 15.058934464875058,
      "grad_norm": 0.590462863445282,
      "learning_rate": 4.845207920226946e-06,
      "loss": 0.181,
      "num_input_tokens_seen": 27657168,
      "step": 31940
    },
    {
      "epoch": 15.06129184347006,
      "grad_norm": 1.6082022190093994,
      "learning_rate": 4.839400920608825e-06,
      "loss": 0.1795,
      "num_input_tokens_seen": 27661248,
      "step": 31945
    },
    {
      "epoch": 15.063649222065063,
      "grad_norm": 2.1558492183685303,
      "learning_rate": 4.83359703000843e-06,
      "loss": 0.1898,
      "num_input_tokens_seen": 27665536,
      "step": 31950
    },
    {
      "epoch": 15.066006600660065,
      "grad_norm": 1.811932921409607,
      "learning_rate": 4.827796249320804e-06,
      "loss": 0.2071,
      "num_input_tokens_seen": 27669392,
      "step": 31955
    },
    {
      "epoch": 15.068363979255068,
      "grad_norm": 1.2653740644454956,
      "learning_rate": 4.82199857944049e-06,
      "loss": 0.1895,
      "num_input_tokens_seen": 27673216,
      "step": 31960
    },
    {
      "epoch": 15.07072135785007,
      "grad_norm": 0.603683590888977,
      "learning_rate": 4.8162040212615695e-06,
      "loss": 0.2405,
      "num_input_tokens_seen": 27677744,
      "step": 31965
    },
    {
      "epoch": 15.073078736445073,
      "grad_norm": 1.6580016613006592,
      "learning_rate": 4.810412575677639e-06,
      "loss": 0.1329,
      "num_input_tokens_seen": 27681504,
      "step": 31970
    },
    {
      "epoch": 15.075436115040075,
      "grad_norm": 1.5934890508651733,
      "learning_rate": 4.804624243581801e-06,
      "loss": 0.1352,
      "num_input_tokens_seen": 27685712,
      "step": 31975
    },
    {
      "epoch": 15.077793493635077,
      "grad_norm": 0.7621966004371643,
      "learning_rate": 4.798839025866703e-06,
      "loss": 0.1339,
      "num_input_tokens_seen": 27690000,
      "step": 31980
    },
    {
      "epoch": 15.08015087223008,
      "grad_norm": 0.6992766857147217,
      "learning_rate": 4.793056923424491e-06,
      "loss": 0.1842,
      "num_input_tokens_seen": 27693968,
      "step": 31985
    },
    {
      "epoch": 15.082508250825082,
      "grad_norm": 0.5695461630821228,
      "learning_rate": 4.78727793714683e-06,
      "loss": 0.1897,
      "num_input_tokens_seen": 27698544,
      "step": 31990
    },
    {
      "epoch": 15.084865629420085,
      "grad_norm": 0.6396686434745789,
      "learning_rate": 4.7815020679249285e-06,
      "loss": 0.1733,
      "num_input_tokens_seen": 27703424,
      "step": 31995
    },
    {
      "epoch": 15.087223008015087,
      "grad_norm": 1.5170581340789795,
      "learning_rate": 4.775729316649483e-06,
      "loss": 0.1896,
      "num_input_tokens_seen": 27707728,
      "step": 32000
    },
    {
      "epoch": 15.087223008015087,
      "eval_loss": 0.2004110962152481,
      "eval_runtime": 22.1659,
      "eval_samples_per_second": 42.543,
      "eval_steps_per_second": 21.294,
      "num_input_tokens_seen": 27707728,
      "step": 32000
    },
    {
      "epoch": 15.08958038661009,
      "grad_norm": 1.0087794065475464,
      "learning_rate": 4.769959684210728e-06,
      "loss": 0.2107,
      "num_input_tokens_seen": 27712640,
      "step": 32005
    },
    {
      "epoch": 15.091937765205092,
      "grad_norm": 1.6743813753128052,
      "learning_rate": 4.764193171498426e-06,
      "loss": 0.1388,
      "num_input_tokens_seen": 27717984,
      "step": 32010
    },
    {
      "epoch": 15.094295143800094,
      "grad_norm": 1.008339524269104,
      "learning_rate": 4.75842977940183e-06,
      "loss": 0.2108,
      "num_input_tokens_seen": 27722592,
      "step": 32015
    },
    {
      "epoch": 15.096652522395097,
      "grad_norm": 0.46439990401268005,
      "learning_rate": 4.752669508809729e-06,
      "loss": 0.184,
      "num_input_tokens_seen": 27727152,
      "step": 32020
    },
    {
      "epoch": 15.099009900990099,
      "grad_norm": 0.8736028671264648,
      "learning_rate": 4.746912360610445e-06,
      "loss": 0.1134,
      "num_input_tokens_seen": 27730688,
      "step": 32025
    },
    {
      "epoch": 15.101367279585101,
      "grad_norm": 0.5855129957199097,
      "learning_rate": 4.741158335691781e-06,
      "loss": 0.1721,
      "num_input_tokens_seen": 27736304,
      "step": 32030
    },
    {
      "epoch": 15.103724658180104,
      "grad_norm": 1.6902676820755005,
      "learning_rate": 4.7354074349410994e-06,
      "loss": 0.2415,
      "num_input_tokens_seen": 27740560,
      "step": 32035
    },
    {
      "epoch": 15.106082036775106,
      "grad_norm": 0.6871082782745361,
      "learning_rate": 4.729659659245245e-06,
      "loss": 0.2103,
      "num_input_tokens_seen": 27745744,
      "step": 32040
    },
    {
      "epoch": 15.108439415370109,
      "grad_norm": 1.030534029006958,
      "learning_rate": 4.723915009490601e-06,
      "loss": 0.2697,
      "num_input_tokens_seen": 27749568,
      "step": 32045
    },
    {
      "epoch": 15.110796793965111,
      "grad_norm": 2.1300032138824463,
      "learning_rate": 4.718173486563077e-06,
      "loss": 0.2157,
      "num_input_tokens_seen": 27754656,
      "step": 32050
    },
    {
      "epoch": 15.113154172560114,
      "grad_norm": 1.4229258298873901,
      "learning_rate": 4.71243509134808e-06,
      "loss": 0.1355,
      "num_input_tokens_seen": 27759248,
      "step": 32055
    },
    {
      "epoch": 15.115511551155116,
      "grad_norm": 0.38564634323120117,
      "learning_rate": 4.706699824730532e-06,
      "loss": 0.1775,
      "num_input_tokens_seen": 27763824,
      "step": 32060
    },
    {
      "epoch": 15.117868929750118,
      "grad_norm": 0.35805121064186096,
      "learning_rate": 4.700967687594901e-06,
      "loss": 0.1382,
      "num_input_tokens_seen": 27768528,
      "step": 32065
    },
    {
      "epoch": 15.12022630834512,
      "grad_norm": 1.1754258871078491,
      "learning_rate": 4.69523868082514e-06,
      "loss": 0.1895,
      "num_input_tokens_seen": 27772736,
      "step": 32070
    },
    {
      "epoch": 15.122583686940123,
      "grad_norm": 0.44963476061820984,
      "learning_rate": 4.689512805304747e-06,
      "loss": 0.1358,
      "num_input_tokens_seen": 27776800,
      "step": 32075
    },
    {
      "epoch": 15.124941065535126,
      "grad_norm": 1.1132287979125977,
      "learning_rate": 4.683790061916707e-06,
      "loss": 0.1725,
      "num_input_tokens_seen": 27781232,
      "step": 32080
    },
    {
      "epoch": 15.127298444130128,
      "grad_norm": 1.6957449913024902,
      "learning_rate": 4.678070451543551e-06,
      "loss": 0.2001,
      "num_input_tokens_seen": 27785968,
      "step": 32085
    },
    {
      "epoch": 15.12965582272513,
      "grad_norm": 2.675936460494995,
      "learning_rate": 4.6723539750673204e-06,
      "loss": 0.1871,
      "num_input_tokens_seen": 27791472,
      "step": 32090
    },
    {
      "epoch": 15.132013201320133,
      "grad_norm": 0.875603973865509,
      "learning_rate": 4.666640633369551e-06,
      "loss": 0.1386,
      "num_input_tokens_seen": 27796064,
      "step": 32095
    },
    {
      "epoch": 15.134370579915135,
      "grad_norm": 0.522096574306488,
      "learning_rate": 4.660930427331323e-06,
      "loss": 0.1346,
      "num_input_tokens_seen": 27801072,
      "step": 32100
    },
    {
      "epoch": 15.136727958510138,
      "grad_norm": 1.2245066165924072,
      "learning_rate": 4.6552233578332244e-06,
      "loss": 0.1948,
      "num_input_tokens_seen": 27805344,
      "step": 32105
    },
    {
      "epoch": 15.13908533710514,
      "grad_norm": 1.3753608465194702,
      "learning_rate": 4.649519425755347e-06,
      "loss": 0.1483,
      "num_input_tokens_seen": 27808880,
      "step": 32110
    },
    {
      "epoch": 15.14144271570014,
      "grad_norm": 0.5763738751411438,
      "learning_rate": 4.64381863197732e-06,
      "loss": 0.1491,
      "num_input_tokens_seen": 27813120,
      "step": 32115
    },
    {
      "epoch": 15.143800094295143,
      "grad_norm": 1.613087773323059,
      "learning_rate": 4.638120977378269e-06,
      "loss": 0.1971,
      "num_input_tokens_seen": 27816880,
      "step": 32120
    },
    {
      "epoch": 15.146157472890145,
      "grad_norm": 0.40279120206832886,
      "learning_rate": 4.632426462836848e-06,
      "loss": 0.1731,
      "num_input_tokens_seen": 27820976,
      "step": 32125
    },
    {
      "epoch": 15.148514851485148,
      "grad_norm": 1.286357045173645,
      "learning_rate": 4.626735089231224e-06,
      "loss": 0.1924,
      "num_input_tokens_seen": 27824944,
      "step": 32130
    },
    {
      "epoch": 15.15087223008015,
      "grad_norm": 1.1743541955947876,
      "learning_rate": 4.621046857439068e-06,
      "loss": 0.1944,
      "num_input_tokens_seen": 27830144,
      "step": 32135
    },
    {
      "epoch": 15.153229608675153,
      "grad_norm": 0.4782865345478058,
      "learning_rate": 4.615361768337587e-06,
      "loss": 0.1537,
      "num_input_tokens_seen": 27834496,
      "step": 32140
    },
    {
      "epoch": 15.155586987270155,
      "grad_norm": 0.80438232421875,
      "learning_rate": 4.6096798228034946e-06,
      "loss": 0.1336,
      "num_input_tokens_seen": 27838944,
      "step": 32145
    },
    {
      "epoch": 15.157944365865157,
      "grad_norm": 1.325225830078125,
      "learning_rate": 4.604001021713008e-06,
      "loss": 0.1724,
      "num_input_tokens_seen": 27842752,
      "step": 32150
    },
    {
      "epoch": 15.16030174446016,
      "grad_norm": 0.5719174742698669,
      "learning_rate": 4.598325365941883e-06,
      "loss": 0.217,
      "num_input_tokens_seen": 27846848,
      "step": 32155
    },
    {
      "epoch": 15.162659123055162,
      "grad_norm": 0.7876774072647095,
      "learning_rate": 4.5926528563653645e-06,
      "loss": 0.1348,
      "num_input_tokens_seen": 27850992,
      "step": 32160
    },
    {
      "epoch": 15.165016501650165,
      "grad_norm": 0.6465445756912231,
      "learning_rate": 4.5869834938582295e-06,
      "loss": 0.1501,
      "num_input_tokens_seen": 27854992,
      "step": 32165
    },
    {
      "epoch": 15.167373880245167,
      "grad_norm": 0.9112846851348877,
      "learning_rate": 4.581317279294772e-06,
      "loss": 0.2337,
      "num_input_tokens_seen": 27859584,
      "step": 32170
    },
    {
      "epoch": 15.16973125884017,
      "grad_norm": 0.6163817644119263,
      "learning_rate": 4.57565421354878e-06,
      "loss": 0.1727,
      "num_input_tokens_seen": 27864304,
      "step": 32175
    },
    {
      "epoch": 15.172088637435172,
      "grad_norm": 1.5091931819915771,
      "learning_rate": 4.569994297493579e-06,
      "loss": 0.1499,
      "num_input_tokens_seen": 27868416,
      "step": 32180
    },
    {
      "epoch": 15.174446016030174,
      "grad_norm": 0.5955219268798828,
      "learning_rate": 4.564337532002002e-06,
      "loss": 0.1187,
      "num_input_tokens_seen": 27873520,
      "step": 32185
    },
    {
      "epoch": 15.176803394625177,
      "grad_norm": 0.695785403251648,
      "learning_rate": 4.55868391794638e-06,
      "loss": 0.236,
      "num_input_tokens_seen": 27878128,
      "step": 32190
    },
    {
      "epoch": 15.179160773220179,
      "grad_norm": 1.142803430557251,
      "learning_rate": 4.553033456198588e-06,
      "loss": 0.2356,
      "num_input_tokens_seen": 27881968,
      "step": 32195
    },
    {
      "epoch": 15.181518151815181,
      "grad_norm": 0.972148060798645,
      "learning_rate": 4.54738614762999e-06,
      "loss": 0.1785,
      "num_input_tokens_seen": 27886368,
      "step": 32200
    },
    {
      "epoch": 15.181518151815181,
      "eval_loss": 0.2011071890592575,
      "eval_runtime": 22.1809,
      "eval_samples_per_second": 42.514,
      "eval_steps_per_second": 21.28,
      "num_input_tokens_seen": 27886368,
      "step": 32200
    },
    {
      "epoch": 15.183875530410184,
      "grad_norm": 2.788815975189209,
      "learning_rate": 4.541741993111465e-06,
      "loss": 0.1994,
      "num_input_tokens_seen": 27891920,
      "step": 32205
    },
    {
      "epoch": 15.186232909005186,
      "grad_norm": 0.9344422221183777,
      "learning_rate": 4.536100993513423e-06,
      "loss": 0.1389,
      "num_input_tokens_seen": 27896736,
      "step": 32210
    },
    {
      "epoch": 15.188590287600189,
      "grad_norm": 0.9539199471473694,
      "learning_rate": 4.530463149705768e-06,
      "loss": 0.16,
      "num_input_tokens_seen": 27901440,
      "step": 32215
    },
    {
      "epoch": 15.190947666195191,
      "grad_norm": 0.9446972608566284,
      "learning_rate": 4.524828462557934e-06,
      "loss": 0.1297,
      "num_input_tokens_seen": 27906208,
      "step": 32220
    },
    {
      "epoch": 15.193305044790193,
      "grad_norm": 1.884371280670166,
      "learning_rate": 4.5191969329388625e-06,
      "loss": 0.1997,
      "num_input_tokens_seen": 27911120,
      "step": 32225
    },
    {
      "epoch": 15.195662423385196,
      "grad_norm": 0.8908267021179199,
      "learning_rate": 4.5135685617169965e-06,
      "loss": 0.1877,
      "num_input_tokens_seen": 27915488,
      "step": 32230
    },
    {
      "epoch": 15.198019801980198,
      "grad_norm": 1.0352294445037842,
      "learning_rate": 4.507943349760313e-06,
      "loss": 0.1889,
      "num_input_tokens_seen": 27920528,
      "step": 32235
    },
    {
      "epoch": 15.2003771805752,
      "grad_norm": 0.4067256450653076,
      "learning_rate": 4.502321297936277e-06,
      "loss": 0.1655,
      "num_input_tokens_seen": 27925120,
      "step": 32240
    },
    {
      "epoch": 15.202734559170203,
      "grad_norm": 0.7257657051086426,
      "learning_rate": 4.496702407111888e-06,
      "loss": 0.2333,
      "num_input_tokens_seen": 27929072,
      "step": 32245
    },
    {
      "epoch": 15.205091937765205,
      "grad_norm": 0.9322750568389893,
      "learning_rate": 4.491086678153653e-06,
      "loss": 0.1964,
      "num_input_tokens_seen": 27932976,
      "step": 32250
    },
    {
      "epoch": 15.207449316360208,
      "grad_norm": 0.5512787699699402,
      "learning_rate": 4.485474111927579e-06,
      "loss": 0.1417,
      "num_input_tokens_seen": 27936912,
      "step": 32255
    },
    {
      "epoch": 15.20980669495521,
      "grad_norm": 0.808140754699707,
      "learning_rate": 4.479864709299197e-06,
      "loss": 0.1978,
      "num_input_tokens_seen": 27941568,
      "step": 32260
    },
    {
      "epoch": 15.212164073550213,
      "grad_norm": 1.0416632890701294,
      "learning_rate": 4.474258471133555e-06,
      "loss": 0.2562,
      "num_input_tokens_seen": 27946512,
      "step": 32265
    },
    {
      "epoch": 15.214521452145215,
      "grad_norm": 0.7085699439048767,
      "learning_rate": 4.4686553982952014e-06,
      "loss": 0.149,
      "num_input_tokens_seen": 27950832,
      "step": 32270
    },
    {
      "epoch": 15.216878830740217,
      "grad_norm": 0.4843701124191284,
      "learning_rate": 4.463055491648191e-06,
      "loss": 0.2404,
      "num_input_tokens_seen": 27955056,
      "step": 32275
    },
    {
      "epoch": 15.21923620933522,
      "grad_norm": 1.3850077390670776,
      "learning_rate": 4.457458752056112e-06,
      "loss": 0.2491,
      "num_input_tokens_seen": 27959280,
      "step": 32280
    },
    {
      "epoch": 15.221593587930222,
      "grad_norm": 0.9250235557556152,
      "learning_rate": 4.451865180382042e-06,
      "loss": 0.1605,
      "num_input_tokens_seen": 27963424,
      "step": 32285
    },
    {
      "epoch": 15.223950966525225,
      "grad_norm": 0.574710488319397,
      "learning_rate": 4.4462747774885936e-06,
      "loss": 0.1279,
      "num_input_tokens_seen": 27966928,
      "step": 32290
    },
    {
      "epoch": 15.226308345120227,
      "grad_norm": 0.8630909323692322,
      "learning_rate": 4.440687544237859e-06,
      "loss": 0.2178,
      "num_input_tokens_seen": 27971360,
      "step": 32295
    },
    {
      "epoch": 15.22866572371523,
      "grad_norm": 0.7451207041740417,
      "learning_rate": 4.435103481491471e-06,
      "loss": 0.1471,
      "num_input_tokens_seen": 27975328,
      "step": 32300
    },
    {
      "epoch": 15.231023102310232,
      "grad_norm": 1.215875506401062,
      "learning_rate": 4.429522590110569e-06,
      "loss": 0.1856,
      "num_input_tokens_seen": 27979840,
      "step": 32305
    },
    {
      "epoch": 15.233380480905234,
      "grad_norm": 1.0244548320770264,
      "learning_rate": 4.423944870955779e-06,
      "loss": 0.1598,
      "num_input_tokens_seen": 27984320,
      "step": 32310
    },
    {
      "epoch": 15.235737859500237,
      "grad_norm": 0.43594756722450256,
      "learning_rate": 4.418370324887272e-06,
      "loss": 0.129,
      "num_input_tokens_seen": 27988320,
      "step": 32315
    },
    {
      "epoch": 15.238095238095237,
      "grad_norm": 1.2927480936050415,
      "learning_rate": 4.412798952764699e-06,
      "loss": 0.2068,
      "num_input_tokens_seen": 27992304,
      "step": 32320
    },
    {
      "epoch": 15.24045261669024,
      "grad_norm": 0.9712690114974976,
      "learning_rate": 4.407230755447245e-06,
      "loss": 0.2343,
      "num_input_tokens_seen": 27996496,
      "step": 32325
    },
    {
      "epoch": 15.242809995285242,
      "grad_norm": 1.5182582139968872,
      "learning_rate": 4.401665733793598e-06,
      "loss": 0.1566,
      "num_input_tokens_seen": 28000576,
      "step": 32330
    },
    {
      "epoch": 15.245167373880244,
      "grad_norm": 0.6976442933082581,
      "learning_rate": 4.3961038886619425e-06,
      "loss": 0.1809,
      "num_input_tokens_seen": 28005040,
      "step": 32335
    },
    {
      "epoch": 15.247524752475247,
      "grad_norm": 0.7382732033729553,
      "learning_rate": 4.39054522091e-06,
      "loss": 0.1904,
      "num_input_tokens_seen": 28008960,
      "step": 32340
    },
    {
      "epoch": 15.24988213107025,
      "grad_norm": 0.9516440033912659,
      "learning_rate": 4.384989731394979e-06,
      "loss": 0.1453,
      "num_input_tokens_seen": 28013392,
      "step": 32345
    },
    {
      "epoch": 15.252239509665252,
      "grad_norm": 0.7223021984100342,
      "learning_rate": 4.379437420973598e-06,
      "loss": 0.1875,
      "num_input_tokens_seen": 28016720,
      "step": 32350
    },
    {
      "epoch": 15.254596888260254,
      "grad_norm": 1.816404104232788,
      "learning_rate": 4.373888290502107e-06,
      "loss": 0.1835,
      "num_input_tokens_seen": 28021424,
      "step": 32355
    },
    {
      "epoch": 15.256954266855256,
      "grad_norm": 0.7275422811508179,
      "learning_rate": 4.36834234083624e-06,
      "loss": 0.2363,
      "num_input_tokens_seen": 28026336,
      "step": 32360
    },
    {
      "epoch": 15.259311645450259,
      "grad_norm": 0.667776346206665,
      "learning_rate": 4.362799572831258e-06,
      "loss": 0.1369,
      "num_input_tokens_seen": 28030656,
      "step": 32365
    },
    {
      "epoch": 15.261669024045261,
      "grad_norm": 2.0240519046783447,
      "learning_rate": 4.35725998734193e-06,
      "loss": 0.1852,
      "num_input_tokens_seen": 28035152,
      "step": 32370
    },
    {
      "epoch": 15.264026402640264,
      "grad_norm": 1.2423646450042725,
      "learning_rate": 4.3517235852225195e-06,
      "loss": 0.2359,
      "num_input_tokens_seen": 28039632,
      "step": 32375
    },
    {
      "epoch": 15.266383781235266,
      "grad_norm": 1.3205918073654175,
      "learning_rate": 4.346190367326822e-06,
      "loss": 0.1446,
      "num_input_tokens_seen": 28043680,
      "step": 32380
    },
    {
      "epoch": 15.268741159830268,
      "grad_norm": 2.104884147644043,
      "learning_rate": 4.340660334508115e-06,
      "loss": 0.2372,
      "num_input_tokens_seen": 28048432,
      "step": 32385
    },
    {
      "epoch": 15.27109853842527,
      "grad_norm": 0.8380875587463379,
      "learning_rate": 4.335133487619206e-06,
      "loss": 0.1565,
      "num_input_tokens_seen": 28052128,
      "step": 32390
    },
    {
      "epoch": 15.273455917020273,
      "grad_norm": 1.0881410837173462,
      "learning_rate": 4.329609827512409e-06,
      "loss": 0.1892,
      "num_input_tokens_seen": 28056176,
      "step": 32395
    },
    {
      "epoch": 15.275813295615276,
      "grad_norm": 0.7586479187011719,
      "learning_rate": 4.324089355039531e-06,
      "loss": 0.2646,
      "num_input_tokens_seen": 28061984,
      "step": 32400
    },
    {
      "epoch": 15.275813295615276,
      "eval_loss": 0.20009984076023102,
      "eval_runtime": 22.18,
      "eval_samples_per_second": 42.516,
      "eval_steps_per_second": 21.28,
      "num_input_tokens_seen": 28061984,
      "step": 32400
    },
    {
      "epoch": 15.278170674210278,
      "grad_norm": 0.6816777586936951,
      "learning_rate": 4.3185720710519075e-06,
      "loss": 0.224,
      "num_input_tokens_seen": 28066528,
      "step": 32405
    },
    {
      "epoch": 15.28052805280528,
      "grad_norm": 0.6242278814315796,
      "learning_rate": 4.3130579764003724e-06,
      "loss": 0.1425,
      "num_input_tokens_seen": 28070688,
      "step": 32410
    },
    {
      "epoch": 15.282885431400283,
      "grad_norm": 1.9026235342025757,
      "learning_rate": 4.307547071935267e-06,
      "loss": 0.1717,
      "num_input_tokens_seen": 28074704,
      "step": 32415
    },
    {
      "epoch": 15.285242809995285,
      "grad_norm": 1.1474766731262207,
      "learning_rate": 4.302039358506435e-06,
      "loss": 0.2634,
      "num_input_tokens_seen": 28078816,
      "step": 32420
    },
    {
      "epoch": 15.287600188590288,
      "grad_norm": 0.6269732713699341,
      "learning_rate": 4.296534836963245e-06,
      "loss": 0.1971,
      "num_input_tokens_seen": 28082736,
      "step": 32425
    },
    {
      "epoch": 15.28995756718529,
      "grad_norm": 1.3560820817947388,
      "learning_rate": 4.291033508154555e-06,
      "loss": 0.167,
      "num_input_tokens_seen": 28087248,
      "step": 32430
    },
    {
      "epoch": 15.292314945780292,
      "grad_norm": 1.067326307296753,
      "learning_rate": 4.285535372928748e-06,
      "loss": 0.2547,
      "num_input_tokens_seen": 28091744,
      "step": 32435
    },
    {
      "epoch": 15.294672324375295,
      "grad_norm": 1.2700697183609009,
      "learning_rate": 4.280040432133695e-06,
      "loss": 0.2272,
      "num_input_tokens_seen": 28095360,
      "step": 32440
    },
    {
      "epoch": 15.297029702970297,
      "grad_norm": 2.116887331008911,
      "learning_rate": 4.274548686616789e-06,
      "loss": 0.2146,
      "num_input_tokens_seen": 28099888,
      "step": 32445
    },
    {
      "epoch": 15.2993870815653,
      "grad_norm": 0.7988749742507935,
      "learning_rate": 4.2690601372249364e-06,
      "loss": 0.1876,
      "num_input_tokens_seen": 28103472,
      "step": 32450
    },
    {
      "epoch": 15.301744460160302,
      "grad_norm": 1.21495521068573,
      "learning_rate": 4.263574784804525e-06,
      "loss": 0.1819,
      "num_input_tokens_seen": 28107680,
      "step": 32455
    },
    {
      "epoch": 15.304101838755304,
      "grad_norm": 1.029961347579956,
      "learning_rate": 4.258092630201479e-06,
      "loss": 0.196,
      "num_input_tokens_seen": 28111872,
      "step": 32460
    },
    {
      "epoch": 15.306459217350307,
      "grad_norm": 0.9532575607299805,
      "learning_rate": 4.252613674261202e-06,
      "loss": 0.1444,
      "num_input_tokens_seen": 28117280,
      "step": 32465
    },
    {
      "epoch": 15.30881659594531,
      "grad_norm": 1.2090163230895996,
      "learning_rate": 4.2471379178286224e-06,
      "loss": 0.1866,
      "num_input_tokens_seen": 28121136,
      "step": 32470
    },
    {
      "epoch": 15.311173974540312,
      "grad_norm": 1.3954849243164062,
      "learning_rate": 4.241665361748181e-06,
      "loss": 0.1819,
      "num_input_tokens_seen": 28127248,
      "step": 32475
    },
    {
      "epoch": 15.313531353135314,
      "grad_norm": 1.9537773132324219,
      "learning_rate": 4.2361960068637994e-06,
      "loss": 0.1303,
      "num_input_tokens_seen": 28131184,
      "step": 32480
    },
    {
      "epoch": 15.315888731730317,
      "grad_norm": 1.088558316230774,
      "learning_rate": 4.230729854018933e-06,
      "loss": 0.1747,
      "num_input_tokens_seen": 28135248,
      "step": 32485
    },
    {
      "epoch": 15.318246110325319,
      "grad_norm": 2.0637154579162598,
      "learning_rate": 4.225266904056521e-06,
      "loss": 0.1959,
      "num_input_tokens_seen": 28139424,
      "step": 32490
    },
    {
      "epoch": 15.320603488920321,
      "grad_norm": 1.095420241355896,
      "learning_rate": 4.21980715781903e-06,
      "loss": 0.1867,
      "num_input_tokens_seen": 28143776,
      "step": 32495
    },
    {
      "epoch": 15.322960867515324,
      "grad_norm": 1.4057101011276245,
      "learning_rate": 4.214350616148416e-06,
      "loss": 0.1648,
      "num_input_tokens_seen": 28148032,
      "step": 32500
    },
    {
      "epoch": 15.325318246110326,
      "grad_norm": 0.7600475549697876,
      "learning_rate": 4.20889727988614e-06,
      "loss": 0.131,
      "num_input_tokens_seen": 28151888,
      "step": 32505
    },
    {
      "epoch": 15.327675624705329,
      "grad_norm": 0.5779590606689453,
      "learning_rate": 4.20344714987318e-06,
      "loss": 0.225,
      "num_input_tokens_seen": 28155760,
      "step": 32510
    },
    {
      "epoch": 15.33003300330033,
      "grad_norm": 0.7280775308609009,
      "learning_rate": 4.198000226950022e-06,
      "loss": 0.2481,
      "num_input_tokens_seen": 28160560,
      "step": 32515
    },
    {
      "epoch": 15.332390381895332,
      "grad_norm": 0.7715330719947815,
      "learning_rate": 4.192556511956635e-06,
      "loss": 0.1956,
      "num_input_tokens_seen": 28164672,
      "step": 32520
    },
    {
      "epoch": 15.334747760490334,
      "grad_norm": 0.5294356346130371,
      "learning_rate": 4.18711600573252e-06,
      "loss": 0.217,
      "num_input_tokens_seen": 28169056,
      "step": 32525
    },
    {
      "epoch": 15.337105139085336,
      "grad_norm": 0.4793950915336609,
      "learning_rate": 4.181678709116671e-06,
      "loss": 0.1592,
      "num_input_tokens_seen": 28173040,
      "step": 32530
    },
    {
      "epoch": 15.339462517680339,
      "grad_norm": 0.8967674970626831,
      "learning_rate": 4.1762446229475785e-06,
      "loss": 0.1246,
      "num_input_tokens_seen": 28176736,
      "step": 32535
    },
    {
      "epoch": 15.341819896275341,
      "grad_norm": 0.7033670544624329,
      "learning_rate": 4.17081374806326e-06,
      "loss": 0.1988,
      "num_input_tokens_seen": 28182240,
      "step": 32540
    },
    {
      "epoch": 15.344177274870344,
      "grad_norm": 1.1141319274902344,
      "learning_rate": 4.165386085301212e-06,
      "loss": 0.1335,
      "num_input_tokens_seen": 28186784,
      "step": 32545
    },
    {
      "epoch": 15.346534653465346,
      "grad_norm": 1.0215359926223755,
      "learning_rate": 4.1599616354984525e-06,
      "loss": 0.1375,
      "num_input_tokens_seen": 28191776,
      "step": 32550
    },
    {
      "epoch": 15.348892032060348,
      "grad_norm": 0.576607346534729,
      "learning_rate": 4.154540399491508e-06,
      "loss": 0.1712,
      "num_input_tokens_seen": 28195568,
      "step": 32555
    },
    {
      "epoch": 15.35124941065535,
      "grad_norm": 1.25444495677948,
      "learning_rate": 4.149122378116394e-06,
      "loss": 0.1904,
      "num_input_tokens_seen": 28199488,
      "step": 32560
    },
    {
      "epoch": 15.353606789250353,
      "grad_norm": 0.4616631269454956,
      "learning_rate": 4.14370757220863e-06,
      "loss": 0.1614,
      "num_input_tokens_seen": 28204384,
      "step": 32565
    },
    {
      "epoch": 15.355964167845356,
      "grad_norm": 1.5359095335006714,
      "learning_rate": 4.138295982603263e-06,
      "loss": 0.1881,
      "num_input_tokens_seen": 28208464,
      "step": 32570
    },
    {
      "epoch": 15.358321546440358,
      "grad_norm": 0.6789297461509705,
      "learning_rate": 4.132887610134814e-06,
      "loss": 0.1842,
      "num_input_tokens_seen": 28212000,
      "step": 32575
    },
    {
      "epoch": 15.36067892503536,
      "grad_norm": 0.6576808094978333,
      "learning_rate": 4.127482455637335e-06,
      "loss": 0.1502,
      "num_input_tokens_seen": 28216064,
      "step": 32580
    },
    {
      "epoch": 15.363036303630363,
      "grad_norm": 0.7666691541671753,
      "learning_rate": 4.1220805199443545e-06,
      "loss": 0.1386,
      "num_input_tokens_seen": 28220320,
      "step": 32585
    },
    {
      "epoch": 15.365393682225365,
      "grad_norm": 0.713118851184845,
      "learning_rate": 4.116681803888925e-06,
      "loss": 0.1463,
      "num_input_tokens_seen": 28224240,
      "step": 32590
    },
    {
      "epoch": 15.367751060820368,
      "grad_norm": 1.4144349098205566,
      "learning_rate": 4.111286308303605e-06,
      "loss": 0.1799,
      "num_input_tokens_seen": 28229184,
      "step": 32595
    },
    {
      "epoch": 15.37010843941537,
      "grad_norm": 0.9449219107627869,
      "learning_rate": 4.105894034020433e-06,
      "loss": 0.2009,
      "num_input_tokens_seen": 28233360,
      "step": 32600
    },
    {
      "epoch": 15.37010843941537,
      "eval_loss": 0.19982516765594482,
      "eval_runtime": 22.1786,
      "eval_samples_per_second": 42.519,
      "eval_steps_per_second": 21.282,
      "num_input_tokens_seen": 28233360,
      "step": 32600
    },
    {
      "epoch": 15.372465818010372,
      "grad_norm": 0.6796343326568604,
      "learning_rate": 4.100504981870975e-06,
      "loss": 0.1532,
      "num_input_tokens_seen": 28237472,
      "step": 32605
    },
    {
      "epoch": 15.374823196605375,
      "grad_norm": 1.3990015983581543,
      "learning_rate": 4.0951191526862915e-06,
      "loss": 0.1286,
      "num_input_tokens_seen": 28242368,
      "step": 32610
    },
    {
      "epoch": 15.377180575200377,
      "grad_norm": 0.6452330946922302,
      "learning_rate": 4.089736547296938e-06,
      "loss": 0.2144,
      "num_input_tokens_seen": 28247520,
      "step": 32615
    },
    {
      "epoch": 15.37953795379538,
      "grad_norm": 0.8788579106330872,
      "learning_rate": 4.08435716653299e-06,
      "loss": 0.1738,
      "num_input_tokens_seen": 28251488,
      "step": 32620
    },
    {
      "epoch": 15.381895332390382,
      "grad_norm": 0.8631284236907959,
      "learning_rate": 4.0789810112240005e-06,
      "loss": 0.1794,
      "num_input_tokens_seen": 28256240,
      "step": 32625
    },
    {
      "epoch": 15.384252710985384,
      "grad_norm": 0.6192811727523804,
      "learning_rate": 4.073608082199057e-06,
      "loss": 0.2851,
      "num_input_tokens_seen": 28260000,
      "step": 32630
    },
    {
      "epoch": 15.386610089580387,
      "grad_norm": 1.7514396905899048,
      "learning_rate": 4.068238380286718e-06,
      "loss": 0.1657,
      "num_input_tokens_seen": 28264800,
      "step": 32635
    },
    {
      "epoch": 15.38896746817539,
      "grad_norm": 0.8380995988845825,
      "learning_rate": 4.062871906315072e-06,
      "loss": 0.1637,
      "num_input_tokens_seen": 28269808,
      "step": 32640
    },
    {
      "epoch": 15.391324846770392,
      "grad_norm": 0.9094882607460022,
      "learning_rate": 4.057508661111686e-06,
      "loss": 0.1307,
      "num_input_tokens_seen": 28273936,
      "step": 32645
    },
    {
      "epoch": 15.393682225365394,
      "grad_norm": 1.3954921960830688,
      "learning_rate": 4.052148645503648e-06,
      "loss": 0.1703,
      "num_input_tokens_seen": 28278048,
      "step": 32650
    },
    {
      "epoch": 15.396039603960396,
      "grad_norm": 2.3167033195495605,
      "learning_rate": 4.046791860317531e-06,
      "loss": 0.1936,
      "num_input_tokens_seen": 28282160,
      "step": 32655
    },
    {
      "epoch": 15.398396982555399,
      "grad_norm": 0.7670348882675171,
      "learning_rate": 4.041438306379431e-06,
      "loss": 0.1911,
      "num_input_tokens_seen": 28286256,
      "step": 32660
    },
    {
      "epoch": 15.400754361150401,
      "grad_norm": 1.2249330282211304,
      "learning_rate": 4.036087984514916e-06,
      "loss": 0.2459,
      "num_input_tokens_seen": 28290256,
      "step": 32665
    },
    {
      "epoch": 15.403111739745404,
      "grad_norm": 2.1646153926849365,
      "learning_rate": 4.030740895549084e-06,
      "loss": 0.2488,
      "num_input_tokens_seen": 28295280,
      "step": 32670
    },
    {
      "epoch": 15.405469118340406,
      "grad_norm": 0.5984014272689819,
      "learning_rate": 4.025397040306531e-06,
      "loss": 0.1988,
      "num_input_tokens_seen": 28299328,
      "step": 32675
    },
    {
      "epoch": 15.407826496935408,
      "grad_norm": 1.5827633142471313,
      "learning_rate": 4.0200564196113285e-06,
      "loss": 0.1437,
      "num_input_tokens_seen": 28303296,
      "step": 32680
    },
    {
      "epoch": 15.41018387553041,
      "grad_norm": 0.5862312912940979,
      "learning_rate": 4.014719034287079e-06,
      "loss": 0.1501,
      "num_input_tokens_seen": 28307712,
      "step": 32685
    },
    {
      "epoch": 15.412541254125413,
      "grad_norm": 1.0607789754867554,
      "learning_rate": 4.0093848851568775e-06,
      "loss": 0.1274,
      "num_input_tokens_seen": 28311696,
      "step": 32690
    },
    {
      "epoch": 15.414898632720416,
      "grad_norm": 0.6455942988395691,
      "learning_rate": 4.004053973043304e-06,
      "loss": 0.1152,
      "num_input_tokens_seen": 28316256,
      "step": 32695
    },
    {
      "epoch": 15.417256011315418,
      "grad_norm": 2.7964460849761963,
      "learning_rate": 3.998726298768465e-06,
      "loss": 0.2212,
      "num_input_tokens_seen": 28320960,
      "step": 32700
    },
    {
      "epoch": 15.41961338991042,
      "grad_norm": 0.4236297905445099,
      "learning_rate": 3.99340186315395e-06,
      "loss": 0.1701,
      "num_input_tokens_seen": 28325232,
      "step": 32705
    },
    {
      "epoch": 15.421970768505423,
      "grad_norm": 0.6869227290153503,
      "learning_rate": 3.988080667020849e-06,
      "loss": 0.1655,
      "num_input_tokens_seen": 28329472,
      "step": 32710
    },
    {
      "epoch": 15.424328147100425,
      "grad_norm": 0.5219728946685791,
      "learning_rate": 3.982762711189766e-06,
      "loss": 0.2529,
      "num_input_tokens_seen": 28333232,
      "step": 32715
    },
    {
      "epoch": 15.426685525695426,
      "grad_norm": 1.4745486974716187,
      "learning_rate": 3.977447996480785e-06,
      "loss": 0.1731,
      "num_input_tokens_seen": 28337312,
      "step": 32720
    },
    {
      "epoch": 15.429042904290428,
      "grad_norm": 1.6292425394058228,
      "learning_rate": 3.97213652371351e-06,
      "loss": 0.2661,
      "num_input_tokens_seen": 28342336,
      "step": 32725
    },
    {
      "epoch": 15.43140028288543,
      "grad_norm": 1.0913313627243042,
      "learning_rate": 3.966828293707042e-06,
      "loss": 0.1542,
      "num_input_tokens_seen": 28347264,
      "step": 32730
    },
    {
      "epoch": 15.433757661480433,
      "grad_norm": 1.0507129430770874,
      "learning_rate": 3.961523307279963e-06,
      "loss": 0.1967,
      "num_input_tokens_seen": 28350816,
      "step": 32735
    },
    {
      "epoch": 15.436115040075435,
      "grad_norm": 1.7249209880828857,
      "learning_rate": 3.956221565250382e-06,
      "loss": 0.1832,
      "num_input_tokens_seen": 28355712,
      "step": 32740
    },
    {
      "epoch": 15.438472418670438,
      "grad_norm": 0.6994402408599854,
      "learning_rate": 3.950923068435883e-06,
      "loss": 0.1854,
      "num_input_tokens_seen": 28359856,
      "step": 32745
    },
    {
      "epoch": 15.44082979726544,
      "grad_norm": 1.3772773742675781,
      "learning_rate": 3.945627817653566e-06,
      "loss": 0.1958,
      "num_input_tokens_seen": 28364464,
      "step": 32750
    },
    {
      "epoch": 15.443187175860443,
      "grad_norm": 0.5447023510932922,
      "learning_rate": 3.9403358137200335e-06,
      "loss": 0.1781,
      "num_input_tokens_seen": 28368704,
      "step": 32755
    },
    {
      "epoch": 15.445544554455445,
      "grad_norm": 1.7269396781921387,
      "learning_rate": 3.9350470574513605e-06,
      "loss": 0.187,
      "num_input_tokens_seen": 28374096,
      "step": 32760
    },
    {
      "epoch": 15.447901933050447,
      "grad_norm": 1.0192173719406128,
      "learning_rate": 3.9297615496631525e-06,
      "loss": 0.1585,
      "num_input_tokens_seen": 28380000,
      "step": 32765
    },
    {
      "epoch": 15.45025931164545,
      "grad_norm": 0.295314222574234,
      "learning_rate": 3.924479291170505e-06,
      "loss": 0.1978,
      "num_input_tokens_seen": 28383792,
      "step": 32770
    },
    {
      "epoch": 15.452616690240452,
      "grad_norm": 1.313003659248352,
      "learning_rate": 3.919200282788002e-06,
      "loss": 0.2345,
      "num_input_tokens_seen": 28388000,
      "step": 32775
    },
    {
      "epoch": 15.454974068835455,
      "grad_norm": 1.5729750394821167,
      "learning_rate": 3.913924525329726e-06,
      "loss": 0.1982,
      "num_input_tokens_seen": 28392400,
      "step": 32780
    },
    {
      "epoch": 15.457331447430457,
      "grad_norm": 2.290383815765381,
      "learning_rate": 3.908652019609279e-06,
      "loss": 0.1971,
      "num_input_tokens_seen": 28396736,
      "step": 32785
    },
    {
      "epoch": 15.45968882602546,
      "grad_norm": 1.6472351551055908,
      "learning_rate": 3.9033827664397364e-06,
      "loss": 0.1414,
      "num_input_tokens_seen": 28401616,
      "step": 32790
    },
    {
      "epoch": 15.462046204620462,
      "grad_norm": 0.6778950095176697,
      "learning_rate": 3.898116766633694e-06,
      "loss": 0.1832,
      "num_input_tokens_seen": 28405552,
      "step": 32795
    },
    {
      "epoch": 15.464403583215464,
      "grad_norm": 1.4663662910461426,
      "learning_rate": 3.8928540210032225e-06,
      "loss": 0.1572,
      "num_input_tokens_seen": 28411200,
      "step": 32800
    },
    {
      "epoch": 15.464403583215464,
      "eval_loss": 0.19959332048892975,
      "eval_runtime": 22.201,
      "eval_samples_per_second": 42.476,
      "eval_steps_per_second": 21.26,
      "num_input_tokens_seen": 28411200,
      "step": 32800
    },
    {
      "epoch": 15.466760961810467,
      "grad_norm": 2.537869453430176,
      "learning_rate": 3.887594530359909e-06,
      "loss": 0.2429,
      "num_input_tokens_seen": 28414832,
      "step": 32805
    },
    {
      "epoch": 15.469118340405469,
      "grad_norm": 1.4069490432739258,
      "learning_rate": 3.88233829551484e-06,
      "loss": 0.2201,
      "num_input_tokens_seen": 28419440,
      "step": 32810
    },
    {
      "epoch": 15.471475719000471,
      "grad_norm": 2.2554824352264404,
      "learning_rate": 3.877085317278581e-06,
      "loss": 0.2442,
      "num_input_tokens_seen": 28423136,
      "step": 32815
    },
    {
      "epoch": 15.473833097595474,
      "grad_norm": 1.0455750226974487,
      "learning_rate": 3.87183559646122e-06,
      "loss": 0.1988,
      "num_input_tokens_seen": 28427616,
      "step": 32820
    },
    {
      "epoch": 15.476190476190476,
      "grad_norm": 1.2633610963821411,
      "learning_rate": 3.866589133872317e-06,
      "loss": 0.158,
      "num_input_tokens_seen": 28432768,
      "step": 32825
    },
    {
      "epoch": 15.478547854785479,
      "grad_norm": 1.0072247982025146,
      "learning_rate": 3.861345930320948e-06,
      "loss": 0.1381,
      "num_input_tokens_seen": 28437488,
      "step": 32830
    },
    {
      "epoch": 15.480905233380481,
      "grad_norm": 1.6848889589309692,
      "learning_rate": 3.856105986615688e-06,
      "loss": 0.1723,
      "num_input_tokens_seen": 28442176,
      "step": 32835
    },
    {
      "epoch": 15.483262611975483,
      "grad_norm": 1.1957882642745972,
      "learning_rate": 3.850869303564589e-06,
      "loss": 0.2184,
      "num_input_tokens_seen": 28446512,
      "step": 32840
    },
    {
      "epoch": 15.485619990570486,
      "grad_norm": 1.6522496938705444,
      "learning_rate": 3.845635881975226e-06,
      "loss": 0.1825,
      "num_input_tokens_seen": 28450288,
      "step": 32845
    },
    {
      "epoch": 15.487977369165488,
      "grad_norm": 0.7985666394233704,
      "learning_rate": 3.840405722654647e-06,
      "loss": 0.1056,
      "num_input_tokens_seen": 28454208,
      "step": 32850
    },
    {
      "epoch": 15.49033474776049,
      "grad_norm": 0.7478682398796082,
      "learning_rate": 3.835178826409419e-06,
      "loss": 0.218,
      "num_input_tokens_seen": 28458736,
      "step": 32855
    },
    {
      "epoch": 15.492692126355493,
      "grad_norm": 1.08069908618927,
      "learning_rate": 3.8299551940455895e-06,
      "loss": 0.2095,
      "num_input_tokens_seen": 28462528,
      "step": 32860
    },
    {
      "epoch": 15.495049504950495,
      "grad_norm": 0.9625383615493774,
      "learning_rate": 3.824734826368703e-06,
      "loss": 0.237,
      "num_input_tokens_seen": 28467744,
      "step": 32865
    },
    {
      "epoch": 15.497406883545498,
      "grad_norm": 0.9415424466133118,
      "learning_rate": 3.819517724183813e-06,
      "loss": 0.1549,
      "num_input_tokens_seen": 28471824,
      "step": 32870
    },
    {
      "epoch": 15.4997642621405,
      "grad_norm": 0.41289976239204407,
      "learning_rate": 3.8143038882954648e-06,
      "loss": 0.1384,
      "num_input_tokens_seen": 28476288,
      "step": 32875
    },
    {
      "epoch": 15.502121640735503,
      "grad_norm": 1.3852834701538086,
      "learning_rate": 3.8090933195076867e-06,
      "loss": 0.1649,
      "num_input_tokens_seen": 28481152,
      "step": 32880
    },
    {
      "epoch": 15.504479019330505,
      "grad_norm": 0.9029893279075623,
      "learning_rate": 3.8038860186240198e-06,
      "loss": 0.1766,
      "num_input_tokens_seen": 28485920,
      "step": 32885
    },
    {
      "epoch": 15.506836397925507,
      "grad_norm": 1.0521725416183472,
      "learning_rate": 3.7986819864475026e-06,
      "loss": 0.2114,
      "num_input_tokens_seen": 28490096,
      "step": 32890
    },
    {
      "epoch": 15.50919377652051,
      "grad_norm": 1.361986517906189,
      "learning_rate": 3.793481223780651e-06,
      "loss": 0.2113,
      "num_input_tokens_seen": 28493312,
      "step": 32895
    },
    {
      "epoch": 15.511551155115512,
      "grad_norm": 1.010658860206604,
      "learning_rate": 3.788283731425496e-06,
      "loss": 0.2348,
      "num_input_tokens_seen": 28497328,
      "step": 32900
    },
    {
      "epoch": 15.513908533710515,
      "grad_norm": 0.9186533689498901,
      "learning_rate": 3.7830895101835488e-06,
      "loss": 0.1428,
      "num_input_tokens_seen": 28501648,
      "step": 32905
    },
    {
      "epoch": 15.516265912305517,
      "grad_norm": 1.7808688879013062,
      "learning_rate": 3.7778985608558274e-06,
      "loss": 0.1387,
      "num_input_tokens_seen": 28505808,
      "step": 32910
    },
    {
      "epoch": 15.518623290900518,
      "grad_norm": 1.717864751815796,
      "learning_rate": 3.7727108842428443e-06,
      "loss": 0.1619,
      "num_input_tokens_seen": 28510592,
      "step": 32915
    },
    {
      "epoch": 15.520980669495522,
      "grad_norm": 0.5536420345306396,
      "learning_rate": 3.7675264811446065e-06,
      "loss": 0.1806,
      "num_input_tokens_seen": 28514336,
      "step": 32920
    },
    {
      "epoch": 15.523338048090523,
      "grad_norm": 1.0803191661834717,
      "learning_rate": 3.7623453523605994e-06,
      "loss": 0.1453,
      "num_input_tokens_seen": 28518512,
      "step": 32925
    },
    {
      "epoch": 15.525695426685525,
      "grad_norm": 0.819144070148468,
      "learning_rate": 3.757167498689834e-06,
      "loss": 0.2011,
      "num_input_tokens_seen": 28522544,
      "step": 32930
    },
    {
      "epoch": 15.528052805280527,
      "grad_norm": 0.7963111400604248,
      "learning_rate": 3.7519929209307914e-06,
      "loss": 0.1576,
      "num_input_tokens_seen": 28526352,
      "step": 32935
    },
    {
      "epoch": 15.53041018387553,
      "grad_norm": 1.0806299448013306,
      "learning_rate": 3.746821619881463e-06,
      "loss": 0.2261,
      "num_input_tokens_seen": 28531200,
      "step": 32940
    },
    {
      "epoch": 15.532767562470532,
      "grad_norm": 0.7730188369750977,
      "learning_rate": 3.74165359633932e-06,
      "loss": 0.1688,
      "num_input_tokens_seen": 28535616,
      "step": 32945
    },
    {
      "epoch": 15.535124941065535,
      "grad_norm": 0.8365828990936279,
      "learning_rate": 3.736488851101341e-06,
      "loss": 0.2,
      "num_input_tokens_seen": 28540560,
      "step": 32950
    },
    {
      "epoch": 15.537482319660537,
      "grad_norm": 1.041605830192566,
      "learning_rate": 3.7313273849640035e-06,
      "loss": 0.2464,
      "num_input_tokens_seen": 28544432,
      "step": 32955
    },
    {
      "epoch": 15.53983969825554,
      "grad_norm": 1.8104106187820435,
      "learning_rate": 3.7261691987232533e-06,
      "loss": 0.174,
      "num_input_tokens_seen": 28548528,
      "step": 32960
    },
    {
      "epoch": 15.542197076850542,
      "grad_norm": 1.305902361869812,
      "learning_rate": 3.7210142931745575e-06,
      "loss": 0.1552,
      "num_input_tokens_seen": 28552448,
      "step": 32965
    },
    {
      "epoch": 15.544554455445544,
      "grad_norm": 0.9506945610046387,
      "learning_rate": 3.7158626691128712e-06,
      "loss": 0.189,
      "num_input_tokens_seen": 28557136,
      "step": 32970
    },
    {
      "epoch": 15.546911834040547,
      "grad_norm": 0.9585821628570557,
      "learning_rate": 3.710714327332629e-06,
      "loss": 0.1658,
      "num_input_tokens_seen": 28561424,
      "step": 32975
    },
    {
      "epoch": 15.549269212635549,
      "grad_norm": 2.0649421215057373,
      "learning_rate": 3.7055692686277815e-06,
      "loss": 0.2584,
      "num_input_tokens_seen": 28565792,
      "step": 32980
    },
    {
      "epoch": 15.551626591230551,
      "grad_norm": 1.0567506551742554,
      "learning_rate": 3.70042749379175e-06,
      "loss": 0.1035,
      "num_input_tokens_seen": 28569376,
      "step": 32985
    },
    {
      "epoch": 15.553983969825554,
      "grad_norm": 1.0727719068527222,
      "learning_rate": 3.6952890036174693e-06,
      "loss": 0.155,
      "num_input_tokens_seen": 28573824,
      "step": 32990
    },
    {
      "epoch": 15.556341348420556,
      "grad_norm": 1.617243766784668,
      "learning_rate": 3.690153798897353e-06,
      "loss": 0.1861,
      "num_input_tokens_seen": 28577872,
      "step": 32995
    },
    {
      "epoch": 15.558698727015559,
      "grad_norm": 0.6744821667671204,
      "learning_rate": 3.6850218804233225e-06,
      "loss": 0.144,
      "num_input_tokens_seen": 28582944,
      "step": 33000
    },
    {
      "epoch": 15.558698727015559,
      "eval_loss": 0.19967150688171387,
      "eval_runtime": 22.2148,
      "eval_samples_per_second": 42.449,
      "eval_steps_per_second": 21.247,
      "num_input_tokens_seen": 28582944,
      "step": 33000
    },
    {
      "epoch": 15.561056105610561,
      "grad_norm": 0.9054422974586487,
      "learning_rate": 3.679893248986779e-06,
      "loss": 0.1497,
      "num_input_tokens_seen": 28587136,
      "step": 33005
    },
    {
      "epoch": 15.563413484205563,
      "grad_norm": 0.5440496206283569,
      "learning_rate": 3.6747679053786147e-06,
      "loss": 0.1776,
      "num_input_tokens_seen": 28591248,
      "step": 33010
    },
    {
      "epoch": 15.565770862800566,
      "grad_norm": 0.5144120454788208,
      "learning_rate": 3.669645850389228e-06,
      "loss": 0.2337,
      "num_input_tokens_seen": 28595344,
      "step": 33015
    },
    {
      "epoch": 15.568128241395568,
      "grad_norm": 1.6629916429519653,
      "learning_rate": 3.664527084808514e-06,
      "loss": 0.161,
      "num_input_tokens_seen": 28600016,
      "step": 33020
    },
    {
      "epoch": 15.57048561999057,
      "grad_norm": 1.2974016666412354,
      "learning_rate": 3.6594116094258337e-06,
      "loss": 0.1584,
      "num_input_tokens_seen": 28603984,
      "step": 33025
    },
    {
      "epoch": 15.572842998585573,
      "grad_norm": 0.9513036608695984,
      "learning_rate": 3.6542994250300665e-06,
      "loss": 0.1782,
      "num_input_tokens_seen": 28608672,
      "step": 33030
    },
    {
      "epoch": 15.575200377180575,
      "grad_norm": 0.5009998679161072,
      "learning_rate": 3.6491905324095825e-06,
      "loss": 0.1895,
      "num_input_tokens_seen": 28613008,
      "step": 33035
    },
    {
      "epoch": 15.577557755775578,
      "grad_norm": 1.3189828395843506,
      "learning_rate": 3.644084932352221e-06,
      "loss": 0.1724,
      "num_input_tokens_seen": 28617552,
      "step": 33040
    },
    {
      "epoch": 15.57991513437058,
      "grad_norm": 1.3458466529846191,
      "learning_rate": 3.6389826256453457e-06,
      "loss": 0.1207,
      "num_input_tokens_seen": 28621200,
      "step": 33045
    },
    {
      "epoch": 15.582272512965583,
      "grad_norm": 1.256820559501648,
      "learning_rate": 3.633883613075781e-06,
      "loss": 0.2421,
      "num_input_tokens_seen": 28625152,
      "step": 33050
    },
    {
      "epoch": 15.584629891560585,
      "grad_norm": 0.4913026690483093,
      "learning_rate": 3.6287878954298693e-06,
      "loss": 0.1387,
      "num_input_tokens_seen": 28629168,
      "step": 33055
    },
    {
      "epoch": 15.586987270155587,
      "grad_norm": 1.1744024753570557,
      "learning_rate": 3.6236954734934354e-06,
      "loss": 0.1862,
      "num_input_tokens_seen": 28633808,
      "step": 33060
    },
    {
      "epoch": 15.58934464875059,
      "grad_norm": 0.9174792766571045,
      "learning_rate": 3.618606348051784e-06,
      "loss": 0.2211,
      "num_input_tokens_seen": 28637760,
      "step": 33065
    },
    {
      "epoch": 15.591702027345592,
      "grad_norm": 1.0505127906799316,
      "learning_rate": 3.6135205198897376e-06,
      "loss": 0.1775,
      "num_input_tokens_seen": 28642960,
      "step": 33070
    },
    {
      "epoch": 15.594059405940595,
      "grad_norm": 0.6301836967468262,
      "learning_rate": 3.6084379897915854e-06,
      "loss": 0.1431,
      "num_input_tokens_seen": 28646848,
      "step": 33075
    },
    {
      "epoch": 15.596416784535597,
      "grad_norm": 1.6244467496871948,
      "learning_rate": 3.6033587585411115e-06,
      "loss": 0.1584,
      "num_input_tokens_seen": 28651952,
      "step": 33080
    },
    {
      "epoch": 15.5987741631306,
      "grad_norm": 0.4422319233417511,
      "learning_rate": 3.5982828269216117e-06,
      "loss": 0.2027,
      "num_input_tokens_seen": 28656064,
      "step": 33085
    },
    {
      "epoch": 15.601131541725602,
      "grad_norm": 0.5385611057281494,
      "learning_rate": 3.593210195715843e-06,
      "loss": 0.1821,
      "num_input_tokens_seen": 28660832,
      "step": 33090
    },
    {
      "epoch": 15.603488920320604,
      "grad_norm": 2.2903780937194824,
      "learning_rate": 3.5881408657060773e-06,
      "loss": 0.1759,
      "num_input_tokens_seen": 28665184,
      "step": 33095
    },
    {
      "epoch": 15.605846298915607,
      "grad_norm": 1.7606574296951294,
      "learning_rate": 3.583074837674075e-06,
      "loss": 0.1503,
      "num_input_tokens_seen": 28668816,
      "step": 33100
    },
    {
      "epoch": 15.608203677510609,
      "grad_norm": 0.7179262042045593,
      "learning_rate": 3.578012112401069e-06,
      "loss": 0.1283,
      "num_input_tokens_seen": 28673296,
      "step": 33105
    },
    {
      "epoch": 15.61056105610561,
      "grad_norm": 0.5467319488525391,
      "learning_rate": 3.5729526906677996e-06,
      "loss": 0.1673,
      "num_input_tokens_seen": 28677376,
      "step": 33110
    },
    {
      "epoch": 15.612918434700614,
      "grad_norm": 0.4120520353317261,
      "learning_rate": 3.5678965732545007e-06,
      "loss": 0.1617,
      "num_input_tokens_seen": 28681648,
      "step": 33115
    },
    {
      "epoch": 15.615275813295614,
      "grad_norm": 0.7780698537826538,
      "learning_rate": 3.562843760940876e-06,
      "loss": 0.1698,
      "num_input_tokens_seen": 28686144,
      "step": 33120
    },
    {
      "epoch": 15.617633191890617,
      "grad_norm": 0.6548003554344177,
      "learning_rate": 3.5577942545061473e-06,
      "loss": 0.1387,
      "num_input_tokens_seen": 28690688,
      "step": 33125
    },
    {
      "epoch": 15.61999057048562,
      "grad_norm": 1.6145857572555542,
      "learning_rate": 3.5527480547289967e-06,
      "loss": 0.1821,
      "num_input_tokens_seen": 28695232,
      "step": 33130
    },
    {
      "epoch": 15.622347949080622,
      "grad_norm": 1.4773184061050415,
      "learning_rate": 3.547705162387624e-06,
      "loss": 0.2351,
      "num_input_tokens_seen": 28699584,
      "step": 33135
    },
    {
      "epoch": 15.624705327675624,
      "grad_norm": 1.0703723430633545,
      "learning_rate": 3.542665578259699e-06,
      "loss": 0.1769,
      "num_input_tokens_seen": 28704720,
      "step": 33140
    },
    {
      "epoch": 15.627062706270626,
      "grad_norm": 0.6979261040687561,
      "learning_rate": 3.5376293031223945e-06,
      "loss": 0.1975,
      "num_input_tokens_seen": 28708528,
      "step": 33145
    },
    {
      "epoch": 15.629420084865629,
      "grad_norm": 1.425107479095459,
      "learning_rate": 3.5325963377523614e-06,
      "loss": 0.1705,
      "num_input_tokens_seen": 28713104,
      "step": 33150
    },
    {
      "epoch": 15.631777463460631,
      "grad_norm": 2.403745412826538,
      "learning_rate": 3.5275666829257536e-06,
      "loss": 0.2193,
      "num_input_tokens_seen": 28718208,
      "step": 33155
    },
    {
      "epoch": 15.634134842055634,
      "grad_norm": 1.2311680316925049,
      "learning_rate": 3.5225403394181955e-06,
      "loss": 0.1994,
      "num_input_tokens_seen": 28722096,
      "step": 33160
    },
    {
      "epoch": 15.636492220650636,
      "grad_norm": 0.6244481205940247,
      "learning_rate": 3.517517308004828e-06,
      "loss": 0.1364,
      "num_input_tokens_seen": 28726368,
      "step": 33165
    },
    {
      "epoch": 15.638849599245638,
      "grad_norm": 0.8425841331481934,
      "learning_rate": 3.512497589460251e-06,
      "loss": 0.1562,
      "num_input_tokens_seen": 28731856,
      "step": 33170
    },
    {
      "epoch": 15.64120697784064,
      "grad_norm": 1.7469443082809448,
      "learning_rate": 3.5074811845585727e-06,
      "loss": 0.1766,
      "num_input_tokens_seen": 28736592,
      "step": 33175
    },
    {
      "epoch": 15.643564356435643,
      "grad_norm": 0.7708004117012024,
      "learning_rate": 3.5024680940733937e-06,
      "loss": 0.1398,
      "num_input_tokens_seen": 28740144,
      "step": 33180
    },
    {
      "epoch": 15.645921735030646,
      "grad_norm": 0.9459390044212341,
      "learning_rate": 3.4974583187777852e-06,
      "loss": 0.1005,
      "num_input_tokens_seen": 28744080,
      "step": 33185
    },
    {
      "epoch": 15.648279113625648,
      "grad_norm": 0.7686309218406677,
      "learning_rate": 3.4924518594443204e-06,
      "loss": 0.181,
      "num_input_tokens_seen": 28748320,
      "step": 33190
    },
    {
      "epoch": 15.65063649222065,
      "grad_norm": 0.7872373461723328,
      "learning_rate": 3.4874487168450682e-06,
      "loss": 0.1584,
      "num_input_tokens_seen": 28752832,
      "step": 33195
    },
    {
      "epoch": 15.652993870815653,
      "grad_norm": 0.8146344423294067,
      "learning_rate": 3.482448891751558e-06,
      "loss": 0.2015,
      "num_input_tokens_seen": 28756240,
      "step": 33200
    },
    {
      "epoch": 15.652993870815653,
      "eval_loss": 0.2004423439502716,
      "eval_runtime": 22.207,
      "eval_samples_per_second": 42.464,
      "eval_steps_per_second": 21.255,
      "num_input_tokens_seen": 28756240,
      "step": 33200
    },
    {
      "epoch": 15.655351249410655,
      "grad_norm": 0.5398359298706055,
      "learning_rate": 3.477452384934843e-06,
      "loss": 0.143,
      "num_input_tokens_seen": 28760752,
      "step": 33205
    },
    {
      "epoch": 15.657708628005658,
      "grad_norm": 1.142638921737671,
      "learning_rate": 3.472459197165434e-06,
      "loss": 0.1588,
      "num_input_tokens_seen": 28764896,
      "step": 33210
    },
    {
      "epoch": 15.66006600660066,
      "grad_norm": 1.1105577945709229,
      "learning_rate": 3.4674693292133518e-06,
      "loss": 0.1825,
      "num_input_tokens_seen": 28769168,
      "step": 33215
    },
    {
      "epoch": 15.662423385195662,
      "grad_norm": 0.6094043254852295,
      "learning_rate": 3.4624827818480977e-06,
      "loss": 0.1917,
      "num_input_tokens_seen": 28772704,
      "step": 33220
    },
    {
      "epoch": 15.664780763790665,
      "grad_norm": 1.1574209928512573,
      "learning_rate": 3.4574995558386474e-06,
      "loss": 0.1498,
      "num_input_tokens_seen": 28776736,
      "step": 33225
    },
    {
      "epoch": 15.667138142385667,
      "grad_norm": 1.110676884651184,
      "learning_rate": 3.452519651953487e-06,
      "loss": 0.1236,
      "num_input_tokens_seen": 28780704,
      "step": 33230
    },
    {
      "epoch": 15.66949552098067,
      "grad_norm": 0.7819063663482666,
      "learning_rate": 3.447543070960585e-06,
      "loss": 0.1941,
      "num_input_tokens_seen": 28784576,
      "step": 33235
    },
    {
      "epoch": 15.671852899575672,
      "grad_norm": 1.1922380924224854,
      "learning_rate": 3.4425698136273778e-06,
      "loss": 0.1539,
      "num_input_tokens_seen": 28789024,
      "step": 33240
    },
    {
      "epoch": 15.674210278170674,
      "grad_norm": 1.079689383506775,
      "learning_rate": 3.437599880720821e-06,
      "loss": 0.1867,
      "num_input_tokens_seen": 28793648,
      "step": 33245
    },
    {
      "epoch": 15.676567656765677,
      "grad_norm": 1.116231918334961,
      "learning_rate": 3.4326332730073267e-06,
      "loss": 0.1693,
      "num_input_tokens_seen": 28798048,
      "step": 33250
    },
    {
      "epoch": 15.67892503536068,
      "grad_norm": 0.5626642107963562,
      "learning_rate": 3.427669991252813e-06,
      "loss": 0.1311,
      "num_input_tokens_seen": 28802128,
      "step": 33255
    },
    {
      "epoch": 15.681282413955682,
      "grad_norm": 0.8573475480079651,
      "learning_rate": 3.42271003622269e-06,
      "loss": 0.2415,
      "num_input_tokens_seen": 28806688,
      "step": 33260
    },
    {
      "epoch": 15.683639792550684,
      "grad_norm": 0.8027072548866272,
      "learning_rate": 3.4177534086818286e-06,
      "loss": 0.1898,
      "num_input_tokens_seen": 28811072,
      "step": 33265
    },
    {
      "epoch": 15.685997171145686,
      "grad_norm": 0.9344469308853149,
      "learning_rate": 3.412800109394612e-06,
      "loss": 0.1699,
      "num_input_tokens_seen": 28815520,
      "step": 33270
    },
    {
      "epoch": 15.688354549740689,
      "grad_norm": 0.805840253829956,
      "learning_rate": 3.4078501391249044e-06,
      "loss": 0.173,
      "num_input_tokens_seen": 28820880,
      "step": 33275
    },
    {
      "epoch": 15.690711928335691,
      "grad_norm": 1.3820695877075195,
      "learning_rate": 3.4029034986360453e-06,
      "loss": 0.182,
      "num_input_tokens_seen": 28825440,
      "step": 33280
    },
    {
      "epoch": 15.693069306930694,
      "grad_norm": 1.374865174293518,
      "learning_rate": 3.397960188690877e-06,
      "loss": 0.1428,
      "num_input_tokens_seen": 28829584,
      "step": 33285
    },
    {
      "epoch": 15.695426685525696,
      "grad_norm": 1.5967825651168823,
      "learning_rate": 3.393020210051717e-06,
      "loss": 0.1991,
      "num_input_tokens_seen": 28833712,
      "step": 33290
    },
    {
      "epoch": 15.697784064120698,
      "grad_norm": 0.4856487214565277,
      "learning_rate": 3.3880835634803655e-06,
      "loss": 0.1462,
      "num_input_tokens_seen": 28837920,
      "step": 33295
    },
    {
      "epoch": 15.700141442715701,
      "grad_norm": 0.8802598714828491,
      "learning_rate": 3.383150249738126e-06,
      "loss": 0.1803,
      "num_input_tokens_seen": 28842176,
      "step": 33300
    },
    {
      "epoch": 15.702498821310703,
      "grad_norm": 0.8564335703849792,
      "learning_rate": 3.3782202695857663e-06,
      "loss": 0.1837,
      "num_input_tokens_seen": 28846192,
      "step": 33305
    },
    {
      "epoch": 15.704856199905706,
      "grad_norm": 0.7652109861373901,
      "learning_rate": 3.373293623783558e-06,
      "loss": 0.185,
      "num_input_tokens_seen": 28851712,
      "step": 33310
    },
    {
      "epoch": 15.707213578500706,
      "grad_norm": 0.6168251633644104,
      "learning_rate": 3.368370313091257e-06,
      "loss": 0.1302,
      "num_input_tokens_seen": 28855472,
      "step": 33315
    },
    {
      "epoch": 15.70957095709571,
      "grad_norm": 1.9617300033569336,
      "learning_rate": 3.363450338268087e-06,
      "loss": 0.2056,
      "num_input_tokens_seen": 28859088,
      "step": 33320
    },
    {
      "epoch": 15.711928335690711,
      "grad_norm": 1.483379602432251,
      "learning_rate": 3.358533700072783e-06,
      "loss": 0.1913,
      "num_input_tokens_seen": 28863456,
      "step": 33325
    },
    {
      "epoch": 15.714285714285714,
      "grad_norm": 0.8629631400108337,
      "learning_rate": 3.3536203992635377e-06,
      "loss": 0.1365,
      "num_input_tokens_seen": 28867360,
      "step": 33330
    },
    {
      "epoch": 15.716643092880716,
      "grad_norm": 0.7022390365600586,
      "learning_rate": 3.348710436598057e-06,
      "loss": 0.1649,
      "num_input_tokens_seen": 28872000,
      "step": 33335
    },
    {
      "epoch": 15.719000471475718,
      "grad_norm": 1.1379904747009277,
      "learning_rate": 3.3438038128335155e-06,
      "loss": 0.1584,
      "num_input_tokens_seen": 28876192,
      "step": 33340
    },
    {
      "epoch": 15.72135785007072,
      "grad_norm": 0.8153406381607056,
      "learning_rate": 3.338900528726571e-06,
      "loss": 0.0952,
      "num_input_tokens_seen": 28879824,
      "step": 33345
    },
    {
      "epoch": 15.723715228665723,
      "grad_norm": 1.3695236444473267,
      "learning_rate": 3.3340005850333812e-06,
      "loss": 0.1224,
      "num_input_tokens_seen": 28883776,
      "step": 33350
    },
    {
      "epoch": 15.726072607260726,
      "grad_norm": 0.657973051071167,
      "learning_rate": 3.329103982509568e-06,
      "loss": 0.0924,
      "num_input_tokens_seen": 28887904,
      "step": 33355
    },
    {
      "epoch": 15.728429985855728,
      "grad_norm": 1.7679401636123657,
      "learning_rate": 3.324210721910259e-06,
      "loss": 0.1669,
      "num_input_tokens_seen": 28892256,
      "step": 33360
    },
    {
      "epoch": 15.73078736445073,
      "grad_norm": 0.9039352536201477,
      "learning_rate": 3.319320803990053e-06,
      "loss": 0.1685,
      "num_input_tokens_seen": 28897184,
      "step": 33365
    },
    {
      "epoch": 15.733144743045733,
      "grad_norm": 0.8745309114456177,
      "learning_rate": 3.3144342295030274e-06,
      "loss": 0.2284,
      "num_input_tokens_seen": 28901040,
      "step": 33370
    },
    {
      "epoch": 15.735502121640735,
      "grad_norm": 0.3760108947753906,
      "learning_rate": 3.309550999202765e-06,
      "loss": 0.138,
      "num_input_tokens_seen": 28905600,
      "step": 33375
    },
    {
      "epoch": 15.737859500235738,
      "grad_norm": 1.4352413415908813,
      "learning_rate": 3.3046711138423197e-06,
      "loss": 0.1791,
      "num_input_tokens_seen": 28909552,
      "step": 33380
    },
    {
      "epoch": 15.74021687883074,
      "grad_norm": 1.4878727197647095,
      "learning_rate": 3.2997945741742255e-06,
      "loss": 0.2211,
      "num_input_tokens_seen": 28913632,
      "step": 33385
    },
    {
      "epoch": 15.742574257425742,
      "grad_norm": 1.2849268913269043,
      "learning_rate": 3.2949213809505082e-06,
      "loss": 0.1672,
      "num_input_tokens_seen": 28917888,
      "step": 33390
    },
    {
      "epoch": 15.744931636020745,
      "grad_norm": 0.7559008598327637,
      "learning_rate": 3.2900515349226834e-06,
      "loss": 0.2248,
      "num_input_tokens_seen": 28922080,
      "step": 33395
    },
    {
      "epoch": 15.747289014615747,
      "grad_norm": 1.1753807067871094,
      "learning_rate": 3.285185036841731e-06,
      "loss": 0.1031,
      "num_input_tokens_seen": 28926208,
      "step": 33400
    },
    {
      "epoch": 15.747289014615747,
      "eval_loss": 0.2006668746471405,
      "eval_runtime": 22.15,
      "eval_samples_per_second": 42.573,
      "eval_steps_per_second": 21.309,
      "num_input_tokens_seen": 28926208,
      "step": 33400
    },
    {
      "epoch": 15.74964639321075,
      "grad_norm": 1.2533067464828491,
      "learning_rate": 3.2803218874581377e-06,
      "loss": 0.1571,
      "num_input_tokens_seen": 28930288,
      "step": 33405
    },
    {
      "epoch": 15.752003771805752,
      "grad_norm": 1.0643240213394165,
      "learning_rate": 3.2754620875218494e-06,
      "loss": 0.262,
      "num_input_tokens_seen": 28935376,
      "step": 33410
    },
    {
      "epoch": 15.754361150400754,
      "grad_norm": 1.0924296379089355,
      "learning_rate": 3.2706056377823146e-06,
      "loss": 0.1264,
      "num_input_tokens_seen": 28939248,
      "step": 33415
    },
    {
      "epoch": 15.756718528995757,
      "grad_norm": 0.9273841977119446,
      "learning_rate": 3.2657525389884647e-06,
      "loss": 0.2054,
      "num_input_tokens_seen": 28943392,
      "step": 33420
    },
    {
      "epoch": 15.75907590759076,
      "grad_norm": 0.8481869697570801,
      "learning_rate": 3.260902791888698e-06,
      "loss": 0.1337,
      "num_input_tokens_seen": 28947808,
      "step": 33425
    },
    {
      "epoch": 15.761433286185762,
      "grad_norm": 0.6117249727249146,
      "learning_rate": 3.2560563972309166e-06,
      "loss": 0.2305,
      "num_input_tokens_seen": 28951984,
      "step": 33430
    },
    {
      "epoch": 15.763790664780764,
      "grad_norm": 0.7895810008049011,
      "learning_rate": 3.251213355762489e-06,
      "loss": 0.1963,
      "num_input_tokens_seen": 28956032,
      "step": 33435
    },
    {
      "epoch": 15.766148043375766,
      "grad_norm": 0.9138057827949524,
      "learning_rate": 3.2463736682302707e-06,
      "loss": 0.1169,
      "num_input_tokens_seen": 28960672,
      "step": 33440
    },
    {
      "epoch": 15.768505421970769,
      "grad_norm": 1.581095814704895,
      "learning_rate": 3.2415373353806124e-06,
      "loss": 0.2356,
      "num_input_tokens_seen": 28965248,
      "step": 33445
    },
    {
      "epoch": 15.770862800565771,
      "grad_norm": 1.3730003833770752,
      "learning_rate": 3.236704357959322e-06,
      "loss": 0.1615,
      "num_input_tokens_seen": 28969824,
      "step": 33450
    },
    {
      "epoch": 15.773220179160774,
      "grad_norm": 1.86228346824646,
      "learning_rate": 3.2318747367117154e-06,
      "loss": 0.1415,
      "num_input_tokens_seen": 28974368,
      "step": 33455
    },
    {
      "epoch": 15.775577557755776,
      "grad_norm": 0.5847036242485046,
      "learning_rate": 3.227048472382585e-06,
      "loss": 0.1328,
      "num_input_tokens_seen": 28978128,
      "step": 33460
    },
    {
      "epoch": 15.777934936350778,
      "grad_norm": 0.9999526739120483,
      "learning_rate": 3.2222255657161915e-06,
      "loss": 0.1119,
      "num_input_tokens_seen": 28982048,
      "step": 33465
    },
    {
      "epoch": 15.78029231494578,
      "grad_norm": 1.811478614807129,
      "learning_rate": 3.2174060174562924e-06,
      "loss": 0.2301,
      "num_input_tokens_seen": 28986336,
      "step": 33470
    },
    {
      "epoch": 15.782649693540783,
      "grad_norm": 1.4623990058898926,
      "learning_rate": 3.2125898283461298e-06,
      "loss": 0.2101,
      "num_input_tokens_seen": 28990560,
      "step": 33475
    },
    {
      "epoch": 15.785007072135786,
      "grad_norm": 0.7577841281890869,
      "learning_rate": 3.207776999128406e-06,
      "loss": 0.1099,
      "num_input_tokens_seen": 28994416,
      "step": 33480
    },
    {
      "epoch": 15.787364450730788,
      "grad_norm": 1.7560397386550903,
      "learning_rate": 3.202967530545331e-06,
      "loss": 0.1382,
      "num_input_tokens_seen": 28998304,
      "step": 33485
    },
    {
      "epoch": 15.78972182932579,
      "grad_norm": 0.8683933615684509,
      "learning_rate": 3.1981614233385778e-06,
      "loss": 0.1667,
      "num_input_tokens_seen": 29002160,
      "step": 33490
    },
    {
      "epoch": 15.792079207920793,
      "grad_norm": 1.3615927696228027,
      "learning_rate": 3.1933586782493115e-06,
      "loss": 0.1864,
      "num_input_tokens_seen": 29006768,
      "step": 33495
    },
    {
      "epoch": 15.794436586515795,
      "grad_norm": 0.801527738571167,
      "learning_rate": 3.188559296018184e-06,
      "loss": 0.2079,
      "num_input_tokens_seen": 29011472,
      "step": 33500
    },
    {
      "epoch": 15.796793965110798,
      "grad_norm": 0.6107223629951477,
      "learning_rate": 3.1837632773853098e-06,
      "loss": 0.1688,
      "num_input_tokens_seen": 29015984,
      "step": 33505
    },
    {
      "epoch": 15.799151343705798,
      "grad_norm": 1.0862910747528076,
      "learning_rate": 3.178970623090294e-06,
      "loss": 0.1964,
      "num_input_tokens_seen": 29020672,
      "step": 33510
    },
    {
      "epoch": 15.801508722300802,
      "grad_norm": 1.1196645498275757,
      "learning_rate": 3.174181333872234e-06,
      "loss": 0.1461,
      "num_input_tokens_seen": 29024608,
      "step": 33515
    },
    {
      "epoch": 15.803866100895803,
      "grad_norm": 1.289502501487732,
      "learning_rate": 3.169395410469686e-06,
      "loss": 0.1859,
      "num_input_tokens_seen": 29029024,
      "step": 33520
    },
    {
      "epoch": 15.806223479490805,
      "grad_norm": 0.3194296061992645,
      "learning_rate": 3.164612853620713e-06,
      "loss": 0.1952,
      "num_input_tokens_seen": 29033280,
      "step": 33525
    },
    {
      "epoch": 15.808580858085808,
      "grad_norm": 0.46292343735694885,
      "learning_rate": 3.1598336640628333e-06,
      "loss": 0.1648,
      "num_input_tokens_seen": 29037584,
      "step": 33530
    },
    {
      "epoch": 15.81093823668081,
      "grad_norm": 0.7410999536514282,
      "learning_rate": 3.155057842533063e-06,
      "loss": 0.1553,
      "num_input_tokens_seen": 29041312,
      "step": 33535
    },
    {
      "epoch": 15.813295615275813,
      "grad_norm": 0.7857690453529358,
      "learning_rate": 3.1502853897678984e-06,
      "loss": 0.1178,
      "num_input_tokens_seen": 29045360,
      "step": 33540
    },
    {
      "epoch": 15.815652993870815,
      "grad_norm": 3.3910911083221436,
      "learning_rate": 3.1455163065033017e-06,
      "loss": 0.2071,
      "num_input_tokens_seen": 29049792,
      "step": 33545
    },
    {
      "epoch": 15.818010372465817,
      "grad_norm": 1.6087571382522583,
      "learning_rate": 3.140750593474734e-06,
      "loss": 0.1798,
      "num_input_tokens_seen": 29054096,
      "step": 33550
    },
    {
      "epoch": 15.82036775106082,
      "grad_norm": 1.9320651292800903,
      "learning_rate": 3.1359882514171294e-06,
      "loss": 0.1395,
      "num_input_tokens_seen": 29057952,
      "step": 33555
    },
    {
      "epoch": 15.822725129655822,
      "grad_norm": 0.5932599902153015,
      "learning_rate": 3.1312292810648903e-06,
      "loss": 0.1308,
      "num_input_tokens_seen": 29062544,
      "step": 33560
    },
    {
      "epoch": 15.825082508250825,
      "grad_norm": 2.605746030807495,
      "learning_rate": 3.1264736831519204e-06,
      "loss": 0.2285,
      "num_input_tokens_seen": 29066992,
      "step": 33565
    },
    {
      "epoch": 15.827439886845827,
      "grad_norm": 0.9876322150230408,
      "learning_rate": 3.1217214584115863e-06,
      "loss": 0.2191,
      "num_input_tokens_seen": 29070704,
      "step": 33570
    },
    {
      "epoch": 15.82979726544083,
      "grad_norm": 1.0787403583526611,
      "learning_rate": 3.116972607576746e-06,
      "loss": 0.2465,
      "num_input_tokens_seen": 29074736,
      "step": 33575
    },
    {
      "epoch": 15.832154644035832,
      "grad_norm": 0.6921275854110718,
      "learning_rate": 3.1122271313797303e-06,
      "loss": 0.1399,
      "num_input_tokens_seen": 29079136,
      "step": 33580
    },
    {
      "epoch": 15.834512022630834,
      "grad_norm": 0.6476067900657654,
      "learning_rate": 3.107485030552343e-06,
      "loss": 0.1851,
      "num_input_tokens_seen": 29083904,
      "step": 33585
    },
    {
      "epoch": 15.836869401225837,
      "grad_norm": 1.9971479177474976,
      "learning_rate": 3.1027463058258848e-06,
      "loss": 0.1457,
      "num_input_tokens_seen": 29088608,
      "step": 33590
    },
    {
      "epoch": 15.839226779820839,
      "grad_norm": 1.0138278007507324,
      "learning_rate": 3.0980109579311273e-06,
      "loss": 0.1506,
      "num_input_tokens_seen": 29092688,
      "step": 33595
    },
    {
      "epoch": 15.841584158415841,
      "grad_norm": 0.5103396773338318,
      "learning_rate": 3.093278987598314e-06,
      "loss": 0.1614,
      "num_input_tokens_seen": 29096816,
      "step": 33600
    },
    {
      "epoch": 15.841584158415841,
      "eval_loss": 0.19996872544288635,
      "eval_runtime": 22.1639,
      "eval_samples_per_second": 42.547,
      "eval_steps_per_second": 21.296,
      "num_input_tokens_seen": 29096816,
      "step": 33600
    },
    {
      "epoch": 15.843941537010844,
      "grad_norm": 1.38894784450531,
      "learning_rate": 3.0885503955571826e-06,
      "loss": 0.1984,
      "num_input_tokens_seen": 29101056,
      "step": 33605
    },
    {
      "epoch": 15.846298915605846,
      "grad_norm": 1.014708161354065,
      "learning_rate": 3.0838251825369313e-06,
      "loss": 0.1968,
      "num_input_tokens_seen": 29105568,
      "step": 33610
    },
    {
      "epoch": 15.848656294200849,
      "grad_norm": 0.43684354424476624,
      "learning_rate": 3.0791033492662517e-06,
      "loss": 0.15,
      "num_input_tokens_seen": 29109104,
      "step": 33615
    },
    {
      "epoch": 15.851013672795851,
      "grad_norm": 0.5316733121871948,
      "learning_rate": 3.0743848964733203e-06,
      "loss": 0.143,
      "num_input_tokens_seen": 29113312,
      "step": 33620
    },
    {
      "epoch": 15.853371051390853,
      "grad_norm": 0.497302770614624,
      "learning_rate": 3.0696698248857625e-06,
      "loss": 0.1119,
      "num_input_tokens_seen": 29117600,
      "step": 33625
    },
    {
      "epoch": 15.855728429985856,
      "grad_norm": 0.4191017150878906,
      "learning_rate": 3.0649581352307192e-06,
      "loss": 0.1213,
      "num_input_tokens_seen": 29122096,
      "step": 33630
    },
    {
      "epoch": 15.858085808580858,
      "grad_norm": 1.359073281288147,
      "learning_rate": 3.060249828234776e-06,
      "loss": 0.1732,
      "num_input_tokens_seen": 29125824,
      "step": 33635
    },
    {
      "epoch": 15.86044318717586,
      "grad_norm": 1.1435312032699585,
      "learning_rate": 3.055544904624025e-06,
      "loss": 0.2133,
      "num_input_tokens_seen": 29130240,
      "step": 33640
    },
    {
      "epoch": 15.862800565770863,
      "grad_norm": 0.6197288036346436,
      "learning_rate": 3.050843365124026e-06,
      "loss": 0.1982,
      "num_input_tokens_seen": 29134912,
      "step": 33645
    },
    {
      "epoch": 15.865157944365865,
      "grad_norm": 0.7204111814498901,
      "learning_rate": 3.0461452104598083e-06,
      "loss": 0.2093,
      "num_input_tokens_seen": 29139456,
      "step": 33650
    },
    {
      "epoch": 15.867515322960868,
      "grad_norm": 0.5400391817092896,
      "learning_rate": 3.0414504413558836e-06,
      "loss": 0.1773,
      "num_input_tokens_seen": 29143600,
      "step": 33655
    },
    {
      "epoch": 15.86987270155587,
      "grad_norm": 1.1027997732162476,
      "learning_rate": 3.0367590585362564e-06,
      "loss": 0.1288,
      "num_input_tokens_seen": 29147712,
      "step": 33660
    },
    {
      "epoch": 15.872230080150873,
      "grad_norm": 1.1871753931045532,
      "learning_rate": 3.0320710627243813e-06,
      "loss": 0.1791,
      "num_input_tokens_seen": 29151776,
      "step": 33665
    },
    {
      "epoch": 15.874587458745875,
      "grad_norm": 0.9665055871009827,
      "learning_rate": 3.027386454643222e-06,
      "loss": 0.1621,
      "num_input_tokens_seen": 29156512,
      "step": 33670
    },
    {
      "epoch": 15.876944837340877,
      "grad_norm": 0.6683578491210938,
      "learning_rate": 3.0227052350151914e-06,
      "loss": 0.1613,
      "num_input_tokens_seen": 29161456,
      "step": 33675
    },
    {
      "epoch": 15.87930221593588,
      "grad_norm": 1.5537006855010986,
      "learning_rate": 3.0180274045621957e-06,
      "loss": 0.1774,
      "num_input_tokens_seen": 29165488,
      "step": 33680
    },
    {
      "epoch": 15.881659594530882,
      "grad_norm": 1.3626151084899902,
      "learning_rate": 3.013352964005625e-06,
      "loss": 0.1402,
      "num_input_tokens_seen": 29169648,
      "step": 33685
    },
    {
      "epoch": 15.884016973125885,
      "grad_norm": 0.86297607421875,
      "learning_rate": 3.0086819140663218e-06,
      "loss": 0.1517,
      "num_input_tokens_seen": 29173344,
      "step": 33690
    },
    {
      "epoch": 15.886374351720887,
      "grad_norm": 0.8383913636207581,
      "learning_rate": 3.0040142554646265e-06,
      "loss": 0.2001,
      "num_input_tokens_seen": 29177312,
      "step": 33695
    },
    {
      "epoch": 15.88873173031589,
      "grad_norm": 0.6753535270690918,
      "learning_rate": 2.999349988920361e-06,
      "loss": 0.1577,
      "num_input_tokens_seen": 29181696,
      "step": 33700
    },
    {
      "epoch": 15.891089108910892,
      "grad_norm": 1.2154649496078491,
      "learning_rate": 2.994689115152796e-06,
      "loss": 0.2727,
      "num_input_tokens_seen": 29185808,
      "step": 33705
    },
    {
      "epoch": 15.893446487505894,
      "grad_norm": 0.4962961673736572,
      "learning_rate": 2.9900316348807105e-06,
      "loss": 0.1879,
      "num_input_tokens_seen": 29190064,
      "step": 33710
    },
    {
      "epoch": 15.895803866100895,
      "grad_norm": 1.582605004310608,
      "learning_rate": 2.985377548822338e-06,
      "loss": 0.1515,
      "num_input_tokens_seen": 29193632,
      "step": 33715
    },
    {
      "epoch": 15.898161244695899,
      "grad_norm": 0.5082992315292358,
      "learning_rate": 2.980726857695404e-06,
      "loss": 0.1981,
      "num_input_tokens_seen": 29197520,
      "step": 33720
    },
    {
      "epoch": 15.9005186232909,
      "grad_norm": 1.1733742952346802,
      "learning_rate": 2.9760795622171017e-06,
      "loss": 0.1683,
      "num_input_tokens_seen": 29202544,
      "step": 33725
    },
    {
      "epoch": 15.902876001885902,
      "grad_norm": 1.6746052503585815,
      "learning_rate": 2.971435663104094e-06,
      "loss": 0.187,
      "num_input_tokens_seen": 29207168,
      "step": 33730
    },
    {
      "epoch": 15.905233380480905,
      "grad_norm": 0.6684556007385254,
      "learning_rate": 2.9667951610725385e-06,
      "loss": 0.2236,
      "num_input_tokens_seen": 29211728,
      "step": 33735
    },
    {
      "epoch": 15.907590759075907,
      "grad_norm": 1.0939873456954956,
      "learning_rate": 2.9621580568380575e-06,
      "loss": 0.1117,
      "num_input_tokens_seen": 29215616,
      "step": 33740
    },
    {
      "epoch": 15.90994813767091,
      "grad_norm": 0.6312682628631592,
      "learning_rate": 2.9575243511157453e-06,
      "loss": 0.1225,
      "num_input_tokens_seen": 29220928,
      "step": 33745
    },
    {
      "epoch": 15.912305516265912,
      "grad_norm": 1.6165639162063599,
      "learning_rate": 2.952894044620186e-06,
      "loss": 0.1641,
      "num_input_tokens_seen": 29224608,
      "step": 33750
    },
    {
      "epoch": 15.914662894860914,
      "grad_norm": 1.7990270853042603,
      "learning_rate": 2.948267138065419e-06,
      "loss": 0.213,
      "num_input_tokens_seen": 29229120,
      "step": 33755
    },
    {
      "epoch": 15.917020273455917,
      "grad_norm": 0.6943824887275696,
      "learning_rate": 2.943643632164983e-06,
      "loss": 0.2625,
      "num_input_tokens_seen": 29233408,
      "step": 33760
    },
    {
      "epoch": 15.919377652050919,
      "grad_norm": 0.5508635640144348,
      "learning_rate": 2.939023527631879e-06,
      "loss": 0.2694,
      "num_input_tokens_seen": 29237008,
      "step": 33765
    },
    {
      "epoch": 15.921735030645921,
      "grad_norm": 0.38460832834243774,
      "learning_rate": 2.934406825178576e-06,
      "loss": 0.2,
      "num_input_tokens_seen": 29240384,
      "step": 33770
    },
    {
      "epoch": 15.924092409240924,
      "grad_norm": 0.4910089373588562,
      "learning_rate": 2.9297935255170357e-06,
      "loss": 0.1456,
      "num_input_tokens_seen": 29244832,
      "step": 33775
    },
    {
      "epoch": 15.926449787835926,
      "grad_norm": 1.13553786277771,
      "learning_rate": 2.925183629358691e-06,
      "loss": 0.1673,
      "num_input_tokens_seen": 29250896,
      "step": 33780
    },
    {
      "epoch": 15.928807166430929,
      "grad_norm": 0.6768810749053955,
      "learning_rate": 2.9205771374144346e-06,
      "loss": 0.2621,
      "num_input_tokens_seen": 29255168,
      "step": 33785
    },
    {
      "epoch": 15.931164545025931,
      "grad_norm": 1.2257373332977295,
      "learning_rate": 2.915974050394657e-06,
      "loss": 0.2247,
      "num_input_tokens_seen": 29258880,
      "step": 33790
    },
    {
      "epoch": 15.933521923620933,
      "grad_norm": 1.363724708557129,
      "learning_rate": 2.9113743690092067e-06,
      "loss": 0.2289,
      "num_input_tokens_seen": 29263200,
      "step": 33795
    },
    {
      "epoch": 15.935879302215936,
      "grad_norm": 0.7035794854164124,
      "learning_rate": 2.906778093967402e-06,
      "loss": 0.1393,
      "num_input_tokens_seen": 29267072,
      "step": 33800
    },
    {
      "epoch": 15.935879302215936,
      "eval_loss": 0.19981051981449127,
      "eval_runtime": 22.138,
      "eval_samples_per_second": 42.596,
      "eval_steps_per_second": 21.321,
      "num_input_tokens_seen": 29267072,
      "step": 33800
    },
    {
      "epoch": 15.938236680810938,
      "grad_norm": 1.2290781736373901,
      "learning_rate": 2.9021852259780656e-06,
      "loss": 0.2238,
      "num_input_tokens_seen": 29272208,
      "step": 33805
    },
    {
      "epoch": 15.94059405940594,
      "grad_norm": 1.1933832168579102,
      "learning_rate": 2.8975957657494583e-06,
      "loss": 0.1361,
      "num_input_tokens_seen": 29276528,
      "step": 33810
    },
    {
      "epoch": 15.942951438000943,
      "grad_norm": 0.7444612979888916,
      "learning_rate": 2.8930097139893417e-06,
      "loss": 0.1658,
      "num_input_tokens_seen": 29280416,
      "step": 33815
    },
    {
      "epoch": 15.945308816595945,
      "grad_norm": 0.9466022849082947,
      "learning_rate": 2.888427071404945e-06,
      "loss": 0.1861,
      "num_input_tokens_seen": 29285088,
      "step": 33820
    },
    {
      "epoch": 15.947666195190948,
      "grad_norm": 1.2784713506698608,
      "learning_rate": 2.8838478387029606e-06,
      "loss": 0.1532,
      "num_input_tokens_seen": 29289168,
      "step": 33825
    },
    {
      "epoch": 15.95002357378595,
      "grad_norm": 0.45573854446411133,
      "learning_rate": 2.8792720165895737e-06,
      "loss": 0.1605,
      "num_input_tokens_seen": 29293296,
      "step": 33830
    },
    {
      "epoch": 15.952380952380953,
      "grad_norm": 0.5792328119277954,
      "learning_rate": 2.874699605770423e-06,
      "loss": 0.1828,
      "num_input_tokens_seen": 29297696,
      "step": 33835
    },
    {
      "epoch": 15.954738330975955,
      "grad_norm": 1.0256019830703735,
      "learning_rate": 2.8701306069506383e-06,
      "loss": 0.1767,
      "num_input_tokens_seen": 29301664,
      "step": 33840
    },
    {
      "epoch": 15.957095709570957,
      "grad_norm": 0.4467410743236542,
      "learning_rate": 2.8655650208348178e-06,
      "loss": 0.1537,
      "num_input_tokens_seen": 29305776,
      "step": 33845
    },
    {
      "epoch": 15.95945308816596,
      "grad_norm": 2.281364679336548,
      "learning_rate": 2.8610028481270257e-06,
      "loss": 0.1651,
      "num_input_tokens_seen": 29309856,
      "step": 33850
    },
    {
      "epoch": 15.961810466760962,
      "grad_norm": 1.2795722484588623,
      "learning_rate": 2.856444089530813e-06,
      "loss": 0.1261,
      "num_input_tokens_seen": 29313328,
      "step": 33855
    },
    {
      "epoch": 15.964167845355965,
      "grad_norm": 0.8288614153862,
      "learning_rate": 2.8518887457491955e-06,
      "loss": 0.1388,
      "num_input_tokens_seen": 29317344,
      "step": 33860
    },
    {
      "epoch": 15.966525223950967,
      "grad_norm": 1.3909512758255005,
      "learning_rate": 2.8473368174846666e-06,
      "loss": 0.16,
      "num_input_tokens_seen": 29321344,
      "step": 33865
    },
    {
      "epoch": 15.96888260254597,
      "grad_norm": 1.5527559518814087,
      "learning_rate": 2.842788305439184e-06,
      "loss": 0.2171,
      "num_input_tokens_seen": 29325200,
      "step": 33870
    },
    {
      "epoch": 15.971239981140972,
      "grad_norm": 0.7102486491203308,
      "learning_rate": 2.8382432103141925e-06,
      "loss": 0.177,
      "num_input_tokens_seen": 29329248,
      "step": 33875
    },
    {
      "epoch": 15.973597359735974,
      "grad_norm": 1.2932488918304443,
      "learning_rate": 2.833701532810598e-06,
      "loss": 0.1574,
      "num_input_tokens_seen": 29333248,
      "step": 33880
    },
    {
      "epoch": 15.975954738330977,
      "grad_norm": 1.3359651565551758,
      "learning_rate": 2.8291632736287877e-06,
      "loss": 0.1903,
      "num_input_tokens_seen": 29337312,
      "step": 33885
    },
    {
      "epoch": 15.978312116925979,
      "grad_norm": 1.997387409210205,
      "learning_rate": 2.824628433468615e-06,
      "loss": 0.1513,
      "num_input_tokens_seen": 29340896,
      "step": 33890
    },
    {
      "epoch": 15.980669495520981,
      "grad_norm": 0.4760560691356659,
      "learning_rate": 2.8200970130294073e-06,
      "loss": 0.1237,
      "num_input_tokens_seen": 29344928,
      "step": 33895
    },
    {
      "epoch": 15.983026874115984,
      "grad_norm": 1.702332615852356,
      "learning_rate": 2.8155690130099775e-06,
      "loss": 0.2129,
      "num_input_tokens_seen": 29349616,
      "step": 33900
    },
    {
      "epoch": 15.985384252710986,
      "grad_norm": 1.4332706928253174,
      "learning_rate": 2.8110444341085895e-06,
      "loss": 0.187,
      "num_input_tokens_seen": 29353760,
      "step": 33905
    },
    {
      "epoch": 15.987741631305987,
      "grad_norm": 0.6783865690231323,
      "learning_rate": 2.806523277022996e-06,
      "loss": 0.2004,
      "num_input_tokens_seen": 29357584,
      "step": 33910
    },
    {
      "epoch": 15.990099009900991,
      "grad_norm": 0.34902727603912354,
      "learning_rate": 2.802005542450409e-06,
      "loss": 0.1416,
      "num_input_tokens_seen": 29362144,
      "step": 33915
    },
    {
      "epoch": 15.992456388495992,
      "grad_norm": 0.8760130405426025,
      "learning_rate": 2.797491231087526e-06,
      "loss": 0.2219,
      "num_input_tokens_seen": 29367296,
      "step": 33920
    },
    {
      "epoch": 15.994813767090994,
      "grad_norm": 1.2470823526382446,
      "learning_rate": 2.7929803436305137e-06,
      "loss": 0.1654,
      "num_input_tokens_seen": 29371600,
      "step": 33925
    },
    {
      "epoch": 15.997171145685996,
      "grad_norm": 2.3500664234161377,
      "learning_rate": 2.788472880774998e-06,
      "loss": 0.1851,
      "num_input_tokens_seen": 29376656,
      "step": 33930
    },
    {
      "epoch": 15.999528524280999,
      "grad_norm": 1.0821192264556885,
      "learning_rate": 2.7839688432160977e-06,
      "loss": 0.1684,
      "num_input_tokens_seen": 29379904,
      "step": 33935
    },
    {
      "epoch": 16.001885902876,
      "grad_norm": 0.7819023728370667,
      "learning_rate": 2.779468231648383e-06,
      "loss": 0.1695,
      "num_input_tokens_seen": 29384608,
      "step": 33940
    },
    {
      "epoch": 16.004243281471005,
      "grad_norm": 0.4686228334903717,
      "learning_rate": 2.774971046765906e-06,
      "loss": 0.161,
      "num_input_tokens_seen": 29388624,
      "step": 33945
    },
    {
      "epoch": 16.006600660066006,
      "grad_norm": 0.9471889734268188,
      "learning_rate": 2.770477289262194e-06,
      "loss": 0.1233,
      "num_input_tokens_seen": 29391808,
      "step": 33950
    },
    {
      "epoch": 16.00895803866101,
      "grad_norm": 1.2059755325317383,
      "learning_rate": 2.765986959830233e-06,
      "loss": 0.1471,
      "num_input_tokens_seen": 29395840,
      "step": 33955
    },
    {
      "epoch": 16.01131541725601,
      "grad_norm": 0.3132414221763611,
      "learning_rate": 2.761500059162492e-06,
      "loss": 0.1312,
      "num_input_tokens_seen": 29400608,
      "step": 33960
    },
    {
      "epoch": 16.013672795851015,
      "grad_norm": 0.9822475910186768,
      "learning_rate": 2.757016587950914e-06,
      "loss": 0.1462,
      "num_input_tokens_seen": 29404592,
      "step": 33965
    },
    {
      "epoch": 16.016030174446016,
      "grad_norm": 0.39518728852272034,
      "learning_rate": 2.752536546886897e-06,
      "loss": 0.1546,
      "num_input_tokens_seen": 29409552,
      "step": 33970
    },
    {
      "epoch": 16.01838755304102,
      "grad_norm": 2.155142068862915,
      "learning_rate": 2.7480599366613234e-06,
      "loss": 0.2495,
      "num_input_tokens_seen": 29413168,
      "step": 33975
    },
    {
      "epoch": 16.02074493163602,
      "grad_norm": 1.2419660091400146,
      "learning_rate": 2.7435867579645473e-06,
      "loss": 0.0863,
      "num_input_tokens_seen": 29418032,
      "step": 33980
    },
    {
      "epoch": 16.023102310231025,
      "grad_norm": 1.2893104553222656,
      "learning_rate": 2.739117011486378e-06,
      "loss": 0.2363,
      "num_input_tokens_seen": 29422608,
      "step": 33985
    },
    {
      "epoch": 16.025459688826025,
      "grad_norm": 1.2895162105560303,
      "learning_rate": 2.7346506979161216e-06,
      "loss": 0.179,
      "num_input_tokens_seen": 29426800,
      "step": 33990
    },
    {
      "epoch": 16.02781706742103,
      "grad_norm": 1.0021971464157104,
      "learning_rate": 2.7301878179425227e-06,
      "loss": 0.1517,
      "num_input_tokens_seen": 29431328,
      "step": 33995
    },
    {
      "epoch": 16.03017444601603,
      "grad_norm": 1.1507562398910522,
      "learning_rate": 2.7257283722538244e-06,
      "loss": 0.1771,
      "num_input_tokens_seen": 29435360,
      "step": 34000
    },
    {
      "epoch": 16.03017444601603,
      "eval_loss": 0.19953377544879913,
      "eval_runtime": 22.1333,
      "eval_samples_per_second": 42.606,
      "eval_steps_per_second": 21.325,
      "num_input_tokens_seen": 29435360,
      "step": 34000
    },
    {
      "epoch": 16.032531824611034,
      "grad_norm": 0.9942497611045837,
      "learning_rate": 2.7212723615377326e-06,
      "loss": 0.2844,
      "num_input_tokens_seen": 29439888,
      "step": 34005
    },
    {
      "epoch": 16.034889203206035,
      "grad_norm": 1.8248616456985474,
      "learning_rate": 2.7168197864814145e-06,
      "loss": 0.1944,
      "num_input_tokens_seen": 29444192,
      "step": 34010
    },
    {
      "epoch": 16.03724658180104,
      "grad_norm": 0.8452807664871216,
      "learning_rate": 2.712370647771509e-06,
      "loss": 0.3082,
      "num_input_tokens_seen": 29448560,
      "step": 34015
    },
    {
      "epoch": 16.03960396039604,
      "grad_norm": 0.5832785964012146,
      "learning_rate": 2.707924946094137e-06,
      "loss": 0.2168,
      "num_input_tokens_seen": 29452960,
      "step": 34020
    },
    {
      "epoch": 16.04196133899104,
      "grad_norm": 1.2057384252548218,
      "learning_rate": 2.7034826821348723e-06,
      "loss": 0.193,
      "num_input_tokens_seen": 29458400,
      "step": 34025
    },
    {
      "epoch": 16.044318717586044,
      "grad_norm": 1.515751600265503,
      "learning_rate": 2.6990438565787786e-06,
      "loss": 0.1599,
      "num_input_tokens_seen": 29462000,
      "step": 34030
    },
    {
      "epoch": 16.046676096181045,
      "grad_norm": 1.8342573642730713,
      "learning_rate": 2.6946084701103714e-06,
      "loss": 0.174,
      "num_input_tokens_seen": 29465824,
      "step": 34035
    },
    {
      "epoch": 16.04903347477605,
      "grad_norm": 0.7138785719871521,
      "learning_rate": 2.6901765234136428e-06,
      "loss": 0.1684,
      "num_input_tokens_seen": 29471280,
      "step": 34040
    },
    {
      "epoch": 16.05139085337105,
      "grad_norm": 0.49854445457458496,
      "learning_rate": 2.685748017172063e-06,
      "loss": 0.2157,
      "num_input_tokens_seen": 29475728,
      "step": 34045
    },
    {
      "epoch": 16.053748231966054,
      "grad_norm": 0.7351404428482056,
      "learning_rate": 2.681322952068549e-06,
      "loss": 0.1244,
      "num_input_tokens_seen": 29479824,
      "step": 34050
    },
    {
      "epoch": 16.056105610561055,
      "grad_norm": 0.7554737329483032,
      "learning_rate": 2.6769013287855137e-06,
      "loss": 0.1349,
      "num_input_tokens_seen": 29484112,
      "step": 34055
    },
    {
      "epoch": 16.05846298915606,
      "grad_norm": 0.4770165681838989,
      "learning_rate": 2.6724831480048286e-06,
      "loss": 0.2799,
      "num_input_tokens_seen": 29488272,
      "step": 34060
    },
    {
      "epoch": 16.06082036775106,
      "grad_norm": 0.4254116117954254,
      "learning_rate": 2.66806841040782e-06,
      "loss": 0.1596,
      "num_input_tokens_seen": 29493088,
      "step": 34065
    },
    {
      "epoch": 16.063177746346064,
      "grad_norm": 1.119785189628601,
      "learning_rate": 2.6636571166753083e-06,
      "loss": 0.2646,
      "num_input_tokens_seen": 29497168,
      "step": 34070
    },
    {
      "epoch": 16.065535124941064,
      "grad_norm": 2.7319138050079346,
      "learning_rate": 2.6592492674875598e-06,
      "loss": 0.2025,
      "num_input_tokens_seen": 29501760,
      "step": 34075
    },
    {
      "epoch": 16.06789250353607,
      "grad_norm": 1.8662391901016235,
      "learning_rate": 2.6548448635243305e-06,
      "loss": 0.1958,
      "num_input_tokens_seen": 29506336,
      "step": 34080
    },
    {
      "epoch": 16.07024988213107,
      "grad_norm": 0.49174752831459045,
      "learning_rate": 2.650443905464828e-06,
      "loss": 0.1357,
      "num_input_tokens_seen": 29511888,
      "step": 34085
    },
    {
      "epoch": 16.072607260726073,
      "grad_norm": 1.711926817893982,
      "learning_rate": 2.646046393987739e-06,
      "loss": 0.1436,
      "num_input_tokens_seen": 29516096,
      "step": 34090
    },
    {
      "epoch": 16.074964639321074,
      "grad_norm": 0.9637434482574463,
      "learning_rate": 2.64165232977121e-06,
      "loss": 0.1519,
      "num_input_tokens_seen": 29520144,
      "step": 34095
    },
    {
      "epoch": 16.077322017916078,
      "grad_norm": 0.6909918189048767,
      "learning_rate": 2.6372617134928695e-06,
      "loss": 0.2143,
      "num_input_tokens_seen": 29524640,
      "step": 34100
    },
    {
      "epoch": 16.07967939651108,
      "grad_norm": 2.0327093601226807,
      "learning_rate": 2.6328745458297943e-06,
      "loss": 0.1594,
      "num_input_tokens_seen": 29529504,
      "step": 34105
    },
    {
      "epoch": 16.082036775106083,
      "grad_norm": 0.8583564758300781,
      "learning_rate": 2.6284908274585546e-06,
      "loss": 0.1678,
      "num_input_tokens_seen": 29533328,
      "step": 34110
    },
    {
      "epoch": 16.084394153701083,
      "grad_norm": 1.5165265798568726,
      "learning_rate": 2.6241105590551595e-06,
      "loss": 0.2031,
      "num_input_tokens_seen": 29538080,
      "step": 34115
    },
    {
      "epoch": 16.086751532296088,
      "grad_norm": 0.7753033638000488,
      "learning_rate": 2.6197337412951105e-06,
      "loss": 0.2198,
      "num_input_tokens_seen": 29542032,
      "step": 34120
    },
    {
      "epoch": 16.08910891089109,
      "grad_norm": 0.8712180852890015,
      "learning_rate": 2.6153603748533705e-06,
      "loss": 0.1151,
      "num_input_tokens_seen": 29546432,
      "step": 34125
    },
    {
      "epoch": 16.091466289486092,
      "grad_norm": 1.8272525072097778,
      "learning_rate": 2.6109904604043585e-06,
      "loss": 0.1683,
      "num_input_tokens_seen": 29550944,
      "step": 34130
    },
    {
      "epoch": 16.093823668081093,
      "grad_norm": 0.8864544630050659,
      "learning_rate": 2.6066239986219765e-06,
      "loss": 0.2279,
      "num_input_tokens_seen": 29555520,
      "step": 34135
    },
    {
      "epoch": 16.096181046676097,
      "grad_norm": 1.0905548334121704,
      "learning_rate": 2.602260990179592e-06,
      "loss": 0.1929,
      "num_input_tokens_seen": 29559856,
      "step": 34140
    },
    {
      "epoch": 16.098538425271098,
      "grad_norm": 0.9910525679588318,
      "learning_rate": 2.5979014357500248e-06,
      "loss": 0.2674,
      "num_input_tokens_seen": 29563696,
      "step": 34145
    },
    {
      "epoch": 16.100895803866102,
      "grad_norm": 3.51308274269104,
      "learning_rate": 2.5935453360055844e-06,
      "loss": 0.1825,
      "num_input_tokens_seen": 29569152,
      "step": 34150
    },
    {
      "epoch": 16.103253182461103,
      "grad_norm": 0.4819304943084717,
      "learning_rate": 2.5891926916180283e-06,
      "loss": 0.2062,
      "num_input_tokens_seen": 29573120,
      "step": 34155
    },
    {
      "epoch": 16.105610561056107,
      "grad_norm": 0.9575157761573792,
      "learning_rate": 2.5848435032585883e-06,
      "loss": 0.1671,
      "num_input_tokens_seen": 29577456,
      "step": 34160
    },
    {
      "epoch": 16.107967939651108,
      "grad_norm": 0.5613216161727905,
      "learning_rate": 2.58049777159797e-06,
      "loss": 0.1546,
      "num_input_tokens_seen": 29581424,
      "step": 34165
    },
    {
      "epoch": 16.11032531824611,
      "grad_norm": 1.4147300720214844,
      "learning_rate": 2.576155497306332e-06,
      "loss": 0.2516,
      "num_input_tokens_seen": 29585872,
      "step": 34170
    },
    {
      "epoch": 16.112682696841112,
      "grad_norm": 1.2680675983428955,
      "learning_rate": 2.57181668105331e-06,
      "loss": 0.227,
      "num_input_tokens_seen": 29589984,
      "step": 34175
    },
    {
      "epoch": 16.115040075436116,
      "grad_norm": 0.8530105948448181,
      "learning_rate": 2.567481323508014e-06,
      "loss": 0.2008,
      "num_input_tokens_seen": 29594672,
      "step": 34180
    },
    {
      "epoch": 16.117397454031117,
      "grad_norm": 1.2480347156524658,
      "learning_rate": 2.5631494253389954e-06,
      "loss": 0.1907,
      "num_input_tokens_seen": 29598992,
      "step": 34185
    },
    {
      "epoch": 16.11975483262612,
      "grad_norm": 0.783078670501709,
      "learning_rate": 2.5588209872142997e-06,
      "loss": 0.1392,
      "num_input_tokens_seen": 29602528,
      "step": 34190
    },
    {
      "epoch": 16.122112211221122,
      "grad_norm": 1.3764334917068481,
      "learning_rate": 2.5544960098014186e-06,
      "loss": 0.1704,
      "num_input_tokens_seen": 29606384,
      "step": 34195
    },
    {
      "epoch": 16.124469589816126,
      "grad_norm": 1.2063758373260498,
      "learning_rate": 2.550174493767318e-06,
      "loss": 0.2065,
      "num_input_tokens_seen": 29610720,
      "step": 34200
    },
    {
      "epoch": 16.124469589816126,
      "eval_loss": 0.20005469024181366,
      "eval_runtime": 22.1757,
      "eval_samples_per_second": 42.524,
      "eval_steps_per_second": 21.285,
      "num_input_tokens_seen": 29610720,
      "step": 34200
    },
    {
      "epoch": 16.126826968411127,
      "grad_norm": 2.426100015640259,
      "learning_rate": 2.545856439778438e-06,
      "loss": 0.2253,
      "num_input_tokens_seen": 29615744,
      "step": 34205
    },
    {
      "epoch": 16.12918434700613,
      "grad_norm": 0.6635550856590271,
      "learning_rate": 2.541541848500667e-06,
      "loss": 0.1482,
      "num_input_tokens_seen": 29619392,
      "step": 34210
    },
    {
      "epoch": 16.13154172560113,
      "grad_norm": 0.42894792556762695,
      "learning_rate": 2.5372307205993733e-06,
      "loss": 0.1696,
      "num_input_tokens_seen": 29623248,
      "step": 34215
    },
    {
      "epoch": 16.133899104196132,
      "grad_norm": 0.9871245622634888,
      "learning_rate": 2.5329230567393917e-06,
      "loss": 0.2007,
      "num_input_tokens_seen": 29626816,
      "step": 34220
    },
    {
      "epoch": 16.136256482791136,
      "grad_norm": 1.6020817756652832,
      "learning_rate": 2.5286188575850164e-06,
      "loss": 0.2337,
      "num_input_tokens_seen": 29632784,
      "step": 34225
    },
    {
      "epoch": 16.138613861386137,
      "grad_norm": 0.48557284474372864,
      "learning_rate": 2.5243181237999984e-06,
      "loss": 0.152,
      "num_input_tokens_seen": 29637200,
      "step": 34230
    },
    {
      "epoch": 16.14097123998114,
      "grad_norm": 0.5544450283050537,
      "learning_rate": 2.520020856047578e-06,
      "loss": 0.161,
      "num_input_tokens_seen": 29641216,
      "step": 34235
    },
    {
      "epoch": 16.14332861857614,
      "grad_norm": 1.3006600141525269,
      "learning_rate": 2.515727054990438e-06,
      "loss": 0.2292,
      "num_input_tokens_seen": 29645856,
      "step": 34240
    },
    {
      "epoch": 16.145685997171146,
      "grad_norm": 0.8612282872200012,
      "learning_rate": 2.511436721290747e-06,
      "loss": 0.1533,
      "num_input_tokens_seen": 29649776,
      "step": 34245
    },
    {
      "epoch": 16.148043375766147,
      "grad_norm": 0.7402763962745667,
      "learning_rate": 2.5071498556101164e-06,
      "loss": 0.1729,
      "num_input_tokens_seen": 29654016,
      "step": 34250
    },
    {
      "epoch": 16.15040075436115,
      "grad_norm": 0.31499651074409485,
      "learning_rate": 2.5028664586096485e-06,
      "loss": 0.1428,
      "num_input_tokens_seen": 29657952,
      "step": 34255
    },
    {
      "epoch": 16.15275813295615,
      "grad_norm": 1.308087706565857,
      "learning_rate": 2.498586530949881e-06,
      "loss": 0.157,
      "num_input_tokens_seen": 29661968,
      "step": 34260
    },
    {
      "epoch": 16.155115511551156,
      "grad_norm": 1.5504364967346191,
      "learning_rate": 2.4943100732908427e-06,
      "loss": 0.2107,
      "num_input_tokens_seen": 29666480,
      "step": 34265
    },
    {
      "epoch": 16.157472890146156,
      "grad_norm": 0.7730907797813416,
      "learning_rate": 2.4900370862920188e-06,
      "loss": 0.1742,
      "num_input_tokens_seen": 29670512,
      "step": 34270
    },
    {
      "epoch": 16.15983026874116,
      "grad_norm": 0.8865915536880493,
      "learning_rate": 2.4857675706123518e-06,
      "loss": 0.1381,
      "num_input_tokens_seen": 29674432,
      "step": 34275
    },
    {
      "epoch": 16.16218764733616,
      "grad_norm": 1.240888237953186,
      "learning_rate": 2.4815015269102543e-06,
      "loss": 0.1804,
      "num_input_tokens_seen": 29678720,
      "step": 34280
    },
    {
      "epoch": 16.164545025931165,
      "grad_norm": 1.6365482807159424,
      "learning_rate": 2.477238955843611e-06,
      "loss": 0.1988,
      "num_input_tokens_seen": 29683696,
      "step": 34285
    },
    {
      "epoch": 16.166902404526166,
      "grad_norm": 0.5067173838615417,
      "learning_rate": 2.4729798580697573e-06,
      "loss": 0.188,
      "num_input_tokens_seen": 29687824,
      "step": 34290
    },
    {
      "epoch": 16.16925978312117,
      "grad_norm": 1.285750389099121,
      "learning_rate": 2.4687242342455034e-06,
      "loss": 0.1657,
      "num_input_tokens_seen": 29692160,
      "step": 34295
    },
    {
      "epoch": 16.17161716171617,
      "grad_norm": 0.9172279834747314,
      "learning_rate": 2.4644720850271196e-06,
      "loss": 0.1885,
      "num_input_tokens_seen": 29696304,
      "step": 34300
    },
    {
      "epoch": 16.173974540311175,
      "grad_norm": 0.9622274041175842,
      "learning_rate": 2.4602234110703364e-06,
      "loss": 0.1576,
      "num_input_tokens_seen": 29700208,
      "step": 34305
    },
    {
      "epoch": 16.176331918906175,
      "grad_norm": 1.4751567840576172,
      "learning_rate": 2.4559782130303576e-06,
      "loss": 0.205,
      "num_input_tokens_seen": 29703920,
      "step": 34310
    },
    {
      "epoch": 16.17868929750118,
      "grad_norm": 2.159778118133545,
      "learning_rate": 2.451736491561843e-06,
      "loss": 0.2307,
      "num_input_tokens_seen": 29708400,
      "step": 34315
    },
    {
      "epoch": 16.18104667609618,
      "grad_norm": 0.9363350868225098,
      "learning_rate": 2.4474982473189163e-06,
      "loss": 0.1013,
      "num_input_tokens_seen": 29712656,
      "step": 34320
    },
    {
      "epoch": 16.183404054691184,
      "grad_norm": 1.475446343421936,
      "learning_rate": 2.4432634809551796e-06,
      "loss": 0.1651,
      "num_input_tokens_seen": 29717376,
      "step": 34325
    },
    {
      "epoch": 16.185761433286185,
      "grad_norm": 0.8824538588523865,
      "learning_rate": 2.439032193123675e-06,
      "loss": 0.1143,
      "num_input_tokens_seen": 29721200,
      "step": 34330
    },
    {
      "epoch": 16.18811881188119,
      "grad_norm": 0.6875060796737671,
      "learning_rate": 2.4348043844769297e-06,
      "loss": 0.1145,
      "num_input_tokens_seen": 29725472,
      "step": 34335
    },
    {
      "epoch": 16.19047619047619,
      "grad_norm": 0.6109195351600647,
      "learning_rate": 2.4305800556669146e-06,
      "loss": 0.2395,
      "num_input_tokens_seen": 29729376,
      "step": 34340
    },
    {
      "epoch": 16.192833569071194,
      "grad_norm": 1.8033708333969116,
      "learning_rate": 2.426359207345083e-06,
      "loss": 0.2183,
      "num_input_tokens_seen": 29733648,
      "step": 34345
    },
    {
      "epoch": 16.195190947666195,
      "grad_norm": 0.8657106161117554,
      "learning_rate": 2.4221418401623396e-06,
      "loss": 0.1781,
      "num_input_tokens_seen": 29737168,
      "step": 34350
    },
    {
      "epoch": 16.1975483262612,
      "grad_norm": 1.0489585399627686,
      "learning_rate": 2.4179279547690557e-06,
      "loss": 0.0781,
      "num_input_tokens_seen": 29742768,
      "step": 34355
    },
    {
      "epoch": 16.1999057048562,
      "grad_norm": 1.0066756010055542,
      "learning_rate": 2.413717551815062e-06,
      "loss": 0.1541,
      "num_input_tokens_seen": 29747328,
      "step": 34360
    },
    {
      "epoch": 16.202263083451204,
      "grad_norm": 0.5375425219535828,
      "learning_rate": 2.409510631949666e-06,
      "loss": 0.15,
      "num_input_tokens_seen": 29751328,
      "step": 34365
    },
    {
      "epoch": 16.204620462046204,
      "grad_norm": 0.4844767451286316,
      "learning_rate": 2.405307195821618e-06,
      "loss": 0.1503,
      "num_input_tokens_seen": 29755504,
      "step": 34370
    },
    {
      "epoch": 16.20697784064121,
      "grad_norm": 0.868802011013031,
      "learning_rate": 2.4011072440791372e-06,
      "loss": 0.1702,
      "num_input_tokens_seen": 29760336,
      "step": 34375
    },
    {
      "epoch": 16.20933521923621,
      "grad_norm": 1.0710006952285767,
      "learning_rate": 2.3969107773699233e-06,
      "loss": 0.2295,
      "num_input_tokens_seen": 29764752,
      "step": 34380
    },
    {
      "epoch": 16.211692597831213,
      "grad_norm": 0.7245571613311768,
      "learning_rate": 2.3927177963411096e-06,
      "loss": 0.1795,
      "num_input_tokens_seen": 29768816,
      "step": 34385
    },
    {
      "epoch": 16.214049976426214,
      "grad_norm": 2.0319621562957764,
      "learning_rate": 2.3885283016393144e-06,
      "loss": 0.3275,
      "num_input_tokens_seen": 29772992,
      "step": 34390
    },
    {
      "epoch": 16.216407355021218,
      "grad_norm": 1.3956694602966309,
      "learning_rate": 2.3843422939106076e-06,
      "loss": 0.1075,
      "num_input_tokens_seen": 29776848,
      "step": 34395
    },
    {
      "epoch": 16.21876473361622,
      "grad_norm": 1.1609503030776978,
      "learning_rate": 2.380159773800525e-06,
      "loss": 0.2042,
      "num_input_tokens_seen": 29781472,
      "step": 34400
    },
    {
      "epoch": 16.21876473361622,
      "eval_loss": 0.2002386599779129,
      "eval_runtime": 22.1628,
      "eval_samples_per_second": 42.549,
      "eval_steps_per_second": 21.297,
      "num_input_tokens_seen": 29781472,
      "step": 34400
    },
    {
      "epoch": 16.221122112211223,
      "grad_norm": 0.7819018363952637,
      "learning_rate": 2.3759807419540675e-06,
      "loss": 0.2084,
      "num_input_tokens_seen": 29786416,
      "step": 34405
    },
    {
      "epoch": 16.223479490806223,
      "grad_norm": 0.25941798090934753,
      "learning_rate": 2.3718051990156835e-06,
      "loss": 0.1692,
      "num_input_tokens_seen": 29790480,
      "step": 34410
    },
    {
      "epoch": 16.225836869401228,
      "grad_norm": 1.5496888160705566,
      "learning_rate": 2.367633145629311e-06,
      "loss": 0.119,
      "num_input_tokens_seen": 29795136,
      "step": 34415
    },
    {
      "epoch": 16.22819424799623,
      "grad_norm": 1.4928135871887207,
      "learning_rate": 2.363464582438316e-06,
      "loss": 0.1222,
      "num_input_tokens_seen": 29800080,
      "step": 34420
    },
    {
      "epoch": 16.23055162659123,
      "grad_norm": 1.1625934839248657,
      "learning_rate": 2.3592995100855526e-06,
      "loss": 0.1329,
      "num_input_tokens_seen": 29804224,
      "step": 34425
    },
    {
      "epoch": 16.232909005186233,
      "grad_norm": 0.9384304881095886,
      "learning_rate": 2.3551379292133273e-06,
      "loss": 0.298,
      "num_input_tokens_seen": 29808896,
      "step": 34430
    },
    {
      "epoch": 16.235266383781234,
      "grad_norm": 0.9652397036552429,
      "learning_rate": 2.3509798404634047e-06,
      "loss": 0.155,
      "num_input_tokens_seen": 29813664,
      "step": 34435
    },
    {
      "epoch": 16.237623762376238,
      "grad_norm": 0.8953929543495178,
      "learning_rate": 2.346825244477019e-06,
      "loss": 0.151,
      "num_input_tokens_seen": 29817184,
      "step": 34440
    },
    {
      "epoch": 16.23998114097124,
      "grad_norm": 0.651710569858551,
      "learning_rate": 2.3426741418948545e-06,
      "loss": 0.1587,
      "num_input_tokens_seen": 29821408,
      "step": 34445
    },
    {
      "epoch": 16.242338519566243,
      "grad_norm": 1.4817639589309692,
      "learning_rate": 2.3385265333570715e-06,
      "loss": 0.1808,
      "num_input_tokens_seen": 29826368,
      "step": 34450
    },
    {
      "epoch": 16.244695898161243,
      "grad_norm": 0.7550337910652161,
      "learning_rate": 2.334382419503278e-06,
      "loss": 0.1843,
      "num_input_tokens_seen": 29830912,
      "step": 34455
    },
    {
      "epoch": 16.247053276756247,
      "grad_norm": 0.5022950172424316,
      "learning_rate": 2.3302418009725465e-06,
      "loss": 0.1013,
      "num_input_tokens_seen": 29835872,
      "step": 34460
    },
    {
      "epoch": 16.249410655351248,
      "grad_norm": 0.6259359121322632,
      "learning_rate": 2.326104678403415e-06,
      "loss": 0.2075,
      "num_input_tokens_seen": 29839936,
      "step": 34465
    },
    {
      "epoch": 16.251768033946252,
      "grad_norm": 0.8234190344810486,
      "learning_rate": 2.321971052433883e-06,
      "loss": 0.16,
      "num_input_tokens_seen": 29844016,
      "step": 34470
    },
    {
      "epoch": 16.254125412541253,
      "grad_norm": 0.8515493273735046,
      "learning_rate": 2.3178409237014004e-06,
      "loss": 0.1876,
      "num_input_tokens_seen": 29848512,
      "step": 34475
    },
    {
      "epoch": 16.256482791136257,
      "grad_norm": 2.025806427001953,
      "learning_rate": 2.313714292842889e-06,
      "loss": 0.3016,
      "num_input_tokens_seen": 29853648,
      "step": 34480
    },
    {
      "epoch": 16.258840169731258,
      "grad_norm": 0.5732306241989136,
      "learning_rate": 2.309591160494734e-06,
      "loss": 0.1855,
      "num_input_tokens_seen": 29857552,
      "step": 34485
    },
    {
      "epoch": 16.261197548326262,
      "grad_norm": 0.7353267669677734,
      "learning_rate": 2.305471527292763e-06,
      "loss": 0.1493,
      "num_input_tokens_seen": 29861472,
      "step": 34490
    },
    {
      "epoch": 16.263554926921262,
      "grad_norm": 1.5660396814346313,
      "learning_rate": 2.3013553938722817e-06,
      "loss": 0.2118,
      "num_input_tokens_seen": 29865616,
      "step": 34495
    },
    {
      "epoch": 16.265912305516267,
      "grad_norm": 1.0086554288864136,
      "learning_rate": 2.297242760868043e-06,
      "loss": 0.2757,
      "num_input_tokens_seen": 29870112,
      "step": 34500
    },
    {
      "epoch": 16.268269684111267,
      "grad_norm": 1.862424373626709,
      "learning_rate": 2.2931336289142735e-06,
      "loss": 0.1825,
      "num_input_tokens_seen": 29873840,
      "step": 34505
    },
    {
      "epoch": 16.27062706270627,
      "grad_norm": 1.763946771621704,
      "learning_rate": 2.289027998644655e-06,
      "loss": 0.2266,
      "num_input_tokens_seen": 29878320,
      "step": 34510
    },
    {
      "epoch": 16.272984441301272,
      "grad_norm": 0.8478330373764038,
      "learning_rate": 2.2849258706923228e-06,
      "loss": 0.217,
      "num_input_tokens_seen": 29883280,
      "step": 34515
    },
    {
      "epoch": 16.275341819896276,
      "grad_norm": 0.7872339487075806,
      "learning_rate": 2.2808272456898705e-06,
      "loss": 0.1156,
      "num_input_tokens_seen": 29887600,
      "step": 34520
    },
    {
      "epoch": 16.277699198491277,
      "grad_norm": 0.45432743430137634,
      "learning_rate": 2.2767321242693707e-06,
      "loss": 0.217,
      "num_input_tokens_seen": 29891840,
      "step": 34525
    },
    {
      "epoch": 16.28005657708628,
      "grad_norm": 1.500696063041687,
      "learning_rate": 2.272640507062329e-06,
      "loss": 0.1523,
      "num_input_tokens_seen": 29896608,
      "step": 34530
    },
    {
      "epoch": 16.28241395568128,
      "grad_norm": 0.8707044720649719,
      "learning_rate": 2.2685523946997382e-06,
      "loss": 0.169,
      "num_input_tokens_seen": 29901792,
      "step": 34535
    },
    {
      "epoch": 16.284771334276286,
      "grad_norm": 1.1573920249938965,
      "learning_rate": 2.2644677878120245e-06,
      "loss": 0.1586,
      "num_input_tokens_seen": 29906048,
      "step": 34540
    },
    {
      "epoch": 16.287128712871286,
      "grad_norm": 0.7355061769485474,
      "learning_rate": 2.2603866870290897e-06,
      "loss": 0.1527,
      "num_input_tokens_seen": 29910000,
      "step": 34545
    },
    {
      "epoch": 16.28948609146629,
      "grad_norm": 0.8523956537246704,
      "learning_rate": 2.256309092980294e-06,
      "loss": 0.1268,
      "num_input_tokens_seen": 29914608,
      "step": 34550
    },
    {
      "epoch": 16.29184347006129,
      "grad_norm": 1.0973013639450073,
      "learning_rate": 2.252235006294448e-06,
      "loss": 0.1713,
      "num_input_tokens_seen": 29918528,
      "step": 34555
    },
    {
      "epoch": 16.294200848656295,
      "grad_norm": 2.1538445949554443,
      "learning_rate": 2.2481644275998333e-06,
      "loss": 0.1528,
      "num_input_tokens_seen": 29923232,
      "step": 34560
    },
    {
      "epoch": 16.296558227251296,
      "grad_norm": 1.3003300428390503,
      "learning_rate": 2.2440973575241832e-06,
      "loss": 0.1816,
      "num_input_tokens_seen": 29927520,
      "step": 34565
    },
    {
      "epoch": 16.2989156058463,
      "grad_norm": 0.838982105255127,
      "learning_rate": 2.240033796694685e-06,
      "loss": 0.228,
      "num_input_tokens_seen": 29932176,
      "step": 34570
    },
    {
      "epoch": 16.3012729844413,
      "grad_norm": 0.6009277105331421,
      "learning_rate": 2.235973745737999e-06,
      "loss": 0.1859,
      "num_input_tokens_seen": 29936880,
      "step": 34575
    },
    {
      "epoch": 16.303630363036305,
      "grad_norm": 0.8560577630996704,
      "learning_rate": 2.2319172052802263e-06,
      "loss": 0.1843,
      "num_input_tokens_seen": 29941360,
      "step": 34580
    },
    {
      "epoch": 16.305987741631306,
      "grad_norm": 1.1267781257629395,
      "learning_rate": 2.2278641759469477e-06,
      "loss": 0.164,
      "num_input_tokens_seen": 29945920,
      "step": 34585
    },
    {
      "epoch": 16.30834512022631,
      "grad_norm": 1.1888340711593628,
      "learning_rate": 2.2238146583631825e-06,
      "loss": 0.1206,
      "num_input_tokens_seen": 29950912,
      "step": 34590
    },
    {
      "epoch": 16.31070249882131,
      "grad_norm": 0.7853846549987793,
      "learning_rate": 2.2197686531534256e-06,
      "loss": 0.1541,
      "num_input_tokens_seen": 29955408,
      "step": 34595
    },
    {
      "epoch": 16.313059877416315,
      "grad_norm": 1.7029612064361572,
      "learning_rate": 2.2157261609416087e-06,
      "loss": 0.196,
      "num_input_tokens_seen": 29959568,
      "step": 34600
    },
    {
      "epoch": 16.313059877416315,
      "eval_loss": 0.20031362771987915,
      "eval_runtime": 22.186,
      "eval_samples_per_second": 42.504,
      "eval_steps_per_second": 21.275,
      "num_input_tokens_seen": 29959568,
      "step": 34600
    },
    {
      "epoch": 16.315417256011315,
      "grad_norm": 0.4116653800010681,
      "learning_rate": 2.211687182351149e-06,
      "loss": 0.2806,
      "num_input_tokens_seen": 29964400,
      "step": 34605
    },
    {
      "epoch": 16.31777463460632,
      "grad_norm": 2.210900068283081,
      "learning_rate": 2.2076517180048993e-06,
      "loss": 0.2072,
      "num_input_tokens_seen": 29968304,
      "step": 34610
    },
    {
      "epoch": 16.32013201320132,
      "grad_norm": 2.018238067626953,
      "learning_rate": 2.2036197685251834e-06,
      "loss": 0.192,
      "num_input_tokens_seen": 29973008,
      "step": 34615
    },
    {
      "epoch": 16.32248939179632,
      "grad_norm": 1.5954220294952393,
      "learning_rate": 2.199591334533771e-06,
      "loss": 0.1691,
      "num_input_tokens_seen": 29977168,
      "step": 34620
    },
    {
      "epoch": 16.324846770391325,
      "grad_norm": 0.5975632071495056,
      "learning_rate": 2.1955664166519036e-06,
      "loss": 0.2405,
      "num_input_tokens_seen": 29981760,
      "step": 34625
    },
    {
      "epoch": 16.327204148986326,
      "grad_norm": 0.20886825025081635,
      "learning_rate": 2.1915450155002793e-06,
      "loss": 0.0846,
      "num_input_tokens_seen": 29985456,
      "step": 34630
    },
    {
      "epoch": 16.32956152758133,
      "grad_norm": 0.7721323370933533,
      "learning_rate": 2.187527131699038e-06,
      "loss": 0.1765,
      "num_input_tokens_seen": 29989616,
      "step": 34635
    },
    {
      "epoch": 16.33191890617633,
      "grad_norm": 0.4571261405944824,
      "learning_rate": 2.18351276586779e-06,
      "loss": 0.1597,
      "num_input_tokens_seen": 29994160,
      "step": 34640
    },
    {
      "epoch": 16.334276284771335,
      "grad_norm": 0.9668379426002502,
      "learning_rate": 2.1795019186256092e-06,
      "loss": 0.1272,
      "num_input_tokens_seen": 29998432,
      "step": 34645
    },
    {
      "epoch": 16.336633663366335,
      "grad_norm": 2.1537210941314697,
      "learning_rate": 2.1754945905910094e-06,
      "loss": 0.1975,
      "num_input_tokens_seen": 30002992,
      "step": 34650
    },
    {
      "epoch": 16.33899104196134,
      "grad_norm": 1.1268161535263062,
      "learning_rate": 2.171490782381977e-06,
      "loss": 0.2123,
      "num_input_tokens_seen": 30006832,
      "step": 34655
    },
    {
      "epoch": 16.34134842055634,
      "grad_norm": 1.5456955432891846,
      "learning_rate": 2.1674904946159425e-06,
      "loss": 0.2365,
      "num_input_tokens_seen": 30010400,
      "step": 34660
    },
    {
      "epoch": 16.343705799151344,
      "grad_norm": 1.947285771369934,
      "learning_rate": 2.16349372790981e-06,
      "loss": 0.2196,
      "num_input_tokens_seen": 30014880,
      "step": 34665
    },
    {
      "epoch": 16.346063177746345,
      "grad_norm": 0.9029040336608887,
      "learning_rate": 2.159500482879928e-06,
      "loss": 0.16,
      "num_input_tokens_seen": 30019328,
      "step": 34670
    },
    {
      "epoch": 16.34842055634135,
      "grad_norm": 1.2191072702407837,
      "learning_rate": 2.155510760142096e-06,
      "loss": 0.2093,
      "num_input_tokens_seen": 30023888,
      "step": 34675
    },
    {
      "epoch": 16.35077793493635,
      "grad_norm": 0.9514925479888916,
      "learning_rate": 2.151524560311588e-06,
      "loss": 0.1685,
      "num_input_tokens_seen": 30030016,
      "step": 34680
    },
    {
      "epoch": 16.353135313531354,
      "grad_norm": 0.6397725939750671,
      "learning_rate": 2.147541884003129e-06,
      "loss": 0.1639,
      "num_input_tokens_seen": 30034256,
      "step": 34685
    },
    {
      "epoch": 16.355492692126354,
      "grad_norm": 1.0361204147338867,
      "learning_rate": 2.1435627318308895e-06,
      "loss": 0.2028,
      "num_input_tokens_seen": 30039488,
      "step": 34690
    },
    {
      "epoch": 16.35785007072136,
      "grad_norm": 1.1155987977981567,
      "learning_rate": 2.139587104408511e-06,
      "loss": 0.1643,
      "num_input_tokens_seen": 30044512,
      "step": 34695
    },
    {
      "epoch": 16.36020744931636,
      "grad_norm": 1.0538123846054077,
      "learning_rate": 2.1356150023490783e-06,
      "loss": 0.1452,
      "num_input_tokens_seen": 30048272,
      "step": 34700
    },
    {
      "epoch": 16.362564827911363,
      "grad_norm": 0.40530553460121155,
      "learning_rate": 2.1316464262651464e-06,
      "loss": 0.1151,
      "num_input_tokens_seen": 30052768,
      "step": 34705
    },
    {
      "epoch": 16.364922206506364,
      "grad_norm": 0.9354806542396545,
      "learning_rate": 2.1276813767687224e-06,
      "loss": 0.1845,
      "num_input_tokens_seen": 30057568,
      "step": 34710
    },
    {
      "epoch": 16.367279585101368,
      "grad_norm": 0.7525047063827515,
      "learning_rate": 2.123719854471254e-06,
      "loss": 0.1624,
      "num_input_tokens_seen": 30061856,
      "step": 34715
    },
    {
      "epoch": 16.36963696369637,
      "grad_norm": 0.9816902279853821,
      "learning_rate": 2.119761859983668e-06,
      "loss": 0.1594,
      "num_input_tokens_seen": 30066704,
      "step": 34720
    },
    {
      "epoch": 16.371994342291373,
      "grad_norm": 0.9178110361099243,
      "learning_rate": 2.1158073939163386e-06,
      "loss": 0.1565,
      "num_input_tokens_seen": 30071040,
      "step": 34725
    },
    {
      "epoch": 16.374351720886374,
      "grad_norm": 1.1991541385650635,
      "learning_rate": 2.111856456879088e-06,
      "loss": 0.1634,
      "num_input_tokens_seen": 30075776,
      "step": 34730
    },
    {
      "epoch": 16.376709099481378,
      "grad_norm": 1.5016785860061646,
      "learning_rate": 2.1079090494811993e-06,
      "loss": 0.2129,
      "num_input_tokens_seen": 30079520,
      "step": 34735
    },
    {
      "epoch": 16.37906647807638,
      "grad_norm": 0.6030968427658081,
      "learning_rate": 2.103965172331418e-06,
      "loss": 0.1498,
      "num_input_tokens_seen": 30083888,
      "step": 34740
    },
    {
      "epoch": 16.381423856671383,
      "grad_norm": 1.2452818155288696,
      "learning_rate": 2.100024826037933e-06,
      "loss": 0.1792,
      "num_input_tokens_seen": 30087936,
      "step": 34745
    },
    {
      "epoch": 16.383781235266383,
      "grad_norm": 1.5842418670654297,
      "learning_rate": 2.0960880112084027e-06,
      "loss": 0.138,
      "num_input_tokens_seen": 30092272,
      "step": 34750
    },
    {
      "epoch": 16.386138613861387,
      "grad_norm": 1.927345633506775,
      "learning_rate": 2.092154728449927e-06,
      "loss": 0.1397,
      "num_input_tokens_seen": 30096144,
      "step": 34755
    },
    {
      "epoch": 16.388495992456388,
      "grad_norm": 2.735520362854004,
      "learning_rate": 2.0882249783690687e-06,
      "loss": 0.3154,
      "num_input_tokens_seen": 30100464,
      "step": 34760
    },
    {
      "epoch": 16.390853371051392,
      "grad_norm": 1.4439148902893066,
      "learning_rate": 2.084298761571851e-06,
      "loss": 0.2135,
      "num_input_tokens_seen": 30104960,
      "step": 34765
    },
    {
      "epoch": 16.393210749646393,
      "grad_norm": 1.207903504371643,
      "learning_rate": 2.080376078663737e-06,
      "loss": 0.217,
      "num_input_tokens_seen": 30109552,
      "step": 34770
    },
    {
      "epoch": 16.395568128241397,
      "grad_norm": 1.046940565109253,
      "learning_rate": 2.0764569302496593e-06,
      "loss": 0.1091,
      "num_input_tokens_seen": 30113648,
      "step": 34775
    },
    {
      "epoch": 16.397925506836398,
      "grad_norm": 1.5091978311538696,
      "learning_rate": 2.0725413169339957e-06,
      "loss": 0.1686,
      "num_input_tokens_seen": 30118576,
      "step": 34780
    },
    {
      "epoch": 16.400282885431402,
      "grad_norm": 0.6699693202972412,
      "learning_rate": 2.068629239320588e-06,
      "loss": 0.2472,
      "num_input_tokens_seen": 30122752,
      "step": 34785
    },
    {
      "epoch": 16.402640264026402,
      "grad_norm": 0.8943654894828796,
      "learning_rate": 2.064720698012726e-06,
      "loss": 0.1162,
      "num_input_tokens_seen": 30127472,
      "step": 34790
    },
    {
      "epoch": 16.404997642621407,
      "grad_norm": 1.3557220697402954,
      "learning_rate": 2.0608156936131522e-06,
      "loss": 0.182,
      "num_input_tokens_seen": 30131008,
      "step": 34795
    },
    {
      "epoch": 16.407355021216407,
      "grad_norm": 0.8103132247924805,
      "learning_rate": 2.056914226724074e-06,
      "loss": 0.1149,
      "num_input_tokens_seen": 30134704,
      "step": 34800
    },
    {
      "epoch": 16.407355021216407,
      "eval_loss": 0.1998654454946518,
      "eval_runtime": 22.1334,
      "eval_samples_per_second": 42.605,
      "eval_steps_per_second": 21.325,
      "num_input_tokens_seen": 30134704,
      "step": 34800
    },
    {
      "epoch": 16.40971239981141,
      "grad_norm": 1.4431648254394531,
      "learning_rate": 2.0530162979471385e-06,
      "loss": 0.2173,
      "num_input_tokens_seen": 30138832,
      "step": 34805
    },
    {
      "epoch": 16.412069778406412,
      "grad_norm": 0.6878511905670166,
      "learning_rate": 2.0491219078834667e-06,
      "loss": 0.132,
      "num_input_tokens_seen": 30143408,
      "step": 34810
    },
    {
      "epoch": 16.414427157001413,
      "grad_norm": 1.4293811321258545,
      "learning_rate": 2.045231057133612e-06,
      "loss": 0.138,
      "num_input_tokens_seen": 30148176,
      "step": 34815
    },
    {
      "epoch": 16.416784535596417,
      "grad_norm": 0.7006293535232544,
      "learning_rate": 2.0413437462975944e-06,
      "loss": 0.2407,
      "num_input_tokens_seen": 30152240,
      "step": 34820
    },
    {
      "epoch": 16.419141914191417,
      "grad_norm": 0.41295313835144043,
      "learning_rate": 2.0374599759748843e-06,
      "loss": 0.1547,
      "num_input_tokens_seen": 30156832,
      "step": 34825
    },
    {
      "epoch": 16.42149929278642,
      "grad_norm": 0.9397628307342529,
      "learning_rate": 2.033579746764419e-06,
      "loss": 0.2188,
      "num_input_tokens_seen": 30161280,
      "step": 34830
    },
    {
      "epoch": 16.423856671381422,
      "grad_norm": 1.1977026462554932,
      "learning_rate": 2.029703059264565e-06,
      "loss": 0.1803,
      "num_input_tokens_seen": 30166240,
      "step": 34835
    },
    {
      "epoch": 16.426214049976426,
      "grad_norm": 1.8940140008926392,
      "learning_rate": 2.02582991407316e-06,
      "loss": 0.1694,
      "num_input_tokens_seen": 30170624,
      "step": 34840
    },
    {
      "epoch": 16.428571428571427,
      "grad_norm": 1.4291698932647705,
      "learning_rate": 2.0219603117874992e-06,
      "loss": 0.1989,
      "num_input_tokens_seen": 30174816,
      "step": 34845
    },
    {
      "epoch": 16.43092880716643,
      "grad_norm": 0.5932263731956482,
      "learning_rate": 2.0180942530043156e-06,
      "loss": 0.103,
      "num_input_tokens_seen": 30179520,
      "step": 34850
    },
    {
      "epoch": 16.433286185761432,
      "grad_norm": 1.5076775550842285,
      "learning_rate": 2.0142317383198107e-06,
      "loss": 0.1386,
      "num_input_tokens_seen": 30183328,
      "step": 34855
    },
    {
      "epoch": 16.435643564356436,
      "grad_norm": 1.1493390798568726,
      "learning_rate": 2.0103727683296243e-06,
      "loss": 0.2162,
      "num_input_tokens_seen": 30187344,
      "step": 34860
    },
    {
      "epoch": 16.438000942951437,
      "grad_norm": 0.7067590355873108,
      "learning_rate": 2.0065173436288636e-06,
      "loss": 0.1214,
      "num_input_tokens_seen": 30191168,
      "step": 34865
    },
    {
      "epoch": 16.44035832154644,
      "grad_norm": 1.9109830856323242,
      "learning_rate": 2.002665464812087e-06,
      "loss": 0.1869,
      "num_input_tokens_seen": 30195776,
      "step": 34870
    },
    {
      "epoch": 16.44271570014144,
      "grad_norm": 0.8581770062446594,
      "learning_rate": 1.998817132473291e-06,
      "loss": 0.1329,
      "num_input_tokens_seen": 30199904,
      "step": 34875
    },
    {
      "epoch": 16.445073078736446,
      "grad_norm": 0.3555780351161957,
      "learning_rate": 1.9949723472059507e-06,
      "loss": 0.166,
      "num_input_tokens_seen": 30204144,
      "step": 34880
    },
    {
      "epoch": 16.447430457331446,
      "grad_norm": 1.198894739151001,
      "learning_rate": 1.9911311096029726e-06,
      "loss": 0.1775,
      "num_input_tokens_seen": 30208304,
      "step": 34885
    },
    {
      "epoch": 16.44978783592645,
      "grad_norm": 1.3850700855255127,
      "learning_rate": 1.9872934202567224e-06,
      "loss": 0.1762,
      "num_input_tokens_seen": 30213600,
      "step": 34890
    },
    {
      "epoch": 16.45214521452145,
      "grad_norm": 1.134480357170105,
      "learning_rate": 1.9834592797590257e-06,
      "loss": 0.2303,
      "num_input_tokens_seen": 30217296,
      "step": 34895
    },
    {
      "epoch": 16.454502593116455,
      "grad_norm": 0.5152111649513245,
      "learning_rate": 1.979628688701149e-06,
      "loss": 0.1249,
      "num_input_tokens_seen": 30222096,
      "step": 34900
    },
    {
      "epoch": 16.456859971711456,
      "grad_norm": 1.3728718757629395,
      "learning_rate": 1.9758016476738193e-06,
      "loss": 0.2094,
      "num_input_tokens_seen": 30226448,
      "step": 34905
    },
    {
      "epoch": 16.45921735030646,
      "grad_norm": 1.230920672416687,
      "learning_rate": 1.971978157267221e-06,
      "loss": 0.1535,
      "num_input_tokens_seen": 30230176,
      "step": 34910
    },
    {
      "epoch": 16.46157472890146,
      "grad_norm": 0.8956285715103149,
      "learning_rate": 1.968158218070973e-06,
      "loss": 0.2167,
      "num_input_tokens_seen": 30234080,
      "step": 34915
    },
    {
      "epoch": 16.463932107496465,
      "grad_norm": 0.45111536979675293,
      "learning_rate": 1.9643418306741682e-06,
      "loss": 0.2009,
      "num_input_tokens_seen": 30238112,
      "step": 34920
    },
    {
      "epoch": 16.466289486091465,
      "grad_norm": 0.7384512424468994,
      "learning_rate": 1.9605289956653337e-06,
      "loss": 0.1722,
      "num_input_tokens_seen": 30242384,
      "step": 34925
    },
    {
      "epoch": 16.46864686468647,
      "grad_norm": 2.098154306411743,
      "learning_rate": 1.9567197136324626e-06,
      "loss": 0.1537,
      "num_input_tokens_seen": 30246880,
      "step": 34930
    },
    {
      "epoch": 16.47100424328147,
      "grad_norm": 0.8276849985122681,
      "learning_rate": 1.9529139851629935e-06,
      "loss": 0.2506,
      "num_input_tokens_seen": 30251216,
      "step": 34935
    },
    {
      "epoch": 16.473361621876474,
      "grad_norm": 0.9544995427131653,
      "learning_rate": 1.949111810843812e-06,
      "loss": 0.2007,
      "num_input_tokens_seen": 30254464,
      "step": 34940
    },
    {
      "epoch": 16.475719000471475,
      "grad_norm": 1.1980594396591187,
      "learning_rate": 1.9453131912612694e-06,
      "loss": 0.1756,
      "num_input_tokens_seen": 30259136,
      "step": 34945
    },
    {
      "epoch": 16.47807637906648,
      "grad_norm": 1.3281657695770264,
      "learning_rate": 1.941518127001149e-06,
      "loss": 0.106,
      "num_input_tokens_seen": 30263200,
      "step": 34950
    },
    {
      "epoch": 16.48043375766148,
      "grad_norm": 1.4679453372955322,
      "learning_rate": 1.9377266186487107e-06,
      "loss": 0.1543,
      "num_input_tokens_seen": 30268096,
      "step": 34955
    },
    {
      "epoch": 16.482791136256484,
      "grad_norm": 0.7382614016532898,
      "learning_rate": 1.9339386667886483e-06,
      "loss": 0.1895,
      "num_input_tokens_seen": 30272480,
      "step": 34960
    },
    {
      "epoch": 16.485148514851485,
      "grad_norm": 0.7467886805534363,
      "learning_rate": 1.9301542720051024e-06,
      "loss": 0.2067,
      "num_input_tokens_seen": 30276624,
      "step": 34965
    },
    {
      "epoch": 16.48750589344649,
      "grad_norm": 0.8646796345710754,
      "learning_rate": 1.926373434881684e-06,
      "loss": 0.1754,
      "num_input_tokens_seen": 30280544,
      "step": 34970
    },
    {
      "epoch": 16.48986327204149,
      "grad_norm": 0.7184978723526001,
      "learning_rate": 1.9225961560014468e-06,
      "loss": 0.1469,
      "num_input_tokens_seen": 30284112,
      "step": 34975
    },
    {
      "epoch": 16.492220650636494,
      "grad_norm": 1.4502986669540405,
      "learning_rate": 1.918822435946885e-06,
      "loss": 0.24,
      "num_input_tokens_seen": 30288672,
      "step": 34980
    },
    {
      "epoch": 16.494578029231494,
      "grad_norm": 0.799526035785675,
      "learning_rate": 1.915052275299961e-06,
      "loss": 0.1041,
      "num_input_tokens_seen": 30293120,
      "step": 34985
    },
    {
      "epoch": 16.4969354078265,
      "grad_norm": 1.0916309356689453,
      "learning_rate": 1.9112856746420854e-06,
      "loss": 0.2271,
      "num_input_tokens_seen": 30297040,
      "step": 34990
    },
    {
      "epoch": 16.4992927864215,
      "grad_norm": 1.5676864385604858,
      "learning_rate": 1.907522634554104e-06,
      "loss": 0.1406,
      "num_input_tokens_seen": 30301328,
      "step": 34995
    },
    {
      "epoch": 16.501650165016503,
      "grad_norm": 1.4235807657241821,
      "learning_rate": 1.9037631556163337e-06,
      "loss": 0.2473,
      "num_input_tokens_seen": 30305200,
      "step": 35000
    },
    {
      "epoch": 16.501650165016503,
      "eval_loss": 0.19977876543998718,
      "eval_runtime": 22.1373,
      "eval_samples_per_second": 42.598,
      "eval_steps_per_second": 21.321,
      "num_input_tokens_seen": 30305200,
      "step": 35000
    },
    {
      "epoch": 16.504007543611504,
      "grad_norm": 0.6613298654556274,
      "learning_rate": 1.9000072384085272e-06,
      "loss": 0.1754,
      "num_input_tokens_seen": 30309360,
      "step": 35005
    },
    {
      "epoch": 16.506364922206508,
      "grad_norm": 1.0866564512252808,
      "learning_rate": 1.8962548835098987e-06,
      "loss": 0.154,
      "num_input_tokens_seen": 30314256,
      "step": 35010
    },
    {
      "epoch": 16.50872230080151,
      "grad_norm": 0.9102181196212769,
      "learning_rate": 1.8925060914991077e-06,
      "loss": 0.2007,
      "num_input_tokens_seen": 30318544,
      "step": 35015
    },
    {
      "epoch": 16.51107967939651,
      "grad_norm": 2.711836099624634,
      "learning_rate": 1.888760862954264e-06,
      "loss": 0.3217,
      "num_input_tokens_seen": 30322432,
      "step": 35020
    },
    {
      "epoch": 16.513437057991514,
      "grad_norm": 0.6073903441429138,
      "learning_rate": 1.8850191984529309e-06,
      "loss": 0.1625,
      "num_input_tokens_seen": 30326176,
      "step": 35025
    },
    {
      "epoch": 16.515794436586514,
      "grad_norm": 0.9895659685134888,
      "learning_rate": 1.8812810985721186e-06,
      "loss": 0.1406,
      "num_input_tokens_seen": 30330944,
      "step": 35030
    },
    {
      "epoch": 16.51815181518152,
      "grad_norm": 0.8447162508964539,
      "learning_rate": 1.8775465638882856e-06,
      "loss": 0.1714,
      "num_input_tokens_seen": 30334944,
      "step": 35035
    },
    {
      "epoch": 16.52050919377652,
      "grad_norm": 0.7165077924728394,
      "learning_rate": 1.8738155949773517e-06,
      "loss": 0.1746,
      "num_input_tokens_seen": 30338688,
      "step": 35040
    },
    {
      "epoch": 16.522866572371523,
      "grad_norm": 1.046647548675537,
      "learning_rate": 1.8700881924146707e-06,
      "loss": 0.1857,
      "num_input_tokens_seen": 30342880,
      "step": 35045
    },
    {
      "epoch": 16.525223950966524,
      "grad_norm": 2.3394715785980225,
      "learning_rate": 1.8663643567750577e-06,
      "loss": 0.2017,
      "num_input_tokens_seen": 30347168,
      "step": 35050
    },
    {
      "epoch": 16.527581329561528,
      "grad_norm": 1.7281910181045532,
      "learning_rate": 1.8626440886327813e-06,
      "loss": 0.1984,
      "num_input_tokens_seen": 30351568,
      "step": 35055
    },
    {
      "epoch": 16.52993870815653,
      "grad_norm": 0.8443936109542847,
      "learning_rate": 1.8589273885615432e-06,
      "loss": 0.1358,
      "num_input_tokens_seen": 30356048,
      "step": 35060
    },
    {
      "epoch": 16.532296086751533,
      "grad_norm": 1.6207983493804932,
      "learning_rate": 1.8552142571345133e-06,
      "loss": 0.1599,
      "num_input_tokens_seen": 30360112,
      "step": 35065
    },
    {
      "epoch": 16.534653465346533,
      "grad_norm": 1.1766784191131592,
      "learning_rate": 1.8515046949243025e-06,
      "loss": 0.2067,
      "num_input_tokens_seen": 30364288,
      "step": 35070
    },
    {
      "epoch": 16.537010843941538,
      "grad_norm": 0.5598084330558777,
      "learning_rate": 1.8477987025029674e-06,
      "loss": 0.1909,
      "num_input_tokens_seen": 30369280,
      "step": 35075
    },
    {
      "epoch": 16.539368222536538,
      "grad_norm": 1.503021001815796,
      "learning_rate": 1.8440962804420232e-06,
      "loss": 0.2147,
      "num_input_tokens_seen": 30373584,
      "step": 35080
    },
    {
      "epoch": 16.541725601131542,
      "grad_norm": 1.1689167022705078,
      "learning_rate": 1.8403974293124265e-06,
      "loss": 0.1794,
      "num_input_tokens_seen": 30378016,
      "step": 35085
    },
    {
      "epoch": 16.544082979726543,
      "grad_norm": 1.1804717779159546,
      "learning_rate": 1.8367021496845854e-06,
      "loss": 0.1664,
      "num_input_tokens_seen": 30382080,
      "step": 35090
    },
    {
      "epoch": 16.546440358321547,
      "grad_norm": 0.6453076601028442,
      "learning_rate": 1.8330104421283662e-06,
      "loss": 0.1896,
      "num_input_tokens_seen": 30386656,
      "step": 35095
    },
    {
      "epoch": 16.548797736916548,
      "grad_norm": 3.580911874771118,
      "learning_rate": 1.8293223072130717e-06,
      "loss": 0.2344,
      "num_input_tokens_seen": 30391568,
      "step": 35100
    },
    {
      "epoch": 16.551155115511552,
      "grad_norm": 1.9084986448287964,
      "learning_rate": 1.8256377455074525e-06,
      "loss": 0.2287,
      "num_input_tokens_seen": 30395712,
      "step": 35105
    },
    {
      "epoch": 16.553512494106553,
      "grad_norm": 0.5173895955085754,
      "learning_rate": 1.8219567575797263e-06,
      "loss": 0.1561,
      "num_input_tokens_seen": 30400624,
      "step": 35110
    },
    {
      "epoch": 16.555869872701557,
      "grad_norm": 0.658839225769043,
      "learning_rate": 1.8182793439975365e-06,
      "loss": 0.1724,
      "num_input_tokens_seen": 30404512,
      "step": 35115
    },
    {
      "epoch": 16.558227251296557,
      "grad_norm": 0.6208469867706299,
      "learning_rate": 1.8146055053279958e-06,
      "loss": 0.2093,
      "num_input_tokens_seen": 30410432,
      "step": 35120
    },
    {
      "epoch": 16.56058462989156,
      "grad_norm": 2.130082845687866,
      "learning_rate": 1.8109352421376486e-06,
      "loss": 0.166,
      "num_input_tokens_seen": 30415120,
      "step": 35125
    },
    {
      "epoch": 16.562942008486562,
      "grad_norm": 0.836676836013794,
      "learning_rate": 1.8072685549924972e-06,
      "loss": 0.1395,
      "num_input_tokens_seen": 30419056,
      "step": 35130
    },
    {
      "epoch": 16.565299387081566,
      "grad_norm": 0.4372972846031189,
      "learning_rate": 1.8036054444579982e-06,
      "loss": 0.1125,
      "num_input_tokens_seen": 30423296,
      "step": 35135
    },
    {
      "epoch": 16.567656765676567,
      "grad_norm": 1.3402754068374634,
      "learning_rate": 1.7999459110990407e-06,
      "loss": 0.1757,
      "num_input_tokens_seen": 30427296,
      "step": 35140
    },
    {
      "epoch": 16.57001414427157,
      "grad_norm": 0.36610284447669983,
      "learning_rate": 1.7962899554799712e-06,
      "loss": 0.2448,
      "num_input_tokens_seen": 30431152,
      "step": 35145
    },
    {
      "epoch": 16.572371522866572,
      "grad_norm": 1.4988073110580444,
      "learning_rate": 1.7926375781645937e-06,
      "loss": 0.1669,
      "num_input_tokens_seen": 30435728,
      "step": 35150
    },
    {
      "epoch": 16.574728901461576,
      "grad_norm": 1.5439804792404175,
      "learning_rate": 1.7889887797161359e-06,
      "loss": 0.2392,
      "num_input_tokens_seen": 30440368,
      "step": 35155
    },
    {
      "epoch": 16.577086280056577,
      "grad_norm": 0.9898727536201477,
      "learning_rate": 1.7853435606973028e-06,
      "loss": 0.15,
      "num_input_tokens_seen": 30444336,
      "step": 35160
    },
    {
      "epoch": 16.57944365865158,
      "grad_norm": 1.902092456817627,
      "learning_rate": 1.781701921670223e-06,
      "loss": 0.1713,
      "num_input_tokens_seen": 30448768,
      "step": 35165
    },
    {
      "epoch": 16.58180103724658,
      "grad_norm": 1.1389875411987305,
      "learning_rate": 1.7780638631964886e-06,
      "loss": 0.1782,
      "num_input_tokens_seen": 30453632,
      "step": 35170
    },
    {
      "epoch": 16.584158415841586,
      "grad_norm": 0.2792969048023224,
      "learning_rate": 1.7744293858371314e-06,
      "loss": 0.1084,
      "num_input_tokens_seen": 30458192,
      "step": 35175
    },
    {
      "epoch": 16.586515794436586,
      "grad_norm": 0.5264815092086792,
      "learning_rate": 1.770798490152631e-06,
      "loss": 0.1249,
      "num_input_tokens_seen": 30461632,
      "step": 35180
    },
    {
      "epoch": 16.58887317303159,
      "grad_norm": 0.5877558588981628,
      "learning_rate": 1.767171176702917e-06,
      "loss": 0.1781,
      "num_input_tokens_seen": 30466064,
      "step": 35185
    },
    {
      "epoch": 16.59123055162659,
      "grad_norm": 0.4696628451347351,
      "learning_rate": 1.7635474460473755e-06,
      "loss": 0.2116,
      "num_input_tokens_seen": 30470256,
      "step": 35190
    },
    {
      "epoch": 16.593587930221595,
      "grad_norm": 0.5133942365646362,
      "learning_rate": 1.7599272987448206e-06,
      "loss": 0.163,
      "num_input_tokens_seen": 30474096,
      "step": 35195
    },
    {
      "epoch": 16.595945308816596,
      "grad_norm": 0.5021618604660034,
      "learning_rate": 1.7563107353535362e-06,
      "loss": 0.1653,
      "num_input_tokens_seen": 30478576,
      "step": 35200
    },
    {
      "epoch": 16.595945308816596,
      "eval_loss": 0.2003721445798874,
      "eval_runtime": 22.1167,
      "eval_samples_per_second": 42.637,
      "eval_steps_per_second": 21.341,
      "num_input_tokens_seen": 30478576,
      "step": 35200
    },
    {
      "epoch": 16.5983026874116,
      "grad_norm": 1.4453283548355103,
      "learning_rate": 1.7526977564312263e-06,
      "loss": 0.1372,
      "num_input_tokens_seen": 30483200,
      "step": 35205
    },
    {
      "epoch": 16.6006600660066,
      "grad_norm": 0.7236012816429138,
      "learning_rate": 1.7490883625350701e-06,
      "loss": 0.14,
      "num_input_tokens_seen": 30486960,
      "step": 35210
    },
    {
      "epoch": 16.603017444601605,
      "grad_norm": 0.8125436902046204,
      "learning_rate": 1.7454825542216807e-06,
      "loss": 0.1385,
      "num_input_tokens_seen": 30491792,
      "step": 35215
    },
    {
      "epoch": 16.605374823196605,
      "grad_norm": 0.9740585088729858,
      "learning_rate": 1.7418803320471105e-06,
      "loss": 0.1641,
      "num_input_tokens_seen": 30495712,
      "step": 35220
    },
    {
      "epoch": 16.607732201791606,
      "grad_norm": 1.8671470880508423,
      "learning_rate": 1.7382816965668737e-06,
      "loss": 0.224,
      "num_input_tokens_seen": 30500096,
      "step": 35225
    },
    {
      "epoch": 16.61008958038661,
      "grad_norm": 3.0054168701171875,
      "learning_rate": 1.7346866483359285e-06,
      "loss": 0.2382,
      "num_input_tokens_seen": 30505152,
      "step": 35230
    },
    {
      "epoch": 16.61244695898161,
      "grad_norm": 0.8352298736572266,
      "learning_rate": 1.7310951879086657e-06,
      "loss": 0.1458,
      "num_input_tokens_seen": 30509504,
      "step": 35235
    },
    {
      "epoch": 16.614804337576615,
      "grad_norm": 0.9138438701629639,
      "learning_rate": 1.7275073158389471e-06,
      "loss": 0.142,
      "num_input_tokens_seen": 30513616,
      "step": 35240
    },
    {
      "epoch": 16.617161716171616,
      "grad_norm": 0.8703700304031372,
      "learning_rate": 1.723923032680061e-06,
      "loss": 0.1484,
      "num_input_tokens_seen": 30517792,
      "step": 35245
    },
    {
      "epoch": 16.61951909476662,
      "grad_norm": 1.2299696207046509,
      "learning_rate": 1.7203423389847428e-06,
      "loss": 0.2129,
      "num_input_tokens_seen": 30522896,
      "step": 35250
    },
    {
      "epoch": 16.62187647336162,
      "grad_norm": 0.2930647134780884,
      "learning_rate": 1.7167652353051928e-06,
      "loss": 0.0934,
      "num_input_tokens_seen": 30527232,
      "step": 35255
    },
    {
      "epoch": 16.624233851956625,
      "grad_norm": 0.6620405316352844,
      "learning_rate": 1.7131917221930333e-06,
      "loss": 0.1592,
      "num_input_tokens_seen": 30531872,
      "step": 35260
    },
    {
      "epoch": 16.626591230551625,
      "grad_norm": 3.0913145542144775,
      "learning_rate": 1.7096218001993513e-06,
      "loss": 0.2188,
      "num_input_tokens_seen": 30536256,
      "step": 35265
    },
    {
      "epoch": 16.62894860914663,
      "grad_norm": 0.6056215167045593,
      "learning_rate": 1.706055469874676e-06,
      "loss": 0.1646,
      "num_input_tokens_seen": 30540528,
      "step": 35270
    },
    {
      "epoch": 16.63130598774163,
      "grad_norm": 0.9327316880226135,
      "learning_rate": 1.702492731768976e-06,
      "loss": 0.201,
      "num_input_tokens_seen": 30544784,
      "step": 35275
    },
    {
      "epoch": 16.633663366336634,
      "grad_norm": 2.0385241508483887,
      "learning_rate": 1.6989335864316724e-06,
      "loss": 0.1895,
      "num_input_tokens_seen": 30548768,
      "step": 35280
    },
    {
      "epoch": 16.636020744931635,
      "grad_norm": 0.7378991842269897,
      "learning_rate": 1.6953780344116265e-06,
      "loss": 0.1886,
      "num_input_tokens_seen": 30552656,
      "step": 35285
    },
    {
      "epoch": 16.63837812352664,
      "grad_norm": 1.7187844514846802,
      "learning_rate": 1.6918260762571497e-06,
      "loss": 0.1471,
      "num_input_tokens_seen": 30556960,
      "step": 35290
    },
    {
      "epoch": 16.64073550212164,
      "grad_norm": 1.4384281635284424,
      "learning_rate": 1.6882777125160093e-06,
      "loss": 0.1301,
      "num_input_tokens_seen": 30561088,
      "step": 35295
    },
    {
      "epoch": 16.643092880716644,
      "grad_norm": 0.9699465036392212,
      "learning_rate": 1.6847329437353899e-06,
      "loss": 0.1963,
      "num_input_tokens_seen": 30564768,
      "step": 35300
    },
    {
      "epoch": 16.645450259311644,
      "grad_norm": 0.811162531375885,
      "learning_rate": 1.6811917704619511e-06,
      "loss": 0.1732,
      "num_input_tokens_seen": 30569632,
      "step": 35305
    },
    {
      "epoch": 16.64780763790665,
      "grad_norm": 0.917967677116394,
      "learning_rate": 1.67765419324179e-06,
      "loss": 0.1894,
      "num_input_tokens_seen": 30574224,
      "step": 35310
    },
    {
      "epoch": 16.65016501650165,
      "grad_norm": 0.331735223531723,
      "learning_rate": 1.6741202126204364e-06,
      "loss": 0.1621,
      "num_input_tokens_seen": 30578512,
      "step": 35315
    },
    {
      "epoch": 16.652522395096653,
      "grad_norm": 1.1603432893753052,
      "learning_rate": 1.6705898291428767e-06,
      "loss": 0.1333,
      "num_input_tokens_seen": 30582864,
      "step": 35320
    },
    {
      "epoch": 16.654879773691654,
      "grad_norm": 0.810802698135376,
      "learning_rate": 1.6670630433535395e-06,
      "loss": 0.1487,
      "num_input_tokens_seen": 30587216,
      "step": 35325
    },
    {
      "epoch": 16.65723715228666,
      "grad_norm": 0.5316689014434814,
      "learning_rate": 1.6635398557962979e-06,
      "loss": 0.1499,
      "num_input_tokens_seen": 30591120,
      "step": 35330
    },
    {
      "epoch": 16.65959453088166,
      "grad_norm": 1.397180199623108,
      "learning_rate": 1.660020267014481e-06,
      "loss": 0.2332,
      "num_input_tokens_seen": 30595472,
      "step": 35335
    },
    {
      "epoch": 16.661951909476663,
      "grad_norm": 1.0813380479812622,
      "learning_rate": 1.6565042775508438e-06,
      "loss": 0.1736,
      "num_input_tokens_seen": 30599776,
      "step": 35340
    },
    {
      "epoch": 16.664309288071664,
      "grad_norm": 0.5328028202056885,
      "learning_rate": 1.6529918879475997e-06,
      "loss": 0.1563,
      "num_input_tokens_seen": 30603616,
      "step": 35345
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 1.4611403942108154,
      "learning_rate": 1.6494830987464043e-06,
      "loss": 0.1453,
      "num_input_tokens_seen": 30607376,
      "step": 35350
    },
    {
      "epoch": 16.66902404526167,
      "grad_norm": 0.6738825440406799,
      "learning_rate": 1.6459779104883555e-06,
      "loss": 0.2542,
      "num_input_tokens_seen": 30611936,
      "step": 35355
    },
    {
      "epoch": 16.671381423856673,
      "grad_norm": 1.4199811220169067,
      "learning_rate": 1.6424763237140013e-06,
      "loss": 0.1631,
      "num_input_tokens_seen": 30615552,
      "step": 35360
    },
    {
      "epoch": 16.673738802451673,
      "grad_norm": 1.8040229082107544,
      "learning_rate": 1.6389783389633207e-06,
      "loss": 0.2214,
      "num_input_tokens_seen": 30619312,
      "step": 35365
    },
    {
      "epoch": 16.676096181046677,
      "grad_norm": 0.8961659669876099,
      "learning_rate": 1.6354839567757546e-06,
      "loss": 0.2084,
      "num_input_tokens_seen": 30623456,
      "step": 35370
    },
    {
      "epoch": 16.678453559641678,
      "grad_norm": 0.7643205523490906,
      "learning_rate": 1.6319931776901831e-06,
      "loss": 0.1472,
      "num_input_tokens_seen": 30627392,
      "step": 35375
    },
    {
      "epoch": 16.680810938236682,
      "grad_norm": 2.045802116394043,
      "learning_rate": 1.6285060022449229e-06,
      "loss": 0.183,
      "num_input_tokens_seen": 30630864,
      "step": 35380
    },
    {
      "epoch": 16.683168316831683,
      "grad_norm": 0.5737906694412231,
      "learning_rate": 1.6250224309777434e-06,
      "loss": 0.1416,
      "num_input_tokens_seen": 30634592,
      "step": 35385
    },
    {
      "epoch": 16.685525695426687,
      "grad_norm": 0.8300437331199646,
      "learning_rate": 1.6215424644258515e-06,
      "loss": 0.187,
      "num_input_tokens_seen": 30639120,
      "step": 35390
    },
    {
      "epoch": 16.687883074021688,
      "grad_norm": 0.9597504138946533,
      "learning_rate": 1.6180661031259036e-06,
      "loss": 0.1985,
      "num_input_tokens_seen": 30643104,
      "step": 35395
    },
    {
      "epoch": 16.690240452616692,
      "grad_norm": 0.6888221502304077,
      "learning_rate": 1.614593347613999e-06,
      "loss": 0.1396,
      "num_input_tokens_seen": 30647744,
      "step": 35400
    },
    {
      "epoch": 16.690240452616692,
      "eval_loss": 0.20051118731498718,
      "eval_runtime": 22.1432,
      "eval_samples_per_second": 42.586,
      "eval_steps_per_second": 21.316,
      "num_input_tokens_seen": 30647744,
      "step": 35400
    },
    {
      "epoch": 16.692597831211692,
      "grad_norm": 1.4729835987091064,
      "learning_rate": 1.6111241984256758e-06,
      "loss": 0.1388,
      "num_input_tokens_seen": 30652144,
      "step": 35405
    },
    {
      "epoch": 16.694955209806697,
      "grad_norm": 0.8703485727310181,
      "learning_rate": 1.6076586560959257e-06,
      "loss": 0.1579,
      "num_input_tokens_seen": 30656496,
      "step": 35410
    },
    {
      "epoch": 16.697312588401697,
      "grad_norm": 0.9619051814079285,
      "learning_rate": 1.604196721159182e-06,
      "loss": 0.1336,
      "num_input_tokens_seen": 30660528,
      "step": 35415
    },
    {
      "epoch": 16.6996699669967,
      "grad_norm": 0.9103407859802246,
      "learning_rate": 1.6007383941493092e-06,
      "loss": 0.1255,
      "num_input_tokens_seen": 30665744,
      "step": 35420
    },
    {
      "epoch": 16.702027345591702,
      "grad_norm": 0.8154021501541138,
      "learning_rate": 1.5972836755996285e-06,
      "loss": 0.1984,
      "num_input_tokens_seen": 30669904,
      "step": 35425
    },
    {
      "epoch": 16.704384724186703,
      "grad_norm": 0.6254956126213074,
      "learning_rate": 1.5938325660429076e-06,
      "loss": 0.1286,
      "num_input_tokens_seen": 30673920,
      "step": 35430
    },
    {
      "epoch": 16.706742102781707,
      "grad_norm": 1.2551851272583008,
      "learning_rate": 1.5903850660113378e-06,
      "loss": 0.2473,
      "num_input_tokens_seen": 30678784,
      "step": 35435
    },
    {
      "epoch": 16.709099481376708,
      "grad_norm": 0.7472854852676392,
      "learning_rate": 1.5869411760365826e-06,
      "loss": 0.1567,
      "num_input_tokens_seen": 30683168,
      "step": 35440
    },
    {
      "epoch": 16.71145685997171,
      "grad_norm": 2.2082974910736084,
      "learning_rate": 1.58350089664972e-06,
      "loss": 0.1869,
      "num_input_tokens_seen": 30687664,
      "step": 35445
    },
    {
      "epoch": 16.713814238566712,
      "grad_norm": 0.800719141960144,
      "learning_rate": 1.5800642283812865e-06,
      "loss": 0.1594,
      "num_input_tokens_seen": 30691904,
      "step": 35450
    },
    {
      "epoch": 16.716171617161717,
      "grad_norm": 0.9423395991325378,
      "learning_rate": 1.5766311717612698e-06,
      "loss": 0.1475,
      "num_input_tokens_seen": 30695600,
      "step": 35455
    },
    {
      "epoch": 16.718528995756717,
      "grad_norm": 1.0241527557373047,
      "learning_rate": 1.5732017273190818e-06,
      "loss": 0.1651,
      "num_input_tokens_seen": 30699968,
      "step": 35460
    },
    {
      "epoch": 16.72088637435172,
      "grad_norm": 0.974482536315918,
      "learning_rate": 1.5697758955835806e-06,
      "loss": 0.2748,
      "num_input_tokens_seen": 30703792,
      "step": 35465
    },
    {
      "epoch": 16.723243752946722,
      "grad_norm": 1.513456106185913,
      "learning_rate": 1.566353677083085e-06,
      "loss": 0.1559,
      "num_input_tokens_seen": 30708096,
      "step": 35470
    },
    {
      "epoch": 16.725601131541726,
      "grad_norm": 1.9526034593582153,
      "learning_rate": 1.562935072345334e-06,
      "loss": 0.2361,
      "num_input_tokens_seen": 30713792,
      "step": 35475
    },
    {
      "epoch": 16.727958510136727,
      "grad_norm": 1.198120355606079,
      "learning_rate": 1.5595200818975281e-06,
      "loss": 0.1551,
      "num_input_tokens_seen": 30718288,
      "step": 35480
    },
    {
      "epoch": 16.73031588873173,
      "grad_norm": 1.1441923379898071,
      "learning_rate": 1.5561087062662905e-06,
      "loss": 0.1913,
      "num_input_tokens_seen": 30722384,
      "step": 35485
    },
    {
      "epoch": 16.73267326732673,
      "grad_norm": 0.632630467414856,
      "learning_rate": 1.5527009459777087e-06,
      "loss": 0.1883,
      "num_input_tokens_seen": 30726592,
      "step": 35490
    },
    {
      "epoch": 16.735030645921736,
      "grad_norm": 0.9414145350456238,
      "learning_rate": 1.5492968015572984e-06,
      "loss": 0.2207,
      "num_input_tokens_seen": 30730784,
      "step": 35495
    },
    {
      "epoch": 16.737388024516736,
      "grad_norm": 0.6513558626174927,
      "learning_rate": 1.5458962735300203e-06,
      "loss": 0.2021,
      "num_input_tokens_seen": 30734608,
      "step": 35500
    },
    {
      "epoch": 16.73974540311174,
      "grad_norm": 0.6114738583564758,
      "learning_rate": 1.54249936242028e-06,
      "loss": 0.115,
      "num_input_tokens_seen": 30738656,
      "step": 35505
    },
    {
      "epoch": 16.74210278170674,
      "grad_norm": 0.7765222787857056,
      "learning_rate": 1.5391060687519222e-06,
      "loss": 0.167,
      "num_input_tokens_seen": 30742720,
      "step": 35510
    },
    {
      "epoch": 16.744460160301745,
      "grad_norm": 1.7208168506622314,
      "learning_rate": 1.5357163930482367e-06,
      "loss": 0.2277,
      "num_input_tokens_seen": 30747392,
      "step": 35515
    },
    {
      "epoch": 16.746817538896746,
      "grad_norm": 0.7519655823707581,
      "learning_rate": 1.532330335831955e-06,
      "loss": 0.1395,
      "num_input_tokens_seen": 30752608,
      "step": 35520
    },
    {
      "epoch": 16.74917491749175,
      "grad_norm": 2.103959083557129,
      "learning_rate": 1.5289478976252491e-06,
      "loss": 0.1537,
      "num_input_tokens_seen": 30757664,
      "step": 35525
    },
    {
      "epoch": 16.75153229608675,
      "grad_norm": 0.6480004787445068,
      "learning_rate": 1.5255690789497345e-06,
      "loss": 0.2044,
      "num_input_tokens_seen": 30761872,
      "step": 35530
    },
    {
      "epoch": 16.753889674681755,
      "grad_norm": 1.1724820137023926,
      "learning_rate": 1.5221938803264641e-06,
      "loss": 0.2057,
      "num_input_tokens_seen": 30766176,
      "step": 35535
    },
    {
      "epoch": 16.756247053276756,
      "grad_norm": 0.8595348596572876,
      "learning_rate": 1.518822302275938e-06,
      "loss": 0.1539,
      "num_input_tokens_seen": 30770704,
      "step": 35540
    },
    {
      "epoch": 16.75860443187176,
      "grad_norm": 0.9849709272384644,
      "learning_rate": 1.5154543453180958e-06,
      "loss": 0.1737,
      "num_input_tokens_seen": 30775024,
      "step": 35545
    },
    {
      "epoch": 16.76096181046676,
      "grad_norm": 1.992353081703186,
      "learning_rate": 1.5120900099723167e-06,
      "loss": 0.2285,
      "num_input_tokens_seen": 30779200,
      "step": 35550
    },
    {
      "epoch": 16.763319189061765,
      "grad_norm": 0.6801621317863464,
      "learning_rate": 1.5087292967574273e-06,
      "loss": 0.1622,
      "num_input_tokens_seen": 30783088,
      "step": 35555
    },
    {
      "epoch": 16.765676567656765,
      "grad_norm": 1.241513967514038,
      "learning_rate": 1.5053722061916908e-06,
      "loss": 0.1077,
      "num_input_tokens_seen": 30787376,
      "step": 35560
    },
    {
      "epoch": 16.76803394625177,
      "grad_norm": 1.7581876516342163,
      "learning_rate": 1.5020187387928124e-06,
      "loss": 0.1285,
      "num_input_tokens_seen": 30791376,
      "step": 35565
    },
    {
      "epoch": 16.77039132484677,
      "grad_norm": 1.6962354183197021,
      "learning_rate": 1.4986688950779343e-06,
      "loss": 0.1403,
      "num_input_tokens_seen": 30795456,
      "step": 35570
    },
    {
      "epoch": 16.772748703441774,
      "grad_norm": 0.7010915279388428,
      "learning_rate": 1.495322675563654e-06,
      "loss": 0.2696,
      "num_input_tokens_seen": 30800320,
      "step": 35575
    },
    {
      "epoch": 16.775106082036775,
      "grad_norm": 2.8560338020324707,
      "learning_rate": 1.4919800807659922e-06,
      "loss": 0.1621,
      "num_input_tokens_seen": 30804560,
      "step": 35580
    },
    {
      "epoch": 16.77746346063178,
      "grad_norm": 0.43469342589378357,
      "learning_rate": 1.4886411112004255e-06,
      "loss": 0.1985,
      "num_input_tokens_seen": 30809232,
      "step": 35585
    },
    {
      "epoch": 16.77982083922678,
      "grad_norm": 1.2382370233535767,
      "learning_rate": 1.4853057673818588e-06,
      "loss": 0.2388,
      "num_input_tokens_seen": 30813760,
      "step": 35590
    },
    {
      "epoch": 16.782178217821784,
      "grad_norm": 0.9418686628341675,
      "learning_rate": 1.481974049824647e-06,
      "loss": 0.2115,
      "num_input_tokens_seen": 30818512,
      "step": 35595
    },
    {
      "epoch": 16.784535596416784,
      "grad_norm": 0.7632461786270142,
      "learning_rate": 1.4786459590425849e-06,
      "loss": 0.1229,
      "num_input_tokens_seen": 30823072,
      "step": 35600
    },
    {
      "epoch": 16.784535596416784,
      "eval_loss": 0.2002013921737671,
      "eval_runtime": 22.1492,
      "eval_samples_per_second": 42.575,
      "eval_steps_per_second": 21.31,
      "num_input_tokens_seen": 30823072,
      "step": 35600
    },
    {
      "epoch": 16.78689297501179,
      "grad_norm": 0.9935822486877441,
      "learning_rate": 1.4753214955489036e-06,
      "loss": 0.2048,
      "num_input_tokens_seen": 30827136,
      "step": 35605
    },
    {
      "epoch": 16.78925035360679,
      "grad_norm": 1.1135532855987549,
      "learning_rate": 1.4720006598562737e-06,
      "loss": 0.2641,
      "num_input_tokens_seen": 30831248,
      "step": 35610
    },
    {
      "epoch": 16.79160773220179,
      "grad_norm": 1.0354697704315186,
      "learning_rate": 1.4686834524768185e-06,
      "loss": 0.1258,
      "num_input_tokens_seen": 30837152,
      "step": 35615
    },
    {
      "epoch": 16.793965110796794,
      "grad_norm": 0.4030141830444336,
      "learning_rate": 1.4653698739220844e-06,
      "loss": 0.187,
      "num_input_tokens_seen": 30841760,
      "step": 35620
    },
    {
      "epoch": 16.796322489391795,
      "grad_norm": 0.6137290596961975,
      "learning_rate": 1.4620599247030715e-06,
      "loss": 0.1619,
      "num_input_tokens_seen": 30846032,
      "step": 35625
    },
    {
      "epoch": 16.7986798679868,
      "grad_norm": 0.721302330493927,
      "learning_rate": 1.4587536053302125e-06,
      "loss": 0.1931,
      "num_input_tokens_seen": 30850496,
      "step": 35630
    },
    {
      "epoch": 16.8010372465818,
      "grad_norm": 1.110918402671814,
      "learning_rate": 1.4554509163133862e-06,
      "loss": 0.1793,
      "num_input_tokens_seen": 30854496,
      "step": 35635
    },
    {
      "epoch": 16.803394625176804,
      "grad_norm": 0.6414738297462463,
      "learning_rate": 1.4521518581619098e-06,
      "loss": 0.1622,
      "num_input_tokens_seen": 30858320,
      "step": 35640
    },
    {
      "epoch": 16.805752003771804,
      "grad_norm": 0.6244027614593506,
      "learning_rate": 1.4488564313845348e-06,
      "loss": 0.1465,
      "num_input_tokens_seen": 30862992,
      "step": 35645
    },
    {
      "epoch": 16.80810938236681,
      "grad_norm": 0.8971337676048279,
      "learning_rate": 1.4455646364894603e-06,
      "loss": 0.1372,
      "num_input_tokens_seen": 30867136,
      "step": 35650
    },
    {
      "epoch": 16.81046676096181,
      "grad_norm": 1.1540195941925049,
      "learning_rate": 1.4422764739843247e-06,
      "loss": 0.1804,
      "num_input_tokens_seen": 30871216,
      "step": 35655
    },
    {
      "epoch": 16.812824139556813,
      "grad_norm": 0.8950682878494263,
      "learning_rate": 1.4389919443762e-06,
      "loss": 0.1812,
      "num_input_tokens_seen": 30875856,
      "step": 35660
    },
    {
      "epoch": 16.815181518151814,
      "grad_norm": 1.5225023031234741,
      "learning_rate": 1.4357110481716063e-06,
      "loss": 0.1687,
      "num_input_tokens_seen": 30880016,
      "step": 35665
    },
    {
      "epoch": 16.817538896746818,
      "grad_norm": 0.4259532690048218,
      "learning_rate": 1.4324337858764941e-06,
      "loss": 0.1809,
      "num_input_tokens_seen": 30884048,
      "step": 35670
    },
    {
      "epoch": 16.81989627534182,
      "grad_norm": 1.983905553817749,
      "learning_rate": 1.4291601579962622e-06,
      "loss": 0.301,
      "num_input_tokens_seen": 30888448,
      "step": 35675
    },
    {
      "epoch": 16.822253653936823,
      "grad_norm": 0.4284643232822418,
      "learning_rate": 1.42589016503574e-06,
      "loss": 0.1227,
      "num_input_tokens_seen": 30892176,
      "step": 35680
    },
    {
      "epoch": 16.824611032531823,
      "grad_norm": 1.0193982124328613,
      "learning_rate": 1.4226238074992099e-06,
      "loss": 0.2357,
      "num_input_tokens_seen": 30896640,
      "step": 35685
    },
    {
      "epoch": 16.826968411126828,
      "grad_norm": 0.7354974150657654,
      "learning_rate": 1.4193610858903778e-06,
      "loss": 0.1629,
      "num_input_tokens_seen": 30900720,
      "step": 35690
    },
    {
      "epoch": 16.82932578972183,
      "grad_norm": 1.7517509460449219,
      "learning_rate": 1.416102000712402e-06,
      "loss": 0.2072,
      "num_input_tokens_seen": 30905616,
      "step": 35695
    },
    {
      "epoch": 16.831683168316832,
      "grad_norm": 0.824403703212738,
      "learning_rate": 1.4128465524678668e-06,
      "loss": 0.1398,
      "num_input_tokens_seen": 30909600,
      "step": 35700
    },
    {
      "epoch": 16.834040546911833,
      "grad_norm": 1.2270914316177368,
      "learning_rate": 1.4095947416588124e-06,
      "loss": 0.1883,
      "num_input_tokens_seen": 30913568,
      "step": 35705
    },
    {
      "epoch": 16.836397925506837,
      "grad_norm": 0.8912440538406372,
      "learning_rate": 1.4063465687866983e-06,
      "loss": 0.1304,
      "num_input_tokens_seen": 30918080,
      "step": 35710
    },
    {
      "epoch": 16.838755304101838,
      "grad_norm": 1.6009536981582642,
      "learning_rate": 1.4031020343524438e-06,
      "loss": 0.2117,
      "num_input_tokens_seen": 30923120,
      "step": 35715
    },
    {
      "epoch": 16.841112682696842,
      "grad_norm": 0.9513965249061584,
      "learning_rate": 1.3998611388563926e-06,
      "loss": 0.1204,
      "num_input_tokens_seen": 30927104,
      "step": 35720
    },
    {
      "epoch": 16.843470061291843,
      "grad_norm": 0.6556822061538696,
      "learning_rate": 1.3966238827983314e-06,
      "loss": 0.1177,
      "num_input_tokens_seen": 30931920,
      "step": 35725
    },
    {
      "epoch": 16.845827439886847,
      "grad_norm": 0.5500525832176208,
      "learning_rate": 1.393390266677483e-06,
      "loss": 0.1163,
      "num_input_tokens_seen": 30936608,
      "step": 35730
    },
    {
      "epoch": 16.848184818481847,
      "grad_norm": 0.8949035406112671,
      "learning_rate": 1.3901602909925204e-06,
      "loss": 0.2131,
      "num_input_tokens_seen": 30941024,
      "step": 35735
    },
    {
      "epoch": 16.85054219707685,
      "grad_norm": 0.5950950980186462,
      "learning_rate": 1.3869339562415373e-06,
      "loss": 0.1634,
      "num_input_tokens_seen": 30945200,
      "step": 35740
    },
    {
      "epoch": 16.852899575671852,
      "grad_norm": 1.8759613037109375,
      "learning_rate": 1.38371126292208e-06,
      "loss": 0.2039,
      "num_input_tokens_seen": 30950000,
      "step": 35745
    },
    {
      "epoch": 16.855256954266856,
      "grad_norm": 0.9816802144050598,
      "learning_rate": 1.3804922115311286e-06,
      "loss": 0.2416,
      "num_input_tokens_seen": 30953760,
      "step": 35750
    },
    {
      "epoch": 16.857614332861857,
      "grad_norm": 0.6484068036079407,
      "learning_rate": 1.3772768025650945e-06,
      "loss": 0.2289,
      "num_input_tokens_seen": 30957376,
      "step": 35755
    },
    {
      "epoch": 16.85997171145686,
      "grad_norm": 0.8358988761901855,
      "learning_rate": 1.3740650365198448e-06,
      "loss": 0.168,
      "num_input_tokens_seen": 30961712,
      "step": 35760
    },
    {
      "epoch": 16.862329090051862,
      "grad_norm": 0.34399035573005676,
      "learning_rate": 1.3708569138906612e-06,
      "loss": 0.201,
      "num_input_tokens_seen": 30966480,
      "step": 35765
    },
    {
      "epoch": 16.864686468646866,
      "grad_norm": 1.0394054651260376,
      "learning_rate": 1.367652435172287e-06,
      "loss": 0.2038,
      "num_input_tokens_seen": 30971152,
      "step": 35770
    },
    {
      "epoch": 16.867043847241867,
      "grad_norm": 0.9389020204544067,
      "learning_rate": 1.364451600858893e-06,
      "loss": 0.1555,
      "num_input_tokens_seen": 30975088,
      "step": 35775
    },
    {
      "epoch": 16.86940122583687,
      "grad_norm": 1.991745948791504,
      "learning_rate": 1.3612544114440823e-06,
      "loss": 0.2026,
      "num_input_tokens_seen": 30979264,
      "step": 35780
    },
    {
      "epoch": 16.87175860443187,
      "grad_norm": 0.5745898485183716,
      "learning_rate": 1.3580608674209072e-06,
      "loss": 0.1956,
      "num_input_tokens_seen": 30982960,
      "step": 35785
    },
    {
      "epoch": 16.874115983026876,
      "grad_norm": 1.0103144645690918,
      "learning_rate": 1.3548709692818434e-06,
      "loss": 0.1625,
      "num_input_tokens_seen": 30987056,
      "step": 35790
    },
    {
      "epoch": 16.876473361621876,
      "grad_norm": 0.8714461922645569,
      "learning_rate": 1.3516847175188223e-06,
      "loss": 0.2663,
      "num_input_tokens_seen": 30991456,
      "step": 35795
    },
    {
      "epoch": 16.87883074021688,
      "grad_norm": 1.4106332063674927,
      "learning_rate": 1.348502112623204e-06,
      "loss": 0.1862,
      "num_input_tokens_seen": 30996032,
      "step": 35800
    },
    {
      "epoch": 16.87883074021688,
      "eval_loss": 0.200471892952919,
      "eval_runtime": 22.1816,
      "eval_samples_per_second": 42.513,
      "eval_steps_per_second": 21.279,
      "num_input_tokens_seen": 30996032,
      "step": 35800
    },
    {
      "epoch": 16.88118811881188,
      "grad_norm": 0.4470083713531494,
      "learning_rate": 1.3453231550857787e-06,
      "loss": 0.1237,
      "num_input_tokens_seen": 30999968,
      "step": 35805
    },
    {
      "epoch": 16.883545497406885,
      "grad_norm": 0.6979115009307861,
      "learning_rate": 1.3421478453967878e-06,
      "loss": 0.1315,
      "num_input_tokens_seen": 31003792,
      "step": 35810
    },
    {
      "epoch": 16.885902876001886,
      "grad_norm": 0.7159786224365234,
      "learning_rate": 1.3389761840459065e-06,
      "loss": 0.1967,
      "num_input_tokens_seen": 31008944,
      "step": 35815
    },
    {
      "epoch": 16.888260254596887,
      "grad_norm": 0.5386729836463928,
      "learning_rate": 1.3358081715222376e-06,
      "loss": 0.1181,
      "num_input_tokens_seen": 31012976,
      "step": 35820
    },
    {
      "epoch": 16.89061763319189,
      "grad_norm": 1.1659141778945923,
      "learning_rate": 1.3326438083143295e-06,
      "loss": 0.1673,
      "num_input_tokens_seen": 31017664,
      "step": 35825
    },
    {
      "epoch": 16.89297501178689,
      "grad_norm": 2.8056490421295166,
      "learning_rate": 1.3294830949101723e-06,
      "loss": 0.2062,
      "num_input_tokens_seen": 31022704,
      "step": 35830
    },
    {
      "epoch": 16.895332390381895,
      "grad_norm": 0.6333066821098328,
      "learning_rate": 1.3263260317971815e-06,
      "loss": 0.1615,
      "num_input_tokens_seen": 31026752,
      "step": 35835
    },
    {
      "epoch": 16.897689768976896,
      "grad_norm": 1.348397135734558,
      "learning_rate": 1.3231726194622208e-06,
      "loss": 0.166,
      "num_input_tokens_seen": 31031136,
      "step": 35840
    },
    {
      "epoch": 16.9000471475719,
      "grad_norm": 1.7318406105041504,
      "learning_rate": 1.3200228583915814e-06,
      "loss": 0.1938,
      "num_input_tokens_seen": 31036208,
      "step": 35845
    },
    {
      "epoch": 16.9024045261669,
      "grad_norm": 1.9279043674468994,
      "learning_rate": 1.3168767490709971e-06,
      "loss": 0.224,
      "num_input_tokens_seen": 31040816,
      "step": 35850
    },
    {
      "epoch": 16.904761904761905,
      "grad_norm": 0.6504449248313904,
      "learning_rate": 1.3137342919856437e-06,
      "loss": 0.125,
      "num_input_tokens_seen": 31044928,
      "step": 35855
    },
    {
      "epoch": 16.907119283356906,
      "grad_norm": 0.8684129118919373,
      "learning_rate": 1.310595487620117e-06,
      "loss": 0.1451,
      "num_input_tokens_seen": 31048736,
      "step": 35860
    },
    {
      "epoch": 16.90947666195191,
      "grad_norm": 1.14115309715271,
      "learning_rate": 1.3074603364584715e-06,
      "loss": 0.1252,
      "num_input_tokens_seen": 31053136,
      "step": 35865
    },
    {
      "epoch": 16.91183404054691,
      "grad_norm": 0.6694037318229675,
      "learning_rate": 1.3043288389841758e-06,
      "loss": 0.1889,
      "num_input_tokens_seen": 31057216,
      "step": 35870
    },
    {
      "epoch": 16.914191419141915,
      "grad_norm": 0.9226470589637756,
      "learning_rate": 1.3012009956801546e-06,
      "loss": 0.1724,
      "num_input_tokens_seen": 31061616,
      "step": 35875
    },
    {
      "epoch": 16.916548797736915,
      "grad_norm": 0.6277981400489807,
      "learning_rate": 1.2980768070287586e-06,
      "loss": 0.1383,
      "num_input_tokens_seen": 31065200,
      "step": 35880
    },
    {
      "epoch": 16.91890617633192,
      "grad_norm": 0.9380372762680054,
      "learning_rate": 1.2949562735117716e-06,
      "loss": 0.3063,
      "num_input_tokens_seen": 31069584,
      "step": 35885
    },
    {
      "epoch": 16.92126355492692,
      "grad_norm": 0.6784023642539978,
      "learning_rate": 1.291839395610428e-06,
      "loss": 0.0926,
      "num_input_tokens_seen": 31074464,
      "step": 35890
    },
    {
      "epoch": 16.923620933521924,
      "grad_norm": 0.4256890118122101,
      "learning_rate": 1.2887261738053852e-06,
      "loss": 0.1388,
      "num_input_tokens_seen": 31079008,
      "step": 35895
    },
    {
      "epoch": 16.925978312116925,
      "grad_norm": 0.8632469177246094,
      "learning_rate": 1.2856166085767396e-06,
      "loss": 0.186,
      "num_input_tokens_seen": 31083264,
      "step": 35900
    },
    {
      "epoch": 16.92833569071193,
      "grad_norm": 1.1673372983932495,
      "learning_rate": 1.2825107004040272e-06,
      "loss": 0.1523,
      "num_input_tokens_seen": 31087792,
      "step": 35905
    },
    {
      "epoch": 16.93069306930693,
      "grad_norm": 1.282408356666565,
      "learning_rate": 1.2794084497662146e-06,
      "loss": 0.2036,
      "num_input_tokens_seen": 31091840,
      "step": 35910
    },
    {
      "epoch": 16.933050447901934,
      "grad_norm": 0.42903655767440796,
      "learning_rate": 1.276309857141711e-06,
      "loss": 0.1078,
      "num_input_tokens_seen": 31095872,
      "step": 35915
    },
    {
      "epoch": 16.935407826496935,
      "grad_norm": 0.4090745449066162,
      "learning_rate": 1.273214923008359e-06,
      "loss": 0.1662,
      "num_input_tokens_seen": 31100464,
      "step": 35920
    },
    {
      "epoch": 16.93776520509194,
      "grad_norm": 0.7648200392723083,
      "learning_rate": 1.2701236478434352e-06,
      "loss": 0.1965,
      "num_input_tokens_seen": 31104544,
      "step": 35925
    },
    {
      "epoch": 16.94012258368694,
      "grad_norm": 1.3139630556106567,
      "learning_rate": 1.2670360321236502e-06,
      "loss": 0.1181,
      "num_input_tokens_seen": 31109072,
      "step": 35930
    },
    {
      "epoch": 16.942479962281944,
      "grad_norm": 1.2252178192138672,
      "learning_rate": 1.2639520763251617e-06,
      "loss": 0.2742,
      "num_input_tokens_seen": 31113824,
      "step": 35935
    },
    {
      "epoch": 16.944837340876944,
      "grad_norm": 0.8820794224739075,
      "learning_rate": 1.2608717809235448e-06,
      "loss": 0.1568,
      "num_input_tokens_seen": 31117440,
      "step": 35940
    },
    {
      "epoch": 16.94719471947195,
      "grad_norm": 1.873049259185791,
      "learning_rate": 1.2577951463938282e-06,
      "loss": 0.2681,
      "num_input_tokens_seen": 31121584,
      "step": 35945
    },
    {
      "epoch": 16.94955209806695,
      "grad_norm": 1.4009698629379272,
      "learning_rate": 1.2547221732104569e-06,
      "loss": 0.2076,
      "num_input_tokens_seen": 31126080,
      "step": 35950
    },
    {
      "epoch": 16.951909476661953,
      "grad_norm": 0.5267327427864075,
      "learning_rate": 1.25165286184733e-06,
      "loss": 0.1136,
      "num_input_tokens_seen": 31130432,
      "step": 35955
    },
    {
      "epoch": 16.954266855256954,
      "grad_norm": 3.2359962463378906,
      "learning_rate": 1.248587212777777e-06,
      "loss": 0.2314,
      "num_input_tokens_seen": 31134464,
      "step": 35960
    },
    {
      "epoch": 16.956624233851958,
      "grad_norm": 1.115225076675415,
      "learning_rate": 1.2455252264745532e-06,
      "loss": 0.2106,
      "num_input_tokens_seen": 31138544,
      "step": 35965
    },
    {
      "epoch": 16.95898161244696,
      "grad_norm": 0.2713083326816559,
      "learning_rate": 1.2424669034098528e-06,
      "loss": 0.1126,
      "num_input_tokens_seen": 31143408,
      "step": 35970
    },
    {
      "epoch": 16.961338991041963,
      "grad_norm": 1.1469566822052002,
      "learning_rate": 1.2394122440553185e-06,
      "loss": 0.091,
      "num_input_tokens_seen": 31147120,
      "step": 35975
    },
    {
      "epoch": 16.963696369636963,
      "grad_norm": 2.281986713409424,
      "learning_rate": 1.2363612488820037e-06,
      "loss": 0.2578,
      "num_input_tokens_seen": 31150800,
      "step": 35980
    },
    {
      "epoch": 16.966053748231968,
      "grad_norm": 0.6630254983901978,
      "learning_rate": 1.2333139183604208e-06,
      "loss": 0.1898,
      "num_input_tokens_seen": 31155008,
      "step": 35985
    },
    {
      "epoch": 16.968411126826968,
      "grad_norm": 1.6995190382003784,
      "learning_rate": 1.2302702529604998e-06,
      "loss": 0.176,
      "num_input_tokens_seen": 31158960,
      "step": 35990
    },
    {
      "epoch": 16.970768505421972,
      "grad_norm": 0.48600295186042786,
      "learning_rate": 1.227230253151615e-06,
      "loss": 0.1381,
      "num_input_tokens_seen": 31163936,
      "step": 35995
    },
    {
      "epoch": 16.973125884016973,
      "grad_norm": 0.540059506893158,
      "learning_rate": 1.2241939194025748e-06,
      "loss": 0.1848,
      "num_input_tokens_seen": 31167328,
      "step": 36000
    },
    {
      "epoch": 16.973125884016973,
      "eval_loss": 0.20079053938388824,
      "eval_runtime": 22.1742,
      "eval_samples_per_second": 42.527,
      "eval_steps_per_second": 21.286,
      "num_input_tokens_seen": 31167328,
      "step": 36000
    },
    {
      "epoch": 16.975483262611977,
      "grad_norm": 0.605492353439331,
      "learning_rate": 1.2211612521816156e-06,
      "loss": 0.1447,
      "num_input_tokens_seen": 31171424,
      "step": 36005
    },
    {
      "epoch": 16.977840641206978,
      "grad_norm": 1.0324732065200806,
      "learning_rate": 1.2181322519564137e-06,
      "loss": 0.2393,
      "num_input_tokens_seen": 31175696,
      "step": 36010
    },
    {
      "epoch": 16.980198019801982,
      "grad_norm": 0.6533383727073669,
      "learning_rate": 1.2151069191940839e-06,
      "loss": 0.1139,
      "num_input_tokens_seen": 31180064,
      "step": 36015
    },
    {
      "epoch": 16.982555398396983,
      "grad_norm": 1.130784511566162,
      "learning_rate": 1.2120852543611644e-06,
      "loss": 0.1698,
      "num_input_tokens_seen": 31184976,
      "step": 36020
    },
    {
      "epoch": 16.984912776991983,
      "grad_norm": 0.4751552641391754,
      "learning_rate": 1.2090672579236379e-06,
      "loss": 0.1691,
      "num_input_tokens_seen": 31189392,
      "step": 36025
    },
    {
      "epoch": 16.987270155586987,
      "grad_norm": 1.1857749223709106,
      "learning_rate": 1.2060529303469126e-06,
      "loss": 0.2629,
      "num_input_tokens_seen": 31193456,
      "step": 36030
    },
    {
      "epoch": 16.989627534181988,
      "grad_norm": 0.6702662706375122,
      "learning_rate": 1.2030422720958445e-06,
      "loss": 0.215,
      "num_input_tokens_seen": 31198288,
      "step": 36035
    },
    {
      "epoch": 16.991984912776992,
      "grad_norm": 1.6033931970596313,
      "learning_rate": 1.200035283634704e-06,
      "loss": 0.1369,
      "num_input_tokens_seen": 31202336,
      "step": 36040
    },
    {
      "epoch": 16.994342291371993,
      "grad_norm": 0.4836793839931488,
      "learning_rate": 1.1970319654272144e-06,
      "loss": 0.1246,
      "num_input_tokens_seen": 31206832,
      "step": 36045
    },
    {
      "epoch": 16.996699669966997,
      "grad_norm": 1.0191075801849365,
      "learning_rate": 1.1940323179365192e-06,
      "loss": 0.1737,
      "num_input_tokens_seen": 31210880,
      "step": 36050
    },
    {
      "epoch": 16.999057048561998,
      "grad_norm": 0.5567862391471863,
      "learning_rate": 1.1910363416252095e-06,
      "loss": 0.1861,
      "num_input_tokens_seen": 31215248,
      "step": 36055
    },
    {
      "epoch": 17.001414427157002,
      "grad_norm": 1.0792062282562256,
      "learning_rate": 1.1880440369552964e-06,
      "loss": 0.1484,
      "num_input_tokens_seen": 31219808,
      "step": 36060
    },
    {
      "epoch": 17.003771805752002,
      "grad_norm": 1.0940254926681519,
      "learning_rate": 1.1850554043882328e-06,
      "loss": 0.1882,
      "num_input_tokens_seen": 31223520,
      "step": 36065
    },
    {
      "epoch": 17.006129184347007,
      "grad_norm": 1.0838295221328735,
      "learning_rate": 1.1820704443849028e-06,
      "loss": 0.1893,
      "num_input_tokens_seen": 31227456,
      "step": 36070
    },
    {
      "epoch": 17.008486562942007,
      "grad_norm": 0.9706223607063293,
      "learning_rate": 1.1790891574056219e-06,
      "loss": 0.1575,
      "num_input_tokens_seen": 31231344,
      "step": 36075
    },
    {
      "epoch": 17.01084394153701,
      "grad_norm": 1.7031818628311157,
      "learning_rate": 1.1761115439101523e-06,
      "loss": 0.1651,
      "num_input_tokens_seen": 31235552,
      "step": 36080
    },
    {
      "epoch": 17.013201320132012,
      "grad_norm": 1.2540249824523926,
      "learning_rate": 1.1731376043576659e-06,
      "loss": 0.177,
      "num_input_tokens_seen": 31240016,
      "step": 36085
    },
    {
      "epoch": 17.015558698727016,
      "grad_norm": 0.8458021283149719,
      "learning_rate": 1.1701673392067875e-06,
      "loss": 0.1567,
      "num_input_tokens_seen": 31244368,
      "step": 36090
    },
    {
      "epoch": 17.017916077322017,
      "grad_norm": 0.862043023109436,
      "learning_rate": 1.1672007489155757e-06,
      "loss": 0.1669,
      "num_input_tokens_seen": 31248432,
      "step": 36095
    },
    {
      "epoch": 17.02027345591702,
      "grad_norm": 1.154882788658142,
      "learning_rate": 1.164237833941506e-06,
      "loss": 0.2243,
      "num_input_tokens_seen": 31252800,
      "step": 36100
    },
    {
      "epoch": 17.02263083451202,
      "grad_norm": 0.37515735626220703,
      "learning_rate": 1.1612785947415022e-06,
      "loss": 0.175,
      "num_input_tokens_seen": 31256992,
      "step": 36105
    },
    {
      "epoch": 17.024988213107026,
      "grad_norm": 1.6372085809707642,
      "learning_rate": 1.1583230317719185e-06,
      "loss": 0.2023,
      "num_input_tokens_seen": 31262096,
      "step": 36110
    },
    {
      "epoch": 17.027345591702026,
      "grad_norm": 0.48437169194221497,
      "learning_rate": 1.1553711454885318e-06,
      "loss": 0.145,
      "num_input_tokens_seen": 31265872,
      "step": 36115
    },
    {
      "epoch": 17.02970297029703,
      "grad_norm": 1.7427947521209717,
      "learning_rate": 1.152422936346567e-06,
      "loss": 0.1503,
      "num_input_tokens_seen": 31270304,
      "step": 36120
    },
    {
      "epoch": 17.03206034889203,
      "grad_norm": 1.1263818740844727,
      "learning_rate": 1.1494784048006718e-06,
      "loss": 0.1273,
      "num_input_tokens_seen": 31274336,
      "step": 36125
    },
    {
      "epoch": 17.034417727487035,
      "grad_norm": 0.8762524127960205,
      "learning_rate": 1.1465375513049326e-06,
      "loss": 0.1473,
      "num_input_tokens_seen": 31278784,
      "step": 36130
    },
    {
      "epoch": 17.036775106082036,
      "grad_norm": 0.37524959444999695,
      "learning_rate": 1.1436003763128616e-06,
      "loss": 0.2039,
      "num_input_tokens_seen": 31282736,
      "step": 36135
    },
    {
      "epoch": 17.03913248467704,
      "grad_norm": 2.1085216999053955,
      "learning_rate": 1.1406668802774106e-06,
      "loss": 0.205,
      "num_input_tokens_seen": 31288480,
      "step": 36140
    },
    {
      "epoch": 17.04148986327204,
      "grad_norm": 1.4416017532348633,
      "learning_rate": 1.137737063650965e-06,
      "loss": 0.2158,
      "num_input_tokens_seen": 31292400,
      "step": 36145
    },
    {
      "epoch": 17.043847241867045,
      "grad_norm": 1.4226363897323608,
      "learning_rate": 1.1348109268853323e-06,
      "loss": 0.1547,
      "num_input_tokens_seen": 31296400,
      "step": 36150
    },
    {
      "epoch": 17.046204620462046,
      "grad_norm": 0.6963545083999634,
      "learning_rate": 1.1318884704317634e-06,
      "loss": 0.2204,
      "num_input_tokens_seen": 31300304,
      "step": 36155
    },
    {
      "epoch": 17.04856199905705,
      "grad_norm": 1.1914584636688232,
      "learning_rate": 1.1289696947409417e-06,
      "loss": 0.1776,
      "num_input_tokens_seen": 31305040,
      "step": 36160
    },
    {
      "epoch": 17.05091937765205,
      "grad_norm": 0.7287025451660156,
      "learning_rate": 1.126054600262974e-06,
      "loss": 0.1419,
      "num_input_tokens_seen": 31310576,
      "step": 36165
    },
    {
      "epoch": 17.053276756247055,
      "grad_norm": 1.284759521484375,
      "learning_rate": 1.1231431874474064e-06,
      "loss": 0.2448,
      "num_input_tokens_seen": 31314544,
      "step": 36170
    },
    {
      "epoch": 17.055634134842055,
      "grad_norm": 1.9278149604797363,
      "learning_rate": 1.12023545674321e-06,
      "loss": 0.2102,
      "num_input_tokens_seen": 31320272,
      "step": 36175
    },
    {
      "epoch": 17.05799151343706,
      "grad_norm": 1.2327567338943481,
      "learning_rate": 1.117331408598804e-06,
      "loss": 0.1478,
      "num_input_tokens_seen": 31325104,
      "step": 36180
    },
    {
      "epoch": 17.06034889203206,
      "grad_norm": 0.7777826189994812,
      "learning_rate": 1.1144310434620191e-06,
      "loss": 0.0824,
      "num_input_tokens_seen": 31329152,
      "step": 36185
    },
    {
      "epoch": 17.062706270627064,
      "grad_norm": 0.6439688801765442,
      "learning_rate": 1.1115343617801365e-06,
      "loss": 0.1661,
      "num_input_tokens_seen": 31332896,
      "step": 36190
    },
    {
      "epoch": 17.065063649222065,
      "grad_norm": 1.2838112115859985,
      "learning_rate": 1.1086413639998515e-06,
      "loss": 0.1406,
      "num_input_tokens_seen": 31336656,
      "step": 36195
    },
    {
      "epoch": 17.06742102781707,
      "grad_norm": 0.7685809135437012,
      "learning_rate": 1.1057520505673103e-06,
      "loss": 0.1349,
      "num_input_tokens_seen": 31341392,
      "step": 36200
    },
    {
      "epoch": 17.06742102781707,
      "eval_loss": 0.2003323882818222,
      "eval_runtime": 22.1413,
      "eval_samples_per_second": 42.59,
      "eval_steps_per_second": 21.318,
      "num_input_tokens_seen": 31341392,
      "step": 36200
    },
    {
      "epoch": 17.06977840641207,
      "grad_norm": 1.000630497932434,
      "learning_rate": 1.1028664219280727e-06,
      "loss": 0.2107,
      "num_input_tokens_seen": 31346032,
      "step": 36205
    },
    {
      "epoch": 17.072135785007074,
      "grad_norm": 0.5364559292793274,
      "learning_rate": 1.0999844785271468e-06,
      "loss": 0.2429,
      "num_input_tokens_seen": 31350432,
      "step": 36210
    },
    {
      "epoch": 17.074493163602074,
      "grad_norm": 0.7620493769645691,
      "learning_rate": 1.097106220808955e-06,
      "loss": 0.1569,
      "num_input_tokens_seen": 31354528,
      "step": 36215
    },
    {
      "epoch": 17.076850542197075,
      "grad_norm": 1.464111566543579,
      "learning_rate": 1.0942316492173698e-06,
      "loss": 0.2088,
      "num_input_tokens_seen": 31358544,
      "step": 36220
    },
    {
      "epoch": 17.07920792079208,
      "grad_norm": 1.1525849103927612,
      "learning_rate": 1.0913607641956841e-06,
      "loss": 0.1795,
      "num_input_tokens_seen": 31363040,
      "step": 36225
    },
    {
      "epoch": 17.08156529938708,
      "grad_norm": 0.6396301984786987,
      "learning_rate": 1.0884935661866213e-06,
      "loss": 0.1452,
      "num_input_tokens_seen": 31367280,
      "step": 36230
    },
    {
      "epoch": 17.083922677982084,
      "grad_norm": 0.4515398442745209,
      "learning_rate": 1.0856300556323418e-06,
      "loss": 0.2265,
      "num_input_tokens_seen": 31371264,
      "step": 36235
    },
    {
      "epoch": 17.086280056577085,
      "grad_norm": 0.99040287733078,
      "learning_rate": 1.0827702329744365e-06,
      "loss": 0.2068,
      "num_input_tokens_seen": 31375488,
      "step": 36240
    },
    {
      "epoch": 17.08863743517209,
      "grad_norm": 1.7038630247116089,
      "learning_rate": 1.0799140986539197e-06,
      "loss": 0.1597,
      "num_input_tokens_seen": 31381280,
      "step": 36245
    },
    {
      "epoch": 17.09099481376709,
      "grad_norm": 0.5846710205078125,
      "learning_rate": 1.0770616531112526e-06,
      "loss": 0.1338,
      "num_input_tokens_seen": 31385120,
      "step": 36250
    },
    {
      "epoch": 17.093352192362094,
      "grad_norm": 0.8959670066833496,
      "learning_rate": 1.0742128967863085e-06,
      "loss": 0.2002,
      "num_input_tokens_seen": 31389888,
      "step": 36255
    },
    {
      "epoch": 17.095709570957094,
      "grad_norm": 0.9009622931480408,
      "learning_rate": 1.071367830118411e-06,
      "loss": 0.144,
      "num_input_tokens_seen": 31393648,
      "step": 36260
    },
    {
      "epoch": 17.0980669495521,
      "grad_norm": 1.3995202779769897,
      "learning_rate": 1.068526453546298e-06,
      "loss": 0.1766,
      "num_input_tokens_seen": 31397888,
      "step": 36265
    },
    {
      "epoch": 17.1004243281471,
      "grad_norm": 0.8631322383880615,
      "learning_rate": 1.0656887675081467e-06,
      "loss": 0.2185,
      "num_input_tokens_seen": 31402640,
      "step": 36270
    },
    {
      "epoch": 17.102781706742103,
      "grad_norm": 1.5720397233963013,
      "learning_rate": 1.0628547724415628e-06,
      "loss": 0.1156,
      "num_input_tokens_seen": 31406928,
      "step": 36275
    },
    {
      "epoch": 17.105139085337104,
      "grad_norm": 0.9860174059867859,
      "learning_rate": 1.0600244687835881e-06,
      "loss": 0.1307,
      "num_input_tokens_seen": 31411104,
      "step": 36280
    },
    {
      "epoch": 17.107496463932108,
      "grad_norm": 0.5187881588935852,
      "learning_rate": 1.0571978569706876e-06,
      "loss": 0.1164,
      "num_input_tokens_seen": 31415632,
      "step": 36285
    },
    {
      "epoch": 17.10985384252711,
      "grad_norm": 0.7642086148262024,
      "learning_rate": 1.0543749374387652e-06,
      "loss": 0.1474,
      "num_input_tokens_seen": 31420240,
      "step": 36290
    },
    {
      "epoch": 17.112211221122113,
      "grad_norm": 1.120782494544983,
      "learning_rate": 1.051555710623142e-06,
      "loss": 0.2064,
      "num_input_tokens_seen": 31424592,
      "step": 36295
    },
    {
      "epoch": 17.114568599717114,
      "grad_norm": 0.709530234336853,
      "learning_rate": 1.0487401769585847e-06,
      "loss": 0.1806,
      "num_input_tokens_seen": 31429568,
      "step": 36300
    },
    {
      "epoch": 17.116925978312118,
      "grad_norm": 0.5696868896484375,
      "learning_rate": 1.0459283368792845e-06,
      "loss": 0.1405,
      "num_input_tokens_seen": 31433520,
      "step": 36305
    },
    {
      "epoch": 17.11928335690712,
      "grad_norm": 1.283246397972107,
      "learning_rate": 1.043120190818858e-06,
      "loss": 0.215,
      "num_input_tokens_seen": 31439184,
      "step": 36310
    },
    {
      "epoch": 17.121640735502123,
      "grad_norm": 0.7335538864135742,
      "learning_rate": 1.0403157392103596e-06,
      "loss": 0.2215,
      "num_input_tokens_seen": 31443600,
      "step": 36315
    },
    {
      "epoch": 17.123998114097123,
      "grad_norm": 0.7014586329460144,
      "learning_rate": 1.0375149824862735e-06,
      "loss": 0.1591,
      "num_input_tokens_seen": 31448288,
      "step": 36320
    },
    {
      "epoch": 17.126355492692127,
      "grad_norm": 0.8983414769172668,
      "learning_rate": 1.034717921078507e-06,
      "loss": 0.1479,
      "num_input_tokens_seen": 31452864,
      "step": 36325
    },
    {
      "epoch": 17.128712871287128,
      "grad_norm": 1.0702909231185913,
      "learning_rate": 1.0319245554184009e-06,
      "loss": 0.176,
      "num_input_tokens_seen": 31457056,
      "step": 36330
    },
    {
      "epoch": 17.131070249882132,
      "grad_norm": 1.706595540046692,
      "learning_rate": 1.0291348859367361e-06,
      "loss": 0.202,
      "num_input_tokens_seen": 31461584,
      "step": 36335
    },
    {
      "epoch": 17.133427628477133,
      "grad_norm": 0.724419116973877,
      "learning_rate": 1.0263489130637016e-06,
      "loss": 0.142,
      "num_input_tokens_seen": 31465328,
      "step": 36340
    },
    {
      "epoch": 17.135785007072137,
      "grad_norm": 1.8654605150222778,
      "learning_rate": 1.0235666372289427e-06,
      "loss": 0.2776,
      "num_input_tokens_seen": 31469744,
      "step": 36345
    },
    {
      "epoch": 17.138142385667138,
      "grad_norm": 2.0007998943328857,
      "learning_rate": 1.0207880588615076e-06,
      "loss": 0.2476,
      "num_input_tokens_seen": 31473824,
      "step": 36350
    },
    {
      "epoch": 17.14049976426214,
      "grad_norm": 0.5278545022010803,
      "learning_rate": 1.0180131783898984e-06,
      "loss": 0.1428,
      "num_input_tokens_seen": 31477648,
      "step": 36355
    },
    {
      "epoch": 17.142857142857142,
      "grad_norm": 1.405354380607605,
      "learning_rate": 1.0152419962420362e-06,
      "loss": 0.13,
      "num_input_tokens_seen": 31481808,
      "step": 36360
    },
    {
      "epoch": 17.145214521452147,
      "grad_norm": 2.8157083988189697,
      "learning_rate": 1.0124745128452685e-06,
      "loss": 0.247,
      "num_input_tokens_seen": 31486016,
      "step": 36365
    },
    {
      "epoch": 17.147571900047147,
      "grad_norm": 1.5702835321426392,
      "learning_rate": 1.0097107286263758e-06,
      "loss": 0.2406,
      "num_input_tokens_seen": 31490192,
      "step": 36370
    },
    {
      "epoch": 17.14992927864215,
      "grad_norm": 0.6051886677742004,
      "learning_rate": 1.00695064401157e-06,
      "loss": 0.1359,
      "num_input_tokens_seen": 31494528,
      "step": 36375
    },
    {
      "epoch": 17.152286657237152,
      "grad_norm": 0.5362810492515564,
      "learning_rate": 1.0041942594264886e-06,
      "loss": 0.1217,
      "num_input_tokens_seen": 31498144,
      "step": 36380
    },
    {
      "epoch": 17.154644035832156,
      "grad_norm": 1.1808366775512695,
      "learning_rate": 1.001441575296208e-06,
      "loss": 0.2077,
      "num_input_tokens_seen": 31502688,
      "step": 36385
    },
    {
      "epoch": 17.157001414427157,
      "grad_norm": 0.4434151351451874,
      "learning_rate": 9.986925920452139e-07,
      "loss": 0.1077,
      "num_input_tokens_seen": 31507296,
      "step": 36390
    },
    {
      "epoch": 17.15935879302216,
      "grad_norm": 0.9832597970962524,
      "learning_rate": 9.959473100974475e-07,
      "loss": 0.2444,
      "num_input_tokens_seen": 31511680,
      "step": 36395
    },
    {
      "epoch": 17.16171617161716,
      "grad_norm": 0.8664050698280334,
      "learning_rate": 9.932057298762564e-07,
      "loss": 0.1819,
      "num_input_tokens_seen": 31515648,
      "step": 36400
    },
    {
      "epoch": 17.16171617161716,
      "eval_loss": 0.20057959854602814,
      "eval_runtime": 22.1529,
      "eval_samples_per_second": 42.568,
      "eval_steps_per_second": 21.306,
      "num_input_tokens_seen": 31515648,
      "step": 36400
    },
    {
      "epoch": 17.164073550212166,
      "grad_norm": 1.2681864500045776,
      "learning_rate": 9.90467851804433e-07,
      "loss": 0.196,
      "num_input_tokens_seen": 31520352,
      "step": 36405
    },
    {
      "epoch": 17.166430928807166,
      "grad_norm": 1.1819881200790405,
      "learning_rate": 9.877336763041895e-07,
      "loss": 0.1416,
      "num_input_tokens_seen": 31524528,
      "step": 36410
    },
    {
      "epoch": 17.16878830740217,
      "grad_norm": 0.957327663898468,
      "learning_rate": 9.850032037971662e-07,
      "loss": 0.1505,
      "num_input_tokens_seen": 31529040,
      "step": 36415
    },
    {
      "epoch": 17.17114568599717,
      "grad_norm": 0.47270381450653076,
      "learning_rate": 9.822764347044406e-07,
      "loss": 0.1737,
      "num_input_tokens_seen": 31533392,
      "step": 36420
    },
    {
      "epoch": 17.173503064592172,
      "grad_norm": 0.6382805109024048,
      "learning_rate": 9.795533694465175e-07,
      "loss": 0.2506,
      "num_input_tokens_seen": 31537488,
      "step": 36425
    },
    {
      "epoch": 17.175860443187176,
      "grad_norm": 0.417219877243042,
      "learning_rate": 9.768340084433197e-07,
      "loss": 0.1329,
      "num_input_tokens_seen": 31541952,
      "step": 36430
    },
    {
      "epoch": 17.178217821782177,
      "grad_norm": 1.2851651906967163,
      "learning_rate": 9.741183521142143e-07,
      "loss": 0.185,
      "num_input_tokens_seen": 31546480,
      "step": 36435
    },
    {
      "epoch": 17.18057520037718,
      "grad_norm": 0.5968962907791138,
      "learning_rate": 9.714064008779889e-07,
      "loss": 0.1802,
      "num_input_tokens_seen": 31551216,
      "step": 36440
    },
    {
      "epoch": 17.18293257897218,
      "grad_norm": 0.5476595163345337,
      "learning_rate": 9.686981551528584e-07,
      "loss": 0.2107,
      "num_input_tokens_seen": 31555744,
      "step": 36445
    },
    {
      "epoch": 17.185289957567186,
      "grad_norm": 1.538398265838623,
      "learning_rate": 9.65993615356467e-07,
      "loss": 0.1354,
      "num_input_tokens_seen": 31559920,
      "step": 36450
    },
    {
      "epoch": 17.187647336162186,
      "grad_norm": 0.4431716799736023,
      "learning_rate": 9.632927819058917e-07,
      "loss": 0.1616,
      "num_input_tokens_seen": 31563328,
      "step": 36455
    },
    {
      "epoch": 17.19000471475719,
      "grad_norm": 1.0324655771255493,
      "learning_rate": 9.605956552176305e-07,
      "loss": 0.2823,
      "num_input_tokens_seen": 31567808,
      "step": 36460
    },
    {
      "epoch": 17.19236209335219,
      "grad_norm": 0.5108117461204529,
      "learning_rate": 9.579022357076223e-07,
      "loss": 0.1583,
      "num_input_tokens_seen": 31572032,
      "step": 36465
    },
    {
      "epoch": 17.194719471947195,
      "grad_norm": 0.7185471057891846,
      "learning_rate": 9.552125237912158e-07,
      "loss": 0.1157,
      "num_input_tokens_seen": 31576048,
      "step": 36470
    },
    {
      "epoch": 17.197076850542196,
      "grad_norm": 0.6979689002037048,
      "learning_rate": 9.525265198832096e-07,
      "loss": 0.1537,
      "num_input_tokens_seen": 31580208,
      "step": 36475
    },
    {
      "epoch": 17.1994342291372,
      "grad_norm": 1.94569993019104,
      "learning_rate": 9.498442243978112e-07,
      "loss": 0.1522,
      "num_input_tokens_seen": 31584976,
      "step": 36480
    },
    {
      "epoch": 17.2017916077322,
      "grad_norm": 1.1850006580352783,
      "learning_rate": 9.471656377486649e-07,
      "loss": 0.2574,
      "num_input_tokens_seen": 31588768,
      "step": 36485
    },
    {
      "epoch": 17.204148986327205,
      "grad_norm": 0.8071943521499634,
      "learning_rate": 9.444907603488456e-07,
      "loss": 0.196,
      "num_input_tokens_seen": 31592832,
      "step": 36490
    },
    {
      "epoch": 17.206506364922205,
      "grad_norm": 0.7915976047515869,
      "learning_rate": 9.418195926108514e-07,
      "loss": 0.149,
      "num_input_tokens_seen": 31596736,
      "step": 36495
    },
    {
      "epoch": 17.20886374351721,
      "grad_norm": 2.2665648460388184,
      "learning_rate": 9.391521349466053e-07,
      "loss": 0.2699,
      "num_input_tokens_seen": 31602352,
      "step": 36500
    },
    {
      "epoch": 17.21122112211221,
      "grad_norm": 1.6265259981155396,
      "learning_rate": 9.364883877674758e-07,
      "loss": 0.2363,
      "num_input_tokens_seen": 31606976,
      "step": 36505
    },
    {
      "epoch": 17.213578500707214,
      "grad_norm": 0.9946079850196838,
      "learning_rate": 9.33828351484231e-07,
      "loss": 0.1997,
      "num_input_tokens_seen": 31610592,
      "step": 36510
    },
    {
      "epoch": 17.215935879302215,
      "grad_norm": 0.5395636558532715,
      "learning_rate": 9.311720265070906e-07,
      "loss": 0.1268,
      "num_input_tokens_seen": 31614768,
      "step": 36515
    },
    {
      "epoch": 17.21829325789722,
      "grad_norm": 0.5400208234786987,
      "learning_rate": 9.285194132456931e-07,
      "loss": 0.1585,
      "num_input_tokens_seen": 31618688,
      "step": 36520
    },
    {
      "epoch": 17.22065063649222,
      "grad_norm": 1.7451497316360474,
      "learning_rate": 9.258705121091032e-07,
      "loss": 0.1567,
      "num_input_tokens_seen": 31623152,
      "step": 36525
    },
    {
      "epoch": 17.223008015087224,
      "grad_norm": 0.65535569190979,
      "learning_rate": 9.232253235058136e-07,
      "loss": 0.1838,
      "num_input_tokens_seen": 31627040,
      "step": 36530
    },
    {
      "epoch": 17.225365393682225,
      "grad_norm": 1.1770778894424438,
      "learning_rate": 9.205838478437478e-07,
      "loss": 0.1901,
      "num_input_tokens_seen": 31632656,
      "step": 36535
    },
    {
      "epoch": 17.22772277227723,
      "grad_norm": 1.419547200202942,
      "learning_rate": 9.179460855302524e-07,
      "loss": 0.1189,
      "num_input_tokens_seen": 31636832,
      "step": 36540
    },
    {
      "epoch": 17.23008015087223,
      "grad_norm": 1.5849002599716187,
      "learning_rate": 9.153120369721046e-07,
      "loss": 0.1473,
      "num_input_tokens_seen": 31640992,
      "step": 36545
    },
    {
      "epoch": 17.232437529467234,
      "grad_norm": 0.6430978775024414,
      "learning_rate": 9.126817025755103e-07,
      "loss": 0.197,
      "num_input_tokens_seen": 31644992,
      "step": 36550
    },
    {
      "epoch": 17.234794908062234,
      "grad_norm": 1.014121174812317,
      "learning_rate": 9.100550827460947e-07,
      "loss": 0.132,
      "num_input_tokens_seen": 31649024,
      "step": 36555
    },
    {
      "epoch": 17.23715228665724,
      "grad_norm": 1.130042314529419,
      "learning_rate": 9.0743217788892e-07,
      "loss": 0.116,
      "num_input_tokens_seen": 31652752,
      "step": 36560
    },
    {
      "epoch": 17.23950966525224,
      "grad_norm": 0.5739092826843262,
      "learning_rate": 9.048129884084683e-07,
      "loss": 0.1232,
      "num_input_tokens_seen": 31657472,
      "step": 36565
    },
    {
      "epoch": 17.241867043847243,
      "grad_norm": 1.20316481590271,
      "learning_rate": 9.021975147086553e-07,
      "loss": 0.1664,
      "num_input_tokens_seen": 31661632,
      "step": 36570
    },
    {
      "epoch": 17.244224422442244,
      "grad_norm": 1.3792779445648193,
      "learning_rate": 8.995857571928141e-07,
      "loss": 0.1942,
      "num_input_tokens_seen": 31666256,
      "step": 36575
    },
    {
      "epoch": 17.246581801037248,
      "grad_norm": 1.6196379661560059,
      "learning_rate": 8.969777162637139e-07,
      "loss": 0.1732,
      "num_input_tokens_seen": 31670368,
      "step": 36580
    },
    {
      "epoch": 17.24893917963225,
      "grad_norm": 0.6283525228500366,
      "learning_rate": 8.943733923235525e-07,
      "loss": 0.1271,
      "num_input_tokens_seen": 31675328,
      "step": 36585
    },
    {
      "epoch": 17.251296558227253,
      "grad_norm": 1.774575114250183,
      "learning_rate": 8.917727857739394e-07,
      "loss": 0.1836,
      "num_input_tokens_seen": 31680224,
      "step": 36590
    },
    {
      "epoch": 17.253653936822253,
      "grad_norm": 0.6935626864433289,
      "learning_rate": 8.891758970159258e-07,
      "loss": 0.2382,
      "num_input_tokens_seen": 31684864,
      "step": 36595
    },
    {
      "epoch": 17.256011315417258,
      "grad_norm": 0.7847061157226562,
      "learning_rate": 8.86582726449986e-07,
      "loss": 0.2701,
      "num_input_tokens_seen": 31690208,
      "step": 36600
    },
    {
      "epoch": 17.256011315417258,
      "eval_loss": 0.2008049339056015,
      "eval_runtime": 22.1801,
      "eval_samples_per_second": 42.516,
      "eval_steps_per_second": 21.28,
      "num_input_tokens_seen": 31690208,
      "step": 36600
    },
    {
      "epoch": 17.25836869401226,
      "grad_norm": 1.7352336645126343,
      "learning_rate": 8.839932744760165e-07,
      "loss": 0.2336,
      "num_input_tokens_seen": 31694384,
      "step": 36605
    },
    {
      "epoch": 17.260726072607262,
      "grad_norm": 0.2372036874294281,
      "learning_rate": 8.814075414933482e-07,
      "loss": 0.2256,
      "num_input_tokens_seen": 31698480,
      "step": 36610
    },
    {
      "epoch": 17.263083451202263,
      "grad_norm": 1.0229891538619995,
      "learning_rate": 8.788255279007257e-07,
      "loss": 0.1887,
      "num_input_tokens_seen": 31702576,
      "step": 36615
    },
    {
      "epoch": 17.265440829797264,
      "grad_norm": 1.146142601966858,
      "learning_rate": 8.762472340963362e-07,
      "loss": 0.2322,
      "num_input_tokens_seen": 31707184,
      "step": 36620
    },
    {
      "epoch": 17.267798208392268,
      "grad_norm": 1.5541106462478638,
      "learning_rate": 8.736726604777811e-07,
      "loss": 0.2574,
      "num_input_tokens_seen": 31711552,
      "step": 36625
    },
    {
      "epoch": 17.27015558698727,
      "grad_norm": 1.5264310836791992,
      "learning_rate": 8.711018074420901e-07,
      "loss": 0.2053,
      "num_input_tokens_seen": 31716048,
      "step": 36630
    },
    {
      "epoch": 17.272512965582273,
      "grad_norm": 1.3443644046783447,
      "learning_rate": 8.685346753857209e-07,
      "loss": 0.1564,
      "num_input_tokens_seen": 31719872,
      "step": 36635
    },
    {
      "epoch": 17.274870344177273,
      "grad_norm": 0.6228989958763123,
      "learning_rate": 8.659712647045654e-07,
      "loss": 0.2126,
      "num_input_tokens_seen": 31724800,
      "step": 36640
    },
    {
      "epoch": 17.277227722772277,
      "grad_norm": 1.0431066751480103,
      "learning_rate": 8.634115757939209e-07,
      "loss": 0.1675,
      "num_input_tokens_seen": 31729344,
      "step": 36645
    },
    {
      "epoch": 17.279585101367278,
      "grad_norm": 2.019035577774048,
      "learning_rate": 8.608556090485387e-07,
      "loss": 0.1932,
      "num_input_tokens_seen": 31734240,
      "step": 36650
    },
    {
      "epoch": 17.281942479962282,
      "grad_norm": 1.5472052097320557,
      "learning_rate": 8.583033648625671e-07,
      "loss": 0.2274,
      "num_input_tokens_seen": 31738464,
      "step": 36655
    },
    {
      "epoch": 17.284299858557283,
      "grad_norm": 0.5432838201522827,
      "learning_rate": 8.557548436295998e-07,
      "loss": 0.1398,
      "num_input_tokens_seen": 31742816,
      "step": 36660
    },
    {
      "epoch": 17.286657237152287,
      "grad_norm": 0.95888751745224,
      "learning_rate": 8.532100457426556e-07,
      "loss": 0.2467,
      "num_input_tokens_seen": 31746960,
      "step": 36665
    },
    {
      "epoch": 17.289014615747288,
      "grad_norm": 0.8066625595092773,
      "learning_rate": 8.506689715941679e-07,
      "loss": 0.1851,
      "num_input_tokens_seen": 31752064,
      "step": 36670
    },
    {
      "epoch": 17.291371994342292,
      "grad_norm": 2.9121487140655518,
      "learning_rate": 8.481316215760011e-07,
      "loss": 0.3503,
      "num_input_tokens_seen": 31756368,
      "step": 36675
    },
    {
      "epoch": 17.293729372937293,
      "grad_norm": 0.9433070421218872,
      "learning_rate": 8.455979960794558e-07,
      "loss": 0.159,
      "num_input_tokens_seen": 31761120,
      "step": 36680
    },
    {
      "epoch": 17.296086751532297,
      "grad_norm": 0.8630431294441223,
      "learning_rate": 8.430680954952364e-07,
      "loss": 0.1996,
      "num_input_tokens_seen": 31766048,
      "step": 36685
    },
    {
      "epoch": 17.298444130127297,
      "grad_norm": 0.7714850902557373,
      "learning_rate": 8.405419202134974e-07,
      "loss": 0.1779,
      "num_input_tokens_seen": 31770368,
      "step": 36690
    },
    {
      "epoch": 17.3008015087223,
      "grad_norm": 0.7977162003517151,
      "learning_rate": 8.380194706237993e-07,
      "loss": 0.1655,
      "num_input_tokens_seen": 31774096,
      "step": 36695
    },
    {
      "epoch": 17.303158887317302,
      "grad_norm": 0.8574299216270447,
      "learning_rate": 8.355007471151366e-07,
      "loss": 0.1641,
      "num_input_tokens_seen": 31778576,
      "step": 36700
    },
    {
      "epoch": 17.305516265912306,
      "grad_norm": 1.1196328401565552,
      "learning_rate": 8.329857500759292e-07,
      "loss": 0.1483,
      "num_input_tokens_seen": 31783184,
      "step": 36705
    },
    {
      "epoch": 17.307873644507307,
      "grad_norm": 1.1000356674194336,
      "learning_rate": 8.304744798940194e-07,
      "loss": 0.2362,
      "num_input_tokens_seen": 31787840,
      "step": 36710
    },
    {
      "epoch": 17.31023102310231,
      "grad_norm": 1.4953525066375732,
      "learning_rate": 8.279669369566756e-07,
      "loss": 0.1448,
      "num_input_tokens_seen": 31792320,
      "step": 36715
    },
    {
      "epoch": 17.31258840169731,
      "grad_norm": 0.8192554116249084,
      "learning_rate": 8.254631216505993e-07,
      "loss": 0.2607,
      "num_input_tokens_seen": 31796272,
      "step": 36720
    },
    {
      "epoch": 17.314945780292316,
      "grad_norm": 1.1173526048660278,
      "learning_rate": 8.229630343619038e-07,
      "loss": 0.1305,
      "num_input_tokens_seen": 31800640,
      "step": 36725
    },
    {
      "epoch": 17.317303158887317,
      "grad_norm": 1.6434496641159058,
      "learning_rate": 8.204666754761392e-07,
      "loss": 0.2045,
      "num_input_tokens_seen": 31805872,
      "step": 36730
    },
    {
      "epoch": 17.31966053748232,
      "grad_norm": 0.9941911101341248,
      "learning_rate": 8.179740453782669e-07,
      "loss": 0.1748,
      "num_input_tokens_seen": 31810608,
      "step": 36735
    },
    {
      "epoch": 17.32201791607732,
      "grad_norm": 0.6466110944747925,
      "learning_rate": 8.154851444526907e-07,
      "loss": 0.1697,
      "num_input_tokens_seen": 31814688,
      "step": 36740
    },
    {
      "epoch": 17.324375294672326,
      "grad_norm": 1.2007535696029663,
      "learning_rate": 8.129999730832283e-07,
      "loss": 0.1382,
      "num_input_tokens_seen": 31819520,
      "step": 36745
    },
    {
      "epoch": 17.326732673267326,
      "grad_norm": 1.4048237800598145,
      "learning_rate": 8.105185316531178e-07,
      "loss": 0.1997,
      "num_input_tokens_seen": 31824912,
      "step": 36750
    },
    {
      "epoch": 17.32909005186233,
      "grad_norm": 1.7465941905975342,
      "learning_rate": 8.08040820545039e-07,
      "loss": 0.1561,
      "num_input_tokens_seen": 31829104,
      "step": 36755
    },
    {
      "epoch": 17.33144743045733,
      "grad_norm": 2.6390585899353027,
      "learning_rate": 8.055668401410782e-07,
      "loss": 0.1885,
      "num_input_tokens_seen": 31833824,
      "step": 36760
    },
    {
      "epoch": 17.333804809052335,
      "grad_norm": 0.8719215393066406,
      "learning_rate": 8.030965908227578e-07,
      "loss": 0.2413,
      "num_input_tokens_seen": 31837712,
      "step": 36765
    },
    {
      "epoch": 17.336162187647336,
      "grad_norm": 0.6484123468399048,
      "learning_rate": 8.006300729710203e-07,
      "loss": 0.1952,
      "num_input_tokens_seen": 31841824,
      "step": 36770
    },
    {
      "epoch": 17.33851956624234,
      "grad_norm": 1.6603071689605713,
      "learning_rate": 7.981672869662337e-07,
      "loss": 0.1668,
      "num_input_tokens_seen": 31845952,
      "step": 36775
    },
    {
      "epoch": 17.34087694483734,
      "grad_norm": 1.42790949344635,
      "learning_rate": 7.957082331881888e-07,
      "loss": 0.2222,
      "num_input_tokens_seen": 31850112,
      "step": 36780
    },
    {
      "epoch": 17.343234323432345,
      "grad_norm": 0.648501992225647,
      "learning_rate": 7.932529120161069e-07,
      "loss": 0.1934,
      "num_input_tokens_seen": 31853936,
      "step": 36785
    },
    {
      "epoch": 17.345591702027345,
      "grad_norm": 0.6947361826896667,
      "learning_rate": 7.908013238286243e-07,
      "loss": 0.1166,
      "num_input_tokens_seen": 31858480,
      "step": 36790
    },
    {
      "epoch": 17.34794908062235,
      "grad_norm": 1.1150749921798706,
      "learning_rate": 7.883534690038136e-07,
      "loss": 0.1569,
      "num_input_tokens_seen": 31863376,
      "step": 36795
    },
    {
      "epoch": 17.35030645921735,
      "grad_norm": 1.1654999256134033,
      "learning_rate": 7.859093479191559e-07,
      "loss": 0.1884,
      "num_input_tokens_seen": 31868288,
      "step": 36800
    },
    {
      "epoch": 17.35030645921735,
      "eval_loss": 0.20062482357025146,
      "eval_runtime": 22.1959,
      "eval_samples_per_second": 42.485,
      "eval_steps_per_second": 21.265,
      "num_input_tokens_seen": 31868288,
      "step": 36800
    },
    {
      "epoch": 17.352663837812354,
      "grad_norm": 1.1903594732284546,
      "learning_rate": 7.834689609515722e-07,
      "loss": 0.2009,
      "num_input_tokens_seen": 31872656,
      "step": 36805
    },
    {
      "epoch": 17.355021216407355,
      "grad_norm": 0.8989871144294739,
      "learning_rate": 7.810323084774002e-07,
      "loss": 0.1611,
      "num_input_tokens_seen": 31877552,
      "step": 36810
    },
    {
      "epoch": 17.35737859500236,
      "grad_norm": 1.9738421440124512,
      "learning_rate": 7.785993908723976e-07,
      "loss": 0.2127,
      "num_input_tokens_seen": 31882400,
      "step": 36815
    },
    {
      "epoch": 17.35973597359736,
      "grad_norm": 1.0088900327682495,
      "learning_rate": 7.761702085117534e-07,
      "loss": 0.2301,
      "num_input_tokens_seen": 31886560,
      "step": 36820
    },
    {
      "epoch": 17.36209335219236,
      "grad_norm": 0.7717536687850952,
      "learning_rate": 7.737447617700844e-07,
      "loss": 0.1481,
      "num_input_tokens_seen": 31890512,
      "step": 36825
    },
    {
      "epoch": 17.364450730787365,
      "grad_norm": 0.7029522061347961,
      "learning_rate": 7.713230510214136e-07,
      "loss": 0.128,
      "num_input_tokens_seen": 31895040,
      "step": 36830
    },
    {
      "epoch": 17.366808109382365,
      "grad_norm": 0.3507051169872284,
      "learning_rate": 7.689050766392092e-07,
      "loss": 0.2367,
      "num_input_tokens_seen": 31899312,
      "step": 36835
    },
    {
      "epoch": 17.36916548797737,
      "grad_norm": 1.4913949966430664,
      "learning_rate": 7.664908389963477e-07,
      "loss": 0.1359,
      "num_input_tokens_seen": 31904112,
      "step": 36840
    },
    {
      "epoch": 17.37152286657237,
      "grad_norm": 0.9847292900085449,
      "learning_rate": 7.64080338465134e-07,
      "loss": 0.2325,
      "num_input_tokens_seen": 31908048,
      "step": 36845
    },
    {
      "epoch": 17.373880245167374,
      "grad_norm": 0.7464202642440796,
      "learning_rate": 7.616735754173043e-07,
      "loss": 0.1722,
      "num_input_tokens_seen": 31911696,
      "step": 36850
    },
    {
      "epoch": 17.376237623762375,
      "grad_norm": 0.9852284789085388,
      "learning_rate": 7.592705502240005e-07,
      "loss": 0.1902,
      "num_input_tokens_seen": 31916944,
      "step": 36855
    },
    {
      "epoch": 17.37859500235738,
      "grad_norm": 1.1183439493179321,
      "learning_rate": 7.568712632558095e-07,
      "loss": 0.2148,
      "num_input_tokens_seen": 31921264,
      "step": 36860
    },
    {
      "epoch": 17.38095238095238,
      "grad_norm": 1.6206867694854736,
      "learning_rate": 7.544757148827297e-07,
      "loss": 0.1823,
      "num_input_tokens_seen": 31925520,
      "step": 36865
    },
    {
      "epoch": 17.383309759547384,
      "grad_norm": 0.5500126481056213,
      "learning_rate": 7.520839054741797e-07,
      "loss": 0.1509,
      "num_input_tokens_seen": 31929104,
      "step": 36870
    },
    {
      "epoch": 17.385667138142384,
      "grad_norm": 1.1644407510757446,
      "learning_rate": 7.496958353990113e-07,
      "loss": 0.1815,
      "num_input_tokens_seen": 31933136,
      "step": 36875
    },
    {
      "epoch": 17.38802451673739,
      "grad_norm": 1.2744765281677246,
      "learning_rate": 7.473115050254941e-07,
      "loss": 0.1899,
      "num_input_tokens_seen": 31936912,
      "step": 36880
    },
    {
      "epoch": 17.39038189533239,
      "grad_norm": 1.0391840934753418,
      "learning_rate": 7.449309147213173e-07,
      "loss": 0.168,
      "num_input_tokens_seen": 31941040,
      "step": 36885
    },
    {
      "epoch": 17.392739273927393,
      "grad_norm": 1.611101508140564,
      "learning_rate": 7.425540648536067e-07,
      "loss": 0.1345,
      "num_input_tokens_seen": 31945024,
      "step": 36890
    },
    {
      "epoch": 17.395096652522394,
      "grad_norm": 1.545087456703186,
      "learning_rate": 7.40180955788894e-07,
      "loss": 0.1873,
      "num_input_tokens_seen": 31949312,
      "step": 36895
    },
    {
      "epoch": 17.397454031117398,
      "grad_norm": 1.3865638971328735,
      "learning_rate": 7.378115878931474e-07,
      "loss": 0.203,
      "num_input_tokens_seen": 31953536,
      "step": 36900
    },
    {
      "epoch": 17.3998114097124,
      "grad_norm": 1.3731306791305542,
      "learning_rate": 7.354459615317527e-07,
      "loss": 0.2301,
      "num_input_tokens_seen": 31957680,
      "step": 36905
    },
    {
      "epoch": 17.402168788307403,
      "grad_norm": 0.5487167239189148,
      "learning_rate": 7.33084077069518e-07,
      "loss": 0.2387,
      "num_input_tokens_seen": 31962800,
      "step": 36910
    },
    {
      "epoch": 17.404526166902404,
      "grad_norm": 0.7201265096664429,
      "learning_rate": 7.307259348706768e-07,
      "loss": 0.17,
      "num_input_tokens_seen": 31967152,
      "step": 36915
    },
    {
      "epoch": 17.406883545497408,
      "grad_norm": 0.7957035303115845,
      "learning_rate": 7.283715352988801e-07,
      "loss": 0.1364,
      "num_input_tokens_seen": 31971056,
      "step": 36920
    },
    {
      "epoch": 17.40924092409241,
      "grad_norm": 1.0484082698822021,
      "learning_rate": 7.260208787172068e-07,
      "loss": 0.2328,
      "num_input_tokens_seen": 31975152,
      "step": 36925
    },
    {
      "epoch": 17.411598302687413,
      "grad_norm": 0.9885473847389221,
      "learning_rate": 7.23673965488167e-07,
      "loss": 0.1015,
      "num_input_tokens_seen": 31979184,
      "step": 36930
    },
    {
      "epoch": 17.413955681282413,
      "grad_norm": 0.6858915090560913,
      "learning_rate": 7.213307959736709e-07,
      "loss": 0.1979,
      "num_input_tokens_seen": 31982544,
      "step": 36935
    },
    {
      "epoch": 17.416313059877417,
      "grad_norm": 0.2312195897102356,
      "learning_rate": 7.189913705350715e-07,
      "loss": 0.1072,
      "num_input_tokens_seen": 31987136,
      "step": 36940
    },
    {
      "epoch": 17.418670438472418,
      "grad_norm": 0.8692322373390198,
      "learning_rate": 7.166556895331411e-07,
      "loss": 0.1657,
      "num_input_tokens_seen": 31992144,
      "step": 36945
    },
    {
      "epoch": 17.421027817067422,
      "grad_norm": 0.5843126773834229,
      "learning_rate": 7.143237533280639e-07,
      "loss": 0.2218,
      "num_input_tokens_seen": 31996064,
      "step": 36950
    },
    {
      "epoch": 17.423385195662423,
      "grad_norm": 0.8892970681190491,
      "learning_rate": 7.119955622794578e-07,
      "loss": 0.1691,
      "num_input_tokens_seen": 32000448,
      "step": 36955
    },
    {
      "epoch": 17.425742574257427,
      "grad_norm": 0.5978298783302307,
      "learning_rate": 7.096711167463577e-07,
      "loss": 0.1698,
      "num_input_tokens_seen": 32005968,
      "step": 36960
    },
    {
      "epoch": 17.428099952852428,
      "grad_norm": 0.8881653547286987,
      "learning_rate": 7.073504170872213e-07,
      "loss": 0.0719,
      "num_input_tokens_seen": 32010976,
      "step": 36965
    },
    {
      "epoch": 17.430457331447432,
      "grad_norm": 1.4122023582458496,
      "learning_rate": 7.05033463659932e-07,
      "loss": 0.2571,
      "num_input_tokens_seen": 32016032,
      "step": 36970
    },
    {
      "epoch": 17.432814710042432,
      "grad_norm": 0.8583844900131226,
      "learning_rate": 7.027202568217928e-07,
      "loss": 0.1607,
      "num_input_tokens_seen": 32020304,
      "step": 36975
    },
    {
      "epoch": 17.435172088637437,
      "grad_norm": 2.078917980194092,
      "learning_rate": 7.004107969295293e-07,
      "loss": 0.2404,
      "num_input_tokens_seen": 32024144,
      "step": 36980
    },
    {
      "epoch": 17.437529467232437,
      "grad_norm": 0.6249929666519165,
      "learning_rate": 6.9810508433929e-07,
      "loss": 0.189,
      "num_input_tokens_seen": 32028864,
      "step": 36985
    },
    {
      "epoch": 17.43988684582744,
      "grad_norm": 0.7269372344017029,
      "learning_rate": 6.958031194066406e-07,
      "loss": 0.2144,
      "num_input_tokens_seen": 32032688,
      "step": 36990
    },
    {
      "epoch": 17.442244224422442,
      "grad_norm": 1.1227132081985474,
      "learning_rate": 6.935049024865776e-07,
      "loss": 0.159,
      "num_input_tokens_seen": 32037008,
      "step": 36995
    },
    {
      "epoch": 17.444601603017446,
      "grad_norm": 1.725088357925415,
      "learning_rate": 6.912104339335118e-07,
      "loss": 0.1824,
      "num_input_tokens_seen": 32041536,
      "step": 37000
    },
    {
      "epoch": 17.444601603017446,
      "eval_loss": 0.20077739655971527,
      "eval_runtime": 22.1483,
      "eval_samples_per_second": 42.577,
      "eval_steps_per_second": 21.311,
      "num_input_tokens_seen": 32041536,
      "step": 37000
    },
    {
      "epoch": 17.446958981612447,
      "grad_norm": 1.5868110656738281,
      "learning_rate": 6.889197141012799e-07,
      "loss": 0.2144,
      "num_input_tokens_seen": 32045568,
      "step": 37005
    },
    {
      "epoch": 17.44931636020745,
      "grad_norm": 1.2257155179977417,
      "learning_rate": 6.866327433431435e-07,
      "loss": 0.1517,
      "num_input_tokens_seen": 32050144,
      "step": 37010
    },
    {
      "epoch": 17.45167373880245,
      "grad_norm": 0.8600960373878479,
      "learning_rate": 6.843495220117735e-07,
      "loss": 0.1501,
      "num_input_tokens_seen": 32055296,
      "step": 37015
    },
    {
      "epoch": 17.454031117397456,
      "grad_norm": 0.7299240231513977,
      "learning_rate": 6.820700504592798e-07,
      "loss": 0.1657,
      "num_input_tokens_seen": 32059872,
      "step": 37020
    },
    {
      "epoch": 17.456388495992456,
      "grad_norm": 0.9556058049201965,
      "learning_rate": 6.797943290371839e-07,
      "loss": 0.1507,
      "num_input_tokens_seen": 32064128,
      "step": 37025
    },
    {
      "epoch": 17.458745874587457,
      "grad_norm": 0.6405320763587952,
      "learning_rate": 6.775223580964274e-07,
      "loss": 0.1467,
      "num_input_tokens_seen": 32069232,
      "step": 37030
    },
    {
      "epoch": 17.46110325318246,
      "grad_norm": 1.1180038452148438,
      "learning_rate": 6.7525413798738e-07,
      "loss": 0.1346,
      "num_input_tokens_seen": 32073760,
      "step": 37035
    },
    {
      "epoch": 17.463460631777462,
      "grad_norm": 0.5357829928398132,
      "learning_rate": 6.729896690598259e-07,
      "loss": 0.1523,
      "num_input_tokens_seen": 32078944,
      "step": 37040
    },
    {
      "epoch": 17.465818010372466,
      "grad_norm": 0.904794454574585,
      "learning_rate": 6.707289516629772e-07,
      "loss": 0.2002,
      "num_input_tokens_seen": 32082992,
      "step": 37045
    },
    {
      "epoch": 17.468175388967467,
      "grad_norm": 1.3772053718566895,
      "learning_rate": 6.684719861454692e-07,
      "loss": 0.1631,
      "num_input_tokens_seen": 32087584,
      "step": 37050
    },
    {
      "epoch": 17.47053276756247,
      "grad_norm": 0.642842710018158,
      "learning_rate": 6.662187728553481e-07,
      "loss": 0.1716,
      "num_input_tokens_seen": 32091600,
      "step": 37055
    },
    {
      "epoch": 17.47289014615747,
      "grad_norm": 0.6495888233184814,
      "learning_rate": 6.639693121400892e-07,
      "loss": 0.1589,
      "num_input_tokens_seen": 32096144,
      "step": 37060
    },
    {
      "epoch": 17.475247524752476,
      "grad_norm": 0.9549074769020081,
      "learning_rate": 6.617236043465868e-07,
      "loss": 0.1755,
      "num_input_tokens_seen": 32100624,
      "step": 37065
    },
    {
      "epoch": 17.477604903347476,
      "grad_norm": 0.9705127477645874,
      "learning_rate": 6.594816498211587e-07,
      "loss": 0.1934,
      "num_input_tokens_seen": 32104768,
      "step": 37070
    },
    {
      "epoch": 17.47996228194248,
      "grad_norm": 1.8882722854614258,
      "learning_rate": 6.572434489095447e-07,
      "loss": 0.1965,
      "num_input_tokens_seen": 32108448,
      "step": 37075
    },
    {
      "epoch": 17.48231966053748,
      "grad_norm": 1.06270170211792,
      "learning_rate": 6.550090019568994e-07,
      "loss": 0.178,
      "num_input_tokens_seen": 32112784,
      "step": 37080
    },
    {
      "epoch": 17.484677039132485,
      "grad_norm": 1.486316204071045,
      "learning_rate": 6.527783093078027e-07,
      "loss": 0.1186,
      "num_input_tokens_seen": 32117104,
      "step": 37085
    },
    {
      "epoch": 17.487034417727486,
      "grad_norm": 1.6991031169891357,
      "learning_rate": 6.5055137130626e-07,
      "loss": 0.2165,
      "num_input_tokens_seen": 32121984,
      "step": 37090
    },
    {
      "epoch": 17.48939179632249,
      "grad_norm": 1.3473559617996216,
      "learning_rate": 6.483281882956854e-07,
      "loss": 0.1913,
      "num_input_tokens_seen": 32126096,
      "step": 37095
    },
    {
      "epoch": 17.49174917491749,
      "grad_norm": 0.6304011344909668,
      "learning_rate": 6.461087606189298e-07,
      "loss": 0.1489,
      "num_input_tokens_seen": 32130016,
      "step": 37100
    },
    {
      "epoch": 17.494106553512495,
      "grad_norm": 1.345908284187317,
      "learning_rate": 6.438930886182554e-07,
      "loss": 0.223,
      "num_input_tokens_seen": 32133632,
      "step": 37105
    },
    {
      "epoch": 17.496463932107496,
      "grad_norm": 1.033267855644226,
      "learning_rate": 6.416811726353417e-07,
      "loss": 0.2031,
      "num_input_tokens_seen": 32137472,
      "step": 37110
    },
    {
      "epoch": 17.4988213107025,
      "grad_norm": 1.5090705156326294,
      "learning_rate": 6.394730130112991e-07,
      "loss": 0.0917,
      "num_input_tokens_seen": 32141472,
      "step": 37115
    },
    {
      "epoch": 17.5011786892975,
      "grad_norm": 1.5930025577545166,
      "learning_rate": 6.372686100866471e-07,
      "loss": 0.2082,
      "num_input_tokens_seen": 32145840,
      "step": 37120
    },
    {
      "epoch": 17.503536067892504,
      "grad_norm": 2.5951151847839355,
      "learning_rate": 6.350679642013413e-07,
      "loss": 0.1772,
      "num_input_tokens_seen": 32149664,
      "step": 37125
    },
    {
      "epoch": 17.505893446487505,
      "grad_norm": 1.4903161525726318,
      "learning_rate": 6.328710756947437e-07,
      "loss": 0.1543,
      "num_input_tokens_seen": 32154208,
      "step": 37130
    },
    {
      "epoch": 17.50825082508251,
      "grad_norm": 0.6676379442214966,
      "learning_rate": 6.306779449056416e-07,
      "loss": 0.1486,
      "num_input_tokens_seen": 32158688,
      "step": 37135
    },
    {
      "epoch": 17.51060820367751,
      "grad_norm": 1.5299878120422363,
      "learning_rate": 6.284885721722422e-07,
      "loss": 0.1908,
      "num_input_tokens_seen": 32162512,
      "step": 37140
    },
    {
      "epoch": 17.512965582272514,
      "grad_norm": 0.7585579752922058,
      "learning_rate": 6.26302957832181e-07,
      "loss": 0.2015,
      "num_input_tokens_seen": 32166288,
      "step": 37145
    },
    {
      "epoch": 17.515322960867515,
      "grad_norm": 0.8864753246307373,
      "learning_rate": 6.241211022224997e-07,
      "loss": 0.1272,
      "num_input_tokens_seen": 32171024,
      "step": 37150
    },
    {
      "epoch": 17.51768033946252,
      "grad_norm": 0.6825023293495178,
      "learning_rate": 6.219430056796732e-07,
      "loss": 0.1581,
      "num_input_tokens_seen": 32174496,
      "step": 37155
    },
    {
      "epoch": 17.52003771805752,
      "grad_norm": 0.7812500596046448,
      "learning_rate": 6.19768668539586e-07,
      "loss": 0.2394,
      "num_input_tokens_seen": 32179120,
      "step": 37160
    },
    {
      "epoch": 17.522395096652524,
      "grad_norm": 1.2399399280548096,
      "learning_rate": 6.175980911375528e-07,
      "loss": 0.1901,
      "num_input_tokens_seen": 32182800,
      "step": 37165
    },
    {
      "epoch": 17.524752475247524,
      "grad_norm": 1.6595323085784912,
      "learning_rate": 6.154312738083034e-07,
      "loss": 0.1661,
      "num_input_tokens_seen": 32187936,
      "step": 37170
    },
    {
      "epoch": 17.52710985384253,
      "grad_norm": 0.9339540004730225,
      "learning_rate": 6.132682168859843e-07,
      "loss": 0.209,
      "num_input_tokens_seen": 32191984,
      "step": 37175
    },
    {
      "epoch": 17.52946723243753,
      "grad_norm": 0.7817550301551819,
      "learning_rate": 6.111089207041704e-07,
      "loss": 0.1612,
      "num_input_tokens_seen": 32196512,
      "step": 37180
    },
    {
      "epoch": 17.531824611032533,
      "grad_norm": 0.6174553632736206,
      "learning_rate": 6.089533855958507e-07,
      "loss": 0.2865,
      "num_input_tokens_seen": 32201168,
      "step": 37185
    },
    {
      "epoch": 17.534181989627534,
      "grad_norm": 0.6154071092605591,
      "learning_rate": 6.068016118934372e-07,
      "loss": 0.1007,
      "num_input_tokens_seen": 32205120,
      "step": 37190
    },
    {
      "epoch": 17.536539368222538,
      "grad_norm": 1.2883553504943848,
      "learning_rate": 6.04653599928759e-07,
      "loss": 0.1663,
      "num_input_tokens_seen": 32209440,
      "step": 37195
    },
    {
      "epoch": 17.53889674681754,
      "grad_norm": 1.3257076740264893,
      "learning_rate": 6.025093500330675e-07,
      "loss": 0.2799,
      "num_input_tokens_seen": 32213584,
      "step": 37200
    },
    {
      "epoch": 17.53889674681754,
      "eval_loss": 0.20078390836715698,
      "eval_runtime": 22.1857,
      "eval_samples_per_second": 42.505,
      "eval_steps_per_second": 21.275,
      "num_input_tokens_seen": 32213584,
      "step": 37200
    },
    {
      "epoch": 17.541254125412543,
      "grad_norm": 1.239426851272583,
      "learning_rate": 6.003688625370291e-07,
      "loss": 0.1681,
      "num_input_tokens_seen": 32217664,
      "step": 37205
    },
    {
      "epoch": 17.543611504007544,
      "grad_norm": 0.5042837858200073,
      "learning_rate": 5.982321377707406e-07,
      "loss": 0.1583,
      "num_input_tokens_seen": 32222288,
      "step": 37210
    },
    {
      "epoch": 17.545968882602544,
      "grad_norm": 0.6293020844459534,
      "learning_rate": 5.96099176063708e-07,
      "loss": 0.2656,
      "num_input_tokens_seen": 32226720,
      "step": 37215
    },
    {
      "epoch": 17.54832626119755,
      "grad_norm": 0.7501979470252991,
      "learning_rate": 5.93969977744857e-07,
      "loss": 0.1162,
      "num_input_tokens_seen": 32230160,
      "step": 37220
    },
    {
      "epoch": 17.55068363979255,
      "grad_norm": 0.9191104173660278,
      "learning_rate": 5.918445431425445e-07,
      "loss": 0.2299,
      "num_input_tokens_seen": 32234784,
      "step": 37225
    },
    {
      "epoch": 17.553041018387553,
      "grad_norm": 1.8737374544143677,
      "learning_rate": 5.897228725845333e-07,
      "loss": 0.225,
      "num_input_tokens_seen": 32239920,
      "step": 37230
    },
    {
      "epoch": 17.555398396982554,
      "grad_norm": 0.9490841031074524,
      "learning_rate": 5.876049663980171e-07,
      "loss": 0.1926,
      "num_input_tokens_seen": 32243888,
      "step": 37235
    },
    {
      "epoch": 17.557755775577558,
      "grad_norm": 2.1173095703125,
      "learning_rate": 5.854908249095959e-07,
      "loss": 0.1557,
      "num_input_tokens_seen": 32248480,
      "step": 37240
    },
    {
      "epoch": 17.56011315417256,
      "grad_norm": 0.10232816636562347,
      "learning_rate": 5.833804484453031e-07,
      "loss": 0.1377,
      "num_input_tokens_seen": 32252848,
      "step": 37245
    },
    {
      "epoch": 17.562470532767563,
      "grad_norm": 0.9402403831481934,
      "learning_rate": 5.81273837330587e-07,
      "loss": 0.2122,
      "num_input_tokens_seen": 32257040,
      "step": 37250
    },
    {
      "epoch": 17.564827911362563,
      "grad_norm": 0.3104332387447357,
      "learning_rate": 5.791709918903071e-07,
      "loss": 0.1718,
      "num_input_tokens_seen": 32260768,
      "step": 37255
    },
    {
      "epoch": 17.567185289957568,
      "grad_norm": 1.7639826536178589,
      "learning_rate": 5.770719124487483e-07,
      "loss": 0.3251,
      "num_input_tokens_seen": 32264688,
      "step": 37260
    },
    {
      "epoch": 17.569542668552568,
      "grad_norm": 1.7234513759613037,
      "learning_rate": 5.749765993296241e-07,
      "loss": 0.2337,
      "num_input_tokens_seen": 32269792,
      "step": 37265
    },
    {
      "epoch": 17.571900047147572,
      "grad_norm": 1.7735072374343872,
      "learning_rate": 5.728850528560509e-07,
      "loss": 0.157,
      "num_input_tokens_seen": 32274320,
      "step": 37270
    },
    {
      "epoch": 17.574257425742573,
      "grad_norm": 1.8057442903518677,
      "learning_rate": 5.707972733505707e-07,
      "loss": 0.2719,
      "num_input_tokens_seen": 32278816,
      "step": 37275
    },
    {
      "epoch": 17.576614804337577,
      "grad_norm": 1.4884535074234009,
      "learning_rate": 5.687132611351509e-07,
      "loss": 0.1625,
      "num_input_tokens_seen": 32282816,
      "step": 37280
    },
    {
      "epoch": 17.578972182932578,
      "grad_norm": 0.5292044281959534,
      "learning_rate": 5.666330165311651e-07,
      "loss": 0.1763,
      "num_input_tokens_seen": 32286432,
      "step": 37285
    },
    {
      "epoch": 17.581329561527582,
      "grad_norm": 0.7966823577880859,
      "learning_rate": 5.645565398594204e-07,
      "loss": 0.1323,
      "num_input_tokens_seen": 32290608,
      "step": 37290
    },
    {
      "epoch": 17.583686940122583,
      "grad_norm": 1.6519352197647095,
      "learning_rate": 5.624838314401304e-07,
      "loss": 0.1899,
      "num_input_tokens_seen": 32295344,
      "step": 37295
    },
    {
      "epoch": 17.586044318717587,
      "grad_norm": 0.8098044991493225,
      "learning_rate": 5.604148915929336e-07,
      "loss": 0.1957,
      "num_input_tokens_seen": 32299728,
      "step": 37300
    },
    {
      "epoch": 17.588401697312587,
      "grad_norm": 0.390343576669693,
      "learning_rate": 5.583497206368887e-07,
      "loss": 0.1916,
      "num_input_tokens_seen": 32303872,
      "step": 37305
    },
    {
      "epoch": 17.59075907590759,
      "grad_norm": 1.9023994207382202,
      "learning_rate": 5.562883188904688e-07,
      "loss": 0.1272,
      "num_input_tokens_seen": 32307600,
      "step": 37310
    },
    {
      "epoch": 17.593116454502592,
      "grad_norm": 1.4946849346160889,
      "learning_rate": 5.542306866715724e-07,
      "loss": 0.1594,
      "num_input_tokens_seen": 32312016,
      "step": 37315
    },
    {
      "epoch": 17.595473833097596,
      "grad_norm": 1.1097692251205444,
      "learning_rate": 5.52176824297504e-07,
      "loss": 0.2673,
      "num_input_tokens_seen": 32316160,
      "step": 37320
    },
    {
      "epoch": 17.597831211692597,
      "grad_norm": 1.4208110570907593,
      "learning_rate": 5.501267320850018e-07,
      "loss": 0.1978,
      "num_input_tokens_seen": 32320480,
      "step": 37325
    },
    {
      "epoch": 17.6001885902876,
      "grad_norm": 1.5651742219924927,
      "learning_rate": 5.480804103502157e-07,
      "loss": 0.1384,
      "num_input_tokens_seen": 32325520,
      "step": 37330
    },
    {
      "epoch": 17.602545968882602,
      "grad_norm": 1.0679144859313965,
      "learning_rate": 5.460378594087101e-07,
      "loss": 0.1814,
      "num_input_tokens_seen": 32330288,
      "step": 37335
    },
    {
      "epoch": 17.604903347477606,
      "grad_norm": 0.46159738302230835,
      "learning_rate": 5.439990795754773e-07,
      "loss": 0.1729,
      "num_input_tokens_seen": 32333824,
      "step": 37340
    },
    {
      "epoch": 17.607260726072607,
      "grad_norm": 0.9982948899269104,
      "learning_rate": 5.419640711649188e-07,
      "loss": 0.2233,
      "num_input_tokens_seen": 32338240,
      "step": 37345
    },
    {
      "epoch": 17.60961810466761,
      "grad_norm": 1.3220844268798828,
      "learning_rate": 5.399328344908583e-07,
      "loss": 0.2158,
      "num_input_tokens_seen": 32342336,
      "step": 37350
    },
    {
      "epoch": 17.61197548326261,
      "grad_norm": 0.8798462748527527,
      "learning_rate": 5.379053698665399e-07,
      "loss": 0.1657,
      "num_input_tokens_seen": 32346512,
      "step": 37355
    },
    {
      "epoch": 17.614332861857616,
      "grad_norm": 1.7254228591918945,
      "learning_rate": 5.358816776046216e-07,
      "loss": 0.2565,
      "num_input_tokens_seen": 32350672,
      "step": 37360
    },
    {
      "epoch": 17.616690240452616,
      "grad_norm": 1.801082730293274,
      "learning_rate": 5.338617580171817e-07,
      "loss": 0.2457,
      "num_input_tokens_seen": 32355408,
      "step": 37365
    },
    {
      "epoch": 17.61904761904762,
      "grad_norm": 1.511832356452942,
      "learning_rate": 5.318456114157239e-07,
      "loss": 0.2363,
      "num_input_tokens_seen": 32359648,
      "step": 37370
    },
    {
      "epoch": 17.62140499764262,
      "grad_norm": 1.670554280281067,
      "learning_rate": 5.298332381111576e-07,
      "loss": 0.2042,
      "num_input_tokens_seen": 32363952,
      "step": 37375
    },
    {
      "epoch": 17.623762376237625,
      "grad_norm": 1.5555421113967896,
      "learning_rate": 5.27824638413818e-07,
      "loss": 0.2218,
      "num_input_tokens_seen": 32368544,
      "step": 37380
    },
    {
      "epoch": 17.626119754832626,
      "grad_norm": 0.7407004237174988,
      "learning_rate": 5.258198126334546e-07,
      "loss": 0.1267,
      "num_input_tokens_seen": 32373184,
      "step": 37385
    },
    {
      "epoch": 17.62847713342763,
      "grad_norm": 1.205657720565796,
      "learning_rate": 5.238187610792367e-07,
      "loss": 0.1661,
      "num_input_tokens_seen": 32377232,
      "step": 37390
    },
    {
      "epoch": 17.63083451202263,
      "grad_norm": 0.6848780512809753,
      "learning_rate": 5.218214840597563e-07,
      "loss": 0.1826,
      "num_input_tokens_seen": 32381008,
      "step": 37395
    },
    {
      "epoch": 17.633191890617635,
      "grad_norm": 2.4926679134368896,
      "learning_rate": 5.198279818830115e-07,
      "loss": 0.192,
      "num_input_tokens_seen": 32385920,
      "step": 37400
    },
    {
      "epoch": 17.633191890617635,
      "eval_loss": 0.19931480288505554,
      "eval_runtime": 22.1745,
      "eval_samples_per_second": 42.526,
      "eval_steps_per_second": 21.286,
      "num_input_tokens_seen": 32385920,
      "step": 37400
    },
    {
      "epoch": 17.635549269212635,
      "grad_norm": 0.9538873434066772,
      "learning_rate": 5.178382548564287e-07,
      "loss": 0.1529,
      "num_input_tokens_seen": 32390528,
      "step": 37405
    },
    {
      "epoch": 17.63790664780764,
      "grad_norm": 0.6887389421463013,
      "learning_rate": 5.15852303286854e-07,
      "loss": 0.2392,
      "num_input_tokens_seen": 32395040,
      "step": 37410
    },
    {
      "epoch": 17.64026402640264,
      "grad_norm": 0.7538397312164307,
      "learning_rate": 5.138701274805396e-07,
      "loss": 0.1554,
      "num_input_tokens_seen": 32399440,
      "step": 37415
    },
    {
      "epoch": 17.64262140499764,
      "grad_norm": 1.6397072076797485,
      "learning_rate": 5.118917277431606e-07,
      "loss": 0.2257,
      "num_input_tokens_seen": 32403184,
      "step": 37420
    },
    {
      "epoch": 17.644978783592645,
      "grad_norm": 1.0567725896835327,
      "learning_rate": 5.099171043798145e-07,
      "loss": 0.1356,
      "num_input_tokens_seen": 32406848,
      "step": 37425
    },
    {
      "epoch": 17.647336162187646,
      "grad_norm": 1.5985381603240967,
      "learning_rate": 5.079462576950133e-07,
      "loss": 0.1685,
      "num_input_tokens_seen": 32411280,
      "step": 37430
    },
    {
      "epoch": 17.64969354078265,
      "grad_norm": 0.6114069223403931,
      "learning_rate": 5.059791879926862e-07,
      "loss": 0.1976,
      "num_input_tokens_seen": 32416000,
      "step": 37435
    },
    {
      "epoch": 17.65205091937765,
      "grad_norm": 0.6045661568641663,
      "learning_rate": 5.040158955761793e-07,
      "loss": 0.2038,
      "num_input_tokens_seen": 32419760,
      "step": 37440
    },
    {
      "epoch": 17.654408297972655,
      "grad_norm": 0.5099557042121887,
      "learning_rate": 5.020563807482559e-07,
      "loss": 0.1644,
      "num_input_tokens_seen": 32423408,
      "step": 37445
    },
    {
      "epoch": 17.656765676567655,
      "grad_norm": 1.3051996231079102,
      "learning_rate": 5.001006438110995e-07,
      "loss": 0.2216,
      "num_input_tokens_seen": 32427728,
      "step": 37450
    },
    {
      "epoch": 17.65912305516266,
      "grad_norm": 0.9606065154075623,
      "learning_rate": 4.981486850663075e-07,
      "loss": 0.1677,
      "num_input_tokens_seen": 32431920,
      "step": 37455
    },
    {
      "epoch": 17.66148043375766,
      "grad_norm": 1.7426106929779053,
      "learning_rate": 4.962005048149005e-07,
      "loss": 0.1476,
      "num_input_tokens_seen": 32435696,
      "step": 37460
    },
    {
      "epoch": 17.663837812352664,
      "grad_norm": 1.1371511220932007,
      "learning_rate": 4.942561033573073e-07,
      "loss": 0.1121,
      "num_input_tokens_seen": 32440208,
      "step": 37465
    },
    {
      "epoch": 17.666195190947665,
      "grad_norm": 1.1910133361816406,
      "learning_rate": 4.923154809933827e-07,
      "loss": 0.1197,
      "num_input_tokens_seen": 32444576,
      "step": 37470
    },
    {
      "epoch": 17.66855256954267,
      "grad_norm": 1.2710044384002686,
      "learning_rate": 4.903786380223957e-07,
      "loss": 0.2117,
      "num_input_tokens_seen": 32448864,
      "step": 37475
    },
    {
      "epoch": 17.67090994813767,
      "grad_norm": 1.0751802921295166,
      "learning_rate": 4.884455747430266e-07,
      "loss": 0.1845,
      "num_input_tokens_seen": 32452800,
      "step": 37480
    },
    {
      "epoch": 17.673267326732674,
      "grad_norm": 0.4758540987968445,
      "learning_rate": 4.865162914533816e-07,
      "loss": 0.2,
      "num_input_tokens_seen": 32456576,
      "step": 37485
    },
    {
      "epoch": 17.675624705327674,
      "grad_norm": 0.997033417224884,
      "learning_rate": 4.845907884509809e-07,
      "loss": 0.1767,
      "num_input_tokens_seen": 32461504,
      "step": 37490
    },
    {
      "epoch": 17.67798208392268,
      "grad_norm": 1.7002713680267334,
      "learning_rate": 4.82669066032762e-07,
      "loss": 0.1748,
      "num_input_tokens_seen": 32465904,
      "step": 37495
    },
    {
      "epoch": 17.68033946251768,
      "grad_norm": 0.7464017271995544,
      "learning_rate": 4.807511244950768e-07,
      "loss": 0.2065,
      "num_input_tokens_seen": 32469888,
      "step": 37500
    },
    {
      "epoch": 17.682696841112683,
      "grad_norm": 1.2659770250320435,
      "learning_rate": 4.788369641336943e-07,
      "loss": 0.1923,
      "num_input_tokens_seen": 32473696,
      "step": 37505
    },
    {
      "epoch": 17.685054219707684,
      "grad_norm": 0.6443919539451599,
      "learning_rate": 4.769265852438032e-07,
      "loss": 0.1496,
      "num_input_tokens_seen": 32478128,
      "step": 37510
    },
    {
      "epoch": 17.68741159830269,
      "grad_norm": 0.789391040802002,
      "learning_rate": 4.750199881200124e-07,
      "loss": 0.1824,
      "num_input_tokens_seen": 32482352,
      "step": 37515
    },
    {
      "epoch": 17.68976897689769,
      "grad_norm": 0.9886108040809631,
      "learning_rate": 4.7311717305633664e-07,
      "loss": 0.11,
      "num_input_tokens_seen": 32486432,
      "step": 37520
    },
    {
      "epoch": 17.692126355492693,
      "grad_norm": 1.8072378635406494,
      "learning_rate": 4.7121814034621623e-07,
      "loss": 0.1813,
      "num_input_tokens_seen": 32490784,
      "step": 37525
    },
    {
      "epoch": 17.694483734087694,
      "grad_norm": 1.0207395553588867,
      "learning_rate": 4.693228902825114e-07,
      "loss": 0.1714,
      "num_input_tokens_seen": 32495744,
      "step": 37530
    },
    {
      "epoch": 17.696841112682698,
      "grad_norm": 1.6589508056640625,
      "learning_rate": 4.6743142315748277e-07,
      "loss": 0.1528,
      "num_input_tokens_seen": 32499408,
      "step": 37535
    },
    {
      "epoch": 17.6991984912777,
      "grad_norm": 0.41480135917663574,
      "learning_rate": 4.655437392628276e-07,
      "loss": 0.1316,
      "num_input_tokens_seen": 32504096,
      "step": 37540
    },
    {
      "epoch": 17.701555869872703,
      "grad_norm": 1.100286841392517,
      "learning_rate": 4.636598388896463e-07,
      "loss": 0.1225,
      "num_input_tokens_seen": 32508144,
      "step": 37545
    },
    {
      "epoch": 17.703913248467703,
      "grad_norm": 1.0903284549713135,
      "learning_rate": 4.6177972232845925e-07,
      "loss": 0.1737,
      "num_input_tokens_seen": 32512352,
      "step": 37550
    },
    {
      "epoch": 17.706270627062707,
      "grad_norm": 0.8271995782852173,
      "learning_rate": 4.5990338986920953e-07,
      "loss": 0.1429,
      "num_input_tokens_seen": 32516576,
      "step": 37555
    },
    {
      "epoch": 17.708628005657708,
      "grad_norm": 1.3095803260803223,
      "learning_rate": 4.5803084180124633e-07,
      "loss": 0.2419,
      "num_input_tokens_seen": 32521264,
      "step": 37560
    },
    {
      "epoch": 17.710985384252712,
      "grad_norm": 0.6881694793701172,
      "learning_rate": 4.561620784133386e-07,
      "loss": 0.1908,
      "num_input_tokens_seen": 32525456,
      "step": 37565
    },
    {
      "epoch": 17.713342762847713,
      "grad_norm": 1.7165199518203735,
      "learning_rate": 4.5429709999367796e-07,
      "loss": 0.1935,
      "num_input_tokens_seen": 32529648,
      "step": 37570
    },
    {
      "epoch": 17.715700141442717,
      "grad_norm": 0.7841071486473083,
      "learning_rate": 4.5243590682986223e-07,
      "loss": 0.1891,
      "num_input_tokens_seen": 32534176,
      "step": 37575
    },
    {
      "epoch": 17.718057520037718,
      "grad_norm": 1.1241283416748047,
      "learning_rate": 4.5057849920891735e-07,
      "loss": 0.1319,
      "num_input_tokens_seen": 32538080,
      "step": 37580
    },
    {
      "epoch": 17.720414898632722,
      "grad_norm": 0.835374653339386,
      "learning_rate": 4.487248774172698e-07,
      "loss": 0.1881,
      "num_input_tokens_seen": 32542848,
      "step": 37585
    },
    {
      "epoch": 17.722772277227723,
      "grad_norm": 0.5177331566810608,
      "learning_rate": 4.4687504174077965e-07,
      "loss": 0.1547,
      "num_input_tokens_seen": 32547456,
      "step": 37590
    },
    {
      "epoch": 17.725129655822727,
      "grad_norm": 0.7886555194854736,
      "learning_rate": 4.450289924647133e-07,
      "loss": 0.1966,
      "num_input_tokens_seen": 32551168,
      "step": 37595
    },
    {
      "epoch": 17.727487034417727,
      "grad_norm": 0.6524661779403687,
      "learning_rate": 4.431867298737513e-07,
      "loss": 0.1362,
      "num_input_tokens_seen": 32555856,
      "step": 37600
    },
    {
      "epoch": 17.727487034417727,
      "eval_loss": 0.20033137500286102,
      "eval_runtime": 22.1656,
      "eval_samples_per_second": 42.543,
      "eval_steps_per_second": 21.294,
      "num_input_tokens_seen": 32555856,
      "step": 37600
    },
    {
      "epoch": 17.72984441301273,
      "grad_norm": 0.3791455030441284,
      "learning_rate": 4.41348254251997e-07,
      "loss": 0.1627,
      "num_input_tokens_seen": 32560112,
      "step": 37605
    },
    {
      "epoch": 17.732201791607732,
      "grad_norm": 1.245448350906372,
      "learning_rate": 4.395135658829652e-07,
      "loss": 0.1705,
      "num_input_tokens_seen": 32563312,
      "step": 37610
    },
    {
      "epoch": 17.734559170202736,
      "grad_norm": 1.5600135326385498,
      "learning_rate": 4.376826650495852e-07,
      "loss": 0.1774,
      "num_input_tokens_seen": 32567104,
      "step": 37615
    },
    {
      "epoch": 17.736916548797737,
      "grad_norm": 1.3443609476089478,
      "learning_rate": 4.358555520342117e-07,
      "loss": 0.155,
      "num_input_tokens_seen": 32570896,
      "step": 37620
    },
    {
      "epoch": 17.739273927392738,
      "grad_norm": 1.361474633216858,
      "learning_rate": 4.3403222711860257e-07,
      "loss": 0.2106,
      "num_input_tokens_seen": 32574768,
      "step": 37625
    },
    {
      "epoch": 17.74163130598774,
      "grad_norm": 0.8758679628372192,
      "learning_rate": 4.3221269058394133e-07,
      "loss": 0.155,
      "num_input_tokens_seen": 32579024,
      "step": 37630
    },
    {
      "epoch": 17.743988684582742,
      "grad_norm": 0.4971490800380707,
      "learning_rate": 4.303969427108173e-07,
      "loss": 0.1533,
      "num_input_tokens_seen": 32584624,
      "step": 37635
    },
    {
      "epoch": 17.746346063177747,
      "grad_norm": 0.7412731647491455,
      "learning_rate": 4.2858498377924825e-07,
      "loss": 0.1868,
      "num_input_tokens_seen": 32589120,
      "step": 37640
    },
    {
      "epoch": 17.748703441772747,
      "grad_norm": 1.8599870204925537,
      "learning_rate": 4.267768140686579e-07,
      "loss": 0.1691,
      "num_input_tokens_seen": 32593872,
      "step": 37645
    },
    {
      "epoch": 17.75106082036775,
      "grad_norm": 0.6625116467475891,
      "learning_rate": 4.2497243385788975e-07,
      "loss": 0.177,
      "num_input_tokens_seen": 32598016,
      "step": 37650
    },
    {
      "epoch": 17.753418198962752,
      "grad_norm": 0.6823185682296753,
      "learning_rate": 4.231718434251991e-07,
      "loss": 0.1211,
      "num_input_tokens_seen": 32602240,
      "step": 37655
    },
    {
      "epoch": 17.755775577557756,
      "grad_norm": 2.9850943088531494,
      "learning_rate": 4.213750430482666e-07,
      "loss": 0.1489,
      "num_input_tokens_seen": 32607216,
      "step": 37660
    },
    {
      "epoch": 17.758132956152757,
      "grad_norm": 0.622117280960083,
      "learning_rate": 4.1958203300417054e-07,
      "loss": 0.1111,
      "num_input_tokens_seen": 32611392,
      "step": 37665
    },
    {
      "epoch": 17.76049033474776,
      "grad_norm": 0.5752894878387451,
      "learning_rate": 4.177928135694259e-07,
      "loss": 0.2164,
      "num_input_tokens_seen": 32614752,
      "step": 37670
    },
    {
      "epoch": 17.76284771334276,
      "grad_norm": 0.6981723308563232,
      "learning_rate": 4.1600738501994807e-07,
      "loss": 0.1283,
      "num_input_tokens_seen": 32618864,
      "step": 37675
    },
    {
      "epoch": 17.765205091937766,
      "grad_norm": 1.1082079410552979,
      "learning_rate": 4.1422574763107237e-07,
      "loss": 0.172,
      "num_input_tokens_seen": 32622656,
      "step": 37680
    },
    {
      "epoch": 17.767562470532766,
      "grad_norm": 0.6062576174736023,
      "learning_rate": 4.124479016775512e-07,
      "loss": 0.1972,
      "num_input_tokens_seen": 32626544,
      "step": 37685
    },
    {
      "epoch": 17.76991984912777,
      "grad_norm": 1.3426668643951416,
      "learning_rate": 4.106738474335514e-07,
      "loss": 0.1665,
      "num_input_tokens_seen": 32632688,
      "step": 37690
    },
    {
      "epoch": 17.77227722772277,
      "grad_norm": 0.7063223719596863,
      "learning_rate": 4.089035851726486e-07,
      "loss": 0.1517,
      "num_input_tokens_seen": 32637120,
      "step": 37695
    },
    {
      "epoch": 17.774634606317775,
      "grad_norm": 0.5607340931892395,
      "learning_rate": 4.0713711516784937e-07,
      "loss": 0.1406,
      "num_input_tokens_seen": 32641328,
      "step": 37700
    },
    {
      "epoch": 17.776991984912776,
      "grad_norm": 1.4326612949371338,
      "learning_rate": 4.05374437691558e-07,
      "loss": 0.1482,
      "num_input_tokens_seen": 32645216,
      "step": 37705
    },
    {
      "epoch": 17.77934936350778,
      "grad_norm": 1.5905051231384277,
      "learning_rate": 4.036155530156044e-07,
      "loss": 0.203,
      "num_input_tokens_seen": 32649456,
      "step": 37710
    },
    {
      "epoch": 17.78170674210278,
      "grad_norm": 1.5230185985565186,
      "learning_rate": 4.018604614112298e-07,
      "loss": 0.192,
      "num_input_tokens_seen": 32654400,
      "step": 37715
    },
    {
      "epoch": 17.784064120697785,
      "grad_norm": 1.3499095439910889,
      "learning_rate": 4.0010916314908996e-07,
      "loss": 0.2362,
      "num_input_tokens_seen": 32659600,
      "step": 37720
    },
    {
      "epoch": 17.786421499292786,
      "grad_norm": 0.7887548804283142,
      "learning_rate": 3.983616584992578e-07,
      "loss": 0.1984,
      "num_input_tokens_seen": 32663968,
      "step": 37725
    },
    {
      "epoch": 17.78877887788779,
      "grad_norm": 1.9791374206542969,
      "learning_rate": 3.9661794773122595e-07,
      "loss": 0.16,
      "num_input_tokens_seen": 32668160,
      "step": 37730
    },
    {
      "epoch": 17.79113625648279,
      "grad_norm": 0.5157480835914612,
      "learning_rate": 3.9487803111388777e-07,
      "loss": 0.1537,
      "num_input_tokens_seen": 32672480,
      "step": 37735
    },
    {
      "epoch": 17.793493635077795,
      "grad_norm": 2.058518886566162,
      "learning_rate": 3.9314190891556747e-07,
      "loss": 0.2112,
      "num_input_tokens_seen": 32676688,
      "step": 37740
    },
    {
      "epoch": 17.795851013672795,
      "grad_norm": 1.2432479858398438,
      "learning_rate": 3.914095814039925e-07,
      "loss": 0.1781,
      "num_input_tokens_seen": 32682080,
      "step": 37745
    },
    {
      "epoch": 17.7982083922678,
      "grad_norm": 1.089514136314392,
      "learning_rate": 3.896810488463104e-07,
      "loss": 0.2359,
      "num_input_tokens_seen": 32686432,
      "step": 37750
    },
    {
      "epoch": 17.8005657708628,
      "grad_norm": 0.7527108192443848,
      "learning_rate": 3.8795631150908565e-07,
      "loss": 0.2206,
      "num_input_tokens_seen": 32690384,
      "step": 37755
    },
    {
      "epoch": 17.802923149457804,
      "grad_norm": 1.3131825923919678,
      "learning_rate": 3.862353696582888e-07,
      "loss": 0.2372,
      "num_input_tokens_seen": 32694912,
      "step": 37760
    },
    {
      "epoch": 17.805280528052805,
      "grad_norm": 0.774832546710968,
      "learning_rate": 3.8451822355931313e-07,
      "loss": 0.1571,
      "num_input_tokens_seen": 32699856,
      "step": 37765
    },
    {
      "epoch": 17.80763790664781,
      "grad_norm": 0.47020700573921204,
      "learning_rate": 3.82804873476969e-07,
      "loss": 0.1809,
      "num_input_tokens_seen": 32704544,
      "step": 37770
    },
    {
      "epoch": 17.80999528524281,
      "grad_norm": 0.48754987120628357,
      "learning_rate": 3.810953196754702e-07,
      "loss": 0.116,
      "num_input_tokens_seen": 32708256,
      "step": 37775
    },
    {
      "epoch": 17.812352663837814,
      "grad_norm": 0.6292479634284973,
      "learning_rate": 3.793895624184529e-07,
      "loss": 0.1743,
      "num_input_tokens_seen": 32712368,
      "step": 37780
    },
    {
      "epoch": 17.814710042432814,
      "grad_norm": 1.0041922330856323,
      "learning_rate": 3.776876019689679e-07,
      "loss": 0.1205,
      "num_input_tokens_seen": 32717088,
      "step": 37785
    },
    {
      "epoch": 17.81706742102782,
      "grad_norm": 1.3645180463790894,
      "learning_rate": 3.7598943858947743e-07,
      "loss": 0.1511,
      "num_input_tokens_seen": 32720816,
      "step": 37790
    },
    {
      "epoch": 17.81942479962282,
      "grad_norm": 2.1736347675323486,
      "learning_rate": 3.742950725418637e-07,
      "loss": 0.2275,
      "num_input_tokens_seen": 32725120,
      "step": 37795
    },
    {
      "epoch": 17.821782178217823,
      "grad_norm": 1.057023048400879,
      "learning_rate": 3.726045040874093e-07,
      "loss": 0.1349,
      "num_input_tokens_seen": 32729024,
      "step": 37800
    },
    {
      "epoch": 17.821782178217823,
      "eval_loss": 0.20056387782096863,
      "eval_runtime": 22.2018,
      "eval_samples_per_second": 42.474,
      "eval_steps_per_second": 21.26,
      "num_input_tokens_seen": 32729024,
      "step": 37800
    },
    {
      "epoch": 17.824139556812824,
      "grad_norm": 1.1910951137542725,
      "learning_rate": 3.709177334868308e-07,
      "loss": 0.2049,
      "num_input_tokens_seen": 32733504,
      "step": 37805
    },
    {
      "epoch": 17.826496935407828,
      "grad_norm": 1.2656841278076172,
      "learning_rate": 3.692347610002478e-07,
      "loss": 0.1828,
      "num_input_tokens_seen": 32738192,
      "step": 37810
    },
    {
      "epoch": 17.82885431400283,
      "grad_norm": 1.7307379245758057,
      "learning_rate": 3.675555868871916e-07,
      "loss": 0.1609,
      "num_input_tokens_seen": 32742304,
      "step": 37815
    },
    {
      "epoch": 17.831211692597833,
      "grad_norm": 0.3935430943965912,
      "learning_rate": 3.658802114066162e-07,
      "loss": 0.2514,
      "num_input_tokens_seen": 32747328,
      "step": 37820
    },
    {
      "epoch": 17.833569071192834,
      "grad_norm": 0.8776196241378784,
      "learning_rate": 3.6420863481688437e-07,
      "loss": 0.113,
      "num_input_tokens_seen": 32751552,
      "step": 37825
    },
    {
      "epoch": 17.835926449787834,
      "grad_norm": 0.4880962371826172,
      "learning_rate": 3.625408573757705e-07,
      "loss": 0.2197,
      "num_input_tokens_seen": 32756144,
      "step": 37830
    },
    {
      "epoch": 17.83828382838284,
      "grad_norm": 2.0190670490264893,
      "learning_rate": 3.608768793404743e-07,
      "loss": 0.1968,
      "num_input_tokens_seen": 32760208,
      "step": 37835
    },
    {
      "epoch": 17.84064120697784,
      "grad_norm": 0.8766478300094604,
      "learning_rate": 3.592167009675934e-07,
      "loss": 0.1762,
      "num_input_tokens_seen": 32764960,
      "step": 37840
    },
    {
      "epoch": 17.842998585572843,
      "grad_norm": 1.342046856880188,
      "learning_rate": 3.575603225131563e-07,
      "loss": 0.1272,
      "num_input_tokens_seen": 32769760,
      "step": 37845
    },
    {
      "epoch": 17.845355964167844,
      "grad_norm": 0.7130065560340881,
      "learning_rate": 3.55907744232592e-07,
      "loss": 0.2075,
      "num_input_tokens_seen": 32774432,
      "step": 37850
    },
    {
      "epoch": 17.847713342762848,
      "grad_norm": 2.04356050491333,
      "learning_rate": 3.5425896638075217e-07,
      "loss": 0.1622,
      "num_input_tokens_seen": 32779056,
      "step": 37855
    },
    {
      "epoch": 17.85007072135785,
      "grad_norm": 0.8923296332359314,
      "learning_rate": 3.5261398921189736e-07,
      "loss": 0.1729,
      "num_input_tokens_seen": 32784080,
      "step": 37860
    },
    {
      "epoch": 17.852428099952853,
      "grad_norm": 0.9359341859817505,
      "learning_rate": 3.509728129797024e-07,
      "loss": 0.2234,
      "num_input_tokens_seen": 32788208,
      "step": 37865
    },
    {
      "epoch": 17.854785478547853,
      "grad_norm": 0.42847633361816406,
      "learning_rate": 3.4933543793725656e-07,
      "loss": 0.1707,
      "num_input_tokens_seen": 32792576,
      "step": 37870
    },
    {
      "epoch": 17.857142857142858,
      "grad_norm": 0.6580623984336853,
      "learning_rate": 3.4770186433707163e-07,
      "loss": 0.1647,
      "num_input_tokens_seen": 32796928,
      "step": 37875
    },
    {
      "epoch": 17.85950023573786,
      "grad_norm": 0.9003836512565613,
      "learning_rate": 3.4607209243105453e-07,
      "loss": 0.1036,
      "num_input_tokens_seen": 32801008,
      "step": 37880
    },
    {
      "epoch": 17.861857614332862,
      "grad_norm": 2.1690561771392822,
      "learning_rate": 3.444461224705431e-07,
      "loss": 0.1921,
      "num_input_tokens_seen": 32805520,
      "step": 37885
    },
    {
      "epoch": 17.864214992927863,
      "grad_norm": 1.2239511013031006,
      "learning_rate": 3.4282395470628116e-07,
      "loss": 0.1168,
      "num_input_tokens_seen": 32809808,
      "step": 37890
    },
    {
      "epoch": 17.866572371522867,
      "grad_norm": 0.5598416328430176,
      "learning_rate": 3.4120558938842417e-07,
      "loss": 0.1345,
      "num_input_tokens_seen": 32814624,
      "step": 37895
    },
    {
      "epoch": 17.868929750117868,
      "grad_norm": 0.7748780846595764,
      "learning_rate": 3.395910267665503e-07,
      "loss": 0.2105,
      "num_input_tokens_seen": 32818208,
      "step": 37900
    },
    {
      "epoch": 17.871287128712872,
      "grad_norm": 0.5203840732574463,
      "learning_rate": 3.3798026708964094e-07,
      "loss": 0.1987,
      "num_input_tokens_seen": 32821760,
      "step": 37905
    },
    {
      "epoch": 17.873644507307873,
      "grad_norm": 0.331057071685791,
      "learning_rate": 3.3637331060609456e-07,
      "loss": 0.1274,
      "num_input_tokens_seen": 32825344,
      "step": 37910
    },
    {
      "epoch": 17.876001885902877,
      "grad_norm": 0.8995531797409058,
      "learning_rate": 3.3477015756372966e-07,
      "loss": 0.175,
      "num_input_tokens_seen": 32829920,
      "step": 37915
    },
    {
      "epoch": 17.878359264497877,
      "grad_norm": 1.141715168952942,
      "learning_rate": 3.3317080820976785e-07,
      "loss": 0.1747,
      "num_input_tokens_seen": 32834112,
      "step": 37920
    },
    {
      "epoch": 17.88071664309288,
      "grad_norm": 1.5171462297439575,
      "learning_rate": 3.315752627908508e-07,
      "loss": 0.1698,
      "num_input_tokens_seen": 32838240,
      "step": 37925
    },
    {
      "epoch": 17.883074021687882,
      "grad_norm": 0.5813401937484741,
      "learning_rate": 3.299835215530317e-07,
      "loss": 0.1451,
      "num_input_tokens_seen": 32842432,
      "step": 37930
    },
    {
      "epoch": 17.885431400282886,
      "grad_norm": 0.6823669075965881,
      "learning_rate": 3.2839558474177245e-07,
      "loss": 0.1274,
      "num_input_tokens_seen": 32846864,
      "step": 37935
    },
    {
      "epoch": 17.887788778877887,
      "grad_norm": 1.0051904916763306,
      "learning_rate": 3.2681145260196056e-07,
      "loss": 0.2013,
      "num_input_tokens_seen": 32851632,
      "step": 37940
    },
    {
      "epoch": 17.89014615747289,
      "grad_norm": 1.2385060787200928,
      "learning_rate": 3.252311253778839e-07,
      "loss": 0.1334,
      "num_input_tokens_seen": 32856464,
      "step": 37945
    },
    {
      "epoch": 17.892503536067892,
      "grad_norm": 0.5398279428482056,
      "learning_rate": 3.2365460331325034e-07,
      "loss": 0.1546,
      "num_input_tokens_seen": 32860912,
      "step": 37950
    },
    {
      "epoch": 17.894860914662896,
      "grad_norm": 1.1102639436721802,
      "learning_rate": 3.2208188665117934e-07,
      "loss": 0.1406,
      "num_input_tokens_seen": 32864816,
      "step": 37955
    },
    {
      "epoch": 17.897218293257897,
      "grad_norm": 0.7853512763977051,
      "learning_rate": 3.205129756342018e-07,
      "loss": 0.1364,
      "num_input_tokens_seen": 32869056,
      "step": 37960
    },
    {
      "epoch": 17.8995756718529,
      "grad_norm": 0.8385063409805298,
      "learning_rate": 3.189478705042659e-07,
      "loss": 0.1938,
      "num_input_tokens_seen": 32872816,
      "step": 37965
    },
    {
      "epoch": 17.9019330504479,
      "grad_norm": 1.4346580505371094,
      "learning_rate": 3.173865715027341e-07,
      "loss": 0.1611,
      "num_input_tokens_seen": 32876128,
      "step": 37970
    },
    {
      "epoch": 17.904290429042906,
      "grad_norm": 0.40758049488067627,
      "learning_rate": 3.158290788703694e-07,
      "loss": 0.0998,
      "num_input_tokens_seen": 32880960,
      "step": 37975
    },
    {
      "epoch": 17.906647807637906,
      "grad_norm": 1.05486261844635,
      "learning_rate": 3.1427539284736297e-07,
      "loss": 0.1354,
      "num_input_tokens_seen": 32885552,
      "step": 37980
    },
    {
      "epoch": 17.90900518623291,
      "grad_norm": 0.9188858270645142,
      "learning_rate": 3.127255136733093e-07,
      "loss": 0.1549,
      "num_input_tokens_seen": 32889664,
      "step": 37985
    },
    {
      "epoch": 17.91136256482791,
      "grad_norm": 0.9385943412780762,
      "learning_rate": 3.1117944158722544e-07,
      "loss": 0.2294,
      "num_input_tokens_seen": 32894832,
      "step": 37990
    },
    {
      "epoch": 17.913719943422915,
      "grad_norm": 1.3139045238494873,
      "learning_rate": 3.0963717682752635e-07,
      "loss": 0.1856,
      "num_input_tokens_seen": 32898880,
      "step": 37995
    },
    {
      "epoch": 17.916077322017916,
      "grad_norm": 1.2011432647705078,
      "learning_rate": 3.080987196320578e-07,
      "loss": 0.1166,
      "num_input_tokens_seen": 32902832,
      "step": 38000
    },
    {
      "epoch": 17.916077322017916,
      "eval_loss": 0.20016683638095856,
      "eval_runtime": 22.1566,
      "eval_samples_per_second": 42.561,
      "eval_steps_per_second": 21.303,
      "num_input_tokens_seen": 32902832,
      "step": 38000
    },
    {
      "epoch": 17.91843470061292,
      "grad_norm": 0.8809848427772522,
      "learning_rate": 3.065640702380607e-07,
      "loss": 0.17,
      "num_input_tokens_seen": 32906896,
      "step": 38005
    },
    {
      "epoch": 17.92079207920792,
      "grad_norm": 2.041506767272949,
      "learning_rate": 3.050332288822011e-07,
      "loss": 0.1814,
      "num_input_tokens_seen": 32911168,
      "step": 38010
    },
    {
      "epoch": 17.92314945780292,
      "grad_norm": 0.4955044686794281,
      "learning_rate": 3.035061958005542e-07,
      "loss": 0.1157,
      "num_input_tokens_seen": 32915664,
      "step": 38015
    },
    {
      "epoch": 17.925506836397926,
      "grad_norm": 0.9049466848373413,
      "learning_rate": 3.019829712286093e-07,
      "loss": 0.1127,
      "num_input_tokens_seen": 32919824,
      "step": 38020
    },
    {
      "epoch": 17.927864214992926,
      "grad_norm": 0.7724030017852783,
      "learning_rate": 3.004635554012647e-07,
      "loss": 0.178,
      "num_input_tokens_seen": 32925040,
      "step": 38025
    },
    {
      "epoch": 17.93022159358793,
      "grad_norm": 1.218166470527649,
      "learning_rate": 2.9894794855283017e-07,
      "loss": 0.1037,
      "num_input_tokens_seen": 32929040,
      "step": 38030
    },
    {
      "epoch": 17.93257897218293,
      "grad_norm": 0.5157722234725952,
      "learning_rate": 2.9743615091703816e-07,
      "loss": 0.1507,
      "num_input_tokens_seen": 32933312,
      "step": 38035
    },
    {
      "epoch": 17.934936350777935,
      "grad_norm": 1.9531913995742798,
      "learning_rate": 2.959281627270216e-07,
      "loss": 0.1501,
      "num_input_tokens_seen": 32937648,
      "step": 38040
    },
    {
      "epoch": 17.937293729372936,
      "grad_norm": 0.5134620070457458,
      "learning_rate": 2.944239842153362e-07,
      "loss": 0.1592,
      "num_input_tokens_seen": 32942048,
      "step": 38045
    },
    {
      "epoch": 17.93965110796794,
      "grad_norm": 1.089339256286621,
      "learning_rate": 2.929236156139381e-07,
      "loss": 0.1963,
      "num_input_tokens_seen": 32946272,
      "step": 38050
    },
    {
      "epoch": 17.94200848656294,
      "grad_norm": 1.973025918006897,
      "learning_rate": 2.9142705715420883e-07,
      "loss": 0.2001,
      "num_input_tokens_seen": 32950704,
      "step": 38055
    },
    {
      "epoch": 17.944365865157945,
      "grad_norm": 0.7045966982841492,
      "learning_rate": 2.8993430906693595e-07,
      "loss": 0.1969,
      "num_input_tokens_seen": 32955328,
      "step": 38060
    },
    {
      "epoch": 17.946723243752945,
      "grad_norm": 1.274095058441162,
      "learning_rate": 2.88445371582316e-07,
      "loss": 0.1502,
      "num_input_tokens_seen": 32958928,
      "step": 38065
    },
    {
      "epoch": 17.94908062234795,
      "grad_norm": 0.6955338716506958,
      "learning_rate": 2.8696024492996796e-07,
      "loss": 0.1574,
      "num_input_tokens_seen": 32963472,
      "step": 38070
    },
    {
      "epoch": 17.95143800094295,
      "grad_norm": 0.5354678630828857,
      "learning_rate": 2.854789293389115e-07,
      "loss": 0.1343,
      "num_input_tokens_seen": 32968128,
      "step": 38075
    },
    {
      "epoch": 17.953795379537954,
      "grad_norm": 0.5639871954917908,
      "learning_rate": 2.8400142503758606e-07,
      "loss": 0.1948,
      "num_input_tokens_seen": 32973120,
      "step": 38080
    },
    {
      "epoch": 17.956152758132955,
      "grad_norm": 0.5956523418426514,
      "learning_rate": 2.8252773225384276e-07,
      "loss": 0.2303,
      "num_input_tokens_seen": 32978080,
      "step": 38085
    },
    {
      "epoch": 17.95851013672796,
      "grad_norm": 2.0082035064697266,
      "learning_rate": 2.8105785121494143e-07,
      "loss": 0.1563,
      "num_input_tokens_seen": 32982240,
      "step": 38090
    },
    {
      "epoch": 17.96086751532296,
      "grad_norm": 1.2093496322631836,
      "learning_rate": 2.795917821475563e-07,
      "loss": 0.1799,
      "num_input_tokens_seen": 32986400,
      "step": 38095
    },
    {
      "epoch": 17.963224893917964,
      "grad_norm": 0.4390227496623993,
      "learning_rate": 2.78129525277776e-07,
      "loss": 0.1785,
      "num_input_tokens_seen": 32990608,
      "step": 38100
    },
    {
      "epoch": 17.965582272512965,
      "grad_norm": 1.0102769136428833,
      "learning_rate": 2.766710808310952e-07,
      "loss": 0.1519,
      "num_input_tokens_seen": 32994368,
      "step": 38105
    },
    {
      "epoch": 17.96793965110797,
      "grad_norm": 1.3084036111831665,
      "learning_rate": 2.7521644903242827e-07,
      "loss": 0.2014,
      "num_input_tokens_seen": 32998592,
      "step": 38110
    },
    {
      "epoch": 17.97029702970297,
      "grad_norm": 1.530592441558838,
      "learning_rate": 2.7376563010609593e-07,
      "loss": 0.1963,
      "num_input_tokens_seen": 33003184,
      "step": 38115
    },
    {
      "epoch": 17.972654408297974,
      "grad_norm": 0.8759222626686096,
      "learning_rate": 2.72318624275833e-07,
      "loss": 0.2193,
      "num_input_tokens_seen": 33006816,
      "step": 38120
    },
    {
      "epoch": 17.975011786892974,
      "grad_norm": 0.7394915223121643,
      "learning_rate": 2.7087543176478324e-07,
      "loss": 0.2178,
      "num_input_tokens_seen": 33011440,
      "step": 38125
    },
    {
      "epoch": 17.97736916548798,
      "grad_norm": 0.7477229833602905,
      "learning_rate": 2.694360527955103e-07,
      "loss": 0.1861,
      "num_input_tokens_seen": 33015744,
      "step": 38130
    },
    {
      "epoch": 17.97972654408298,
      "grad_norm": 0.8785243630409241,
      "learning_rate": 2.680004875899811e-07,
      "loss": 0.1903,
      "num_input_tokens_seen": 33019536,
      "step": 38135
    },
    {
      "epoch": 17.982083922677983,
      "grad_norm": 1.4771125316619873,
      "learning_rate": 2.665687363695768e-07,
      "loss": 0.1277,
      "num_input_tokens_seen": 33024384,
      "step": 38140
    },
    {
      "epoch": 17.984441301272984,
      "grad_norm": 0.8717460632324219,
      "learning_rate": 2.6514079935509584e-07,
      "loss": 0.2065,
      "num_input_tokens_seen": 33028816,
      "step": 38145
    },
    {
      "epoch": 17.986798679867988,
      "grad_norm": 0.922614574432373,
      "learning_rate": 2.6371667676673983e-07,
      "loss": 0.2057,
      "num_input_tokens_seen": 33033296,
      "step": 38150
    },
    {
      "epoch": 17.98915605846299,
      "grad_norm": 1.1625365018844604,
      "learning_rate": 2.6229636882412755e-07,
      "loss": 0.158,
      "num_input_tokens_seen": 33037552,
      "step": 38155
    },
    {
      "epoch": 17.991513437057993,
      "grad_norm": 0.7179512977600098,
      "learning_rate": 2.6087987574628935e-07,
      "loss": 0.2391,
      "num_input_tokens_seen": 33041776,
      "step": 38160
    },
    {
      "epoch": 17.993870815652993,
      "grad_norm": 1.0584789514541626,
      "learning_rate": 2.5946719775166437e-07,
      "loss": 0.1741,
      "num_input_tokens_seen": 33046288,
      "step": 38165
    },
    {
      "epoch": 17.996228194247998,
      "grad_norm": 0.6735996603965759,
      "learning_rate": 2.5805833505810616e-07,
      "loss": 0.1439,
      "num_input_tokens_seen": 33050224,
      "step": 38170
    },
    {
      "epoch": 17.998585572842998,
      "grad_norm": 0.9908170700073242,
      "learning_rate": 2.566532878828798e-07,
      "loss": 0.1738,
      "num_input_tokens_seen": 33054960,
      "step": 38175
    },
    {
      "epoch": 18.000942951438002,
      "grad_norm": 0.5039419531822205,
      "learning_rate": 2.552520564426619e-07,
      "loss": 0.1687,
      "num_input_tokens_seen": 33059456,
      "step": 38180
    },
    {
      "epoch": 18.003300330033003,
      "grad_norm": 0.5699019432067871,
      "learning_rate": 2.5385464095353803e-07,
      "loss": 0.1473,
      "num_input_tokens_seen": 33065056,
      "step": 38185
    },
    {
      "epoch": 18.005657708628007,
      "grad_norm": 1.261415719985962,
      "learning_rate": 2.5246104163100804e-07,
      "loss": 0.1947,
      "num_input_tokens_seen": 33068992,
      "step": 38190
    },
    {
      "epoch": 18.008015087223008,
      "grad_norm": 1.8079063892364502,
      "learning_rate": 2.510712586899833e-07,
      "loss": 0.1221,
      "num_input_tokens_seen": 33072944,
      "step": 38195
    },
    {
      "epoch": 18.010372465818012,
      "grad_norm": 0.8346419930458069,
      "learning_rate": 2.4968529234478124e-07,
      "loss": 0.1965,
      "num_input_tokens_seen": 33076912,
      "step": 38200
    },
    {
      "epoch": 18.010372465818012,
      "eval_loss": 0.19926267862319946,
      "eval_runtime": 22.43,
      "eval_samples_per_second": 42.042,
      "eval_steps_per_second": 21.043,
      "num_input_tokens_seen": 33076912,
      "step": 38200
    },
    {
      "epoch": 18.012729844413013,
      "grad_norm": 0.9768953323364258,
      "learning_rate": 2.483031428091448e-07,
      "loss": 0.1732,
      "num_input_tokens_seen": 33081536,
      "step": 38205
    },
    {
      "epoch": 18.015087223008017,
      "grad_norm": 1.0002237558364868,
      "learning_rate": 2.469248102962091e-07,
      "loss": 0.2091,
      "num_input_tokens_seen": 33085600,
      "step": 38210
    },
    {
      "epoch": 18.017444601603017,
      "grad_norm": 1.1528762578964233,
      "learning_rate": 2.4555029501853455e-07,
      "loss": 0.1705,
      "num_input_tokens_seen": 33089824,
      "step": 38215
    },
    {
      "epoch": 18.019801980198018,
      "grad_norm": 1.137803077697754,
      "learning_rate": 2.441795971880906e-07,
      "loss": 0.2611,
      "num_input_tokens_seen": 33093936,
      "step": 38220
    },
    {
      "epoch": 18.022159358793022,
      "grad_norm": 0.790336549282074,
      "learning_rate": 2.4281271701625255e-07,
      "loss": 0.1892,
      "num_input_tokens_seen": 33098208,
      "step": 38225
    },
    {
      "epoch": 18.024516737388023,
      "grad_norm": 0.9966343641281128,
      "learning_rate": 2.4144965471381007e-07,
      "loss": 0.2131,
      "num_input_tokens_seen": 33102528,
      "step": 38230
    },
    {
      "epoch": 18.026874115983027,
      "grad_norm": 1.1033174991607666,
      "learning_rate": 2.400904104909674e-07,
      "loss": 0.2637,
      "num_input_tokens_seen": 33106416,
      "step": 38235
    },
    {
      "epoch": 18.029231494578028,
      "grad_norm": 0.618076741695404,
      "learning_rate": 2.3873498455733725e-07,
      "loss": 0.1425,
      "num_input_tokens_seen": 33110800,
      "step": 38240
    },
    {
      "epoch": 18.031588873173032,
      "grad_norm": 1.0403759479522705,
      "learning_rate": 2.3738337712194137e-07,
      "loss": 0.2285,
      "num_input_tokens_seen": 33115136,
      "step": 38245
    },
    {
      "epoch": 18.033946251768032,
      "grad_norm": 1.9813883304595947,
      "learning_rate": 2.3603558839321305e-07,
      "loss": 0.1903,
      "num_input_tokens_seen": 33119488,
      "step": 38250
    },
    {
      "epoch": 18.036303630363037,
      "grad_norm": 0.5325500965118408,
      "learning_rate": 2.3469161857900267e-07,
      "loss": 0.1248,
      "num_input_tokens_seen": 33123568,
      "step": 38255
    },
    {
      "epoch": 18.038661008958037,
      "grad_norm": 1.9429079294204712,
      "learning_rate": 2.3335146788656393e-07,
      "loss": 0.1683,
      "num_input_tokens_seen": 33127952,
      "step": 38260
    },
    {
      "epoch": 18.04101838755304,
      "grad_norm": 1.6562687158584595,
      "learning_rate": 2.3201513652256757e-07,
      "loss": 0.1752,
      "num_input_tokens_seen": 33132976,
      "step": 38265
    },
    {
      "epoch": 18.043375766148042,
      "grad_norm": 0.7119272947311401,
      "learning_rate": 2.3068262469308766e-07,
      "loss": 0.1493,
      "num_input_tokens_seen": 33137184,
      "step": 38270
    },
    {
      "epoch": 18.045733144743046,
      "grad_norm": 0.6509285569190979,
      "learning_rate": 2.2935393260362093e-07,
      "loss": 0.1147,
      "num_input_tokens_seen": 33141200,
      "step": 38275
    },
    {
      "epoch": 18.048090523338047,
      "grad_norm": 0.6831954121589661,
      "learning_rate": 2.2802906045906458e-07,
      "loss": 0.1651,
      "num_input_tokens_seen": 33146288,
      "step": 38280
    },
    {
      "epoch": 18.05044790193305,
      "grad_norm": 1.5245437622070312,
      "learning_rate": 2.2670800846373018e-07,
      "loss": 0.1847,
      "num_input_tokens_seen": 33150992,
      "step": 38285
    },
    {
      "epoch": 18.05280528052805,
      "grad_norm": 3.1700782775878906,
      "learning_rate": 2.2539077682134367e-07,
      "loss": 0.2421,
      "num_input_tokens_seen": 33154768,
      "step": 38290
    },
    {
      "epoch": 18.055162659123056,
      "grad_norm": 0.8700650334358215,
      "learning_rate": 2.2407736573503423e-07,
      "loss": 0.2524,
      "num_input_tokens_seen": 33158480,
      "step": 38295
    },
    {
      "epoch": 18.057520037718056,
      "grad_norm": 0.8134238123893738,
      "learning_rate": 2.2276777540735093e-07,
      "loss": 0.1199,
      "num_input_tokens_seen": 33162528,
      "step": 38300
    },
    {
      "epoch": 18.05987741631306,
      "grad_norm": 0.5252361297607422,
      "learning_rate": 2.2146200604024613e-07,
      "loss": 0.1776,
      "num_input_tokens_seen": 33166736,
      "step": 38305
    },
    {
      "epoch": 18.06223479490806,
      "grad_norm": 0.9019278883934021,
      "learning_rate": 2.2016005783508375e-07,
      "loss": 0.1502,
      "num_input_tokens_seen": 33170624,
      "step": 38310
    },
    {
      "epoch": 18.064592173503065,
      "grad_norm": 1.2698742151260376,
      "learning_rate": 2.1886193099264763e-07,
      "loss": 0.1584,
      "num_input_tokens_seen": 33176048,
      "step": 38315
    },
    {
      "epoch": 18.066949552098066,
      "grad_norm": 1.196964144706726,
      "learning_rate": 2.175676257131165e-07,
      "loss": 0.1168,
      "num_input_tokens_seen": 33180768,
      "step": 38320
    },
    {
      "epoch": 18.06930693069307,
      "grad_norm": 0.8689922094345093,
      "learning_rate": 2.162771421960974e-07,
      "loss": 0.154,
      "num_input_tokens_seen": 33185248,
      "step": 38325
    },
    {
      "epoch": 18.07166430928807,
      "grad_norm": 0.5772242546081543,
      "learning_rate": 2.1499048064059224e-07,
      "loss": 0.1376,
      "num_input_tokens_seen": 33189696,
      "step": 38330
    },
    {
      "epoch": 18.074021687883075,
      "grad_norm": 1.2705223560333252,
      "learning_rate": 2.1370764124502285e-07,
      "loss": 0.1895,
      "num_input_tokens_seen": 33193312,
      "step": 38335
    },
    {
      "epoch": 18.076379066478076,
      "grad_norm": 0.9941864609718323,
      "learning_rate": 2.1242862420721988e-07,
      "loss": 0.1601,
      "num_input_tokens_seen": 33198416,
      "step": 38340
    },
    {
      "epoch": 18.07873644507308,
      "grad_norm": 0.8928539752960205,
      "learning_rate": 2.1115342972442276e-07,
      "loss": 0.1669,
      "num_input_tokens_seen": 33202912,
      "step": 38345
    },
    {
      "epoch": 18.08109382366808,
      "grad_norm": 0.5581601858139038,
      "learning_rate": 2.0988205799328252e-07,
      "loss": 0.1629,
      "num_input_tokens_seen": 33206768,
      "step": 38350
    },
    {
      "epoch": 18.083451202263085,
      "grad_norm": 0.877923846244812,
      "learning_rate": 2.0861450920986182e-07,
      "loss": 0.2155,
      "num_input_tokens_seen": 33210896,
      "step": 38355
    },
    {
      "epoch": 18.085808580858085,
      "grad_norm": 0.6222615242004395,
      "learning_rate": 2.07350783569632e-07,
      "loss": 0.1509,
      "num_input_tokens_seen": 33215168,
      "step": 38360
    },
    {
      "epoch": 18.08816595945309,
      "grad_norm": 1.9687936305999756,
      "learning_rate": 2.060908812674761e-07,
      "loss": 0.1152,
      "num_input_tokens_seen": 33218752,
      "step": 38365
    },
    {
      "epoch": 18.09052333804809,
      "grad_norm": 0.5028272867202759,
      "learning_rate": 2.0483480249768317e-07,
      "loss": 0.1314,
      "num_input_tokens_seen": 33223248,
      "step": 38370
    },
    {
      "epoch": 18.092880716643094,
      "grad_norm": 1.666025161743164,
      "learning_rate": 2.035825474539621e-07,
      "loss": 0.2376,
      "num_input_tokens_seen": 33227696,
      "step": 38375
    },
    {
      "epoch": 18.095238095238095,
      "grad_norm": 0.737824022769928,
      "learning_rate": 2.0233411632942235e-07,
      "loss": 0.1117,
      "num_input_tokens_seen": 33231392,
      "step": 38380
    },
    {
      "epoch": 18.0975954738331,
      "grad_norm": 1.071795105934143,
      "learning_rate": 2.0108950931658764e-07,
      "loss": 0.2743,
      "num_input_tokens_seen": 33235472,
      "step": 38385
    },
    {
      "epoch": 18.0999528524281,
      "grad_norm": 0.5853174328804016,
      "learning_rate": 1.998487266073934e-07,
      "loss": 0.1711,
      "num_input_tokens_seen": 33239872,
      "step": 38390
    },
    {
      "epoch": 18.102310231023104,
      "grad_norm": 1.9023174047470093,
      "learning_rate": 1.986117683931865e-07,
      "loss": 0.113,
      "num_input_tokens_seen": 33244768,
      "step": 38395
    },
    {
      "epoch": 18.104667609618105,
      "grad_norm": 1.2254739999771118,
      "learning_rate": 1.9737863486471442e-07,
      "loss": 0.1437,
      "num_input_tokens_seen": 33248832,
      "step": 38400
    },
    {
      "epoch": 18.104667609618105,
      "eval_loss": 0.2001124620437622,
      "eval_runtime": 22.1737,
      "eval_samples_per_second": 42.528,
      "eval_steps_per_second": 21.286,
      "num_input_tokens_seen": 33248832,
      "step": 38400
    },
    {
      "epoch": 18.10702498821311,
      "grad_norm": 2.230266809463501,
      "learning_rate": 1.9614932621215e-07,
      "loss": 0.2016,
      "num_input_tokens_seen": 33253296,
      "step": 38405
    },
    {
      "epoch": 18.10938236680811,
      "grad_norm": 0.864616334438324,
      "learning_rate": 1.9492384262506102e-07,
      "loss": 0.2068,
      "num_input_tokens_seen": 33257136,
      "step": 38410
    },
    {
      "epoch": 18.111739745403113,
      "grad_norm": 0.8647833466529846,
      "learning_rate": 1.9370218429243524e-07,
      "loss": 0.1904,
      "num_input_tokens_seen": 33261504,
      "step": 38415
    },
    {
      "epoch": 18.114097123998114,
      "grad_norm": 1.0964874029159546,
      "learning_rate": 1.9248435140267197e-07,
      "loss": 0.1901,
      "num_input_tokens_seen": 33265600,
      "step": 38420
    },
    {
      "epoch": 18.116454502593115,
      "grad_norm": 1.6570937633514404,
      "learning_rate": 1.9127034414356814e-07,
      "loss": 0.1859,
      "num_input_tokens_seen": 33269104,
      "step": 38425
    },
    {
      "epoch": 18.11881188118812,
      "grad_norm": 1.75380277633667,
      "learning_rate": 1.9006016270234627e-07,
      "loss": 0.2042,
      "num_input_tokens_seen": 33273632,
      "step": 38430
    },
    {
      "epoch": 18.12116925978312,
      "grad_norm": 0.36379021406173706,
      "learning_rate": 1.888538072656293e-07,
      "loss": 0.1346,
      "num_input_tokens_seen": 33278704,
      "step": 38435
    },
    {
      "epoch": 18.123526638378124,
      "grad_norm": 0.9279085993766785,
      "learning_rate": 1.8765127801944893e-07,
      "loss": 0.1483,
      "num_input_tokens_seen": 33282768,
      "step": 38440
    },
    {
      "epoch": 18.125884016973124,
      "grad_norm": 1.7742921113967896,
      "learning_rate": 1.8645257514925406e-07,
      "loss": 0.1547,
      "num_input_tokens_seen": 33286416,
      "step": 38445
    },
    {
      "epoch": 18.12824139556813,
      "grad_norm": 0.45469173789024353,
      "learning_rate": 1.8525769883989685e-07,
      "loss": 0.1456,
      "num_input_tokens_seen": 33290400,
      "step": 38450
    },
    {
      "epoch": 18.13059877416313,
      "grad_norm": 1.1359978914260864,
      "learning_rate": 1.8406664927564654e-07,
      "loss": 0.1847,
      "num_input_tokens_seen": 33295120,
      "step": 38455
    },
    {
      "epoch": 18.132956152758133,
      "grad_norm": 1.0271108150482178,
      "learning_rate": 1.8287942664017566e-07,
      "loss": 0.1889,
      "num_input_tokens_seen": 33299280,
      "step": 38460
    },
    {
      "epoch": 18.135313531353134,
      "grad_norm": 1.135151743888855,
      "learning_rate": 1.8169603111656552e-07,
      "loss": 0.156,
      "num_input_tokens_seen": 33303856,
      "step": 38465
    },
    {
      "epoch": 18.137670909948138,
      "grad_norm": 0.7015843391418457,
      "learning_rate": 1.805164628873146e-07,
      "loss": 0.1243,
      "num_input_tokens_seen": 33308048,
      "step": 38470
    },
    {
      "epoch": 18.14002828854314,
      "grad_norm": 0.8301371335983276,
      "learning_rate": 1.793407221343274e-07,
      "loss": 0.152,
      "num_input_tokens_seen": 33312288,
      "step": 38475
    },
    {
      "epoch": 18.142385667138143,
      "grad_norm": 0.4917954206466675,
      "learning_rate": 1.781688090389172e-07,
      "loss": 0.1498,
      "num_input_tokens_seen": 33316864,
      "step": 38480
    },
    {
      "epoch": 18.144743045733144,
      "grad_norm": 1.3673911094665527,
      "learning_rate": 1.770007237818061e-07,
      "loss": 0.2009,
      "num_input_tokens_seen": 33321184,
      "step": 38485
    },
    {
      "epoch": 18.147100424328148,
      "grad_norm": 1.417060375213623,
      "learning_rate": 1.7583646654313059e-07,
      "loss": 0.174,
      "num_input_tokens_seen": 33325168,
      "step": 38490
    },
    {
      "epoch": 18.14945780292315,
      "grad_norm": 1.7595878839492798,
      "learning_rate": 1.7467603750242757e-07,
      "loss": 0.2475,
      "num_input_tokens_seen": 33329104,
      "step": 38495
    },
    {
      "epoch": 18.151815181518153,
      "grad_norm": 0.5740852952003479,
      "learning_rate": 1.7351943683865944e-07,
      "loss": 0.1684,
      "num_input_tokens_seen": 33333424,
      "step": 38500
    },
    {
      "epoch": 18.154172560113153,
      "grad_norm": 0.41528189182281494,
      "learning_rate": 1.723666647301808e-07,
      "loss": 0.2051,
      "num_input_tokens_seen": 33337008,
      "step": 38505
    },
    {
      "epoch": 18.156529938708157,
      "grad_norm": 0.790101945400238,
      "learning_rate": 1.712177213547661e-07,
      "loss": 0.1413,
      "num_input_tokens_seen": 33341840,
      "step": 38510
    },
    {
      "epoch": 18.158887317303158,
      "grad_norm": 1.468207836151123,
      "learning_rate": 1.7007260688959581e-07,
      "loss": 0.1776,
      "num_input_tokens_seen": 33346752,
      "step": 38515
    },
    {
      "epoch": 18.161244695898162,
      "grad_norm": 1.4994553327560425,
      "learning_rate": 1.68931321511262e-07,
      "loss": 0.1901,
      "num_input_tokens_seen": 33350400,
      "step": 38520
    },
    {
      "epoch": 18.163602074493163,
      "grad_norm": 0.46377742290496826,
      "learning_rate": 1.6779386539576835e-07,
      "loss": 0.1557,
      "num_input_tokens_seen": 33354624,
      "step": 38525
    },
    {
      "epoch": 18.165959453088167,
      "grad_norm": 0.6457148194313049,
      "learning_rate": 1.666602387185162e-07,
      "loss": 0.2046,
      "num_input_tokens_seen": 33358816,
      "step": 38530
    },
    {
      "epoch": 18.168316831683168,
      "grad_norm": 2.0370547771453857,
      "learning_rate": 1.655304416543352e-07,
      "loss": 0.2312,
      "num_input_tokens_seen": 33363568,
      "step": 38535
    },
    {
      "epoch": 18.17067421027817,
      "grad_norm": 0.6627401113510132,
      "learning_rate": 1.6440447437744698e-07,
      "loss": 0.0992,
      "num_input_tokens_seen": 33367920,
      "step": 38540
    },
    {
      "epoch": 18.173031588873172,
      "grad_norm": 0.7484326958656311,
      "learning_rate": 1.6328233706149332e-07,
      "loss": 0.2179,
      "num_input_tokens_seen": 33372208,
      "step": 38545
    },
    {
      "epoch": 18.175388967468177,
      "grad_norm": 0.9563989043235779,
      "learning_rate": 1.6216402987951906e-07,
      "loss": 0.2456,
      "num_input_tokens_seen": 33377040,
      "step": 38550
    },
    {
      "epoch": 18.177746346063177,
      "grad_norm": 1.31964910030365,
      "learning_rate": 1.6104955300398627e-07,
      "loss": 0.184,
      "num_input_tokens_seen": 33380976,
      "step": 38555
    },
    {
      "epoch": 18.18010372465818,
      "grad_norm": 0.8318966627120972,
      "learning_rate": 1.5993890660675748e-07,
      "loss": 0.1653,
      "num_input_tokens_seen": 33385504,
      "step": 38560
    },
    {
      "epoch": 18.182461103253182,
      "grad_norm": 0.2983452379703522,
      "learning_rate": 1.5883209085910678e-07,
      "loss": 0.1284,
      "num_input_tokens_seen": 33389408,
      "step": 38565
    },
    {
      "epoch": 18.184818481848186,
      "grad_norm": 0.4805051386356354,
      "learning_rate": 1.5772910593172264e-07,
      "loss": 0.1398,
      "num_input_tokens_seen": 33393344,
      "step": 38570
    },
    {
      "epoch": 18.187175860443187,
      "grad_norm": 0.8673396110534668,
      "learning_rate": 1.5662995199469954e-07,
      "loss": 0.183,
      "num_input_tokens_seen": 33397856,
      "step": 38575
    },
    {
      "epoch": 18.18953323903819,
      "grad_norm": 1.4791797399520874,
      "learning_rate": 1.5553462921753802e-07,
      "loss": 0.2158,
      "num_input_tokens_seen": 33402176,
      "step": 38580
    },
    {
      "epoch": 18.19189061763319,
      "grad_norm": 0.8738508820533752,
      "learning_rate": 1.544431377691502e-07,
      "loss": 0.1898,
      "num_input_tokens_seen": 33406656,
      "step": 38585
    },
    {
      "epoch": 18.194247996228196,
      "grad_norm": 1.2393605709075928,
      "learning_rate": 1.5335547781785975e-07,
      "loss": 0.2017,
      "num_input_tokens_seen": 33411088,
      "step": 38590
    },
    {
      "epoch": 18.196605374823196,
      "grad_norm": 1.781603455543518,
      "learning_rate": 1.5227164953139917e-07,
      "loss": 0.2157,
      "num_input_tokens_seen": 33416176,
      "step": 38595
    },
    {
      "epoch": 18.1989627534182,
      "grad_norm": 0.8077720999717712,
      "learning_rate": 1.511916530769042e-07,
      "loss": 0.1477,
      "num_input_tokens_seen": 33420800,
      "step": 38600
    },
    {
      "epoch": 18.1989627534182,
      "eval_loss": 0.1995776891708374,
      "eval_runtime": 22.1843,
      "eval_samples_per_second": 42.508,
      "eval_steps_per_second": 21.276,
      "num_input_tokens_seen": 33420800,
      "step": 38600
    },
    {
      "epoch": 18.2013201320132,
      "grad_norm": 0.7317941784858704,
      "learning_rate": 1.5011548862092773e-07,
      "loss": 0.2051,
      "num_input_tokens_seen": 33424832,
      "step": 38605
    },
    {
      "epoch": 18.203677510608205,
      "grad_norm": 1.6226979494094849,
      "learning_rate": 1.490431563294231e-07,
      "loss": 0.1604,
      "num_input_tokens_seen": 33428672,
      "step": 38610
    },
    {
      "epoch": 18.206034889203206,
      "grad_norm": 1.2177025079727173,
      "learning_rate": 1.4797465636776365e-07,
      "loss": 0.2397,
      "num_input_tokens_seen": 33432528,
      "step": 38615
    },
    {
      "epoch": 18.208392267798207,
      "grad_norm": 0.7754179239273071,
      "learning_rate": 1.4690998890072027e-07,
      "loss": 0.2489,
      "num_input_tokens_seen": 33436816,
      "step": 38620
    },
    {
      "epoch": 18.21074964639321,
      "grad_norm": 1.2712966203689575,
      "learning_rate": 1.4584915409248112e-07,
      "loss": 0.1853,
      "num_input_tokens_seen": 33441536,
      "step": 38625
    },
    {
      "epoch": 18.21310702498821,
      "grad_norm": 0.4298529028892517,
      "learning_rate": 1.4479215210663754e-07,
      "loss": 0.1435,
      "num_input_tokens_seen": 33445664,
      "step": 38630
    },
    {
      "epoch": 18.215464403583216,
      "grad_norm": 1.0009944438934326,
      "learning_rate": 1.4373898310619528e-07,
      "loss": 0.1237,
      "num_input_tokens_seen": 33450240,
      "step": 38635
    },
    {
      "epoch": 18.217821782178216,
      "grad_norm": 0.9210699796676636,
      "learning_rate": 1.4268964725356604e-07,
      "loss": 0.2297,
      "num_input_tokens_seen": 33454448,
      "step": 38640
    },
    {
      "epoch": 18.22017916077322,
      "grad_norm": 0.608596920967102,
      "learning_rate": 1.4164414471056764e-07,
      "loss": 0.1746,
      "num_input_tokens_seen": 33458720,
      "step": 38645
    },
    {
      "epoch": 18.22253653936822,
      "grad_norm": 0.8315104246139526,
      "learning_rate": 1.4060247563843497e-07,
      "loss": 0.2057,
      "num_input_tokens_seen": 33462880,
      "step": 38650
    },
    {
      "epoch": 18.224893917963225,
      "grad_norm": 1.8691933155059814,
      "learning_rate": 1.3956464019780068e-07,
      "loss": 0.2282,
      "num_input_tokens_seen": 33466944,
      "step": 38655
    },
    {
      "epoch": 18.227251296558226,
      "grad_norm": 1.712908387184143,
      "learning_rate": 1.385306385487145e-07,
      "loss": 0.1458,
      "num_input_tokens_seen": 33470752,
      "step": 38660
    },
    {
      "epoch": 18.22960867515323,
      "grad_norm": 0.771830677986145,
      "learning_rate": 1.3750047085063222e-07,
      "loss": 0.1951,
      "num_input_tokens_seen": 33475504,
      "step": 38665
    },
    {
      "epoch": 18.23196605374823,
      "grad_norm": 0.8816052675247192,
      "learning_rate": 1.3647413726242119e-07,
      "loss": 0.1892,
      "num_input_tokens_seen": 33480128,
      "step": 38670
    },
    {
      "epoch": 18.234323432343235,
      "grad_norm": 1.5348974466323853,
      "learning_rate": 1.3545163794235205e-07,
      "loss": 0.1724,
      "num_input_tokens_seen": 33484144,
      "step": 38675
    },
    {
      "epoch": 18.236680810938235,
      "grad_norm": 1.1449682712554932,
      "learning_rate": 1.3443297304810698e-07,
      "loss": 0.1884,
      "num_input_tokens_seen": 33488608,
      "step": 38680
    },
    {
      "epoch": 18.23903818953324,
      "grad_norm": 0.6642941832542419,
      "learning_rate": 1.3341814273677977e-07,
      "loss": 0.118,
      "num_input_tokens_seen": 33492864,
      "step": 38685
    },
    {
      "epoch": 18.24139556812824,
      "grad_norm": 0.8813267350196838,
      "learning_rate": 1.324071471648647e-07,
      "loss": 0.1624,
      "num_input_tokens_seen": 33496512,
      "step": 38690
    },
    {
      "epoch": 18.243752946723244,
      "grad_norm": 0.884408175945282,
      "learning_rate": 1.3139998648827312e-07,
      "loss": 0.1368,
      "num_input_tokens_seen": 33502368,
      "step": 38695
    },
    {
      "epoch": 18.246110325318245,
      "grad_norm": 0.925818920135498,
      "learning_rate": 1.3039666086232526e-07,
      "loss": 0.1589,
      "num_input_tokens_seen": 33506272,
      "step": 38700
    },
    {
      "epoch": 18.24846770391325,
      "grad_norm": 0.947136640548706,
      "learning_rate": 1.2939717044174183e-07,
      "loss": 0.1656,
      "num_input_tokens_seen": 33511584,
      "step": 38705
    },
    {
      "epoch": 18.25082508250825,
      "grad_norm": 0.7077450752258301,
      "learning_rate": 1.284015153806578e-07,
      "loss": 0.1685,
      "num_input_tokens_seen": 33516128,
      "step": 38710
    },
    {
      "epoch": 18.253182461103254,
      "grad_norm": 1.252299427986145,
      "learning_rate": 1.274096958326171e-07,
      "loss": 0.2052,
      "num_input_tokens_seen": 33520928,
      "step": 38715
    },
    {
      "epoch": 18.255539839698255,
      "grad_norm": 0.5811238884925842,
      "learning_rate": 1.2642171195056952e-07,
      "loss": 0.1998,
      "num_input_tokens_seen": 33527792,
      "step": 38720
    },
    {
      "epoch": 18.25789721829326,
      "grad_norm": 1.4297064542770386,
      "learning_rate": 1.2543756388687377e-07,
      "loss": 0.1328,
      "num_input_tokens_seen": 33531728,
      "step": 38725
    },
    {
      "epoch": 18.26025459688826,
      "grad_norm": 0.8406440019607544,
      "learning_rate": 1.2445725179330014e-07,
      "loss": 0.1795,
      "num_input_tokens_seen": 33535664,
      "step": 38730
    },
    {
      "epoch": 18.262611975483264,
      "grad_norm": 1.2302502393722534,
      "learning_rate": 1.2348077582102212e-07,
      "loss": 0.1558,
      "num_input_tokens_seen": 33539664,
      "step": 38735
    },
    {
      "epoch": 18.264969354078264,
      "grad_norm": 0.7656831741333008,
      "learning_rate": 1.2250813612062762e-07,
      "loss": 0.2026,
      "num_input_tokens_seen": 33543664,
      "step": 38740
    },
    {
      "epoch": 18.26732673267327,
      "grad_norm": 0.7494287490844727,
      "learning_rate": 1.215393328421105e-07,
      "loss": 0.1709,
      "num_input_tokens_seen": 33547456,
      "step": 38745
    },
    {
      "epoch": 18.26968411126827,
      "grad_norm": 1.1055959463119507,
      "learning_rate": 1.2057436613486796e-07,
      "loss": 0.2511,
      "num_input_tokens_seen": 33551024,
      "step": 38750
    },
    {
      "epoch": 18.272041489863273,
      "grad_norm": 0.5450553297996521,
      "learning_rate": 1.1961323614771424e-07,
      "loss": 0.1416,
      "num_input_tokens_seen": 33555792,
      "step": 38755
    },
    {
      "epoch": 18.274398868458274,
      "grad_norm": 0.8964176177978516,
      "learning_rate": 1.1865594302886418e-07,
      "loss": 0.1539,
      "num_input_tokens_seen": 33559872,
      "step": 38760
    },
    {
      "epoch": 18.276756247053278,
      "grad_norm": 0.49842244386672974,
      "learning_rate": 1.1770248692594687e-07,
      "loss": 0.2107,
      "num_input_tokens_seen": 33564096,
      "step": 38765
    },
    {
      "epoch": 18.27911362564828,
      "grad_norm": 0.7721198201179504,
      "learning_rate": 1.167528679859975e-07,
      "loss": 0.1844,
      "num_input_tokens_seen": 33567840,
      "step": 38770
    },
    {
      "epoch": 18.281471004243283,
      "grad_norm": 1.217712163925171,
      "learning_rate": 1.1580708635545446e-07,
      "loss": 0.1923,
      "num_input_tokens_seen": 33573280,
      "step": 38775
    },
    {
      "epoch": 18.283828382838283,
      "grad_norm": 1.4450627565383911,
      "learning_rate": 1.1486514218017885e-07,
      "loss": 0.2136,
      "num_input_tokens_seen": 33577424,
      "step": 38780
    },
    {
      "epoch": 18.286185761433288,
      "grad_norm": 0.933501124382019,
      "learning_rate": 1.1392703560542117e-07,
      "loss": 0.1661,
      "num_input_tokens_seen": 33581184,
      "step": 38785
    },
    {
      "epoch": 18.28854314002829,
      "grad_norm": 0.4976312518119812,
      "learning_rate": 1.129927667758518e-07,
      "loss": 0.1633,
      "num_input_tokens_seen": 33585248,
      "step": 38790
    },
    {
      "epoch": 18.290900518623292,
      "grad_norm": 1.2430620193481445,
      "learning_rate": 1.1206233583554992e-07,
      "loss": 0.2498,
      "num_input_tokens_seen": 33589664,
      "step": 38795
    },
    {
      "epoch": 18.293257897218293,
      "grad_norm": 1.3996394872665405,
      "learning_rate": 1.1113574292799523e-07,
      "loss": 0.2379,
      "num_input_tokens_seen": 33594000,
      "step": 38800
    },
    {
      "epoch": 18.293257897218293,
      "eval_loss": 0.20038294792175293,
      "eval_runtime": 22.1729,
      "eval_samples_per_second": 42.529,
      "eval_steps_per_second": 21.287,
      "num_input_tokens_seen": 33594000,
      "step": 38800
    },
    {
      "epoch": 18.295615275813297,
      "grad_norm": 0.5185421109199524,
      "learning_rate": 1.1021298819608449e-07,
      "loss": 0.1731,
      "num_input_tokens_seen": 33598000,
      "step": 38805
    },
    {
      "epoch": 18.297972654408298,
      "grad_norm": 1.2452195882797241,
      "learning_rate": 1.0929407178211226e-07,
      "loss": 0.2043,
      "num_input_tokens_seen": 33602512,
      "step": 38810
    },
    {
      "epoch": 18.300330033003302,
      "grad_norm": 1.053628921508789,
      "learning_rate": 1.0837899382779293e-07,
      "loss": 0.2028,
      "num_input_tokens_seen": 33607040,
      "step": 38815
    },
    {
      "epoch": 18.302687411598303,
      "grad_norm": 0.9706258177757263,
      "learning_rate": 1.0746775447423862e-07,
      "loss": 0.1493,
      "num_input_tokens_seen": 33611728,
      "step": 38820
    },
    {
      "epoch": 18.305044790193303,
      "grad_norm": 0.33106696605682373,
      "learning_rate": 1.0656035386197583e-07,
      "loss": 0.1553,
      "num_input_tokens_seen": 33615488,
      "step": 38825
    },
    {
      "epoch": 18.307402168788308,
      "grad_norm": 1.2335566282272339,
      "learning_rate": 1.0565679213093982e-07,
      "loss": 0.1151,
      "num_input_tokens_seen": 33619968,
      "step": 38830
    },
    {
      "epoch": 18.309759547383308,
      "grad_norm": 0.6475567817687988,
      "learning_rate": 1.0475706942046638e-07,
      "loss": 0.2211,
      "num_input_tokens_seen": 33623424,
      "step": 38835
    },
    {
      "epoch": 18.312116925978312,
      "grad_norm": 0.8531498908996582,
      "learning_rate": 1.0386118586930282e-07,
      "loss": 0.1932,
      "num_input_tokens_seen": 33627392,
      "step": 38840
    },
    {
      "epoch": 18.314474304573313,
      "grad_norm": 0.5484543442726135,
      "learning_rate": 1.0296914161561367e-07,
      "loss": 0.1422,
      "num_input_tokens_seen": 33631392,
      "step": 38845
    },
    {
      "epoch": 18.316831683168317,
      "grad_norm": 1.0005927085876465,
      "learning_rate": 1.0208093679695552e-07,
      "loss": 0.1646,
      "num_input_tokens_seen": 33636064,
      "step": 38850
    },
    {
      "epoch": 18.319189061763318,
      "grad_norm": 1.306164264678955,
      "learning_rate": 1.0119657155030493e-07,
      "loss": 0.2445,
      "num_input_tokens_seen": 33639936,
      "step": 38855
    },
    {
      "epoch": 18.321546440358322,
      "grad_norm": 1.2342500686645508,
      "learning_rate": 1.003160460120417e-07,
      "loss": 0.2171,
      "num_input_tokens_seen": 33644160,
      "step": 38860
    },
    {
      "epoch": 18.323903818953323,
      "grad_norm": 1.1943793296813965,
      "learning_rate": 9.943936031795165e-08,
      "loss": 0.121,
      "num_input_tokens_seen": 33647936,
      "step": 38865
    },
    {
      "epoch": 18.326261197548327,
      "grad_norm": 0.5502345561981201,
      "learning_rate": 9.856651460323219e-08,
      "loss": 0.1644,
      "num_input_tokens_seen": 33652208,
      "step": 38870
    },
    {
      "epoch": 18.328618576143327,
      "grad_norm": 1.3712879419326782,
      "learning_rate": 9.769750900248953e-08,
      "loss": 0.2134,
      "num_input_tokens_seen": 33655792,
      "step": 38875
    },
    {
      "epoch": 18.33097595473833,
      "grad_norm": 0.7793319225311279,
      "learning_rate": 9.683234364973038e-08,
      "loss": 0.1404,
      "num_input_tokens_seen": 33660176,
      "step": 38880
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 0.42659881711006165,
      "learning_rate": 9.597101867837854e-08,
      "loss": 0.1197,
      "num_input_tokens_seen": 33664768,
      "step": 38885
    },
    {
      "epoch": 18.335690711928336,
      "grad_norm": 0.4577949345111847,
      "learning_rate": 9.511353422125835e-08,
      "loss": 0.1116,
      "num_input_tokens_seen": 33668832,
      "step": 38890
    },
    {
      "epoch": 18.338048090523337,
      "grad_norm": 1.189680576324463,
      "learning_rate": 9.42598904106029e-08,
      "loss": 0.1318,
      "num_input_tokens_seen": 33673408,
      "step": 38895
    },
    {
      "epoch": 18.34040546911834,
      "grad_norm": 2.718334197998047,
      "learning_rate": 9.341008737806245e-08,
      "loss": 0.2024,
      "num_input_tokens_seen": 33678672,
      "step": 38900
    },
    {
      "epoch": 18.34276284771334,
      "grad_norm": 0.7548535466194153,
      "learning_rate": 9.256412525467661e-08,
      "loss": 0.0963,
      "num_input_tokens_seen": 33683120,
      "step": 38905
    },
    {
      "epoch": 18.345120226308346,
      "grad_norm": 0.5264550447463989,
      "learning_rate": 9.172200417091326e-08,
      "loss": 0.1449,
      "num_input_tokens_seen": 33687504,
      "step": 38910
    },
    {
      "epoch": 18.347477604903347,
      "grad_norm": 0.6927786469459534,
      "learning_rate": 9.088372425663239e-08,
      "loss": 0.1632,
      "num_input_tokens_seen": 33691952,
      "step": 38915
    },
    {
      "epoch": 18.34983498349835,
      "grad_norm": 1.6544088125228882,
      "learning_rate": 9.004928564110837e-08,
      "loss": 0.156,
      "num_input_tokens_seen": 33696016,
      "step": 38920
    },
    {
      "epoch": 18.35219236209335,
      "grad_norm": 0.7418910264968872,
      "learning_rate": 8.92186884530244e-08,
      "loss": 0.1712,
      "num_input_tokens_seen": 33700144,
      "step": 38925
    },
    {
      "epoch": 18.354549740688356,
      "grad_norm": 0.787574291229248,
      "learning_rate": 8.83919328204641e-08,
      "loss": 0.1229,
      "num_input_tokens_seen": 33704144,
      "step": 38930
    },
    {
      "epoch": 18.356907119283356,
      "grad_norm": 0.8431331515312195,
      "learning_rate": 8.756901887093105e-08,
      "loss": 0.2589,
      "num_input_tokens_seen": 33710192,
      "step": 38935
    },
    {
      "epoch": 18.35926449787836,
      "grad_norm": 1.420302391052246,
      "learning_rate": 8.674994673132098e-08,
      "loss": 0.198,
      "num_input_tokens_seen": 33714960,
      "step": 38940
    },
    {
      "epoch": 18.36162187647336,
      "grad_norm": 0.8911357522010803,
      "learning_rate": 8.593471652794949e-08,
      "loss": 0.178,
      "num_input_tokens_seen": 33719136,
      "step": 38945
    },
    {
      "epoch": 18.363979255068365,
      "grad_norm": 1.3594200611114502,
      "learning_rate": 8.512332838653548e-08,
      "loss": 0.1812,
      "num_input_tokens_seen": 33722768,
      "step": 38950
    },
    {
      "epoch": 18.366336633663366,
      "grad_norm": 1.0784324407577515,
      "learning_rate": 8.431578243220106e-08,
      "loss": 0.167,
      "num_input_tokens_seen": 33727312,
      "step": 38955
    },
    {
      "epoch": 18.36869401225837,
      "grad_norm": 0.8245601058006287,
      "learning_rate": 8.351207878948552e-08,
      "loss": 0.1855,
      "num_input_tokens_seen": 33731296,
      "step": 38960
    },
    {
      "epoch": 18.37105139085337,
      "grad_norm": 1.5245875120162964,
      "learning_rate": 8.271221758232583e-08,
      "loss": 0.1391,
      "num_input_tokens_seen": 33735200,
      "step": 38965
    },
    {
      "epoch": 18.373408769448375,
      "grad_norm": 0.7410275340080261,
      "learning_rate": 8.191619893407332e-08,
      "loss": 0.1998,
      "num_input_tokens_seen": 33740064,
      "step": 38970
    },
    {
      "epoch": 18.375766148043375,
      "grad_norm": 0.8047024607658386,
      "learning_rate": 8.112402296748534e-08,
      "loss": 0.2005,
      "num_input_tokens_seen": 33744464,
      "step": 38975
    },
    {
      "epoch": 18.37812352663838,
      "grad_norm": 1.5430450439453125,
      "learning_rate": 8.033568980471973e-08,
      "loss": 0.2822,
      "num_input_tokens_seen": 33749792,
      "step": 38980
    },
    {
      "epoch": 18.38048090523338,
      "grad_norm": 1.8405067920684814,
      "learning_rate": 7.955119956735146e-08,
      "loss": 0.2176,
      "num_input_tokens_seen": 33754096,
      "step": 38985
    },
    {
      "epoch": 18.382838283828384,
      "grad_norm": 1.8708691596984863,
      "learning_rate": 7.877055237636155e-08,
      "loss": 0.1667,
      "num_input_tokens_seen": 33758160,
      "step": 38990
    },
    {
      "epoch": 18.385195662423385,
      "grad_norm": 0.6173863410949707,
      "learning_rate": 7.79937483521287e-08,
      "loss": 0.2006,
      "num_input_tokens_seen": 33761584,
      "step": 38995
    },
    {
      "epoch": 18.38755304101839,
      "grad_norm": 1.2932281494140625,
      "learning_rate": 7.722078761444873e-08,
      "loss": 0.1929,
      "num_input_tokens_seen": 33765936,
      "step": 39000
    },
    {
      "epoch": 18.38755304101839,
      "eval_loss": 0.1996067613363266,
      "eval_runtime": 22.189,
      "eval_samples_per_second": 42.498,
      "eval_steps_per_second": 21.272,
      "num_input_tokens_seen": 33765936,
      "step": 39000
    },
    {
      "epoch": 18.38991041961339,
      "grad_norm": 0.379110723733902,
      "learning_rate": 7.645167028252631e-08,
      "loss": 0.1131,
      "num_input_tokens_seen": 33770512,
      "step": 39005
    },
    {
      "epoch": 18.392267798208394,
      "grad_norm": 0.906917154788971,
      "learning_rate": 7.568639647496379e-08,
      "loss": 0.1941,
      "num_input_tokens_seen": 33773904,
      "step": 39010
    },
    {
      "epoch": 18.394625176803395,
      "grad_norm": 0.348568856716156,
      "learning_rate": 7.492496630977508e-08,
      "loss": 0.1671,
      "num_input_tokens_seen": 33777920,
      "step": 39015
    },
    {
      "epoch": 18.396982555398395,
      "grad_norm": 1.6968398094177246,
      "learning_rate": 7.416737990438571e-08,
      "loss": 0.1729,
      "num_input_tokens_seen": 33781344,
      "step": 39020
    },
    {
      "epoch": 18.3993399339934,
      "grad_norm": 0.7160608172416687,
      "learning_rate": 7.341363737562445e-08,
      "loss": 0.185,
      "num_input_tokens_seen": 33785200,
      "step": 39025
    },
    {
      "epoch": 18.4016973125884,
      "grad_norm": 0.5036654472351074,
      "learning_rate": 7.266373883972887e-08,
      "loss": 0.2158,
      "num_input_tokens_seen": 33790320,
      "step": 39030
    },
    {
      "epoch": 18.404054691183404,
      "grad_norm": 0.9801011085510254,
      "learning_rate": 7.191768441233981e-08,
      "loss": 0.1411,
      "num_input_tokens_seen": 33794992,
      "step": 39035
    },
    {
      "epoch": 18.406412069778405,
      "grad_norm": 0.46766987442970276,
      "learning_rate": 7.11754742085069e-08,
      "loss": 0.108,
      "num_input_tokens_seen": 33799728,
      "step": 39040
    },
    {
      "epoch": 18.40876944837341,
      "grad_norm": 2.010944128036499,
      "learning_rate": 7.043710834269413e-08,
      "loss": 0.1753,
      "num_input_tokens_seen": 33803936,
      "step": 39045
    },
    {
      "epoch": 18.41112682696841,
      "grad_norm": 0.6366557478904724,
      "learning_rate": 6.970258692876319e-08,
      "loss": 0.1955,
      "num_input_tokens_seen": 33808144,
      "step": 39050
    },
    {
      "epoch": 18.413484205563414,
      "grad_norm": 0.6438245177268982,
      "learning_rate": 6.897191007998738e-08,
      "loss": 0.2224,
      "num_input_tokens_seen": 33813280,
      "step": 39055
    },
    {
      "epoch": 18.415841584158414,
      "grad_norm": 0.504860520362854,
      "learning_rate": 6.824507790904599e-08,
      "loss": 0.1193,
      "num_input_tokens_seen": 33817440,
      "step": 39060
    },
    {
      "epoch": 18.41819896275342,
      "grad_norm": 2.127075672149658,
      "learning_rate": 6.752209052802439e-08,
      "loss": 0.1625,
      "num_input_tokens_seen": 33822800,
      "step": 39065
    },
    {
      "epoch": 18.42055634134842,
      "grad_norm": 0.8508981466293335,
      "learning_rate": 6.680294804841946e-08,
      "loss": 0.1815,
      "num_input_tokens_seen": 33827376,
      "step": 39070
    },
    {
      "epoch": 18.422913719943423,
      "grad_norm": 0.5480726361274719,
      "learning_rate": 6.608765058112865e-08,
      "loss": 0.1177,
      "num_input_tokens_seen": 33831248,
      "step": 39075
    },
    {
      "epoch": 18.425271098538424,
      "grad_norm": 0.9177204966545105,
      "learning_rate": 6.537619823646368e-08,
      "loss": 0.1606,
      "num_input_tokens_seen": 33835056,
      "step": 39080
    },
    {
      "epoch": 18.427628477133428,
      "grad_norm": 1.1736301183700562,
      "learning_rate": 6.466859112413404e-08,
      "loss": 0.2001,
      "num_input_tokens_seen": 33839040,
      "step": 39085
    },
    {
      "epoch": 18.42998585572843,
      "grad_norm": 1.9994735717773438,
      "learning_rate": 6.39648293532663e-08,
      "loss": 0.2041,
      "num_input_tokens_seen": 33843024,
      "step": 39090
    },
    {
      "epoch": 18.432343234323433,
      "grad_norm": 1.2966618537902832,
      "learning_rate": 6.32649130323848e-08,
      "loss": 0.226,
      "num_input_tokens_seen": 33847232,
      "step": 39095
    },
    {
      "epoch": 18.434700612918434,
      "grad_norm": 1.9063270092010498,
      "learning_rate": 6.256884226943094e-08,
      "loss": 0.1955,
      "num_input_tokens_seen": 33851760,
      "step": 39100
    },
    {
      "epoch": 18.437057991513438,
      "grad_norm": 0.8859390616416931,
      "learning_rate": 6.187661717174386e-08,
      "loss": 0.1824,
      "num_input_tokens_seen": 33855824,
      "step": 39105
    },
    {
      "epoch": 18.43941537010844,
      "grad_norm": 1.5778988599777222,
      "learning_rate": 6.118823784607708e-08,
      "loss": 0.1206,
      "num_input_tokens_seen": 33860128,
      "step": 39110
    },
    {
      "epoch": 18.441772748703443,
      "grad_norm": 1.4788471460342407,
      "learning_rate": 6.050370439858178e-08,
      "loss": 0.0919,
      "num_input_tokens_seen": 33864352,
      "step": 39115
    },
    {
      "epoch": 18.444130127298443,
      "grad_norm": 1.5540004968643188,
      "learning_rate": 5.98230169348235e-08,
      "loss": 0.1673,
      "num_input_tokens_seen": 33869280,
      "step": 39120
    },
    {
      "epoch": 18.446487505893447,
      "grad_norm": 0.6655845642089844,
      "learning_rate": 5.914617555977664e-08,
      "loss": 0.1923,
      "num_input_tokens_seen": 33873472,
      "step": 39125
    },
    {
      "epoch": 18.448844884488448,
      "grad_norm": 0.6646893620491028,
      "learning_rate": 5.8473180377816017e-08,
      "loss": 0.1943,
      "num_input_tokens_seen": 33877344,
      "step": 39130
    },
    {
      "epoch": 18.451202263083452,
      "grad_norm": 0.2991706132888794,
      "learning_rate": 5.780403149272251e-08,
      "loss": 0.145,
      "num_input_tokens_seen": 33881600,
      "step": 39135
    },
    {
      "epoch": 18.453559641678453,
      "grad_norm": 1.4156543016433716,
      "learning_rate": 5.7138729007694126e-08,
      "loss": 0.1771,
      "num_input_tokens_seen": 33885920,
      "step": 39140
    },
    {
      "epoch": 18.455917020273457,
      "grad_norm": 0.75637286901474,
      "learning_rate": 5.64772730253238e-08,
      "loss": 0.1364,
      "num_input_tokens_seen": 33890192,
      "step": 39145
    },
    {
      "epoch": 18.458274398868458,
      "grad_norm": 1.3044885396957397,
      "learning_rate": 5.5819663647618814e-08,
      "loss": 0.2788,
      "num_input_tokens_seen": 33895168,
      "step": 39150
    },
    {
      "epoch": 18.460631777463462,
      "grad_norm": 1.8602378368377686,
      "learning_rate": 5.5165900975989723e-08,
      "loss": 0.223,
      "num_input_tokens_seen": 33899120,
      "step": 39155
    },
    {
      "epoch": 18.462989156058462,
      "grad_norm": 0.932961642742157,
      "learning_rate": 5.451598511125311e-08,
      "loss": 0.2456,
      "num_input_tokens_seen": 33902800,
      "step": 39160
    },
    {
      "epoch": 18.465346534653467,
      "grad_norm": 1.5081273317337036,
      "learning_rate": 5.3869916153637124e-08,
      "loss": 0.1743,
      "num_input_tokens_seen": 33906704,
      "step": 39165
    },
    {
      "epoch": 18.467703913248467,
      "grad_norm": 1.5442746877670288,
      "learning_rate": 5.322769420277318e-08,
      "loss": 0.1766,
      "num_input_tokens_seen": 33911184,
      "step": 39170
    },
    {
      "epoch": 18.47006129184347,
      "grad_norm": 0.8017455339431763,
      "learning_rate": 5.258931935769873e-08,
      "loss": 0.183,
      "num_input_tokens_seen": 33915376,
      "step": 39175
    },
    {
      "epoch": 18.472418670438472,
      "grad_norm": 1.7566661834716797,
      "learning_rate": 5.19547917168628e-08,
      "loss": 0.2056,
      "num_input_tokens_seen": 33919600,
      "step": 39180
    },
    {
      "epoch": 18.474776049033476,
      "grad_norm": 0.5869095921516418,
      "learning_rate": 5.13241113781121e-08,
      "loss": 0.1279,
      "num_input_tokens_seen": 33923520,
      "step": 39185
    },
    {
      "epoch": 18.477133427628477,
      "grad_norm": 2.269312620162964,
      "learning_rate": 5.0697278438707755e-08,
      "loss": 0.1531,
      "num_input_tokens_seen": 33927808,
      "step": 39190
    },
    {
      "epoch": 18.47949080622348,
      "grad_norm": 0.8754624724388123,
      "learning_rate": 5.0074292995316854e-08,
      "loss": 0.1481,
      "num_input_tokens_seen": 33932240,
      "step": 39195
    },
    {
      "epoch": 18.48184818481848,
      "grad_norm": 0.7794775366783142,
      "learning_rate": 4.945515514400978e-08,
      "loss": 0.0899,
      "num_input_tokens_seen": 33936896,
      "step": 39200
    },
    {
      "epoch": 18.48184818481848,
      "eval_loss": 0.20044204592704773,
      "eval_runtime": 22.1965,
      "eval_samples_per_second": 42.484,
      "eval_steps_per_second": 21.265,
      "num_input_tokens_seen": 33936896,
      "step": 39200
    },
    {
      "epoch": 18.484205563413486,
      "grad_norm": 1.2193721532821655,
      "learning_rate": 4.883986498026571e-08,
      "loss": 0.1905,
      "num_input_tokens_seen": 33940832,
      "step": 39205
    },
    {
      "epoch": 18.486562942008486,
      "grad_norm": 1.1095986366271973,
      "learning_rate": 4.822842259896987e-08,
      "loss": 0.2031,
      "num_input_tokens_seen": 33944624,
      "step": 39210
    },
    {
      "epoch": 18.48892032060349,
      "grad_norm": 0.45837467908859253,
      "learning_rate": 4.762082809441626e-08,
      "loss": 0.1075,
      "num_input_tokens_seen": 33948960,
      "step": 39215
    },
    {
      "epoch": 18.49127769919849,
      "grad_norm": 0.6826955676078796,
      "learning_rate": 4.7017081560302156e-08,
      "loss": 0.138,
      "num_input_tokens_seen": 33954144,
      "step": 39220
    },
    {
      "epoch": 18.493635077793492,
      "grad_norm": 0.8845247030258179,
      "learning_rate": 4.6417183089730866e-08,
      "loss": 0.1567,
      "num_input_tokens_seen": 33958944,
      "step": 39225
    },
    {
      "epoch": 18.495992456388496,
      "grad_norm": 0.7736167311668396,
      "learning_rate": 4.5821132775217265e-08,
      "loss": 0.1525,
      "num_input_tokens_seen": 33963248,
      "step": 39230
    },
    {
      "epoch": 18.498349834983497,
      "grad_norm": 1.7422754764556885,
      "learning_rate": 4.5228930708679504e-08,
      "loss": 0.1666,
      "num_input_tokens_seen": 33967392,
      "step": 39235
    },
    {
      "epoch": 18.5007072135785,
      "grad_norm": 0.9127885103225708,
      "learning_rate": 4.464057698144175e-08,
      "loss": 0.1882,
      "num_input_tokens_seen": 33972096,
      "step": 39240
    },
    {
      "epoch": 18.5030645921735,
      "grad_norm": 0.9992003440856934,
      "learning_rate": 4.4056071684236974e-08,
      "loss": 0.1154,
      "num_input_tokens_seen": 33977808,
      "step": 39245
    },
    {
      "epoch": 18.505421970768506,
      "grad_norm": 0.7603959441184998,
      "learning_rate": 4.347541490719864e-08,
      "loss": 0.2158,
      "num_input_tokens_seen": 33982192,
      "step": 39250
    },
    {
      "epoch": 18.507779349363506,
      "grad_norm": 0.4094744622707367,
      "learning_rate": 4.2898606739877336e-08,
      "loss": 0.1467,
      "num_input_tokens_seen": 33986544,
      "step": 39255
    },
    {
      "epoch": 18.51013672795851,
      "grad_norm": 1.8237513303756714,
      "learning_rate": 4.232564727122135e-08,
      "loss": 0.1306,
      "num_input_tokens_seen": 33991456,
      "step": 39260
    },
    {
      "epoch": 18.51249410655351,
      "grad_norm": 0.38282448053359985,
      "learning_rate": 4.1756536589585004e-08,
      "loss": 0.1649,
      "num_input_tokens_seen": 33995264,
      "step": 39265
    },
    {
      "epoch": 18.514851485148515,
      "grad_norm": 0.9641935229301453,
      "learning_rate": 4.119127478273976e-08,
      "loss": 0.1664,
      "num_input_tokens_seen": 33999584,
      "step": 39270
    },
    {
      "epoch": 18.517208863743516,
      "grad_norm": 1.0490328073501587,
      "learning_rate": 4.062986193784923e-08,
      "loss": 0.2497,
      "num_input_tokens_seen": 34002752,
      "step": 39275
    },
    {
      "epoch": 18.51956624233852,
      "grad_norm": 1.1791476011276245,
      "learning_rate": 4.007229814149416e-08,
      "loss": 0.1135,
      "num_input_tokens_seen": 34007200,
      "step": 39280
    },
    {
      "epoch": 18.52192362093352,
      "grad_norm": 1.6914809942245483,
      "learning_rate": 3.951858347965576e-08,
      "loss": 0.2315,
      "num_input_tokens_seen": 34011072,
      "step": 39285
    },
    {
      "epoch": 18.524280999528525,
      "grad_norm": 1.323285698890686,
      "learning_rate": 3.896871803772684e-08,
      "loss": 0.149,
      "num_input_tokens_seen": 34015216,
      "step": 39290
    },
    {
      "epoch": 18.526638378123526,
      "grad_norm": 1.168974757194519,
      "learning_rate": 3.842270190050068e-08,
      "loss": 0.2102,
      "num_input_tokens_seen": 34020048,
      "step": 39295
    },
    {
      "epoch": 18.52899575671853,
      "grad_norm": 0.7963993549346924,
      "learning_rate": 3.7880535152179376e-08,
      "loss": 0.2489,
      "num_input_tokens_seen": 34024352,
      "step": 39300
    },
    {
      "epoch": 18.53135313531353,
      "grad_norm": 0.697597861289978,
      "learning_rate": 3.734221787637382e-08,
      "loss": 0.1679,
      "num_input_tokens_seen": 34027952,
      "step": 39305
    },
    {
      "epoch": 18.533710513908535,
      "grad_norm": 1.3260483741760254,
      "learning_rate": 3.680775015609817e-08,
      "loss": 0.1618,
      "num_input_tokens_seen": 34031856,
      "step": 39310
    },
    {
      "epoch": 18.536067892503535,
      "grad_norm": 1.3610552549362183,
      "learning_rate": 3.627713207377537e-08,
      "loss": 0.1452,
      "num_input_tokens_seen": 34036608,
      "step": 39315
    },
    {
      "epoch": 18.53842527109854,
      "grad_norm": 0.9562854170799255,
      "learning_rate": 3.575036371123164e-08,
      "loss": 0.2232,
      "num_input_tokens_seen": 34040656,
      "step": 39320
    },
    {
      "epoch": 18.54078264969354,
      "grad_norm": 0.6218854784965515,
      "learning_rate": 3.5227445149704776e-08,
      "loss": 0.1779,
      "num_input_tokens_seen": 34044832,
      "step": 39325
    },
    {
      "epoch": 18.543140028288544,
      "grad_norm": 1.0109275579452515,
      "learning_rate": 3.470837646983027e-08,
      "loss": 0.1687,
      "num_input_tokens_seen": 34049808,
      "step": 39330
    },
    {
      "epoch": 18.545497406883545,
      "grad_norm": 1.6117889881134033,
      "learning_rate": 3.419315775165799e-08,
      "loss": 0.2535,
      "num_input_tokens_seen": 34054528,
      "step": 39335
    },
    {
      "epoch": 18.54785478547855,
      "grad_norm": 0.6770122051239014,
      "learning_rate": 3.368178907464103e-08,
      "loss": 0.2316,
      "num_input_tokens_seen": 34058912,
      "step": 39340
    },
    {
      "epoch": 18.55021216407355,
      "grad_norm": 2.3149070739746094,
      "learning_rate": 3.317427051763855e-08,
      "loss": 0.1702,
      "num_input_tokens_seen": 34063504,
      "step": 39345
    },
    {
      "epoch": 18.552569542668554,
      "grad_norm": 1.8057191371917725,
      "learning_rate": 3.267060215891571e-08,
      "loss": 0.2276,
      "num_input_tokens_seen": 34067856,
      "step": 39350
    },
    {
      "epoch": 18.554926921263554,
      "grad_norm": 1.4561383724212646,
      "learning_rate": 3.217078407614649e-08,
      "loss": 0.1655,
      "num_input_tokens_seen": 34071872,
      "step": 39355
    },
    {
      "epoch": 18.55728429985856,
      "grad_norm": 1.3862606287002563,
      "learning_rate": 3.1674816346405345e-08,
      "loss": 0.2121,
      "num_input_tokens_seen": 34076736,
      "step": 39360
    },
    {
      "epoch": 18.55964167845356,
      "grad_norm": 0.8242679834365845,
      "learning_rate": 3.11826990461811e-08,
      "loss": 0.1775,
      "num_input_tokens_seen": 34081408,
      "step": 39365
    },
    {
      "epoch": 18.561999057048563,
      "grad_norm": 0.6338306069374084,
      "learning_rate": 3.069443225136304e-08,
      "loss": 0.1939,
      "num_input_tokens_seen": 34085552,
      "step": 39370
    },
    {
      "epoch": 18.564356435643564,
      "grad_norm": 2.0394177436828613,
      "learning_rate": 3.021001603724372e-08,
      "loss": 0.207,
      "num_input_tokens_seen": 34090320,
      "step": 39375
    },
    {
      "epoch": 18.566713814238568,
      "grad_norm": 0.6640616059303284,
      "learning_rate": 2.9729450478532818e-08,
      "loss": 0.2064,
      "num_input_tokens_seen": 34094208,
      "step": 39380
    },
    {
      "epoch": 18.56907119283357,
      "grad_norm": 1.1846332550048828,
      "learning_rate": 2.9252735649337726e-08,
      "loss": 0.1909,
      "num_input_tokens_seen": 34098960,
      "step": 39385
    },
    {
      "epoch": 18.571428571428573,
      "grad_norm": 0.8842399716377258,
      "learning_rate": 2.8779871623171863e-08,
      "loss": 0.1713,
      "num_input_tokens_seen": 34102944,
      "step": 39390
    },
    {
      "epoch": 18.573785950023574,
      "grad_norm": 0.5605196356773376,
      "learning_rate": 2.8310858472957448e-08,
      "loss": 0.1594,
      "num_input_tokens_seen": 34106752,
      "step": 39395
    },
    {
      "epoch": 18.576143328618578,
      "grad_norm": 1.0313173532485962,
      "learning_rate": 2.784569627101996e-08,
      "loss": 0.1772,
      "num_input_tokens_seen": 34110592,
      "step": 39400
    },
    {
      "epoch": 18.576143328618578,
      "eval_loss": 0.20034067332744598,
      "eval_runtime": 22.1782,
      "eval_samples_per_second": 42.519,
      "eval_steps_per_second": 21.282,
      "num_input_tokens_seen": 34110592,
      "step": 39400
    },
    {
      "epoch": 18.57850070721358,
      "grad_norm": 0.9206286072731018,
      "learning_rate": 2.738438508909924e-08,
      "loss": 0.1696,
      "num_input_tokens_seen": 34114256,
      "step": 39405
    },
    {
      "epoch": 18.580858085808583,
      "grad_norm": 0.9592415690422058,
      "learning_rate": 2.692692499833005e-08,
      "loss": 0.1431,
      "num_input_tokens_seen": 34118240,
      "step": 39410
    },
    {
      "epoch": 18.583215464403583,
      "grad_norm": 1.4797699451446533,
      "learning_rate": 2.647331606926151e-08,
      "loss": 0.2363,
      "num_input_tokens_seen": 34122896,
      "step": 39415
    },
    {
      "epoch": 18.585572842998587,
      "grad_norm": 1.3508158922195435,
      "learning_rate": 2.6023558371843225e-08,
      "loss": 0.2321,
      "num_input_tokens_seen": 34127008,
      "step": 39420
    },
    {
      "epoch": 18.587930221593588,
      "grad_norm": 0.9912437796592712,
      "learning_rate": 2.557765197543638e-08,
      "loss": 0.1446,
      "num_input_tokens_seen": 34131632,
      "step": 39425
    },
    {
      "epoch": 18.59028760018859,
      "grad_norm": 1.0897772312164307,
      "learning_rate": 2.513559694880263e-08,
      "loss": 0.1568,
      "num_input_tokens_seen": 34136000,
      "step": 39430
    },
    {
      "epoch": 18.592644978783593,
      "grad_norm": 1.0582655668258667,
      "learning_rate": 2.469739336011523e-08,
      "loss": 0.1874,
      "num_input_tokens_seen": 34140672,
      "step": 39435
    },
    {
      "epoch": 18.595002357378593,
      "grad_norm": 0.865576446056366,
      "learning_rate": 2.4263041276947894e-08,
      "loss": 0.1815,
      "num_input_tokens_seen": 34144512,
      "step": 39440
    },
    {
      "epoch": 18.597359735973598,
      "grad_norm": 0.9602841734886169,
      "learning_rate": 2.3832540766283164e-08,
      "loss": 0.1734,
      "num_input_tokens_seen": 34148288,
      "step": 39445
    },
    {
      "epoch": 18.599717114568598,
      "grad_norm": 1.0949794054031372,
      "learning_rate": 2.3405891894512366e-08,
      "loss": 0.2118,
      "num_input_tokens_seen": 34152128,
      "step": 39450
    },
    {
      "epoch": 18.602074493163602,
      "grad_norm": 0.8776617646217346,
      "learning_rate": 2.29830947274301e-08,
      "loss": 0.2252,
      "num_input_tokens_seen": 34156192,
      "step": 39455
    },
    {
      "epoch": 18.604431871758603,
      "grad_norm": 1.0528937578201294,
      "learning_rate": 2.2564149330231432e-08,
      "loss": 0.1433,
      "num_input_tokens_seen": 34160064,
      "step": 39460
    },
    {
      "epoch": 18.606789250353607,
      "grad_norm": 0.8491555452346802,
      "learning_rate": 2.2149055767528572e-08,
      "loss": 0.2464,
      "num_input_tokens_seen": 34164160,
      "step": 39465
    },
    {
      "epoch": 18.609146628948608,
      "grad_norm": 0.8705281615257263,
      "learning_rate": 2.1737814103334197e-08,
      "loss": 0.1742,
      "num_input_tokens_seen": 34168944,
      "step": 39470
    },
    {
      "epoch": 18.611504007543612,
      "grad_norm": 0.6430046558380127,
      "learning_rate": 2.1330424401064253e-08,
      "loss": 0.1667,
      "num_input_tokens_seen": 34173856,
      "step": 39475
    },
    {
      "epoch": 18.613861386138613,
      "grad_norm": 0.7918540835380554,
      "learning_rate": 2.092688672354348e-08,
      "loss": 0.152,
      "num_input_tokens_seen": 34178624,
      "step": 39480
    },
    {
      "epoch": 18.616218764733617,
      "grad_norm": 1.2261419296264648,
      "learning_rate": 2.0527201133005435e-08,
      "loss": 0.1986,
      "num_input_tokens_seen": 34184048,
      "step": 39485
    },
    {
      "epoch": 18.618576143328617,
      "grad_norm": 1.4284464120864868,
      "learning_rate": 2.0131367691084148e-08,
      "loss": 0.2597,
      "num_input_tokens_seen": 34188256,
      "step": 39490
    },
    {
      "epoch": 18.62093352192362,
      "grad_norm": 1.2077770233154297,
      "learning_rate": 1.9739386458819675e-08,
      "loss": 0.2459,
      "num_input_tokens_seen": 34192544,
      "step": 39495
    },
    {
      "epoch": 18.623290900518622,
      "grad_norm": 1.030333399772644,
      "learning_rate": 1.9351257496666442e-08,
      "loss": 0.2304,
      "num_input_tokens_seen": 34196896,
      "step": 39500
    },
    {
      "epoch": 18.625648279113626,
      "grad_norm": 1.9588446617126465,
      "learning_rate": 1.896698086447657e-08,
      "loss": 0.1856,
      "num_input_tokens_seen": 34200512,
      "step": 39505
    },
    {
      "epoch": 18.628005657708627,
      "grad_norm": 0.4172164797782898,
      "learning_rate": 1.8586556621505436e-08,
      "loss": 0.1024,
      "num_input_tokens_seen": 34204672,
      "step": 39510
    },
    {
      "epoch": 18.63036303630363,
      "grad_norm": 1.3222460746765137,
      "learning_rate": 1.820998482642833e-08,
      "loss": 0.1681,
      "num_input_tokens_seen": 34209472,
      "step": 39515
    },
    {
      "epoch": 18.632720414898632,
      "grad_norm": 1.1335541009902954,
      "learning_rate": 1.7837265537309912e-08,
      "loss": 0.1543,
      "num_input_tokens_seen": 34214064,
      "step": 39520
    },
    {
      "epoch": 18.635077793493636,
      "grad_norm": 0.579404354095459,
      "learning_rate": 1.7468398811629206e-08,
      "loss": 0.1387,
      "num_input_tokens_seen": 34218864,
      "step": 39525
    },
    {
      "epoch": 18.637435172088637,
      "grad_norm": 0.913057267665863,
      "learning_rate": 1.710338470627404e-08,
      "loss": 0.136,
      "num_input_tokens_seen": 34222768,
      "step": 39530
    },
    {
      "epoch": 18.63979255068364,
      "grad_norm": 1.0144129991531372,
      "learning_rate": 1.6742223277529945e-08,
      "loss": 0.1368,
      "num_input_tokens_seen": 34226896,
      "step": 39535
    },
    {
      "epoch": 18.64214992927864,
      "grad_norm": 0.4987218677997589,
      "learning_rate": 1.6384914581094036e-08,
      "loss": 0.1676,
      "num_input_tokens_seen": 34231472,
      "step": 39540
    },
    {
      "epoch": 18.644507307873646,
      "grad_norm": 1.1486434936523438,
      "learning_rate": 1.6031458672069455e-08,
      "loss": 0.1955,
      "num_input_tokens_seen": 34235696,
      "step": 39545
    },
    {
      "epoch": 18.646864686468646,
      "grad_norm": 1.7743651866912842,
      "learning_rate": 1.5681855604962602e-08,
      "loss": 0.2212,
      "num_input_tokens_seen": 34240256,
      "step": 39550
    },
    {
      "epoch": 18.64922206506365,
      "grad_norm": 0.4767345190048218,
      "learning_rate": 1.5336105433683135e-08,
      "loss": 0.1281,
      "num_input_tokens_seen": 34243760,
      "step": 39555
    },
    {
      "epoch": 18.65157944365865,
      "grad_norm": 0.8507617115974426,
      "learning_rate": 1.499420821155506e-08,
      "loss": 0.218,
      "num_input_tokens_seen": 34249200,
      "step": 39560
    },
    {
      "epoch": 18.653936822253655,
      "grad_norm": 0.7313666343688965,
      "learning_rate": 1.4656163991302874e-08,
      "loss": 0.1709,
      "num_input_tokens_seen": 34252656,
      "step": 39565
    },
    {
      "epoch": 18.656294200848656,
      "grad_norm": 0.700857937335968,
      "learning_rate": 1.4321972825051544e-08,
      "loss": 0.1407,
      "num_input_tokens_seen": 34256688,
      "step": 39570
    },
    {
      "epoch": 18.65865157944366,
      "grad_norm": 1.542574167251587,
      "learning_rate": 1.3991634764345951e-08,
      "loss": 0.1477,
      "num_input_tokens_seen": 34261440,
      "step": 39575
    },
    {
      "epoch": 18.66100895803866,
      "grad_norm": 1.9268170595169067,
      "learning_rate": 1.3665149860120352e-08,
      "loss": 0.2571,
      "num_input_tokens_seen": 34265776,
      "step": 39580
    },
    {
      "epoch": 18.663366336633665,
      "grad_norm": 0.4252636432647705,
      "learning_rate": 1.3342518162728912e-08,
      "loss": 0.1621,
      "num_input_tokens_seen": 34270784,
      "step": 39585
    },
    {
      "epoch": 18.665723715228665,
      "grad_norm": 1.2576943635940552,
      "learning_rate": 1.30237397219235e-08,
      "loss": 0.1867,
      "num_input_tokens_seen": 34274736,
      "step": 39590
    },
    {
      "epoch": 18.66808109382367,
      "grad_norm": 1.7697067260742188,
      "learning_rate": 1.2708814586862016e-08,
      "loss": 0.1217,
      "num_input_tokens_seen": 34280352,
      "step": 39595
    },
    {
      "epoch": 18.67043847241867,
      "grad_norm": 0.6585330367088318,
      "learning_rate": 1.2397742806111168e-08,
      "loss": 0.1775,
      "num_input_tokens_seen": 34284208,
      "step": 39600
    },
    {
      "epoch": 18.67043847241867,
      "eval_loss": 0.20034067332744598,
      "eval_runtime": 22.1425,
      "eval_samples_per_second": 42.588,
      "eval_steps_per_second": 21.316,
      "num_input_tokens_seen": 34284208,
      "step": 39600
    },
    {
      "epoch": 18.672795851013674,
      "grad_norm": 0.9547841548919678,
      "learning_rate": 1.209052442764369e-08,
      "loss": 0.1924,
      "num_input_tokens_seen": 34288928,
      "step": 39605
    },
    {
      "epoch": 18.675153229608675,
      "grad_norm": 0.557896614074707,
      "learning_rate": 1.17871594988328e-08,
      "loss": 0.1592,
      "num_input_tokens_seen": 34293136,
      "step": 39610
    },
    {
      "epoch": 18.677510608203676,
      "grad_norm": 0.7036988139152527,
      "learning_rate": 1.1487648066466072e-08,
      "loss": 0.1762,
      "num_input_tokens_seen": 34297984,
      "step": 39615
    },
    {
      "epoch": 18.67986798679868,
      "grad_norm": 1.289360761642456,
      "learning_rate": 1.1191990176728784e-08,
      "loss": 0.179,
      "num_input_tokens_seen": 34301904,
      "step": 39620
    },
    {
      "epoch": 18.68222536539368,
      "grad_norm": 0.5189416408538818,
      "learning_rate": 1.0900185875215018e-08,
      "loss": 0.2056,
      "num_input_tokens_seen": 34306368,
      "step": 39625
    },
    {
      "epoch": 18.684582743988685,
      "grad_norm": 1.257908582687378,
      "learning_rate": 1.0612235206924891e-08,
      "loss": 0.1434,
      "num_input_tokens_seen": 34310880,
      "step": 39630
    },
    {
      "epoch": 18.686940122583685,
      "grad_norm": 1.3460776805877686,
      "learning_rate": 1.0328138216264549e-08,
      "loss": 0.1532,
      "num_input_tokens_seen": 34314592,
      "step": 39635
    },
    {
      "epoch": 18.68929750117869,
      "grad_norm": 0.4687715470790863,
      "learning_rate": 1.004789494704339e-08,
      "loss": 0.2288,
      "num_input_tokens_seen": 34318864,
      "step": 39640
    },
    {
      "epoch": 18.69165487977369,
      "grad_norm": 1.1610796451568604,
      "learning_rate": 9.771505442482397e-09,
      "loss": 0.1264,
      "num_input_tokens_seen": 34324000,
      "step": 39645
    },
    {
      "epoch": 18.694012258368694,
      "grad_norm": 0.8195088505744934,
      "learning_rate": 9.498969745200259e-09,
      "loss": 0.1309,
      "num_input_tokens_seen": 34327840,
      "step": 39650
    },
    {
      "epoch": 18.696369636963695,
      "grad_norm": 1.1583518981933594,
      "learning_rate": 9.230287897230017e-09,
      "loss": 0.205,
      "num_input_tokens_seen": 34331664,
      "step": 39655
    },
    {
      "epoch": 18.6987270155587,
      "grad_norm": 0.5088052153587341,
      "learning_rate": 8.965459940002419e-09,
      "loss": 0.1822,
      "num_input_tokens_seen": 34336016,
      "step": 39660
    },
    {
      "epoch": 18.7010843941537,
      "grad_norm": 0.6986492276191711,
      "learning_rate": 8.704485914357019e-09,
      "loss": 0.1658,
      "num_input_tokens_seen": 34341072,
      "step": 39665
    },
    {
      "epoch": 18.703441772748704,
      "grad_norm": 0.5702382922172546,
      "learning_rate": 8.447365860539402e-09,
      "loss": 0.1294,
      "num_input_tokens_seen": 34345824,
      "step": 39670
    },
    {
      "epoch": 18.705799151343705,
      "grad_norm": 1.2982231378555298,
      "learning_rate": 8.194099818201184e-09,
      "loss": 0.2393,
      "num_input_tokens_seen": 34350784,
      "step": 39675
    },
    {
      "epoch": 18.70815652993871,
      "grad_norm": 0.7465348243713379,
      "learning_rate": 7.944687826400011e-09,
      "loss": 0.1757,
      "num_input_tokens_seen": 34354624,
      "step": 39680
    },
    {
      "epoch": 18.71051390853371,
      "grad_norm": 1.5830048322677612,
      "learning_rate": 7.699129923599557e-09,
      "loss": 0.2308,
      "num_input_tokens_seen": 34359088,
      "step": 39685
    },
    {
      "epoch": 18.712871287128714,
      "grad_norm": 0.9249418377876282,
      "learning_rate": 7.457426147663982e-09,
      "loss": 0.2091,
      "num_input_tokens_seen": 34363008,
      "step": 39690
    },
    {
      "epoch": 18.715228665723714,
      "grad_norm": 1.9754085540771484,
      "learning_rate": 7.219576535871797e-09,
      "loss": 0.132,
      "num_input_tokens_seen": 34367504,
      "step": 39695
    },
    {
      "epoch": 18.71758604431872,
      "grad_norm": 0.926866352558136,
      "learning_rate": 6.985581124896445e-09,
      "loss": 0.1743,
      "num_input_tokens_seen": 34371760,
      "step": 39700
    },
    {
      "epoch": 18.71994342291372,
      "grad_norm": 1.155371069908142,
      "learning_rate": 6.755439950828501e-09,
      "loss": 0.1622,
      "num_input_tokens_seen": 34375440,
      "step": 39705
    },
    {
      "epoch": 18.722300801508723,
      "grad_norm": 1.5554827451705933,
      "learning_rate": 6.5291530491562444e-09,
      "loss": 0.1508,
      "num_input_tokens_seen": 34379440,
      "step": 39710
    },
    {
      "epoch": 18.724658180103724,
      "grad_norm": 0.9865996837615967,
      "learning_rate": 6.3067204547739845e-09,
      "loss": 0.1909,
      "num_input_tokens_seen": 34383552,
      "step": 39715
    },
    {
      "epoch": 18.727015558698728,
      "grad_norm": 0.5987734198570251,
      "learning_rate": 6.088142201987612e-09,
      "loss": 0.1709,
      "num_input_tokens_seen": 34387136,
      "step": 39720
    },
    {
      "epoch": 18.72937293729373,
      "grad_norm": 0.5582821369171143,
      "learning_rate": 5.873418324503499e-09,
      "loss": 0.1523,
      "num_input_tokens_seen": 34391536,
      "step": 39725
    },
    {
      "epoch": 18.731730315888733,
      "grad_norm": 1.1782606840133667,
      "learning_rate": 5.6625488554340465e-09,
      "loss": 0.0684,
      "num_input_tokens_seen": 34395200,
      "step": 39730
    },
    {
      "epoch": 18.734087694483733,
      "grad_norm": 1.8061250448226929,
      "learning_rate": 5.455533827297688e-09,
      "loss": 0.2439,
      "num_input_tokens_seen": 34399584,
      "step": 39735
    },
    {
      "epoch": 18.736445073078738,
      "grad_norm": 1.06574547290802,
      "learning_rate": 5.252373272018885e-09,
      "loss": 0.1432,
      "num_input_tokens_seen": 34405296,
      "step": 39740
    },
    {
      "epoch": 18.738802451673738,
      "grad_norm": 0.6083155870437622,
      "learning_rate": 5.053067220925356e-09,
      "loss": 0.1551,
      "num_input_tokens_seen": 34409360,
      "step": 39745
    },
    {
      "epoch": 18.741159830268742,
      "grad_norm": 1.1293972730636597,
      "learning_rate": 4.857615704759177e-09,
      "loss": 0.1594,
      "num_input_tokens_seen": 34413136,
      "step": 39750
    },
    {
      "epoch": 18.743517208863743,
      "grad_norm": 0.6515628695487976,
      "learning_rate": 4.666018753654577e-09,
      "loss": 0.1523,
      "num_input_tokens_seen": 34417488,
      "step": 39755
    },
    {
      "epoch": 18.745874587458747,
      "grad_norm": 1.846292495727539,
      "learning_rate": 4.478276397162917e-09,
      "loss": 0.1877,
      "num_input_tokens_seen": 34421680,
      "step": 39760
    },
    {
      "epoch": 18.748231966053748,
      "grad_norm": 2.1800994873046875,
      "learning_rate": 4.294388664233262e-09,
      "loss": 0.1646,
      "num_input_tokens_seen": 34425296,
      "step": 39765
    },
    {
      "epoch": 18.750589344648752,
      "grad_norm": 1.309867262840271,
      "learning_rate": 4.114355583223484e-09,
      "loss": 0.1962,
      "num_input_tokens_seen": 34430256,
      "step": 39770
    },
    {
      "epoch": 18.752946723243753,
      "grad_norm": 0.8243728876113892,
      "learning_rate": 3.9381771818974845e-09,
      "loss": 0.1775,
      "num_input_tokens_seen": 34435024,
      "step": 39775
    },
    {
      "epoch": 18.755304101838757,
      "grad_norm": 0.9068992137908936,
      "learning_rate": 3.765853487427973e-09,
      "loss": 0.1976,
      "num_input_tokens_seen": 34440256,
      "step": 39780
    },
    {
      "epoch": 18.757661480433757,
      "grad_norm": 0.8770253658294678,
      "learning_rate": 3.5973845263825857e-09,
      "loss": 0.1971,
      "num_input_tokens_seen": 34444752,
      "step": 39785
    },
    {
      "epoch": 18.76001885902876,
      "grad_norm": 0.9406282901763916,
      "learning_rate": 3.4327703247488684e-09,
      "loss": 0.1788,
      "num_input_tokens_seen": 34449952,
      "step": 39790
    },
    {
      "epoch": 18.762376237623762,
      "grad_norm": 2.015291452407837,
      "learning_rate": 3.2720109079037443e-09,
      "loss": 0.2485,
      "num_input_tokens_seen": 34454112,
      "step": 39795
    },
    {
      "epoch": 18.764733616218766,
      "grad_norm": 1.417697548866272,
      "learning_rate": 3.1151063006468193e-09,
      "loss": 0.157,
      "num_input_tokens_seen": 34458576,
      "step": 39800
    },
    {
      "epoch": 18.764733616218766,
      "eval_loss": 0.20034067332744598,
      "eval_runtime": 22.1634,
      "eval_samples_per_second": 42.548,
      "eval_steps_per_second": 21.296,
      "num_input_tokens_seen": 34458576,
      "step": 39800
    },
    {
      "epoch": 18.767090994813767,
      "grad_norm": 0.4762636423110962,
      "learning_rate": 2.962056527169854e-09,
      "loss": 0.1226,
      "num_input_tokens_seen": 34462688,
      "step": 39805
    },
    {
      "epoch": 18.76944837340877,
      "grad_norm": 0.6564816832542419,
      "learning_rate": 2.8128616110761898e-09,
      "loss": 0.1957,
      "num_input_tokens_seen": 34466688,
      "step": 39810
    },
    {
      "epoch": 18.77180575200377,
      "grad_norm": 0.3600189685821533,
      "learning_rate": 2.6675215753724223e-09,
      "loss": 0.1247,
      "num_input_tokens_seen": 34471056,
      "step": 39815
    },
    {
      "epoch": 18.774163130598772,
      "grad_norm": 0.7892587780952454,
      "learning_rate": 2.5260364424739557e-09,
      "loss": 0.1695,
      "num_input_tokens_seen": 34476880,
      "step": 39820
    },
    {
      "epoch": 18.776520509193777,
      "grad_norm": 1.645873785018921,
      "learning_rate": 2.3884062341994475e-09,
      "loss": 0.1543,
      "num_input_tokens_seen": 34480576,
      "step": 39825
    },
    {
      "epoch": 18.778877887788777,
      "grad_norm": 0.5833899974822998,
      "learning_rate": 2.25463097177081e-09,
      "loss": 0.1699,
      "num_input_tokens_seen": 34484960,
      "step": 39830
    },
    {
      "epoch": 18.78123526638378,
      "grad_norm": 1.408219337463379,
      "learning_rate": 2.1247106758215397e-09,
      "loss": 0.1836,
      "num_input_tokens_seen": 34489392,
      "step": 39835
    },
    {
      "epoch": 18.783592644978782,
      "grad_norm": 1.1617603302001953,
      "learning_rate": 1.998645366382834e-09,
      "loss": 0.1829,
      "num_input_tokens_seen": 34494096,
      "step": 39840
    },
    {
      "epoch": 18.785950023573786,
      "grad_norm": 0.7321072816848755,
      "learning_rate": 1.876435062897475e-09,
      "loss": 0.2708,
      "num_input_tokens_seen": 34499232,
      "step": 39845
    },
    {
      "epoch": 18.788307402168787,
      "grad_norm": 0.5982344746589661,
      "learning_rate": 1.758079784211497e-09,
      "loss": 0.1478,
      "num_input_tokens_seen": 34503120,
      "step": 39850
    },
    {
      "epoch": 18.79066478076379,
      "grad_norm": 0.5653239488601685,
      "learning_rate": 1.6435795485797434e-09,
      "loss": 0.3063,
      "num_input_tokens_seen": 34506656,
      "step": 39855
    },
    {
      "epoch": 18.79302215935879,
      "grad_norm": 0.9366185665130615,
      "learning_rate": 1.5329343736547596e-09,
      "loss": 0.1617,
      "num_input_tokens_seen": 34509840,
      "step": 39860
    },
    {
      "epoch": 18.795379537953796,
      "grad_norm": 1.2121479511260986,
      "learning_rate": 1.4261442765006739e-09,
      "loss": 0.1687,
      "num_input_tokens_seen": 34514320,
      "step": 39865
    },
    {
      "epoch": 18.797736916548796,
      "grad_norm": 0.6052267551422119,
      "learning_rate": 1.3232092735876445e-09,
      "loss": 0.1761,
      "num_input_tokens_seen": 34518480,
      "step": 39870
    },
    {
      "epoch": 18.8000942951438,
      "grad_norm": 1.2152436971664429,
      "learning_rate": 1.2241293807918607e-09,
      "loss": 0.1294,
      "num_input_tokens_seen": 34522592,
      "step": 39875
    },
    {
      "epoch": 18.8024516737388,
      "grad_norm": 0.30286020040512085,
      "learning_rate": 1.128904613387216e-09,
      "loss": 0.1319,
      "num_input_tokens_seen": 34527728,
      "step": 39880
    },
    {
      "epoch": 18.804809052333805,
      "grad_norm": 1.3432954549789429,
      "learning_rate": 1.0375349860591853e-09,
      "loss": 0.1457,
      "num_input_tokens_seen": 34531584,
      "step": 39885
    },
    {
      "epoch": 18.807166430928806,
      "grad_norm": 1.581465721130371,
      "learning_rate": 9.5002051290205e-10,
      "loss": 0.2255,
      "num_input_tokens_seen": 34535776,
      "step": 39890
    },
    {
      "epoch": 18.80952380952381,
      "grad_norm": 0.3367459774017334,
      "learning_rate": 8.663612074077954e-10,
      "loss": 0.1506,
      "num_input_tokens_seen": 34540704,
      "step": 39895
    },
    {
      "epoch": 18.81188118811881,
      "grad_norm": 1.7535662651062012,
      "learning_rate": 7.865570824799884e-10,
      "loss": 0.2187,
      "num_input_tokens_seen": 34545504,
      "step": 39900
    },
    {
      "epoch": 18.814238566713815,
      "grad_norm": 1.0376657247543335,
      "learning_rate": 7.106081504254514e-10,
      "loss": 0.1957,
      "num_input_tokens_seen": 34549600,
      "step": 39905
    },
    {
      "epoch": 18.816595945308816,
      "grad_norm": 1.1236668825149536,
      "learning_rate": 6.385144229570372e-10,
      "loss": 0.1717,
      "num_input_tokens_seen": 34553872,
      "step": 39910
    },
    {
      "epoch": 18.81895332390382,
      "grad_norm": 0.6876412630081177,
      "learning_rate": 5.70275911190854e-10,
      "loss": 0.1647,
      "num_input_tokens_seen": 34558400,
      "step": 39915
    },
    {
      "epoch": 18.82131070249882,
      "grad_norm": 1.3965567350387573,
      "learning_rate": 5.058926256490403e-10,
      "loss": 0.2215,
      "num_input_tokens_seen": 34562400,
      "step": 39920
    },
    {
      "epoch": 18.823668081093825,
      "grad_norm": 1.4038293361663818,
      "learning_rate": 4.4536457626254134e-10,
      "loss": 0.1725,
      "num_input_tokens_seen": 34567024,
      "step": 39925
    },
    {
      "epoch": 18.826025459688825,
      "grad_norm": 1.0282695293426514,
      "learning_rate": 3.88691772365557e-10,
      "loss": 0.1729,
      "num_input_tokens_seen": 34571696,
      "step": 39930
    },
    {
      "epoch": 18.82838283828383,
      "grad_norm": 0.9390859007835388,
      "learning_rate": 3.358742226955425e-10,
      "loss": 0.1813,
      "num_input_tokens_seen": 34575904,
      "step": 39935
    },
    {
      "epoch": 18.83074021687883,
      "grad_norm": 1.8882423639297485,
      "learning_rate": 2.8691193539875925e-10,
      "loss": 0.2595,
      "num_input_tokens_seen": 34579440,
      "step": 39940
    },
    {
      "epoch": 18.833097595473834,
      "grad_norm": 0.9665737748146057,
      "learning_rate": 2.418049180274995e-10,
      "loss": 0.1767,
      "num_input_tokens_seen": 34583632,
      "step": 39945
    },
    {
      "epoch": 18.835454974068835,
      "grad_norm": 0.5446557402610779,
      "learning_rate": 2.005531775373104e-10,
      "loss": 0.2455,
      "num_input_tokens_seen": 34588304,
      "step": 39950
    },
    {
      "epoch": 18.83781235266384,
      "grad_norm": 1.0432910919189453,
      "learning_rate": 1.6315672028699435e-10,
      "loss": 0.1515,
      "num_input_tokens_seen": 34592464,
      "step": 39955
    },
    {
      "epoch": 18.84016973125884,
      "grad_norm": 0.8507328629493713,
      "learning_rate": 1.2961555204693555e-10,
      "loss": 0.1881,
      "num_input_tokens_seen": 34597024,
      "step": 39960
    },
    {
      "epoch": 18.842527109853844,
      "grad_norm": 1.0281962156295776,
      "learning_rate": 9.992967798799768e-11,
      "loss": 0.1338,
      "num_input_tokens_seen": 34601488,
      "step": 39965
    },
    {
      "epoch": 18.844884488448844,
      "grad_norm": 0.4385526776313782,
      "learning_rate": 7.409910268707521e-11,
      "loss": 0.1277,
      "num_input_tokens_seen": 34606048,
      "step": 39970
    },
    {
      "epoch": 18.84724186704385,
      "grad_norm": 1.0242528915405273,
      "learning_rate": 5.212383012986877e-11,
      "loss": 0.2619,
      "num_input_tokens_seen": 34611376,
      "step": 39975
    },
    {
      "epoch": 18.84959924563885,
      "grad_norm": 0.35856783390045166,
      "learning_rate": 3.400386370533415e-11,
      "loss": 0.2161,
      "num_input_tokens_seen": 34616288,
      "step": 39980
    },
    {
      "epoch": 18.851956624233853,
      "grad_norm": 1.7590653896331787,
      "learning_rate": 1.9739206205682258e-11,
      "loss": 0.1823,
      "num_input_tokens_seen": 34620368,
      "step": 39985
    },
    {
      "epoch": 18.854314002828854,
      "grad_norm": 0.6212170720100403,
      "learning_rate": 9.329859829154685e-12,
      "loss": 0.1508,
      "num_input_tokens_seen": 34624976,
      "step": 39990
    },
    {
      "epoch": 18.85667138142386,
      "grad_norm": 0.5000754594802856,
      "learning_rate": 2.7758261855748148e-12,
      "loss": 0.1509,
      "num_input_tokens_seen": 34629024,
      "step": 39995
    },
    {
      "epoch": 18.85902876001886,
      "grad_norm": 1.822178602218628,
      "learning_rate": 7.710628524559838e-14,
      "loss": 0.2026,
      "num_input_tokens_seen": 34633072,
      "step": 40000
    },
    {
      "epoch": 18.85902876001886,
      "eval_loss": 0.20034067332744598,
      "eval_runtime": 22.1688,
      "eval_samples_per_second": 42.537,
      "eval_steps_per_second": 21.291,
      "num_input_tokens_seen": 34633072,
      "step": 40000
    },
    {
      "epoch": 18.85902876001886,
      "num_input_tokens_seen": 34633072,
      "step": 40000,
      "total_flos": 1.450306979895337e+17,
      "train_loss": 0.19550500297993423,
      "train_runtime": 18418.8323,
      "train_samples_per_second": 8.687,
      "train_steps_per_second": 2.172
    }
  ],
  "logging_steps": 5,
  "max_steps": 40000,
  "num_input_tokens_seen": 34633072,
  "num_train_epochs": 19,
  "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": 1.450306979895337e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}