{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.6222222222222222,
  "eval_steps": 500,
  "global_step": 672,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004629629629629629,
      "grad_norm": 36.40811538696289,
      "learning_rate": 1.6634586466165415e-06,
      "loss": 0.3689,
      "mean_token_accuracy": 0.8758680820465088,
      "num_tokens": 463066.0,
      "step": 5
    },
    {
      "epoch": 0.009259259259259259,
      "grad_norm": 1.924208164215088,
      "learning_rate": 3.7427819548872183e-06,
      "loss": 0.1388,
      "mean_token_accuracy": 0.9592014193534851,
      "num_tokens": 925139.0,
      "step": 10
    },
    {
      "epoch": 0.013888888888888888,
      "grad_norm": 0.7631484866142273,
      "learning_rate": 5.822105263157896e-06,
      "loss": 0.1049,
      "mean_token_accuracy": 0.9600694298744201,
      "num_tokens": 1384956.0,
      "step": 15
    },
    {
      "epoch": 0.018518518518518517,
      "grad_norm": 0.6139844059944153,
      "learning_rate": 7.901428571428572e-06,
      "loss": 0.0921,
      "mean_token_accuracy": 0.9657985925674438,
      "num_tokens": 1845242.0,
      "step": 20
    },
    {
      "epoch": 0.023148148148148147,
      "grad_norm": 0.7837924361228943,
      "learning_rate": 9.98075187969925e-06,
      "loss": 0.0817,
      "mean_token_accuracy": 0.9684027791023254,
      "num_tokens": 2303765.0,
      "step": 25
    },
    {
      "epoch": 0.027777777777777776,
      "grad_norm": 0.8705114722251892,
      "learning_rate": 1.2060075187969928e-05,
      "loss": 0.0713,
      "mean_token_accuracy": 0.9748264074325561,
      "num_tokens": 2766067.0,
      "step": 30
    },
    {
      "epoch": 0.032407407407407406,
      "grad_norm": 0.96305912733078,
      "learning_rate": 1.4139398496240603e-05,
      "loss": 0.0685,
      "mean_token_accuracy": 0.9763888955116272,
      "num_tokens": 3228390.0,
      "step": 35
    },
    {
      "epoch": 0.037037037037037035,
      "grad_norm": 0.6377480030059814,
      "learning_rate": 1.4555167719763352e-05,
      "loss": 0.0536,
      "mean_token_accuracy": 0.9795138955116272,
      "num_tokens": 3690524.0,
      "step": 40
    },
    {
      "epoch": 0.041666666666666664,
      "grad_norm": 1.0052807331085205,
      "learning_rate": 1.4554780008074628e-05,
      "loss": 0.0456,
      "mean_token_accuracy": 0.9828125238418579,
      "num_tokens": 4153822.0,
      "step": 45
    },
    {
      "epoch": 0.046296296296296294,
      "grad_norm": 0.809347927570343,
      "learning_rate": 1.455409407911408e-05,
      "loss": 0.043,
      "mean_token_accuracy": 0.9848958373069763,
      "num_tokens": 4614618.0,
      "step": 50
    },
    {
      "epoch": 0.05092592592592592,
      "grad_norm": 0.6592745780944824,
      "learning_rate": 1.4553109970361657e-05,
      "loss": 0.0371,
      "mean_token_accuracy": 0.9871527791023255,
      "num_tokens": 5081882.0,
      "step": 55
    },
    {
      "epoch": 0.05555555555555555,
      "grad_norm": 0.8141903877258301,
      "learning_rate": 1.4551827735590205e-05,
      "loss": 0.0436,
      "mean_token_accuracy": 0.9833333373069764,
      "num_tokens": 5545243.0,
      "step": 60
    },
    {
      "epoch": 0.06018518518518518,
      "grad_norm": 0.4772146940231323,
      "learning_rate": 1.455024744486252e-05,
      "loss": 0.045,
      "mean_token_accuracy": 0.9826389074325561,
      "num_tokens": 6005131.0,
      "step": 65
    },
    {
      "epoch": 0.06481481481481481,
      "grad_norm": 0.6070070266723633,
      "learning_rate": 1.4548369184527518e-05,
      "loss": 0.0448,
      "mean_token_accuracy": 0.9829861164093018,
      "num_tokens": 6466053.0,
      "step": 70
    },
    {
      "epoch": 0.06944444444444445,
      "grad_norm": 0.4639335572719574,
      "learning_rate": 1.4546193057215521e-05,
      "loss": 0.0413,
      "mean_token_accuracy": 0.9847222208976746,
      "num_tokens": 6925273.0,
      "step": 75
    },
    {
      "epoch": 0.07407407407407407,
      "grad_norm": 0.42907801270484924,
      "learning_rate": 1.4543719181832652e-05,
      "loss": 0.0362,
      "mean_token_accuracy": 0.9866319179534913,
      "num_tokens": 7387597.0,
      "step": 80
    },
    {
      "epoch": 0.0787037037037037,
      "grad_norm": 0.6013121008872986,
      "learning_rate": 1.4540947693554328e-05,
      "loss": 0.0355,
      "mean_token_accuracy": 0.9855902671813965,
      "num_tokens": 7851122.0,
      "step": 85
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.4883110225200653,
      "learning_rate": 1.4537878743817883e-05,
      "loss": 0.0387,
      "mean_token_accuracy": 0.9855902791023254,
      "num_tokens": 8312588.0,
      "step": 90
    },
    {
      "epoch": 0.08796296296296297,
      "grad_norm": 0.45554330945014954,
      "learning_rate": 1.453451250031429e-05,
      "loss": 0.0368,
      "mean_token_accuracy": 0.9848958492279053,
      "num_tokens": 8770713.0,
      "step": 95
    },
    {
      "epoch": 0.09259259259259259,
      "grad_norm": 0.47422030568122864,
      "learning_rate": 1.4530849146978997e-05,
      "loss": 0.0396,
      "mean_token_accuracy": 0.9859375,
      "num_tokens": 9233053.0,
      "step": 100
    },
    {
      "epoch": 0.09722222222222222,
      "grad_norm": 0.39943304657936096,
      "learning_rate": 1.4526888883981882e-05,
      "loss": 0.0459,
      "mean_token_accuracy": 0.9833333492279053,
      "num_tokens": 9692467.0,
      "step": 105
    },
    {
      "epoch": 0.10185185185185185,
      "grad_norm": 0.29724475741386414,
      "learning_rate": 1.45226319277163e-05,
      "loss": 0.0374,
      "mean_token_accuracy": 0.9862847208976746,
      "num_tokens": 10154258.0,
      "step": 110
    },
    {
      "epoch": 0.10648148148148148,
      "grad_norm": 0.37697604298591614,
      "learning_rate": 1.4518078510787285e-05,
      "loss": 0.033,
      "mean_token_accuracy": 0.9878472328186035,
      "num_tokens": 10614925.0,
      "step": 115
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 0.36144766211509705,
      "learning_rate": 1.4513228881998816e-05,
      "loss": 0.0283,
      "mean_token_accuracy": 0.9885416626930237,
      "num_tokens": 11077558.0,
      "step": 120
    },
    {
      "epoch": 0.11574074074074074,
      "grad_norm": 0.534447968006134,
      "learning_rate": 1.4508083306340234e-05,
      "loss": 0.032,
      "mean_token_accuracy": 0.986631965637207,
      "num_tokens": 11538490.0,
      "step": 125
    },
    {
      "epoch": 0.12037037037037036,
      "grad_norm": 0.5760793089866638,
      "learning_rate": 1.450264206497176e-05,
      "loss": 0.0291,
      "mean_token_accuracy": 0.9899305582046509,
      "num_tokens": 11996473.0,
      "step": 130
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.44985342025756836,
      "learning_rate": 1.449690545520913e-05,
      "loss": 0.0354,
      "mean_token_accuracy": 0.9871527791023255,
      "num_tokens": 12454602.0,
      "step": 135
    },
    {
      "epoch": 0.12962962962962962,
      "grad_norm": 0.2975083887577057,
      "learning_rate": 1.4490873790507355e-05,
      "loss": 0.0313,
      "mean_token_accuracy": 0.9881944537162781,
      "num_tokens": 12913475.0,
      "step": 140
    },
    {
      "epoch": 0.13425925925925927,
      "grad_norm": 0.3903728127479553,
      "learning_rate": 1.4484547400443593e-05,
      "loss": 0.0337,
      "mean_token_accuracy": 0.9880208253860474,
      "num_tokens": 13373358.0,
      "step": 145
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 0.639337956905365,
      "learning_rate": 1.4477926630699127e-05,
      "loss": 0.0326,
      "mean_token_accuracy": 0.9869791746139527,
      "num_tokens": 13833882.0,
      "step": 150
    },
    {
      "epoch": 0.14351851851851852,
      "grad_norm": 0.440522164106369,
      "learning_rate": 1.4471011843040499e-05,
      "loss": 0.0283,
      "mean_token_accuracy": 0.989062511920929,
      "num_tokens": 14294543.0,
      "step": 155
    },
    {
      "epoch": 0.14814814814814814,
      "grad_norm": 0.4374408721923828,
      "learning_rate": 1.446380341529972e-05,
      "loss": 0.0305,
      "mean_token_accuracy": 0.9878472208976745,
      "num_tokens": 14754981.0,
      "step": 160
    },
    {
      "epoch": 0.1527777777777778,
      "grad_norm": 0.42126211524009705,
      "learning_rate": 1.4456301741353646e-05,
      "loss": 0.0299,
      "mean_token_accuracy": 0.9892360925674438,
      "num_tokens": 15216018.0,
      "step": 165
    },
    {
      "epoch": 0.1574074074074074,
      "grad_norm": 0.4279370903968811,
      "learning_rate": 1.4448507231102438e-05,
      "loss": 0.0277,
      "mean_token_accuracy": 0.9895833373069763,
      "num_tokens": 15680195.0,
      "step": 170
    },
    {
      "epoch": 0.16203703703703703,
      "grad_norm": 0.5686227083206177,
      "learning_rate": 1.4440420310447175e-05,
      "loss": 0.0291,
      "mean_token_accuracy": 0.9892361283302307,
      "num_tokens": 16142235.0,
      "step": 175
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.27839693427085876,
      "learning_rate": 1.4432041421266582e-05,
      "loss": 0.0285,
      "mean_token_accuracy": 0.9894097447395325,
      "num_tokens": 16604297.0,
      "step": 180
    },
    {
      "epoch": 0.1712962962962963,
      "grad_norm": 0.27903345227241516,
      "learning_rate": 1.4423371021392878e-05,
      "loss": 0.0279,
      "mean_token_accuracy": 0.9885416746139526,
      "num_tokens": 17063789.0,
      "step": 185
    },
    {
      "epoch": 0.17592592592592593,
      "grad_norm": 0.39598026871681213,
      "learning_rate": 1.4414409584586765e-05,
      "loss": 0.0302,
      "mean_token_accuracy": 0.9878472089767456,
      "num_tokens": 17527365.0,
      "step": 190
    },
    {
      "epoch": 0.18055555555555555,
      "grad_norm": 0.36333873867988586,
      "learning_rate": 1.4405157600511548e-05,
      "loss": 0.0299,
      "mean_token_accuracy": 0.9888888835906983,
      "num_tokens": 17990157.0,
      "step": 195
    },
    {
      "epoch": 0.18518518518518517,
      "grad_norm": 0.3383961319923401,
      "learning_rate": 1.4395615574706358e-05,
      "loss": 0.0275,
      "mean_token_accuracy": 0.9887152791023255,
      "num_tokens": 18453485.0,
      "step": 200
    },
    {
      "epoch": 0.18981481481481483,
      "grad_norm": 0.27447694540023804,
      "learning_rate": 1.4385784028558558e-05,
      "loss": 0.0277,
      "mean_token_accuracy": 0.9885416626930237,
      "num_tokens": 18913478.0,
      "step": 205
    },
    {
      "epoch": 0.19444444444444445,
      "grad_norm": 0.45147305727005005,
      "learning_rate": 1.4375663499275225e-05,
      "loss": 0.0296,
      "mean_token_accuracy": 0.9878472208976745,
      "num_tokens": 19375959.0,
      "step": 210
    },
    {
      "epoch": 0.19907407407407407,
      "grad_norm": 0.32271087169647217,
      "learning_rate": 1.4365254539853818e-05,
      "loss": 0.0297,
      "mean_token_accuracy": 0.9887152791023255,
      "num_tokens": 19835198.0,
      "step": 215
    },
    {
      "epoch": 0.2037037037037037,
      "grad_norm": 0.4061264991760254,
      "learning_rate": 1.435455771905195e-05,
      "loss": 0.027,
      "mean_token_accuracy": 0.9899305582046509,
      "num_tokens": 20296629.0,
      "step": 220
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 0.42220982909202576,
      "learning_rate": 1.4343573621356315e-05,
      "loss": 0.0307,
      "mean_token_accuracy": 0.9890625,
      "num_tokens": 20756961.0,
      "step": 225
    },
    {
      "epoch": 0.21296296296296297,
      "grad_norm": 0.263630211353302,
      "learning_rate": 1.4332302846950745e-05,
      "loss": 0.0299,
      "mean_token_accuracy": 0.9894097208976745,
      "num_tokens": 21218202.0,
      "step": 230
    },
    {
      "epoch": 0.2175925925925926,
      "grad_norm": 0.35219112038612366,
      "learning_rate": 1.4320746011683426e-05,
      "loss": 0.0206,
      "mean_token_accuracy": 0.9911458492279053,
      "num_tokens": 21680433.0,
      "step": 235
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.3776715397834778,
      "learning_rate": 1.4308903747033241e-05,
      "loss": 0.0285,
      "mean_token_accuracy": 0.9897569417953491,
      "num_tokens": 22144020.0,
      "step": 240
    },
    {
      "epoch": 0.22685185185185186,
      "grad_norm": 0.30106329917907715,
      "learning_rate": 1.429677670007526e-05,
      "loss": 0.0224,
      "mean_token_accuracy": 0.9916666746139526,
      "num_tokens": 22605744.0,
      "step": 245
    },
    {
      "epoch": 0.23148148148148148,
      "grad_norm": 0.4496817886829376,
      "learning_rate": 1.4284365533445395e-05,
      "loss": 0.025,
      "mean_token_accuracy": 0.9911458373069764,
      "num_tokens": 23067418.0,
      "step": 250
    },
    {
      "epoch": 0.2361111111111111,
      "grad_norm": 0.22386600077152252,
      "learning_rate": 1.427167092530418e-05,
      "loss": 0.0265,
      "mean_token_accuracy": 0.9907986402511597,
      "num_tokens": 23528890.0,
      "step": 255
    },
    {
      "epoch": 0.24074074074074073,
      "grad_norm": 0.3611874580383301,
      "learning_rate": 1.4258693569299725e-05,
      "loss": 0.0281,
      "mean_token_accuracy": 0.9894097328186036,
      "num_tokens": 23989710.0,
      "step": 260
    },
    {
      "epoch": 0.24537037037037038,
      "grad_norm": 0.43193748593330383,
      "learning_rate": 1.4245434174529815e-05,
      "loss": 0.0296,
      "mean_token_accuracy": 0.9878472089767456,
      "num_tokens": 24449509.0,
      "step": 265
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.30856630206108093,
      "learning_rate": 1.423189346550315e-05,
      "loss": 0.03,
      "mean_token_accuracy": 0.9883680582046509,
      "num_tokens": 24909615.0,
      "step": 270
    },
    {
      "epoch": 0.25462962962962965,
      "grad_norm": 0.5478079319000244,
      "learning_rate": 1.4218072182099782e-05,
      "loss": 0.0254,
      "mean_token_accuracy": 0.9901041746139526,
      "num_tokens": 25373094.0,
      "step": 275
    },
    {
      "epoch": 0.25925925925925924,
      "grad_norm": 0.3937035799026489,
      "learning_rate": 1.4203971079530656e-05,
      "loss": 0.0241,
      "mean_token_accuracy": 0.989756953716278,
      "num_tokens": 25833929.0,
      "step": 280
    },
    {
      "epoch": 0.2638888888888889,
      "grad_norm": 0.4984318017959595,
      "learning_rate": 1.4189590928296375e-05,
      "loss": 0.0281,
      "mean_token_accuracy": 0.9901041626930237,
      "num_tokens": 26293705.0,
      "step": 285
    },
    {
      "epoch": 0.26851851851851855,
      "grad_norm": 0.3481650948524475,
      "learning_rate": 1.4174932514145072e-05,
      "loss": 0.0265,
      "mean_token_accuracy": 0.9899305582046509,
      "num_tokens": 26752738.0,
      "step": 290
    },
    {
      "epoch": 0.27314814814814814,
      "grad_norm": 0.3228101432323456,
      "learning_rate": 1.4159996638029501e-05,
      "loss": 0.0299,
      "mean_token_accuracy": 0.9887152910232544,
      "num_tokens": 27211937.0,
      "step": 295
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 0.24106641113758087,
      "learning_rate": 1.4144784116063246e-05,
      "loss": 0.0293,
      "mean_token_accuracy": 0.9888888835906983,
      "num_tokens": 27674565.0,
      "step": 300
    },
    {
      "epoch": 0.2824074074074074,
      "grad_norm": 0.16628754138946533,
      "learning_rate": 1.4129295779476157e-05,
      "loss": 0.026,
      "mean_token_accuracy": 0.9887152791023255,
      "num_tokens": 28137015.0,
      "step": 305
    },
    {
      "epoch": 0.28703703703703703,
      "grad_norm": 0.2680685520172119,
      "learning_rate": 1.41135324745689e-05,
      "loss": 0.0245,
      "mean_token_accuracy": 0.9914930462837219,
      "num_tokens": 28599562.0,
      "step": 310
    },
    {
      "epoch": 0.2916666666666667,
      "grad_norm": 0.21587631106376648,
      "learning_rate": 1.4097495062666736e-05,
      "loss": 0.0245,
      "mean_token_accuracy": 0.9907986164093018,
      "num_tokens": 29061733.0,
      "step": 315
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 0.2370392084121704,
      "learning_rate": 1.4081184420072449e-05,
      "loss": 0.0204,
      "mean_token_accuracy": 0.9939236164093017,
      "num_tokens": 29525030.0,
      "step": 320
    },
    {
      "epoch": 0.30092592592592593,
      "grad_norm": 0.6963658332824707,
      "learning_rate": 1.4064601438018467e-05,
      "loss": 0.0233,
      "mean_token_accuracy": 0.9907986283302307,
      "num_tokens": 29984716.0,
      "step": 325
    },
    {
      "epoch": 0.3055555555555556,
      "grad_norm": 0.47941067814826965,
      "learning_rate": 1.4047747022618158e-05,
      "loss": 0.0254,
      "mean_token_accuracy": 0.990625011920929,
      "num_tokens": 30444022.0,
      "step": 330
    },
    {
      "epoch": 0.3101851851851852,
      "grad_norm": 0.3066806495189667,
      "learning_rate": 1.4030622094816324e-05,
      "loss": 0.0305,
      "mean_token_accuracy": 0.9880208253860474,
      "num_tokens": 30902071.0,
      "step": 335
    },
    {
      "epoch": 0.3148148148148148,
      "grad_norm": 0.5959760546684265,
      "learning_rate": 1.401322759033888e-05,
      "loss": 0.0285,
      "mean_token_accuracy": 0.989062488079071,
      "num_tokens": 31359469.0,
      "step": 340
    },
    {
      "epoch": 0.3194444444444444,
      "grad_norm": 0.37421202659606934,
      "learning_rate": 1.3995564459641721e-05,
      "loss": 0.0304,
      "mean_token_accuracy": 0.9885416746139526,
      "num_tokens": 31818587.0,
      "step": 345
    },
    {
      "epoch": 0.32407407407407407,
      "grad_norm": 0.2207397222518921,
      "learning_rate": 1.3977633667858792e-05,
      "loss": 0.0234,
      "mean_token_accuracy": 0.9909722328186035,
      "num_tokens": 32278929.0,
      "step": 350
    },
    {
      "epoch": 0.3287037037037037,
      "grad_norm": 0.5324211120605469,
      "learning_rate": 1.3959436194749348e-05,
      "loss": 0.0236,
      "mean_token_accuracy": 0.9916666865348815,
      "num_tokens": 32741545.0,
      "step": 355
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.2752709984779358,
      "learning_rate": 1.3940973034644425e-05,
      "loss": 0.0235,
      "mean_token_accuracy": 0.9920138955116272,
      "num_tokens": 33204799.0,
      "step": 360
    },
    {
      "epoch": 0.33796296296296297,
      "grad_norm": 0.43308961391448975,
      "learning_rate": 1.39222451963925e-05,
      "loss": 0.0252,
      "mean_token_accuracy": 0.9902777791023254,
      "num_tokens": 33664016.0,
      "step": 365
    },
    {
      "epoch": 0.3425925925925926,
      "grad_norm": 0.18644089996814728,
      "learning_rate": 1.3903253703304376e-05,
      "loss": 0.0238,
      "mean_token_accuracy": 0.9918402910232544,
      "num_tokens": 34124419.0,
      "step": 370
    },
    {
      "epoch": 0.3472222222222222,
      "grad_norm": 0.3096047639846802,
      "learning_rate": 1.388399959309725e-05,
      "loss": 0.0294,
      "mean_token_accuracy": 0.9897569298744202,
      "num_tokens": 34586289.0,
      "step": 375
    },
    {
      "epoch": 0.35185185185185186,
      "grad_norm": 0.21771998703479767,
      "learning_rate": 1.3864483917838042e-05,
      "loss": 0.0284,
      "mean_token_accuracy": 0.9899305582046509,
      "num_tokens": 35046458.0,
      "step": 380
    },
    {
      "epoch": 0.35648148148148145,
      "grad_norm": 0.19050583243370056,
      "learning_rate": 1.384470774388587e-05,
      "loss": 0.0248,
      "mean_token_accuracy": 0.9894097328186036,
      "num_tokens": 35506030.0,
      "step": 385
    },
    {
      "epoch": 0.3611111111111111,
      "grad_norm": 0.4737027585506439,
      "learning_rate": 1.3824672151833819e-05,
      "loss": 0.0262,
      "mean_token_accuracy": 0.9894097208976745,
      "num_tokens": 35968659.0,
      "step": 390
    },
    {
      "epoch": 0.36574074074074076,
      "grad_norm": 0.19767142832279205,
      "learning_rate": 1.3804378236449872e-05,
      "loss": 0.0277,
      "mean_token_accuracy": 0.9894097208976745,
      "num_tokens": 36428594.0,
      "step": 395
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 0.38778188824653625,
      "learning_rate": 1.3783827106617095e-05,
      "loss": 0.0279,
      "mean_token_accuracy": 0.988194465637207,
      "num_tokens": 36888791.0,
      "step": 400
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.4036060571670532,
      "learning_rate": 1.3763019885273058e-05,
      "loss": 0.0255,
      "mean_token_accuracy": 0.9899305701255798,
      "num_tokens": 37351215.0,
      "step": 405
    },
    {
      "epoch": 0.37962962962962965,
      "grad_norm": 0.2737887501716614,
      "learning_rate": 1.3741957709348464e-05,
      "loss": 0.0274,
      "mean_token_accuracy": 0.9899305462837219,
      "num_tokens": 37814308.0,
      "step": 410
    },
    {
      "epoch": 0.38425925925925924,
      "grad_norm": 0.29210758209228516,
      "learning_rate": 1.3720641729705027e-05,
      "loss": 0.0231,
      "mean_token_accuracy": 0.9909722328186035,
      "num_tokens": 38277748.0,
      "step": 415
    },
    {
      "epoch": 0.3888888888888889,
      "grad_norm": 0.2634587585926056,
      "learning_rate": 1.3699073111072596e-05,
      "loss": 0.0217,
      "mean_token_accuracy": 0.9920139074325561,
      "num_tokens": 38741634.0,
      "step": 420
    },
    {
      "epoch": 0.39351851851851855,
      "grad_norm": 0.38451331853866577,
      "learning_rate": 1.3677253031985506e-05,
      "loss": 0.024,
      "mean_token_accuracy": 0.991319453716278,
      "num_tokens": 39200312.0,
      "step": 425
    },
    {
      "epoch": 0.39814814814814814,
      "grad_norm": 0.30776146054267883,
      "learning_rate": 1.3655182684718178e-05,
      "loss": 0.0214,
      "mean_token_accuracy": 0.9918402910232544,
      "num_tokens": 39661895.0,
      "step": 430
    },
    {
      "epoch": 0.4027777777777778,
      "grad_norm": 0.4174097776412964,
      "learning_rate": 1.3632863275219987e-05,
      "loss": 0.0233,
      "mean_token_accuracy": 0.9916666626930237,
      "num_tokens": 40121380.0,
      "step": 435
    },
    {
      "epoch": 0.4074074074074074,
      "grad_norm": 0.45231279730796814,
      "learning_rate": 1.3610296023049342e-05,
      "loss": 0.0248,
      "mean_token_accuracy": 0.9918402910232544,
      "num_tokens": 40582106.0,
      "step": 440
    },
    {
      "epoch": 0.41203703703703703,
      "grad_norm": 0.21521063148975372,
      "learning_rate": 1.3587482161307078e-05,
      "loss": 0.0226,
      "mean_token_accuracy": 0.9918403029441833,
      "num_tokens": 41040490.0,
      "step": 445
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.19439320266246796,
      "learning_rate": 1.3564422936569055e-05,
      "loss": 0.0248,
      "mean_token_accuracy": 0.9911458492279053,
      "num_tokens": 41502281.0,
      "step": 450
    },
    {
      "epoch": 0.4212962962962963,
      "grad_norm": 0.36206671595573425,
      "learning_rate": 1.354111960881805e-05,
      "loss": 0.0307,
      "mean_token_accuracy": 0.9901041626930237,
      "num_tokens": 41962684.0,
      "step": 455
    },
    {
      "epoch": 0.42592592592592593,
      "grad_norm": 0.34299176931381226,
      "learning_rate": 1.3517573451374918e-05,
      "loss": 0.022,
      "mean_token_accuracy": 0.9914930701255799,
      "num_tokens": 42425390.0,
      "step": 460
    },
    {
      "epoch": 0.4305555555555556,
      "grad_norm": 0.3069056570529938,
      "learning_rate": 1.349378575082901e-05,
      "loss": 0.0213,
      "mean_token_accuracy": 0.9918403029441833,
      "num_tokens": 42887841.0,
      "step": 465
    },
    {
      "epoch": 0.4351851851851852,
      "grad_norm": 0.34767967462539673,
      "learning_rate": 1.346975780696786e-05,
      "loss": 0.0235,
      "mean_token_accuracy": 0.9914930462837219,
      "num_tokens": 43350012.0,
      "step": 470
    },
    {
      "epoch": 0.4398148148148148,
      "grad_norm": 0.22474323213100433,
      "learning_rate": 1.3445490932706194e-05,
      "loss": 0.0224,
      "mean_token_accuracy": 0.9927083373069763,
      "num_tokens": 43815142.0,
      "step": 475
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.2890940308570862,
      "learning_rate": 1.3420986454014157e-05,
      "loss": 0.0251,
      "mean_token_accuracy": 0.9902777910232544,
      "num_tokens": 44273281.0,
      "step": 480
    },
    {
      "epoch": 0.44907407407407407,
      "grad_norm": 0.2148568481206894,
      "learning_rate": 1.3396245709844881e-05,
      "loss": 0.0205,
      "mean_token_accuracy": 0.9918403029441833,
      "num_tokens": 44734476.0,
      "step": 485
    },
    {
      "epoch": 0.4537037037037037,
      "grad_norm": 0.42016416788101196,
      "learning_rate": 1.337127005206132e-05,
      "loss": 0.0231,
      "mean_token_accuracy": 0.9909722328186035,
      "num_tokens": 45198706.0,
      "step": 490
    },
    {
      "epoch": 0.4583333333333333,
      "grad_norm": 0.30800846219062805,
      "learning_rate": 1.3346060845362378e-05,
      "loss": 0.028,
      "mean_token_accuracy": 0.9881944417953491,
      "num_tokens": 45658711.0,
      "step": 495
    },
    {
      "epoch": 0.46296296296296297,
      "grad_norm": 0.14300820231437683,
      "learning_rate": 1.3320619467208342e-05,
      "loss": 0.0237,
      "mean_token_accuracy": 0.9916666746139526,
      "num_tokens": 46119675.0,
      "step": 500
    },
    {
      "epoch": 0.4675925925925926,
      "grad_norm": 0.13899363577365875,
      "learning_rate": 1.3294947307745615e-05,
      "loss": 0.0203,
      "mean_token_accuracy": 0.991319453716278,
      "num_tokens": 46578591.0,
      "step": 505
    },
    {
      "epoch": 0.4722222222222222,
      "grad_norm": 0.19077682495117188,
      "learning_rate": 1.3269045769730765e-05,
      "loss": 0.0195,
      "mean_token_accuracy": 0.9925347447395325,
      "num_tokens": 47041069.0,
      "step": 510
    },
    {
      "epoch": 0.47685185185185186,
      "grad_norm": 0.26572269201278687,
      "learning_rate": 1.324291626845386e-05,
      "loss": 0.0198,
      "mean_token_accuracy": 0.9927083492279053,
      "num_tokens": 47504342.0,
      "step": 515
    },
    {
      "epoch": 0.48148148148148145,
      "grad_norm": 0.2430904060602188,
      "learning_rate": 1.321656023166116e-05,
      "loss": 0.0275,
      "mean_token_accuracy": 0.9902778029441833,
      "num_tokens": 47964322.0,
      "step": 520
    },
    {
      "epoch": 0.4861111111111111,
      "grad_norm": 0.20157483220100403,
      "learning_rate": 1.318997909947707e-05,
      "loss": 0.0256,
      "mean_token_accuracy": 0.9911458492279053,
      "num_tokens": 48425703.0,
      "step": 525
    },
    {
      "epoch": 0.49074074074074076,
      "grad_norm": 0.35652822256088257,
      "learning_rate": 1.316317432432548e-05,
      "loss": 0.0212,
      "mean_token_accuracy": 0.9914930701255799,
      "num_tokens": 48886298.0,
      "step": 530
    },
    {
      "epoch": 0.49537037037037035,
      "grad_norm": 0.3549620509147644,
      "learning_rate": 1.3136147370850387e-05,
      "loss": 0.0279,
      "mean_token_accuracy": 0.9911458611488342,
      "num_tokens": 49349292.0,
      "step": 535
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.2592863142490387,
      "learning_rate": 1.3108899715835875e-05,
      "loss": 0.0267,
      "mean_token_accuracy": 0.9887152791023255,
      "num_tokens": 49811354.0,
      "step": 540
    },
    {
      "epoch": 0.5046296296296297,
      "grad_norm": 0.20696547627449036,
      "learning_rate": 1.3081432848125409e-05,
      "loss": 0.0213,
      "mean_token_accuracy": 0.9925347447395325,
      "num_tokens": 50272656.0,
      "step": 545
    },
    {
      "epoch": 0.5092592592592593,
      "grad_norm": 0.29152804613113403,
      "learning_rate": 1.305374826854049e-05,
      "loss": 0.0221,
      "mean_token_accuracy": 0.9920138835906982,
      "num_tokens": 50731855.0,
      "step": 550
    },
    {
      "epoch": 0.5138888888888888,
      "grad_norm": 0.2276608943939209,
      "learning_rate": 1.302584748979865e-05,
      "loss": 0.0289,
      "mean_token_accuracy": 0.9894097328186036,
      "num_tokens": 51191643.0,
      "step": 555
    },
    {
      "epoch": 0.5185185185185185,
      "grad_norm": 0.21999682486057281,
      "learning_rate": 1.29977320364308e-05,
      "loss": 0.0235,
      "mean_token_accuracy": 0.9904513955116272,
      "num_tokens": 51652573.0,
      "step": 560
    },
    {
      "epoch": 0.5231481481481481,
      "grad_norm": 0.24313662946224213,
      "learning_rate": 1.2969403444697913e-05,
      "loss": 0.0189,
      "mean_token_accuracy": 0.9939236402511596,
      "num_tokens": 52116690.0,
      "step": 565
    },
    {
      "epoch": 0.5277777777777778,
      "grad_norm": 0.24149328470230103,
      "learning_rate": 1.2940863262507095e-05,
      "loss": 0.0285,
      "mean_token_accuracy": 0.9897569298744202,
      "num_tokens": 52576193.0,
      "step": 570
    },
    {
      "epoch": 0.5324074074074074,
      "grad_norm": 0.21016943454742432,
      "learning_rate": 1.291211304932699e-05,
      "loss": 0.0198,
      "mean_token_accuracy": 0.9932291865348816,
      "num_tokens": 53036452.0,
      "step": 575
    },
    {
      "epoch": 0.5370370370370371,
      "grad_norm": 0.20628230273723602,
      "learning_rate": 1.2883154376102591e-05,
      "loss": 0.0232,
      "mean_token_accuracy": 0.9907986044883728,
      "num_tokens": 53498982.0,
      "step": 580
    },
    {
      "epoch": 0.5416666666666666,
      "grad_norm": 0.20213210582733154,
      "learning_rate": 1.2853988825169387e-05,
      "loss": 0.0237,
      "mean_token_accuracy": 0.991319453716278,
      "num_tokens": 53959847.0,
      "step": 585
    },
    {
      "epoch": 0.5462962962962963,
      "grad_norm": 0.26332107186317444,
      "learning_rate": 1.2824617990166898e-05,
      "loss": 0.0241,
      "mean_token_accuracy": 0.990625,
      "num_tokens": 54422565.0,
      "step": 590
    },
    {
      "epoch": 0.5509259259259259,
      "grad_norm": 0.2062518149614334,
      "learning_rate": 1.279504347595161e-05,
      "loss": 0.0205,
      "mean_token_accuracy": 0.9918402671813965,
      "num_tokens": 54884710.0,
      "step": 595
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.14048855006694794,
      "learning_rate": 1.2765266898509277e-05,
      "loss": 0.0227,
      "mean_token_accuracy": 0.990625,
      "num_tokens": 55345608.0,
      "step": 600
    },
    {
      "epoch": 0.5601851851851852,
      "grad_norm": 0.20836009085178375,
      "learning_rate": 1.2735289884866619e-05,
      "loss": 0.025,
      "mean_token_accuracy": 0.9911458373069764,
      "num_tokens": 55808782.0,
      "step": 605
    },
    {
      "epoch": 0.5648148148148148,
      "grad_norm": 0.24805429577827454,
      "learning_rate": 1.2705114073002424e-05,
      "loss": 0.0199,
      "mean_token_accuracy": 0.9923611283302307,
      "num_tokens": 56269602.0,
      "step": 610
    },
    {
      "epoch": 0.5694444444444444,
      "grad_norm": 0.2596973776817322,
      "learning_rate": 1.2674741111758047e-05,
      "loss": 0.0185,
      "mean_token_accuracy": 0.9916666746139526,
      "num_tokens": 56729421.0,
      "step": 615
    },
    {
      "epoch": 0.5740740740740741,
      "grad_norm": 0.4484080672264099,
      "learning_rate": 1.264417266074731e-05,
      "loss": 0.024,
      "mean_token_accuracy": 0.9914930582046508,
      "num_tokens": 57189796.0,
      "step": 620
    },
    {
      "epoch": 0.5787037037037037,
      "grad_norm": 0.18526935577392578,
      "learning_rate": 1.2613410390265825e-05,
      "loss": 0.0189,
      "mean_token_accuracy": 0.9930555820465088,
      "num_tokens": 57652021.0,
      "step": 625
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.3614274859428406,
      "learning_rate": 1.2582455981199726e-05,
      "loss": 0.0209,
      "mean_token_accuracy": 0.9920138835906982,
      "num_tokens": 58116218.0,
      "step": 630
    },
    {
      "epoch": 0.5879629629629629,
      "grad_norm": 0.16941134631633759,
      "learning_rate": 1.255131112493382e-05,
      "loss": 0.0212,
      "mean_token_accuracy": 0.9928819537162781,
      "num_tokens": 58576965.0,
      "step": 635
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 0.6090588569641113,
      "learning_rate": 1.251997752325917e-05,
      "loss": 0.0221,
      "mean_token_accuracy": 0.9920138955116272,
      "num_tokens": 59038218.0,
      "step": 640
    },
    {
      "epoch": 0.5972222222222222,
      "grad_norm": 0.43704843521118164,
      "learning_rate": 1.2488456888280105e-05,
      "loss": 0.0212,
      "mean_token_accuracy": 0.9916666865348815,
      "num_tokens": 59500213.0,
      "step": 645
    },
    {
      "epoch": 0.6018518518518519,
      "grad_norm": 0.33717143535614014,
      "learning_rate": 1.2456750942320677e-05,
      "loss": 0.0202,
      "mean_token_accuracy": 0.9914930820465088,
      "num_tokens": 59964684.0,
      "step": 650
    },
    {
      "epoch": 0.6064814814814815,
      "grad_norm": 0.3672013580799103,
      "learning_rate": 1.2424861417830537e-05,
      "loss": 0.0226,
      "mean_token_accuracy": 0.9914930582046508,
      "num_tokens": 60425807.0,
      "step": 655
    },
    {
      "epoch": 0.6111111111111112,
      "grad_norm": 0.24480049312114716,
      "learning_rate": 1.2392790057290287e-05,
      "loss": 0.0241,
      "mean_token_accuracy": 0.9911458253860473,
      "num_tokens": 60884792.0,
      "step": 660
    },
    {
      "epoch": 0.6157407407407407,
      "grad_norm": 0.18146894872188568,
      "learning_rate": 1.2360538613116256e-05,
      "loss": 0.03,
      "mean_token_accuracy": 0.9871527791023255,
      "num_tokens": 61343755.0,
      "step": 665
    },
    {
      "epoch": 0.6203703703703703,
      "grad_norm": 0.22168032824993134,
      "learning_rate": 1.2328108847564752e-05,
      "loss": 0.0239,
      "mean_token_accuracy": 0.9920138955116272,
      "num_tokens": 61802407.0,
      "step": 670
    }
  ],
  "logging_steps": 5,
  "max_steps": 2160,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.258746787334193e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}