{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.99972616910579,
  "eval_steps": 500,
  "global_step": 24648,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06,
      "grad_norm": 17.19582176208496,
      "learning_rate": 1.0101419878296147e-05,
      "loss": 3.88,
      "step": 500
    },
    {
      "epoch": 0.12,
      "grad_norm": 22.53339385986328,
      "learning_rate": 2.0223123732251522e-05,
      "loss": 1.9753,
      "step": 1000
    },
    {
      "epoch": 0.18,
      "grad_norm": 20.645118713378906,
      "learning_rate": 3.0344827586206897e-05,
      "loss": 1.4051,
      "step": 1500
    },
    {
      "epoch": 0.24,
      "grad_norm": 11.275988578796387,
      "learning_rate": 4.0486815415821506e-05,
      "loss": 1.2663,
      "step": 2000
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.870928764343262,
      "learning_rate": 4.993012667357887e-05,
      "loss": 1.2106,
      "step": 2500
    },
    {
      "epoch": 0.37,
      "grad_norm": 13.5437650680542,
      "learning_rate": 4.880313753775414e-05,
      "loss": 1.179,
      "step": 3000
    },
    {
      "epoch": 0.43,
      "grad_norm": 14.121590614318848,
      "learning_rate": 4.767614840192941e-05,
      "loss": 1.1044,
      "step": 3500
    },
    {
      "epoch": 0.49,
      "grad_norm": 11.260817527770996,
      "learning_rate": 4.6549159266104676e-05,
      "loss": 1.043,
      "step": 4000
    },
    {
      "epoch": 0.55,
      "grad_norm": 15.443611145019531,
      "learning_rate": 4.542217013027995e-05,
      "loss": 1.0212,
      "step": 4500
    },
    {
      "epoch": 0.61,
      "grad_norm": 14.637772560119629,
      "learning_rate": 4.429518099445522e-05,
      "loss": 0.9764,
      "step": 5000
    },
    {
      "epoch": 0.67,
      "grad_norm": 18.14031410217285,
      "learning_rate": 4.316819185863049e-05,
      "loss": 0.9848,
      "step": 5500
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.293859481811523,
      "learning_rate": 4.204120272280576e-05,
      "loss": 0.9378,
      "step": 6000
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.198644638061523,
      "learning_rate": 4.0916467565252675e-05,
      "loss": 0.9363,
      "step": 6500
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.705594062805176,
      "learning_rate": 3.978947842942794e-05,
      "loss": 0.9482,
      "step": 7000
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.531152725219727,
      "learning_rate": 3.866248929360321e-05,
      "loss": 0.9033,
      "step": 7500
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.644086837768555,
      "learning_rate": 3.753550015777848e-05,
      "loss": 0.9042,
      "step": 8000
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.761592864990234,
      "learning_rate": 3.640851102195375e-05,
      "loss": 0.7736,
      "step": 8500
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.461353302001953,
      "learning_rate": 3.528152188612902e-05,
      "loss": 0.6728,
      "step": 9000
    },
    {
      "epoch": 1.16,
      "grad_norm": 7.326571941375732,
      "learning_rate": 3.415453275030429e-05,
      "loss": 0.6811,
      "step": 9500
    },
    {
      "epoch": 1.22,
      "grad_norm": 7.815227031707764,
      "learning_rate": 3.3027543614479556e-05,
      "loss": 0.6769,
      "step": 10000
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.3690824508667,
      "learning_rate": 3.190055447865483e-05,
      "loss": 0.6725,
      "step": 10500
    },
    {
      "epoch": 1.34,
      "grad_norm": 16.308076858520508,
      "learning_rate": 3.077581932110174e-05,
      "loss": 0.6596,
      "step": 11000
    },
    {
      "epoch": 1.4,
      "grad_norm": 8.466190338134766,
      "learning_rate": 2.9651084163548663e-05,
      "loss": 0.6683,
      "step": 11500
    },
    {
      "epoch": 1.46,
      "grad_norm": 15.437102317810059,
      "learning_rate": 2.8524095027723934e-05,
      "loss": 0.6908,
      "step": 12000
    },
    {
      "epoch": 1.52,
      "grad_norm": 7.6999406814575195,
      "learning_rate": 2.7397105891899204e-05,
      "loss": 0.6875,
      "step": 12500
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.09485912322998,
      "learning_rate": 2.6270116756074475e-05,
      "loss": 0.6514,
      "step": 13000
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.5618314743042,
      "learning_rate": 2.514312762024974e-05,
      "loss": 0.6452,
      "step": 13500
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.317076683044434,
      "learning_rate": 2.401613848442501e-05,
      "loss": 0.6474,
      "step": 14000
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.310576438903809,
      "learning_rate": 2.288914934860028e-05,
      "loss": 0.645,
      "step": 14500
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.67317008972168,
      "learning_rate": 2.176216021277555e-05,
      "loss": 0.6259,
      "step": 15000
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.449134826660156,
      "learning_rate": 2.0635171076950818e-05,
      "loss": 0.6342,
      "step": 15500
    },
    {
      "epoch": 1.95,
      "grad_norm": 5.870083808898926,
      "learning_rate": 1.9510435919397738e-05,
      "loss": 0.6219,
      "step": 16000
    },
    {
      "epoch": 2.01,
      "grad_norm": 12.512371063232422,
      "learning_rate": 1.838344678357301e-05,
      "loss": 0.6211,
      "step": 16500
    },
    {
      "epoch": 2.07,
      "grad_norm": 7.509356498718262,
      "learning_rate": 1.7256457647748276e-05,
      "loss": 0.416,
      "step": 17000
    },
    {
      "epoch": 2.13,
      "grad_norm": 28.715145111083984,
      "learning_rate": 1.6129468511923546e-05,
      "loss": 0.4208,
      "step": 17500
    },
    {
      "epoch": 2.19,
      "grad_norm": 5.760591983795166,
      "learning_rate": 1.5002479376098815e-05,
      "loss": 0.4239,
      "step": 18000
    },
    {
      "epoch": 2.25,
      "grad_norm": 7.888522624969482,
      "learning_rate": 1.3877744218545735e-05,
      "loss": 0.4135,
      "step": 18500
    },
    {
      "epoch": 2.31,
      "grad_norm": 16.289827346801758,
      "learning_rate": 1.2750755082721002e-05,
      "loss": 0.4387,
      "step": 19000
    },
    {
      "epoch": 2.37,
      "grad_norm": 12.066337585449219,
      "learning_rate": 1.1623765946896273e-05,
      "loss": 0.4299,
      "step": 19500
    },
    {
      "epoch": 2.43,
      "grad_norm": 12.49575424194336,
      "learning_rate": 1.0496776811071542e-05,
      "loss": 0.4255,
      "step": 20000
    },
    {
      "epoch": 2.49,
      "grad_norm": 9.139410972595215,
      "learning_rate": 9.369787675246811e-06,
      "loss": 0.4237,
      "step": 20500
    },
    {
      "epoch": 2.56,
      "grad_norm": 18.495920181274414,
      "learning_rate": 8.24505251769373e-06,
      "loss": 0.4293,
      "step": 21000
    },
    {
      "epoch": 2.62,
      "grad_norm": 22.488752365112305,
      "learning_rate": 7.118063381868999e-06,
      "loss": 0.4201,
      "step": 21500
    },
    {
      "epoch": 2.68,
      "grad_norm": 3.9392380714416504,
      "learning_rate": 5.9910742460442686e-06,
      "loss": 0.3984,
      "step": 22000
    },
    {
      "epoch": 2.74,
      "grad_norm": 5.216261386871338,
      "learning_rate": 4.8640851102195375e-06,
      "loss": 0.4006,
      "step": 22500
    },
    {
      "epoch": 2.8,
      "grad_norm": 8.506802558898926,
      "learning_rate": 3.7393499526664564e-06,
      "loss": 0.4085,
      "step": 23000
    },
    {
      "epoch": 2.86,
      "grad_norm": 15.280343055725098,
      "learning_rate": 2.6123608168417258e-06,
      "loss": 0.4029,
      "step": 23500
    },
    {
      "epoch": 2.92,
      "grad_norm": 22.39061164855957,
      "learning_rate": 1.485371681016995e-06,
      "loss": 0.4135,
      "step": 24000
    },
    {
      "epoch": 2.98,
      "grad_norm": 4.789835453033447,
      "learning_rate": 3.583825451922644e-07,
      "loss": 0.4149,
      "step": 24500
    },
    {
      "epoch": 3.0,
      "step": 24648,
      "total_flos": 3.86452500250624e+16,
      "train_loss": 0.7860646006427233,
      "train_runtime": 13844.0651,
      "train_samples_per_second": 28.489,
      "train_steps_per_second": 1.78
    }
  ],
  "logging_steps": 500,
  "max_steps": 24648,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 5000,
  "total_flos": 3.86452500250624e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}