{
  "best_metric": 3.033102035522461,
  "best_model_checkpoint": "/home/lily/nos6/bloom-adapt/exps_ru/bigscience/bloom-560m_ru_adpt_continual-pretrain-reinit_original-frozen_100_000samples/checkpoint-25000",
  "epoch": 1.462073805485701,
  "global_step": 25000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "learning_rate": 9.8e-05,
      "loss": 8.1452,
      "step": 500
    },
    {
      "epoch": 0.06,
      "learning_rate": 9.6e-05,
      "loss": 6.7637,
      "step": 1000
    },
    {
      "epoch": 0.09,
      "learning_rate": 9.4e-05,
      "loss": 6.1828,
      "step": 1500
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.200000000000001e-05,
      "loss": 5.7689,
      "step": 2000
    },
    {
      "epoch": 0.15,
      "learning_rate": 9e-05,
      "loss": 5.4343,
      "step": 2500
    },
    {
      "epoch": 0.18,
      "learning_rate": 8.800000000000001e-05,
      "loss": 5.1783,
      "step": 3000
    },
    {
      "epoch": 0.2,
      "learning_rate": 8.6e-05,
      "loss": 4.9451,
      "step": 3500
    },
    {
      "epoch": 0.23,
      "learning_rate": 8.4e-05,
      "loss": 4.7916,
      "step": 4000
    },
    {
      "epoch": 0.26,
      "learning_rate": 8.2e-05,
      "loss": 4.6364,
      "step": 4500
    },
    {
      "epoch": 0.29,
      "learning_rate": 8e-05,
      "loss": 4.4733,
      "step": 5000
    },
    {
      "epoch": 0.29,
      "eval_loss": 4.440802097320557,
      "eval_runtime": 488.2752,
      "eval_samples_per_second": 13.628,
      "eval_steps_per_second": 3.408,
      "step": 5000
    },
    {
      "epoch": 0.32,
      "learning_rate": 7.800000000000001e-05,
      "loss": 4.3792,
      "step": 5500
    },
    {
      "epoch": 0.35,
      "learning_rate": 7.6e-05,
      "loss": 4.2641,
      "step": 6000
    },
    {
      "epoch": 0.38,
      "learning_rate": 7.4e-05,
      "loss": 4.1555,
      "step": 6500
    },
    {
      "epoch": 0.41,
      "learning_rate": 7.2e-05,
      "loss": 4.0775,
      "step": 7000
    },
    {
      "epoch": 0.44,
      "learning_rate": 7e-05,
      "loss": 3.9886,
      "step": 7500
    },
    {
      "epoch": 0.47,
      "learning_rate": 6.800000000000001e-05,
      "loss": 3.9199,
      "step": 8000
    },
    {
      "epoch": 0.5,
      "learning_rate": 6.6e-05,
      "loss": 3.8676,
      "step": 8500
    },
    {
      "epoch": 0.53,
      "learning_rate": 6.400000000000001e-05,
      "loss": 3.7797,
      "step": 9000
    },
    {
      "epoch": 0.56,
      "learning_rate": 6.2e-05,
      "loss": 3.7232,
      "step": 9500
    },
    {
      "epoch": 0.58,
      "learning_rate": 6e-05,
      "loss": 3.7008,
      "step": 10000
    },
    {
      "epoch": 0.58,
      "eval_loss": 3.6708006858825684,
      "eval_runtime": 488.5846,
      "eval_samples_per_second": 13.619,
      "eval_steps_per_second": 3.406,
      "step": 10000
    },
    {
      "epoch": 0.61,
      "learning_rate": 5.8e-05,
      "loss": 3.6294,
      "step": 10500
    },
    {
      "epoch": 0.64,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 3.5787,
      "step": 11000
    },
    {
      "epoch": 0.67,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 3.5317,
      "step": 11500
    },
    {
      "epoch": 0.7,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 3.4919,
      "step": 12000
    },
    {
      "epoch": 0.73,
      "learning_rate": 5e-05,
      "loss": 3.4581,
      "step": 12500
    },
    {
      "epoch": 0.76,
      "learning_rate": 4.8e-05,
      "loss": 3.4151,
      "step": 13000
    },
    {
      "epoch": 0.79,
      "learning_rate": 4.600000000000001e-05,
      "loss": 3.3759,
      "step": 13500
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 3.3627,
      "step": 14000
    },
    {
      "epoch": 0.85,
      "learning_rate": 4.2e-05,
      "loss": 3.3041,
      "step": 14500
    },
    {
      "epoch": 0.88,
      "learning_rate": 4e-05,
      "loss": 3.2953,
      "step": 15000
    },
    {
      "epoch": 0.88,
      "eval_loss": 3.304645299911499,
      "eval_runtime": 489.0469,
      "eval_samples_per_second": 13.606,
      "eval_steps_per_second": 3.403,
      "step": 15000
    },
    {
      "epoch": 0.91,
      "learning_rate": 3.8e-05,
      "loss": 3.2512,
      "step": 15500
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.6e-05,
      "loss": 3.2418,
      "step": 16000
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 3.2232,
      "step": 16500
    },
    {
      "epoch": 0.99,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 3.1974,
      "step": 17000
    },
    {
      "epoch": 1.02,
      "learning_rate": 3e-05,
      "loss": 3.093,
      "step": 17500
    },
    {
      "epoch": 1.05,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 3.0629,
      "step": 18000
    },
    {
      "epoch": 1.08,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 3.0435,
      "step": 18500
    },
    {
      "epoch": 1.11,
      "learning_rate": 2.4e-05,
      "loss": 3.0299,
      "step": 19000
    },
    {
      "epoch": 1.14,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 3.0134,
      "step": 19500
    },
    {
      "epoch": 1.17,
      "learning_rate": 2e-05,
      "loss": 3.0173,
      "step": 20000
    },
    {
      "epoch": 1.17,
      "eval_loss": 3.116018772125244,
      "eval_runtime": 488.9966,
      "eval_samples_per_second": 13.607,
      "eval_steps_per_second": 3.403,
      "step": 20000
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.8e-05,
      "loss": 2.9757,
      "step": 20500
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.9725,
      "step": 21000
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 2.9563,
      "step": 21500
    },
    {
      "epoch": 1.29,
      "learning_rate": 1.2e-05,
      "loss": 2.9656,
      "step": 22000
    },
    {
      "epoch": 1.32,
      "learning_rate": 1e-05,
      "loss": 2.9446,
      "step": 22500
    },
    {
      "epoch": 1.35,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.941,
      "step": 23000
    },
    {
      "epoch": 1.37,
      "learning_rate": 6e-06,
      "loss": 2.9236,
      "step": 23500
    },
    {
      "epoch": 1.4,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.9237,
      "step": 24000
    },
    {
      "epoch": 1.43,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.922,
      "step": 24500
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0,
      "loss": 2.926,
      "step": 25000
    },
    {
      "epoch": 1.46,
      "eval_loss": 3.033102035522461,
      "eval_runtime": 488.912,
      "eval_samples_per_second": 13.61,
      "eval_steps_per_second": 3.403,
      "step": 25000
    },
    {
      "epoch": 1.46,
      "step": 25000,
      "total_flos": 3.7147907956565606e+17,
      "train_loss": 3.8248685107421876,
      "train_runtime": 44867.3669,
      "train_samples_per_second": 4.458,
      "train_steps_per_second": 0.557
    }
  ],
  "max_steps": 25000,
  "num_train_epochs": 2,
  "total_flos": 3.7147907956565606e+17,
  "trial_name": null,
  "trial_params": null
}