{
  "best_metric": 2.6958820819854736,
  "best_model_checkpoint": "/users/zyong2/data/zyong2/bigscience/data/processed/024/bloom-560m_de_continual-pretrain_100000samples_-1vocab_original_bsz32/checkpoint-5000",
  "epoch": 8.794834455282318,
  "global_step": 20000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.1,
      "learning_rate": 9e-05,
      "loss": 2.8663,
      "step": 2500
    },
    {
      "epoch": 2.2,
      "learning_rate": 8e-05,
      "loss": 2.4246,
      "step": 5000
    },
    {
      "epoch": 2.2,
      "eval_loss": 2.6958820819854736,
      "eval_runtime": 84.553,
      "eval_samples_per_second": 9.509,
      "eval_steps_per_second": 4.754,
      "step": 5000
    },
    {
      "epoch": 3.3,
      "learning_rate": 7e-05,
      "loss": 2.1988,
      "step": 7500
    },
    {
      "epoch": 4.4,
      "learning_rate": 6e-05,
      "loss": 1.9906,
      "step": 10000
    },
    {
      "epoch": 4.4,
      "eval_loss": 2.8127527236938477,
      "eval_runtime": 84.2148,
      "eval_samples_per_second": 9.547,
      "eval_steps_per_second": 4.774,
      "step": 10000
    },
    {
      "epoch": 5.5,
      "learning_rate": 5e-05,
      "loss": 1.7836,
      "step": 12500
    },
    {
      "epoch": 6.6,
      "learning_rate": 4e-05,
      "loss": 1.572,
      "step": 15000
    },
    {
      "epoch": 6.6,
      "eval_loss": 3.2299082279205322,
      "eval_runtime": 84.3434,
      "eval_samples_per_second": 9.532,
      "eval_steps_per_second": 4.766,
      "step": 15000
    },
    {
      "epoch": 7.7,
      "learning_rate": 3e-05,
      "loss": 1.354,
      "step": 17500
    },
    {
      "epoch": 8.79,
      "learning_rate": 2e-05,
      "loss": 1.1367,
      "step": 20000
    },
    {
      "epoch": 8.79,
      "eval_loss": 4.094446182250977,
      "eval_runtime": 84.5086,
      "eval_samples_per_second": 9.514,
      "eval_steps_per_second": 4.757,
      "step": 20000
    },
    {
      "epoch": 8.79,
      "step": 20000,
      "total_flos": 1.1890718469185864e+18,
      "train_loss": 1.915824853515625,
      "train_runtime": 214194.7633,
      "train_samples_per_second": 3.735,
      "train_steps_per_second": 0.117
    }
  ],
  "max_steps": 25000,
  "num_train_epochs": 11,
  "total_flos": 1.1890718469185864e+18,
  "trial_name": null,
  "trial_params": null
}