{
  "best_metric": 3.2361934185028076,
  "best_model_checkpoint": "/users/zyong2/data/zyong2/bigscience/data/processed/024/bloom-350m_az_fish_1000samples_-1vocab_original-frozen/checkpoint-5000",
  "epoch": 144.9225225225225,
  "global_step": 20000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 18.12,
      "learning_rate": 9e-05,
      "loss": 2.7987,
      "step": 2500
    },
    {
      "epoch": 36.23,
      "learning_rate": 8e-05,
      "loss": 2.0754,
      "step": 5000
    },
    {
      "epoch": 36.23,
      "eval_loss": 3.2361934185028076,
      "eval_runtime": 515.1551,
      "eval_samples_per_second": 9.523,
      "eval_steps_per_second": 4.762,
      "step": 5000
    },
    {
      "epoch": 54.35,
      "learning_rate": 7e-05,
      "loss": 1.628,
      "step": 7500
    },
    {
      "epoch": 72.46,
      "learning_rate": 6e-05,
      "loss": 1.2791,
      "step": 10000
    },
    {
      "epoch": 72.46,
      "eval_loss": 4.527117729187012,
      "eval_runtime": 515.0288,
      "eval_samples_per_second": 9.526,
      "eval_steps_per_second": 4.763,
      "step": 10000
    },
    {
      "epoch": 90.58,
      "learning_rate": 5e-05,
      "loss": 1.0056,
      "step": 12500
    },
    {
      "epoch": 108.69,
      "learning_rate": 4e-05,
      "loss": 0.7878,
      "step": 15000
    },
    {
      "epoch": 108.69,
      "eval_loss": 5.602742671966553,
      "eval_runtime": 515.0516,
      "eval_samples_per_second": 9.525,
      "eval_steps_per_second": 4.763,
      "step": 15000
    },
    {
      "epoch": 126.81,
      "learning_rate": 3e-05,
      "loss": 0.6143,
      "step": 17500
    },
    {
      "epoch": 144.92,
      "learning_rate": 2e-05,
      "loss": 0.4749,
      "step": 20000
    },
    {
      "epoch": 144.92,
      "eval_loss": 6.468493938446045,
      "eval_runtime": 515.9317,
      "eval_samples_per_second": 9.509,
      "eval_steps_per_second": 4.755,
      "step": 20000
    },
    {
      "epoch": 144.92,
      "step": 20000,
      "total_flos": 2.9879104118666035e+17,
      "train_loss": 1.3329723815917969,
      "train_runtime": 48886.3278,
      "train_samples_per_second": 4.091,
      "train_steps_per_second": 0.511
    }
  ],
  "max_steps": 25000,
  "num_train_epochs": 182,
  "total_flos": 2.9879104118666035e+17,
  "trial_name": null,
  "trial_params": null
}