checkpoint-45/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.824742268041237,
  "eval_steps": 500,
  "global_step": 45,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.41237113402061853,
      "grad_norm": 29.422929763793945,
      "learning_rate": 3.6e-05,
      "loss": 2.1003,
      "memory/device_reserved (GiB)": 0,
      "memory/max_active (GiB)": 4.94,
      "memory/max_allocated (GiB)": 8.44,
      "ppl": 8.16862,
      "step": 10,
      "tokens/total": 40960,
      "tokens/train_per_sec_per_gpu": 0.03,
      "tokens/trainable": 732
    },
    {
      "epoch": 0.8247422680412371,
      "grad_norm": 20.819232940673828,
      "learning_rate": 7.6e-05,
      "loss": 1.3313,
      "memory/device_reserved (GiB)": 0,
      "memory/max_active (GiB)": 4.94,
      "memory/max_allocated (GiB)": 8.44,
      "ppl": 3.78596,
      "step": 20,
      "tokens/total": 81920,
      "tokens/train_per_sec_per_gpu": 0.01,
      "tokens/trainable": 1375
    },
    {
      "epoch": 1.2061855670103092,
      "grad_norm": 10.385747909545898,
      "learning_rate": 0.000116,
      "loss": 0.8107,
      "memory/device_reserved (GiB)": 0,
      "memory/max_active (GiB)": 4.94,
      "memory/max_allocated (GiB)": 8.44,
      "ppl": 2.24948,
      "step": 30,
      "tokens/total": 119808,
      "tokens/train_per_sec_per_gpu": 0.03,
      "tokens/trainable": 1990
    },
    {
      "epoch": 1.6185567010309279,
      "grad_norm": 9.247703552246094,
      "learning_rate": 0.00015600000000000002,
      "loss": 0.4243,
      "memory/device_reserved (GiB)": 0,
      "memory/max_active (GiB)": 4.94,
      "memory/max_allocated (GiB)": 8.44,
      "ppl": 1.52852,
      "step": 40,
      "tokens/total": 160768,
      "tokens/train_per_sec_per_gpu": 0.01,
      "tokens/trainable": 2671
    }
  ],
  "logging_steps": 10,
  "max_steps": 49,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 15,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 542038679617536.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}