checkpoint-740/trainer_state.json

{
  "best_metric": 3.880638837814331,
  "best_model_checkpoint": "buscuit_wrappers_image_detection/checkpoint-740",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 740,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.19543147208121828,
      "eval_loss": 4.442927360534668,
      "eval_runtime": 108.6903,
      "eval_samples_per_second": 29.0,
      "eval_steps_per_second": 3.625,
      "step": 148
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7385786802030457,
      "eval_loss": 4.219130039215088,
      "eval_runtime": 108.4785,
      "eval_samples_per_second": 29.056,
      "eval_steps_per_second": 3.632,
      "step": 296
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8911802030456852,
      "eval_loss": 4.033923625946045,
      "eval_runtime": 108.6113,
      "eval_samples_per_second": 29.021,
      "eval_steps_per_second": 3.628,
      "step": 444
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.740444302558899,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 4.2571,
      "step": 500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9216370558375635,
      "eval_loss": 3.919619560241699,
      "eval_runtime": 108.434,
      "eval_samples_per_second": 29.068,
      "eval_steps_per_second": 3.634,
      "step": 592
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.930520304568528,
      "eval_loss": 3.880638837814331,
      "eval_runtime": 108.0954,
      "eval_samples_per_second": 29.159,
      "eval_steps_per_second": 3.645,
      "step": 740
    }
  ],
  "logging_steps": 500,
  "max_steps": 740,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 1.832760491724288e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}