trainer_state.json

{
  "best_metric": 0.9875,
  "best_model_checkpoint": "vit-large-patch16-224-dungeon-geo-morphs-0-4-28Nov24-003/checkpoint-50",
  "epoch": 24.0,
  "eval_steps": 10,
  "global_step": 60,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 4.0,
      "grad_norm": 26.687711715698242,
      "learning_rate": 9.444444444444445e-06,
      "loss": 1.4912,
      "step": 10
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6785714285714286,
      "eval_loss": 1.1069196462631226,
      "eval_runtime": 7.7142,
      "eval_samples_per_second": 72.594,
      "eval_steps_per_second": 9.074,
      "step": 10
    },
    {
      "epoch": 8.0,
      "grad_norm": 21.99604034423828,
      "learning_rate": 7.592592592592594e-06,
      "loss": 0.6967,
      "step": 20
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9232142857142858,
      "eval_loss": 0.5143486857414246,
      "eval_runtime": 7.4404,
      "eval_samples_per_second": 75.264,
      "eval_steps_per_second": 9.408,
      "step": 20
    },
    {
      "epoch": 12.0,
      "grad_norm": 9.932448387145996,
      "learning_rate": 5.740740740740741e-06,
      "loss": 0.2492,
      "step": 30
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9767857142857143,
      "eval_loss": 0.25464263558387756,
      "eval_runtime": 8.0137,
      "eval_samples_per_second": 69.88,
      "eval_steps_per_second": 8.735,
      "step": 30
    },
    {
      "epoch": 16.0,
      "grad_norm": 3.4819257259368896,
      "learning_rate": 3.88888888888889e-06,
      "loss": 0.0819,
      "step": 40
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.975,
      "eval_loss": 0.16488894820213318,
      "eval_runtime": 8.2824,
      "eval_samples_per_second": 67.614,
      "eval_steps_per_second": 8.452,
      "step": 40
    },
    {
      "epoch": 20.0,
      "grad_norm": 1.605717420578003,
      "learning_rate": 2.037037037037037e-06,
      "loss": 0.0326,
      "step": 50
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.9875,
      "eval_loss": 0.10863279551267624,
      "eval_runtime": 7.0802,
      "eval_samples_per_second": 79.094,
      "eval_steps_per_second": 9.887,
      "step": 50
    },
    {
      "epoch": 24.0,
      "grad_norm": 0.9297263622283936,
      "learning_rate": 1.8518518518518518e-07,
      "loss": 0.0167,
      "step": 60
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.9857142857142858,
      "eval_loss": 0.09427731484174728,
      "eval_runtime": 8.2708,
      "eval_samples_per_second": 67.708,
      "eval_steps_per_second": 8.463,
      "step": 60
    },
    {
      "epoch": 24.0,
      "step": 60,
      "total_flos": 5.259589576438579e+17,
      "train_loss": 0.4280624141295751,
      "train_runtime": 348.1499,
      "train_samples_per_second": 6.894,
      "train_steps_per_second": 0.172
    }
  ],
  "logging_steps": 10,
  "max_steps": 60,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 30,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.259589576438579e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}