checkpoint-592/trainer_state.json

{
  "best_global_step": 592,
  "best_metric": 0.1268095225095749,
  "best_model_checkpoint": "siglip2-finetune-full/checkpoint-592",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 592,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.800612525081846,
      "eval_loss": 0.6111576557159424,
      "eval_model_preparation_time": 0.0024,
      "eval_runtime": 147.7559,
      "eval_samples_per_second": 64.085,
      "eval_steps_per_second": 8.013,
      "step": 296
    },
    {
      "epoch": 1.689189189189189,
      "grad_norm": 7.893478870391846,
      "learning_rate": 3.3948339483394833e-05,
      "loss": 0.6928,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9590241841799556,
      "eval_loss": 0.1268095225095749,
      "eval_model_preparation_time": 0.0024,
      "eval_runtime": 148.2374,
      "eval_samples_per_second": 63.877,
      "eval_steps_per_second": 7.987,
      "step": 592
    }
  ],
  "logging_steps": 500,
  "max_steps": 592,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.5862668048752394e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}