trainer_state.json

{
  "best_metric": 0.09891285747289658,
  "best_model_checkpoint": "output/output_minicpmv26_upsampled_3686400_pixels/checkpoint-200",
  "epoch": 0.704225352112676,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.176056338028169,
      "eval_loss": 0.1100488156080246,
      "eval_runtime": 190.5006,
      "eval_samples_per_second": 5.953,
      "eval_steps_per_second": 0.373,
      "step": 50
    },
    {
      "epoch": 0.352112676056338,
      "eval_loss": 0.10178369283676147,
      "eval_runtime": 190.517,
      "eval_samples_per_second": 5.952,
      "eval_steps_per_second": 0.373,
      "step": 100
    },
    {
      "epoch": 0.528169014084507,
      "eval_loss": 0.09937668591737747,
      "eval_runtime": 189.9102,
      "eval_samples_per_second": 5.971,
      "eval_steps_per_second": 0.374,
      "step": 150
    },
    {
      "epoch": 0.704225352112676,
      "eval_loss": 0.09891285747289658,
      "eval_runtime": 191.1583,
      "eval_samples_per_second": 5.932,
      "eval_steps_per_second": 0.371,
      "step": 200
    },
    {
      "epoch": 0.704225352112676,
      "grad_norm": 1.8547923564910889,
      "learning_rate": 1e-06,
      "loss": 0.139,
      "step": 200
    },
    {
      "epoch": 0.704225352112676,
      "step": 200,
      "total_flos": 3.598989492318372e+17,
      "train_loss": 0.1389956569671631,
      "train_runtime": 3953.2203,
      "train_samples_per_second": 2.024,
      "train_steps_per_second": 0.126
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 50,
  "total_flos": 3.598989492318372e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}