trainer_state.json

{
  "best_metric": 0.10972103476524353,
  "best_model_checkpoint": "output/output_minicpmv26_20240908_220534/checkpoint-200",
  "epoch": 1.4925373134328357,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.373134328358209,
      "eval_loss": 0.11530657857656479,
      "eval_runtime": 57.3132,
      "eval_samples_per_second": 9.317,
      "eval_steps_per_second": 0.593,
      "step": 50
    },
    {
      "epoch": 0.746268656716418,
      "eval_loss": 0.11028832942247391,
      "eval_runtime": 56.3749,
      "eval_samples_per_second": 9.472,
      "eval_steps_per_second": 0.603,
      "step": 100
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.6975690126419067,
      "learning_rate": 1e-06,
      "loss": 0.1504,
      "step": 134
    },
    {
      "epoch": 1.1194029850746268,
      "eval_loss": 0.11502116918563843,
      "eval_runtime": 56.3478,
      "eval_samples_per_second": 9.477,
      "eval_steps_per_second": 0.603,
      "step": 150
    },
    {
      "epoch": 1.4925373134328357,
      "eval_loss": 0.10972103476524353,
      "eval_runtime": 56.3066,
      "eval_samples_per_second": 9.484,
      "eval_steps_per_second": 0.604,
      "step": 200
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 1.4877771139144897,
      "learning_rate": 1e-06,
      "loss": 0.0774,
      "step": 200
    },
    {
      "epoch": 1.4925373134328357,
      "step": 200,
      "total_flos": 2.772040612315136e+17,
      "train_loss": 0.12631385326385497,
      "train_runtime": 2854.968,
      "train_samples_per_second": 2.802,
      "train_steps_per_second": 0.175
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 50,
  "total_flos": 2.772040612315136e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}