trainer_state.json

{
  "epoch": 0.03262034577566522,
  "global_step": 350,
  "max_steps": 3000,
  "logging_steps": 50,
  "eval_steps": 50,
  "save_steps": 50,
  "train_batch_size": 8,
  "num_train_epochs": 1,
  "num_input_tokens_seen": 0,
  "total_flos": 1.6004947053699072e+17,
  "log_history": [
    {
      "loss": 11.5092,
      "grad_norm": 27.41529655456543,
      "learning_rate": 4.666666666666667e-05,
      "epoch": 0.004660049396523603,
      "step": 50
    },
    {
      "eval_loss": 6.5413079261779785,
      "eval_runtime": 484.9209,
      "eval_samples_per_second": 1.862,
      "eval_steps_per_second": 0.466,
      "epoch": 0.004660049396523603,
      "step": 50
    },
    {
      "loss": 0.907,
      "grad_norm": 0.3267304003238678,
      "learning_rate": 9.999988344964554e-05,
      "epoch": 0.009320098793047207,
      "step": 100
    },
    {
      "eval_loss": 0.2963584065437317,
      "eval_runtime": 482.3744,
      "eval_samples_per_second": 1.872,
      "eval_steps_per_second": 0.469,
      "epoch": 0.009320098793047207,
      "step": 100
    },
    {
      "loss": 0.2887,
      "grad_norm": 0.23999030888080597,
      "learning_rate": 9.992123261946325e-05,
      "epoch": 0.01398014818957081,
      "step": 150
    },
    {
      "eval_loss": 0.26408717036247253,
      "eval_runtime": 483.0929,
      "eval_samples_per_second": 1.869,
      "eval_steps_per_second": 0.468,
      "epoch": 0.01398014818957081,
      "step": 150
    },
    {
      "loss": 0.273,
      "grad_norm": 0.22597914934158325,
      "learning_rate": 9.96971586146684e-05,
      "epoch": 0.018640197586094413,
      "step": 200
    },
    {
      "eval_loss": 0.2530948221683502,
      "eval_runtime": 483.695,
      "eval_samples_per_second": 1.867,
      "eval_steps_per_second": 0.467,
      "epoch": 0.018640197586094413,
      "step": 200
    },
    {
      "loss": 0.2655,
      "grad_norm": 0.2108173966407776,
      "learning_rate": 9.932831417461484e-05,
      "epoch": 0.023300246982618015,
      "step": 250
    },
    {
      "eval_loss": 0.2521112263202667,
      "eval_runtime": 484.1571,
      "eval_samples_per_second": 1.865,
      "eval_steps_per_second": 0.467,
      "epoch": 0.023300246982618015,
      "step": 250
    },
    {
      "loss": 0.2635,
      "grad_norm": 0.16866298019886017,
      "learning_rate": 9.881577376254393e-05,
      "epoch": 0.02796029637914162,
      "step": 300
    },
    {
      "eval_loss": 0.24954503774642944,
      "eval_runtime": 482.9845,
      "eval_samples_per_second": 1.87,
      "eval_steps_per_second": 0.468,
      "epoch": 0.02796029637914162,
      "step": 300
    },
    {
      "loss": 0.2622,
      "grad_norm": 0.19963641464710236,
      "learning_rate": 9.816103043561648e-05,
      "epoch": 0.03262034577566522,
      "step": 350
    },
    {
      "eval_loss": 0.24790766835212708,
      "eval_runtime": 482.7586,
      "eval_samples_per_second": 1.87,
      "eval_steps_per_second": 0.468,
      "epoch": 0.03262034577566522,
      "step": 350
    }
  ],
  "best_metric": 0.24790766835212708,
  "best_model_checkpoint": "./multimodal-phi3_5-mini-instruct-llava_adapter/checkpoint-350",
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "is_hyper_param_search": false,
  "trial_name": null,
  "trial_params": null,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.01
      },
      "attributes": {
        "early_stopping_patience_counter": 3
      }
    },
    "TrainerControl": {
      "args": {
        "should_training_stop": true,
        "should_epoch_stop": false,
        "should_save": true,
        "should_evaluate": false,
        "should_log": false
      },
      "attributes": {}
    }
  }
}