last-checkpoint/trainer_state.json

{
  "best_metric": 0.8979436755180359,
  "best_model_checkpoint": "./opt_trained3/checkpoint-2720",
  "epoch": 1.99898924928788,
  "eval_steps": 500,
  "global_step": 2720,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3675457134981163,
      "grad_norm": 1.3510414361953735,
      "learning_rate": 9.189154158168292e-05,
      "loss": 0.9069,
      "step": 500
    },
    {
      "epoch": 0.7350914269962326,
      "grad_norm": 1.1864418983459473,
      "learning_rate": 7.019605024359474e-05,
      "loss": 0.9081,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.9017300605773926,
      "eval_runtime": 764.7068,
      "eval_samples_per_second": 50.601,
      "eval_steps_per_second": 12.651,
      "step": 1361
    },
    {
      "epoch": 1.1021777083524764,
      "grad_norm": 1.2064156532287598,
      "learning_rate": 4.195020556092935e-05,
      "loss": 0.8997,
      "step": 1500
    },
    {
      "epoch": 1.4697234218505928,
      "grad_norm": 1.2276240587234497,
      "learning_rate": 1.631521781767214e-05,
      "loss": 0.8969,
      "step": 2000
    },
    {
      "epoch": 1.837269135348709,
      "grad_norm": 1.193297266960144,
      "learning_rate": 1.6054963006338742e-06,
      "loss": 0.8947,
      "step": 2500
    },
    {
      "epoch": 1.99898924928788,
      "eval_loss": 0.8979436755180359,
      "eval_runtime": 761.1017,
      "eval_samples_per_second": 50.841,
      "eval_steps_per_second": 12.711,
      "step": 2720
    }
  ],
  "logging_steps": 500,
  "max_steps": 2720,
  "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": 1.37491227277056e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}