trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.4347452223300934,
  "best_model_checkpoint": "bert_base_rand_50_v2_sst2/checkpoint-264",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1584,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 7.811213970184326,
      "learning_rate": 4.9e-05,
      "loss": 0.3864,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8027522935779816,
      "eval_loss": 0.4347452223300934,
      "eval_runtime": 0.5741,
      "eval_samples_per_second": 1518.777,
      "eval_steps_per_second": 6.967,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.949086666107178,
      "learning_rate": 4.8e-05,
      "loss": 0.2165,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8027522935779816,
      "eval_loss": 0.5241801738739014,
      "eval_runtime": 0.5788,
      "eval_samples_per_second": 1506.68,
      "eval_steps_per_second": 6.911,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.619488716125488,
      "learning_rate": 4.7e-05,
      "loss": 0.1588,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7924311926605505,
      "eval_loss": 0.553471565246582,
      "eval_runtime": 0.5832,
      "eval_samples_per_second": 1495.164,
      "eval_steps_per_second": 6.859,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.747478485107422,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1206,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8119266055045872,
      "eval_loss": 0.5647808313369751,
      "eval_runtime": 0.5838,
      "eval_samples_per_second": 1493.58,
      "eval_steps_per_second": 6.851,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.65807580947876,
      "learning_rate": 4.5e-05,
      "loss": 0.0928,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7970183486238532,
      "eval_loss": 0.6491648554801941,
      "eval_runtime": 0.5817,
      "eval_samples_per_second": 1499.02,
      "eval_steps_per_second": 6.876,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 6.006173610687256,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0724,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8245412844036697,
      "eval_loss": 0.7118688225746155,
      "eval_runtime": 0.5648,
      "eval_samples_per_second": 1543.853,
      "eval_steps_per_second": 7.082,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 5.316079940232192e+16,
      "train_loss": 0.17456337297805632,
      "train_runtime": 670.6712,
      "train_samples_per_second": 5021.015,
      "train_steps_per_second": 19.682
    }
  ],
  "logging_steps": 1,
  "max_steps": 13200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 5
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.316079940232192e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}