trainer_state.json

{
  "best_metric": 0.6791029572486877,
  "best_model_checkpoint": "./results/checkpoint-342",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 684,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7781569965870307,
      "eval_confusion_matrix": [
        [
          58,
          0,
          1,
          1,
          10,
          9,
          1
        ],
        [
          2,
          81,
          0,
          0,
          4,
          2,
          0
        ],
        [
          0,
          0,
          87,
          0,
          2,
          6,
          0
        ],
        [
          1,
          0,
          0,
          67,
          5,
          2,
          0
        ],
        [
          1,
          0,
          3,
          4,
          69,
          6,
          4
        ],
        [
          1,
          0,
          6,
          1,
          12,
          56,
          0
        ],
        [
          0,
          0,
          0,
          4,
          31,
          11,
          38
        ]
      ],
      "eval_f1": 0.7777605036882297,
      "eval_loss": 0.6863528490066528,
      "eval_precision": 0.8141265117353299,
      "eval_recall": 0.7752230960697254,
      "eval_runtime": 414.2099,
      "eval_samples_per_second": 1.415,
      "eval_steps_per_second": 0.179,
      "step": 171
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8003412969283277,
      "eval_confusion_matrix": [
        [
          72,
          0,
          0,
          1,
          4,
          2,
          1
        ],
        [
          2,
          84,
          0,
          0,
          0,
          0,
          3
        ],
        [
          2,
          0,
          86,
          0,
          3,
          4,
          0
        ],
        [
          7,
          0,
          0,
          65,
          2,
          1,
          0
        ],
        [
          10,
          1,
          2,
          2,
          53,
          8,
          11
        ],
        [
          5,
          0,
          10,
          1,
          9,
          51,
          0
        ],
        [
          1,
          0,
          0,
          3,
          17,
          5,
          58
        ]
      ],
      "eval_f1": 0.7977110437689846,
      "eval_loss": 0.6791029572486877,
      "eval_precision": 0.8015628505641682,
      "eval_recall": 0.7980677533763562,
      "eval_runtime": 417.5215,
      "eval_samples_per_second": 1.404,
      "eval_steps_per_second": 0.177,
      "step": 342
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.1340806484222412,
      "learning_rate": 5.3801169590643275e-05,
      "loss": 0.3724,
      "step": 500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8225255972696246,
      "eval_confusion_matrix": [
        [
          72,
          0,
          1,
          0,
          3,
          3,
          1
        ],
        [
          1,
          81,
          1,
          0,
          2,
          0,
          4
        ],
        [
          1,
          0,
          86,
          0,
          0,
          8,
          0
        ],
        [
          5,
          0,
          0,
          67,
          1,
          2,
          0
        ],
        [
          10,
          0,
          3,
          4,
          47,
          5,
          18
        ],
        [
          1,
          0,
          5,
          1,
          1,
          67,
          1
        ],
        [
          2,
          0,
          1,
          3,
          11,
          5,
          62
        ]
      ],
      "eval_f1": 0.819196982637564,
      "eval_loss": 0.805719256401062,
      "eval_precision": 0.821570223887049,
      "eval_recall": 0.8240875601856804,
      "eval_runtime": 415.3032,
      "eval_samples_per_second": 1.411,
      "eval_steps_per_second": 0.178,
      "step": 513
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8242320819112628,
      "eval_confusion_matrix": [
        [
          71,
          0,
          0,
          1,
          3,
          4,
          1
        ],
        [
          1,
          82,
          1,
          0,
          2,
          1,
          2
        ],
        [
          1,
          0,
          85,
          0,
          1,
          8,
          0
        ],
        [
          4,
          0,
          0,
          67,
          3,
          0,
          1
        ],
        [
          5,
          0,
          2,
          6,
          50,
          4,
          20
        ],
        [
          1,
          0,
          5,
          1,
          2,
          66,
          1
        ],
        [
          1,
          0,
          0,
          3,
          15,
          3,
          62
        ]
      ],
      "eval_f1": 0.8226354635422165,
      "eval_loss": 0.7115229368209839,
      "eval_precision": 0.8223101749006074,
      "eval_recall": 0.8254496320643309,
      "eval_runtime": 409.1206,
      "eval_samples_per_second": 1.432,
      "eval_steps_per_second": 0.181,
      "step": 684
    }
  ],
  "logging_steps": 500,
  "max_steps": 684,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 4.234352059828961e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}