checkpoint-600/trainer_state.json

{
  "best_metric": 0.9875688610221299,
  "best_model_checkpoint": "distilhubert-finetuned-cry-detector/checkpoint-600",
  "epoch": 13.953488372093023,
  "eval_steps": 100,
  "global_step": 600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.1627906976744187,
      "grad_norm": 0.9607858657836914,
      "learning_rate": 1.744186046511628e-05,
      "loss": 0.6124,
      "step": 50
    },
    {
      "epoch": 2.3255813953488373,
      "grad_norm": 2.0142862796783447,
      "learning_rate": 2.997578868976474e-05,
      "loss": 0.3124,
      "step": 100
    },
    {
      "epoch": 2.3255813953488373,
      "eval_accuracy": 0.9641025641025641,
      "eval_confusion_matrix": [
        [
          948,
          22
        ],
        [
          27,
          368
        ]
      ],
      "eval_f1": 0.9640343817214121,
      "eval_loss": 0.2739163041114807,
      "eval_precision": 0.9639973701512163,
      "eval_recall": 0.9641025641025641,
      "eval_runtime": 3.48,
      "eval_samples_per_second": 392.238,
      "eval_steps_per_second": 6.322,
      "step": 100
    },
    {
      "epoch": 3.488372093023256,
      "grad_norm": 1.6254416704177856,
      "learning_rate": 2.9496736466239833e-05,
      "loss": 0.2509,
      "step": 150
    },
    {
      "epoch": 4.651162790697675,
      "grad_norm": 1.0478523969650269,
      "learning_rate": 2.8422656841613377e-05,
      "loss": 0.2337,
      "step": 200
    },
    {
      "epoch": 4.651162790697675,
      "eval_accuracy": 0.9736263736263736,
      "eval_confusion_matrix": [
        [
          950,
          20
        ],
        [
          16,
          379
        ]
      ],
      "eval_f1": 0.9736655604528286,
      "eval_loss": 0.23845618963241577,
      "eval_precision": 0.9737247242968067,
      "eval_recall": 0.9736263736263736,
      "eval_runtime": 3.3782,
      "eval_samples_per_second": 404.066,
      "eval_steps_per_second": 6.512,
      "step": 200
    },
    {
      "epoch": 5.813953488372093,
      "grad_norm": 0.19202682375907898,
      "learning_rate": 2.679763606999547e-05,
      "loss": 0.213,
      "step": 250
    },
    {
      "epoch": 6.976744186046512,
      "grad_norm": 3.5125396251678467,
      "learning_rate": 2.4688374121322346e-05,
      "loss": 0.2064,
      "step": 300
    },
    {
      "epoch": 6.976744186046512,
      "eval_accuracy": 0.9831501831501831,
      "eval_confusion_matrix": [
        [
          958,
          12
        ],
        [
          11,
          384
        ]
      ],
      "eval_f1": 0.9831564891330211,
      "eval_loss": 0.2294875830411911,
      "eval_precision": 0.983164066755398,
      "eval_recall": 0.9831501831501831,
      "eval_runtime": 3.421,
      "eval_samples_per_second": 399.01,
      "eval_steps_per_second": 6.431,
      "step": 300
    },
    {
      "epoch": 8.13953488372093,
      "grad_norm": 0.06433115154504776,
      "learning_rate": 2.218144694035814e-05,
      "loss": 0.2044,
      "step": 350
    },
    {
      "epoch": 9.30232558139535,
      "grad_norm": 0.1391066163778305,
      "learning_rate": 1.9379752884171956e-05,
      "loss": 0.2023,
      "step": 400
    },
    {
      "epoch": 9.30232558139535,
      "eval_accuracy": 0.9868131868131869,
      "eval_confusion_matrix": [
        [
          957,
          13
        ],
        [
          5,
          390
        ]
      ],
      "eval_f1": 0.9868519842367506,
      "eval_loss": 0.22771421074867249,
      "eval_precision": 0.986971783497838,
      "eval_recall": 0.9868131868131869,
      "eval_runtime": 3.414,
      "eval_samples_per_second": 399.822,
      "eval_steps_per_second": 6.444,
      "step": 400
    },
    {
      "epoch": 10.465116279069768,
      "grad_norm": 0.03317234292626381,
      "learning_rate": 1.6398289196234188e-05,
      "loss": 0.201,
      "step": 450
    },
    {
      "epoch": 11.627906976744185,
      "grad_norm": 0.0141974575817585,
      "learning_rate": 1.3359431874331888e-05,
      "loss": 0.2003,
      "step": 500
    },
    {
      "epoch": 11.627906976744185,
      "eval_accuracy": 0.9875457875457876,
      "eval_confusion_matrix": [
        [
          960,
          10
        ],
        [
          7,
          388
        ]
      ],
      "eval_f1": 0.9875597008648797,
      "eval_loss": 0.22541972994804382,
      "eval_precision": 0.9875850990732533,
      "eval_recall": 0.9875457875457876,
      "eval_runtime": 3.4267,
      "eval_samples_per_second": 398.347,
      "eval_steps_per_second": 6.42,
      "step": 500
    },
    {
      "epoch": 12.790697674418604,
      "grad_norm": 0.020028484985232353,
      "learning_rate": 1.0387912673007252e-05,
      "loss": 0.2005,
      "step": 550
    },
    {
      "epoch": 13.953488372093023,
      "grad_norm": 0.013127407990396023,
      "learning_rate": 7.605699412521431e-06,
      "loss": 0.2002,
      "step": 600
    },
    {
      "epoch": 13.953488372093023,
      "eval_accuracy": 0.9875457875457876,
      "eval_confusion_matrix": [
        [
          959,
          11
        ],
        [
          6,
          389
        ]
      ],
      "eval_f1": 0.9875688610221299,
      "eval_loss": 0.2259330302476883,
      "eval_precision": 0.9876237449942113,
      "eval_recall": 0.9875457875457876,
      "eval_runtime": 3.3216,
      "eval_samples_per_second": 410.95,
      "eval_steps_per_second": 6.623,
      "step": 600
    }
  ],
  "logging_steps": 50,
  "max_steps": 860,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 100,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.001
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.73355980821248e+17,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}