trainer_state.json

{
  "best_metric": 0.8325659152041666,
  "best_model_checkpoint": "/home/mesay/project_gpt/low-resource-hate/finetuned-models/full-sample/multilingual-models/bert-base-uncased_HSpeech_32122_full/checkpoint-5859",
  "epoch": 3.0,
  "global_step": 5859,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.26,
      "learning_rate": 4.573306024918929e-05,
      "loss": 0.5365,
      "step": 500
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.146612049837856e-05,
      "loss": 0.488,
      "step": 1000
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.719918074756785e-05,
      "loss": 0.4706,
      "step": 1500
    },
    {
      "epoch": 1.0,
      "eval_hate_f1": 0.6494623655913979,
      "eval_hate_precision": 0.7864583333333334,
      "eval_hate_recall": 0.5531135531135531,
      "eval_loss": 0.4354419410228729,
      "eval_macro_f1": 0.7606084447941254,
      "eval_runtime": 5.6262,
      "eval_samples_per_second": 308.556,
      "eval_steps_per_second": 9.776,
      "step": 1953
    },
    {
      "epoch": 1.02,
      "learning_rate": 3.293224099675713e-05,
      "loss": 0.4372,
      "step": 2000
    },
    {
      "epoch": 1.28,
      "learning_rate": 2.866530124594641e-05,
      "loss": 0.393,
      "step": 2500
    },
    {
      "epoch": 1.54,
      "learning_rate": 2.439836149513569e-05,
      "loss": 0.3942,
      "step": 3000
    },
    {
      "epoch": 1.79,
      "learning_rate": 2.013142174432497e-05,
      "loss": 0.3718,
      "step": 3500
    },
    {
      "epoch": 2.0,
      "eval_hate_f1": 0.7334558823529411,
      "eval_hate_precision": 0.7361623616236163,
      "eval_hate_recall": 0.7307692307692307,
      "eval_loss": 0.4192083179950714,
      "eval_macro_f1": 0.8059057935254639,
      "eval_runtime": 5.6449,
      "eval_samples_per_second": 307.537,
      "eval_steps_per_second": 9.743,
      "step": 3906
    },
    {
      "epoch": 2.05,
      "learning_rate": 1.5864481993514253e-05,
      "loss": 0.3504,
      "step": 4000
    },
    {
      "epoch": 2.3,
      "learning_rate": 1.1597542242703534e-05,
      "loss": 0.2862,
      "step": 4500
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.330602491892815e-06,
      "loss": 0.285,
      "step": 5000
    },
    {
      "epoch": 2.82,
      "learning_rate": 3.063662741082096e-06,
      "loss": 0.2801,
      "step": 5500
    },
    {
      "epoch": 3.0,
      "eval_hate_f1": 0.7647628267182962,
      "eval_hate_precision": 0.811088295687885,
      "eval_hate_recall": 0.7234432234432234,
      "eval_loss": 0.4106697142124176,
      "eval_macro_f1": 0.8325659152041666,
      "eval_runtime": 5.6284,
      "eval_samples_per_second": 308.438,
      "eval_steps_per_second": 9.772,
      "step": 5859
    },
    {
      "epoch": 3.0,
      "step": 5859,
      "total_flos": 4140989789224320.0,
      "train_loss": 0.38279720344810403,
      "train_runtime": 1347.9523,
      "train_samples_per_second": 69.545,
      "train_steps_per_second": 4.347
    }
  ],
  "max_steps": 5859,
  "num_train_epochs": 3,
  "total_flos": 4140989789224320.0,
  "trial_name": null,
  "trial_params": null
}