FacebookAI/roberta_base_twitter/checkpoint-50/trainer_state.json

{
  "best_metric": 0.47586047649383545,
  "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/FacebookAI/roberta_base_twitter/checkpoint-50",
  "epoch": 0.36764705882352944,
  "eval_steps": 50,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07,
      "grad_norm": 0.8274936676025391,
      "learning_rate": 1.950980392156863e-05,
      "loss": 0.6101,
      "step": 10
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.8833539485931396,
      "learning_rate": 1.9019607843137255e-05,
      "loss": 0.6154,
      "step": 20
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.267473220825195,
      "learning_rate": 1.8529411764705884e-05,
      "loss": 0.4968,
      "step": 30
    },
    {
      "epoch": 0.29,
      "grad_norm": 6.414277076721191,
      "learning_rate": 1.8039215686274513e-05,
      "loss": 0.5094,
      "step": 40
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.6427950859069824,
      "learning_rate": 1.7549019607843138e-05,
      "loss": 0.4867,
      "step": 50
    },
    {
      "epoch": 0.37,
      "eval_accuracy": 0.7711397058823529,
      "eval_f1_macro": 0.7371988324345717,
      "eval_f1_micro": 0.7711397058823529,
      "eval_loss": 0.47586047649383545,
      "eval_runtime": 1.0164,
      "eval_samples_per_second": 1070.408,
      "eval_steps_per_second": 16.725,
      "step": 50
    }
  ],
  "logging_steps": 10,
  "max_steps": 408,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "total_flos": 210488839372800.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}