trainer_state.json

{
  "best_metric": 0.691364049911499,
  "best_model_checkpoint": "distilbert_lda_5_v1_qnli/checkpoint-2870",
  "epoch": 12.0,
  "eval_steps": 500,
  "global_step": 4920,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.16467422246932983,
      "learning_rate": 0.00098,
      "loss": 0.7218,
      "step": 410
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4946000366099213,
      "eval_loss": 0.6934226751327515,
      "eval_runtime": 2.347,
      "eval_samples_per_second": 2327.7,
      "eval_steps_per_second": 9.374,
      "step": 410
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2183358371257782,
      "learning_rate": 0.00096,
      "loss": 0.6932,
      "step": 820
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4946000366099213,
      "eval_loss": 0.6935070157051086,
      "eval_runtime": 2.3054,
      "eval_samples_per_second": 2369.647,
      "eval_steps_per_second": 9.543,
      "step": 820
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.1270582526922226,
      "learning_rate": 0.00094,
      "loss": 0.6932,
      "step": 1230
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5053999633900788,
      "eval_loss": 0.6933382749557495,
      "eval_runtime": 2.3726,
      "eval_samples_per_second": 2302.498,
      "eval_steps_per_second": 9.272,
      "step": 1230
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.01812651753425598,
      "learning_rate": 0.00092,
      "loss": 0.6931,
      "step": 1640
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5053999633900788,
      "eval_loss": 0.6933382749557495,
      "eval_runtime": 2.3609,
      "eval_samples_per_second": 2313.915,
      "eval_steps_per_second": 9.318,
      "step": 1640
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.05285927280783653,
      "learning_rate": 0.0009000000000000001,
      "loss": 0.6932,
      "step": 2050
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4946000366099213,
      "eval_loss": 0.6934226751327515,
      "eval_runtime": 2.4032,
      "eval_samples_per_second": 2273.215,
      "eval_steps_per_second": 9.154,
      "step": 2050
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.05455218255519867,
      "learning_rate": 0.00088,
      "loss": 0.6932,
      "step": 2460
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5053999633900788,
      "eval_loss": 0.6933382749557495,
      "eval_runtime": 2.399,
      "eval_samples_per_second": 2277.166,
      "eval_steps_per_second": 9.17,
      "step": 2460
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.019881071522831917,
      "learning_rate": 0.00086,
      "loss": 0.6931,
      "step": 2870
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5053999633900788,
      "eval_loss": 0.691364049911499,
      "eval_runtime": 2.3314,
      "eval_samples_per_second": 2343.275,
      "eval_steps_per_second": 9.437,
      "step": 2870
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.13051092624664307,
      "learning_rate": 0.00084,
      "loss": 0.693,
      "step": 3280
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.4946000366099213,
      "eval_loss": 0.6934226751327515,
      "eval_runtime": 2.3863,
      "eval_samples_per_second": 2289.292,
      "eval_steps_per_second": 9.219,
      "step": 3280
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.05455218255519867,
      "learning_rate": 0.00082,
      "loss": 0.6932,
      "step": 3690
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.5053999633900788,
      "eval_loss": 0.6933382749557495,
      "eval_runtime": 2.3723,
      "eval_samples_per_second": 2302.849,
      "eval_steps_per_second": 9.274,
      "step": 3690
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.057659585028886795,
      "learning_rate": 0.0008,
      "loss": 0.6932,
      "step": 4100
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.4946000366099213,
      "eval_loss": 0.6934226751327515,
      "eval_runtime": 2.3678,
      "eval_samples_per_second": 2307.24,
      "eval_steps_per_second": 9.291,
      "step": 4100
    },
    {
      "epoch": 11.0,
      "grad_norm": 0.09045995026826859,
      "learning_rate": 0.0007800000000000001,
      "loss": 0.6932,
      "step": 4510
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.5053999633900788,
      "eval_loss": 0.6933382749557495,
      "eval_runtime": 2.3728,
      "eval_samples_per_second": 2302.381,
      "eval_steps_per_second": 9.272,
      "step": 4510
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.01812651753425598,
      "learning_rate": 0.00076,
      "loss": 0.6932,
      "step": 4920
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.4946000366099213,
      "eval_loss": 0.6933804750442505,
      "eval_runtime": 2.3791,
      "eval_samples_per_second": 2296.263,
      "eval_steps_per_second": 9.247,
      "step": 4920
    },
    {
      "epoch": 12.0,
      "step": 4920,
      "total_flos": 8.325019642455245e+16,
      "train_loss": 0.6955604677277852,
      "train_runtime": 1216.889,
      "train_samples_per_second": 4303.72,
      "train_steps_per_second": 16.846
    }
  ],
  "logging_steps": 1,
  "max_steps": 20500,
  "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": 8.325019642455245e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}