trainer_state.json

{
  "best_metric": 0.9584930821803634,
  "best_model_checkpoint": "/scratch/camembertv2/runs/results/flue-CLS/camembertav2-base-bf16-p2-17000/max_seq_length-1024-gradient_accumulation_steps-4-precision-fp32-learning_rate-3e-05-epochs-6-lr_scheduler-linear-warmup_steps-0/SEED-1/checkpoint-562",
  "epoch": 5.984,
  "eval_steps": 500,
  "global_step": 1122,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 4.058077812194824,
      "learning_rate": 2.732620320855615e-05,
      "loss": 0.3373,
      "step": 100
    },
    {
      "epoch": 0.9973333333333333,
      "eval_accuracy": 0.9546591098516419,
      "eval_loss": 0.13871271908283234,
      "eval_runtime": 86.0844,
      "eval_samples_per_second": 69.687,
      "eval_steps_per_second": 8.712,
      "step": 187
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 14.420208930969238,
      "learning_rate": 2.4652406417112303e-05,
      "loss": 0.1886,
      "step": 200
    },
    {
      "epoch": 1.6,
      "grad_norm": 5.297502517700195,
      "learning_rate": 2.197860962566845e-05,
      "loss": 0.1315,
      "step": 300
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.953492248708118,
      "eval_loss": 0.15359961986541748,
      "eval_runtime": 85.7484,
      "eval_samples_per_second": 69.961,
      "eval_steps_per_second": 8.747,
      "step": 375
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 5.223259925842285,
      "learning_rate": 1.93048128342246e-05,
      "loss": 0.1186,
      "step": 400
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.2902381420135498,
      "learning_rate": 1.663101604278075e-05,
      "loss": 0.0632,
      "step": 500
    },
    {
      "epoch": 2.997333333333333,
      "eval_accuracy": 0.9584930821803634,
      "eval_loss": 0.16531726717948914,
      "eval_runtime": 85.8419,
      "eval_samples_per_second": 69.884,
      "eval_steps_per_second": 8.737,
      "step": 562
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.11738034337759018,
      "learning_rate": 1.39572192513369e-05,
      "loss": 0.0659,
      "step": 600
    },
    {
      "epoch": 3.7333333333333334,
      "grad_norm": 1.2773685455322266,
      "learning_rate": 1.1283422459893049e-05,
      "loss": 0.0421,
      "step": 700
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9574929154859143,
      "eval_loss": 0.19776488840579987,
      "eval_runtime": 86.0955,
      "eval_samples_per_second": 69.678,
      "eval_steps_per_second": 8.711,
      "step": 750
    },
    {
      "epoch": 4.266666666666667,
      "grad_norm": 0.0813373252749443,
      "learning_rate": 8.609625668449198e-06,
      "loss": 0.0377,
      "step": 800
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.8819140791893005,
      "learning_rate": 5.935828877005348e-06,
      "loss": 0.0193,
      "step": 900
    },
    {
      "epoch": 4.997333333333334,
      "eval_accuracy": 0.9553258876479414,
      "eval_loss": 0.22088374197483063,
      "eval_runtime": 85.8351,
      "eval_samples_per_second": 69.89,
      "eval_steps_per_second": 8.738,
      "step": 937
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 12.588311195373535,
      "learning_rate": 3.2620320855614974e-06,
      "loss": 0.0196,
      "step": 1000
    },
    {
      "epoch": 5.866666666666667,
      "grad_norm": 0.04329814016819,
      "learning_rate": 5.882352941176471e-07,
      "loss": 0.0222,
      "step": 1100
    },
    {
      "epoch": 5.984,
      "eval_accuracy": 0.9554925820970162,
      "eval_loss": 0.2227182686328888,
      "eval_runtime": 86.0509,
      "eval_samples_per_second": 69.715,
      "eval_steps_per_second": 8.716,
      "step": 1122
    },
    {
      "epoch": 5.984,
      "step": 1122,
      "total_flos": 6620583341429724.0,
      "train_loss": 0.09330896472760913,
      "train_runtime": 1923.7045,
      "train_samples_per_second": 18.705,
      "train_steps_per_second": 0.583
    }
  ],
  "logging_steps": 100,
  "max_steps": 1122,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6620583341429724.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}