trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.4546952545642853,
  "best_model_checkpoint": "tiny_bert_rand_100_v2_sst2/checkpoint-264",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1584,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 6.941708564758301,
      "learning_rate": 4.9e-05,
      "loss": 0.4211,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8096330275229358,
      "eval_loss": 0.4546952545642853,
      "eval_runtime": 0.2664,
      "eval_samples_per_second": 3272.979,
      "eval_steps_per_second": 15.014,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.13948917388916,
      "learning_rate": 4.8e-05,
      "loss": 0.2373,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.805045871559633,
      "eval_loss": 0.5264467000961304,
      "eval_runtime": 0.2785,
      "eval_samples_per_second": 3130.648,
      "eval_steps_per_second": 14.361,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.722019195556641,
      "learning_rate": 4.7e-05,
      "loss": 0.1858,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7970183486238532,
      "eval_loss": 0.5311426520347595,
      "eval_runtime": 0.2625,
      "eval_samples_per_second": 3322.149,
      "eval_steps_per_second": 15.239,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.655911445617676,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1548,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8004587155963303,
      "eval_loss": 0.6056620478630066,
      "eval_runtime": 0.3067,
      "eval_samples_per_second": 2842.915,
      "eval_steps_per_second": 13.041,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 7.7325334548950195,
      "learning_rate": 4.5e-05,
      "loss": 0.1289,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7970183486238532,
      "eval_loss": 0.6140044331550598,
      "eval_runtime": 0.2918,
      "eval_samples_per_second": 2988.568,
      "eval_steps_per_second": 13.709,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.185094356536865,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1065,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7901376146788991,
      "eval_loss": 0.7775940299034119,
      "eval_runtime": 0.2731,
      "eval_samples_per_second": 3193.272,
      "eval_steps_per_second": 14.648,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 1.0596790573590528e+16,
      "train_loss": 0.2057432032594777,
      "train_runtime": 220.1611,
      "train_samples_per_second": 15295.387,
      "train_steps_per_second": 59.956
    }
  ],
  "logging_steps": 1,
  "max_steps": 13200,
  "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": 1.0596790573590528e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}