{
  "best_global_step": 5688,
  "best_metric": 0.4044477343559265,
  "best_model_checkpoint": "tiny_bert_km_5_v2_qqp/checkpoint-5688",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 12798,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.8056795597076416,
      "learning_rate": 4.9e-05,
      "loss": 0.4976,
      "step": 1422
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7705169428642098,
      "eval_combined_score": 0.7026638087539974,
      "eval_f1": 0.6348106746437849,
      "eval_loss": 0.4731883108615875,
      "eval_runtime": 11.6835,
      "eval_samples_per_second": 3460.449,
      "eval_steps_per_second": 13.523,
      "step": 1422
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.5427074432373047,
      "learning_rate": 4.8e-05,
      "loss": 0.4095,
      "step": 2844
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8045263418253772,
      "eval_combined_score": 0.7676065002703305,
      "eval_f1": 0.7306866587152837,
      "eval_loss": 0.4118152856826782,
      "eval_runtime": 11.4911,
      "eval_samples_per_second": 3518.379,
      "eval_steps_per_second": 13.75,
      "step": 2844
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.615072727203369,
      "learning_rate": 4.7e-05,
      "loss": 0.3541,
      "step": 4266
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8160771704180064,
      "eval_combined_score": 0.7802986003261612,
      "eval_f1": 0.7445200302343159,
      "eval_loss": 0.40586817264556885,
      "eval_runtime": 11.5081,
      "eval_samples_per_second": 3513.184,
      "eval_steps_per_second": 13.729,
      "step": 4266
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.066542148590088,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.3088,
      "step": 5688
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8236210734603018,
      "eval_combined_score": 0.7876160332643387,
      "eval_f1": 0.7516109930683758,
      "eval_loss": 0.4044477343559265,
      "eval_runtime": 11.4853,
      "eval_samples_per_second": 3520.139,
      "eval_steps_per_second": 13.757,
      "step": 5688
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.02803897857666,
      "learning_rate": 4.5e-05,
      "loss": 0.2715,
      "step": 7110
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8228790502102399,
      "eval_combined_score": 0.7795562144859443,
      "eval_f1": 0.7362333787616486,
      "eval_loss": 0.4359254240989685,
      "eval_runtime": 12.0762,
      "eval_samples_per_second": 3347.906,
      "eval_steps_per_second": 13.084,
      "step": 7110
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.2800185680389404,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2395,
      "step": 8532
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8291862478357656,
      "eval_combined_score": 0.7956308142644686,
      "eval_f1": 0.7620753806931716,
      "eval_loss": 0.4360813498497009,
      "eval_runtime": 11.5074,
      "eval_samples_per_second": 3513.38,
      "eval_steps_per_second": 13.73,
      "step": 8532
    },
    {
      "epoch": 7.0,
      "grad_norm": 5.218929767608643,
      "learning_rate": 4.3e-05,
      "loss": 0.2125,
      "step": 9954
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8311402423942617,
      "eval_combined_score": 0.8026412827414455,
      "eval_f1": 0.7741423230886294,
      "eval_loss": 0.44221463799476624,
      "eval_runtime": 11.4819,
      "eval_samples_per_second": 3521.193,
      "eval_steps_per_second": 13.761,
      "step": 9954
    },
    {
      "epoch": 8.0,
      "grad_norm": 2.9345195293426514,
      "learning_rate": 4.2e-05,
      "loss": 0.1876,
      "step": 11376
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8353450408112788,
      "eval_combined_score": 0.8054393208237531,
      "eval_f1": 0.7755336008362276,
      "eval_loss": 0.48129960894584656,
      "eval_runtime": 11.4472,
      "eval_samples_per_second": 3531.863,
      "eval_steps_per_second": 13.802,
      "step": 11376
    },
    {
      "epoch": 9.0,
      "grad_norm": 6.157193660736084,
      "learning_rate": 4.1e-05,
      "loss": 0.1677,
      "step": 12798
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.833044768736087,
      "eval_combined_score": 0.8069800279902442,
      "eval_f1": 0.7809152872444012,
      "eval_loss": 0.5044159293174744,
      "eval_runtime": 11.4623,
      "eval_samples_per_second": 3527.226,
      "eval_steps_per_second": 13.784,
      "step": 12798
    },
    {
      "epoch": 9.0,
      "step": 12798,
      "total_flos": 8.587209601564877e+16,
      "train_loss": 0.294309814901422,
      "train_runtime": 1807.9441,
      "train_samples_per_second": 10062.424,
      "train_steps_per_second": 39.326
    }
  ],
  "logging_steps": 1,
  "max_steps": 71100,
  "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": 8.587209601564877e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}