{
  "best_global_step": 70,
  "best_metric": 80.06465424867399,
  "best_model_checkpoint": "./finetuned_models/roberta-base-squad2-nq-nasa\\checkpoint-70",
  "epoch": 4.5,
  "eval_steps": 10,
  "global_step": 90,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.25,
      "grad_norm": 78.77288818359375,
      "learning_rate": 2.88e-05,
      "loss": 2.4514,
      "step": 5
    },
    {
      "epoch": 0.5,
      "grad_norm": 38.788047790527344,
      "learning_rate": 2.7300000000000003e-05,
      "loss": 2.5046,
      "step": 10
    },
    {
      "epoch": 0.5,
      "eval_HasAns_f1": 72.05322567724542,
      "eval_NoAns_f1": 0.0,
      "eval_exact": 58.0,
      "eval_f1": 72.05322567724542,
      "eval_loss": 1.7548620700836182,
      "eval_runtime": 0.281,
      "eval_samples_per_second": 177.92,
      "eval_steps_per_second": 14.234,
      "step": 10
    },
    {
      "epoch": 0.75,
      "grad_norm": 39.58235168457031,
      "learning_rate": 2.58e-05,
      "loss": 2.161,
      "step": 15
    },
    {
      "epoch": 1.0,
      "grad_norm": 27.151378631591797,
      "learning_rate": 2.43e-05,
      "loss": 1.5071,
      "step": 20
    },
    {
      "epoch": 1.0,
      "eval_HasAns_f1": 79.20518851115948,
      "eval_NoAns_f1": 0.0,
      "eval_exact": 66.0,
      "eval_f1": 79.20518851115948,
      "eval_loss": 1.5741324424743652,
      "eval_runtime": 0.2825,
      "eval_samples_per_second": 176.976,
      "eval_steps_per_second": 14.158,
      "step": 20
    },
    {
      "epoch": 1.25,
      "grad_norm": 22.792789459228516,
      "learning_rate": 2.2800000000000002e-05,
      "loss": 1.1212,
      "step": 25
    },
    {
      "epoch": 1.5,
      "grad_norm": 16.58112907409668,
      "learning_rate": 2.13e-05,
      "loss": 1.0362,
      "step": 30
    },
    {
      "epoch": 1.5,
      "eval_HasAns_f1": 72.06465424867399,
      "eval_NoAns_f1": 0.0,
      "eval_exact": 58.0,
      "eval_f1": 72.06465424867399,
      "eval_loss": 1.6612951755523682,
      "eval_runtime": 0.2746,
      "eval_samples_per_second": 182.093,
      "eval_steps_per_second": 14.567,
      "step": 30
    },
    {
      "epoch": 1.75,
      "grad_norm": 28.810911178588867,
      "learning_rate": 1.98e-05,
      "loss": 1.0579,
      "step": 35
    },
    {
      "epoch": 2.0,
      "grad_norm": 18.168163299560547,
      "learning_rate": 1.83e-05,
      "loss": 0.7972,
      "step": 40
    },
    {
      "epoch": 2.0,
      "eval_HasAns_f1": 71.7789399629597,
      "eval_NoAns_f1": 0.0,
      "eval_exact": 56.0,
      "eval_f1": 71.7789399629597,
      "eval_loss": 1.7468838691711426,
      "eval_runtime": 0.2855,
      "eval_samples_per_second": 175.114,
      "eval_steps_per_second": 14.009,
      "step": 40
    },
    {
      "epoch": 2.25,
      "grad_norm": 12.084095001220703,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.7772,
      "step": 45
    },
    {
      "epoch": 2.5,
      "grad_norm": 15.065386772155762,
      "learning_rate": 1.53e-05,
      "loss": 0.3204,
      "step": 50
    },
    {
      "epoch": 2.5,
      "eval_HasAns_f1": 78.35036853438828,
      "eval_NoAns_f1": 0.0,
      "eval_exact": 66.0,
      "eval_f1": 78.35036853438828,
      "eval_loss": 1.749760389328003,
      "eval_runtime": 0.2837,
      "eval_samples_per_second": 176.263,
      "eval_steps_per_second": 14.101,
      "step": 50
    },
    {
      "epoch": 2.75,
      "grad_norm": 14.943965911865234,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 0.5864,
      "step": 55
    },
    {
      "epoch": 3.0,
      "grad_norm": 69.38665771484375,
      "learning_rate": 1.2299999999999999e-05,
      "loss": 0.4019,
      "step": 60
    },
    {
      "epoch": 3.0,
      "eval_HasAns_f1": 76.06465424867399,
      "eval_NoAns_f1": 0.0,
      "eval_exact": 62.0,
      "eval_f1": 76.06465424867399,
      "eval_loss": 1.8023927211761475,
      "eval_runtime": 0.2736,
      "eval_samples_per_second": 182.767,
      "eval_steps_per_second": 14.621,
      "step": 60
    },
    {
      "epoch": 3.25,
      "grad_norm": 4.20263671875,
      "learning_rate": 1.08e-05,
      "loss": 0.3497,
      "step": 65
    },
    {
      "epoch": 3.5,
      "grad_norm": 19.26348876953125,
      "learning_rate": 9.3e-06,
      "loss": 0.4772,
      "step": 70
    },
    {
      "epoch": 3.5,
      "eval_HasAns_f1": 80.06465424867399,
      "eval_NoAns_f1": 0.0,
      "eval_exact": 66.0,
      "eval_f1": 80.06465424867399,
      "eval_loss": 1.7305716276168823,
      "eval_runtime": 0.2777,
      "eval_samples_per_second": 180.029,
      "eval_steps_per_second": 14.402,
      "step": 70
    },
    {
      "epoch": 3.75,
      "grad_norm": 14.338303565979004,
      "learning_rate": 7.8e-06,
      "loss": 0.268,
      "step": 75
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.6691396236419678,
      "learning_rate": 6.3e-06,
      "loss": 0.3622,
      "step": 80
    },
    {
      "epoch": 4.0,
      "eval_HasAns_f1": 78.06465424867399,
      "eval_NoAns_f1": 0.0,
      "eval_exact": 64.0,
      "eval_f1": 78.06465424867399,
      "eval_loss": 1.7988979816436768,
      "eval_runtime": 0.276,
      "eval_samples_per_second": 181.159,
      "eval_steps_per_second": 14.493,
      "step": 80
    },
    {
      "epoch": 4.25,
      "grad_norm": 7.720816135406494,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.1835,
      "step": 85
    },
    {
      "epoch": 4.5,
      "grad_norm": 13.70312786102295,
      "learning_rate": 3.3e-06,
      "loss": 0.4416,
      "step": 90
    },
    {
      "epoch": 4.5,
      "eval_HasAns_f1": 78.16991740656873,
      "eval_NoAns_f1": 0.0,
      "eval_exact": 66.0,
      "eval_f1": 78.16991740656873,
      "eval_loss": 1.8446077108383179,
      "eval_runtime": 0.2905,
      "eval_samples_per_second": 172.105,
      "eval_steps_per_second": 13.768,
      "step": 90
    }
  ],
  "logging_steps": 5,
  "max_steps": 100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 141100248637440.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}