{
  "best_metric": 0.9724182168056447,
  "best_model_checkpoint": "/data3/akoudounas/speech_unlearning/models/fsc/facebook/wav2vec2-base/4/_retain/checkpoint-2800",
  "epoch": 3.872752420470263,
  "eval_steps": 400,
  "global_step": 2800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.13831258644536654,
      "grad_norm": 10.856929779052734,
      "learning_rate": 3.571428571428572e-05,
      "loss": 13.3808,
      "step": 100
    },
    {
      "epoch": 0.2766251728907331,
      "grad_norm": 47.9949836730957,
      "learning_rate": 7.142857142857143e-05,
      "loss": 11.8806,
      "step": 200
    },
    {
      "epoch": 0.4149377593360996,
      "grad_norm": 24.35114097595215,
      "learning_rate": 9.920634920634922e-05,
      "loss": 8.8407,
      "step": 300
    },
    {
      "epoch": 0.5532503457814661,
      "grad_norm": 63.59002685546875,
      "learning_rate": 9.523809523809524e-05,
      "loss": 6.359,
      "step": 400
    },
    {
      "epoch": 0.5532503457814661,
      "eval_accuracy": 0.5381654906991661,
      "eval_f1_macro": 0.42880994677778567,
      "eval_loss": 1.3551914691925049,
      "eval_runtime": 15.4249,
      "eval_samples_per_second": 202.141,
      "eval_steps_per_second": 25.284,
      "step": 400
    },
    {
      "epoch": 0.6915629322268326,
      "grad_norm": 81.84086608886719,
      "learning_rate": 9.126984126984128e-05,
      "loss": 4.7754,
      "step": 500
    },
    {
      "epoch": 0.8298755186721992,
      "grad_norm": 30.993141174316406,
      "learning_rate": 8.730158730158731e-05,
      "loss": 3.5042,
      "step": 600
    },
    {
      "epoch": 0.9681881051175657,
      "grad_norm": 53.2839469909668,
      "learning_rate": 8.333333333333334e-05,
      "loss": 2.5084,
      "step": 700
    },
    {
      "epoch": 1.1065006915629323,
      "grad_norm": 19.756397247314453,
      "learning_rate": 7.936507936507937e-05,
      "loss": 1.6743,
      "step": 800
    },
    {
      "epoch": 1.1065006915629323,
      "eval_accuracy": 0.9339320076972418,
      "eval_f1_macro": 0.9281889642509323,
      "eval_loss": 0.3620944619178772,
      "eval_runtime": 15.3176,
      "eval_samples_per_second": 203.557,
      "eval_steps_per_second": 25.461,
      "step": 800
    },
    {
      "epoch": 1.2448132780082988,
      "grad_norm": 94.57241821289062,
      "learning_rate": 7.53968253968254e-05,
      "loss": 1.3345,
      "step": 900
    },
    {
      "epoch": 1.3831258644536653,
      "grad_norm": 33.056297302246094,
      "learning_rate": 7.142857142857143e-05,
      "loss": 1.2201,
      "step": 1000
    },
    {
      "epoch": 1.5214384508990317,
      "grad_norm": 59.42671585083008,
      "learning_rate": 6.746031746031747e-05,
      "loss": 1.0254,
      "step": 1100
    },
    {
      "epoch": 1.6597510373443982,
      "grad_norm": 42.64637756347656,
      "learning_rate": 6.349206349206349e-05,
      "loss": 0.955,
      "step": 1200
    },
    {
      "epoch": 1.6597510373443982,
      "eval_accuracy": 0.9464400256574728,
      "eval_f1_macro": 0.9447013212949815,
      "eval_loss": 0.2722650468349457,
      "eval_runtime": 15.1717,
      "eval_samples_per_second": 205.514,
      "eval_steps_per_second": 25.706,
      "step": 1200
    },
    {
      "epoch": 1.798063623789765,
      "grad_norm": 8.055658340454102,
      "learning_rate": 5.9523809523809524e-05,
      "loss": 0.6846,
      "step": 1300
    },
    {
      "epoch": 1.9363762102351314,
      "grad_norm": 49.374576568603516,
      "learning_rate": 5.555555555555556e-05,
      "loss": 0.671,
      "step": 1400
    },
    {
      "epoch": 2.074688796680498,
      "grad_norm": 31.296615600585938,
      "learning_rate": 5.158730158730159e-05,
      "loss": 0.7596,
      "step": 1500
    },
    {
      "epoch": 2.2130013831258646,
      "grad_norm": 25.859086990356445,
      "learning_rate": 4.761904761904762e-05,
      "loss": 0.6647,
      "step": 1600
    },
    {
      "epoch": 2.2130013831258646,
      "eval_accuracy": 0.9554201411161001,
      "eval_f1_macro": 0.9536541863078912,
      "eval_loss": 0.2484109252691269,
      "eval_runtime": 15.2579,
      "eval_samples_per_second": 204.353,
      "eval_steps_per_second": 25.561,
      "step": 1600
    },
    {
      "epoch": 2.351313969571231,
      "grad_norm": 1.0564467906951904,
      "learning_rate": 4.3650793650793655e-05,
      "loss": 0.5915,
      "step": 1700
    },
    {
      "epoch": 2.4896265560165975,
      "grad_norm": 5.982853412628174,
      "learning_rate": 3.968253968253968e-05,
      "loss": 0.5167,
      "step": 1800
    },
    {
      "epoch": 2.627939142461964,
      "grad_norm": 0.10894600301980972,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.4604,
      "step": 1900
    },
    {
      "epoch": 2.7662517289073305,
      "grad_norm": 29.308454513549805,
      "learning_rate": 3.1746031746031745e-05,
      "loss": 0.5138,
      "step": 2000
    },
    {
      "epoch": 2.7662517289073305,
      "eval_accuracy": 0.9541372674791533,
      "eval_f1_macro": 0.9495278002552265,
      "eval_loss": 0.25238117575645447,
      "eval_runtime": 14.8444,
      "eval_samples_per_second": 210.046,
      "eval_steps_per_second": 26.273,
      "step": 2000
    },
    {
      "epoch": 2.904564315352697,
      "grad_norm": 124.017578125,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.5587,
      "step": 2100
    },
    {
      "epoch": 3.0428769017980635,
      "grad_norm": 11.674728393554688,
      "learning_rate": 2.380952380952381e-05,
      "loss": 0.4878,
      "step": 2200
    },
    {
      "epoch": 3.18118948824343,
      "grad_norm": 0.5218621492385864,
      "learning_rate": 1.984126984126984e-05,
      "loss": 0.3506,
      "step": 2300
    },
    {
      "epoch": 3.3195020746887964,
      "grad_norm": 10.221122741699219,
      "learning_rate": 1.5873015873015872e-05,
      "loss": 0.3697,
      "step": 2400
    },
    {
      "epoch": 3.3195020746887964,
      "eval_accuracy": 0.9666452854393842,
      "eval_f1_macro": 0.9636798531055095,
      "eval_loss": 0.19276651740074158,
      "eval_runtime": 14.4155,
      "eval_samples_per_second": 216.295,
      "eval_steps_per_second": 27.054,
      "step": 2400
    },
    {
      "epoch": 3.4578146611341634,
      "grad_norm": 162.04151916503906,
      "learning_rate": 1.1904761904761905e-05,
      "loss": 0.2989,
      "step": 2500
    },
    {
      "epoch": 3.59612724757953,
      "grad_norm": 0.0918952003121376,
      "learning_rate": 7.936507936507936e-06,
      "loss": 0.3396,
      "step": 2600
    },
    {
      "epoch": 3.7344398340248963,
      "grad_norm": 0.07032816857099533,
      "learning_rate": 3.968253968253968e-06,
      "loss": 0.3549,
      "step": 2700
    },
    {
      "epoch": 3.872752420470263,
      "grad_norm": 0.07183802127838135,
      "learning_rate": 0.0,
      "loss": 0.3393,
      "step": 2800
    },
    {
      "epoch": 3.872752420470263,
      "eval_accuracy": 0.9724182168056447,
      "eval_f1_macro": 0.9697252951761609,
      "eval_loss": 0.15897579491138458,
      "eval_runtime": 15.3271,
      "eval_samples_per_second": 203.431,
      "eval_steps_per_second": 25.445,
      "step": 2800
    }
  ],
  "logging_steps": 100,
  "max_steps": 2800,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 2800,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.253602560365056e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}