Hugging Face's logo

fydhfzh
/
hubert-classifier-aug-fold-2

Audio Classification
Transformers
TensorBoard
Safetensors
hubert
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
hubert-classifier-aug-fold-2 / README.md
fydhfzh's picture
fydhfzh
End of training
bc29397 verified
preview code
|
raw
history blame
11.4 kB
metadata
license: apache-2.0
base_model: facebook/hubert-base-ls960
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: hubert-classifier-aug-fold-2
    results: []

hubert-classifier-aug-fold-2

This model is a fine-tuned version of facebook/hubert-base-ls960 on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.7308
  • Accuracy: 0.8585
  • Precision: 0.8738
  • Recall: 0.8585
  • F1: 0.8578
  • Binary: 0.8997

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 0.0001
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 128
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 30
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy Precision Recall F1 Binary
No log 0.13 50 3.8576 0.0621 0.0074 0.0621 0.0112 0.3362
No log 0.27 100 3.4091 0.0931 0.0349 0.0931 0.0334 0.3617
No log 0.4 150 3.2224 0.1053 0.0231 0.1053 0.0330 0.3698
No log 0.54 200 3.0715 0.1215 0.0509 0.1215 0.0558 0.3800
No log 0.67 250 2.7013 0.2416 0.1533 0.2416 0.1549 0.4665
No log 0.81 300 2.4186 0.3077 0.2216 0.3077 0.2178 0.5124
No log 0.94 350 2.2086 0.3725 0.2827 0.3725 0.2882 0.5592
3.209 1.08 400 1.9624 0.4534 0.4314 0.4534 0.3907 0.6155
3.209 1.21 450 1.7552 0.5344 0.4957 0.5344 0.4788 0.6735
3.209 1.35 500 1.5887 0.5668 0.5359 0.5668 0.5127 0.6957
3.209 1.48 550 1.4249 0.6073 0.5999 0.6073 0.5671 0.7248
3.209 1.62 600 1.2562 0.6397 0.6424 0.6397 0.6007 0.7486
3.209 1.75 650 1.1504 0.6680 0.6724 0.6680 0.6379 0.7683
3.209 1.89 700 1.0797 0.6991 0.7111 0.6991 0.6753 0.7896
1.7186 2.02 750 1.0473 0.7314 0.7634 0.7314 0.7173 0.8139
1.7186 2.16 800 0.9383 0.7503 0.7777 0.7503 0.7343 0.8271
1.7186 2.29 850 0.9288 0.7679 0.7842 0.7679 0.7567 0.8393
1.7186 2.43 900 0.8284 0.7746 0.7897 0.7746 0.7679 0.8432
1.7186 2.56 950 0.8246 0.7800 0.8108 0.7800 0.7729 0.8474
1.7186 2.7 1000 0.8581 0.7665 0.7893 0.7665 0.7586 0.8370
1.7186 2.83 1050 0.7753 0.7962 0.8151 0.7962 0.7919 0.8583
1.7186 2.96 1100 0.7556 0.8043 0.8205 0.8043 0.7965 0.8641
1.0562 3.1 1150 0.8130 0.7962 0.8149 0.7962 0.7913 0.8567
1.0562 3.23 1200 0.7633 0.7895 0.8189 0.7895 0.7880 0.8528
1.0562 3.37 1250 0.6852 0.8178 0.8352 0.8178 0.8135 0.8742
1.0562 3.5 1300 0.8465 0.7638 0.8001 0.7638 0.7588 0.8347
1.0562 3.64 1350 0.7130 0.8111 0.8215 0.8111 0.8026 0.8686
1.0562 3.77 1400 0.7425 0.8057 0.8242 0.8057 0.8040 0.8657
1.0562 3.91 1450 0.7719 0.8057 0.8220 0.8057 0.8030 0.8634
0.782 4.04 1500 0.6581 0.8354 0.8497 0.8354 0.8332 0.8850
0.782 4.18 1550 0.7134 0.8124 0.8281 0.8124 0.8095 0.8682
0.782 4.31 1600 0.6923 0.8259 0.8455 0.8259 0.8234 0.8795
0.782 4.45 1650 0.6699 0.8367 0.8539 0.8367 0.8347 0.8868
0.782 4.58 1700 0.6169 0.8475 0.8597 0.8475 0.8456 0.8926
0.782 4.72 1750 0.6198 0.8381 0.8491 0.8381 0.8349 0.8870
0.782 4.85 1800 0.6939 0.8313 0.8502 0.8313 0.8318 0.8823
0.782 4.99 1850 0.7710 0.8205 0.8408 0.8205 0.8147 0.8729
0.6114 5.12 1900 0.6556 0.8381 0.8495 0.8381 0.8362 0.8880
0.6114 5.26 1950 0.7667 0.8151 0.8332 0.8151 0.8129 0.8698
0.6114 5.39 2000 0.7299 0.8232 0.8413 0.8232 0.8200 0.8748
0.6114 5.53 2050 0.7309 0.8327 0.8419 0.8327 0.8266 0.8835
0.6114 5.66 2100 0.7464 0.8192 0.8344 0.8192 0.8165 0.8750
0.6114 5.8 2150 0.7440 0.8340 0.8484 0.8340 0.8323 0.8849
0.6114 5.93 2200 0.7002 0.8475 0.8594 0.8475 0.8455 0.8935
0.5068 6.06 2250 0.7030 0.8448 0.8599 0.8448 0.8421 0.8912
0.5068 6.2 2300 0.7355 0.8421 0.8560 0.8421 0.8398 0.8889
0.5068 6.33 2350 0.7511 0.8246 0.8339 0.8246 0.8209 0.8771
0.5068 6.47 2400 0.6739 0.8421 0.8534 0.8421 0.8408 0.8903
0.5068 6.6 2450 0.6982 0.8475 0.8654 0.8475 0.8441 0.8931
0.5068 6.74 2500 0.7522 0.8408 0.8531 0.8408 0.8394 0.8880
0.5068 6.87 2550 0.7261 0.8354 0.8479 0.8354 0.8323 0.8856
0.4429 7.01 2600 0.7445 0.8421 0.8557 0.8421 0.8402 0.8896
0.4429 7.14 2650 0.7698 0.8327 0.8406 0.8327 0.8290 0.8823
0.4429 7.28 2700 0.8428 0.8273 0.8433 0.8273 0.8254 0.8806
0.4429 7.41 2750 0.8403 0.8259 0.8443 0.8259 0.8242 0.8789
0.4429 7.55 2800 0.7045 0.8381 0.8527 0.8381 0.8372 0.8866
0.4429 7.68 2850 0.8546 0.8327 0.8441 0.8327 0.8309 0.8825
0.4429 7.82 2900 0.7722 0.8340 0.8511 0.8340 0.8350 0.8830
0.4429 7.95 2950 0.7198 0.8435 0.8590 0.8435 0.8432 0.8927
0.3872 8.09 3000 0.7239 0.8408 0.8510 0.8408 0.8382 0.8914
0.3872 8.22 3050 0.7778 0.8475 0.8572 0.8475 0.8452 0.8950
0.3872 8.36 3100 0.8277 0.8394 0.8513 0.8394 0.8378 0.8892
0.3872 8.49 3150 0.7813 0.8462 0.8587 0.8462 0.8459 0.8912
0.3872 8.63 3200 0.7736 0.8394 0.8503 0.8394 0.8373 0.8880
0.3872 8.76 3250 0.7917 0.8394 0.8530 0.8394 0.8371 0.8883
0.3872 8.89 3300 0.7909 0.8475 0.8565 0.8475 0.8456 0.8930
0.3527 9.03 3350 0.7729 0.8529 0.8659 0.8529 0.8506 0.8977
0.3527 9.16 3400 0.8406 0.8475 0.8635 0.8475 0.8443 0.8937
0.3527 9.3 3450 0.7908 0.8435 0.8561 0.8435 0.8406 0.8910
0.3527 9.43 3500 0.8294 0.8300 0.8432 0.8300 0.8279 0.8808
0.3527 9.57 3550 0.8214 0.8421 0.8559 0.8421 0.8418 0.8897
0.3527 9.7 3600 0.8200 0.8435 0.8571 0.8435 0.8426 0.8918
0.3527 9.84 3650 0.7658 0.8502 0.8594 0.8502 0.8487 0.8969
0.3527 9.97 3700 0.8817 0.8354 0.8505 0.8354 0.8350 0.8842
0.3278 10.11 3750 0.8310 0.8489 0.8660 0.8489 0.8493 0.8950
0.3278 10.24 3800 0.6998 0.8664 0.8765 0.8664 0.8654 0.9072
0.3278 10.38 3850 0.7762 0.8435 0.8588 0.8435 0.8439 0.8904
0.3278 10.51 3900 0.9134 0.8435 0.8563 0.8435 0.8427 0.8920
0.3278 10.65 3950 0.7972 0.8556 0.8722 0.8556 0.8542 0.8988
0.3278 10.78 4000 0.8311 0.8610 0.8727 0.8610 0.8598 0.9043
0.3278 10.92 4050 0.8660 0.8543 0.8647 0.8543 0.8540 0.8993
0.2997 11.05 4100 0.8472 0.8556 0.8679 0.8556 0.8544 0.8988
0.2997 11.19 4150 0.7564 0.8556 0.8672 0.8556 0.8544 0.9001
0.2997 11.32 4200 0.7832 0.8529 0.8684 0.8529 0.8527 0.9005
0.2997 11.46 4250 0.8058 0.8623 0.8735 0.8623 0.8591 0.9049
0.2997 11.59 4300 0.7588 0.8623 0.8788 0.8623 0.8613 0.9054
0.2997 11.73 4350 0.8209 0.8462 0.8658 0.8462 0.8472 0.8926
0.2997 11.86 4400 0.7649 0.8650 0.8815 0.8650 0.8660 0.9043
0.2997 11.99 4450 0.7985 0.8529 0.8671 0.8529 0.8514 0.8966
0.283 12.13 4500 0.7353 0.8543 0.8640 0.8543 0.8529 0.8977
0.283 12.26 4550 0.7714 0.8543 0.8658 0.8543 0.8529 0.8992
0.283 12.4 4600 0.8393 0.8502 0.8629 0.8502 0.8476 0.8945
0.283 12.53 4650 0.7847 0.8556 0.8652 0.8556 0.8534 0.9001
0.283 12.67 4700 0.8041 0.8623 0.8722 0.8623 0.8590 0.9043
0.283 12.8 4750 0.8478 0.8475 0.8592 0.8475 0.8458 0.8939

Framework versions

  • Transformers 4.38.2
  • Pytorch 2.3.0
  • Datasets 2.19.1
  • Tokenizers 0.15.1