<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
model_syllable_onSet2
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4231
- 0 Precision: 1.0
- 0 Recall: 0.96
- 0 F1-score: 0.9796
- 0 Support: 25
- 1 Precision: 0.9643
- 1 Recall: 0.9643
- 1 F1-score: 0.9643
- 1 Support: 28
- 2 Precision: 1.0
- 2 Recall: 0.9643
- 2 F1-score: 0.9818
- 2 Support: 28
- 3 Precision: 0.8889
- 3 Recall: 1.0
- 3 F1-score: 0.9412
- 3 Support: 16
- Accuracy: 0.9691
- Macro avg Precision: 0.9633
- Macro avg Recall: 0.9721
- Macro avg F1-score: 0.9667
- Macro avg Support: 97
- Weighted avg Precision: 0.9714
- Weighted avg Recall: 0.9691
- Weighted avg F1-score: 0.9695
- Weighted avg Support: 97
- Wer: 0.2827
- Mtrix: [[0, 1, 2, 3], [0, 24, 1, 0, 0], [1, 0, 27, 0, 1], [2, 0, 0, 27, 1], [3, 0, 0, 0, 16]]
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: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 200
- num_epochs: 70
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | 0 Precision | 0 Recall | 0 F1-score | 0 Support | 1 Precision | 1 Recall | 1 F1-score | 1 Support | 2 Precision | 2 Recall | 2 F1-score | 2 Support | 3 Precision | 3 Recall | 3 F1-score | 3 Support | Accuracy | Macro avg Precision | Macro avg Recall | Macro avg F1-score | Macro avg Support | Weighted avg Precision | Weighted avg Recall | Weighted avg F1-score | Weighted avg Support | Wer | Mtrix |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1.3102 | 4.16 | 100 | 1.2133 | 0.125 | 0.04 | 0.0606 | 25 | 0.0 | 0.0 | 0.0 | 28 | 0.3146 | 1.0 | 0.4786 | 28 | 0.0 | 0.0 | 0.0 | 16 | 0.2990 | 0.1099 | 0.26 | 0.1348 | 97 | 0.1230 | 0.2990 | 0.1538 | 97 | 0.9676 | [[0, 1, 2, 3], [0, 1, 0, 24, 0], [1, 7, 0, 21, 0], [2, 0, 0, 28, 0], [3, 0, 0, 16, 0]] |
0.7368 | 8.33 | 200 | 0.7100 | 1.0 | 0.72 | 0.8372 | 25 | 0.3333 | 0.0357 | 0.0645 | 28 | 0.3684 | 1.0 | 0.5385 | 28 | 0.0 | 0.0 | 0.0 | 16 | 0.4845 | 0.4254 | 0.4389 | 0.3600 | 97 | 0.4603 | 0.4845 | 0.3898 | 97 | 0.8227 | [[0, 1, 2, 3], [0, 18, 2, 5, 0], [1, 0, 1, 27, 0], [2, 0, 0, 28, 0], [3, 0, 0, 16, 0]] |
0.3813 | 12.49 | 300 | 0.3802 | 0.8519 | 0.92 | 0.8846 | 25 | 0.7333 | 0.7857 | 0.7586 | 28 | 0.9231 | 0.8571 | 0.8889 | 28 | 0.9286 | 0.8125 | 0.8667 | 16 | 0.8454 | 0.8592 | 0.8438 | 0.8497 | 97 | 0.8509 | 0.8454 | 0.8465 | 97 | 0.7694 | [[0, 1, 2, 3], [0, 23, 2, 0, 0], [1, 4, 22, 2, 0], [2, 0, 3, 24, 1], [3, 0, 3, 0, 13]] |
0.2761 | 16.65 | 400 | 0.2263 | 1.0 | 1.0 | 1.0 | 25 | 1.0 | 0.9643 | 0.9818 | 28 | 1.0 | 0.9643 | 0.9818 | 28 | 0.8889 | 1.0 | 0.9412 | 16 | 0.9794 | 0.9722 | 0.9821 | 0.9762 | 97 | 0.9817 | 0.9794 | 0.9798 | 97 | 0.4392 | [[0, 1, 2, 3], [0, 25, 0, 0, 0], [1, 0, 27, 0, 1], [2, 0, 0, 27, 1], [3, 0, 0, 0, 16]] |
0.1596 | 20.82 | 500 | 0.2283 | 1.0 | 0.96 | 0.9796 | 25 | 0.9310 | 0.9643 | 0.9474 | 28 | 0.9643 | 0.9643 | 0.9643 | 28 | 0.9375 | 0.9375 | 0.9375 | 16 | 0.9588 | 0.9582 | 0.9565 | 0.9572 | 97 | 0.9595 | 0.9588 | 0.9589 | 97 | 0.4971 | [[0, 1, 2, 3], [0, 24, 1, 0, 0], [1, 0, 27, 1, 0], [2, 0, 0, 27, 1], [3, 0, 1, 0, 15]] |
0.124 | 24.98 | 600 | 0.1841 | 1.0 | 0.96 | 0.9796 | 25 | 0.9655 | 1.0 | 0.9825 | 28 | 1.0 | 0.9643 | 0.9818 | 28 | 0.9412 | 1.0 | 0.9697 | 16 | 0.9794 | 0.9767 | 0.9811 | 0.9784 | 97 | 0.9803 | 0.9794 | 0.9794 | 97 | 0.2955 | [[0, 1, 2, 3], [0, 24, 1, 0, 0], [1, 0, 28, 0, 0], [2, 0, 0, 27, 1], [3, 0, 0, 0, 16]] |
0.1162 | 29.16 | 700 | 0.2286 | 1.0 | 0.96 | 0.9796 | 25 | 0.9333 | 1.0 | 0.9655 | 28 | 1.0 | 0.9286 | 0.9630 | 28 | 0.9412 | 1.0 | 0.9697 | 16 | 0.9691 | 0.9686 | 0.9721 | 0.9694 | 97 | 0.9711 | 0.9691 | 0.9691 | 97 | 0.3627 | [[0, 1, 2, 3], [0, 24, 1, 0, 0], [1, 0, 28, 0, 0], [2, 0, 1, 26, 1], [3, 0, 0, 0, 16]] |
0.1576 | 33.33 | 800 | 0.2259 | 1.0 | 0.92 | 0.9583 | 25 | 0.9333 | 1.0 | 0.9655 | 28 | 1.0 | 0.9643 | 0.9818 | 28 | 0.9412 | 1.0 | 0.9697 | 16 | 0.9691 | 0.9686 | 0.9711 | 0.9688 | 97 | 0.9711 | 0.9691 | 0.9691 | 97 | 0.3210 | [[0, 1, 2, 3], [0, 23, 2, 0, 0], [1, 0, 28, 0, 0], [2, 0, 0, 27, 1], [3, 0, 0, 0, 16]] |
0.0957 | 37.49 | 900 | 0.2757 | 1.0 | 0.96 | 0.9796 | 25 | 0.9643 | 0.9643 | 0.9643 | 28 | 0.9643 | 0.9643 | 0.9643 | 28 | 0.9412 | 1.0 | 0.9697 | 16 | 0.9691 | 0.9674 | 0.9721 | 0.9695 | 97 | 0.9697 | 0.9691 | 0.9691 | 97 | 0.3499 | [[0, 1, 2, 3], [0, 24, 1, 0, 0], [1, 0, 27, 1, 0], [2, 0, 0, 27, 1], [3, 0, 0, 0, 16]] |
0.1145 | 41.65 | 1000 | 0.2951 | 1.0 | 0.96 | 0.9796 | 25 | 1.0 | 0.9643 | 0.9818 | 28 | 1.0 | 0.9643 | 0.9818 | 28 | 0.8421 | 1.0 | 0.9143 | 16 | 0.9691 | 0.9605 | 0.9721 | 0.9644 | 97 | 0.9740 | 0.9691 | 0.9701 | 97 | 0.3024 | [[0, 1, 2, 3], [0, 24, 0, 0, 1], [1, 0, 27, 0, 1], [2, 0, 0, 27, 1], [3, 0, 0, 0, 16]] |
0.121 | 45.82 | 1100 | 0.3262 | 1.0 | 0.96 | 0.9796 | 25 | 1.0 | 0.9643 | 0.9818 | 28 | 1.0 | 0.9643 | 0.9818 | 28 | 0.8421 | 1.0 | 0.9143 | 16 | 0.9691 | 0.9605 | 0.9721 | 0.9644 | 97 | 0.9740 | 0.9691 | 0.9701 | 97 | 0.2885 | [[0, 1, 2, 3], [0, 24, 0, 0, 1], [1, 0, 27, 0, 1], [2, 0, 0, 27, 1], [3, 0, 0, 0, 16]] |
0.079 | 49.98 | 1200 | 0.3615 | 1.0 | 0.96 | 0.9796 | 25 | 0.9643 | 0.9643 | 0.9643 | 28 | 1.0 | 0.9643 | 0.9818 | 28 | 0.8889 | 1.0 | 0.9412 | 16 | 0.9691 | 0.9633 | 0.9721 | 0.9667 | 97 | 0.9714 | 0.9691 | 0.9695 | 97 | 0.3615 | [[0, 1, 2, 3], [0, 24, 1, 0, 0], [1, 0, 27, 0, 1], [2, 0, 0, 27, 1], [3, 0, 0, 0, 16]] |
0.0733 | 54.16 | 1300 | 0.3891 | 1.0 | 0.96 | 0.9796 | 25 | 0.9643 | 0.9643 | 0.9643 | 28 | 1.0 | 0.9643 | 0.9818 | 28 | 0.8889 | 1.0 | 0.9412 | 16 | 0.9691 | 0.9633 | 0.9721 | 0.9667 | 97 | 0.9714 | 0.9691 | 0.9695 | 97 | 0.3082 | [[0, 1, 2, 3], [0, 24, 1, 0, 0], [1, 0, 27, 0, 1], [2, 0, 0, 27, 1], [3, 0, 0, 0, 16]] |
0.0962 | 58.33 | 1400 | 0.3620 | 1.0 | 0.96 | 0.9796 | 25 | 0.9643 | 0.9643 | 0.9643 | 28 | 1.0 | 0.9643 | 0.9818 | 28 | 0.8889 | 1.0 | 0.9412 | 16 | 0.9691 | 0.9633 | 0.9721 | 0.9667 | 97 | 0.9714 | 0.9691 | 0.9695 | 97 | 0.2851 | [[0, 1, 2, 3], [0, 24, 1, 0, 0], [1, 0, 27, 0, 1], [2, 0, 0, 27, 1], [3, 0, 0, 0, 16]] |
0.0628 | 62.49 | 1500 | 0.4084 | 1.0 | 0.96 | 0.9796 | 25 | 0.9630 | 0.9286 | 0.9455 | 28 | 0.9643 | 0.9643 | 0.9643 | 28 | 0.8889 | 1.0 | 0.9412 | 16 | 0.9588 | 0.9540 | 0.9632 | 0.9576 | 97 | 0.9607 | 0.9588 | 0.9590 | 97 | 0.3001 | [[0, 1, 2, 3], [0, 24, 1, 0, 0], [1, 0, 26, 1, 1], [2, 0, 0, 27, 1], [3, 0, 0, 0, 16]] |
0.0675 | 66.65 | 1600 | 0.4231 | 1.0 | 0.96 | 0.9796 | 25 | 0.9643 | 0.9643 | 0.9643 | 28 | 1.0 | 0.9643 | 0.9818 | 28 | 0.8889 | 1.0 | 0.9412 | 16 | 0.9691 | 0.9633 | 0.9721 | 0.9667 | 97 | 0.9714 | 0.9691 | 0.9695 | 97 | 0.2827 | [[0, 1, 2, 3], [0, 24, 1, 0, 0], [1, 0, 27, 0, 1], [2, 0, 0, 27, 1], [3, 0, 0, 0, 16]] |
Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0+cu116
- Datasets 2.8.0
- Tokenizers 0.13.2