metadata
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: wav2vec2-large-xlsr-coraa-exp-2
results: []
wav2vec2-large-xlsr-coraa-exp-2
This model is a fine-tuned version of Edresson/wav2vec2-large-xlsr-coraa-portuguese on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 9.0840
- Wer: 1.0
- Cer: 0.9619
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: 3e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 150
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 38.902 | 1.0 | 14 | 30.9943 | 0.9994 | 0.9583 |
| 38.902 | 2.0 | 28 | 13.1446 | 1.0 | 0.9619 |
| 38.902 | 3.0 | 42 | 10.9579 | 1.0 | 0.9619 |
| 38.902 | 4.0 | 56 | 10.3935 | 1.0 | 0.9619 |
| 38.902 | 5.0 | 70 | 9.9203 | 1.0 | 0.9619 |
| 38.902 | 6.0 | 84 | 9.4185 | 1.0 | 0.9619 |
| 38.902 | 7.0 | 98 | 9.0840 | 1.0 | 0.9619 |
| 11.1539 | 8.0 | 112 | 9.0912 | 0.9994 | 0.9610 |
| 11.1539 | 9.0 | 126 | 9.1239 | 0.9850 | 0.9137 |
| 11.1539 | 10.0 | 140 | 9.1313 | 0.9809 | 0.9035 |
| 11.1539 | 11.0 | 154 | 9.2100 | 0.9868 | 0.8705 |
| 11.1539 | 12.0 | 168 | 9.1043 | 0.9829 | 0.8711 |
| 11.1539 | 13.0 | 182 | 9.4617 | 0.9793 | 0.9264 |
| 11.1539 | 14.0 | 196 | 9.6228 | 0.9783 | 0.9235 |
| 3.9522 | 15.0 | 210 | 9.3262 | 0.9774 | 0.8737 |
| 3.9522 | 16.0 | 224 | 9.4366 | 0.9764 | 0.8773 |
| 3.9522 | 17.0 | 238 | 9.7235 | 0.9758 | 0.9005 |
| 3.9522 | 18.0 | 252 | 9.9012 | 0.9750 | 0.8971 |
| 3.9522 | 19.0 | 266 | 9.7519 | 0.9846 | 0.8567 |
| 3.9522 | 20.0 | 280 | 9.8620 | 0.9764 | 0.8976 |
| 3.9522 | 21.0 | 294 | 9.8945 | 0.9783 | 0.8681 |
| 3.3723 | 22.0 | 308 | 9.9917 | 0.9748 | 0.8706 |
| 3.3723 | 23.0 | 322 | 10.0580 | 0.9801 | 0.8577 |
| 3.3723 | 24.0 | 336 | 10.1575 | 0.9764 | 0.8708 |
| 3.3723 | 25.0 | 350 | 10.1982 | 0.9819 | 0.8555 |
| 3.3723 | 26.0 | 364 | 10.4087 | 0.9791 | 0.9087 |
| 3.3723 | 27.0 | 378 | 10.5064 | 0.9799 | 0.8902 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.4.1+cu121
- Datasets 3.2.0
- Tokenizers 0.13.3