|
--- |
|
library_name: transformers |
|
license: apache-2.0 |
|
base_model: facebook/wav2vec2-large-xlsr-53 |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- wer |
|
model-index: |
|
- name: xlsr-a-nomo |
|
results: [] |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# xlsr-a-nomo |
|
|
|
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.6748 |
|
- Wer: 0.3390 |
|
|
|
## 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: 0.0004 |
|
- train_batch_size: 8 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 2 |
|
- total_train_batch_size: 16 |
|
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 132 |
|
- num_epochs: 100 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Wer | |
|
|:-------------:|:-------:|:----:|:---------------:|:------:| |
|
| 4.3851 | 2.3529 | 200 | 2.4270 | 1.0 | |
|
| 1.3933 | 4.7059 | 400 | 0.5187 | 0.6193 | |
|
| 0.2417 | 7.0588 | 600 | 0.4328 | 0.4403 | |
|
| 0.1174 | 9.4118 | 800 | 0.4113 | 0.4186 | |
|
| 0.08 | 11.7647 | 1000 | 0.3914 | 0.3589 | |
|
| 0.0509 | 14.1176 | 1200 | 0.4065 | 0.3665 | |
|
| 0.0445 | 16.4706 | 1400 | 0.4714 | 0.3636 | |
|
| 0.0329 | 18.8235 | 1600 | 0.3930 | 0.3561 | |
|
| 0.0315 | 21.1765 | 1800 | 0.5501 | 0.3655 | |
|
| 0.0298 | 23.5294 | 2000 | 0.4462 | 0.3561 | |
|
| 0.0366 | 25.8824 | 2200 | 0.4993 | 0.3447 | |
|
| 0.0259 | 28.2353 | 2400 | 0.5077 | 0.3561 | |
|
| 0.0197 | 30.5882 | 2600 | 0.5029 | 0.3466 | |
|
| 0.0229 | 32.9412 | 2800 | 0.4760 | 0.3400 | |
|
| 0.0174 | 35.2941 | 3000 | 0.5118 | 0.3475 | |
|
| 0.0102 | 37.6471 | 3200 | 0.5630 | 0.3428 | |
|
| 0.0104 | 40.0 | 3400 | 0.5598 | 0.3400 | |
|
| 0.015 | 42.3529 | 3600 | 0.5226 | 0.3428 | |
|
| 0.0102 | 44.7059 | 3800 | 0.5421 | 0.3513 | |
|
| 0.012 | 47.0588 | 4000 | 0.5936 | 0.3456 | |
|
| 0.0101 | 49.4118 | 4200 | 0.5772 | 0.3485 | |
|
| 0.009 | 51.7647 | 4400 | 0.5759 | 0.3438 | |
|
| 0.0104 | 54.1176 | 4600 | 0.5755 | 0.3400 | |
|
| 0.0118 | 56.4706 | 4800 | 0.5868 | 0.3362 | |
|
| 0.0118 | 58.8235 | 5000 | 0.6174 | 0.3456 | |
|
| 0.0109 | 61.1765 | 5200 | 0.6037 | 0.3390 | |
|
| 0.0086 | 63.5294 | 5400 | 0.5903 | 0.3447 | |
|
| 0.0039 | 65.8824 | 5600 | 0.5894 | 0.3428 | |
|
| 0.0058 | 68.2353 | 5800 | 0.6388 | 0.3428 | |
|
| 0.0044 | 70.5882 | 6000 | 0.6378 | 0.3390 | |
|
| 0.0032 | 72.9412 | 6200 | 0.6868 | 0.3409 | |
|
| 0.0057 | 75.2941 | 6400 | 0.6439 | 0.3419 | |
|
| 0.0034 | 77.6471 | 6600 | 0.6888 | 0.3381 | |
|
| 0.0022 | 80.0 | 6800 | 0.7217 | 0.3381 | |
|
| 0.0032 | 82.3529 | 7000 | 0.5946 | 0.3371 | |
|
| 0.0033 | 84.7059 | 7200 | 0.6650 | 0.3390 | |
|
| 0.0018 | 87.0588 | 7400 | 0.6844 | 0.3371 | |
|
| 0.0026 | 89.4118 | 7600 | 0.7199 | 0.3409 | |
|
| 0.0028 | 91.7647 | 7800 | 0.6868 | 0.3381 | |
|
| 0.002 | 94.1176 | 8000 | 0.6752 | 0.3409 | |
|
| 0.0013 | 96.4706 | 8200 | 0.6788 | 0.3390 | |
|
| 0.0011 | 98.8235 | 8400 | 0.6748 | 0.3390 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.47.0.dev0 |
|
- Pytorch 2.4.0 |
|
- Datasets 3.0.1 |
|
- Tokenizers 0.20.0 |
|
|
