|
--- |
|
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-nomi |
|
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-nomi |
|
|
|
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: 0.3688 |
|
- Wer: 0.3324 |
|
|
|
## 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 | |
|
|:-------------:|:-------:|:----:|:---------------:|:------:| |
|
| 4.7382 | 2.2727 | 200 | 2.5107 | 1.0 | |
|
| 1.3613 | 4.5455 | 400 | 0.3782 | 0.5943 | |
|
| 0.2498 | 6.8182 | 600 | 0.2562 | 0.4209 | |
|
| 0.1205 | 9.0909 | 800 | 0.2575 | 0.3548 | |
|
| 0.0772 | 11.3636 | 1000 | 0.2902 | 0.3432 | |
|
| 0.0625 | 13.6364 | 1200 | 0.3199 | 0.3458 | |
|
| 0.0461 | 15.9091 | 1400 | 0.2814 | 0.3351 | |
|
| 0.0348 | 18.1818 | 1600 | 0.3389 | 0.3396 | |
|
| 0.0323 | 20.4545 | 1800 | 0.3000 | 0.3423 | |
|
| 0.0341 | 22.7273 | 2000 | 0.3097 | 0.3342 | |
|
| 0.0271 | 25.0 | 2200 | 0.3270 | 0.3342 | |
|
| 0.0236 | 27.2727 | 2400 | 0.3370 | 0.3423 | |
|
| 0.0245 | 29.5455 | 2600 | 0.3201 | 0.3387 | |
|
| 0.0143 | 31.8182 | 2800 | 0.3483 | 0.3315 | |
|
| 0.0183 | 34.0909 | 3000 | 0.3245 | 0.3333 | |
|
| 0.0149 | 36.3636 | 3200 | 0.3269 | 0.3342 | |
|
| 0.0128 | 38.6364 | 3400 | 0.3180 | 0.3324 | |
|
| 0.0121 | 40.9091 | 3600 | 0.3465 | 0.3387 | |
|
| 0.0145 | 43.1818 | 3800 | 0.3465 | 0.3378 | |
|
| 0.014 | 45.4545 | 4000 | 0.3181 | 0.3342 | |
|
| 0.0101 | 47.7273 | 4200 | 0.3438 | 0.3333 | |
|
| 0.0057 | 50.0 | 4400 | 0.3405 | 0.3387 | |
|
| 0.0101 | 52.2727 | 4600 | 0.3508 | 0.3396 | |
|
| 0.0084 | 54.5455 | 4800 | 0.3602 | 0.3360 | |
|
| 0.0057 | 56.8182 | 5000 | 0.3369 | 0.3378 | |
|
| 0.0143 | 59.0909 | 5200 | 0.3584 | 0.3387 | |
|
| 0.0062 | 61.3636 | 5400 | 0.3748 | 0.3360 | |
|
| 0.0068 | 63.6364 | 5600 | 0.3625 | 0.3369 | |
|
| 0.006 | 65.9091 | 5800 | 0.3773 | 0.3369 | |
|
| 0.0059 | 68.1818 | 6000 | 0.3666 | 0.3351 | |
|
| 0.008 | 70.4545 | 6200 | 0.3597 | 0.3378 | |
|
| 0.0061 | 72.7273 | 6400 | 0.3703 | 0.3396 | |
|
| 0.0041 | 75.0 | 6600 | 0.3843 | 0.3387 | |
|
| 0.0055 | 77.2727 | 6800 | 0.3829 | 0.3360 | |
|
| 0.0028 | 79.5455 | 7000 | 0.3877 | 0.3378 | |
|
| 0.0025 | 81.8182 | 7200 | 0.3898 | 0.3333 | |
|
| 0.0021 | 84.0909 | 7400 | 0.3910 | 0.3342 | |
|
| 0.0021 | 86.3636 | 7600 | 0.3889 | 0.3360 | |
|
| 0.0025 | 88.6364 | 7800 | 0.3871 | 0.3342 | |
|
| 0.0025 | 90.9091 | 8000 | 0.3787 | 0.3333 | |
|
| 0.0016 | 93.1818 | 8200 | 0.3676 | 0.3307 | |
|
| 0.0017 | 95.4545 | 8400 | 0.3651 | 0.3307 | |
|
| 0.0015 | 97.7273 | 8600 | 0.3685 | 0.3324 | |
|
| 0.0015 | 100.0 | 8800 | 0.3688 | 0.3324 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.47.0.dev0 |
|
- Pytorch 2.4.0 |
|
- Datasets 3.0.1 |
|
- Tokenizers 0.20.0 |
|
|
