|
|
--- |
|
|
library_name: transformers |
|
|
license: apache-2.0 |
|
|
base_model: facebook/wav2vec2-xls-r-1b |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
model-index: |
|
|
- name: wav2vec2-1b-Yspeed_pause |
|
|
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. --> |
|
|
|
|
|
# wav2vec2-1b-Yspeed_pause |
|
|
|
|
|
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on an unknown dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 1.2373 |
|
|
- Cer: 33.3529 |
|
|
|
|
|
## 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: 2 |
|
|
- eval_batch_size: 8 |
|
|
- seed: 42 |
|
|
- gradient_accumulation_steps: 8 |
|
|
- 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: 50 |
|
|
- num_epochs: 5 |
|
|
- mixed_precision_training: Native AMP |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Cer | |
|
|
|:-------------:|:------:|:----:|:---------------:|:-------:| |
|
|
| 9.4479 | 0.2581 | 200 | 3.2207 | 66.0714 | |
|
|
| 1.7868 | 0.5161 | 400 | 2.8650 | 64.3856 | |
|
|
| 1.3863 | 0.7742 | 600 | 2.1014 | 51.3158 | |
|
|
| 1.2139 | 1.0323 | 800 | 1.9113 | 48.1849 | |
|
|
| 1.1064 | 1.2903 | 1000 | 1.8469 | 48.8193 | |
|
|
| 0.9283 | 1.5484 | 1200 | 1.6672 | 43.8087 | |
|
|
| 0.874 | 1.8065 | 1400 | 1.6412 | 44.4608 | |
|
|
| 0.8219 | 2.0645 | 1600 | 1.7808 | 45.8882 | |
|
|
| 0.7202 | 2.3226 | 1800 | 1.6440 | 43.6384 | |
|
|
| 0.6672 | 2.5806 | 2000 | 1.5708 | 41.7411 | |
|
|
| 0.6327 | 2.8387 | 2200 | 1.4854 | 42.5223 | |
|
|
| 0.57 | 3.0968 | 2400 | 1.4387 | 39.2152 | |
|
|
| 0.469 | 3.3548 | 2600 | 1.3790 | 38.0169 | |
|
|
| 0.4253 | 3.6129 | 2800 | 1.4179 | 38.1520 | |
|
|
| 0.398 | 3.8710 | 3000 | 1.2919 | 36.1020 | |
|
|
| 0.3415 | 4.1290 | 3200 | 1.4286 | 37.2357 | |
|
|
| 0.3044 | 4.3871 | 3400 | 1.4068 | 36.1666 | |
|
|
| 0.2848 | 4.6452 | 3600 | 1.2308 | 33.3353 | |
|
|
| 0.277 | 4.9032 | 3800 | 1.2373 | 33.3529 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.45.2 |
|
|
- Pytorch 2.3.1.post100 |
|
|
- Datasets 2.19.1 |
|
|
- Tokenizers 0.20.1 |
|
|
|
