|
|
--- |
|
|
library_name: transformers |
|
|
license: apache-2.0 |
|
|
base_model: facebook/wav2vec2-xls-r-1b |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
model-index: |
|
|
- name: wav2vec2-1b-E50_freq |
|
|
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-E50_freq |
|
|
|
|
|
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: 0.4231 |
|
|
- Cer: 11.7951 |
|
|
|
|
|
## 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 | |
|
|
|:-------------:|:------:|:----:|:---------------:|:-------:| |
|
|
| 9.4581 | 0.2580 | 200 | 3.9157 | 68.4387 | |
|
|
| 2.071 | 0.5160 | 400 | 1.6011 | 37.3766 | |
|
|
| 1.307 | 0.7741 | 600 | 1.1918 | 28.6889 | |
|
|
| 1.1363 | 1.0321 | 800 | 1.0170 | 23.9192 | |
|
|
| 0.8977 | 1.2901 | 1000 | 0.8161 | 20.3771 | |
|
|
| 0.8219 | 1.5481 | 1200 | 0.7277 | 18.9908 | |
|
|
| 0.7857 | 1.8062 | 1400 | 0.7119 | 19.1494 | |
|
|
| 0.6967 | 2.0642 | 1600 | 0.6717 | 17.8748 | |
|
|
| 0.5744 | 2.3222 | 1800 | 0.7313 | 19.1259 | |
|
|
| 0.5676 | 2.5802 | 2000 | 0.6092 | 16.3593 | |
|
|
| 0.5362 | 2.8383 | 2200 | 0.5840 | 16.5120 | |
|
|
| 0.4879 | 3.0963 | 2400 | 0.5039 | 13.8745 | |
|
|
| 0.4144 | 3.3543 | 2600 | 0.5449 | 14.6558 | |
|
|
| 0.3875 | 3.6123 | 2800 | 0.4790 | 13.3400 | |
|
|
| 0.3723 | 3.8703 | 3000 | 0.4497 | 12.3238 | |
|
|
| 0.3326 | 4.1284 | 3200 | 0.4624 | 12.3825 | |
|
|
| 0.2905 | 4.3864 | 3400 | 0.4345 | 12.2004 | |
|
|
| 0.2745 | 4.6444 | 3600 | 0.4479 | 12.2768 | |
|
|
| 0.2668 | 4.9024 | 3800 | 0.4231 | 11.7951 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.45.2 |
|
|
- Pytorch 2.3.1.post100 |
|
|
- Datasets 2.19.1 |
|
|
- Tokenizers 0.20.1 |
|
|
|
