|
|
--- |
|
|
library_name: transformers |
|
|
license: mit |
|
|
base_model: openai/whisper-large-v3-turbo |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
datasets: |
|
|
- fleurs |
|
|
metrics: |
|
|
- wer |
|
|
model-index: |
|
|
- name: whisper-large-v3-turbo-serbian-serbia |
|
|
results: |
|
|
- task: |
|
|
name: Automatic Speech Recognition |
|
|
type: automatic-speech-recognition |
|
|
dataset: |
|
|
name: fleurs |
|
|
type: fleurs |
|
|
config: sr_rs |
|
|
split: None |
|
|
args: sr_rs |
|
|
metrics: |
|
|
- name: Wer |
|
|
type: wer |
|
|
value: 18.343921455385505 |
|
|
--- |
|
|
|
|
|
<!-- 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. --> |
|
|
|
|
|
# whisper-large-v3-turbo-serbian-serbia |
|
|
|
|
|
This model is a fine-tuned version of [openai/whisper-large-v3-turbo](https://huggingface.co/openai/whisper-large-v3-turbo) on the fleurs dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 0.5289 |
|
|
- Model Preparation Time: 0.0069 |
|
|
- Wer Ortho: 34.0354 |
|
|
- Wer: 18.3439 |
|
|
|
|
|
## 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: 1e-05 |
|
|
- train_batch_size: 64 |
|
|
- eval_batch_size: 64 |
|
|
- seed: 42 |
|
|
- gradient_accumulation_steps: 2 |
|
|
- total_train_batch_size: 128 |
|
|
- 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_ratio: 0.06 |
|
|
- num_epochs: 1 |
|
|
- mixed_precision_training: Native AMP |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Model Preparation Time | Wer Ortho | Wer | |
|
|
|:-------------:|:------:|:----:|:---------------:|:----------------------:|:---------:|:-------:| |
|
|
| 0.1827 | 0.0263 | 32 | 0.8340 | 0.0069 | 36.3524 | 21.6142 | |
|
|
| 0.1531 | 0.0527 | 64 | 0.6341 | 0.0069 | 36.4759 | 21.7225 | |
|
|
| 0.1551 | 0.0790 | 96 | 0.6893 | 0.0069 | 36.0038 | 21.6647 | |
|
|
| 0.1495 | 0.1054 | 128 | 0.5745 | 0.0069 | 35.4518 | 21.2388 | |
|
|
| 0.1388 | 0.1317 | 160 | 0.5572 | 0.0069 | 35.5680 | 20.5241 | |
|
|
| 0.1359 | 0.1581 | 192 | 0.6313 | 0.0069 | 36.2217 | 20.6974 | |
|
|
| 0.138 | 0.1844 | 224 | 0.5451 | 0.0069 | 35.9384 | 20.4014 | |
|
|
| 0.1332 | 0.2108 | 256 | 0.6109 | 0.0069 | 35.5171 | 20.2209 | |
|
|
| 0.131 | 0.2371 | 288 | 0.5748 | 0.0069 | 35.7859 | 20.6324 | |
|
|
| 0.1316 | 0.2635 | 320 | 0.5705 | 0.0069 | 34.9942 | 20.1559 | |
|
|
| 0.1296 | 0.2898 | 352 | 0.6122 | 0.0069 | 35.3791 | 20.0332 | |
|
|
| 0.1274 | 0.3162 | 384 | 0.5848 | 0.0069 | 35.1322 | 19.8960 | |
|
|
| 0.1185 | 0.3425 | 416 | 0.5458 | 0.0069 | 35.4227 | 19.8239 | |
|
|
| 0.1207 | 0.3689 | 448 | 0.5714 | 0.0069 | 34.8053 | 19.6289 | |
|
|
| 0.1193 | 0.3952 | 480 | 0.6553 | 0.0069 | 34.7327 | 19.3546 | |
|
|
| 0.119 | 0.4216 | 512 | 0.5913 | 0.0069 | 35.0523 | 19.4629 | |
|
|
| 0.122 | 0.4479 | 544 | 0.5890 | 0.0069 | 34.8489 | 19.4846 | |
|
|
| 0.1203 | 0.4743 | 576 | 0.5554 | 0.0069 | 35.0886 | 19.5206 | |
|
|
| 0.1164 | 0.5006 | 608 | 0.5637 | 0.0069 | 34.7690 | 19.4124 | |
|
|
| 0.1173 | 0.5270 | 640 | 0.5492 | 0.0069 | 34.6601 | 19.2319 | |
|
|
| 0.1175 | 0.5533 | 672 | 0.5847 | 0.0069 | 34.4785 | 19.3474 | |
|
|
| 0.1154 | 0.5797 | 704 | 0.5900 | 0.0069 | 34.7618 | 18.8926 | |
|
|
| 0.1136 | 0.6060 | 736 | 0.6104 | 0.0069 | 34.6092 | 19.0586 | |
|
|
| 0.1142 | 0.6324 | 768 | 0.5393 | 0.0069 | 34.2533 | 18.8276 | |
|
|
| 0.1156 | 0.6587 | 800 | 0.5473 | 0.0069 | 34.5221 | 18.8926 | |
|
|
| 0.1125 | 0.6851 | 832 | 0.5747 | 0.0069 | 34.4858 | 18.8276 | |
|
|
| 0.1107 | 0.7114 | 864 | 0.5470 | 0.0069 | 34.1880 | 18.8709 | |
|
|
| 0.1088 | 0.7378 | 896 | 0.5620 | 0.0069 | 34.4785 | 18.8926 | |
|
|
| 0.1111 | 0.7641 | 928 | 0.5832 | 0.0069 | 34.4059 | 18.8493 | |
|
|
| 0.1081 | 0.7904 | 960 | 0.5430 | 0.0069 | 34.3623 | 18.8204 | |
|
|
| 0.108 | 0.8168 | 992 | 0.5560 | 0.0069 | 34.4567 | 18.7626 | |
|
|
| 0.106 | 0.8431 | 1024 | 0.5344 | 0.0069 | 34.3550 | 18.8637 | |
|
|
| 0.1051 | 0.8695 | 1056 | 0.5496 | 0.0069 | 34.4567 | 18.5894 | |
|
|
| 0.1096 | 0.8958 | 1088 | 0.5341 | 0.0069 | 34.1008 | 18.4811 | |
|
|
| 0.1098 | 0.9222 | 1120 | 0.5339 | 0.0069 | 34.2025 | 18.4306 | |
|
|
| 0.1045 | 0.9485 | 1152 | 0.5289 | 0.0069 | 34.0354 | 18.3439 | |
|
|
| 0.1072 | 0.9749 | 1184 | 0.5265 | 0.0069 | 34.0863 | 18.4089 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.51.3 |
|
|
- Pytorch 2.5.1 |
|
|
- Datasets 3.6.0 |
|
|
- Tokenizers 0.21.1 |
|
|
|
