|
--- |
|
license: apache-2.0 |
|
library_name: peft |
|
tags: |
|
- generated_from_trainer |
|
base_model: openai/whisper-small |
|
datasets: |
|
- librispeech |
|
model-index: |
|
- name: Whisper Small English 1h |
|
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. --> |
|
|
|
# Whisper Small English 1h |
|
|
|
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the librispeech dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.3666 |
|
|
|
## 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: 8 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 50 |
|
- num_epochs: 50 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 1.8938 | 1.0 | 39 | 2.5946 | |
|
| 1.4637 | 2.0 | 78 | 2.0459 | |
|
| 1.3003 | 3.0 | 117 | 1.6739 | |
|
| 0.9415 | 4.0 | 156 | 1.2729 | |
|
| 0.8165 | 5.0 | 195 | 1.0158 | |
|
| 0.6326 | 6.0 | 234 | 0.9033 | |
|
| 0.5716 | 7.0 | 273 | 0.7272 | |
|
| 0.4662 | 8.0 | 312 | 0.6731 | |
|
| 0.4133 | 9.0 | 351 | 0.6433 | |
|
| 0.4 | 10.0 | 390 | 0.6248 | |
|
| 0.3862 | 11.0 | 429 | 0.6103 | |
|
| 0.3901 | 12.0 | 468 | 0.5962 | |
|
| 0.3661 | 13.0 | 507 | 0.5841 | |
|
| 0.3609 | 14.0 | 546 | 0.5739 | |
|
| 0.3439 | 15.0 | 585 | 0.5660 | |
|
| 0.3391 | 16.0 | 624 | 0.5581 | |
|
| 0.3231 | 17.0 | 663 | 0.5510 | |
|
| 0.3097 | 18.0 | 702 | 0.5441 | |
|
| 0.2994 | 19.0 | 741 | 0.5373 | |
|
| 0.2991 | 20.0 | 780 | 0.5304 | |
|
| 0.2972 | 21.0 | 819 | 0.5240 | |
|
| 0.2898 | 22.0 | 858 | 0.5187 | |
|
| 0.2809 | 23.0 | 897 | 0.5142 | |
|
| 0.2845 | 24.0 | 936 | 0.5119 | |
|
| 0.269 | 25.0 | 975 | 0.5074 | |
|
| 0.2721 | 26.0 | 1014 | 0.5033 | |
|
| 0.2633 | 27.0 | 1053 | 0.5006 | |
|
| 0.2623 | 28.0 | 1092 | 0.4984 | |
|
| 0.2492 | 29.0 | 1131 | 0.4931 | |
|
| 0.25 | 30.0 | 1170 | 0.4861 | |
|
| 0.2479 | 31.0 | 1209 | 0.4833 | |
|
| 0.2416 | 32.0 | 1248 | 0.4777 | |
|
| 0.2356 | 33.0 | 1287 | 0.4794 | |
|
| 0.2281 | 34.0 | 1326 | 0.4663 | |
|
| 0.2191 | 35.0 | 1365 | 0.4605 | |
|
| 0.2218 | 36.0 | 1404 | 0.4600 | |
|
| 0.2078 | 37.0 | 1443 | 0.4545 | |
|
| 0.2122 | 38.0 | 1482 | 0.4470 | |
|
| 0.2076 | 39.0 | 1521 | 0.4510 | |
|
| 0.2004 | 40.0 | 1560 | 0.4326 | |
|
| 0.2004 | 41.0 | 1599 | 0.4280 | |
|
| 0.1901 | 42.0 | 1638 | 0.4342 | |
|
| 0.1856 | 43.0 | 1677 | 0.4107 | |
|
| 0.1802 | 44.0 | 1716 | 0.4060 | |
|
| 0.1677 | 45.0 | 1755 | 0.4029 | |
|
| 0.1658 | 46.0 | 1794 | 0.3922 | |
|
| 0.1589 | 47.0 | 1833 | 0.3845 | |
|
| 0.152 | 48.0 | 1872 | 0.3790 | |
|
| 0.1493 | 49.0 | 1911 | 0.3691 | |
|
| 0.1426 | 50.0 | 1950 | 0.3666 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.10.0 |
|
- Transformers 4.40.1 |
|
- Pytorch 2.3.0+cu121 |
|
- Datasets 2.19.0 |
|
- Tokenizers 0.19.1 |
