shuaijiang's picture
Update README.md
fc52a0c verified
|
raw
history blame
2.98 kB
---
license: apache-2.0
metrics:
- cer
---
## Welcome
If you find this model helpful, please *like* this model and star us on https://github.com/LianjiaTech/BELLE and https://github.com/shuaijiang/Whisper-Finetune
# Belle-whisper-large-v3-turbo-zh
Fine tune whisper-large-v3-turbo-zh to enhance Chinese speech recognition capabilities,
Belle-whisper-large-v3-turbo-zh demonstrates a 31~64% relative improvement in performance to whisper-large-v3-turbo on Chinese ASR benchmarks, including AISHELL1, AISHELL2, WENETSPEECH, and HKUST.
Same to Belle-whisper-large-v3-zh-punct, the punctuation marks come from model [punc_ct-transformer_cn-en-common-vocab471067-large](https://www.modelscope.cn/models/iic/punc_ct-transformer_cn-en-common-vocab471067-large/),
and are added to the training datasets.
## Usage
```python
from transformers import pipeline
transcriber = pipeline(
"automatic-speech-recognition",
model="BELLE-2/Belle-whisper-large-v3-turbo-zh"
)
transcriber.model.config.forced_decoder_ids = (
transcriber.tokenizer.get_decoder_prompt_ids(
language="zh",
task="transcribe"
)
)
transcription = transcriber("my_audio.wav")
```
## Fine-tuning
| Model | (Re)Sample Rate | Train Datasets | Fine-tuning (full or peft) |
|:----------------:|:-------:|:----------------------------------------------------------:|:-----------:|
| Belle-whisper-large-v3-turbo-zh | 16KHz | [AISHELL-1](https://openslr.magicdatatech.com/resources/33/) [AISHELL-2](https://www.aishelltech.com/aishell_2) [WenetSpeech](https://wenet.org.cn/WenetSpeech/) [HKUST](https://catalog.ldc.upenn.edu/LDC2005S15) | [full fine-tuning](https://github.com/shuaijiang/Whisper-Finetune) |
To incorporate punctuation marks without compromising performance, Lora fine-tuning was employed.
If you want to fine-thuning the model on your datasets, please reference to the [github repo](https://github.com/shuaijiang/Whisper-Finetune)
## CER(%) ↓
| Model | Language Tag | aishell_1_test(↓) |aishell_2_test(↓)| wenetspeech_net(↓) | wenetspeech_meeting(↓) | HKUST_dev(↓)|
|:----------------:|:-------:|:-----------:|:-----------:|:--------:|:-----------:|:-------:|
| whisper-large-v3 | Chinese | 8.085 | 5.475 | 11.72 | 20.15 | 28.597 |
| whisper-large-v3-turbo | Chinese | 8.639 | 6.014 | 13.507 | 20.313 | 37.324 |
| Belle-whisper-large-v3-turbo-zh | Chinese | 3.07 | 4.114 | 8.x | 13.357 | 18.944 |
It is worth mentioning that compared to whisper-large-v3 and whisper-large-v3-turbo, Belle-whisper-large-v3-turbo-zh has a significant improvement.
## Citation
Please cite our paper and github when using our code, data or model.
```
@misc{BELLE,
author = {BELLEGroup},
title = {BELLE: Be Everyone's Large Language model Engine},
year = {2023},
publisher = {GitHub},
journal = {GitHub repository},
howpublished = {\url{https://github.com/LianjiaTech/BELLE}},
}
```