README.md

---
license: apache-2.0
datasets:
- pongjin/en_corpora_parliament_processed
language:
- en
pipeline_tag: automatic-speech-recognition
metrics:
- wer
---
**This model has been referred to the following links**  
1) https://huggingface.co/blog/wav2vec2-with-ngram
2) https://huggingface.co/blog/fine-tune-wav2vec2-english  

Thanks to [patrickvonplaten Patrick von Platen](https://huggingface.co/patrickvonplaten)

해당 모델은 한국인의 영어 발화 인식 성능 개선을 위해 facebook/wav2vec2-large-960h 로 파인튜닝한 모델에 KenLM 5-gram 을 붙인 ASR + LM 모델입니다.

If you want to use LM, you must have kenlm installed https://github.com/kpu/kenlm  
```python
pip install https://github.com/kpu/kenlm/archive/master.zip
```
학습 데이터 출처 : https://aiopen.etri.re.kr/voiceModel

>transformers==4.24.0  
>huggingface_hub==0.13.2

 | wer | epoch | batch | lr | weight_decay| warmup_steps|
 | --- | --- | --- | --- | --- | --- |
 | 0.17 | 10 | 16 | 1e-4 | 0.005 | 1000 |