khanhld/wav2vec2-base-vietnamese-160h
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity
Files
main
wav2vec2-base-vietnamese-160h/README.md
ModelHub XC 4c37b13fbc 初始化项目，由ModelHub XC社区提供模型 
Model: khanhld/wav2vec2-base-vietnamese-160h
Source: Original Platform
2026-05-14 22:35:55 +08:00

179 lines
6.6 KiB
Markdown
Raw Permalink Blame History

---
language: vi
datasets:
- vivos
- common_voice
- FOSD
- VLSP
metrics:
- wer
pipeline_tag: automatic-speech-recognition
tags:
- audio
- speech
- Transformer
- wav2vec2
- automatic-speech-recognition
- vietnamese
license: cc-by-nc-4.0
widget:
- example_title: common_voice_vi_30519758.mp3
src: https://huggingface.co/khanhld/wav2vec2-base-vietnamese-160h/raw/main/examples/common_voice_vi_30519758.mp3
- example_title: VIVOSDEV15_020.wav
src: https://huggingface.co/khanhld/wav2vec2-base-vietnamese-160h/raw/main/examples/VIVOSDEV15_020.wav
model-index:
- name: Wav2vec2 Base Vietnamese 160h
results:
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: common-voice-vietnamese
type: common_voice
args: vi
metrics:
- name: Test WER
type: wer
value: 10.78
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: VIVOS
type: vivos
args: vi
metrics:
- name: Test WER
type: wer
value: 15.05
---
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/wav2vec2-base-vietnamese-160h/speech-recognition-on-common-voice-vi)](https://paperswithcode.com/sota/speech-recognition-on-common-voice-vi?p=wav2vec2-base-vietnamese-160h)
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/wav2vec2-base-vietnamese-160h/speech-recognition-on-vivos)](https://paperswithcode.com/sota/speech-recognition-on-vivos?p=wav2vec2-base-vietnamese-160h)
# Vietnamese Speech Recognition using Wav2vec 2.0
### Table of contents
1. [Model Description](#description)
2. [Implementation](#implementation)
3. [Benchmark Result](#benchmark)
4. [Example Usage](#example)
5. [Evaluation](#evaluation)
6. [Citation](#citation)
7. [Contact](#contact)
<a name = "description" ></a>
### Model Description
Fine-tuned the Wav2vec2-based model on about 160 hours of Vietnamese speech dataset from different resources, including [VIOS](https://huggingface.co/datasets/vivos), [COMMON VOICE](https://huggingface.co/datasets/mozilla-foundation/common_voice_8_0), [FOSD](https://data.mendeley.com/datasets/k9sxg2twv4/4) and [VLSP 100h](https://drive.google.com/file/d/1vUSxdORDxk-ePUt-bUVDahpoXiqKchMx/view). We have not yet incorporated the Language Model into our ASR system but still gained a promising result.
<a name = "implementation" ></a>
### Implementation
We also provide code for Pre-training and Fine-tuning the Wav2vec2 model. If you wish to train on your dataset, check it out here:
- [Pre-train code](https://github.com/khanld/Wav2vec2-Pretraining)
- [Fine-tune code](https://github.com/khanld/ASR-Wa2vec-Finetune)
<a name = "benchmark" ></a>
### Benchmark WER Result
| | [VIVOS](https://huggingface.co/datasets/vivos) | [COMMON VOICE 8.0](https://huggingface.co/datasets/mozilla-foundation/common_voice_8_0) |
|---|---|---|
|without LM| 15.05 | 10.78 |
|with LM| in progress | in progress |
<a name = "example" ></a>
### Example Usage [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1blz1KclnIfbOp8o2fW3WJgObOQ9SMGBo?usp=sharing)
```python
from transformers import Wav2Vec2Processor, Wav2Vec2ForCTC
import librosa
import torch
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
processor = Wav2Vec2Processor.from_pretrained("khanhld/wav2vec2-base-vietnamese-160h")
model = Wav2Vec2ForCTC.from_pretrained("khanhld/wav2vec2-base-vietnamese-160h")
model.to(device)
def transcribe(wav):
input_values = processor(wav, sampling_rate=16000, return_tensors="pt").input_values
logits = model(input_values.to(device)).logits
pred_ids = torch.argmax(logits, dim=-1)
pred_transcript = processor.batch_decode(pred_ids)[0]
return pred_transcript
wav, _ = librosa.load('path/to/your/audio/file', sr = 16000)
print(f"transcript: {transcribe(wav)}")
```
<a name = "evaluation"></a>
### Evaluation [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1XQCq4YGLnl23tcKmYeSwaksro4IgC_Yi?usp=sharing)
```python
from transformers import Wav2Vec2Processor, Wav2Vec2ForCTC
from datasets import load_dataset
import torch
import re
from datasets import load_dataset, load_metric, Audio
wer = load_metric("wer")
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
# load processor and model
processor = Wav2Vec2Processor.from_pretrained("khanhld/wav2vec2-base-vietnamese-160h")
model = Wav2Vec2ForCTC.from_pretrained("khanhld/wav2vec2-base-vietnamese-160h")
model.to(device)
model.eval()
# Load dataset
test_dataset = load_dataset("mozilla-foundation/common_voice_8_0", "vi", split="test", use_auth_token="your_huggingface_auth_token")
test_dataset = test_dataset.cast_column("audio", Audio(sampling_rate=16000))
chars_to_ignore = r'[,?.!\-;:"“%\'<EFBFBD>]' # ignore special characters
# preprocess data
def preprocess(batch):
audio = batch["audio"]
batch["input_values"] = audio["array"]
batch["transcript"] = re.sub(chars_to_ignore, '', batch["sentence"]).lower()
return batch
# run inference
def inference(batch):
input_values = processor(batch["input_values"],
sampling_rate=16000,
return_tensors="pt").input_values
logits = model(input_values.to(device)).logits
pred_ids = torch.argmax(logits, dim=-1)
batch["pred_transcript"] = processor.batch_decode(pred_ids)
return batch
test_dataset = test_dataset.map(preprocess)
result = test_dataset.map(inference, batched=True, batch_size=1)
print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_transcript"], references=result["transcript"])))
```
**Test Result**: 10.78%
<a name = "citation" ></a>
### Citation
[![DOI](https://zenodo.org/badge/491468343.svg)](https://zenodo.org/badge/latestdoi/491468343)
<strong>BibTeX</strong>
```
@mics{Duy_Khanh_Finetune_Wav2vec_2_0_2022,
author = {Duy Khanh, Le},
doi = {10.5281/zenodo.6542357},
license = {CC-BY-NC-4.0},
month = {5},
title = {{Finetune Wav2vec 2.0 For Vietnamese Speech Recognition}},
url = {https://github.com/khanld/ASR-Wa2vec-Finetune},
year = {2022}
}
```
<strong>APA</strong>
```
Duy Khanh, L. (2022). Finetune Wav2vec 2.0 For Vietnamese Speech Recognition [Data set]. https://doi.org/10.5281/zenodo.6542357
```
<a name = "contact"></a>
### Contact
- khanhld218@gmail.com
- [![GitHub](https://img.shields.io/badge/github-%23121011.svg?style=for-the-badge&logo=github&logoColor=white)](https://github.com/)
- [![LinkedIn](https://img.shields.io/badge/linkedin-%230077B5.svg?style=for-the-badge&logo=linkedin&logoColor=white)](https://www.linkedin.com/in/khanhld257/)
Reference in New Issue View Git Blame Copy Permalink