v0.8.0

游雁

2023-10-10 580b11b57ac4b62f7e2acda73813a4e10e8e4cd3

 egs_modelscope/vad/TEMPLATE/README.md

@@ -1,3 +1,5 @@
([简体中文](./README_zh.md)|English)

# Voice Activity Detection

> **Note**: 
@@ -22,7 +24,7 @@
#### [FSMN-VAD-online model](https://modelscope.cn/models/damo/speech_fsmn_vad_zh-cn-16k-common-pytorch/summary)
```python
inference_pipeline = pipeline(
    task=Tasks.auto_speech_recognition,
    task=Tasks.voice_activity_detection,
    model='damo/speech_fsmn_vad_zh-cn-16k-common-pytorch',
    )
import soundfile
@@ -83,7 +85,7 @@
#### Decode with multi GPUs:
```shell
    bash infer.sh \
    --model "damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch" \
    --model "damo/speech_fsmn_vad_zh-cn-16k-common-pytorch" \
    --data_dir "./data/test" \
    --output_dir "./results" \
    --batch_size 1 \
@@ -93,11 +95,11 @@
#### Decode with multi-thread CPUs:
```shell
    bash infer.sh \
    --model "damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch" \
    --model "damo/speech_fsmn_vad_zh-cn-16k-common-pytorch" \
    --data_dir "./data/test" \
    --output_dir "./results" \
    --gpu_inference false \
    --njob 1
    --njob 64
```

## Finetune with pipeline