FunASR安装

1. 创建和激活conda环境
   conda create -n funasr python=3.8
   conda activate funasr
2. 安装pytorch torchaudio等
   conda install pytorch=1.12.1 torchvision torchaudio cudatoolkit=11.3 -c pytorch -c conda-forge
   linux系统上cuda是11.6
3. 安装modelscope相关工具
   pip install “modelscope[audio_asr]” -f https://modelscope.oss-cn-beijing.aliyuncs.com/releases/repo.html
4. 克隆FunAsr项目并安装依赖
   git clone https://github.com/alibaba/FunASR.git
   cd FunASR
   pip install --editable ./
5. 测试
   python

import torch
torch.__version__
torch.cuda.is_available()

import modelscope
import funasr
modelscope.__version__
funasr.__version__
1
2
3
4
5
6
7
8

测试推理代码

from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks

inference_pipeline = pipeline(
    task=Tasks.auto_speech_recognition,
    model='damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch')

rec_result = inference_pipeline(audio_in='https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_zh.wav')
print(rec_result)
1
2
3
4
5
6
7
8
9

预期输出：

{'text': '欢迎大家来体验达摩院推出的语音识别模型'}
1

6. 最终环境记录

python=3.8 pytorch=1.12.1 cudatoolkit=11.3 modelscope=1.3.0 funasr=0.2.1

备注：如果克隆不下来FunAsr，
apt-get update
apt-get install libcurl4-openssl-dev

7. 卸载
   funasr pip uninstall funasr

8. 更新

pip install “modelscope[audio]” --upgrade -f https://modelscope.oss-cn-beijing.aliyuncs.com/releases/repo.html
pip install “funasr” --upgrade -f https://modelscope.oss-cn-beijing.aliyuncs.com/releases/repo.html