Paraformer-large模型介绍
使用方法
下载模型
git clone https://github.com/manyeyes/AliParaformerAsr.git
cd /AliParaformerAsr/AliParaformerAsr.Examples
git clone https://huggingface.co/manyeyes/AliParaformerAsr-large-onnx-offline.git
编辑 Program.cs 中的 modelName 值:
string modelName="AliParaformerAsr-large-onnx-offline";
//[fp32模型/int8模型],二选一
//fp32模型
//string modelFilePath = applicationBase + "./"+ modelName + "/model.onnx";
//int8模型
string modelFilePath = applicationBase + "./"+ modelName + "/model_quant.onnx";
将文件夹 AliParaformerAsr-large-onnx-offline 中的文件属性-》复制到输出目录-》如果较新则复制。 或者 编辑 AliParaformerAsr.Examples.csproj ,添加
<ItemGroup>
<None Update="AliParaformerAsr-large-onnx-offline\am.mvn">
<CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
</None>
<None Update="AliParaformerAsr-large-onnx-offline\asr.yaml">
<CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
</None>
<None Update="AliParaformerAsr-large-onnx-offline\example\0.wav">
<CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
</None>
<None Update="AliParaformerAsr-large-onnx-offline\example\1.wav">
<CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
</None>
<None Update="AliParaformerAsr-large-onnx-offline\example\2.wav">
<CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
</None>
<None Update="AliParaformerAsr-large-onnx-offline\example\3.wav">
<CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
</None>
<None Update="AliParaformerAsr-large-onnx-offline\example\4.wav">
<CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
</None>
<None Update="AliParaformerAsr-large-onnx-offline\model_quant.onnx">
<CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
</None>
<None Update="AliParaformerAsr-large-onnx-offline\model.onnx">
<CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
</None>
<None Update="AliParaformerAsr-large-onnx-offline\tokens.txt">
<CopyToOutputDirectory>PreserveNewest</CopyToOutputDirectory>
</None>
</ItemGroup>
启动 AliParaformerAsr.Examples
通过modelscope了解更多
原模型
- pytorch https://www.modelscope.cn/models/damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch/summary
- onnx https://www.modelscope.cn/models/damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-onnx/summary
Highlights
- 热词版本:Paraformer-large热词版模型支持热词定制功能,基于提供的热词列表进行激励增强,提升热词的召回率和准确率。
- 长音频版本:Paraformer-large长音频模型,集成VAD、ASR、标点与时间戳功能,可直接对时长为数小时音频进行识别,并输出带标点文字与时间戳。