paddlespeech 语音识别 web流服务部署(Streaming Speech Recognition)

安装

首先需要安装 paddlepaddle、paddlespeech 参考这两篇

https://blog.csdn.net/weixin_48185819/article/details/126405989?spm=1001.2014.3001.5502

https://blog.csdn.net/weixin_48185819/article/details/126406615?spm=1001.2014.3001.5502

项目下载

https://github.com/PaddlePaddle/PaddleSpeech

有时候git clone 无法下载，需要直接从网站上下载

下载后 是这个项目 PaddleSpeech-develop   (2022/8/19) 在这个时间点

因为项目一直在更新，给大家列出下载时间点，可能以后会变，因为在部署过程中遇到了版本问题

基本上有这些文件夹，只展示了一部分

部署普通ASR服务

这个文件夹下 有一些部署文件

主要使用application.yaml 部署(也可以使用其他yaml部署)

/PaddleSpeech-develop/paddlespeech/server/conf

 运行(在上面这个路径)

paddlespeech_server start --config_file application.yaml

会出现一些error和检查 但是没有什么影响

还要warm up

 not modify any behavior and is safe. If you specifically wanted the numpy scalar type, use `np.complex128` here.
Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations
  dtype=np.complex,
[nltk_data] Error loading averaged_perceptron_tagger: <urlopen error
[nltk_data]     [Errno 111] Connection refused>
[nltk_data] Error loading cmudict: <urlopen error [Errno 111]
[nltk_data]     Connection refused>
[2022-08-19 17:11:58,447] [    INFO] - start to init the model
[2022-08-19 17:11:58,447] [    INFO] - File /root/.paddlespeech/models/conformer_wenetspeech-zh-16k/asr1_conformer_wenetspeech_ckpt_0.1.1.model.tar.gz md5 checking...
[2022-08-19 17:12:02,646] [    INFO] - Use pretrained model stored in: /root/.paddlespeech/models/conformer_wenetspeech-zh-16k/asr1_conformer_wenetspeech_ckpt_0.1.1.model.tar
[2022-08-19 17:12:02,647] [    INFO] - /root/.paddlespeech/models/conformer_wenetspeech-zh-16k/asr1_conformer_wenetspeech_ckpt_0.1.1.model.tar
[2022-08-19 17:12:02,647] [    INFO] - /root/.paddlespeech/models/conformer_wenetspeech-zh-16k/asr1_conformer_wenetspeech_ckpt_0.1.1.model.tar/model.yaml
[2022-08-19 17:12:02,647] [    INFO] - /root/.paddlespeech/models/conformer_wenetspeech-zh-16k/asr1_conformer_wenetspeech_ckpt_0.1.1.model.tar/exp/conformer/checkpoints/wenetspeech.pdparams
[2022-08-19 17:12:06,447] [    INFO] - Initialize ASR server engine successfully on device: cpu.
[2022-08-19 17:12:06,449] [    INFO] - File /root/.paddlespeech/models/fastspeech2_csmsc-zh/fastspeech2_nosil_baker_ckpt_0.4.zip md5 checking...
[2022-08-19 17:12:07,467] [    INFO] - Use pretrained model stored in: /root/.paddlespeech/models/fastspeech2_csmsc-zh/fastspeech2_nosil_baker_ckpt_0.4
[2022-08-19 17:12:07,467] [    INFO] - /root/.paddlespeech/models/fastspeech2_csmsc-zh/fastspeech2_nosil_baker_ckpt_0.4
[2022-08-19 17:12:07,467] [    INFO] - /root/.paddlespeech/models/fastspeech2_csmsc-zh/fastspeech2_nosil_baker_ckpt_0.4/default.yaml
[2022-08-19 17:12:07,467] [    INFO] - /root/.paddlespeech/models/fastspeech2_csmsc-zh/fastspeech2_nosil_baker_ckpt_0.4/snapshot_iter_76000.pdz
[2022-08-19 17:12:07,467] [    INFO] - File /root/.paddlespeech/models/pwgan_csmsc-zh/pwg_baker_ckpt_0.4.zip md5 checking...
[2022-08-19 17:12:07,500] [    INFO] - Use pretrained model stored in: /root/.paddlespeech/models/pwgan_csmsc-zh/pwg_baker_ckpt_0.4
[2022-08-19 17:12:07,501] [    INFO] - /root/.paddlespeech/models/pwgan_csmsc-zh/pwg_baker_ckpt_0.4
[2022-08-19 17:12:07,501] [    INFO] - /root/.paddlespeech/models/pwgan_csmsc-zh/pwg_baker_ckpt_0.4/pwg_default.yaml
[2022-08-19 17:12:07,501] [    INFO] - /root/.paddlespeech/models/pwgan_csmsc-zh/pwg_baker_ckpt_0.4/pwg_snapshot_iter_400000.pdz
vocab_size: 268
frontend done!
encoder_type is transformer
decoder_type is transformer
acoustic model done!
voc done!
[2022-08-19 17:12:15,269] [    INFO] - Start to