PaddlePaddle / PaddleSpeech

Easy-to-use Speech Toolkit including Self-Supervised Learning model, SOTA/Streaming ASR with punctuation, Streaming TTS with text frontend, Speaker Verification System, End-to-End Speech Translation and Keyword Spotting. Won NAACL2022 Best Demo Award.
https://paddlespeech.readthedocs.io
Apache License 2.0
11.14k stars 1.85k forks source link

ASR模型测试异常 #3859

Closed zhangxiangchn closed 1 month ago

zhangxiangchn commented 1 month ago

General Question

环境和相关包: 昇腾910B,安装的paddle相关包如下 paddle2onnx 1.2.9 paddleaudio 1.0.2 paddlefsl 1.1.0 paddlenlp 2.5.2 paddlepaddle 2.5.2 paddleslim 2.3.4 paddlespeech 1.3.0 paddlespeech-feat 0.1.0

image paddlespeech asr --lang zh --input zh.wav 可以正确进行语音识别

根据/examples/aishell/asr1/README.md 中的步骤可以正常训练,在执行模型测试的时候出现如下异常

image

除了上述报错还有一个问题,上述./local/test.sh脚本只传了两个参数,但是test.sh 脚本中接收了3个参数? image

zhangxiangchn commented 1 month ago

source utils/parse_options.sh 这个通过添加环境变量指定绝对路径解决了