NameError: name 'ClusterBackend' is not defined

modelscope / FunClip

Open-source, accurate and easy-to-use video speech recognition & clipping tool, LLM based AI clipping intergrated.

MIT License

3.47k stars 377 forks source link

NameError: name 'ClusterBackend' is not defined #22

Closed andywen0090 closed 5 months ago

andywen0090 commented 6 months ago

Traceback (most recent call last): File "D:\2 github\FunClip\funclipper\launch.py", line 7, in funasr_model = AutoModel(model="iic/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch", File "C:\Users\whyuan\anaconda3\envs\py39\lib\site-packages\funasr\auto\auto_model.py", line 129, in init self.cb_model = ClusterBackend().to(kwargs["device"]) NameError: name 'ClusterBackend' is not defined

请问上述报错该如何解决呢？

andywen0090 commented 6 months ago

解决了，发现好像切了python版本到更新就好了，python版本最好不要用3.9，用新一些。如果不行from funasr.models.campplus.cluster_backend import ClusterBackend这样也可以。零散有挺多问题，requirmenets需要增加modelscope，否则模型下载不了。另外还需要install torchaudio，还需要install hdbscan，这个有依赖windows的windows 10 sdk和MSVC V143 x64/x86生成工具等。

R1ckShi commented 6 months ago

感谢反馈，modelscope已经添加到requirments，torch和torchaudio因为与环境、cuda等相关没有添加至requirments funasr的相关报错在最新的funasr中应该已经修正