"AssertionError: iic/SenseVoiceSmall is not registered" while deploying it on server via fastapi

[abysee]

Hi, I encountered a problem when deploying SenseVoice on server via fastapi, it raised "AssertionError: iic/SenseVoiceSmall is not registered".

I was following exactlly the instruction:

export SENSEVOICE_DEVICE=cuda:0
fastapi run --port 

And I've looked up at #https://github.com/FunAudioLLM/SenseVoice/issues/52 and tried to add the following into api.py but it doesn't work.

    trust_remote_code=True,
    remote_code="./model.py",

Could you please give me some suggestions?

• OS (e.g., Linux):
• FunASR Version (e.g., 1.0.0): 1.1.5
• ModelScope Version (e.g., 1.11.0): 1.15.0
• PyTorch Version (e.g., 2.0.0): pytorch-lightning 2.3.3 pytorch-wpe 0.0.1 torch 2.0.1+cu118 torch-complex 0.4.4 torchaudio 2.0.2+cu118 torchmetrics 1.4.1 torchvision 0.17.0+cu118
• How you installed funasr (pip, source): pip
• Python version: 3.8
• GPU (e.g., V100M32) RTX 4090
• CUDA/cuDNN version (e.g., cuda11.7): cuda11.8
• Docker version (e.g., funasr-runtime-sdk-cpu-0.4.1) None
• Any other relevant information:

[abysee]

I just fixed it myself, it seemed to be related to modelscope when loading the model. I upgraded the modelscope module to the newest version via pip isntall -U modelscope and then redid the whole process and it worked.