vad_model和punc_model设置本地路径报错

[LIUKAI0815]

Traceback (most recent call last):
File "/workspace/model/llm/audio_model/asr.py", line 6, in
model = AutoModel(model="./ASR/speech_paraformer-large-contextual_asr_nat-zh-cn-16k-common-vocab8404",vad_model="./ASR/iic/speech_fsmn_vad_zh-cn-16k-common-onnx",
File "/root/miniforge3/envs/audio/lib/python3.10/site-packages/funasr/auto/auto_model.py", line 135, in init
vad_model, vad_kwargs = self.build_model(**vad_kwargs)
File "/root/miniforge3/envs/audio/lib/python3.10/site-packages/funasr/auto/auto_model.py", line 259, in build_model
assert model_class is not None, f'{kwargs["model"]} is not registered'
AssertionError: ./ASR/iic/speech_fsmn_vad_zh-cn-16k-common-onnx is not registered

代码：import os
from funasr import AutoModel
os.environ['CUDA_VISIBLE_DEVICES'] = "3"

paraformer-zh is a multi-functional asr model

use vad, punc, spk or not as you need

model = AutoModel(model="./ASR/speech_paraformer-large-contextual_asr_nat-zh-cn-16k-common-vocab8404",vad_model="./ASR/iic/speech_fsmn_vad_zh-cn-16k-common-onnx",
punc_model="./ASR/iic/punc_ct-transformer_zh-cn-common-vocab272727-onnx")
res = model.generate(input=f"video.wav",
batch_size_s=492,
hotword='外呼')
print(res)

[LIUKAI0815]

已经解决，是因为路径里用了onnx模型