离线paraformer推理报错

[dospeech]

Help

3.5号刚git clone的新代码，运行infer.shhou 报错，不知道什么原因，情况如下：

root@9df7eea98999:/workspace/20240305_funasr/FunASR/examples/industrial_data_pretraining/paraformer# ./infer.sh

If you want to use hugging, please pip install -U transformers /usr/local/lib/python3.8/runpy.py:127: RuntimeWarning: 'funasr.bin.inference' found in sys.modules after import of package 'funasr.bin', but prior to execution of 'funasr.bin.inference'; this may result in unpredictable behaviour warn(RuntimeWarning(msg)) [2024-03-05 13:20:05,566][root][INFO] - download models from model hub: ms [2024-03-05 13:20:11,621][root][INFO] - Loading pretrained params from exp/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch/model.pt ckpt: exp/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch/model.pt Error executing job with overrides: ['++model=exp/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch', '++model_revision=v2.0.4', '++input=exp/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch/example/asr_example.wav', '++output_dir=./outputs/debug', '++device=cuda:0'] Traceback (most recent call last): File "/workspace/20240305_funasr/FunASR/funasr/bin/inference.py", line 25, in main_hydra model = AutoModel(kwargs) File "/workspace/20240305_funasr/FunASR/funasr/auto/auto_model.py", line 102, in init model, kwargs = self.build_model(kwargs) File "/workspace/20240305_funasr/FunASR/funasr/auto/auto_model.py", line 191, in build_model load_pretrained_model( File "/workspace/20240305_funasr/FunASR/funasr/train_utils/load_pretrained_model.py", line 68, in load_pretrained_model src_state = torch.load(path, map_location=map_location) File "/usr/local/lib/python3.8/site-packages/torch/serialization.py", line 795, in load return _legacy_load(opened_file, map_location, pickle_module, pickle_load_args) File "/usr/local/lib/python3.8/site-packages/torch/serialization.py", line 1002, in _legacy_load magic_number = pickle_module.load(f, pickle_load_args) _pickle.UnpicklingError: invalid load key, 'v'.

Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace.

#########. infer.sh Code 如下： input="exp/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch/example/asr_example.wav"

output_dir="./outputs/debug"

model="exp/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"

model="exp/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch" model_revision="v2.0.4"

device="cuda:0" # "cuda:0" for gpu0, "cuda:1" for gpu1, "cpu"

python -m funasr.bin.inference \ ++model=${model} \ ++model_revision=${model_revision} \ ++input="${input}" \ ++output_dir="${output_dir}" \ ++device="${device}" \

What have you tried?

只改变音频输入路径

What's your environment?

• OS :Debian GNU/Linux 10 (buster)

• FunASR Version (1.0.0):

• ModelScope Version (e.g., 1.11.0):

• PyTorch Version :1.13.0+cu117

• How you installed funasr (pip, source): https://pypi.tuna.tsinghua.edu.cn/simple

• Python version:Python 3.8.10

• GPU (e.g., V100M32)：V100M32

• Any other relevant information:

[LauraGPT]

Please check your model path.