speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch 用onnx推理时候报错

[acetotot]

Code

model_dir = "iic/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch/" model = Paraformer(model_dir, batch_size=1, quantize=True) result = model("norm.wav")

What's your environment?

• OS (e.g., Linux):
• FunASR Version (e.g., 1.0.0):1.0.16
• funasr-onnx Version：0.3.1
• ModelScope Version (e.g., 1.11.0):1.12.0
• PyTorch Version (e.g., 2.0.0):1.11.0
• How you installed funasr (pip, source):source
• Python version:3.8.18

报错： return self.session.run(self.get_output_names(), input_dict) "Required inputs (['bias_embed']) are missing from input feed (['speech', 'speech_lengths'])."

[LauraGPT]

Update the code and try it again.

[TengLi931128]

+1, same error.

显示 ‘input wav is silence or noise’，但我在代码定位了下，报错的信息还是"Required inputs (['bias_embed']) are missing from input feed (['speech', 'speech_lengths'])."

funasr 1.0.24 funasr-onnx 0.3.1

[R1ckShi]

可以在现在最新的代码中通过如下的方法导出：

from funasr import AutoModel

model = AutoModel(
    model="iic/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch",
)

res = model.export(type="onnx", quantize=False)
print(res)