11721206

when inference use phoneme，there is an error "can't convert np.ndarray of type numpy.str_ ........."

> 把调用代码贴上来，以及prompt音频 ``` from cosyvoice.cli.cosyvoice import CosyVoice from cosyvoice.utils.file_utils import load_wav import torchaudio import sys, psutil from time import time as ttime import torch import threading, os from modelscope import...

> Could you also share the code you used in the `To reproduce` section? CUDA kernel for `ScatterND` shouldn't be missing. I just use this code (https://github.com/RVC-Boss/GPT-SoVITS/blob/main/GPT_SoVITS/onnx_export.py) convert torch model...

> @11721206, Could you try 1.18.0 or 1.18.1? > > 1.17.3 supports ScatterND up to opset 13 but your model is opset 17: > > https://github.com/microsoft/onnxruntime/blob/0453cd761860e68d3852e7f81a5092c98369bb75/onnxruntime/core/providers/cuda/tensor/scatter_nd.cc#L23 > > 1.18.1 supports...

> https://github.com/microsoft/onnxruntime/blob/0453cd761860e68d3852e7f81a5092c98369bb75/onnxruntime/core/providers/cuda/tensor/scatter_nd.cc#L21-L23 > > > Such lines indicate that the operator is supported since opset version 13, not up to 13. > `ScatterND` is updated since opset 16 (new attribute...

how can i finetune the qwen3-omni talker part? Is there any examples or sugguestions for dataset preparation and finetune talker scripts? Thanks