For debugging consider passing CUDA_LAUNCH_BLOCKING=1.

本文主要是介绍For debugging consider passing CUDA_LAUNCH_BLOCKING=1.，希望对大家解决编程问题提供一定的参考价值，需要的开发者们随着小编来一起学习吧！

环境：

wsl ubuntu22.04

vits2

问题描述：

RuntimeError: CUDA error: unknown error
[rank0]: CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
[rank0]: For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
[rank0]: Compile with TORCH_USE_CUDA_DSA to enable device-side assertions

解决方案：

1.在train_ms.py开头代码添加

import os
os.environ['CUDA_LAUNCH_BLOCKING'] = '1'

2.重新训练 这边报错消失

这篇关于For debugging consider passing CUDA_LAUNCH_BLOCKING=1.的文章就介绍到这儿，希望我们推荐的文章对编程师们有所帮助！

---