admin管理员组

文章数量:1576394

今天用xgboost跑了一下分类发生如下报错:

A worker process managed by the executor was unexpectedly terminated.

This could be caused by a segmentation fault while calling the function or by an excessive memory usage causing the Operating System to kill the worker.

 

这是显存不足造成的。

 

解决方案

制定富裕空间的显卡

import os

os.environ["CUDA_VISIBLE_DEVICES"] = "0"

注意,用jupyter发生报错时候可以尝试本地脚本跑!

转载于:https://wwwblogs/Tw1st-Fate/p/11087371.html

本文标签: 报错Xgboostnjobs