当前位置:网站首页>Pytorch, specify the problem of the graphics card
Pytorch, specify the problem of the graphics card
2022-07-22 21:33:00 【Lost in code】
Run on the server PyTorch Training code for , It turned out that there was no problem . But reinstall cuda after , Cannot specify which blocks GPU Run on , Only in order from 0 Block start using .
resolvent : stay os Followed by the designated graphics card .( Because when importing torch Behind the library , It has called the graphics card .)
import os
os.environ["CUDA_VISIBLE_DEVICES"] = "4,5,6,7"
import torch
Game Over!
边栏推荐
- Bash variable -- user defined variable
- Web编程入门 3.1 选择框select
- YOLO9000: Better, Faster, Stronger
- 快速查找论文的源代码网站资料
- Wechat applet comprehensive case practice 1
- Interview of 2020 Central South University information and communication summer camp
- [lttng learning journey] - lttng features
- [lttng learning journey] ----- core concepts
- 如何做好研发精益需求管理
- Condition judgment
猜你喜欢
随机推荐
pytorch, 指定显卡问题
蓝桥杯-省赛-翻硬币
Solve the problems that tracepass website cannot be opened and stackoverflow is not fully displayed
蓝桥杯-省赛-带分数
英语四六级听力有线传输无线发射系统方案
[lttng learning journey] ----- core concepts
[lttng learning journey] - trace control - Preliminary
物理层2.1
Makefile中如何自动生成头文件依赖
Juc-6.2-concurrency container copyonwritearraylist
JUC-8.0-Future和Callable
NVIDIA CLARA资料
TIME的新封面:元宇宙将改变一切
Juc-6.3-concurrency container queue
FPN资料
Buuctf breakthrough diary 04 -- [strong net cup 2019] casual note 1
Web编程入门 AdminLTE框架的使用
PKG config lookup library and for compilation
Juc-7.3-thread collaboration cyclicbarrier
[lttng learning journey] - trace control -- advanced