On the version alignment of python, CUDA and python


In the process of using deep learning model training, the preparation of tools is also a good start. It’s hard to finish at the beginning. Sharpening the knife doesn’t miss the firewood cutting. If we can solve the problems in the early stage, we can save a lot of energy for the later stage.

Take the python tool as an example

The python version is 1.0.1, and the native Python version is 3.6.2

CUDA of GPU on server_ VERSION=9000

Note: due to CUDA on GPU_ Version is 9000, so you should at least install CUDA version > = 9.0. Although CUDA = 7.0 ~ 8.0 can run, you may encounter various problems at the beginning. I have CUDA version of 10.0, and the command to install CUDA is as follows:

conda install cudatoolkit=10.0

Note: GPU graphics card driver generally has no problem, so try not to move the cudnn version.

The above discussion on the version alignment of python, CUDA and Python is the whole content shared by Xiaobian. I hope it can give you a reference, and I hope you can support developeppaer more.

Recommended Today

Division without division

preface At the beginning of July, we challenged leetcode‘s question 29 (medium difficulty, which seems nothing to boast about). The problem requires that the function of integer division be realized without division, multiplication and modular operation. Since the divisor and divisor are integers, subtraction can be used to achieve division (what a naive idea). A […]