search

d-ailin / GDN

Implementation code for the paper "Graph Neural Network-Based Anomaly Detection in Multivariate Time Series" (AAAI 2021)
MIT License
482 stars 141 forks source link

CUDA version? #3

Closed CaffreyR closed 3 years ago

CaffreyR commented 3 years ago

Thank you for sharing the code. One little question, does the CUDA version have to be 10.2? Does the OSError has to do with the problem？ Thanks again

d-ailin commented 3 years ago

Hi, I haven't tried it with the latest CUDA version(e.g. 11.1), but I think it would be best to be 10.2 as some packages(e.g., torch-geometric) require a specific environment.

Maybe you could set up the environment with conda:

# CUDA 10.2
conda install pytorch==1.5.1 torchvision==0.6.1 cudatoolkit=10.2 -c pytorch
CaffreyR commented 3 years ago

Hi, I haven't tried it with the latest CUDA version(e.g. 11.1), but I think it would be best to be 10.2 as some packages(e.g., torch-geometric) require a specific environment.

Maybe you could set up the environment with conda:

# CUDA 10.2
conda install pytorch==1.5.1 torchvision==0.6.1 cudatoolkit=10.2 -c p

Hi, I haven't tried it with the latest CUDA version(e.g. 11.1), but I think it would be best to be 10.2 as some packages(e.g., torch-geometric) require a specific environment.

Maybe you could set up the environment with conda:

CUDA 10.2

conda install pytorch==1.5.1 torchvision==0.6.1 cudatoolkit=10.2 -c pytorch

Thanks for replying.OSError occurs when importing libs, probably the version problem. I will try again later. Thanks again.