localcolabfold
localcolabfold copied to clipboard
Question: Out of memory in 3080 when using the wsl2.
What is your question? I only use a short sequence to predict the structure(len 98). However, it is out of memory in 3080 when I use the wsl2.
Computational environment
- OS: [Ubuntu 20.04, Windows10 & WSL2]
- CUDA version: CUDA #11.7

Did you try these environment variables before prediction? https://github.com/YoshitakaMo/localcolabfold#for-wsl2-in-windows