search

median-research-group / LibMTL

A PyTorch Library for Multi-Task Learning
MIT License
1.94k stars 181 forks source link

QM9 applicability of the architectures #77

Closed PaoloBonettiPolimi closed 5 months ago

PaoloBonettiPolimi commented 5 months ago

Hi guys, thank you for your work! I have a question regarding the QM9 dataset: all the architectures that are not based on HPS require a resnet or an image, leading to an error when the "main.py" is called for any of them. Am I wrong? Do you have any variant of the "main.py" that works for the QM9 dataset with the other architectures?

Baijiong-Lin commented 5 months ago

Yes, you are right. Actually, there are very few general architectures in multi-task learning.