View Code? Open in Web Editor
NEW
Official PyTorch implementation of the ICML 2023 paper "Adaptive IMLE for Few-shot Pretraining-free Generative Modelling "
Home Page: https://mehranagh20.github.io/AdaIMLE/
License: MIT License
Python 16.86%
Jupyter Notebook 82.92%
Shell 0.22%
adaimle's People
Contributors
adaimle's Issues
Thank you for your great work!
I would like to know what make the results of EDM so bad in your paper, especially the distortions in generated images.
Is it caused merely by the small size of training set?
Would like to know more details on the experiments :)