search

AstraZeneca / DiffAbXL

The official implementation of DiffAbXL benchmarked in the paper "Exploring Log-Likelihood Scores for Ranking Antibody Sequence Designs", formerly titled "Benchmarking Generative Models for Antibody Design".
Apache License 2.0
39 stars 3 forks source link

Model weights #6

Closed yx0516 closed 1 week ago

yx0516 commented 1 week ago

Appreciate the outstanding work!

Will the trained weights and log probability inference script be available for use? I was unable to find any mention of checkpoints in the repo.

Thanks a lot.

talipucar commented 1 week ago

Hi @yx0516

At this time, there are no plans to release the weights.