How to reproduce the P@L results?
#1
by
libokj
- opened
Hi,
Thank you for another wonderful update of ESM. I have a question about reproducing the P@L results. ESM-C models don't seem to have released a trained contact prediction head like in ESM-2. Is it implemented in the same way as ESM-2 (https://doi.org/10.1101/2020.12.15.422761)?