BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Language based Pre-training for Drug Discovery - Bharath Ramsundar
DTSTART:20211115T163000Z
DTEND:20211115T170000Z
UID:TALK162331@talks.cam.ac.uk
CONTACT:Dr Christoph Schran
DESCRIPTION:Pretraining has taken the NLP world by storm as ever larger la
 nguage models have broken successive benchmarks. In this talk\, I’ll rev
 iew some recent work applying pretraining to scientific challenges\, and i
 n particular will discuss the challenges of pretraining for molecular mach
 ine learning. I’ll introduce our new architecture\, ChemBERTa\, which ex
 plores the use of BERT -style pretraining for machine learning problems in
 spired by drug discovery applications.
LOCATION:https://zoom.us/j/92447982065?pwd=RkhaYkM5VTZPZ3pYSHptUXlRSkppQT0
 9
END:VEVENT
END:VCALENDAR