Why Adam work in optimizing language models: heavy-tailed data imbalance
ETH AI Center, Andreasstrasse 5, 8050 Zürich, Kreis 11
About
A talk at the ETH AI Center on optimizing language models, focusing on heavy-tailed data imbalance. The title points to the Adam optimizer and why it works with such data. No further details about the speakers are given.
Written by a model, sources: forms.gle
Venue
ETH AI Center is in Zürich’s Kreis 11.
Andreasstrasse 5, 8050 Zürich
ETH AI Center 1 more event