Getting ViT in Shape: Scaling Laws for Compute-Optimal Model Design.

Ibrahim Alabdulmohsin Xiaohua Zhai Alexander Kolesnikov Lucas Beyer

Published in: CoRR (2023)

Keyphrases

conceptual framework
formal model
computational model
mathematical model
statistical model
closed form
objective function
simulation model
prior information
conceptual model
prior knowledge
neural network
energy function
parameter space
probabilistic model
shape parameters
shape change