Skip to content

Tags: lucidrains/Mega-pytorch

Tags

0.1.0

Toggle 0.1.0's commit message
add sub-groupnorm in multihead ema, for use in audio modeling

0.0.15

Toggle 0.0.15's commit message
fix laplace activation function thanks to @boweny-cerebras

0.0.14

Toggle 0.0.14's commit message
fix residual within mega layer, thanks to @VHellendoorn

0.0.12

Toggle 0.0.12's commit message
prenorm requires a final layernorm

0.0.11

Toggle 0.0.11's commit message
fix residual for prenorm

0.0.10

Toggle 0.0.10's commit message
expose multi-headed learned EMA, for use outside of repo

0.0.9

Toggle 0.0.9's commit message
offer prenorm architecture

0.0.8

Toggle 0.0.8's commit message
handle bidirectional better

0.0.7

Toggle 0.0.7's commit message
setup enwik8 autoregressive training

0.0.6

Toggle 0.0.6's commit message
improvise on bidirectional for multi-head learned ema