Baidu trained ERNIE 5.5 Flash Base on FlashMask-4 kernels, reporting speedups on NVIDIA SM100 GPUs of up to 175% in forward pass and 74% in backward pass over FlashAttention-4 MaskMod
Agree
Disagree
Overview
Debates
Related claims
Topics
Sources
Best
Space 41e851
Loading debates…
Opening sign-in…