xFormers：一个模块化和可编程的Transformer建模

xFormers: A modular and hackable Transformer modelling library

https://github.com/facebookresearch/xformers

https://facebookresearch.github.io/xformers/components/ops.html#xformers.ops.memory_efficient_attention

Self-attention Does Not Need O(n^2) Memory

https://arxiv.org/abs/2112.05682