Animations for Flash Attention, Flash Attention2, and Standard Attention #736
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Thank you very much, I find this topic very interesting. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
I made a series of
manimanimations illustrating how the implementations of Standard Attention, Flash Attention, and Flash Attention 2 work. They show how, asymptotically, Flash Attention and Flash Attention 2 perform better than Standard Attention with respect to the number of Input/Output accesses between HMB and SRAM.Note that the animations only show the Forward Pass for each algorithm.
Find them all in this playlist