Expand AMD ROCm Tips readme section#12349
Open
alexheretic wants to merge 1 commit intoComfy-Org:masterfrom
Open
Expand AMD ROCm Tips readme section#12349alexheretic wants to merge 1 commit intoComfy-Org:masterfrom
alexheretic wants to merge 1 commit intoComfy-Org:masterfrom
Conversation
alexheretic
requested review from
Kosinkadink,
comfyanonymous and
guill
as code owners
alexheretic
force-pushed
the
rocm-tips
branch
from
550d96d to
d0f0286
Compare
Add suggestion to disable online tuning Add miopen info Add flash attention info Add vram oom suggestion
alexheretic
force-pushed
the
rocm-tips
branch
from
d0f0286 to
cb213ae
Compare
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 have verified all these points to be effective, testing with a gfx1100 on Linux. I suspect they are generally worth at least noting for all AMD users, and this readme is one of the first things amd users might read.
For example I recently tested flash-attention on a 360x640 wan2.2 workflow:
--use-flash-attention: 26.06s/it, 26.02s/it--use-flash-attention(tuned attn_fwd): 15.56s/it, 15.12s/it--pytorch-cross-attention: 30.69s/it, 30.97s/it