Add disable_tensor_cache=True to HPUGraph capture #252
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
michalkuligowski
approved these changes
Sep 10, 2024
zhouyu5
pushed a commit
to zhouyu5/vllm-fork
that referenced
this pull request
Sep 13, 2024
RuntimeErrors are not observed anymore on habana_main when disable_tensor_cache is used. This PR enables disable_tensor_cache.
michalkuligowski
added a commit
that referenced
this pull request
Sep 17, 2024
After #252, HPUGraph capture takes much less memory, and we can reduce the memory reserved for HPUGraphs. On Llama3.1-8b-Instruct (G2), capturing 100% of prefill and decode graphs on BS=256 now takes 1.566 GB of HBM, which is far less than 40% (~30 GB) we reserve by default. This results in lots of unused (==wasted) memory, which could be used instead for more KV cache blocks.
zhouyu5
pushed a commit
to zhouyu5/vllm-fork
that referenced
this pull request
Sep 20, 2024
RuntimeErrors are not observed anymore on habana_main when disable_tensor_cache is used. This PR enables disable_tensor_cache.
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Outdated suggestions cannot be applied.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.
RuntimeErrors are not observed anymore on habana_main when disable_tensor_cache is used. This PR enables disable_tensor_cache.