[Unity] PagedKVCache supporting on-the-fly RoPE calculation - #16396
Merged
tqchen merged 1 commit intoJan 15, 2024
Merged
Conversation
MasterJH5574
force-pushed
the
unity-dev/2024-01-13-kv-cache-rope
branch
3 times, most recently
from
January 13, 2024 22:49
b67d71e to
281ddc3
Compare
tqchen
approved these changes
Jan 14, 2024
MasterJH5574
force-pushed
the
unity-dev/2024-01-13-kv-cache-rope
branch
3 times, most recently
from
January 14, 2024 19:01
f162975 to
bd3b958
Compare
This PR enhances PagedKVCache with the inline RoPE compute, which unblocks the movement towards sliding window and attention sink. Both FlashInfer and TIR kernels are updated in this PR with the RoPE calculation. Note that FlashInfer is bumped in order to include the RoPE update. The previous standalone kernel used for RoPE application are thereby removed. --- Co-authored-by: Bohan Hou <spectrometerh@gmail.com> Co-authored-by: Hongyi Jin <jinhongyi02@gmail.com>
MasterJH5574
force-pushed
the
unity-dev/2024-01-13-kv-cache-rope
branch
from
January 14, 2024 21:10
bd3b958 to
6f180d4
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR enhances PagedKVCache with the inline RoPE compute, which unblocks the movement towards sliding window and attention sink.
Both FlashInfer and TIR kernels are updated in this PR with the RoPE calculation. Note that FlashInfer is bumped in order to include the RoPE update.
The previous standalone kernel used for RoPE application are thereby removed.
Co-authored-by: Bohan Hou spectrometerh@gmail.com
Co-authored-by: Hongyi Jin jinhongyi02@gmail.com