Conversation
Prefix-cache extra identity previously keyed embedding spans by token range only. Requests with the same token ids and different input_embeddings could reuse the wrong KV. Hash embedding content into the trie identity, matching the multimodal-span path. Fixes InternLM#4933 Signed-off-by: GokayAI <60583610+gokay-ai@users.noreply.github.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Motivation
Prefix-cache extra identity keyed embedding spans by token range only. Two requests with the same token ids but different
input_embeddingscould reuse the wrong KV.Fixes #4933
Modification
Hash embedding content into the trie identity (same idea as the multimodal-span path). Unit coverage in the block-trie tests.