Making sure summarization leverages KV cache #152
Conversation
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
eddierichter-amd
force-pushed
the
eddie/summarization-kv-cache-eddie-fork
branch
from
4e2b5ae to
7a33fa9
Compare
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Outdated suggestions cannot be applied.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.
The initial implementation of summarization used the .generate() method in the ChatSDK which bypassed sending previously stored messages to the LLM. It therefore needed to send the entire prompt to the agent which was formatted slightly differently and bypassed the KV cache and had to be processed. This resulted in very high TTFT for summarization. This change reduces the TTFT of summarization to be similar to other requests.