feat: drastically decrease token use through pre-processing#31
Merged
Conversation
During distillation, the vast majority of JSONL transcript content is tool machinery invisible to the user (tool results, invocations, progress events, thinking blocks, file snapshots). Typically only 2-15% is user/assistant text. Filter each log down to the conversational flow — user messages and assistant text responses — before feeding it to the distilling agent. Uses serde_json::Value for resilience to schema changes; unknown types and non-JSON lines fall through as raw content (fail-useful).
df09627 to
dcf3f31
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
During distillation, the vast majority of JSONL transcript content is tool
machinery invisible to the user (tool results, invocations, progress events,
thinking blocks, file snapshots). Typically only 2-15% is user/assistant text.
Always fail-useful - even to the extreme of preserving the raw input if it doesn't parse as valid JSON.