Skip to content

Fix MT-Bench multi-turn evaluation logic#1129

Draft
pjavanrood wants to merge 1 commit intohuggingface:mainfrom
pjavanrood:mt_bench_bug_fix_upstream_fix
Draft

Fix MT-Bench multi-turn evaluation logic#1129
pjavanrood wants to merge 1 commit intohuggingface:mainfrom
pjavanrood:mt_bench_bug_fix_upstream_fix

Conversation

@pjavanrood
Copy link
Contributor

This PR addresses the issue identified in #1128.

Re-created from fork PR #36.

Fixes #1128

- Handle individual turns in batch evaluation
- Return individual scores and judgements for each document in the batch

Fixes #35
@pjavanrood pjavanrood marked this pull request as draft January 23, 2026 00:11
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Fix MT-Bench multi-turn evaluation logic

1 participant