Use DeepSeek agent for BuffBench - #622
Conversation
Co-authored-by: Claude Opus 4.7 <[email protected]>
Co-authored-by: Claude Opus 4.7 <[email protected]>
Co-authored-by: Claude Opus 4.7 <[email protected]>
Co-authored-by: Claude Opus 4.7 <[email protected]>
Opus 4.7's tokenizer differs from most models; using 4.6 gives more representative counts for open-source models routed through the Anthropic count_tokens endpoint. Co-Authored-By: Claude Opus 4.7 (1M context) <[email protected]>
Co-authored-by: Claude Opus 4.7 <[email protected]>
…on, remove redundant IS_FREEBUFF guard, extract END_SESSION_MESSAGE constant
This reverts commit c11a473.
Co-authored-by: James Grugett <[email protected]>
Co-authored-by: James Grugett <[email protected]>
Co-authored-by: James Grugett <[email protected]>
Co-authored-by: Claude Opus 4.7 (1M context) <[email protected]>
Co-authored-by: James Grugett <[email protected]>
commented
Aug 20, 2026
|
The stated change — swapping However the diff is much larger than the description suggests. As it stands this reads like two unrelated changes squashed together: a trivial eval config change, and a meaningful behavioral change to how base2/base-deep orchestrate sub-agents. The latter deserves its own PR with a description of why those agents are being removed and what testing (if any) shows the direct-tool approach performs as well. Please split these so each can be reviewed and ported on its own merits. |
421a0a5 to
d20fdc8
Compare
Summary
base2-free-evalstobase2-free-deepseek.Validation
git diff origin/main...HEAD.