Handle explain option for plaintext judges
295a884
ping98k
commited on
Spaces:
Sleeping
Sleeping
Commit History
Add Elo bar plot
e6b4ba0
ping98k
commited on
Use Elo rating
b936324
ping98k
commited on
Switch to chat_template_kwargs for thinking
6d50a75
ping98k
commited on
Add configurable thinking budget
3404ee0
ping98k
commited on
Add judge instruction options and temperature settings
02aebba
ping98k
commited on
Fix pairwise tournament loop
2b53c20
ping98k
commited on
Log scoring and pairwise completions
3bd9ad6
ping98k
commited on
Add full loop test for run_tournament
249284d
ping98k
commited on
Add unit tests for completion-based utils
aebdef3
ping98k
commited on