Add fewshot_as_multiturn column

#868
by djstrong - opened

Now the results can't be distinguished:
image.png

Open LLM Leaderboard org
edited Aug 1

Hi! Thanks for the report, this is a bug I believe - all chat models are launched with fewshot as multiturn, but gemma had to be relaunched entirely to include a fix in the harness regarding token management - however, the new results should have overcome the old ones. I'll take a look asap

Open LLM Leaderboard org

Hi!
I'm not finding both these results in the leaderboard so I'm going to assume it was a cache problem.
Tell me if you still have the issue.

image.png

clefourrier changed discussion status to closed

Sign up or log in to comment