is it possible to publish the bfcl multiturn evaluation handler so that we could directly measure your model?
#2
by
tongliuphysics
- opened
Thanks for this amazing model. Is it possible to publish the bfcl multiturn evaluation handler so that we could directly measure your model?
tongliuphysics
changed discussion title from
is it possible to publish the bfcl evaluation handler so that we could directly measure your model?
to is it possible to publish the bfcl multiturn evaluation handler so that we could directly measure your model?
Hey, thanks for your feedback! We're going to explore this idea and how to make it as reproducible as possible.