sprint 1 follow-up finetuning experiments #223
Description
Investigate how to improve 7B CUA models' performance on perturbation eval data with finetuning
{{ message }}
Investigate how to improve 7B CUA models' performance on perturbation eval data with finetuning