MeckyWu comments

Repositories
Issues
Comments

Results 2 comments of


                                            MeckyWu

Scores and probability calcuations

Hi @namdw @xukp20 I'm sorry for not getting back to you sooner. You are right. The correct implementation should be to calculate the win-rate matrix first and then take the...

Discrepancy in rm_scores Length for Mistral-7B-SPPO Iter1 Dataset

We included the original two responses in the UltraFeedback dataset and the 5 generated responses in the first iteration's dataset. This is not done in the released Github repo though....