In Technical Report

Open univa-JASON opened this issue 1 year ago • 2 comments

How did you proceed with DPO learning? Using CPMTrainer, or HF DPOTrainer? Does CPM Trainer support DPO finetuning?

Aug 08 '24 04:08 univa-JASON

We write the training code based on the RLAIF-V project. The code implement a trainer for DPO by itself.

Aug 08 '24 15:08 yiranyyu

thank you for your answer! i have 1 more question, Can I apply wsd scheduler in This repo's finetuning code?

Aug 09 '24 01:08 univa-JASON

Hello! By any chance, could you point me to the code you used for MiniCPM-V DPO learning? Thank you and greatly appreciate it.

Nov 21 '24 10:11 DarioPTWR