Online RLHF
Collection
Datasets, code, and models for online RLHF (i.e., iterative DPO)
•
19 items
•
Updated
•
4
This is the SFT checkpoint used for the project Online-RLHF. Also check our technical report here.
The model is trained from meta-llama/Meta-Llama-3-8B on a mixture of diverse open-source high-quality data for 1 epoch with detailed parameters in the report. It has not been trained by RLHF and can serve as a good starting point for the RLHF research.
- Downloads last month
- 4,562