|
--- |
|
license: apache-2.0 |
|
language: |
|
- en |
|
--- |
|
|
|
# Starling-LM-7B-alpha-ExPO |
|
|
|
The extrapolated (ExPO) model based on [`berkeley-nest/Starling-LM-7B-alpha`](https://huggingface.co/berkeley-nest/Starling-LM-7B-alpha) and [`openchat/openchat_3.5`](https://huggingface.co/openchat/openchat_3.5), as in the "[Weak-to-Strong Extrapolation Expedites Alignment](https://arxiv.org/abs/2404.16792)" paper. |
|
|
|
Specifically, we obtain this model by extrapolating **(alpha = 0.5)** from the weights of the SFT and DPO/RLHF checkpoints, achieving superior alignment with human preference. |
