Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -9,4 +9,4 @@ language:
9
 
10
  The extrapolated (ExPO) model based on [`shenzhi-wang/Llama3-8B-Chinese-Chat`](https://huggingface.co/shenzhi-wang/Llama3-8B-Chinese-Chat) and [`meta-llama/Meta-Llama-3-8B-Instruct`](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct), as in the "[Weak-to-Strong Extrapolation Expedites Alignment](https://arxiv.org/abs/2404.16792)" paper.
11
 
12
- Specifically, we obtain this model by extrapolating from the weights of the SFT and DPO/RLHF checkpoints, achieving superior alignment with human preference.
 
9
 
10
  The extrapolated (ExPO) model based on [`shenzhi-wang/Llama3-8B-Chinese-Chat`](https://huggingface.co/shenzhi-wang/Llama3-8B-Chinese-Chat) and [`meta-llama/Meta-Llama-3-8B-Instruct`](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct), as in the "[Weak-to-Strong Extrapolation Expedites Alignment](https://arxiv.org/abs/2404.16792)" paper.
11
 
12
+ Specifically, we obtain this model by extrapolating **(alpha = 0.3)** from the weights of the SFT and DPO/RLHF checkpoints, achieving superior alignment with human preference.