PolyAlign Chinese HDPO - Qwen2.5-1.5B

This model is the Chinese HDPO checkpoint for Qwen2.5-1.5B.

Base model used for HDPO policy training:

sathiiiii/polyalign-qwen2.5-1.5b-zh-dist-sft

The HDPO preference data was scored offline with the PolyAlign MLP critic.

Downloads last month
7
Safetensors
Model size
2B params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for sathiiiii/polyalign-qwen2.5-1.5b-zh-hdpo

Finetuned
(1)
this model