Hugging Face's logo

wh-zhu
/
DeepSeek-R1-TrRa-iter2-1.5B-lambda_2

Safetensors
qwen2
Model card Files
xet
Community
DeepSeek-R1-TrRa-iter2-1.5B-lambda_2 / README.md
wh-zhu's picture
wh-zhu
Upload folder using huggingface_hub
a16cc13 verified
preview code
|
raw
history blame contribute delete
964 Bytes

🛠️ ReAligner

arXiv Paper   Homepage   Models

A flexible realignment framework is proposed to quantitatively control alignment during training and inference, combining Training-time Realignment (TrRa) and Inference-time Realignment (InRa).
  • We realign DeepScaleR-1.5B model and reduce token usage without performance loss and even enhance reasoning capabilities.

img