v0.1.0 - Initial Release of RewardUQ

We are excited to announce the first release of RewardUQ, a unified framework for training and evaluating uncertainty-aware reward models. Built on the Hugging Face ecosystem, it bridges the gap between research and production for uncertainty quantification.

Includes implementations for:

• MLP Head Ensemble
• LoRA Ensemble
• DPO-based MC Dropout
• Bayesian Linear Head