janoserdody

### 🥰 Feature Description We are running an on-premise LLM and would like to further fine-tune the model using Reinforcement Learning from Human Feedback (RLHF). To facilitate this, we propose...