Reward Modeling — Roadmap Step & Resources
Training a model to predict which outputs humans would prefer
- Level: beginner
- Estimated time: 1 days
- Roadmap: Fine-Tuning & Model Customization
Before this step
Study resources
- PyTorch Documentation (Article) — Official documentation for the deep learning framework underlying most modern fine-tuning workflows. Background reading/viewing for: Reward Modeling.
- Hugging Face Documentation (Article) — Official documentation on RLHF and preference-based training methods. Relevant to: Reward Modeling.
Part of
- RLHF and Preference-Based Alignment — section
- Fine-Tuning & Model Customization roadmap — the full learning path