Key points
- Open Educational Language Models combine generative AI with Open Educational Resources.
- OELMs separate learning content from activities using content files, learning activity files and a coordination service to merge them.
- That separation allows flexibility in mixing content and activities.
- The architecture avoids Retrieval Augmented Generation but includes references to files for alignment.
The article describes the architecture of Open Educational Language Models (OELMs), which combine generative AI with Open Educational Resources (OER). OELMs separate learning content from activities, using content files, learning activity files, and a coordination service to merge them. This allows flexibility in mixing content and activities. The system avoids Retrieval Augmented Generation (RAG) but includes references to files for alignment.
Leave a Reply