Oi-Wiki
CreatedShwStone
Augments large language models with OI-Wiki content for enhanced problem-solving in competitive programming and ICPC.
About
Empowers large language models with the comprehensive knowledge of OI-Wiki. It works by creating vector embeddings of OI-Wiki's pages using Deepseek-V3, building a vector database for efficient similarity search. During a query, the tool finds the most relevant OI-Wiki content based on semantic similarity and returns it, providing LLMs with valuable context for tasks related to competitive programming and ICPC.
Key Features
- Semantic search of OI-Wiki content
- Utilizes fastembed for CPU vector embeddings
- Uses Deepseek-V3 for document summarization
- Integration with large language models
- Employs Milvus-lite for vector database management
- 24 GitHub stars
Use Cases
- Improving LLM performance in competitive programming tasks
- Providing LLMs with relevant context for OI/ICPC problems
- Creating a knowledge base for LLMs in algorithm development
