AI implementations using java, stores and either of Langchain4j or springai framework
- LLM - Large Language Model
- Embedding Store - database equivalent to store vector embeddings
- Embeddings - Document converted to vector
- Document - information that is required for AI to process
- Java 17+
- Docker Engine
Below is the summary of implementations in this repository
| Category | Module | Description |
|---|---|---|
| playground | langchain4j playground | AI playground using Langchain4j |
| spring ai playground | AI playground using SpringAI & SpringBoot | |
| embeddingstores | neo4j store with springai | Embedding store implementation using springai, spring boot and neo4j as store |
| opensearch store with langchain4j | Embedding store implementation using langchain4j and opensearch as store | |
| pgvector store with lanchain4j | Embedding store implementation using langchain4j and pgvector as store | |
| pgvector store with springai | Embedding store implementation using springai, spring boot and pgvector as store | |
| RAG(Retrieval-Augmented-Generation) | rag implementation using langchain4j and AllMiniLmL6V2 | RAG Implementation using Langchain4j, PGVector store and allMiniLmL6V2 LLM with spring boot |
| rag implementation using springai with llama llm | RAG Implementation using springai, Redis store, PDF document reader and ollama LLM with llama2 model. Ollama LLM is offline version of LLM which will be downloaded once and can be used locally, as a result it will be very slow in responding based on your system configuration |
|
| rag implementation using springai with openai llm | RAG Implementation using springai, PGVector store, Tika document reader and openai LLM with spring boot | |
| ChatBot | chatbox using ollama | ChatBox using Ollama3 LLM and chromadb |
Thanks to langchain4j for providing an openAI compatible API for learning and demo purposes.
Note All examples contains testcontainers, so can run directly in dev mode.