Caicai Spring AI Exploring Project Getting Started

ACAI

A Spring AI Exploring Project[Caicai]. Functions include:

1. Support Ollama Chat, using the Spring AI Chat SDK.
2. Support RAG(Retrieval Augmented Generation), able to retrieve and generate content based on user queries and input documents.
3. TESTS: Embedding, Vector Store, ETL pipeline

TODO:

1. Image
2. Audio
3. Function Calling
4. Multimodality
5. User Interface
6. Evaluation Testing

1. Project preparation

Basic development environment preparation

Environment	Version
java	OpenJDK version 21.0.1
gradle	Gradle 8.6
pgvector	pgvector:pg16
ollama	0.1.38
httpie	3.2.2

2. Running Models

2.1 for chat

ollama run llama3

OR

ollama run wangshenzhi/llama3-8b-chinese-chat-ollama-q8:latest

2.2 for text embedding

ollama rum nomic-embed-text

3. Running the Spring Boot Application

./gradlew bootRun

4、Retrieval Augmented Generation(RAG)

4.1 Uploading documents

http -f POST :8080/rga/upload file@~/Desktop/upload_test.txt 

4.2 Searching documents

http :8080/rga/search keyword=="Caicai AI test"

4.3 Chatting with documents

http :8080/rga/chat message=="Caicai AI test"

5. Chat(Default)

http :8080/chat message=="Tell me a funny story."

http :8080/chat/generic-options message=="what is spring AI? Give a short answer."

http :8080/chat/generic-options message=="what is spring AI? Give a short answer."

6. API Document

API DOC: http://localhost:8080/swagger-ui/index.html