README.md

🤔 What is this?

Inspired by langchain to provide composability when building Large Language Model Application. GChain mission is to bring langchain concept to Go in idiomatic way.

This Library will help on many usecases, such as :

❓ Question Answering over specific documents

• Building chatbot with vector databased backed knowledge base

💬 Chatbots

• Streaming Chatbot with Go and WebSocket

📄 Document Summarization

Install and Importing

$ go get github.com/wejick/gchain

import "github.com/wejick/gchain

Example

llmModel = _openai.NewOpenAIModel(authToken, "", "text-davinci-003",callback.NewManager(), true)
chain, err := llm_chain.NewLLMChain(llmModel, nil)
if err != nil {
    //handle error
}
outputMap, err := chain.Run(context.Background(), map[string]string{"input": "Indonesia Capital is Jakarta\nJakarta is the capital of "})
fmt.Println(outputMap["output"])

More example in the example folder

As our documentation is not yet complete, please refer to examples and integration test for reference.

Notice

1. Don't use it if you have better option
2. GChain priority is golang idiomatic. So eventhough it happily use many langchain concept, don't expect exactly the same behavior as this is not reimplementation.