Chatbots

Empire Chain provides several types of chatbots that can be easily integrated into your applications.

Simple Chatbot

The basic chatbot implementation using Streamlit:

from empire_chain.streamlit import Chatbot
from empire_chain.llms.llms import OpenAILLM

# Create and run chatbot
chatbot = Chatbot(
    title="Empire Chatbot",
    llm=OpenAILLM("gpt-4o-mini")
)
chatbot.chat()

Vision Chatbot

Chat with images using multimodal models:

from empire_chain.streamlit import VisionChatbot

# Create and run vision chatbot
chatbot = VisionChatbot(title="Empire Vision Chatbot")
chatbot.chat()

PDF Chatbot

Chat with PDF documents using RAG:

from empire_chain.streamlit import PDFChatbot
from empire_chain.llms.llms import OpenAILLM
from empire_chain.vector_stores import QdrantVectorStore
from empire_chain.embeddings import OpenAIEmbeddings

# Create and run PDF chatbot
pdf_chatbot = PDFChatbot(
    title="PDF Chatbot",
    llm=OpenAILLM("gpt-4o-mini"),
    vector_store=QdrantVectorStore(":memory:"),
    embeddings=OpenAIEmbeddings("text-embedding-3-small")
)
pdf_chatbot.chat()

Features

Simple Chatbot: Basic text-based conversation
Vision Chatbot: Image understanding and discussion
PDF Chatbot: Document-based conversation using RAG
Customizable UI: Built with Streamlit for easy deployment
Multiple LLM Support: OpenAI, Anthropic, Groq

Running the Chatbots

Install dependencies:
```
pip install empire-chain streamlit
```
Run the chatbot:
```
streamlit run app.py
```

For more examples and advanced usage, check out the chatbot cookbooks in the repository.