Quick Start Guide
This guide will help you get started with Empire Chain by walking through some common use cases.
Basic Setup
First, make sure you have Empire Chain installed and your environment configured:
1. Simple LLM Integration
from empire_chain.llms import OpenAILLM
# Initialize the LLM
llm = OpenAILLM("gpt-4")
# Generate text
response = llm.generate("What are the key principles of AI safety?")
print(response)
2. Document Processing
from empire_chain.file_reader import DocumentReader
# Initialize the document reader
reader = DocumentReader()
# Read a PDF file
text = reader.read("document.pdf")
print(text)
3. Building a Simple Chatbot
from empire_chain.streamlit import Chatbot
from empire_chain.llms import OpenAILLM
# Create a chatbot
chatbot = Chatbot(
llm=OpenAILLM("gpt-4"),
title="My First Chatbot"
)
# Launch the chatbot
chatbot.chat()
4. RAG Implementation
from empire_chain.vector_stores import QdrantVectorStore
from empire_chain.embeddings import OpenAIEmbeddings
from empire_chain.llms import OpenAILLM
from empire_chain.file_reader import DocumentReader
# Initialize components
vector_store = QdrantVectorStore(":memory:")
embeddings = OpenAIEmbeddings("text-embedding-3-small")
llm = OpenAILLM("gpt-4")
reader = DocumentReader()
# Process document
text = reader.read("knowledge_base.pdf")
text_embedding = embeddings.embed(text)
vector_store.add(text, text_embedding)
# Query the system
query = "What are the main points in the document?"
query_embedding = embeddings.embed(query)
relevant_texts = vector_store.query(query_embedding, k=3)
# Generate response
context = "\n".join(relevant_texts)
response = llm.generate(f"Based on this context, {query}\n\nContext: {context}")
print(response)
5. Web Crawling
from empire_chain.crawl4ai import Crawler
# Initialize crawler
crawler = Crawler()
# Crawl a website
data = crawler.crawl("https://example.com")
print(data)
6. Data Visualization
from empire_chain.visualizer import DataAnalyzer, ChartFactory
# Analyze data
analyzer = DataAnalyzer()
data = """
The company saw revenue growth of $1M in Q1, $1.5M in Q2,
$2M in Q3, and $2.5M in Q4 of 2023.
"""
analyzed_data = analyzer.analyze(data)
# Create and display chart
chart = ChartFactory.create_chart('Line Graph', analyzed_data)
chart.show()
7. Using PhiData Agents
from empire_chain.phidata_agents import PhiWebAgent, PhiFinanceAgent
# Create agents
web_agent = PhiWebAgent()
finance_agent = PhiFinanceAgent()
# Use agents
news = web_agent.generate("What are the latest developments in AI?")
stock_analysis = finance_agent.generate("Analyze recent NVIDIA stock performance")
Next Steps
- Explore more examples in our Cookbooks
- Learn about Core Concepts
- Check out the API Reference
For more detailed examples, visit our GitHub repository.