Skip to main content

LocalLab: Run language models locally or in Google Collab with a friendly API

Project description

🚀 LocalLab: Your Personal AI Lab

LocalLab lets you run AI language models on your computer or Google Colab - no cloud services needed! Think of it as having ChatGPT-like capabilities right on your machine.

🤔 What is LocalLab?

LocalLab consists of two parts working together:

graph LR
    A[LocalLab Server] -->|Runs| B[AI Models]
    C[Your Code] -->|Uses| D[LocalLab Client] -->|Talks to| A

The Server (Your AI Engine)

Think of the server as your personal AI engine. It:

  • Downloads and runs AI models on your computer
  • Manages memory and resources automatically
  • Optimizes performance based on your hardware
  • Provides a simple API for accessing models

You can run it:

  • On your computer (local mode)
  • On Google Colab (free GPU mode)

The Client (Your AI Controller)

The client is how your code talks to the AI. It:

  • Connects to your LocalLab server
  • Sends requests for text generation
  • Handles chat conversations
  • Processes multiple requests at once
  • Streams responses in real-time

✨ How It Works Together

When you use LocalLab:

  1. Server Setup

    from locallab import start_server
    start_server()  # Server starts and loads AI model
    
  2. Client Connection

    from locallab.client import LocalLabClient
    server_url = "http://localhost:8000" # or "https://your-ngrok-url.ngrok.app"
    client = LocalLabClient(server_url)
    
  3. AI Interaction

    # Your code sends requests through the client
    response = await client.generate("Write a story")
    print(response)  # Server processes and returns AI response
    

💡 Quick Examples

# Generate text
response = await client.generate("Hello!")

# Chat with AI
response = await client.chat([
    {"role": "user", "content": "Hi!"}
])

# Process multiple prompts
responses = await client.batch_generate([
    "Write a joke",
    "Tell a story"
])

➡️ See More Examples

💻 Requirements

Local Computer:

  • Python 3.8+
  • 4GB RAM minimum
  • GPU optional (but recommended)

Google Colab:

  • Just a Google account!
  • Free tier works fine

📚 Getting Started

1. Choose Your Path

New to AI/Programming?

  1. Start with our Getting Started Guide
  2. Try the Basic Examples
  3. Join our Community

Developer?

  1. Check API Reference
  2. See Client Libraries
  3. Read Advanced Features

2. Read the Docs

Our Documentation Guide will help you:

  • Understand LocalLab's features
  • Learn best practices
  • Find solutions to common issues
  • Master advanced features

🌟 Features

  • Easy Setup: Just pip install and run
  • Multiple Models: Use any Hugging Face model
  • Resource Efficient: Automatic optimization
  • Privacy First: All local, no data sent to cloud
  • Free GPU: Google Colab integration

➡️ See All Features

🔍 Need Help?

📖 Additional Resources


Made with ❤️ by Utkarsh Tiwari
GitHubTwitterLinkedIn

Project details


Release history Release notifications | RSS feed

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

locallab-0.4.43.tar.gz (53.4 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

locallab-0.4.43-py3-none-any.whl (57.5 kB view details)

Uploaded Python 3

File details

Details for the file locallab-0.4.43.tar.gz.

File metadata

  • Download URL: locallab-0.4.43.tar.gz
  • Upload date:
  • Size: 53.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.1.0 CPython/3.9.21

File hashes

Hashes for locallab-0.4.43.tar.gz
Algorithm Hash digest
SHA256 56952c39416f791db89f17ed0951d598dc526edb9d0d5e4ed34d02bbf4eabc8e
MD5 6c1ae85bcc2ddd05af45af4600fa58f8
BLAKE2b-256 b985cc5af94c68e5fdc3f04a35d4009a0d721f8a573a86221a38114d2f723080

See more details on using hashes here.

File details

Details for the file locallab-0.4.43-py3-none-any.whl.

File metadata

  • Download URL: locallab-0.4.43-py3-none-any.whl
  • Upload date:
  • Size: 57.5 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.1.0 CPython/3.9.21

File hashes

Hashes for locallab-0.4.43-py3-none-any.whl
Algorithm Hash digest
SHA256 c1a15c7988e659ec77a5336df1f4c1c4e8df70dfdc7254e5d091a05890f98c93
MD5 efe68a13b5dcdd7cb5a995f3ccd55031
BLAKE2b-256 a20e17e853d783125a030cf0b3fe4c90c2638482d700ad90e946b32cd85a3183

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page