Infera Lite is an easy-to-use chrome extension UI that connects users to a decentralized AI inference network, allowing them to contribute their GPU for AI model inference. This guide covers the extension’s features and navigation.

​
Download Infera lite

  1. Go to the Chrome Web Store
  2. Click “Add to your Chrome” to install onto your browser.

​
Onboarding pages will pop up once you add extension

​
Get started

​
Choose your operating system

​
Install Ollama

​
Follow the infrastruction of your Operating system

​
You’re all set!

​
Infera Lite UI

Upon opening the extension, users will see a dashboard that includes key components related to their node’s performance, points earned, and system status. Below is a breakdown of the UI elements:

​
Home screen

The Home screen is the landing page of the extension and gives an overview of the node’s current status.

  • Start/Stop Button: Allows you to toggle the node on or off.
  • Active Status: Indicates whether the node is running, displaying either “Active” (active and awaiting jobs) or “Inactive” (node deactivated).
  • Points: Displays the points earned from processed tasks and uptime on the network.
  • Uptime: Shows the node’s total uptime.

​
Reputation & Node details

This section provides a real-time snapshot of your node’s performance:

  • Device details: Displays information about the node’s GPU, CPU, RAM, VRAM, and available RAM.
  • Tasks Completed: Shows the number of inference tasks your node has processed.
  • Reputation: A pie chart representing your node’s reputation on the Infera network based on speed and reliability.

​
Models page

This page lists all the LLM models that users can install to support the Infera network, starting with every state-of-the-art open-source model:

  • Llama 3.1 8B
  • Llama 3.2 1B / 3B
  • Phi 3.5 4B
  • Mistral 7B
  • Qwen 2 8B
  • Gemma 2 9B
  • Yi large 9B

​
Privacy Assurance

  • No Personal Data: We don’t collect or store your personal information.
  • Encryption: All communications are encrypted for security.
  • Local Processing: All inference happens on your device; nothing is sent externally.
  • No Tracking: Your activities are never tracked.
WindowsDocker Setup