AntNet

AntNet is a decentralized AI grid and a Distributed Intelligence Swarm that distributes heavy text-processing tasks across a network of volunteer devices. It accepts large .txt files, splits them into manageable chunks, and processes user-defined requests in parallel using local phi3 models. This design enables scalable, low-cost AI inference without relying on cloud GPUs or external APIs.

---

Features

• 📄 Text File Input
  Accepts .txt files as the primary data source.

• ✂️ Configurable Chunking
  Supports text chunking with adjustable sizes ranging from 500 to 5000 characters.

• 📝 Task Request Interface
  Allows users to specify what actions or analysis should be performed on the text data via the master web server.

• 🌐 Distributed Worker Execution
  User-defined requests are forwarded from the master server to one or more worker applications running on separate machines.

• ⚡ Parallel Processing
  Workers process text chunks in parallel, reducing processing time and LLM computation cost, and publish results back to the master server.

• 🧠 Master–Worker Architecture
  Implements a master–worker model where the master server coordinates tasks and worker nodes perform distributed computation.

• 💬 Integrated Chat Interface
  The master web server provides a chat feature that compiles processed data and forwards it to a swarm-based AI model for final responses.

---

MASTER–WEB SERVER

Live Application:
https://ant-net-frontend-lsn2.vercel.app/

You are required to install and setup the AntNet_Lite_Setup and run the worker or else it will show "Waiting for Connection".

The master web server acts as the central control layer of the AntNet system and provides the primary user interface.

• Users upload their data in .txt format through the web interface.
• Users can configure:
  • Chunk size range
  • Number of fragments to distribute across workers
• Users submit a custom request or instruction describing how the uploaded text should be processed.
• The master server distributes the workload to multiple worker machines running the worker application.
• Worker nodes process the assigned text chunks in parallel and return results to the master server.
• The master server compiles the processed chunks and presents the final output to the user through the web interface or chat system.

---

AntNet_Lite_Setup Installation Guide

Setup Instructions

Step 1: Install and Run the Worker

• Install the setup package and run it on your local machine.
• It is strongly recommended to run the worker on a separate machine from the main server to avoid excessive CPU and RAM usage.

Step 2: Automatic Ollama & Model Setup

• If Ollama is not already installed, the setup will automatically install it.
• The phi3:mini (3.8B parameters) AI model will also be pulled automatically.

Step 3: Using an Existing Ollama Installation

• If Ollama is already installed or running in environments such as:
  • Docker
  • Docker Compose
  • Kubernetes
  • Virtual Machines
• Ensure Ollama is running before starting the worker setup.
• The worker will not detect Ollama if it is not active at startup.

Step 4: Parallel Processing (Recommended)

• Run the worker on multiple machines for improved performance.
• Enables parallel processing and faster overall data handling.

Step 5: Use GPU-Enabled Machines (Recommended)

• For faster data processing and improved inference performance, it is recommended to run workers on GPU-enabled machines.
• More powerful GPUs significantly reduce processing time, especially when handling large text datasets.