Skip to content

installation.md

Latest commit

 

History

History
69 lines (54 loc) · 2.87 KB

installation.md

File metadata and controls

69 lines (54 loc) · 2.87 KB

NeuroKey Installation Guide

Welcome to NeuroKey! This project is transitioning to a high-performance Rust + Tauri architecture with a PyTorch GPU Backend.

Prerequisites

To run or develop NeuroKey, you need the following dependencies installed on your system:

1. Python & Machine Learning

NeuroKey uses Python 3 for data scraping and PyTorch for GPU-accelerated layout evaluation.

  • Python 3.10+
  • PyTorch: Follow the official guide at pytorch.org to install the correct version for your GPU.

2. Rust & Tauri

The core CPU engine and the Desktop GUI are built with Rust and Tauri.

  • Rust: Install via rustup (rustup.rs)
  • Node.js: Version 18+ (nodejs.org)
  • Tauri Prerequisites: Depending on your OS (Linux/Windows/macOS), you must install the OS-specific Tauri dependencies. See the Tauri Prerequisites Guide.

Installation Steps

Step 1: Clone the Repository

git clone https://github.com/yourusername/neurokey.git
cd neurokey

Step 2: Set up Python Virtual Environment (Optional but Recommended)

python3 -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

Step 3: Install Python Dependencies

pip install -r requirements.txt
# If you haven't installed PyTorch globally, install it here.
# Example for CUDA 12.1:
# pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121

Step 4: Download Corpus Data

Before running any optimizations, you need to download the corpus data which will be used to generate the n-gram frequencies.

bash get_data.sh

Note: This script will download gigabytes of code from various GitHub repositories to build a mathematically accurate model of modern programming languages. It may take some time.

Step 5: Start the Web UI (Legacy Python Mode)

Currently, NeuroKey runs via a Flask web interface.

python3 gui.py

Open http://localhost:5000 in your web browser. From here, you can select between the Simulated Annealing (CPU) engine or the PyTorch Tensor Batching (GPU) engine.

Future Release: Tauri Native App

We are actively working on NeuroKey v2.0 which bundles everything into a single desktop application.

Key GUI Features

  • Advanced Ergonomic Weights: Fine-tune SFB penalties, Roll bonuses, and Effort weights directly from the UI.
  • Live Convergence Charts: Monitor optimization progress with real-time score tracking.
  • Layout Presets: Instantly switch between QWERTY, Dvorak, and Colemak.
  • System Telemetry: Real-time monitoring of CPU/GPU load and optimization logs.

Note: If you are a developer, ensure you run npm install to fetch the new recharts and lucide-react dependencies.