Markdown
A fully offline, portable AI toolkit that runs entirely from a USB drive without installing any software on the host computer. It is designed to provide a secure, localized environment for writing, studying, note organization, and automated workflow processing.
Developed and maintained by Blake Von Jett. (J0hnWIcks)
Copy your AI_USB_setup.ps1 script to the root directory of your USB drive (e.g., E:\), open PowerShell 5.1, and run the initialization command:
PowerShell -NoProfile -ExecutionPolicy Bypass -File "E:\AI_USB_setup.ps1"
⚡ Note: This setup script automatically handles the architecture build. It carves out your directory skeletons (notes/, study/, runtimes/, etc.) and dynamically compiles your execution wrappers (Run_Toolkit.bat) and system documentation (HELP.txt) on the fly.
🧠 Adding the AI Programs & Models
Because this repository remains completely lightweight, it does not bundle heavy runtime binaries or multi-gigabyte model weights. Follow these quick steps to fully arm your toolkit:
Step 1: Download the Inference Engine
Download: llama-b9701-bin-win-cpu-x64.zip (16.4 MB)
Placement: Extract the contents of this zip file (including llama-cli.exe and its supporting .dll files) directly into:
AI_USB\runtimes\llama.cpp\
Step 2: Download the AI Models (GGUF)
Download your preferred quantized linguistic models directly from Hugging Face:
📥 Phi-3-Mini (Q4_K_M) (~2.4 GB) — Highly optimized, fast CPU performance.
📥 Mistral-7B-Instruct-v0.3 (Q4_K_M) (~4.4 GB) — Deep reasoning and complex structured text generation.
Placement: Drop your downloaded .gguf files straight into:
AI_USB\runtimes\llama.cpp\models\
🛠️ Execution
Once your runtimes and models are dropped into place:
Open your AI_USB\config\settings.json and verify your default_model path points to the file you want to use.
Double-click Run_Toolkit.bat in the root directory to launch your localized, completely offline AI session.