How to Install & Use PrivateGPT

PrivateGPT allows you to chat with your local documents, pdfs, cvs, etc.. The chat result is more relevant and private. It allows you to localize GPT and no need to go to openAI which incures additional cost (embedding).

The official documentation for privateGPT is available from Official privateGPT Guide. The following is a simplified step-by-step-guide.

Step-by-step Guide

Installation

1. Clone the repo: sh git clone https://github.com/imartinez/privateGPT

2. Install dependency cd privateGPT a. On MacOS, make sure you have 'pyenv' installed: brew install pyenv See more info about pyenv and install instructions here b. Install python 3.11 using pyenv which is required pyenv install 3.11 Note: to use 3.11 python, do this: pyenv local 3.11 c. Install Poetry: poetry install --with ui,local
3. Download Embedding and LLM models poetry run python scripts/setup
4. MacOS support: cmd CMAKE_ARGS="-DLLAMA_METAL=on" pip install --force-reinstall --no-cache-dir llama-cpp-python # Running Enter the following comamnd to run: PGPT_PROFILES=local make run # Possible Issues The best place to report the issues you are running into is through the repo's github issue report: https://github.com/imartinez/privateGPT/issues As of Oct. 26, 2023, there are 448 open issues