PrivateGPT

Run ChatGPT offline on your local documents

Ever wondered if you have a massive knowledge base or document or Notion database which you want to use as a Question Answer system?

Ivan Martinez has made this possible where you can interact provately with you documents using the power of GPT, 100% private;y, no data leaks.

You can avoid having to upload any data to the cloud by downloading the GPT model and conducting file inspections on your local machine using Python.

A Python script known as PrivateGPT can be used to investigate local files by employing GPT4ALL, which is a free and open-source language model. The configuration of it is not overly complicated.

Step 1: Clone the repo

git clone git@github.com:imartinez/privateGPT.git

Step 2: Download the LLM, which is approximately 10 gigabytes in size, and save it in a new folder called “models.”

Step 3: Put the documents you want to investigate into the source_documents folder. By default, there’s a text of the last US state of the union in there

Step 4: Execute the ‘ingest.py’ script, which may take a significant amount of time.

Step 5: When you run the ‘privateGPT.py’ script, you will see a prompt appear in your terminal.

Let me know in comments what did you tried or you face any challenge.