https://github.com/ngerakines/llm_fundamentals

https://github.com/ngerakines/llm_fundamentals

Last synced: 4 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/ngerakines/llm_fundamentals
• Owner: ngerakines
• License: mit
• Created: 2024-06-04T15:10:50.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2024-07-18T13:51:36.000Z (12 months ago)
• Last Synced: 2025-03-06T21:38:17.883Z (4 months ago)
• Language: Python
• Size: 29.3 KB
• Stars: 2
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# llm_fundamentals

This repository contains research and experiments for AI / ML / CS / LLM topics.

# Setup

This project was developed against Python 3.12.2.

First, create a virtual environment:

    $ python3 -m venv ./.venv

Then, install the required packages:

    $ ./.venv/bin/pip install -r requirements.txt

macOS Users: You'll need to do a few things first.

* `brew install gfortran openblas`

* `export OPENBLAS="$(brew --prefix openblas)"`

## Ollama

Some examples require ollama to be running.

https://ollama.com/

# Usage

The `cache.py` script can be used to download and cache the models and datasets used by the project.

    $ ./.venv/bin/python3 cache.py -h

Once everything is cached, you can set some environment flags to ensure everything runs in offline mode.

    $ export HF_DATASETS_OFFLINE=1

    $ export TRANSFORMERS_OFFLINE=1

Each of the scripts has some usage documentation and help text to get you started, but generally work with defaults:

    $ ./.venv/bin/python3 coach1.py

# Tips

The tools `htop` and `nvtop` are pretty handy.

    $ brew install nvtop htop