Skip to content

Latest commit

 

History

History
29 lines (18 loc) · 1.19 KB

README.md

File metadata and controls

29 lines (18 loc) · 1.19 KB

Diffusion with FastAI

Overview

This course repository tracks my progress following part 2 of fast.ai's 2022-23 course. During the course we will be building a deep learning framework called "miniai" from scratch. It is created from Jupyter notebooks using nbdev.

As a side note, I plan to adapt parts of this library to aid students in NUS CS2109S (Introduction to AI and Machine Learning) in their learning of the heavy syllabus.

Install

Clone this repo, cd to it, and run:

pip install -e .

(You may optionally create a virtual environment first)

This installs the miniai library which I will be learning to create through the fast.ai course.

Using Jupyter notebooks

  1. Notebooks can be found in the nbs subfolder.
  2. Copy the notebook from the nbs subfolder into the miniai subfolder.
  3. Run the notebooks using the local environment (with miniai module installed) as a Jupyter kernel (VSCode functionality).
  4. Running the cells with export functionality periodically exports the code within the notebooks to the files in miniai (preliminary versions are already initialised but feel free to iterate)