MNIST Workshop

Welcome to the MNIST Workshop repository! This project contains all the materials, workbooks, and reference guides needed to learn how to build, train, and evaluate a machine learning model on the classic MNIST handwritten digit dataset using PyTorch. This repo, contains all the material discussed during the workshop.

Repository Structure

Here is a breakdown of the files included in this workshop:

• MNIST_WORKSHOP_WorkBook.ipynb: The main workbook for the workshop. This Jupyter Notebook contains the code skeleton and instructions you will follow to build your model.
• MNIST_WORKSHOP_WorkBook-Solution.ipynb: The completed version of the workbook with full solutions. Use this if you get stuck or want to compare your results.
• MNIST Workshop.pptx: The presentation slides accompanying the workshop.
• Documentation & Cheatsheets:
  • pytorch_documentation.md: Quick reference for PyTorch commands.
  • numpy_cheatsheet.md: Common NumPy operations.
  • pandas_cheatsheet.md: Common Pandas data manipulation commands.
  • matplotlib_cheatsheet.md: Guide for plotting with Matplotlib.
  • python_documentation.md: General Python syntax reference.

Why Run on Google Colab?

For this workshop, we recommend using Google Colab. It is an incredibly useful tool for Deep Learning for several reasons:

• No Installation Required: Colab runs entirely in the cloud, so you don't need to install Python, Anaconda, or Jupyter locally.
• Zero Dependency Issues: Essential libraries like PyTorch, NumPy, Pandas, and Matplotlib come pre-installed. You can skip the headache of version conflicts and environment management.
• Free GPU Access: Colab provides free access to GPUs (Graphics Processing Units), which significantly speeds up the training of neural networks compared to running on a standard laptop CPU.

Getting Started with Google Colab

Follow these simple steps to get the workbook running in your browser:

Step 1: Download the Material

1. Click the green <> Code button at the top right of this repository.
2. Select Download ZIP.
3. Extract the ZIP file to a folder on your computer.

Step 2: Upload to Colab

1. Navigate to Google Colab.
2. If prompted with a "Welcome to Colaboratory" window, click the Upload tab. (If you are already in a notebook, go to File > Upload notebook).
3. Click Browse or drag-and-drop the MNIST_WORKSHOP_WorkBook.ipynb file from the folder you just extracted.

Step 3: Start Coding

Once the notebook loads, you are ready to go! You can run code cells by clicking the "Play" button on the left side of each cell or by pressing Shift + Enter.

Note: The MNIST dataset is built into the torchvision library, so the code in the notebook will automatically download the data into the Colab environment for you. You do not need to manually upload dataset files.

Additional Resources

• Google Slides Presentation
• PyTorch Official Documentation
• MNIST Dataset

Feel free to reach out to the education team on WEAP's discord server, regarding any help you need, whether it be regarding solving the workbook or understaning how a piece of code works or understanding a concept. The education team will be more than happy to assist.

---