channing/MNISTTest
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
Simple use of a CNN model to recognize handwritten digits from the MNIST dataset.
22 commits 1 branch 0 tags 41 KiB
Find a file
Exact
Channing Bellamy abeec8b676 update README
2026-01-11 15:59:08 +02:00
.gitignore Update file timestamps 2024-04-18 20:58:59 +02:00
classifier.py Improve evaluation efficiency with no_grad() 2024-04-19 17:08:24 +02:00
log.txt Add training run to log 2024-04-19 14:32:14 +02:00
README.md update README 2026-01-11 15:59:08 +02:00

README.md

Channing Bellamy BLLCHA013 19 April 2024 CSC3022F ML Assignment 1

This program uses a CNN model to recognize handwritten digits from the MNIST dataset.

Please ensure PyTorch and TorchVison are available, for example, using pip3: pip3 install torch torchvision

This program consists of a single Python file named "classifier.py". Example invocation: python3 classifier.py

"classifier.py" contains the program logic, including a CNN model, tranforms, and file reading logic.

Tested to work on:

  • Linux 6.8.7-arch1-1 (x86_64) with Python 3.11.8, torch 2.2.2, and torchvision 0.17.2.
  • macOS 14.4.1 Sonoma (x86_64) with Python 3.12.3, torch 2.2.2, and torchvision 0.17.2.