This repository provides structured code examples and project templates designed for classroom instruction in machine learning and neural networks. It offers reference implementations of deep learning models for both computer vision and natural language processing tasks, built using PyTorch as the core framework.
The codebase is organized as a modular project template with separate directories for data handling, model definitions, and training scripts, promoting reusability and clarity. It includes predefined pipelines for image classification and text processing, along with a command-line interface for initiating training with customizable hyperparameters. Checkpoint-based model persistence allows training to be resumed, and a separate evaluation script computes accuracy and loss on held-out test data.
The examples cover training vision models for image classification and NLP models for text processing, with configuration-driven experiment setup to manage hyperparameters and dataset paths. The documentation and code are structured to support academic instruction in deep learning concepts.
