Advanced Deep Learning

Computer Exercises & Implementations in Python (PyTorch)

This repository contains my computer exercise solutions for the Advanced Deep Learning course.
The work focuses on implementing and experimenting with advanced neural architectures, optimization techniques, and modern learning paradigms using Python and PyTorch.
To get started with the exercises, check out the Setup & Getting Started Guide

---

Topics Covered

• Loss functions & optimization
• Residual Networks (ResNets) & 3D CNNs
• Sequence models: RNNs, LSTMs, ConvLSTMs
• Transformers & attention-based architectures
• Contrastive learning (Triplet Loss, N‑pair Loss)
• Graph Neural Networks (GCN, Spatio‑Temporal GCN)
• Generative models: GANs, VAEs, Diffusion Models
• Self‑supervised learning & representation learning
• Uncertainty estimation & Bayesian neural networks

---

Repository Structure

Each folder contains:

• Jupyter notebooks
• Python scripts
• Model training outputs (if generated)
• Visualizations and analysis

---

Technologies Used

• Python 3
• PyTorch
• NumPy, Pandas
• Matplotlib, Seaborn
• Jupyter Notebook
• Google Colab

GPU acceleration is recommended but not required.

---

Notes

• This repository contains only my own exercise implementations.
• No confidential course materials or restricted content are included.
• All notebooks follow the required structure and only modify designated code blocks.