Deep Generative Models - Course Assignments

University of Tehran, Fall 2023

This repository contains my homework submissions for the "Deep Generative Models" course offered at the University of Tehran in Fall 2023. The course was instructed by Dr. Tavassolipour and Dr. Sadeghi. The assignments cover a range of topics in deep generative modeling, including probabilistic graphical models, variational autoencoders, generative adversarial networks, and more.

Homework 1

Topics Covered:

• Probabilistic Graphical Model
• Causality
• Variational Inference

Link to HW1 Folder

Homework 2 - Part 1

Topics Covered:

• Variational Auto Encoder (VAE)
• Normalizing Flow

Link to HW2 Part 1 Folder

Homework 2 - Part 2

Topics Covered:

• Generative Adversarial Networks (GAN)
• Diffusion Models

Link to HW2 Part 2 Folder

Homework 3

Topics Covered:

• Fine-tuning Large Language Models (LLM) using LoRA PEFT
• Prompt Engineering
• SpeechT5 Text-to-Speech (TTS) Fine-tuning

Link to HW3 Folder

Acknowledgments

Special thanks to Dr. Tavassolipour and Dr. Sadeghi for their guidance and instruction throughout the course.

---