nighting0le01/generative-Models

Generative Models Implemented Pytorch

This repo includes From scratch Implementation of Various Generative Model, Highlighted as below:

1. Generative Adversarial Networks: CycleGAN, Pix2pix, LSGAN, WGAN,WGAN-GP, Vanilla GAN, patch Discriminator.
2. Variational Autoencoder: AE, VAE, Beta- VAE
3. Diffusion: End to End implmentation of DDPM(Stochastic) and DDIM (deterministic) Diffusion Models with Cosine and Linear Noise Schedule and 25 block (12 Encoder, 12 Decoder, 1 latent) Unet conditioned on Timesteps and Text.

Dataset

CUB-200 Bird Species Dataset The CUB-200 dataset, short for "Caltech-UCSD Birds-200," is a widely used benchmark dataset for fine-grained visual categorization and bird species recognition. This dataset is specifically designed for training and evaluating computer vision models on the task of identifying and classifying bird species from images.

Software setup

Please use Python 3.8.

Run the following script to setup all dependencies:

chmod +x setup.sh
./setup.sh /path/to/python_env/lib/python3.8/site-packages # e.g. ./download_cub.sh ~/miniconda3/envs/gen_ai/lib/python3.8/site-packages

##  Directories

.
├── attributes.txt
├── cub_clean_custom_na.npz
├── data_gan
├── datasets
│   ├── CUB_200_2011
│   ├── CUB_200_2011_32
│   └── CUB_200_2011.tgz
├── diffusion
│   ├── data_ddim
│   ├── data_ddpm
│   ├── epoch_199.pth
│   ├── inference.py
│   ├── model.py
│   ├── __pycache__
│   ├── README.md
│   ├── unet.py
│   └── utils.py
├── gan
│   ├── attributes.txt
│   ├── data_gan
│   ├── data_ls_gan
│   ├── data_wgan_gp
│   ├── __init__.py
│   ├── LSGAN.py
│   ├── networks.py
│   ├── __pycache__
│   ├── README.md
│   ├── train.py
│   ├── utils.py
│   ├── Vanilla_Losses.py
│   └── WGAN.py
├── image-1.png
├── image-2.png
├── image-3.png
├── image-4.png
├── image-5.png
├── image.png
├── README.md
├── requirements.txt
├── resize_dataset.py
├── setup.sh
└── vae
    ├── data
    ├── model.py
    ├── __pycache__
    ├── README.md
    ├── train.py
    └── utils.py

Results