https://github.com/birdx0810/rcgan-pytorch

This repository is a non-official implementation of Recurrent (Conditional) GAN (Esteban et al., 2017) using PyTorch.
https://github.com/birdx0810/rcgan-pytorch

deep-learning generative-adversarial-network hacktoberfest pytorch time-series

Last synced: about 2 months ago
JSON representation

This repository is a non-official implementation of Recurrent (Conditional) GAN (Esteban et al., 2017) using PyTorch.

• Host: GitHub
• URL: https://github.com/birdx0810/rcgan-pytorch
• Owner: birdx0810
• License: mit
• Created: 2020-10-11T11:00:22.000Z (over 4 years ago)
• Default Branch: main
• Last Pushed: 2022-07-26T05:40:48.000Z (over 2 years ago)
• Last Synced: 2024-10-30T03:57:55.768Z (3 months ago)
• Topics: deep-learning, generative-adversarial-network, hacktoberfest, pytorch, time-series
• Language: Jupyter Notebook
• Homepage:
• Size: 3.7 MB
• Stars: 4
• Watchers: 2
• Forks: 1
• Open Issues: 2
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# RCGAN_PyTorch

This is a PyTorch implementation (kinda) of Recurrent (Conditional) GAN (Esteban et al., 2017). 

> :warning: WARNING!!!

> - This implementation is written for other purposes, not for experiments in the original paper.

> - There are some known issues that I've haven't got time to resolve (see issue [#1](https://github.com/d9n13lt4n/rcgan-pytorch/issues/1#issuecomment-935364784)).

## Getting Started

### Installing Requirements

This implementation assumes Python3.8 and a Linux environment with a GPU is used.

```bash

cat requirements.txt | xargs -n 1 pip install --upgrade

```

### Directory Hierarchy

```bash

data/                         # the folder holding the datasets and preprocessing files

  └ data_preprocessing.py     # the data preprocessing functions

metrics/                      # the folder holding the metric functions for evaluating the model

  ├ dataset.py                # the dataset class for feature predicting and one-step ahead predicting

  ├ general_rnn.py            # the model for fitting the dataset during TSTR evaluation

  ├ metric_utils.py           # the main function for evaluating TSTR

  └ visualization.py          # PCA and t-SNE implementation for time series taken from the original repo

models/                       # the code for the model

output/                       # the output of the model

main.py                       # the main code for training and evaluating TSTR of the model

requirements.txt              # requirements for running code

run.sh                        # the bash script for running model

visualization.ipynb           # jupyter notebook for running visualization of original and synthetic data

```