Skip to content

Latest commit

 

History

History
23 lines (18 loc) · 1.44 KB

README.md

File metadata and controls

23 lines (18 loc) · 1.44 KB
Raw

FuzzyAdventure

This repository consists of some explorational work done on LeNet and DCGAN. More details for each section can be found in the README files of the folders themselves.

Files

i) 4.1 SNN Model

  • The markdown file contains a systematic approach to modelling a deep convolutional SNN.
  • The Images folder contains Reference images and comparisons to past work.

ii) 4.2.1 LeNet CV

  • Contains the main notebook LeNet_75%, in which I have trained on the CIFAR10 dataset.
  • Contains the DCGAN notebook, in which I have trained an adversarial net on the CIFAR10 dataset.
  • Contains the EfNet-B0 notebook, in which I retrained a pretrained model (EfficientNet-B0) on the CIFAR10 dataset, to label images generated by the GAN generator.
  • Contains the MNIST notebook, in which I tested data augmentation techniques using the MNIST dataset.
  • Contains Data.zip, which has 1500(due to size limitations) of the 20000 images generated by the GAN.
  • Contains a GAN images folder which shows the progression of the Generated images at 8 different times during training.
  • Contains the benchmarking results for the numerous tests I performed on LeNet_75% training.
  • Lastly, for reproducibility, I have kept manual seeds, and all the .pth files can also be found in the folder.

iii) 4.2.2 NLP Reddit Flair Detector

  • Contains the main notebook NLP_Reddit, in which I have implemented the RNN Network (LSTM) and it's results.