Here is a compilation of a range of Deep Learning algorithms for interesting tasks such as:
- Computer Vision - Image classification
- Natural Language Processing - Summarization
- Generative Adversarial Networks - Style transfer
- Transfer Learning using pretrained models
- Predictions
- Cancer Image Classifier from Google Scraped data
- Generating Characters using Long-Short Term Memory/Recurrent Neural Networks
- Building Multi Layer Perceptron fronm scratch and testing it on MNIST Database
- Quick Pet Classification using FAST.AI
- Plant Disease Image classification using FAST.AI
- Flower Species Indentification using Pytorch
- Style Transfer Algorithm via pre-trained models (VGG-19) using Pytorch
- World Class Image Classifier of Cats and Dogs using FAST.AI
- Classifying CIFAR-10 database using Pytorch
- Make sure you have Jupyter Notebooks in your computer or in an online space.
- Make sure the computer has a GPU for training the models both locally via graphic cards such as NVIDIA or online spaces such as Google Colab
Just open the notebooks in your gpu space of choice such as FloydHub, Google Colaboratory, PaperSpace or any other workspace of your choice including a local setup!
These are implemented with either or libraries
- PyTorch - The python framework used
- Fast.ai - This is also a python framework built on top of Pytorch
Herman Njoroge Chege // HURU School Artificial Intelligence Department
This project is licensed under the MIT License, however some specific algorithms might be subject to their respective licenses.
These resources are courtesy of Udacity, Facebook, Jeremy Howard and other prolific researchers in the fast evolving deep learning space.