This recipe contains Deep learining solutions using PyTorch1.x
This repository contains jupyter notebooks and python files for Deep Learning solutions in PyTorch.
- Chapter 1 - Working with Tensors using PyTorch:
- Installing PyTorch
- Creating Tensors in PyTorch
- Interoperating NumPy bridge
- Gradients and no gradients
- Viewing tensors in PyTorch
- Chapter2 - Dealing with Neural Networks:
- Defining the neural network class
- Creating a fully connected network
- Defining the Loss Function
- Implementing Optimizer
- Implementing Dropouts
- Implementing Functional APIs
- Chapter3 - Convolutional Neural Networks for Computer Vision:
- Exploring Convolutions
- Exploring Pooling
- Exploring Transform
- Performing Data Augmentation
- Loading image data
- Defining CNN Architecture
- Training an image classifier
- Chapter4 - Recurrent neural networks for NLP:
- Performing Tokenization
- Creating Fields
- Developing Dataset
- Developing Iterators
- Exploring Word embeddings
- Building a LSTM network
- Building a Multilayer LSTM network
- Building a Bidirectional LSTM network
- Chapter5 - Transfer Learning and TensorBoard:
- Adapting a pretrained model
- Implementing model training
- Implementing model testing
- Loading the dataset
- Defining the TensorBoard writer
- Training the model and unfreezing layers
- Chapter6 - Exploring Generative Adversarial Networks:
- Creating a DCGAN Generator
- Creating a DCGAN Discriminator
- Training DCGAN model
- Visualizing DCGAN results
- Running PGGAN with PyTorch hub
- Chapter7 - Deep Reinforcement Learning:
- Introducing Deep Reinforcement Learning
- Introducing OpenAI gym - Cartpole
- Introducing DQN
- Implementing DQN Class
- Training DQN
- Introducing DeepGA
- Generating Agents
- Selecting Agents
- Mutating Agents
- Training DeepGA
- Chapter8 - Productionizing AI models in PyTorch: