Things you will learn in Notebook
- How to train a pretrained model in Pytorch
- How to work with pytorch like tensor, numpy to tensor, named tensor, storage in pytorch etc.
- How to implement One-hot encoding in Pytorch.
- How to handle Time series data in Pytorch
- How to handle Images in Pytorch, changing dimension as required in pytorch module.
- How to perform Word to Index conversion
- How to split data between train and validation
- How to build model in Pytorch
- How to find derivative in pytorch
- How to move from training mode to evaluation mode
- How to train a network with help of Optimizer and Criterion (Loss Function).
- Creating standard project using Pytorch for both NLP and CV tasks via torchtext and torchvision
- Implementing Image Classification using Pytorch from Scratch
- Loading and preprocessing Textdata in torchtext
- Using Pretrained Word Embedding in torchtext
- DataLoaders
- How parameters i.e. coefficients of a neural network is initialized
- Training GANs Network.
- Saving and Loading GANs Network for Image Generation.
- state_dict() - save the model
- What happens when you don't normalize data
- How to avoid making gradient descent from reaching nan or inf
- Difference between save() and state_dict()