Some notes, tutorials, and some experimentation with the fast ai library (https://github.com/fastai/fastai). The notebooks are primarily aimed for people who want to use fastai as a library and perhaps don't want to go through all the lectures to understand its functioning. While I definitely encourage everyone to see the lectures (which are absolutely fantastic), I do feel there is a need for tutorials to use it out of the box. While the courses folder in the library does contain more or less everything, I feel there is a need to put everything together nicely and compactly which is the main aim for creating the repository. This should allow people who are already familiar with pytorch write their own code for research or any other purposes much faster.

Also those who have seen the lectures might find few interesting things from the notebooks.

• Installation
• Quick Introduction to FAI
• Using Cadence Model
• Creating your own model
• Using Visdom with FAI
• Logging with FAI
• Using forward hooks for feature extraction
• Using CAM for visualization
• Using Grad-CAM for visualization
• Using backward hooks for unit testing
• Creating own DataSet for Siamese
• Using Different number of inputs for the model using ColumnarModelData
• Appending a New Feature with CNN features
• Using Sampler for Class Imbalance