Experimentation with key aspects of fastai v3 code

I too often find myself searching though the lesson notebooks for working examples of core aspects of fastai.

Here I have put together notes grouping examples around particular topics.

creating models: models.ipynb

augmentation/transforms: transforms.ipynb

learner: fit_one_cycle.ipynb

transfer learning: models_transfer_learning.ipynb

TODO

loading data - DataBunch - ImageDataBunch, tabular, text: dataloader.ipynb

inference: web_apps.ipynb

loss functions - mse, cross entropy, bce, dice, lovasz: loss_functions.ipynb - see lesson3 Create a regression model [1:38:59], lesson5 - Going back to Lesson2 SGD notebook [1:19:16], MNIST SGD [1:23:59] - lesson5-sgd-mnist.ipynb

NLP - see lesson3 IMDB [1:41:07] lesson3-imdb.ipynb, lesson4 Natural Language Processing (NLP) [2:00]

tabular - see lesson 4 Tabular [33:10] - lesson4-tabular.ipynb

collaborative filtering - see lesson4 - Collaborative Filtering [53:08] - lesson4-collab.ipynb