Ravin Kohli

Reg cocktails

1

## Types of changes - [x] Breaking change (fix or feature that would cause existing functionality to not work as expected) Note that a Pull Request should only contain one...

Updates embedding layers to pytorch embedding. ## Types of changes - [x] Breaking change (fix or feature that would cause existing functionality to not work as expected) - [ ]...

[FIX] Docker publish error

1

## Types of changes - [x] Bug fix (non-breaking change which fixes an issue) Note that a Pull Request should only contain one of refactoring, new features or documentation changes....

[ADD] fit ensemble

1

This PR enables fitting an ensemble after `search` has finished. It also fixes issue #299. ## Types of changes - [x] New feature (non-breaking change which adds functionality) Note that...

Due to our current implementation of Embedding module, we are forced to one-hot encode all categorical columns. This leads to an explosion in the memory usage. We can avoid this...

[REFACTOR] Reduce walltime of examples

1

Currently, the examples are all searching for 5 minutes which I think for the datasets we use is too long. Especially for the purpose of demonstrating the functionality of AutoPyTorch,...

We should make a docker image for running autopytorch on github

[REFACTOR] Too many variables in `search` and `fit_pipeline`.

4

Why do we need that many variables now? Most information is in the BaseDataset, isn't it? Let's try reducing variable numbers as much as possible, because it will be painful...

We should add functionality to display training, testing and validation curves for the incumbent pipeline directly from the BaseTask API or accessible from it. We should also add an example...

Hey, if we plan to remove support for python 3.6 and even python 3.7, we should add type annotations to the dataset properties and the get_properties methods. This can be...