pytorch-lightning

Proposed refactoring or deprecation

Aim SDK includes logging adapters for popular ML frameworks, such as pytorch_lightning, keras, tf.keras, etc.
At the moment there are no unit-tests covering the Logger classes. Need to add tests for this part.

Motivation

Improve overall quality of the product, make sure no regressions occur when code modifications are made.

Pitch

As a

🐛 Bug

found another paper-cut with instance segmentation...
it is not trivial/intuitive how to match mask with input image as output can be any size and mask is 128x128
so it is just scaled equally in each dimension is applied padding uniformly?

To Reproduce

https://www.kaggle.com/jirkaborovec/cell-instance-segm-lightning-flash#Training-with-Flash-Lightning

Additional cont

📓 New <Tutorial/Example>

Is this a request for a tutorial or for an example?
Either really.

What is the task?
Object detection

Is this example for a specific model?
Any detection model

Is this example for a specific dataset?
Any LightningDataModule that uses nvidia DALI. Maybe COCO?

I have a LightningDataModule that uses NVIDIA DALI. The train and val dataload

Add a function and module that permits spike-time decoding, as suggested by @schmitts https://twitter.com/sbstnschmtthd/status/1432343373072019461

It is always good to have more options to choose. So it would be a good idea to add more schedulers. The steps are the following:

• in conf/scheduler add a config for a new scheduler
• if this scheduler requires some other library, update requirements
• run tests to check that everything works

Example: https://github.com/Erlemar/pytorch_tempest/blob/master/conf/scheduler/cyclic.yaml
`