Victor Turrisi

Hi, `--scale_loss` was a parameter that barlow had in their original version. It is not mentioned in the paper, but you can check the commit that removed it in their...

Exactly, our implementation should match the original implementation from barlow.

Hmm, I think that's not exactly what I need. What I want to do is access the indexes outside of the pipeline, something like ``` train_loader = Loader(....) for idx,...

@GuillaumeLeclerc sorry, I didn't get it. If `with_indices` is problematic, what's the other option?

Personally speaking, I don't think it's something urgent. I'm waiting for implementations of common transformations (like color jitter), or for me to have time to implement them myself, to switch...

Hey. I'm a bit busy this week, but I'll try to share the checkpoints as soon as possible.

Indeed having end-to-end tests on the methods themselves is something that we need. However, using a toy dataset is not something that can correctly evaluate all methods. A decent middle...

@turian I think the most important tests would be to validate the performance of the methods (as the other features are easily tested by the unit tests). Linear evaluation is...

@turian been quite busy this week, but I'll try to get back here as soon as possible. Regardless, the end to end tests that you mentioned can be easily done...

The latest commit has tests for all scripts in `tests/scripts`. I think they are sufficient for the purpose of checking if there's something wrong with any script or method. About...