Frédéric Branchaud-Charron

Results 137 comments of Frédéric Branchaud-Charron

Hi @mariosasko thank you for your review. I agree that `load_dataset('sql',...)` is a bit weird and I would be happy to remove it. To be honest, I only added it...

Thank you so much! I was missing a lot of things sorry about that. LGTM

I found an issue where `active_learning_process.ipynb` generates a search map of 32Mb which kills the search bar. Investigating how to reduce the final animation size.

Everything is setup! Could @GeorgePearse and @parmidaatg re-review and then we deploy?

Oh good point! If you could make the PR that would be great, otherwise I'll have time over the weekend

Hmm good question. I think for now we can keep Heuristic and Strategy together. So yes something like "power_bald", "power_entropy" would make sense to me.

@Borda we will deprecate our Lightning integration in favor of Lightning-flash integration. We did a [blogpost](https://devblog.pytorchlightning.ai/active-learning-made-simple-using-flash-and-baal-2216df6f872c) with Thomas a few months ago.

Would it be labelled randomly or according to BALD too? This can be done by wrapping val_set in ActiveLearningDataset.

I feel we still need to do this no? We should meet this week to settle on a set of topic we need to discuss. @rafapi would you be interested...