Frédéric Branchaud-Charron

Oh I like that! I think we should do this. There is a lack of leaderboards in active learning which makes research more difficult. And websites such as [`paperswithcode`](https://paperswithcode.com/sota/active-learning-on-cifar10-10000) do...

Hello, As far as I know, this is not a subject that academia has worked on. BaaL should be able to handle it, as we do not impose any format....

This is a quick example on how to use our heuristics (BALD, Entropy) with NER and in segmentation. I hope this helps you, if I missed anything, please tell me....

*Reopening for visibility* Yes I think that would be it. @parmidaatg worked on multilabel in the past and might give more insight.

We've just moved to a new documentation system (mkdocs) that should help us better structure our tutorials. If someone could run some experiements showing that BALD is at least better...

I think we made good progress on this on #262. I'll close this one, but for now the code should work for y'all to run experiment with.

What do we want to log? 1. Performance metrics 2. Uncertainty over time

From my current understanding, this timeout is only used when we need to get the results. One of: 1. All tasks are done 2. One worker died Your function should...

Hello! This is due to how pandas write numpy arrays to csv. [Source](https://stackoverflow.com/questions/54753179/to-csv-saves-np-array-as-string-instead-of-as-a-list) To fix this, you can convert them to list yourselves. ```python df = ds.to_pandas() df['int'] = df['int'].apply(lambda...

Adding to the conversation, Dask also uses `fsspec` for this feature. [Dask: How to connect to remote data](https://docs.dask.org/en/stable/how-to/connect-to-remote-data.html) Happy to help on this feature :D