Pavel Izmailov

Hi @Parskatt, thank you for noticing this! I believe you are correct, and this was not intended. Practically, I believe this does not make a difference, but we need to...

Right :) We are also typically not training these covariance parameters and just keep them fixed at `1.` throughout. Definitely a mistake on our side though.

Hey @petteriTeikari thank you for pointing this out, I ran into this issue myself. In this project, we don't really `_assess_dimension_` in this work, I believe it has been added...

Hey @nilsleh we actually have an example of using SWAG for a regression problem here: [notebook](https://github.com/wjmaddox/drbayes/blob/master/experiments/synthetic_regression/visualizing_uncertainty.ipynb). Would you mind comparing your implementation to that to see if there are any...

Hey @olccihyeon! I believe, this line is trying to load [this file](https://github.com/izmailovpavel/spurious_feature_learning/tree/main/data). When you are running the script, are you running it from the `group_dro` folder? With `python3 run_expt.py ...`?...

Hi, currently we don't have any keras implementation or examples. We will try to provide an example some time soon, probably after may 18. If you come up with an...

@simon-larsson that's great :) Do the results you are getting look similar to what we get in PyTorch?

@simon-larsson great! Sure, feel free to email me at [email protected].