Tomasz Bartczak

Hello! Thanks for the thorough explanation. I would then call this metric 'WorstCaseExecutorMemUsage'. And this is actually what I am looking for. Although in my case the worst case didn't...

Given that we've implemented an `"auto"` strategy as default in https://github.com/keras-team/keras/pull/16664 - no updates to the docs are needed.

it works without escaping

test class shouldn't expect any kind of ordering - that's right - still I would like to be able to have the same behaviour i had for junit 4.10, as...

I'm not doing it at the moment. I just assumed that one can do sth like (psuedocode) `inputs, targets = zip(*ds.take(10))` and pass this to `Experiment`, but I don't see...

Hello. Sorry for the late reply! > * we need to check the interaction with the formats functionality (setting different formats for each `tick`). For example, if we have multiple...

And furthermore there is yet another architecture where GAP is followed by a dense layer before given to softmax layer: http://cnnlocalization.csail.mit.edu/ supporting this as well would be really cool. Actually...

From what I know about the internals of quiver - the activations at each layer are normalized to fill the whole space of color intensity, since they are first normalized...

I don't have access to Mac M1 to reproduce. Please provide full logs from the process, starting from the command being executed. We might debug it this way.

Ok, that's another issue. If you try and checkout the branch used in this PR https://github.com/allegro/allRank/pull/51 it should work - it's updated to support current versions of all libs.