Benjamin Bossan

Could you elaborate on that?

We started out with the premise that it is easy to make an error with caching enabled and not noticing it. If users have to use a context manager, they...

Good catch, we haven't tried out sparse matrices yet. Unfortunately, they won't work for numerous reasons. I tried to make it work with skorch, but then pytorch makes trouble (no...

I believe you are taking the wrong approach here. `CountVectorizer` is not a tokenizer, it creates row-wise word counts. `nn.Embedding` expects indices, though. Here is a short snippet that creates...

hi @akzaidi , we just merged a [notebook](https://nbviewer.jupyter.org/github/dnouri/skorch/blob/master/examples/rnn_classifer/RNN_sentiment_classification.ipynb) that shows how to best use skorch and RNNs for a classification task. I hope this helps with your task.

Glad to hear that you make progress. > is there an easy way to implement multi-GPU hyperparameter search when using GridSearchCV or RandomizedSearchCV? This is not possible yet. We have...

@ottonemo do you have an opinion on that?

Does the [image segmentation example](https://github.com/dnouri/skorch/tree/master/examples/nuclei_image_segmentation) solve this issue?

Yes, we thought about that but haven't reached a conclusion yet. You could post the gist of your current solution, maybe we can find a way to simplify it. In...

In general, ottonemo's 2nd suggestion looks good enough to me for prototyping. For bringing your model to production, I see why you would want a more robust solution. > For...