[FEATURE REQUEST] Should learner.teach call model.fit at every query?

[aryamansriram]

I am working with the ActiveLearner and I want to visualize the loss,accuracy after every query. I was wondering if teach is calling the fit method at every query because if that's the case, tensorboard has issues visualizing the same since a new graph is created every time. I checked the code and indeed it is. Would having the uncertainty sampling a part of a callback in the model.fit() method be a better option? So that you only call fit once and have the sampling be done in the callback after every epoch? Apologies if this is a stupid doubt in any way

[cosmic-cortex]

Hi! I am not sure I understand the issue. Are you using Keras and its scikit-learn wrapper? I suspect that your problem can be solved with writing a custom adapter class, but I don't understand your usecase well enough.