Any way to see the learning rate schedule which will be used?

In the earlier version we could do learn.fit_opt_sched as described in https://github.com/sgugger/Deep-Learning/blob/master/Understanding%20the%20new%20fastai%20API%20for%20scheduling%20training.ipynb

Is there similar for this version?

I believe learn.recorder.plot_lr() would do the required, however, recorder is not created until train begin.

The same API is now usable via the callback GeneralScheduler.

I mean how do I plot the lr curve before starting the training?

There won’t be anything to plot until you’ve started training (even if it’s a mock training), I’m not sure I understand the question

Ah, I got it. I wanted to know how my learning rate changed, but that required me to train first. I was under the impression that we could have a visualisation for the learning rate schedule before training began.