Hi, thanks for sharing this codebase! I notice that there might be a small mistake in the implementation of Glister, in line 131 of the file deepcore/methods/glister.py,
flatten(), upadate_state=self.update_val_gradients
I believe this updadate_state is not right, which should be update_state, and I would like to ask what such self.update_val_gradients function works for? I am confused about the usage and effect of such function
Hi, thanks for sharing this codebase! I notice that there might be a small mistake in the implementation of Glister, in line 131 of the file deepcore/methods/glister.py,
flatten(), upadate_state=self.update_val_gradientsI believe this updadate_state is not right, which should be update_state, and I would like to ask what such self.update_val_gradients function works for? I am confused about the usage and effect of such function