sweetdream33

Thank you for your answer. I'm running the cifar10_macro_final.sh. tail of cifar10_macro_final.sh(in the stdout file) looks like : `epoch=306 ch_step=153050 loss=0.001949 lr=0.0011 |g|=0.2311 tr_acc=100/100 mins=2119.42` `epoch=306 ch_step=153100 loss=0.001135 lr=0.0011 |g|=0.0480...

Because takes a long time, I set to batch_size = 20 and ran it again After 15 epochs, loss does not fall. It takes 16 hours. Should I try to...

I didn't share my code somewhere. I just modified the part of source code that resizes the image. ``` models.py : def _pre_process(x) micro_child.py : self.pool_layers = [2,4,6,8] def _pre_process(x):...

Hello. Have you resolved the issue? I am also experiencing the same problem.