didiwai

iterate_mini_batches

1

I use your code, but when I run do_train func, there is a bug ValueError: Cannot feed value of shape (10, 2048) for Tensor u'pool_3/_reshape:0', which has shape '(1, 2048)'