I find it’s quite inconvenient to know how training going, so do we have some interfaces like Keras?
E.g.
Epoch 1/15
22500/22500 [==============================] - 0s - loss: 0.2122 - acc: 0.9257 - val_loss: 0.0627 - val_acc: 0.9872
Epoch 2/15
22500/22500 [==============================] - 0s - loss: 0.0598 - acc: 0.9860 - val_loss: 0.0412 - val_acc: 0.9912
It will be helpful for us if we have this kind of interface.