Trainable param for predefined layers

[iesahin]

TF: 1.13.0-dev20190204
Keras: 2.2.4
Keras-VGGFace 0.5

Keras has a trainable param for layers. I would like to set this to False when I load the predefined weights for fine tuning, so that only the top layers I define are trained. I probably can do this myself with a PR but is there another mechanism to achieve this?

[rcmalli]

This is mostly about customization. It can be added to the library if you would like to contribute.

[sherlockchou86]

no need to add to library, just set it if you want to fine-tuning by yourself.