Pre-trained model: is Input Shape parameter useful?

Pre-trained mode layer, the description of the “Input Shape” parameter says :slight_smile:
Optional shape tuple, only to be specified if include_top is false (otherwise the input shape has to be (224, 224, 3) (with tf dim ordering) or (3, 224, 244) (with th dim ordering). it should have exactly 3 inputs channels, and width and height should be no smaller than 197. e.g. (200, 200, 3) would be one valid value.

When include_top is false, not specifying the input shape does not stop the Model to be valid and it can be trained. It seem like DLS automatically fills that parameter.

You are right. DLS has knowledge of the input data so it automatically fills it instead of requiring user to fill. In some cases it may not be trivial for user to know the shape as well.

Since we directly show help text from Keras as it is, that is why the help text is that way even though it is not really required.

OK, I understand. Isn’t there a way to hide the parameter in the display ? since it is never useful.