Transfer learning consists of taking features learned on one problem, and leveraging them on a new, similar problem. For instance, features from a model that has learned to identify racoons may be useful to kick-start a model meant to identify tanukis.
Layers & models have three weight attributes:
weightsis the list of all weights variables of the layer.
trainable_weightsis the list of those that are meant to be updated (via gradient descent) to minimize the loss during training.
non_trainable_weightsis the list of those that aren't meant to be trained. Typically they are updated by the model during the forward pass.
MORE LIKE THIS
❤️ Brainstash Inc.