Replace SequentialLayer with Sequential Model

The SequentialLayer was introducing more problems than the ones it solved.

Unfortunately, the old checkpoints will not be compatible but they can be easily converted using:

old_model = tf.keras.models.load_model(".../old_checkpoint")
new_model = bob.learn.tensorflow.models.incepetion_resenet_v2.IncepetionResnetV2()
new_model.set_weights(old_model.get_weights)
new_model.save(".../new_checkpoint")

added 1 commit

• a61605d2 - Replace SequentialLayer with Sequential Model

Compare with previous version

Thanks @amohammadi

This week I'm busy.

Reviewing by the end of the week.

Hey @amohammadi, have you checked if our current v1 checkpoints works with this update?

Thanks

If you call old_model = tf.keras.models.load_model(".../old_checkpoint"), then nothing needs to be changed. But if you do:

model = bob.learn.tensorflow.models.incepetion_resenet_v2.IncepetionResnetV2()
model.load_weights(".../old_checkpoint").assert_consumed()

it will fail.

merged

mentioned in commit 7a498fd9

mentioned in issue bob#269 (closed)