Bucketed model with shared weights

[ishalyminov]

Is there a way to share internal layers' weights of a Seq2Seq (preferably AttentionSeq2Seq) model in order to make a bucketed training of those, or it's supposed to be implemented by the user?:)

[ishalyminov]

@farizrahman4u I've just managed to:

1. create a model with explicitly specified input_length, output_length for one bucket
2. save its weights
3. create a model for another bucket (again, setting its input and output lengths explicitly)
4. loaded the 1st model's weights into it
5. fitted some data -- and nothing crashed:)

What is the meaning of input_length, output_length if they don't affect the topology? And is the way described above OK for training bucketed models?