sequence_generator forward decoder not model

Created by: jxmsML

❓ Questions and Help

This is probably very tiny, but why is this that SequenceGenerator._generate only call self.model.decoder(src_tokens, **kwargs) instead of self.model(src_tokens, **kwargs) https://github.com/facebookresearch/metaseq/blob/4629c56c467c1c40ef518a86f12799062e2551fa/metaseq/sequence_generator.py#L169-L172

The reason I asked is this is probably not friendly for arch where model is a wrapper of Decoder + something else.

Before asking:

search the issues.
search the docs.

What is your question?

Code

What have you tried?

What's your environment?

metaseq Version (e.g., 1.0 or master):
PyTorch Version (e.g., 1.0)
OS (e.g., Linux):
How you installed metaseq (pip, source):
Build command you used (if compiling from source):
Python version:
CUDA/cuDNN version:
GPU models and configuration:
Any other relevant information: