Some questions about the code and paper.

[lucky-23]

Hi, thanks for the code. I have some questions regarding your paper and the code.

1. In Table 3 of the paper, the last two rows indicate the simon model with and without context embedding right? When I run the github code, I obtain similar results as the one without context embedding with the same hyper-parameter settings. In the train.py file, as per my understanding, the context embedding part seems to be missing. Can you please elaborate more about this ? And how do I obtain the same results as the paper with the context embeddings?