Could you share the evaluation code on Visdial-Q dataset?

[wanyao1992]

Hi Zilong,
Thanks for your sharing of this copy of code for the task of visual dialog. In your paper, you have conducted experiments on the Visdial-Q dataset, and I am also plan to evaluate the question generation in my paper. Do you plan to share this part of code as well in this repos? Highly appreciated if you can release this part of code. Thanks.

[zilongzheng]

Hi Yao,
Sorry for late reply. I didn't share the part of Visdial-Q code to protect original author's (@unnat) originality. The part of VisDial-Q evaluation is a subtle modification of the VisDial task, while we change the input to be dialog history and the queried image and output to be the predicted question. Please feel free to email me (zilongzheng0318@ucla.edu) for details.