The difference between experiments section of the paper and options

[nitec427]

Hi everyone,
Thanks for the great work. In your paper, you have mentioned that both stages are trained for 300 epochs. I aim to use your paper as a baseline for my study but in the configs, whole network and encoder is trained for 600 epoch. Could you explain the difference.
Wish the best.

[order-a-lemonade]

same question, Which version would you like to use? Have you reproduce the result？