Training Settings

[lxtGH]

@sud0301 Hi! Thanks for releasing your code. It seems your training setting is a little different from baseline. They use 20k iteration while you use 40k.

[sud0301]

The performance of the baseline method deteriorates if trained longer. Secondly, the baseline method uses a batch size of 10, whereas we use a batch size of 8.

[mra1a]

@sud0301 Hi！why the baseline method uses a batch size of 10, and you use a batchsize of 8