Hi, Xingyi
I enlarged the training batch size in order to accelerate the training process. On the contrary, the validation accuracy was dropping. Thank you!!!
Hi,
I also observed that. It seems a problem of HourglassNet. Increasing the learning rate with the same ratio (https://arxiv.org/pdf/1706.02677.pdf) will help.
Hi, Xingyi I enlarged the training batch size in order to accelerate the training process. On the contrary, the validation accuracy was dropping. Thank you!!!