I just run the train.py as your Readme, but this error occurs. It seems due to index of testing set, but I don't know how to solve it.
I change the --loadckpt to False while training becauese I want it train from the begining. Besides, while training at the Iter 210, the loss is Nan.
BTW, if I change the batch size from 1 to 2, this error disappers, but the result was terrible. I don't know why.
I just run the train.py as your Readme, but this error occurs. It seems due to index of testing set, but I don't know how to solve it. I change the --loadckpt to False while training becauese I want it train from the begining. Besides, while training at the Iter 210, the loss is Nan. BTW, if I change the batch size from 1 to 2, this error disappers, but the result was terrible. I don't know why.