


请问这种情况还是训练次数还不够吗?但我已经训练很久了,但是gpu不行,所以我batch-size为1进行训练,我已经训练600000了?
Could it be that the number of training iterations is still insufficient in this case? But I've been training for a long time. However, since the GPU isn't powerful enough, I set the batch - size to 1 for training. I've already carried out 600,000 training iterations.