Fork自 PaddlePaddle / Paddle
* add soft_label and axis for CrossEntropyLoss and improve performance,test=develop * fix conflict in nn/functional/loss.py, test=develop
拖放文件到此处或点击上传