未验证 提交 ba053d9e 编写于 作者: S SunAhong1993 提交者: GitHub

Update SofmaxWithLoss.md

上级 62cecbc8
......@@ -83,7 +83,7 @@ layer {
# label输入shape:(100,1)
# 输出shape:(10,1)
softmaxwithloss = fluid.layers.softmax_with_cross_entropy(logits=logs, label=labels,
soft_label=True, ignore_index=-100,
numeric_stable_mode=False,
soft_label=False, ignore_index=-100,
numeric_stable_mode=True,
return_softmax=False)
```
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册