drop out, learning rate in nn

1. use different initial learning rates, says: 1e-3, 1e-4, 1e-5, if 1e-5 is the best one, that means your network is too complicate. you may want reduce to the layers.

• 点赞
• 收藏
• 分享
• • 文章举报

seamanj 发布了748 篇原创文章 · 获赞 191 · 访问量 101万+ 他的留言板 关注

标签：says,nn,748,drop,reduce,rate,1e,learning,文章
来源： https://blog.csdn.net/seamanj/article/details/103982094