deepID deepid_class

原创 2015年07月08日 09:42:43
Training the model ...
    epoch 0, minibatch_index 199/200, error 0.980000
epoch 0, train_score 0.981920, valid_score 0.985120
    epoch 1, minibatch_index 199/200, error 0.970000
epoch 1, train_score 0.966220, valid_score 0.978880
    epoch 2, minibatch_index 199/200, error 0.936000
epoch 2, train_score 0.939180, valid_score 0.962000
    epoch 3, minibatch_index 199/200, error 0.902000
epoch 3, train_score 0.909490, valid_score 0.951280
    epoch 4, minibatch_index 199/200, error 0.860000
epoch 4, train_score 0.880050, valid_score 0.936960
    epoch 5, minibatch_index 199/200, error 0.816000
epoch 5, train_score 0.850390, valid_score 0.923360
    epoch 6, minibatch_index 199/200, error 0.842000
epoch 6, train_score 0.821610, valid_score 0.921840
    epoch 7, minibatch_index 199/200, error 0.818000
epoch 7, train_score 0.794550, valid_score 0.909680
    epoch 8, minibatch_index 199/200, error 0.760000
epoch 8, train_score 0.767890, valid_score 0.905840
    epoch 9, minibatch_index 199/200, error 0.754000
epoch 9, train_score 0.746010, valid_score 0.886640
    epoch 10, minibatch_index 199/200, error 0.712000
epoch 10, train_score 0.725360, valid_score 0.873920
    epoch 11, minibatch_index 199/200, error 0.708000
epoch 11, train_score 0.706780, valid_score 0.872000
    epoch 12, minibatch_index 199/200, error 0.684000
epoch 12, train_score 0.685420, valid_score 0.868400
    epoch 13, minibatch_index 199/200, error 0.706000
epoch 13, train_score 0.670460, valid_score 0.871440
    epoch 14, minibatch_index 199/200, error 0.658000
epoch 14, train_score 0.657160, valid_score 0.847920
    epoch 15, minibatch_index 199/200, error 0.642000
epoch 15, train_score 0.639900, valid_score 0.844800
    epoch 16, minibatch_index 199/200, error 0.640000
epoch 16, train_score 0.626590, valid_score 0.844640
    epoch 17, minibatch_index 199/200, error 0.596000
epoch 17, train_score 0.612180, valid_score 0.830720
    epoch 18, minibatch_index 199/200, error 0.638000
epoch 18, train_score 0.601290, valid_score 0.841360
    epoch 19, minibatch_index 199/200, error 0.592000
epoch 19, train_score 0.587440, valid_score 0.823840

0 0.98192 0.98512
1 0.96622 0.97888
2 0.93918 0.962
3 0.90949 0.95128
4 0.88005 0.93696
5 0.85039 0.92336
6 0.82161 0.92184
7 0.79455 0.90968
8 0.76789 0.90584
9 0.74601 0.88664
10 0.72536 0.87392
11 0.70678 0.872
12 0.68542 0.8684
13 0.67046 0.87144
14 0.65716 0.84792
15 0.6399 0.8448
16 0.62659 0.84464
17 0.61218 0.83072
18 0.60129 0.84136
19 0.58744 0.82384


版权声明:本文为博主原创文章,未经博主允许不得转载。

相关文章推荐

DeepID人脸识别算法之三代

 DeepID人脸识别算法之三代 转载请注明:http://blog.csdn.net/stdcoutzyx/article/details/42091205 DeepID,目前最强人脸识别...

deepID 深度人脸识别

  • 2016-07-08 10:47
  • 2.50MB
  • 下载

DeepID训练过程中loss居高不下解决方法

1.      DeepID架构参考: http://blog.csdn.net/a_1937/article/details/50334919 2.      问题一: 用...

专访DeepID发明者孙祎:关于深度学习与人脸算法的深层思考

摘要:DeepID人脸算法发明者、人脸识别创业公司Linkface研究员孙祎深度解读DeepID的研发心得,人脸识别应用的现状、难点与未来,深度学习的实践经验等问题。 人脸识别是近两年计算机...

DeepID算法实践

转载请注明:http://blog.csdn.net/stdcoutzyx/article/details/45570221 好久没有写博客了,I have failed my blog. 目前...

专访DeepID发明者孙祎:关于深度学习与人脸算法的深层思考

摘要:DeepID人脸算法发明者、人脸识别创业公司Linkface研究员孙祎深度解读DeepID的研发心得,人脸识别应用的现状、难点与未来,深度学习的实践经验等问题。 人脸识别是近两年计算机...

DeepID3 face recognition

DeepID3 face recognition刚看完DeepID3[1],总结一下,还是先简单介绍一下网络结构吧。Network ArchitectureDeepID3有两种不同的结构,分别为Dee...
内容举报
返回顶部
收藏助手
不良信息举报
您举报文章:深度学习:神经网络中的前向传播和反向传播算法推导
举报原因:
原因补充:

(最多只允许输入30个字)