double-vane

@yjxiong flow modality not work now?

  @yjxiong Like #7 ,I don't know where I can download the correct flow caffemodel

@goodluckcwl Thank you very much. **Does the lambda remain the same?** layer { name: "fc6" type: "MarginInnerProduct" bottom: "fc5" bottom: "label" top: "fc6" top: "lambda" param { lr_mult: 1 decay_mult:...

@ShangxuanWu 你好，你按照hieudt-pixelz做的有效，是指tensorflow和caffe都可以正常运行吗？还是只是解决了caffe这个问题，tensorflow还能正常运行吗？谢谢！

> For large and difficult datasets, you should first try to set lambda_min as 5 or 10 Is this one? But, when I use sphereface-20, I can get a small...

Ok, How to determine lambda_min? train softmax_loss ? train accuracy? or test results?

Okay, I down it from 1000, and my boss wants me to drop it to 5. I'll try it for more numbers! Thanks again. Best!

额，我的意思是您这里的vgg16.ckpt的链接已经失效，所以我看不到里面网络的名称。 我说的网络代码是指构造vgg16的网络代码，是用slim构造的，在tensorflow里的/usr/local/lib/python2.7/dist-packages/tensorflow/contrib/slim/python/slim/nets/vgg.py构造的吗？ 您能帮我看一下，如果我的网络显示是这样的，那我写.prototxt时，第一层的conv的名称是随便取还是local/Conv?十分感谢 Converted 28 variables to const ops. 102 ops in the final graph. (u'Placeholder', (,)) (u'Placeholder_1', (,)) (u'Placeholder_2', (,)) (u'Placeholder_3', (,)) (u'local/Conv/weights/Initializer/random_uniform/shape', (,)) (u'local/Conv/weights/Initializer/random_uniform/min', (,)) (u'local/Conv/weights/Initializer/random_uniform/max', (,)) (u'local/Conv/weights/Initializer/random_uniform/RandomUniform',...