search

yule-li / CosFace

Tensorflow implementation for paper CosFace: Large Margin Cosine Loss for Deep Face Recognition
285 stars 99 forks source link

Could you tell me your trained model is trained by which network structure? #3

Open lina0309 opened 6 years ago

lina0309 commented 6 years ago

I can test your upload model in the test.sh. but i retrain the model by train.sh, i found i could not test in my correct model path. So i want to know your model in which structures... --network_type sphere_network in the test.sh mean i should train the model by #NETWORK=sphere_network right? Thanks very much~

yule-li commented 6 years ago

In train.sh, the default network is sphere_network. So if you run train.sh with default hyper-parameters, you just need to change the MODEL_DIR to where your trained model is saved. Notice the full name of saved model file is /path/to/xxx.ckpt-yyy.data-00000-of-00001 and you should set MODEL_DIR=/path/to/xxx.ckpt-yyy.

lina0309 commented 5 years ago

Oh, thank you very much, i will have a try following your reply~

leacornelio commented 5 years ago

How do you download the model from Baidu? I have an account on Baidu but it's asking me to install a software.

DecentMakeover commented 5 years ago

@leacornelio Have you downloaded the model?

DecentMakeover commented 5 years ago

Btw, the software it is asking you to download is the baiduclient disk. Once you install it, you will be able to download the model.

If you have succeeded in downloading the model, could you please upload it to google drive/dropbox? .I am using a linux machine.Thanks