without GPU

[kingrain213]

No description provided.

[hx173149]

Hi If you want to training from the beginning, you could replace the mean file; If you want to fine-tuning from the sports-1M / UCF-101 model, you'd better not replace the mean file. I think you should write a simple code to calculate the mean file. 2017-10-24 10:08 GMT+08:00 kingrain213 <notifications@github.com>:

…

If I want to train a new model on another dateset, shall I have to calculate the crop_mean.npy, then replace your file, train and test? I saw the crop_mean.npy's size is (16, 112, 112, 3) and type is float, Is it just a sum and then divide the image query number? Do you have the python script or other function to calculate the mean value， can you share it to me ? my email is ***@***.***, thank you very much. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#32 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFD3BBynppTKv1v9E8H6FS_vS9vp3gcPks5svUaWgaJpZM4QCOlf> .

-- 侯鑫祝您： 工作顺利，身体健康，万事如意~~~