-
Notifications
You must be signed in to change notification settings - Fork 707
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
How can I reproduce the results on Caffe? #2
Comments
I suggest not training this model from scratch using caffe, since caffe use |
Thanks for your advice! |
Hi @shicai If I would like to fine tune the pretrained model, What number would you suggest for Convolution, BatchNorm and Scale layer? According to your above suggestion, I guess that would be For BatchNorm, would be shown as below Scale layer would be Please help me to check param { lr_mult and decay_mult }. Thanks :) |
If you use the pretrained weights for detection, I sugguest you fixing all the BN parameters by setting lr_mult = 0 and decay_mult = 0. |
Thanks for your suggestion. I will finetune Convolution layer only and fix all the BN parameters 👍 |
@ryusaeba btw, to fix all the BN parameters, you should also set |
wow, that is a great helpful reminder. many many thanks :) |
@shicai |
It's ok to fine tune conv layers when fixing bn parameters, since bn mean/var parameters are not stable during detection training stage. |
@shicai |
yes. |
@shicai |
@shicai |
I think it is mainly because batch size for training detection models is very small. |
Could you share me a rough value about batch size? what number is belonging to small or large? |
if you want stable bn training, you'd better set batch size to 16 or even larger. but for detection tasks, batch size is always set to 1 or 2, due to memory reasons. |
Hi, thanks for sharing this MobileNets!
I am just wandering if I can reproduce same results on Caffe from scatch. Is it possible if you can sharing the solver.prototxt you used to achieve this accuracy rate? And how many days did it take to complete training?
Thanks~
The text was updated successfully, but these errors were encountered: