mini-batch size for RetinaNet.

[nonstop1962]

Hi, I want to train the retinaNet with retinanet_R-50-FPN_1x.yaml configuration setting.

I want to ask where can I set the mini-batch size for training.
I think there is no information of the batch size in .yaml file, and it is hard to find that information on the train_net.py either.

[rbgirshick]

For questions like this, please consult detectron/core/config.py (e.g., https://github.com/facebookresearch/Detectron/blob/master/detectron/core/config.py#L93).

[austinmw]

@rbgirshick I have the same question and this link is broken

[ir413]

Thanks for reporting @austinmw. The link is now fixed.