Multi-GPUs training

[leepp15]

Hello Xdever! I expect to train the network by multi-gpus, but in the version, it seems that when I open several GPUs visible such as GPU0,1, the result is that the training is not accelerated while two GPUs are fully occupied.
How to achieve multi-gpus training function? Thanks.

[RobertCsordas]

The problem is that training with batch size of >1 is not supported. Therefore it can't be parallelized efficiently. Supporting batch of >1 is not trivial because of varying number of bounding boxes per image, and varying image size. Boxes could be represented by TensorArrays, or a padded tensor and a count vector. The end of the network must probably be replicated many times and the output of the neural part should be split between them among the batch dimension.