Open chungjin opened 8 years ago
I don't think this is possible, since it used python layer, and caffe does not support multi-gpu if python layer is used
I think so, so they must have used the C++ and Cuda to rewrite the python layer
https://github.com/bharatsingh430/py-R-FCN-multiGPU
You can use this branch
I wander whether your project can be trained by multiple GPUs? I have tried to use the command like --gpu 0,1,2 but it seems it does not support this way. But in Dai's paper, they use 8 GPUs to train the model.