I’ve been trying to run DeepLabCut on a laptop using the GPU (it is a Quadro P2000 4GB device). The train_network function runs for me for a few (~200-300) iterations, then bails with an error indicating a memory allocation problem. Error message looks like this:
ResourceExhaustedError: OOM when allocating tensor with shape[1,2048,64,114] and type float on /job:localhost/replica:0/task:0/device:GPU:0 by allocator GPU_0_bfc
Any advice on a workaround? Does this machine just simply have insufficient video card memory?