search

CMU-Perceptual-Computing-Lab / caffe_rtpose

Realtime C++ code for multi-person pose estimation
Other
356 stars 207 forks source link

Error while running on a video #40

Closed tchinchuthomas closed 7 years ago

tchinchuthomas commented 7 years ago

Hi,

I am getting the following error while testing it with video. Please help.

F0427 11:36:23.180558 8473 syncedmem.cpp:56] Check failed: error == cudaSuccess (2 vs. 0) out of memory Check failure stack trace: @ 0x7f8f2bad15cd google::LogMessage::Fail() @ 0x7f8f2bad3433 google::LogMessage::SendToLog() @ 0x7f8f2bad115b google::LogMessage::Flush() @ 0x7f8f2bad3e1e google::LogMessageFatal::~LogMessageFatal() @ 0x7f8f2f88cf40 caffe::SyncedMemory::to_gpu() @ 0x7f8f2f88bf29 caffe::SyncedMemory::mutable_gpu_data() @ 0x7f8f2f88e382 caffe::Blob<>::mutable_gpu_data() @ 0x7f8f2f7a79b8 caffe::BaseConvolutionLayer<>::forward_gpu_gemm() @ 0x7f8f2f93f796 caffe::ConvolutionLayer<>::Forward_gpu() @ 0x7f8f2f8788e2 caffe::Net<>::ForwardFromTo() @ 0x40a81b warmup() @ 0x411204 processFrame() @ 0x7f8f2a9d26ba start_thread @ 0x7f8f2a70882d clone @ (nil) (unknown) Aborted (core dumped)

ZheC commented 7 years ago

As the error message states: error == cudaSuccess (2 vs. 0) out of memory. How many GBs GPU memory does your computer have?

tchinchuthomas commented 7 years ago

It has 2 GPUs (Nvidia GTX 1080 )each of 8GB GPU memory, 32 CPU core, 64 GB RAM. Unable to understand why is it still out of memory despite having good GPU resources. By any case, are these resources still insufficient to run the code?

tsimk commented 7 years ago

That's sufficient---we've used similar setups. Looks like you're not using cuDNN, you must use it to avoid this issue, e.g., https://github.com/CMU-Perceptual-Computing-Lab/caffe_rtpose/issues/36