Open hhertogh opened 8 years ago
Same problem here.
Same here.
exactly same error here
I found that when I use vgg16, it may run out of memory, so I use ./tools/demo.py --cpu --net=zf
, then problem solved.
GPU: Nvidia Gtx 1050 ti 4gb System Memory: 8gb
./tools/demo.py --cpu --net vgg16
System RAM: 5.5gb
./tools/demo.py --gpu 0 --net vgg16
Crashed (Out of gpu memory)
./tools/demo.py --cpu --net zf
System RAM: 1.4gb
./tools/demo.py --gpu 0 --net zf
System RAM: 0.8gb
Hello,
what is the minimum amount of memory required to execute the "demo" script ?
I've got the following error: Loaded network /home/henry/py-faster-rcnn/data/faster_rcnn_models/VGG16_faster_rcnn_final.caffemodel F1009 00:02:59.708518 11833 syncedmem.hpp:25] Check failed: ptr host allocation of size 172800000 failed ** Check failure stack trace: *** Aborted (core dumped)
I am trying to make it work on a droplet (digital ocean) so indeed the memory is pretty limited and only CPU mode is used. My goal is to create an API that could use the framework to detect objects...
Thanks for your help.