Making GPU works

[enzyme69]

Previously, I have the blob thing issue, but resolved after changing that line.

But I cannot make the GPU working.

[py27] C:\Users\Jimmy\style-transfer-master>python style.py -s C:\Users\Jimmy\st
yle-transfer-master\images\style\starry_night.jpg -c C:\Users\Jimmy\style-transf
er-master\images\content\blackdog.jpg -m VGG16 -g 1
style.py:main:11:46:43.924 -- Starting style transfer.
WARNING: Logging before InitGoogleLogging() is written to STDERR
F0406 11:46:43.947576 25172 common.cpp:145] Check failed: error == cudaSuccess (
10 vs. 0) invalid device ordinal
*** Check failure stack trace: ***

[enzyme69]

I seems to have issue with memory.

[py27] C:\Users\Jimmy\style-transfer-master>python style.py -s C:\Users\Jimmy\s
yle-transfer-master\images\style\starry_night.jpg -c C:\Users\Jimmy\style-trans
er-master\images\content\nanjing.jpg -m VGG16 -g 0
style.py:main:13:19:38.180 -- Starting style transfer.
WARNING: Logging before InitGoogleLogging() is written to STDERR
I0406 13:19:38.391417 23468 common.cpp:23] System entropy source not available,
using fallback algorithm to generate seed instead.
style.py:main:13:19:38.392 -- Running net on GPU 0.
style.py:main:13:19:38.447 -- Successfully loaded images.
style.py:main:13:19:40.099 -- Successfully loaded model VGG16.
F0406 13:19:45.972851 23468 syncedmem.cpp:51] Check failed: error == cudaSucces
(2 vs. 0) out of memory
*** Check failure stack trace: ***

Is there option to make the image smaller?

[enzyme69]

Still running out of memory with small image...

[py27] C:_PYTOOLS\chainer-master>python
Python 2.7.11 |Anaconda 2.5.0 (64-bit)| (default, Feb 16 2016, 09:58:36) [MSC v.
1500 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
Anaconda is brought to you by Continuum Analytics.
Please check out: http://continuum.io/thanks and https://anaconda.org

import sys, platform
platform.architecture()
('64bit', 'WindowsPE')
sys.maxint
2147483647

[enzyme69]

Apparently it's the error of RAM out of memory. Somewhat I can only process image under 400 pixels. Weird. Because I tried "deepdream-windows-cuda_v75_nocudnn" and can process big image up to 1280 x 720.

Any idea how to increase the GPU memory? Or can we make it somewhat so CPU and GPU are used together?

[ttoinou]

I also have this "entropy source" warning I don't know what it means.
And yes there is option to make the image smaller you can use --length x

[ssdutHB]

That's it. The error information occurs when GPU runs out of memory. Trying to use image with low resolution can avoid this.