Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

GPU not working in Tensorflow? #294

Closed
ckchow opened this issue Sep 23, 2016 · 2 comments
Closed

GPU not working in Tensorflow? #294

ckchow opened this issue Sep 23, 2016 · 2 comments
Labels
enhancement

Comments

@ckchow
Copy link

ckchow commented Sep 23, 2016

I'm using a prebuilt version of javacpp-presets's tensorflow bindings. (for SparkNet, specifically here https://github.com/amplab/SparkNet). However, whenever I create a graph with GPU nodes and load it with javacpp-tensorflow, the placement log says they are all placed on CPU. nvidia-smi does not report any GPU memory usage by the JVM either. Has anyone gotten GPU stuff to work on javacpp-tensorflow?
@saudet
Copy link
Member

saudet commented Sep 24, 2016

The binaries are not compiled with CUDA enabled. Rebuilding them with CUDA enabled will work, but the binaries might stop working on machines without GPU. That's something I'd like to check. Could you help out with that?

saudet added a commit that referenced this issue Oct 2, 2016
@saudet
* Enable CUDA in cppbuild.sh script for TensorFlow (issue #294)
cef8d6a
@saudet
Copy link
Member

saudet commented Dec 9, 2016

Version 1.3 now contains a build with GPU support enabled. Enjoy!

@saudet saudet closed this as completed Dec 9, 2016
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
enhancement
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@saudet @ckchow