Skip to content
This repository has been archived by the owner on Jan 3, 2023. It is now read-only.

Question: How I can use TF + nGraph + PlaidML ? #534

Open
jaebaek opened this issue Apr 25, 2019 · 3 comments
Open

Question: How I can use TF + nGraph + PlaidML ? #534

jaebaek opened this issue Apr 25, 2019 · 3 comments

Comments

@jaebaek
Copy link

jaebaek commented Apr 25, 2019

I installed TF and nGraph using
python3 build_ngtf.py --build_plaidml_backend --use_prebuilt_tensorflow

I tested TF mnist hello world.
It uses only CPU. How can I use PlaidML + OpenCL for its backend?

@cepera
Copy link

cepera commented Apr 29, 2019

hi, you need to set NGRAPH_TF_BACKEND env variable to PLAIDML.
export NGRAPH_TF_BACKEND="PLAIDML"

@SleepProgger
Copy link

SleepProgger commented May 27, 2019

@cepera Did you get it to run with plaidml ?
When i try i get an
tensorflow.python.framework.errors_impl.InternalError: Caught exception while compiling op_backend: get_shape() must be called on a node with exactly one output ()
error or a segfault (random)

@cepera
Copy link

cepera commented May 28, 2019

hi @SleepProgger, well, I did run it with plaidml like... in March, it worked fine for me. I would recommend to create a new issue here with some steps to reproduce

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants