search

TheFoundryVisionmongers / nuke-ML-server

A Nuke client plug-in which connects to a Python server to allow Machine Learning inference in Nuke.
Apache License 2.0
134 stars 36 forks source link

Docker image crashes without AVX instructions #10

Closed samhodge closed 5 years ago

samhodge commented 5 years ago 
[kognat@vxfhost Server]$ sudo docker run --runtime=nvidia -v /home/kognat/dev/nuke-ML-server/Models:/workspace/ml-server/models:ro -it nuke-ml-magic:latest
[sudo] password for kognat: 
root@a1b2b7bf4646:/workspace/ml-server# python
Python 2.7.16 |Anaconda, Inc.| (default, Mar 14 2019, 21:00:58) 
[GCC 7.3.0] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow
2019-05-25 07:49:46.583944: F tensorflow/core/platform/cpu_feature_guard.cc:37] The TensorFlow library was compiled to use AVX instructions, but these aren't available on your machine.
Aborted (core dumped)

Party over dude.

samhodge commented 5 years ago

Workaround is here: https://github.com/yaroslavvb/tensorflow-community-wheels/issues/69#issuecomment-478763464

samhodge commented 5 years ago

doesnt effect the Caffe2 models

ringdk commented 5 years ago

Thanks for heads-up! We haven't seen that issue before.

At the moment we're not planning on rebuilding and supporting our own TF libraries. We'll add it to our list of known issues. Thanks again!

samhodge commented 5 years ago

Lack of AVX instructions is a curiosity relevant only to ancient hardware, like mine.

ringdk commented 5 years ago

Lol! It's not as uncommon as you think, we've hit similar issues before with Blink.