0
$\begingroup$

I am unable to identify this Warning Below. I am using M1 MacBook Air

CODE

from keras import models from keras.layers import Dense model=models.Sequential() model.add(Dense(16,activation='relu',input_shape=(10000,))) model.add(Dense(16,activation='relu')) model.add(Dense(16,activation='sigmoid'))

OUTPUT

Metal device set to: Apple M1 2021-09-24 18:18:51.205006: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:305] Could not identify NUMA node of platform GPU ID 0, defaulting to 0. Your kernel may not have been built with NUMA support. 2021-09-24 18:18:51.205728: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:271] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 0 MB memory) -> physical PluggableDevice (device: 0, name: METAL, pci bus id: <undefined>)

Does it mean I am not using GPU of m1 ?

Improve this question
$\endgroup$

2 Answers 2

Reset to default
2
$\begingroup$

Apple silicon is UMA, not NUMA. It's a unified memory architecture. So ignore this message. TF will work despite this annoying informational message.

Improve this answer
$\endgroup$
4
  • $\begingroup$ Can you please explain in a bit more detail what is the reason behind this? $\endgroup$ Commented Jul 4, 2022 at 18:09
  • $\begingroup$ Apple silicon uses unified memory. Memory is shared by the different subsystems of the system on a chip. NUMA is non-uniform memory architecture. That's when the CPU has RAM, and the GPU has VRAM, for example. $\endgroup$ Commented Jul 5, 2022 at 0:09
  • $\begingroup$ Thanks Michael, this is very clear. Could be worth adding to original response even. Thanks! $\endgroup$ Commented Jul 5, 2022 at 8:24
  • $\begingroup$ That should be non-unified memory architecture for NUMA. I think spell check messed it up. $\endgroup$ Commented Jul 15, 2022 at 5:00
1
$\begingroup$

You are running on GPU but you may find issues if you try and use the Adam optimiser. If you do try SGD instead.

You can add %GPU and GPU Time to the Activity Monitor if you want to check you are using GPU when training.

Improve this answer
$\endgroup$

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.