search

ctrahey / m1-tensorflow-config

Conda env files for Tensorflow on Apple M1 silicon
35 stars 9 forks source link

Works but throws error #3

Open RuizSerra opened 2 years ago

RuizSerra commented 2 years ago

Output: 

Init Plugin
Init Graph Optimizer
Init Kernel
Running TensorFlow 2.5.0 with 1 GPUs recognized
Metal device set to: Apple M1

systemMemory: 16.00 GB
maxCacheSize: 5.33 GB

2022-08-04 12:15:40.822101: 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.
2022-08-04 12:15:40.822188: 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>)
Traceback (most recent call last):
  File "confirm.py", line 8, in <module>
    d = tf.random_normal([])
AttributeError: module 'tensorflow' has no attribute 'random_normal'

Checked GPU usage with Activity Monitor (cmd+3, cmd+4):

image
GNU-Linuxer commented 2 years ago

Same issue here.

ctrahey commented 1 year ago

Apologies for the delayed response here - Moved to Europe in October :-)

Just got a new work-issued laptop (M1 Pro) so I hope to revisit this soon and I'll walk through all the issues at that time. Hopefully within a week or two.