Unable to Use M1 Mac Pro Max GPU for TensorFlow Model Training

Hi, Just starting at Dev/ML, so I could be wrong, and I do not know how it works through Conda, but I tensorflow-metal, version 1.1.0/Sept2023 does not install with python 3.12. At the document/link that you points even specify explicitly python 3.9

I managed install it with python 3.10/3.11 and tensorflow 2.17.x

Regards

Hi, I had a similar problem in M1Pro(14.7) to do ML/AI, after checking different combinations of python, tensorflow, keras environments, these steps allowed me to operate with more current versions TF2.17-2.18,keras 3.8 & python3.12

conda create -n <Environment Name> python 3.11.11 or 3.12.9 
conda activate <Environment name> 
pip install tensorflow==2.17 or 2.18 tensorflow-metal 
conda install <other package name>

Automatic installation of tensorflow-metal 1.2 to check installation status:

python --version
python -c "import tensorflow as tf; import keras ; print(tf.__version__); print(keras.__version__)”
 python -c "import tensorflow as tf; print(tf.config.list_physical_devices('CPU'))”
 python -c "import tensorflow as tf; print(tf.config.list_physical_devices(‘GPU'))"

In the tensorflow documentation from version 2.16 onwards it is recommended to use "pip install tensorflow" and not other methods like conda or poetry. I have tested it for CIfar100 and MNIST and the code in keras2 and keras3 works on the local machine.

You should see a similar result:

I hope you find it useful and successful!!