I was excited to see the version 2.7 was out but.... As soon as I start to run my model: It prints: Metal device set to: AMD Radeon Pro 5600M 2022-01-13 17:02:36.447465: I tensorflow/core/platform/cpu_feature_guard.cc:151] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations: AVX2 FMA To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags. 2022-01-13 17:02:36.448221: 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. 2022-01-13 17:02:36.448581: 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>) ...and the kernel dies Also: print("Num GPUs Available: ", len(tf.config.experimental.list_physical_devices('GPU'))) Num GPUs Available: 1 Things are in place - why doesn't this work yet??? MacOS 12.1 MacBook Pro (16-inch, 2019) 8-Core Intel Core i9 64 GB

I'm running this on Intell MBPro Metal device set to: AMD Radeon Pro 5600M 2022-03-29 11:46:55.397695: I tensorflow/core/platform/cpu_feature_guard.cc:151] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations: SSE4.2 AVX AVX2 FMA To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags. 2022-03-29 11:46:55.398290: 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. 2022-03-29 11:46:55.398545: 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: ) My model gets created and as soon as it starts the training on the first epoch the kernel dies. Epoch 1/3000 2022-03-29 11:47:10.011230: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled. I would give almost anything to get this working. Tensorflow-metal has never worked for me other than saying I have a gpu.

It's been years, and I keep trying, keep trying different hacks to get things installed. Things do install but nothing ever runs to completion. I would expect Apple would want to play a bigger part in this rather than having us have to move to linux with Nvidia. I wish Apple would just put some resources behind this.