tensorflow_macos

Incorrect Pytorch gradients with Apple MPS backend... Yep this kind of thing can happen. I found and reported incorrect gradients for Apple's Metal-backed tensorflow conv2d in 2021 [1]. (Pretty sure I've seen incorrect gradients with another Pytorch backend, but that was a few years ago and I don't seem to have raised an issue to refer to... ) One might think this class of errors would be caught by a test suite. Autodiff can be tested quite comprehensively against numerical differentiation [2]. (Although this example is from a much simpler lib than Pytorch, so I could be missing something.) [1] https://github.com/apple/tensorflow_macos/issues/230 [2] https://github.com/sradc/SmallPebble/blob/2cd915c4ba72bf2d92...

Tensorflow has a CoreML enabled version which run on ANE. https://github.com/apple/tensorflow_macos

Someone has benchmarked Tensorflow using this code: https://github.com/apple/tensorflow_macos/issues/25 https://forums.macrumors.com/threads/apple-silicon-deep-lear... It is expected that the M1 Max should have similar performance to a RTX-2080 or Titan X for that task at least.

I was able to install this fairly easily (much more so then the crap they dumped out here - https://github.com/apple/tensorflow_macos. Just take a look at the 200 github issues that were ignored for the most part...) I also noticed that in my project I got a decent speedup immediately when executing my model, but I have not run any benchmarks. But, where do you go to file bugs? Ask questions? etc. I am not a big Mac developer, so is there something I don't know?

You might be interested in this for your M1 MBA: https://github.com/apple/tensorflow_macos

Just tried out apples native tensorflow for M1 branch yesterday: https://github.com/apple/tensorflow_macos installation was a non-issue, The training kept the Mac mini completely silent, having 40% of colab gpu speed is very satisfying for small tests. pytorch on the other hand is not ready yet (there are installation instructions but they didn't work)

Tensorflow is natively accelerated on M1. The branch is different - https://github.com/apple/tensorflow_macos Not sure what the confusion was.

There are some very preliminary benchmarks on a very small model here: https://github.com/apple/tensorflow_macos/issues/10 It doesn't really prove much beyond that you can probably get enough speed on an M1 to debug your training loop. That's impressive but we need to see more to see if the "only 4x slower than a colab GPU (ie at least a K80)" numbers hold up.

I can't answer your question, but the source code is here for those interested: https://github.com/apple/tensorflow_macos/archive/v0.1alpha0...

From https://github.com/apple/tensorflow_macos/blob/master/script... INSTALLER_PATH=https://github.com/apple/tensorflow_macos/releases/download/v0.1alpha0/tensorflow_… So, see: https://github.com/apple/tensorflow_macos/releases/

Sounds like even though it runs on all Macs with macOS 11.x, AMD GPU on Intel Macs are not optimized: https://github.com/apple/tensorflow_macos/issues/7

From: https://github.com/apple/tensorflow_macos#device-selection-o... "There is an optional mlcompute.set_mlc_device(device_name=’any') API for ML Compute device selection. The default value for device_name is 'any’, which means ML Compute will select the best available device on your system, including multiple GPUs on multi-GPU configurations." Hope this is not a stupid question: Does this TF version just work with M1 GPU-Cores / Intels iGPU or also with AMD and supported Nvidia GPUs?