coremltools VS hummingbird

Instead of trying to integrate the whole stack of, say, pytorch, Apple's primary approach has been converting models to work with Apple's stack.

https://github.com/apple/coremltools

Clearly no one is going to be doing training or even fine tuning on Apple hardware at any scale (it competes at the low end, but at scale you invariably will be using nvidia hardware), but once you have a decent model it's a robust way of using it on Apple devices.

There is one guy who was able to run it on iOS. See this thread for more information. Basically, the idea is to convert torch models to CoreMl. Only the CLIP tokenizer's implementation is currently missing. I guess this guy will keep modifications private, but he is trying to optimize model for lower RAM requirements.

Afaik (correct me if I’m wrong) both PyTorch and tensorflow only use the gpu when training and not the neural engine. I think the neural engines can be used for inference if the model is in the CoreML format (https://github.com/apple/coremltools)

CoreML Tools

And Python.

There's literally an Apple provided tool, called [coremltools[(https://github.com/apple/coremltools) to convert many common PyTorch and TensorFlow models to CoreML.

If are interested in just predictions you can try Hummingbird. It is part of the PyTorch ecosystem. We get already trained scikit-learn models and translate them into PyTorch models. From them you can run your model on any hardware support by PyTorch, export it into TVM, ONNX, etc. Performance on hardware acceleration is quite good (orders of magnitude better than scikit-learn is some cases)

I think Rapids AI's cuML tried to go into this direction (essentially scikit-learn on the GPU): https://docs.rapids.ai/api/cuml/stable/api.html#logistic-reg.... For some reason it never took really off though.

Btw., going on a tangent, you might like Hummingbird (https://github.com/microsoft/hummingbird). It allows you trained scikit-learn tree-based models to PyTorch. I watched the SciPy talk last year, and it's a super smart & elegant idea.

ONNX opens an avenue for direct inference using a number of languages and platforms. For example, a model could be run directly on Android to limit data sent to a third party service. ONNX is an exciting development with a lot of promise. Microsoft has also released Hummingbird which enables exporting traditional models (sklearn, decision trees, logistical regression..) to ONNX.

And Python.

For inference, you can convert your models to other formats that support GPU acceleration. See Hummingbird https://github.com/microsoft/hummingbird

The surprising thing is that Hummingbird can be faster than the GPU implementation of LightGBM (and XGBoost) if you use tensor compilers such as TVM. [The paper](https://www.usenix.org/conference/osdi20/presentation/nakandala) describes our findings. We have also open sourced the [benchmark code](https://github.com/microsoft/hummingbird/tree/main/benchmarks) so you try yourself!

I took their sample code from Github and tweaked it to spit out times for each model's prediction, as well as increase the number of rows to 5 million. I used Google's Colab and selected GPU for my hardware accelerator. This gives an option to run code on GPU, not that all computations will happen on the GPU.