Apple Releases MLX Framework for Machine Learning on Apple Silicon

0
Apple Releases MLX Framework
Apple Releases MLX Framework

The news that Apple Released MLX Framework for Machine Learning on Apple Silicon has been proven authentic. ML Explore or MLX is the name of Apple’s new machine learning (ML) framework for Apple Silicon computers.

This new framework is intended to facilitate training and execution of machine learning models on Macs with M series chips running M1, M2 and M3.

In this article we will tell you about Apple Releases MLX Framework for Machine Learning on Apple Silicon. We will also tell you about what exactly this framework is and how it’s helpful for users.

What is this MLX?

According to the company, MLX has a unified memory model. Apart from that, Apple showed how this open-source framework can be utilized which allows anyone interested in machine learning to run it on a laptop or desktop.

Apple has disclosed details about the MLX framework on GitHub, the code hosting platform. The data makes it known that the Python API of MLX framework closely mimics the NumPy C++ API, which is a scientific computing library for Python.

How Is This Framework Helpful To Users?

Apple stated that end users may also leverage higher level packages for building and running more complex models on their machines. Before MLX, developers had to rely on an interpreter to convert and optimize their models (via CoreML).

This simplified training and running ML models on computers. However, this has been replaced by MLX which enables users to train and perform their own models directly onto computers made from Apple Silicon.

The Framework Of MLX

Apple says that the design of MLX is based on other widely used frameworks such as PyTorch, ArrayFire, Jax, and NumPy. For a long time Apple has stressed the unified memory model in its framework where operations on MLX arrays are in shared memory and can be done on any type of device (currently Apple supports both CPU and GPU) with no data copies necessary.

Apple has also given examples of MLX in operation with Stable Diffusion image generation on Apple Silicon hardware. According to Apple, while generating batches of images with sizes 6, 8, 12 and 16, MLX may achieve up to 40% more throughput than PyTorch.

The tests were carried out on a Mac with the M2 Ultra chip which is the fastest processor ever made by this company. MLX which can generate all these images within 90 seconds will do that work in approximately 120 seconds if it were PyTorch.

Conclusion

In this article we discussed about Apple Released MLX Framework for Machine Learning on Apple Silicon. Other machine learning applications include text generation using Mistral large language model and Meta’s open source LLaMA language model.

OpenAI’s open-source Whisper tool can also allow AI/ML researchers to run the speech recognition models on their computer using MLX. Apple’s new MLX framework has the potential to simplify ML research and development on their hardware.

This advancement will empower developers to create tools, which can be used for services and apps that offer on device ML capabilities ensuring seamless operation, on users computers.

Also, Read WhatsApp Lets iPhone Users Share Music on Video Calls

LEAVE A REPLY

Please enter your comment!
Please enter your name here