Arm-Examples/ML-examples

Arm Machine Learning tutorials and examples

What it solves

This repository provides a collection of practical machine learning examples and tutorials designed to help developers optimize and deploy AI models on Arm-based hardware, specifically targeting Neural Processing Units (NPUs) and Cortex-based platforms.

How it works

The project organizes a series of specialized examples that demonstrate different aspects of the ML pipeline on Arm hardware:

  • Hardware Acceleration: Examples for the Ethos-U NPU on Corstone-300 and Corstone-1000 platforms.
  • Library Integration: Demonstrations of how to use the KleidiAI library.
  • Model Optimization: Tutorials on pruning and clustering to reduce model weight size for Ethos NPUs.
  • Framework Integration: Guides on training Conformer networks for speech recognition using PyTorch/ExecuTorch and exporting them to Ethos-U.
  • TensorFlow Lite Deployment: A Jupyter notebook demonstrating RNN unrolling for .tflite format export.

Who it’s for

Embedded systems developers and ML engineers who are targeting Arm Cortex and Ethos NPU hardware for edge AI deployment.

Highlights

  • Support for Arm Ethos-U NPUs and Corstone reference platforms.
  • Examples for model compression techniques like pruning and clustering.
  • Integration with PyTorch, ExecuTorch, and TensorFlow Lite.
  • Specific implementations for speech recognition using the Conformer network.

Related

  • Project
    qualcomm/ai-hub-apps

    Qualcomm AI Hub Apps provides sample Android, Windows, and Ubuntu applications plus tutorials that demonstrate how to run vision, audio, and language models on Snapdragon devices using TensorFlow Lite, ONNX, and Qualcomm’s Genie SDK, with optional CPU/GPU/NPU acceleration.

  • Project
    CerebriumAI/examples

    A collection of ready-to-use deployment examples for the Cerebrium platform, covering LLMs, voice agents, image generation, and inference optimization.

  • Project
    qualcomm/ai-hub-models

    Qualcomm AI Hub Models is a Python package and CLI that provides a catalog of pre‑optimised machine‑learning models (vision, language, generative AI) ready for deployment on Snapdragon CPUs, GPUs, and NPUs. It handles quantisation, compilation, and performance profiling on real Qualcomm devices via the AI Hub Workbench, and includes sample apps and demos for both local PyTorch inference and on‑device accelerated runtimes.

  • Project
    google-ai-edge/litert-samples

    Google AI Edge LiteRT Samples is a genuine open‑source collection of on‑device machine‑learning demos, model‑conversion recipes, and automation tools for LiteRT (the successor to TensorFlow Lite) and LiteRT‑LM. It provides Android, iOS, and Web/WASM examples covering speech, vision, text‑to‑image, and multimodal apps, plus scripts to convert Hugging Face checkpoints into LiteRT‑compatible bundles. The repo is aimed at developers building high‑performance edge AI applications.

  • Project
    awslabs/awsome-distributed-ai

    A collection of reference architectures and examples for deploying and operating distributed AI training and inference on AWS services like SageMaker HyperPod and EKS.