Qualcomm AI Hub Models
A collection of machine learning models optimized and ready to deploy on Qualcomm devices, with a CLI to browse, compile, quantize, profile and run them on real hardware.
Qualcomm AI Hub Models is a collection of state-of-the-art machine learning models optimized for deployment on Qualcomm devices, distributed as a Python package (qai_hub_models) with a command-line interface for browsing and downloading the collection, plus a Python API. Model READMEs list any extra dependencies a specific model needs and cover supported on-device runtimes, hardware targets and precisions, chipsets, and devices.
Using Qualcomm AI Hub Workbench, which requires a free Qualcomm account and an API token, a model can be exported end to end: compiled for a chosen device and target runtime, optionally quantized, profiled on a real device in the cloud, run with sample input so its on-device output can be compared with the original PyTorch output, and downloaded as a compiled model. Most models also ship a CLI demo that runs the full pipeline end to end, for example predicting and drawing bounding boxes for an object-detection model.
It is released under the BSD-3-Clause license.