v0.32.2

✨ What's New

• New Apps:

  • GenieX Chat Android: A sample Android chat app that runs Large Language Models and Vision-Language Models on-device with the Qualcomm® GenieX SDK, using a single pluggable runtime that targets the Snapdragon® NPU, GPU, or CPU.
  • Posenet Pose Estimation: A sample app for Ubuntu integrating the Posenet MobileNet pose-estimation model via TFLite Runtime. It uses GStreamer for input video and provides a minimal web interface to stream output frames suitable for embedded applications.

• Extended CLI support for Windows C++ and Python Apps

• Moving forward, Nightly builds will be available at http://qaihub-public-python-wheels.s3-website-us-west-2.amazonaws.com/qai-hub-apps

• CLI Updates:

  • Users can now bundle locally exported models directly with an app during fetch
    qai-hub-apps fetch --model now accepts either a model ID or a local model path. Added --model-id and --model-path flags for explicit model source selection

⚠️ Deprecations

• LLM Chat Android has been deprecated in favor of GenieX Chat Android
• LLM Chat Windows has been deprecated in favor of GenieX Chat Windows

🛠️ Get started

pip install qai-hub-apps

qai-hub-apps list
qai-hub-apps fetch geniex_chat_android

PyPi: https://pypi.org/project/qai-hub-apps/
GitHub: https://github.com/qualcomm/ai-hub-apps

🗒️ Additional Notes:

• New Tutorial: https://github.com/qualcomm/ai-hub-apps/tree/main/tutorials/geniex
• Updated qai-hub-models-cli to 0.57.0