VENTUNO Q Instant Camera Models
Compiled local image-editing runtime for the VENTUNO Q Instant Camera.
This repository contains checksummed ARM64 Python runtime archives and Qualcomm QNN context binaries for the Arduino VENTUNO Q. The contexts are compiled for the target Qualcomm QCS8275 / HTP v75 profile and are not portable to unrelated hardware.
Install
Use the installer from the source repository:
./scripts/install-models.sh
The installer pins the immutable v1 tag, downloads one archive at a time, verifies SHA256SUMS, and extracts into ~/instant-camera-ai.
Contents
- Stable Diffusion 1.5 text encoder and VAE QNN contexts.
- InstructPix2Pix VAE encoder context.
- IP-Adapter Plus reference encoder and editor U-Net contexts.
- Stable Diffusion 1.5 CLIP tokenizer.
- ARM64 ONNX Runtime QNN Python environment.
- Upstream license texts and build provenance.
The complete experimental model tree is intentionally excluded. This repository contains only the components required by the supported standard local editor.
Licensing
The compiled contexts are modified forms of upstream model components:
- Stable Diffusion 1.5: CreativeML OpenRAIL-M.
- InstructPix2Pix: MIT, with Stable Diffusion restrictions applying to derived model weights.
- IP-Adapter: Apache-2.0.
- ONNX Runtime QNN and its bundled Python dependencies retain their included licenses.
The model-license archive is installed under ~/instant-camera-ai/model-licenses. Use of these artifacts is subject to CreativeML OpenRAIL-M, including its use restrictions. This repository is not affiliated with or endorsed by Arduino, Qualcomm, Stability AI, RunwayML, Timothy Brooks, or Tencent AI Lab.
Provenance
The QNN contexts were compiled through Qualcomm AI Hub for the Arduino VENTUNO Q target. See PROVENANCE.txt in the license archive for component details. Checksums cover the exact downloadable archives; immutable repository tags identify published bundle versions.