Instructions to use subhashhhhhh/lumen-npu with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- LiteRT
How to use subhashhhhhh/lumen-npu with LiteRT:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
Lumen NPU models
Models for Lumen, compiled ahead of time for Qualcomm Hexagon NPUs with
LiteRT's Qualcomm compiler (QAIRT). One folder per chip (SM7450/, SM8650/, โฆ);
each phone downloads only its own chip's files.
| File | Model | License |
|---|---|---|
siglip_visual.tflite |
Image encoder of SigLIP 2 B/16 | Apache 2.0 |
ocr_det.tflite |
Text detector of PP-OCRv6 small | Apache 2.0 |
They run only with LiteRT's Qualcomm dispatch library and Qualcomm's runtime on a phone with that chip. Built by a GitHub Actions workflow; the source models are the same ones Lumen uses on CPU and GPU.
- Downloads last month
- 644
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐ Ask for provider support