PrismML brings its tiny LLMs to Qualcomm-powered smart glasses
PrismML showcased a 2-billion-parameter 1-bit Bonsai model that runs locally on Qualcomm Snapdragon smart glasses.
At Qualcomm's Snapdragon Summit, Qualcomm showcased PrismML's 1-bit Bonsai LLM running locally on smart glasses built on the Snapdragon AR1 Gen 1 platform. The glasses version is a 2-billion-parameter vision-language model so a wearer can ask about what they see. PrismML, founded by Caltech researchers and advised by Ion Stoica, says its method shrinks larger models by about 4x while retaining nearly all standard-benchmark performance. No smart glasses product using PrismML has been announced.
- PrismML showed 1-bit Bonsai on Snapdragon AR1 Gen 1 smart glasses.
- The glasses version is a 2-billion-parameter vision-language model run locally.
- PrismML says compression shrinks models about 4x with most benchmark performance retained.
- No smart glasses product shipping PrismML has been announced yet.
- Vendors
- Qualcomm
- Products
- Snapdragon AR1 Gen 1
- AI models
- Bonsai
From the original article
1 min read at techcrunch.com
Prism's larger goal is open-weight AI that runs on devices and makes better use of the computing power they already have.
By Julie Bort · techcrunch.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.