jevIMAGE
DiffusionGemma on DGX Spark delivers Jev-like structured output at 0.20s
The author ran a Jev-like structured output setup with Google DeepMind's DiffusionGemma on a DGX Spark, getting 0.20s median results untuned, with a vLLM branch tested on real hardware.
Matt Mastracci
@mmastracArchived 2026.09.18Originally posted 2026.09.17

Matt Mastracci @mmastracOriginally posted 2026.09.17
Now you can run a Jev-like on your own hardware using DiffusionGemma from @GoogleDeepMind. Untuned, on a DGX Spark, 0.20s median structured results! Branch is in progress, but tested and working on real hardware. https://github.com/mmastrac/vllm/tree/structured-reads
x.com/mmastrac/status/2100349903507689744Archived 2026.09.18 10:16