Enntity / sparkscore
Typed, calibrated decisions on one NVIDIA DGX Spark: d3-ultra 27B NVFP4 on Atlas, JevBench public 90.9% at p50 146 ms, every question in one engine pass.
Why this job: works in the areas we cover, but does not say which job it does
What it does
SparkScore serves d3-ultra (27B, by the vLLM Semantic Router team) quantized to NVFP4, on the Atlas inference engine. The API has the shape of TypeSafe's Jev decision API, and it accepts the model name jev-latest, so a Jev client needs only a new base URL. It runs on your own hardware, so the data you score stays on the box.
Get started
git clone https://github.com/Enntity/sparkscore.git
cd sparkscore
./start.sh - Maintenance
- Active
- Last commit 3 days ago
- Stars
- 10
- +10 this week
- Licence
- AGPL-3.0
- Share-alike, even hosted
- Downloads for
- None listed
- No release files
- Latest version
- None yet
- No releases
- Built with
- Python
- 94% of the code
Is it looked after?
- Contributors
- 1
- Last 90 days
- Issues
- 0% closed
- 4 still open
- Pull requests
- 2 merged
- 0 waiting
- Releases
- None
- Never published
- Contributing guide (opens in new tab)
- Security policy (opens in new tab)
- No code of conduct
Similar repos
| Repo | Stars | Last commit | Licence | Downloads for |
|---|---|---|---|---|
| nokia-applied-research / AnyJev | 1,606 | 23 days ago | Apache-2.0 | No downloads |
| Shanghua-Gao / RSI-Jev | 137 | 3 days ago | MIT | Mac |
| malevrigns / agent-jev | 398 | 14 days ago | Apache-2.0 | No downloads |
More: languages, topics and project record
Languages
- Python 94%
- Shell 5%
- Dockerfile 1%
Topics
- calibration
- decision-model
- dgx-spark
- gb10
- jev
- llm-inference
- nvfp4
Project record
- JobOther AI tools
- StatusNew to our list
- Commits15
- Forks0
- Watching0
- CreatedNov 6, 2026
- First seen by usNov 7, 2026
- CheckedNov 13, 2026