raullenchai / twinspark
Production-grade recipe for DeepSeek-V4-Flash-0731 (284B MoE) on 2x NVIDIA DGX Spark: self-healing 2-node vLLM cluster, reboot-verified, tuned DSpark speculative decoding (~75 tok/s), full benchmarks, OpenAI Codex CLI integration
Why this job: topics: local-llm, llm-inference, vllm
What it does
Assumes both Sparks are through DGX OS first-boot, the QSFP DAC is plugged (before power-on!), and you can SSH between them.
Get started
No single install command. Read the steps in the README (opens in new tab)
- Maintenance
- Active
- Last commit 2 months ago
- Stars
- 176
- On GitHub
- Licence
- MIT
- Allows business use
- Downloads for
- None listed
- No release files
- Latest version
- None yet
- No releases
- Built with
- Shell
- 81% of the code
Is it looked after?
- Contributors
- 1
- Last 90 days
- Issues
- None
- No issues filed
- Pull requests
- 0 merged
- 0 waiting
- Releases
- None
- Never published
- No contributing guide
- No security policy
- No code of conduct
Similar repos
| Repo | Stars | Last commit | Licence | Downloads for |
|---|---|---|---|---|
| chadhurley25075-png / pd-bridge | 151 | 32 days ago | Apache-2.0 | No downloads |
| Entrpi / ds4-on-spark | 414 | 2 months ago | MIT | No downloads |
| hasso5703 / alzeo | 252 | yesterday | MIT | No downloads |
More: languages, topics and project record
Languages
- Shell 81%
- Python 19%
Topics
- deepseek
- dgx-spark
- llm-inference
- local-llm
- vllm
Project record
- JobLocal LLMs
- StatusListed
- Commits7
- Forks20
- Watching17
- Stars in 28 days0
- CreatedAug 11, 2026
- First seen by usNov 7, 2026
- CheckedNov 7, 2026