Shell · MIT
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
★ 176
Stars
20 forks · 0 open issues
flat
Growth · measured
Stars counted between two of our own snapshots roughly a week apart. The repository page gives the exact span.
14.6
Momentum / 100
see the breakdown below
2mo ago
Created
last push 2mo ago
Star history
How this score was built
| Component | Weight | Earned | Why |
|---|---|---|---|
| Velocity | 55 | 0.0 | flat, log-scaled against a 750/day ceiling |
| Acceleration | 15 | 0.0 | needs 7-day and 30-day rates side by side |
| Freshness | 15 | 8.1 | last push 2mo ago; decays to zero at 180 days |
| Scale | 15 | 6.5 | 176 stars, log-scaled against 150k — stops a 3-star repo topping the board |
| Momentum | 100 | 14.6 | sum of the above |
Detail
| Topics | deepseekdgx-sparkllm-inferencelocal-llmvllm |
| Matched on | LLM & Inference (llm-inference) |
| Short window | +0 stars over 7 days = 0.0/day |
| Long window | +0 stars over 26 days = 0.0/day |
| Lifetime average | 2.05 stars/day over 86 days |
| First seen here | 2026-10-11 |
| Last read | 2d ago |