# MiaAI-Lab/DeepSeek-v4.1-Flash-EXL3-2x-DGX-Sparks

> A local EXL3 2.9 bpw checkpoint of DeepSeek-V4.1-Flash (196 GiB, 39 shards) served via an OpenAI-compatible vLLM API on a 2x NVIDIA GB10 (DGX Spark) kit with tensor-parallel 2. Includes DSpark speculative decoding and memory tuning for the node's tight RAM budget.

- Magnitude: 0.8 out of 10 — Steady
- Stars: 248 total · +2 stars on 2026-09-29 (UTC)
- Star trust: star growth looks organic
- Category: Language models · Language: Python · License: AGPL-3.0 · Created: 2026-09-13 · Last push: 2026-09-19
- GitHub: https://github.com/MiaAI-Lab/DeepSeek-v4.1-Flash-EXL3-2x-DGX-Sparks · Homepage: https://x.com/MiaAI_lab · Page: https://gitnova.dev/en/r/MiaAI-Lab/DeepSeek-v4.1-Flash-EXL3-2x-DGX-Sparks

## Useful for

- Deploy a local OpenAI-compatible DeepSeek-V4.1-Flash server on two DGX Sparks
- Run inference with up to 600k token context via vLLM
- Enable DSpark speculative decoding to speed up generation

## Why it’s here

- 2 stars today.
- The repository is 16 days old and already has 248 stars.
- Recent forks include notable developers: @0xSojalSec (1,003 followers).

## Star trust

Star growth looks organic.

Star-trust labels are heuristics based on the repository’s behavior, not a check of every stargazer.

## Numbers

- Forks: 34
- Issues and pull requests: 28
- Watchers: 4
- Average over the last week: 3 per day
- Usual pace: 12 per day

## Stars per day, last 30 days (oldest → newest, today is partial)

2026-09-30 … 2026-10-29: None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None, None

## Magnitude by day, last 14 days

09-16 3.6, 09-17 3.9, 09-18 3.1, 09-19 1.7, 09-20 1.7, 09-21 1.3, 09-22 1.4, 09-23 1.1, 09-24 1.1, 09-25 1.3, 09-26 0.8, 09-27 0.9, 09-28 0.6, 09-29 0.9

---

Magnitude (0–10) measures how fast and how unusually interest in a repository is growing right now. It is not a quality score.
Days are UTC. “So far today” is a fact; “expected by the end of the day” is a forecast.
Summaries and use cases are written by an LLM (DeepSeek V4.1 Flash) from the README and may be inaccurate: verify specific claims (benchmarks, speed, hardware) in the repository itself.
Data as of 2026-09-29 23:45 UTC, updated every 30 minutes.