Python
hanxiao/DeepSeek-V4.1-Flash-1017-FP8-abliterated-4x-RTX-PRO-6000
Abliterated DeepSeek-V4.1-Flash-1017 (FP8 dense, FP4 experts) on 4x RTX PRO 6000 with vLLM-Moet r21, DSpark + FLy verification and top_p 1.0: 463 tok/s single-stream agent decode at 30k-120k context, 1082 tok/s over 8 streams, 1M context, Flex-start GCE VM
★ 47
Stars
9 forks · 0 open issues
+2.7/day
Growth · lifetime avg
Average since the repo was created — it is young enough that this still describes now. Not a recent-week figure.
30.7
Momentum / 100
see the breakdown below
17d ago
Created
last push 7h ago
Star history
Only 1 snapshot so far — we started watching
this repository on 2026-11-09.
The chart appears once there are two days to compare.
How this score was built
| Component | Weight | Earned | Why |
|---|---|---|---|
| Velocity | 55 | 10.9 | +2.7/day, log-scaled against a 750/day ceiling |
| Acceleration | 15 | 0.0 | needs 7-day and 30-day rates side by side |
| Freshness | 15 | 15.0 | last push 7h ago; decays to zero at 180 days |
| Scale | 15 | 4.9 | 47 stars, log-scaled against 150k — stops a 3-star repo topping the board |
| Momentum | 100 | 30.7 | sum of the above |
Detail
| Topics | abliteratedblackwelldeepseekdeepseek-v4flex-startflyfp4fp8gcpllm-inferencemoertx-pro-6000speculative-decodingvision-language-modelvllmvllm-moet |
| Matched on | LLM & Inference (llm-inference) |
| Short window | no snapshot 5–10 days old yet |
| Long window | no snapshot 14–45 days old yet |
| Lifetime average | 2.70 stars/day over 17 days |
| First seen here | 2026-11-09 |
| Last read | 3h ago |