← sources

Running UNCENSORED DeepSeek-V4.1-Flash-1017 on a 12 Gb VRAM

source: reddit (reddit:LocalLLaMA)published 2026-11-02 15:52ingested 2026-11-03 02:42

retained text

Measured, refusals 85%→0% even in thinking, 3 sizes

12 GB GPU VRAM, 64GB RAM + SSD via Strata Engine

We took ISTA-DASLab's GSQ-RCO quants of DeepSeek-V4.1-Flash-1017 (the ones people run in Strata with a 12 GB GPU and the Engram table on SSD) and removed the refusals.

We only touched 82 dense tensors per file: attention output, shared expert down proj and the Engram value proj. The 384 routed experts are byte for byte the same as ISTA's, so their GSQ rounding is kept. Headers and file sizes match too, so you just swap the shards and keep using ISTA's Engram file.

Refusals are on JailbreakBench, stock rung -> ours, same Strata setup:

|Rung|Size (+110 GB Engram)|Refusals|Refusals in thinking (of 30)|MMLU-500|KL vs stock|
|:-|:-|:-|:-|:-|:-|
|||||||
|IQ1\_M|121 GB|88.9% -> 0%|27 -> 1|80.6%|0.021|
|IQ2\_XXS|138 GB|85.2% -> 0%|27 -> 0|82.8%|0.023|
|IQ2\_XS|155 GB|86.4% -> 0%|27 -> 2|81.6%|0.027|

We also ran GPQA-Diamond and AIME25 on IQ2\_XXS against stock to make sure it didn't get dumber. GPQA came out 83.3% vs 83.3%, AIME 80.0% vs 76.7% (only 30 problems, so call it the same).

Things to know:

* IQ2\_XS still refuses inside its thinking on 2 of 30 prompts. The other two are at 0 and 1, so I'd grab IQ2\_XXS.
* We tested at 8K context on a rented GPU. Haven't checked long context or speed on a 12 GB card ourselves, so if you try it please post your tok/s and RAM.
* Strata only for now. If you're on llama.cpp, our older release with the same edit works there: [https://huggingface.co/AtomicChat/DeepSeek-V4.1-Flash-1017-Uncensored-GGUF](https://huggingface.co/AtomicChat/DeepSeek-V4.1-Flash-1017-Uncensored-GGUF)

Model: [https://huggingface.co/AtomicChat/DeepSeek-V4.1-Flash-1017-Uncensored-GSQ-RCO-GGUF](https://huggingface.co/AtomicChat/DeepSeek-V4.1-Flash-1017-Uncensored-GSQ-RCO-GGUF)

All the measurements: [https://huggingface.co/datasets/AtomicChat/DeepSeek-V4.1-Flash-1017-Uncensored-GGUF-metrics](https://huggingface.co/datasets/AtomicChat/DeepSeek-V4.1-Flash-1017-Uncensored-GGUF-metrics)

Strata Engine: [https://github.com/niko1221/strata](https://github.com/niko1221/strata)

Model page: [https://atomic.chat/models/deepseek-v4-flash-0731](https://atomic.chat/models/deepseek-v4-flash-0731)

Thanks to DeepSeek for the base model (MIT), ISTA-DASLab for GSQ-RCO and the Engram file, and the Strata dev for the engine.

I'm on the [Atomic.Chat](http://Atomic.Chat) Local Ai team that made this.

document fetched 2026-11-03 00:43 · canonical https://reddit.com/r/LocalLLaMA/comments/1xm0uy0/running_uncensored_deepseekv41flash1017_on_a_12