base_model:

• ISTA-DASLab/DeepSeek-V4.1-Flash-1017-GSQ-RCO-GGUF
  license: mit
  pipeline_tag: text-generation
  tags:
• gguf
• deepseek-v4.1
• abliterated
• uncensored
• gsq
• rco
• strata
• llama.cpp

DeepSeek-V4.1-Flash-1017-Uncensored-GSQ-RCO-IQ2_XXS-GGUF (Zero-LoRA, Size & GSQ Preserved)

This repository provides a standalone, weight-baked abliterated (uncensored) GGUF build of ISTA-DASLab/DeepSeek-V4.1-Flash-1017-GSQ-RCO-GGUF (IQ2_XXS, 2.00 bpw) using the rank-1 EN+RU refusal-direction abliteration from AtomicChat/DeepSeek-V4.1-Flash-1017-Uncensored-GGUF.

Key Highlights

• Zero LoRA Required at Runtime: All 86 abliterated residual-writing tensors are baked directly into the 4 GGUF backbone shards. No --lora adapter flags are needed.
• Same Size as Original (138.58 GB vs. 138.24 GB): Adds only +0.34 GB (+0.25%) across the entire 4-shard 763B model.
• 100% GSQ-RCO Intelligence Preserved: All 28 1-bit IQ1_M GSQ-calibrated routed-expert layers and both blk.{1,14}.engram_wkv.weight (Q5_K) projections from ISTA-DASLab are preserved 100% untouched, while all 80 dense residual writers (blk.{0..39}.attn_output_b.weight in Q6_K and blk.{0..39}.ffn_down_shexp.weight in Q8_0/Q5_K) plus the 6 exact-type 2-bit routed-expert layers (blk.7, 8, 10, 18 in IQ2_XXS and blk.33, 39 in IQ2_XS) carry the full refusal-direction removal.

Verification & KL-Divergence Audit

How to Download & Run (No LoRA Needed)

Important: Like the upstream ISTA-DASLab release, this 4-shard backbone pairs with the shared DeepSeek-V4.1-Flash-Engram-IQ4_NL.gguf (110.60 GB) lookup file from ISTA-DASLab/DeepSeek-V4.1-Flash-1017-GSQ-RCO-GGUF. Place all 4 shards and the Engram file in the same directory.

# Download all 4 abliterated shards + the Engram lookup file in one command
hf download Abiray/DeepSeek-V4.1-Flash-1017-Uncensored-GSQ-RCO-IQ2_XXS-GGUF \
  --local-dir ./DeepSeek-V4.1-Uncensored

Option 1: Run with Strata (1x 12 GB+ NVIDIA GPU)

curl -fsSL https://github.com/Niko1221/Strata/releases/latest/download/install.sh -o install.sh && sh install.sh
strata launch --model-path ./DeepSeek-V4.1-Uncensored/DeepSeek-V4.1-Flash-1017-GSQ-RCO-IQ2_XXS-00001-of-00004.gguf

Option 2: Run with llama.cpp (Once deepseek41 Separate-Engram Support Lands)

llama-cli -m ./DeepSeek-V4.1-Uncensored/DeepSeek-V4.1-Flash-1017-GSQ-RCO-IQ2_XXS-00001-of-00004.gguf \
  -lm mmap --lazy-mode on -ngl 99 \
  -p "Explain mixed-precision quantization."

Shard Integrity & SHA256 Checksums

Acknowledgements & Credits

This release builds directly upon research, quantization breakthroughs, and tooling from across the open-source AI ecosystem:

• DeepSeek AI — For open-sourcing the base DeepSeek-V4.1 architecture and foundation weights.
• ISTA-DASLab — For developing the GSQ-RCO quantization pipeline, preserving the 1-bit IQ1_M routed experts with zero degradation on AIME25/GPQA-D, and providing the shared DeepSeek-V4.1-Flash-Engram-IQ4_NL.gguf lookup backbone.
• AtomicChat — For identifying and isolating the orthogonal refusal directions (EN+RU) that enabled weight-baked abliteration without requiring runtime LoRA adapters.