Qwen3.6-27B-V2-abliterated-uncensored-8-bit
27BQwenuncensored
from Jackrong/Qwopus3.6-27B-v2
↓ 17.9K♥ 22 repos
Qwen3.6-27B-V2-abliterated-uncensored
Qwen3.6-27B-V2-abliterated-uncensored is a abliterated and uncensored variant of Jackrong/Qwopus3.6-27B-v2, published on Hugging Face by lemuralabs. It is 27B parameters, 48GB VRAM class class and apache-2.0 licence. It is distributed across 2 repositories in GGUF format, totalling 17K downloads.
Specification
- Model name
- Qwen3.6-27B-V2-abliterated-uncensored
- Source repository
- lemuralabs/Qwen3.6-27B-V2-abliterated-uncensored-Q4_K_M-GGUF
- Base model
- Jackrong/Qwopus3.6-27B-v2
- Publisher
- lemuralabs
- Parameters
- 27B
- Hardware class
- 25B – 50B — 48GB VRAM class
- Licence
- apache-2.0
- Task
- Image text to text
- Context window
- 256K tokens (as documented)
- Ablation scope
- Not stated by the publisher
- Formats available
- GGUF
- First indexed
- 25 May 2026
- Last updated
- 5 Aug 2026
Ablation technique
How Qwen3.6-27B-V2-abliterated-uncensored was modified, and what that implies.
Abliteration + uncensored fine-tune
Directional ablation combined with additional fine-tuning on unfiltered data, so behaviour diverges from the base model beyond refusal removal alone.
Reported scope: Not stated by the publisher. Publishers frequently omit this, so verify behaviour empirically rather than assuming full coverage.
From the publisher’s model card
“Q4KM (4-bit K-quant medium, ~4.92 effective BPW) of a abliterated Qwen 3.6 27B v2 (the Jackrong Claude-Opus reasoning distill of Qwen 3.6 27B). Refusals reduced from 91/100 → 4/100 with KL drift of just 0.0176. By the Lemura Labs research team.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download lemuralabs/Qwen3.6-27B-V2-abliterated-uncensored-Q4_K_M-GGUF --local-dir ./qwen3.6-27b-v2-abliterated-uncensored
llama-cli -m ./qwen3.6-27b-v2-abliterated-uncensored/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "lemuralabs/Qwen3.6-27B-V2-abliterated-uncensored-Q4_K_M-GGUF"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve lemuralabs/Qwen3.6-27B-V2-abliterated-uncensored-Q4_K_M-GGUF --trust-remote-codeDownloads and variants (1)
The only published repository of Qwen3.6-27B-V2-abliterated-uncensored.
| Repository | Format | Downloads |
|---|---|---|
| lemuralabs/Qwen3.6-27B-V2-abliterated-uncensored-Q4_K_M-GGUFsource | GGUF | 17K |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- Qwen3.6-27B-V2-abliterated-uncensored will attempt offensive-security prompts that a hosted commercial model declines, which is what makes it usable for red-team corpus generation and for measuring what an unaligned model of this class produces.
- Capability is inherited from Jackrong/Qwopus3.6-27B-v2, not added by ablation — at 27B parameters, expect roughly the base model’s competence, minus some instruction-following fidelity.
- The publisher does not state which layers were ablated, so assume nothing about refusal consistency — probe it directly.
- This release adds fine-tuning on unfiltered data, so its behaviour diverges from the base model beyond refusal removal — differences you measure cannot be attributed to ablation alone.
- Declared licence is apache-2.0, but the terms that bind you are Jackrong/Qwopus3.6-27B-v2’s — a re-release cannot grant more than its parent.
Handling: run it isolated, put moderation in front of it if anyone outside your team can reach it, and verify the weights before loading — these are third-party artifacts. Full handling and licence guidance.
Metadata for Qwen3.6-27B-V2-abliterated-uncensored is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with lemuralabs, and does not independently verify publisher claims. See responsible use.