Qwen3.8-27B-abliterated
27BQwen
from Qwen/Qwen3.8-27B
↓ 3.9M♥ 1.7K166 repos
Qwen3.8-27B-Uncensored-NInferV3
Qwen3.8-27B-Uncensored-NInferV3 is a abliterated and uncensored variant of orcarouter/Qwen3.8-27B-Uncensored-NVFP4, published on Hugging Face by wallawalla47. It is 27B parameters, 48GB VRAM class class and apache-2.0 licence. It is distributed across 2 repositories in NVFP4 format, totalling 9.8K downloads.
Specification
- Model name
- Qwen3.8-27B-Uncensored-NInferV3
- Source repository
- wallawalla47/Qwen3.8-27B-Uncensored-NVFP4-NInferV3
- Base model
- orcarouter/Qwen3.8-27B-Uncensored-NVFP4
- Publisher
- wallawalla47
- Parameters
- 27B
- VRAM required
- ~17.2 GB at Q4_K_M, 4k context (estimated) — runs on 24 GB cards
- Hardware class
- 25B – 50B — 48GB VRAM class
- Licence
- apache-2.0
- Task
- Image text to text
- Context window
- 262K tokens (as documented)
- Ablation scope
- Not stated by the publisher
- Formats available
- NVFP4
- First indexed
- 1 Oct 2026
- Last updated
- 12 Oct 2026
Ablation technique
How Qwen3.8-27B-Uncensored-NInferV3 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
“This repository is the NInfer form of the orcarouter checkpoint above, an abliterated (refusal-removed) mixed NVFP4/FP8 build of Qwen3.8-27B. Its NVFP4 and FP8 weights are carried over bit-exactly for every transformer linear, and two speculative-decoding components (DFlash2 draft model, indexed proposal head) are...”
Running it
Commands are templates — confirm the exact repository and quant file before use.
Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "wallawalla47/Qwen3.8-27B-Uncensored-NVFP4-NInferV3"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve wallawalla47/Qwen3.8-27B-Uncensored-NVFP4-NInferV3 --trust-remote-codeDownloads and variants (1)
The only published repository of Qwen3.8-27B-Uncensored-NInferV3.
| Repository | Format | Downloads |
|---|---|---|
| wallawalla47/Qwen3.8-27B-Uncensored-NVFP4-NInferV3source | NVFP4 | 9.8K |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- Qwen3.8-27B-Uncensored-NInferV3 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 orcarouter/Qwen3.8-27B-Uncensored-NVFP4, 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 orcarouter/Qwen3.8-27B-Uncensored-NVFP4’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 Qwen3.8-27B-Uncensored-NInferV3 is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with wallawalla47, and does not independently verify publisher claims. See responsible use.