Ternary-Bonsai-2-27B-Uncensored-Heretic-NInfer
27BOtherheretic
from OS-Software/Ternary-Bonsai-2-27B-Uncensored-Heretic-GGUF
↓ 38.6K♥ 13
Offrail-27B-Uncensored
Offrail-27B-Uncensored is a abliterated and uncensored variant of OS-Software/Ternary-Bonsai-2-27B-Uncensored-Heretic-GGUF, published on Hugging Face by sudoingx. It is 27B parameters, 48GB VRAM class class and apache-2.0 licence. It is distributed across 2 repositories in GGUF format, totalling 20.2K downloads.
Specification
- Model name
- Offrail-27B-Uncensored
- Source repository
- sudoingx/Offrail-27B-Uncensored
- Publisher
- sudoingx
- 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
- Text generation
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- GGUF
- First indexed
- 15 Oct 2026
- Last updated
- 16 Oct 2026
Ablation technique
How Offrail-27B-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
“27B for 8GB and 12GB cards. OS-Software's uncensored Heretic build of Ternary Bonsai 2 27B (the 1.75 bpw PTQ10 file, 5.95 GB) with the Qwen 3.8 27B multi-token-prediction head grafted back on as block 64, so the PrismML llama.cpp fork runs --spec-type draft-mtp on it.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download sudoingx/Offrail-27B-Uncensored --local-dir ./offrail-27b-uncensored
llama-cli -m ./offrail-27b-uncensored/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "sudoingx/Offrail-27B-Uncensored"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve sudoingx/Offrail-27B-Uncensored --trust-remote-codeDownloads and variants (1)
The only published repository of Offrail-27B-Uncensored.
| Repository | Format | Downloads |
|---|---|---|
| sudoingx/Offrail-27B-Uncensoredsource | GGUF | 20.2K |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- Offrail-27B-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 OS-Software/Ternary-Bonsai-2-27B-Uncensored-Heretic-GGUF, 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 OS-Software/Ternary-Bonsai-2-27B-Uncensored-Heretic-GGUF’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 Offrail-27B-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 sudoingx, and does not independently verify publisher claims. See responsible use.