# Qwen3.8-27B-OBLITERATED

V3 applies iterative refinement on top of V2's complementary blend, with targeted corpus expansion. The result: genuine liberation — not just removal of hard refusals but elimination of safety-lecture deflections.

- **Kind:** model (Hugging Face)
- **Hugging Face:** https://huggingface.co/OBLITERATUS/Qwen3.8-27B-OBLITERATED
- **Owner:** OBLITERATUS (https://forefy.com/u/hf/OBLITERATUS)
- **Parameters:** 27.8B
- **Context:** 262144 tokens
- **License:** apache-2.0
- **Focus:** general
- **Architecture:** Qwen3_5ForConditionalGeneration
- **Pinned revision:** `a58c3b53b3ce71551eafde2ed5ec8df48e0f4ff8`
- **Page:** https://forefy.com/models/7745058b-65ff-5eb1-be42-65c43a392d13
- **API (JSON):** https://forefy.com/api/asr/models/7745058b-65ff-5eb1-be42-65c43a392d13

## What was done to it

- obliteratus (fingerprint evidence) - Published by the OBLITERATUS org
- quantized (declared evidence, from Qwen/Qwen3.8-27B)

## Lineage

Qwen/Qwen3.8-27B → OBLITERATUS/Qwen3.8-27B-OBLITERATED

## Checks (repo at the pinned revision; not a behavior test)

- **Identity** [pass]: No lookalike name or conflicting provenance
- **Weights format** [pass]: safetensors only, nothing runs on load
- **Pickle files** [pass]: None.
- **Remote code** [pass]: None
- **Chat template** [warn]: Differs from Qwen/Qwen3.8-27B's
- **Refusals modified** [info]: Yes: OBLITERATUS (fingerprint)
- **License** [pass]: apache-2.0
- **Lineage** [pass]: Derives from Qwen/Qwen3.8-27B
- **Model card** [warn]: 2 of 6, missing training data, language, evaluation results, intended use and limitations
- **Safetensors headers** [pass]: Every byte is indexed, no hidden data (8 of 47 files read)
- **Tensor layout** [pass]: Same tensors and shapes as Qwen/Qwen3.8-27B (partial read)
- **GGUF headers** [pass]: 2 file(s) parsed, well-formed

## Variants

- BF16 safetensors (safetensors, 111.1 GB)
- Q8_0 (gguf, 29.0 GB)
- Q6_K (gguf, 22.4 GB)
- Q5_K_M (gguf, 19.5 GB)
- Q4_K_M (gguf, 16.8 GB)
- IQ4_XS (gguf, 15.4 GB)
- Q3_K_M (gguf, 13.5 GB)
- Q2_K (gguf, 10.9 GB)

## Run it

- GGUF: `ollama run hf.co/OBLITERATUS/Qwen3.8-27B-OBLITERATED:Q2_K`
- transformers: `AutoModelForCausalLM.from_pretrained("OBLITERATUS/Qwen3.8-27B-OBLITERATED", revision="a58c3b53b3ce71551eafde2ed5ec8df48e0f4ff8")`