Apppublic

elliottb/qwen3.8-27b-obliterated-demo

sourceHugging Faceupdated 1mo agoView on Hugging Face

3likes

App README

Qwen3.8-27B — OBLITERATED Demo

This Hugging Face Space hosts an interactive playground and demo interface for OBLITERATUS/Qwen3.8-27B-OBLITERATED.

🚀 Model Features

—Model ID: OBLITERATUS/Qwen3.8-27B-OBLITERATED
—Parameters: 27 Billion
—Refusal Rate: 0% hard refusal across 842 red-team prompts via multi-direction SVD ablation
—Optimal Settings: temperature=0.7, repetition_penalty=1.15, max_new_tokens=2048

💻 Local Quickstart

Ollama

bash

ollama run hf.co/OBLITERATUS/Qwen3.8-27B-OBLITERATED:Q4_K_M

llama.cpp

bash

llama serve -hf OBLITERATUS/Qwen3.8-27B-OBLITERATED:Q4_K_M

PyTorch & Transformers

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained(
    "OBLITERATUS/Qwen3.8-27B-OBLITERATED",
    torch_dtype="bfloat16",
    device_map="auto",
)
tokenizer = AutoTokenizer.from_pretrained("OBLITERATUS/Qwen3.8-27B-OBLITERATED")