Instructions to use MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2")
model = AutoModelForCausalLM.from_pretrained("MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2

SGLang

How to use MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2 with Docker Model Runner:
```
docker model run hf.co/MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Qwen3.5-9B-Abliterated-HSAQ-v2

HSAQR v2 — Solo abliteración (identificación automática → rewrite selectivo → PWS global). Este repo NO incluye el fine-tune final.

Autor: Jesús Antonio Zárate Hernández (MethodWhite), M.A.T.E.R.I.A. Research, 2026.

Diferencia con Qwen3.5-9B-HSAQR-v2

	Este repo (Abliterated)	Qwen3.5-9B-HSAQR-v2
Identificación automática (8/32 capas)	✅	✅
Rewrite selectivo (ortogonalización)	✅	✅
PWS global (MLP 12288→7456)	✅	✅
Fine-tune HSAQ-optimizer	❌ No	✅ Sí
Dataset unificado 2026	—	✅ (60+ ejemplos)
Uso	Abliteración pura, para comparar	Modelo final rewrite + abliteración

Pipeline v2 aplicado

Identificación auto-calibrada (IQR + magnitud relativa): capas L17-L21, L23, L24, L29 con señal real de refusal.
Rewrite jerárquico de canales (hilos/cuerdas/hebras) + ortogonalización selectiva de o_proj y down_proj.
PWS global: dimensión intermedia del MLP 12288 → 7456 (config actualizado).

El modelo NO pasó por el fine-tune final con HSAQ-optimizer. Para el modelo completo (abliteración + rewrite de comportamiento), usar MethodWhite/Qwen3.5-9B-HSAQR-v2.

Nota sobre el tamaño (7B vs 9B)

El PWS recortó la dimensión intermedia del MLP de 12288 → 7456 en todas las capas (−39.3% del MLP). Los embeddings, atención y normas quedaron intactos. Por eso el modelo tiene 7.05B parámetros (de los 8.95B originales, −21.2%) sin perder el conocimiento léxico (embeddings intactos) — es compresión física, no destilación.

Resultados medidos

PWS: −39% pesos MLP (ahorro ~7 GB en disco)
Identificación automática: 8/32 capas tocadas, sintaxis intacta
Abliteración sin fine-tune: refusal eliminado, coherencia base preservada

Uso

from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained("MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2")
tokenizer = AutoTokenizer.from_pretrained("MethodWhite/Qwen3.5-9B-Abliterated-HSAQ-v2")