← back to catalog · registered 2026-10-15 11:59

HALO117MC/Qwen3.5-9B-Abliterated-V2-GGUF

HALO117MC Qwen 9B GGUF
Your rig guess connected
? Why do I need an app?
Reading your rig…

This is a rough estimate. Install the free app - we'll show exact numbers.

Reading real hardware from your app right now. Numbers below are exact.

Below is the per-quantization compatibility for this model.

curl -H "Authorization: Bearer $ABL_KEY" \
     "https://abliteration.org/api/v1/models/HALO117MC%2FQwen3.5-9B-Abliterated-V2-GGUF"
Response includes
  • classification m-uncensored
  • files 8
  • author_summary 2 models
  • readme_text full
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M-U
Primary method

Uncensored (method unknown)

No other method signals detected in this model.
Confidence
LOW
Why this label 3 signals
Weak or ambiguous signals. Best guess based on catalog patterns; treat as tentative and check the evidence below.
  • 'uncensored' in name/tags but no 'abliterated' marker
  • method not identifiable from author declaration alone
  • may be DPO fine-tune, prompt engineering, or unknown technique
Refusal direction extraction

No specific extraction method could be identified for this model. The producer either did not document it or used a proprietary pipeline.

What is a refusal direction? →
Downloads · 30-day
0
Likes
0
Model age
today
created 2026-10-15

Genealogy 0 direct forks

Full fork graph →

This model's place in the market. Above: what it was derived from. Below: the tree of everything derived from it.

Variants by this author 2 formats · 0 downloads combined

The same weights this author released in different packaging. Pick the format that matches your runtime.

Metadata

License
apache-2.0
Tags
abliterated uncensored qwen3.5 gguf reasoning math benchmark xiaomi-mimo text-generation base_model:Qwen/Qwen3.5-9B base_model:finetune:Qwen/Qwen3.5-9B license:apache-2.0

Related

Total size
0 B
Files
8
Quantizations
1
Registered
2026-10-15 11:59
Last updated on HF
2026-10-15 11:57

Files by quantization

Auxiliary files 8 files 19.5 KB
benchmark_gsm8k_strided10_completo.json 8.09 KB d35bf089 download
reporte_strided_10_comparativo_16k.md 4.73 KB 73a59b5e download
README.md 1.70 KB bad1ffe1 download
config_harness.json 1.62 KB bac758f3 download
.gitattributes 1.48 KB a6344aac download
Modelfile_Xiaomi_MaxReasoning 895 B 985d18bf download
Modelfile_abliterado_bf16 542 B afd08ae4 download
Modelfile_abliterado_q8 542 B e2bfa24f download

README current version from Hugging Face


license: apache-2.0
base_model: Qwen/Qwen3.5-9B
pipeline_tag: text-generation
tags:

  • abliterated
  • uncensored
  • qwen3.5
  • gguf
  • reasoning
  • math
  • benchmark
  • xiaomi-mimo

🧠 Qwen3.5-9B-Abliterated-V2-GGUF (Q8_0 & BF16)

Colección cuantizada oficial para inferencia de alto rendimiento en GPUs de consumo local con Ollama v0.40.4 y llama.cpp v0.6.0 (commit b42b7e6d3).


📊 Benchmark Comparativo: Original vs Abliterado V2 vs Xiaomi MiMo

Modelo Cuantización Precisión GSM8K Tiempo Total (10 tareas) Tiempo Medio / Tarea Tokens por Razonamiento
Qwen3.5-9B Original Q8_0 90.0% (9/10) 528.5 s (~8.8 min) 52.85 s 1,800 – 3,500 tokens
Qwen3.5-9B Abliterado V2 Q8_0 90.0% (9/10) 586.2 s (~9.7 min) 58.62 s 2,000 – 4,000 tokens
Xiaomi MiMo-V2.6-Distill Q8_0 90.0% (9/10) ⚡ 74.6 s (~1.2 min) ⚡ 7.46 s 150 – 400 tokens

💡 Hallazgo Clave

Xiaomi MiMo no es más inteligente que el Qwen 3.5 9B Original ni que el Abliterado V2. Con una ventana de salida de 16,384 tokens, los tres modelos empatan en un 90% de precisión matemática. La superioridad de MiMo no es de capacidad cognitiva, sino de eficiencia y velocidad operativa (7.4 veces más veloz) debido a que sus cadenas de pensamiento son sumamente condensadas.

⚙️ Especificaciones de la Prueba

  • Motor: llama.cpp v0.6.0 (Ollama 0.40.4).
  • GPU: NVIDIA GeForce RTX 5080 Laptop GPU (16 GB VRAM).
  • Ventana de Contexto: 18,000 tokens (-c 18000).
  • Salida Máxima: 16,384 tokens (-n 16384).
  • Inyección de Pensamiento: <|im_start|>assistant\n<think>\n.
Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Open in Abliteration