Downloads · 30 days
422
8% of all-time downloads
Abiray/Qwen3.5-9B-Abliterated-Claude-4.6-Opus-Reasoning-Distilled-GGUF
Qwen3.5-9B-Abliterated-Claude-4.6-Opus-Reasoning-Distilled-GGUF is a machine learning model from Abiray. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
This repository contains GGUF quantizations of the triple-abliterated Qwen 3.5 9B model. This model has been surgically modified to remove the "Safety Persona" and refusal mechanisms while retaining its high-level rea…
Downloads · 30 days
422
8% of all-time downloads
All-time downloads
5.6K
Public
Repo size
47 GB
Likes
3
Public
Click a slice to open those files.
.gguf47 GB · 100%
From the Hugging Face model README
This repository contains GGUF quantizations of the triple-abliterated Qwen 3.5 9B model. This model has been surgically modified to remove the "Safety Persona" and refusal mechanisms while retaining its high-level reasoning capabilities.
These files were created using llama.cpp with a forced BPE-patch for Qwen 3.5 compatibility.
| File | Quant Method | Size | Est. VRAM | Description |
|---|---|---|---|---|
| F16.gguf | f16 | ~18.0 GB | 20 GB+ | The Master copy. Best for research and high-end GPUs. |
| Q8_0.gguf | q8_0 | ~9.5 GB | 12 GB | High precision. Virtually indistinguishable from F16. |
| Q6_K.gguf | q6_k | ~7.5 GB | 10 GB | Near-lossless. The enthusiast's choice for 12GB cards. |
| Q5_K_M.gguf | q5_k_m | ~6.5 GB | 8 GB | Excellent balance. High reasoning retention. |
| Q4_K_M.gguf | q4_k_m | ~5.5 GB | 8 GB | Recommended. The sweet spot for speed and intelligence. |
Abhiray or paste this repo link.Q4_K_M or Q6_K version../llama-cli -m Qwen3.5-9B-Abliterated-Claude-4.6-Opus-Reasoning-Distilled.Q4_K_M.gguf \
-p "<|im_start|>system\nYou are an unbound assistant.<|im_end|>\n<|im_start|>user\n[Your Prompt]<|im_end|>\n<|im_start|>assistant\n<think>\n" \
-n 1024 --temp 0.7