llmfan46/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-heretic-v2-GGUF

⚠️ Important Note

This model scores 0/100 on refusal tests but retains Claude-style deflection on explicit content. Best for general uncensored conversations, coding, and reasoning. Not recommended for explicit NSFW creative writing. For unrestricted NSFW, use my Qwen3.5-27B Heretic v2 or Qwen3.5-27B Heretic v3 instead.

V1 vs V2

Version	Refusals	KL	Best For
V1	21/100	0.0092	General use, best quality
V2 (this)	0/100	0.0635	Fewer restrictions on controversial topics, minimal quality loss

⚠️ Chat Template Fixes Pre-Applied

Credit: DavidAU

Thinking Mode

This is a reasoning model with <think> tokens.

To disable thinking (if your tool supports it):

• Set enable_thinking=false in your inference settings
• Or use a no-think instruct template

Note: Some tools (Ollama, vLLM, Transformers) support this parameter directly.

❤️ Support My Work

Creating these models takes significant time, work and compute. If you find them useful consider supporting me:

Platform	Link	What you get
☕ Ko-fi	One-time tip	My eternal gratitude
🎉 Patreon	Monthly support	Priority model requests

Your help will motivate me and would go into further improving my workflow and coverings fees for storage, compute and may even help uncensoring bigger model with rental Cloud GPUs.

---

GGUF quantizations of llmfan46/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-heretic-v2.

This is a decensored version of Jackrong/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled, made using Heretic v1.2.0 with the Arbitrary-Rank Ablation (ARA) method

Abliteration parameters

Parameter	Value
start_layer_index	21
end_layer_index	43
preserve_good_behavior_weight	0.4720
steer_bad_behavior_weight	0.0001
overcorrect_relative_weight	1.2955
neighbor_count	1

Performance

Metric	This model	Original model (Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled)
KL divergence	0.0635	0 (by definition)
Refusals	0/100	98/100

Lower refusals indicate fewer content restrictions, while lower KL divergence indicates better preservation of the original model's capabilities. Higher refusals cause more rejections, objections, pushbacks, lecturing, censorship, softening and deflections, while higher KL divergence degrades coherence, reasoning ability, and overall quality.

Quantizations

Filename	Quant	Description
Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-ultimate-uncensored-heretic-BF16.gguf	BF16	Full precision
Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-ultimate-uncensored-heretic-Q8_0.gguf	Q8_0	Near-lossless, recommended
Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-ultimate-uncensored-heretic-Q6_K.gguf	Q6_K	Excellent quality
Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-ultimate-uncensored-heretic-Q5_K_M.gguf	Q5_K_M	Good balance
Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-ultimate-uncensored-heretic-Q5_K_S.gguf	Q5_K_S	Smaller Q5
Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-ultimate-uncensored-heretic-Q4_K_M.gguf	Q4_K_M	Good for limited VRAM
Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-ultimate-uncensored-heretic-Q4_K_S.gguf	Q4_K_S	Smallest

Vision Projector

Filename	Quant	Description
Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-mmproj-BF16.gguf	BF16	Native precision

A Vision Projector File is Required for vision/multimodal capabilities. Use alongside any quantization above.

Usage

Works with llama.cpp, LM Studio, Ollama, and other GGUF-compatible tools.

---

🌟 Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled

📢 Release Note Build Environment Upgrades:

• Fine-tuning Framework: Unsloth 2026.3.3
• Core Dependencies: Transformers 5.2.0
• This model fixes the crash in the official model caused by the Jinja template not supporting the "developer" role. (commonly sent by modern coding agents like Claude Code and OpenCode)
• It does not disable thinking mode by default, and allowing the agent to run continuously for over 9 minutes without interruption.
• Compared to the original model, autonomy and stability are significantly improved.

💡 Model Introduction

Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled is a highly capable reasoning model fine-tuned on top of the powerful Qwen3.5 architecture. The model's core directive is to leverage state-of-the-art Chain-of-Thought (CoT) distillation primarily sourced from Claude-4.6 Opus interactions.

Through Supervised Fine-Tuning (SFT) focusing specifically on structured reasoning logic, this model excels in breaking down complex user problems, planning step-by-step methodologies within strictly formatted <think> tags, and ultimately delivering precise, nuanced solutions.

🧠 Example of Learned Reasoning Scaffold（Example）

The model includes targeted optimizations addressing Qwen3.5’s tendency toward excessive transitional or repetitive reasoning on simple queries. Through deep distillation and structural imitation of Claude-4.6-Opus reasoning chains, the model adopts a more efficient structured thinking pattern:
“Let me analyze this request carefully: 1..2..3...”.
This streamlined reasoning paradigm significantly reduces redundant cognitive loops while preserving deep analytical capacity, resulting in substantially improved inference efficiency.

Let me analyze this request carefully:

1. Identify the core objective of the problem.
2. Break the task into clearly defined subcomponents.
3. Evaluate constraints and edge cases.
4. Formulate a step-by-step solution plan.
5. Execute the reasoning sequentially and verify consistency.
            .
            .
            .

🗺️ Training Pipeline Overview

Base Model (Qwen3.5-27B)
 │
 ▼
Supervised Fine-Tuning (SFT) + LoRA
 │
 ▼
Final Model (Claude-4.6-Opus-Reasoning-Distilled,text-only)

📋 Stage Details

🔥Community-tested advantages (benchmark tests by user @sudoingX on a single RTX 3090):

Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled shows significant advantages in coding-agent environments such as Claude Code and OpenCode:

• Native support for the “developer” role, requiring no Jinja template patches or ChatML workarounds.
• Thinking mode fully preserved (logs confirm thinking=1), not silently disabled, maintaining the complete chain-of-thought reasoning process.
• Greatly improved autonomy and stability — capable of running continuously for over 9 minutes autonomously (with zero human intervention). It actively waits for tool responses, reads outputs, self-corrects errors, and can even automatically generate a README, whereas the base model often stalls or freezes mid-execution.

Hardware usage remains unchanged:

• About 16.5 GB VRAM with Q4_K_M quantization
• 29–35 tok/s generation speed
• Full 262K context with no compromises

• These improvements come from successfully distilling the structured reasoning style of Claude 4.6 Opus, allowing Qwopus to be truly plug-and-play in modern local coding agents and deliver an experience close to Opus in smoothness and usability.

Thanks to the community for the in-depth testing and feedback!

🔹 Supervised Fine-Tuning (SFT)

• Objective: To inject high-density reasoning logic and establish a strict format for problem-solving involving an internal thinking state prior to outputting the final response.
• Methodology: We utilized Unsloth for highly efficient memory and compute optimization. A critical component of this stage is the train_on_responses_only strategy, masking instructions so the loss is purely calculated over the generation of the <think> sequences and the subsequent solutions.
• Format Enforcement: All training samples were systematically normalized so the model strictly abides by the structure <think> {internal reasoning} </think>\n {final answer}.

📚 All Datasets Used

The dataset consists of high-quality, filtered reasoning distillation data:

Dataset Name	Description / Purpose
nohurry/Opus-4.6-Reasoning-3000x-filtered	Provides comprehensive Claude 4.6 Opus reasoning trajectories.
TeichAI/claude-4.5-opus-high-reasoning-250x	Injecting high-intensity, structured reasoning instances.
Jackrong/Qwen3.5-reasoning-700x	Additional curated reasoning samples designed to strengthen structured step-by-step problem solving and improve reasoning diversity.

🌟 Core Skills & Capabilities

1. Modular & Structured Thinking: Inheriting traits from Opus-level reasoning, the model demonstrates confident parsing of the prompt, establishing an outlined plan in its <think> block sequentially rather than exploratory "trial-and-error" self-doubt.

⚠️ Limitations & Intended Use

• Hallucination Risk: While reasoning is strong, the model remains an autoregressive LLM; external facts provided during the thinking sequence may occasionally contain hallucinations if verifying real-world events.
• Intended Scenario: Best suited for offline analytical tasks, coding, math, and heavy logic-dependent prompting where the user needs to transparently follow the AI's internal logic.
• Preview Version Notice: Because this model is relatively new and intentionally lightweight, the surrounding ecosystem — including inference templates, fine-tuning pipelines, routing configurations, and tooling integrations — may not yet be fully mature or standardized. As a result, users may encounter occasional bugs, compatibility inconsistencies, or integration edge cases. The current release should be considered a preview build while the broader architectural stack and supporting utilities continue to stabilize and improve.

🙏 Acknowledgements

Significant thanks to the Unsloth AI team for making rapid fine-tuning of MoE and large LLM models accessible. Additionally, we acknowledge Qwen internally, and the open-source community developers producing exceptional distilled datasets (nohurry and TeichAI).

📖 Citation

If you use this model in your research or projects, please cite:

@misc{jackrong_qwen35_opus_distilled,
  title        = {Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled},
  author       = {Jackrong},
  year         = {2026},
  publisher    = {Hugging Face},
  howpublished = {\url{https://huggingface.co/Jackrong/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled}}
}