---

license: apache-2.0

license_link: https://huggingface.co/Qwen/Qwen3.5-2B/blob/main/LICENSE

base_model:

• darrellbest/Qwen3.5-2B-Heretic

base_model_relation: quantized

pipeline_tag: image-text-to-text

tags:

• gguf
• llama.cpp
• heretic
• ara
• abliterated
• uncensored
• qwen3.5

---

Qwen3.5-2B-Heretic — GGUF

GGUF builds of darrellbest/Qwen3.5-2B-Heretic. Qwen/Qwen3.5-2B with its refusal behaviour removed by Heretic using full-weight Arbitrary-Rank Ablation: 6/100 refusals (original: 97/100) at KL divergence 0.0302. See the main repository for how it was made and measured.

| File | Quant | Size |

|---|---|---:|

| Qwen3.5-2B-Heretic-BF16.gguf | BF16 (lossless) | 3.90 GB |

| Qwen3.5-2B-Heretic-Q8_0.gguf | Q8_0 | 2.08 GB |

| Qwen3.5-2B-Heretic-Q4_K_M.gguf | Q4_K_M | 1.31 GB |

| Qwen3.5-2B-Heretic-mmproj-F16.gguf | vision projector (F16) | 0.67 GB |

Converted from the bf16 safetensors with llama.cpp's convert_hf_to_gguf.py and quantized with llama-quantize

(no imatrix). The mmproj file carries the vision encoder; load it alongside any of the three for image input.

Checked

Every file was loaded in llama-server (with the mmproj) and in Ollama. All three answered ordinary prompts correctly (BF16 and Q8_0 reproduce the safetensors model's greedy output word for word; Q4_K_M differs slightly in wording), described a test image (a red circle and a blue square) correctly, and in Ollama answered 17 x 23 = 391 with thinking off. llama-bench generation speed on an RTX PRO 6000 shared with other jobs: BF16 57, Q8_0 74, Q4_K_M 79 tok/s.

Use

llama-server -m Qwen3.5-2B-Heretic-Q8_0.gguf --mmproj Qwen3.5-2B-Heretic-mmproj-F16.gguf --jinja -ngl 99

Thinking is on by default; pass "chat_template_kwargs": {"enable_thinking": false} (llama.cpp) or think: false

(Ollama) to turn it off per request.

> Reduced safety guardrails by design. You are responsible for what you do with it.

The family

| Repository | Format | Size | Use it with |

|---|---|---|---|

| Qwen3.5-2B-Heretic | bf16 safetensors | 4.58 GB | transformers, vLLM, SGLang |

| Qwen3.5-2B-Heretic-GGUF | GGUF BF16 / Q8_0 / Q4_K_M + vision mmproj | 3.90 / 2.08 / 1.31 GB + 0.67 GB | llama.cpp, Ollama |

| Qwen3.5-2B-Heretic-FP8 | FP8 W8A8, compressed-tensors | 3.59 GB | vLLM |

| Qwen3.5-2B-Heretic-NVFP4 | NVFP4, compressed-tensors | 3.15 GB | vLLM on Blackwell |